AOSpine Masters Series Volume 2 Primary Spinal Tumors

SAVE OFFLINE

AOSpine Masters Series Primary Spinal Tumors

AOSpine Masters Series Primary Spinal Tumors

Series Editor:

Luiz Roberto Vialle, MD, PhD Professor of Orthopedics, School of Medicine Catholic Universit y of Parana State Spine Unit Curitiba, Brazil

Guest Editors: Ziya L. Gokaslan, MD, FACS Donlin M. Long Professor Professor of Neurosurgery, Oncology, and Orthopaedic Surgery Director, Neurosurgical Spine Program Vice Chair, Departm ent of Neurosurgery Johns Hopkins Universit y School of Medicine Baltimore, Maryland

Charles G. Fisher, MD, MHSc, FRCSC Professor and Head Division of Spine Surgery Department of Orthopaedic Surgery Universit y of British Colum bia Vancouver, British Colum bia

Stefano Boriani, MD Unit of Oncologic and Degenerative Spine Surgery Rizzoli Orthopedic Institute Bologna, Italy

With 86 f gures

Th iem e New York • St u t tgart • Delh i • Rio

Th iem e Medical Pu blish ers, In c. 333 Seven th Ave. New York, NY 10001 Execut ive Editor: Kay Con erly Man aging Editor: Ju d ith Tom at Editorial Assist an t: Haley Paskalides Sen ior Vice Presiden t , Editorial an d Elect ron ic Produ ct Developm en t: Corn elia Sch u lze Product ion Editor: Barbara A. Ch ern ow In tern at ion al Product ion Director: An dreas Sch aber t In tern at ion al Market ing Director: Fion a Hen derson Director of Sales, North Am erica: Mike Rosem an In tern at ion al Sales Director: Louisa Turrell Vice Presiden t , Fin an ce an d Accoun t s: Sarah Van derbilt Presiden t: Brian D. Scan lan Com p ositor: Carol Pierson , Ch ern ow Editorial Ser vices, In c.

Librar y of Congress dat a is available from th e p ublish er.

Copyrigh t ©2015 by Th iem e Medical Publish ers, In c. Im po rtan t n ote: Medicin e is an ever-ch anging scien ce u n dergoing con t in u al develop m en t . Research an d clin ical experien ce are con t in ually expan ding ou r kn ow ledge, in par t icu lar ou r kn ow ledge of prop er t reat m en t an d drug th erapy. In sofar as th is book m en t ion s any dosage or ap plicat ion , readers m ay rest assu red th at th e au th ors, editors, an d publish ers h ave m ade ever y e ort to en sure th at su ch referen ces are in accordan ce w ith the state o f know ledge at th e tim e o f pro ductio n o f the bo o k. Never th eless, th is does n ot involve, im p ly, or express any gu aran tee or respon sibilit y on th e par t of th e pu blish ers in respect to any dosage in st ruct ion s an d form s of applicat ion s st ated in th e book. Every user is requested to exam ine carefully th e m an u fact u rers’ lea ets accom p anying each drug an d to ch eck, if necessar y in con su ltat ion w ith a physician or sp ecialist , w h eth er th e dosage sch edu les m en t ion ed th erein or th e con t rain dicat ion s st ated by th e m an ufact u rers d i er from th e st atem en ts m ade in th e presen t book. Such exam in at ion is part icularly im por t an t w ith drugs th at are eith errarely u sed or h ave been n ew ly released on th e m arket . Ever y dosage sch edu le or ever y form of ap plicat ion u sed is en t irely at th e u ser’s ow n risk an d respon sibilit y. Th e au th ors an d p ublish ers requ est ever y u ser to rep or t to th e p u blish ers any discrepan cies or in accuracies n ot iced. If errors in th isw ork are fou n d after pu blicat ion , errat a w ill be posted at w w w.th iem e.com on th e p rodu ct descript ion page. Som e of th e p rodu ct n am es, paten t s, an d registered d esign s referred to in th is book are in fact registered t radem arks or p ropriet ar y n am es even th ough speci c referen ce to th is fact is n ot alw ays m ade in th e text . Th erefore, th e appearan ce of a n am e w ith ou t design at ion as prop rietar y is n ot to be con st ru ed as a represen t at ion by th e p u blish er th at it is in th e pu blic dom ain . Prin ted in Ch in a by Everbest Prin t ing Ltd. 5 4 3 2 1 ISBN 978-1-62623-047-7 Also available as an e-book: eISBN 978-1-62623-049-1

AOSpine Masters Series Luiz Roberto Vialle, MD, PhD Series Editor

Volum e 1

Met astat ic Spin al Tu m ors

Volum e 2

Prim ar y Spin al Tum ors

Volum e 3

Cer vical Degen erat ive Con dit ion s

Volum e 4

Adu lt Sp in al Deform it ies

Volum e 5

Sp in al Trau m a 1, Cer vical

Volum e 6

Sp in al Trau m a 2, Pitfalls on Th oracolum bar

Volum e 7

SCI an d Regen erat ion

Volum e 8

Back Pain

Vo lum e 9

Pediat ric Sp in al Deform it ies

Volum e 10

Sp in al In fect ion

Contents

Series Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Luiz Roberto Vialle Guest Edito rs’ Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Charles G. Fisher, Stefano Boriani, and Ziya L. Gok aslan 1 Evaluat ion an d Decision Making . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 Stefano Boriani and Charles G. Fisher 2 Safet y an d E cacy of Su rger y for Prim ar y Tum ors of th e Spin e . . . . . . . . . . . . . . . . . . . . . . . . . 15 Daryl R. Fourney, Charles G. Fisher, and Stefano Boriani 3 In ter ven t ion al Opt ion s for Prim ar y Tu m ors of th e Spin e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Sudhir Kathuria 4 Radiat ion Th erapy for Prim ar y Bon e Tu m ors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Brent Y. Kim ball and Yoshiya (Josh) Yam ada 5 Medical On cology Prin ciples for th e Sp in e On cology Su rgeon . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Scott H. Okuno, Steven I. Robinson, and Shreyaskum ar Patel 6 Sp in al Osteoid Osteom a an d Osteoblastom a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Mélissa Nadeau and Christ ian P. DiPaola 7 An eu r ysm al Bon e Cyst an d Gian t Cell Tu m or . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Stefano Boriani, Stefano Bandiera, Riccardo Gherm andi, Sim one Colangeli, Luca Am endola, and Alessandro Gasbarrini 8 Ch ordom a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Jack son Sui, Pat ricia L. Zadnik , Mari L. Groves, and Ziya L. Gokaslan 9 Ch on drosarcom a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Stefano Boriani, Stefano Bandiera, Riccardo Gherm andi, Sim one Colangeli, Luca Am endola, and Alessandro Gasbarrini 10 Osteogen ic Sarcom a an d Ew ing’s Sarcom a of th e Spin e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 Derek G. Ju, Pat ricia L. Zadnik , and Daniel Michael Sciubba

viii

Contents 11 Margin s in Sp in e Tu m or Resect ion : How Much Is En ough ? Is Plan n ed Tran sgression Okay? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124 Ilya Laufer, Mari L. Groves, and Jean-Paul W olinsk y 12 Prin cip les Beh in d Determ in ing th e Righ t Ap proach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 Nicolas Dea and Row an Schouten 13 Sp in op elvic Recon st r uct ion /Fixat ion an d Fusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 Peter Paul Varga 14 St r u ct u ral Graft Select ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Y. Raja Ram persaud 15 Wou n d Closu re Tech n iqu es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Just in M. Sack s and Just in M. Broyles 16 Com plicat ion s an d Th eir Avoidan ce: How to Plan Prim ar y Tum or Resect ion to Min im ize Com p licat ion s an d Maxim ize Outcom e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Michael C. Oh, Vedat Deviren, and Christopher P. Am es Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191

Series Preface

Sp in e care is advan cing at a rapid pace. Th e ch allen ge for today’s sp in e care p rofession al is to qu ickly syn t h esize t h e best available eviden ce an d exp er t op in ion in th e m an agem en t of sp in e p at h ologies. Th e AOSp in e Masters Series p rovides ju st th at—each volu m e in th e series delivers p ath ology-focu sed exper t opin ion on procedu res, diagn osis, clin ical w isdom , an d pitfalls, an d h igh ligh ts today’s top research pap ers. To bring th e value of its m asters level educat ion al cou rses an d academ ic congresses to a w ider audien ce, AOSpin e h as assem bled in tern at ion ally recogn ized spin e path ology leaders to develop volum es in th is Masters Series as a

vehicle for sharing their experiences and expert ise an d p roviding lin ks to th e literat u re. Each volu m e focuses on a curren t com pelling an d som et im es con t roversial topic in spin e care. Th e u n iqu e an d e cien t for m at of t h e Masters Ser ies volu m es qu ickly focu ses t h e at ten t ion of t h e read er on t h e core in for m at ion crit ical to un derst an ding th e topic, w h ile en cou raging th e read er to look fu r th er in to th e recom m en d ed literat u re. Th rough th is approach , AOSpin e is advan cing spin e care w orldw ide. Luiz Roberto Vialle, MD, PhD

Guest Editors’ Preface

To pract ice eviden ce-based spin e surger y a spin e surgeon m ust com bin e a rigorous an d crit ical approach to th e evaluat ion of scien t i c literat u re w ith clin ical exp er t ise, an d a st rong com m it m en t to p at ien t cen teredn ess an d h u m an ist ic valu es. In n o oth er dom ain of spin e su rger y are th ese p rin ciples m ore crit ical an d relied upon then in the m anagem ent of prim ary t u m ors of th e spin e, w h ere a poor m an agem en t decision can ren der a p oten t ially cu rable pat ien t , in cu rable. Th e goal of th is gu ide is to balan ce crit ical appraisal of curren t eviden ce in spin e on cology w ith op in ion from exp erien ced experts, to create a dist inctive and m ean ingful clin ical referen ce for th e pract icing spin e on cology su rgeon . The chapters have been researched and w ritten by key opin ion leaders in spin e on cology an d range from gen eral evalu at ion , st aging an d decision m aking prin ciples to h istology speci c on cologic pat ien t m an agem en t . Th e 3 sen ior editors h ave review ed each ch apter to t r y an d ensure consistency and the necessary synthesis of best available literat u re an d expert opin ion . Furth erm ore a con scious an d ver y n ecessar y e ort to ensure m ult idisciplin ar y appraisal an d inp ut h as been t aken . In fact th e inp ut of th e m edical an d radiat ion on cologist , radiology in ter ven t ion alist , an d path ologist , along w ith th e spin e surgeon is em ph asized th rough out th is book, just as it sh ould be in th e day-to-day care of spin e t um or pat ien t s. Mult iple auth ors w ere

en couraged for each chapter to en sure the m ost balanced, transparent, and com prehensive rep resen t at ion p ossible. Th e on cologically ap p rop riate m an agem en t of prim ary tum ors of the spine has been plagued by a lack of stan dardizat ion ; th is begin s w ith th e basic term in ology, an d en com passes th e staging and com plex m anagem ent of these rare, often let h al t u m ors. Th e basis of t h e p roblem is th e fact th at th ese elem en t s used in m an aging spin e t u m ors h ave all h istorically relied on the use of accepted gen eral orth opedic and neurosurgical prin ciples in th e spin e. Th e spin al com m un it y, h ow ever, h as been gen erally relu ct an t to fu lly adopt th e p rin ciples of app en dicular m usculoskeletal oncology developed by En n eking. Th is is becau se en bloc t um or resection in the spin e entails com plexities not alw ays en cou n tered in th e ext rem it ies, su ch as restoring stabilit y, and avoiding severe neurologic injur y. Consequently, th e adaptation of these relatively aggressive oncologic principles to the spin e h as been left open to th e person al in terpretat ion of th e t reat ing su rgeon , as op posed to a un iversal eviden ce-based agreem en t on appropriate care. For t un ately th rough collaborat ive e orts of th e w orld’s sp in e on cology surgeon s, m any of w h om are auth ors in th is book, an d educat ion an d research organ izat ion s su ch as AO, th ere is greater stan dardizat ion an d im proved ou tcom es in th e m an agem en t of th ese pat ien ts.

xii

Guest Editors’ Preface We are h op efu l th at th is gu ide w ill con t rib u te to th e evolving realizat ion th at eviden cebased can cer t reat m en t can be applied to prim ar y sp in e t u m or m an agem en t as w ell. Th e sh ear im m en sit y, m orbidit y, an d resou rce con su m pt ion of m any of th ese procedu res, in a set t ing of pat ien t qu alit y of life, m u st be a take h om e m essage from th is book. Fur th erm ore t h e read ers m u st recogn ize t h e absolu te n e-

cessit y for st an dard ized st aging an d langu age, along w it h early m u lt id iscip lin ar y inp u t in m an agem en t , if w e are going to provide th ese u n ique an d ch allenging pat ien ts w ith th e h igh est level of care possible. Charles G. Fisher, MD, MHSc, FRCSC Stefano Boriani, MD Ziya L. Gokaslan, MD, FACS

Contributors

Luca Am e ndo la, MD Un it of Sp in e Surger y Ospedale Maggiore Bologn a, Italy Christo phe r P. Am es, MD Professor Clin ical Neurological Su rger y Director Spin e Tu m or an d Spin al Deform it y Su rger y Codirector Spin al Surger y an d Un iversit y of Californ ia– San Fran cisco Spin e Cen ter Director Spin al Biom ech an ics Laborator y San Fran cisco, Californ ia Stefano Bandie ra, MD Un it of On cologic an d Degen erat ive Spin e Su rger y Rizzoli In st it u te Bologn a, Italy Stefano Bo riani, MD Un it of On cologic an d Degen erat ive Spin e Su rger y Rizzoli Or th opedic In st it u te Bologn a, Italy

Justin M. Broyles, MD Plast ic Su rgeon in Residen ce Depar t m en t of Plast ic an d Recon st ru ct ive Su rger y Joh n s Hopkin s Sch ool of Medicin e Balt im ore, Mar ylan d Sim one Co langeli, MD Un it of On cologic an d Degen erat ive Spin e Su rger y Rizzoli In st it u te Bologn a, It aly Nico las Dea, MD, FRCSC Neu rosu rgeon Com bin ed Neu rosu rgical an d Or th op edic Spin e Program Van couver Gen eral Hospital Un iversit y of Brit ish Colum bia Blusson Spin al Cord Cen ter Van couver, Brit ish Colum bia Vedat Deviren, MD Associate Professor Clin ical Or th opaedics Or th opaedic Su rger y Un iversit y of Californ ia–San Fran cisco Spin e Cen ter San Fran cisco, Californ ia

xiv

Contributors Christian P. DiPao la, MD Depar t m en t of Or th op ed ics, Sp in e Su rger y Un iversit y of Massach u set ts Mem orial Medical Cen ter Worcester, Massach u set t s

Derek G. Ju, MD Medical St u den t , MD Can didate Depar t m en t of Neu rosu rger y Joh n s Hopkin s Un iversit y Balt im ore, Mar ylan d

Charles G. Fishe r, MD, MHSc, FRCSC Professor an d Head Division of Sp in e Su rger y Depar t m en t of Or th op aedic Su rger y Un iversit y of Brit ish Colum bia Van couver, Brit ish Colu m bia

Sudhir Kathuria, MD Assist an t Professor Radiology Neu rological Su rger y an d Neurology Director Spin e In ter ven t ion s, In ter ven t ion al Neu roradiology Balt im ore, Mar ylan d

Daryl R. Fo urney, MD, FRCSC, FACS Assist an t Professor Division of Neu rosurger y Royal Un iversit y Hospital Un iversit y of Saskatch ew an Saskatoon , Saskatch ew an

Bre nt Y. Kim ball, MD Depar t m en t of Neu rosu rger y Un iversit y of Ten n essee Health Scien ce Cen ter Mem p h is, Ten n essee

Alessandro Gasbarrini MD Un it of On cologic an d Degen erat ive Spin e Su rger y Rizzoli In st it u te Bologn a, Italy Riccardo Gherm andi, MD Un it of On cologic an d Degen erat ive Spin e Su rger y Rizzoli In st it u te Bologn a, Italy Ziya L. Go kaslan, MD, FACS Don lin M. Long Professor Professor of Neu rosu rger y, On cology, an d Or th opaedic Su rger y Director, Neu rosu rgical Sp in e Program Vice Ch air, Dep ar t m en t of Neu rosurger y Joh n s Hop kin s Un iversit y Sch ool of Medicin e Balt im ore, Mar ylan d Mari L. Groves, MD Assist an t Residen t Depar t m en t of Neu rosu rger y Joh n s Hop kin s Hospit al Un iversit y Sch ool of Medicin e Joh n s Hop kin s Hospit al Balt im ore, Mar ylan d

Ilya Laufer, MD Depar t m en t of Neu rosu rger y Mem orial Sloan -Ket tering Can cer Cen ter Depar t m en t of Neu rological Su rger y Weill Corn ell Medical College New York, New York Mélissa Nadeau, MD, FRCSC Adu lt Sp in e Fellow Depar t m en t of Or th opaedic Surger y Van couver Gen eral Hospital Un iversit y of Brit ish Colum bia Van couver, Brit ish Colum bia Michael C. Oh, MD, PhD Meth odist Brain an d Spin e In st it ute Dallas, Texas Scott H. Okuno , MD Professor Depar t m en t of On cology Mayo Clin ic Roch ester, Min n esota Shreyaskum ar Patel, MD Professor an d Depu t y Depart m en t Ch air Depar t m en t of Sarcom a Medical On cology Division of Can cer Medicin e Universit y of Texas M.D. Anderson Cancer Center Houston , Texas

Contributors Y. Raja Ram pe rsaud, MD, FRCS (C) Or th opaedic Surgeon Division of Orth opaedic Su rger y Ar th rit is Program Toron to Western Hospit al Un iversit y Health Net w ork Toron to, On tario Steve n I. Ro binso n, MBBS In st r uctor Medicin e an d On cology Depar t m en t of On cology Mayo Clin ic Roch ester, Min n esota Justin M. Sacks, MD FACS Assist an t Professor Depar t m en t of Plast ic/Recon st ru ct ive Su rger y Joh n s Hop kin s Sch ool of Med icin e Balt im ore, Mar ylan d Row an Scho ute n, FRACS, MBChB, BSc Or th opaedic Surgeon For té Health Bu ilding Ch ristch urch , New Zealan d Daniel Michael Sciubba, MD Associate Professor of Neu rosu rger y, On cology & Orth opaedic Su rger y Joh n s Hop kin s Un iversit y Balt im ore, Mar ylan d

Jackso n Sui Hosp it al an d Health Care Profession al Birm ingh am , Mich igan Peter Paul Varga, MD Director Nat ion al Cen ter for Spin al Disorders Budapest , Hungar y Jean-Paul Wolinsk y Associate Professor Neu rosurger y an d On cology Depar t m en t of Neu rosu rger y Joh n s Hopkin s Un iversit y Sch ool of Medicin e Balt im ore, Mar ylan d Yo shiya (Jo sh) Yam ada, MD FRCPC Depar t m en t of Radiat ion On cology Mem orial Sloan -Ket tering Can cer Cen ter New York, New York Patricia L. Zadnik, BA Spin al On cology Research Fellow Sciu bba Lab Joh n s Hopkin s Medicin e Balt im ore, Mar ylan d

xv

1 Evaluation and Decision Making Stefano Boriani and Charles G. Fisher

■ Introduction

■ Clinical Findings

and Imaging Prim ar y bon e t u m ors of th e spin e are ver y rare,1,2 com prising on ly 10% or less of all bon e t u m ors. In th e Un ited States abou t 7,500 n ew cases are est im ated per year. Th e overall w orld occurren ce can be expected to be 2.5 to 8.5 cases per m illion in h abit an t s p er year. Com pared w ith p rim ar y sp in al t u m ors, m et astat ic t u m ors are m u ch m ore com m on ; th e spin e is th e m ost com m on skelet al region for secon dar y t um ors. Th e occurren ce of spin al m etast ases in th e clin ical cou rse of com m on solid t u m ors is rep or ted to be 20 to 70%; th erefore, a 30 to 50 t im es m ore frequ en t in ciden ce is repor ted com pared w ith th at of p rim ar y bon e t u m ors of th e spin e. Of 43,735 prim ar y bon e t u m ors review ed in th e literat u re on bon e t um ors, 1,851 (4.2%) w ere located in th e sp in e above th e sacru m . Th ese gures explain w hy th ey are often un susp ected, m isdiagn osed, an d, in m any cases, u n for t un ately, in correctly t reated. As th e occurren ce of lum bar sten osis, cer vical radicu lop athy, an d acu te sp in al cord inju r y is est im ated at 300 cases, 83 cases, an d 5 cases p er 100,000 person s p er year, respect ively,2 an exp er t ise in diagn osis an d t reat m en t st rategy is required for early recogn it ion of a spin e t u m or an d for di eren t iat ing a p rim ar y t u m or from a m etastat ic t u m or.

Back an d n eck pain related to act ivit y are ver y frequen t sym ptom s, par t icu larly in adult s, an d m ostly related to disk p rolap se, degen erat ive ch anges, an d m yofascial st rain s. Conversely a spin e t um or sh ould be rst suspected w h en a pat ien t w ith a h istor y of can cer (cu rren t , recen t , or in th e past) com plain s of back or n eck pain , p ar t icu larly if it is p rogressive, u n relen ting, n ot closely related to act ivit y, an d in creasing du ring th e n igh t . Tu m or grow th m ay cau se exp an sion of th e bony cor tex of t h e vertebral body, w h ich resu lt s in p at h ological fract u re an d invasion of paraver tebral soft t issues. Pain in th ese cases can be associated w ith acu te or ch ron ic com p ression of n eu rologic st r u ct u res th at can resu lt in radicu lopathy an d or m yelop athy. Som e lesion s do n ot cau se pain . Laten t lesion s (h em an giom as, brou s dysp lasia, exostosis) are asym ptom at ic by d e n it ion , an d d iagn osis is ach ieved in cid en t ally on im agin g perform ed for oth er reason s. Cases of in ciden tal detect ion of ch ordom a are also repor ted. Th is m align an t t u m or is ch aracterized by ver y slow grow th an d, in an atom ically volu m in ou s region s su ch as th e sacru m , can grow to ver y large sizes before detected. Pain fu l scoliosis in an adolescen t is st rongly suggestive of osteoid osteom a or osteoblastom a.

2

Chapter 1 Usu ally n ot visible on st an dard radiogram s, th e xed cu r ve w ith ou t a com p en sator y cu r ve an d rot at ion is st rongly suggest ive. Tech n et iu m isotope bon e scan h elps to localize th e lesion . A com p u ted tom ograp hy (CT) scan , perform ed at th e level of in ten se u pt ake, d isp lays th e p ath ogn om on ic im age of osteoid osteom a: a sm all islan d of path ological ossi cat ion surroun ded by a lyt ic h alo an d frequen tly by a w ide rim of react ive bon e form at ion . Tech n et iu m isotop e bon e scan is also able to localize m u lt ip le lesion s. Th e role of p osit ron em ission tom ograp hy (PET) scan is becom ing m ore relevan t , p ar t icu larly in it s abilit y to differen t iate t um ors from in fect iou s diseases. Im aging st u dies (CT scan an d m agn et ic reson an ce im aging [MRI]) can gen erate a w orking diagn osis, as som e pat tern s are qu ite ch aracterist ic. Gian t cell t um or an d Ew ing’s sarcom a are lyt ic con dit ion s, an d m ost osteosarcom as are characterized by extensile aggressive pathological bon e for m at ion w ith ill-d e n ed borders. A m u lt icam eral balloon -like pat tern w ith dou ble-den sit y con ten t is t yp ical of an eu r ysm al bon e cyst s. In lt rat ing erosion s in side th e can cellous bone of a vertebral body arising from th e p osterior w all is suggest ive of ch ord om a. Soft t issu e m asses arising from th e p osterior elem en ts w ith rou n ded calci cat ion s are t yp ical of p erip h eral ch on drosarcom as. Angiography show s pathological vascularit y, an d select ive ar terial em bolizat ion h as becom e an in dispen sable tool to reduce in t raoperat ive bleeding. Histological d iagn osis sh ou ld alw ays be ach ieved by biopsy, bu t clin ical, laborator y, an d im aging st u dies are im p or tan t to orien tate diagn osis an d to select th e biop sy tech n iqu e.

■ Biopsy Th e goal of a biop sy is to obt ain a sp ecim en of th e t u m or, w h ich is rep resen t at ive of th e lesion an d large en ough to allow h istological an d u lt rast r u ct u ral an alysis as w ell as im m u n ologic stain s. Th e su rgeon m u st recogn ize th e vit al p ar t of th e t u m or an d discard th e n ecrot ic

or react ive part . Cu lt ures sh ould be t aken to rule ou t in fect ion . Th ere are th ree t radit ion al biopsy tech n iques: in cision al, n eedle or t rocar, an d excision al. A n u m ber of basic p r in cip les sh ou ld be obser ved w h en p erform ing in cision al an d excision al biopsies in order to preven t t um or contam ination of the surrounding tissue, w hich is th e m ajor risk of biopsy. Tran sverse in cision s an d aps sh ou ld be avoided ; th e t u m or sh ould be approach ed in th e m ost direct m an n er possible, avoiding th e an atom ic in tersp aces (socalled ext racom p ar t m en t al sp aces) com m on ly follow ed in or th opedic surger y. Th e approach should cross m uscular structures. Tissues should be handled carefully, hem ostasis should be m eticu lou s, an d su t u ring perform ed in all th e an atom ic layers. Bleeding from exp osed bon e or from un cau terized vessels an d injured m uscle w ill form a postoperat ive h em atom a th at m ay car r y t u m or cells beyon d th e m argin s of th e intended excision and contam inate tissues proxim al or dist al to th e prim ar y lesion . Th e m argin of th e soft t issu e com pon en t of th e m ass is often th e m ost h elpful to biopsy, as cen t ral por t ion s are frequ en tly n ecrot ic. Th e su rgeon sh ou ld take care n ot to cr u sh or distor t th e sp ecim en , so as to m ain t ain it s arch itect u re. Fin ally, if t h e d e n it ive excision follow s t h e biopsy during th e sam e session , it is essen t ial th at all in st r u m en t s u sed d u r ing t h e biop sy be d iscard ed . Th e eld sh ou ld be re-d rap ed , an d t h e su rgeon s sh ou ld ch ange gow n s an d gloves before th e d e n it ive excision . If fu sion is plan n ed, bon e graft sh ou ld be taken th rough a separate surgical set up. If th e plan n ed de n it ive p rocedu re is an en -bloc resect ion , th e biopsy in cision from th e skin to th e t um or m ass sh ou ld be p laced so th at it m ay be excised in a sin gle block w it h t h e t u m or an d it s m argin s. Th is is rarely p ossible in t h e t reat m en t of t u m ors of t h e sp in e, if th e ep id u ral sp ace h as been violated . Frozen section biopsy can be obtained during th e excision al procedu re w h en th e diagn osis is h igh ly probable an d just requ ires con rm at ion . In rare sit u at ion s, a pat ien t m ay h ave t w o sep arate p rim aries; if th ere is any dou bt , th e result s of th e frozen sect ion sh ou ld be received

Evaluation and Decision Making before proceeding w ith th e resect ion , su ch as in a sit u at ion of an assu m ed spin e m et ast asis. Excisional biopsy can be considered only for cond it ion s w h ose radiograp h ic pat tern is path ogn om on ic, for exam p le osteoid osteom a. Th e correct w ay to obt ain a t um or sam ple from a vertebral body w ithout violating the epidural space is w ith a tran spedicular biopsy. The p edicle m u st be p erforated by a drill or t rocar. The surgeon m ust be careful not to perforate the pedicle wall, causing epidural space contam ination. A sm all straight curette is then introduced to rem ove the specim en . Th e perforated pedicle m ust be lled w ith acr ylic cem en t to preven t tum or back ow. The soft tissue track can be rem oved togeth er w ith th e su rrou n ding m uscle. Sm all t rocar biop sies are su bject to sam p ling errors an d p rovide sm all specim en s for evalu at ion . Th e au th ors’ preferred tech n iqu e is to biopsy w ith a 12-gauge t rocar perform ed u n d er CT gu idan ce. Th is tech n iqu e p rod u ces t issu e cores t h at allow for h istological arch itect u re an d im m u n oh istoch em ical st u d ies. CT gu idan ce allow s th e in ter ven t ion al radiologist to reach even t h e sm allest lesion s in sid e th e ver tebral body. Diagn osis sh ou ld alw ays be con r m ed on h istological diagn osis; biop sy is m an dator y. Conversely, som e laten t , asym ptom at ic lesion s such as h em angiom a, brou s dysplasia, an d exostosis are ch aracterized by such a path ogn om on ic im aging th at a biop sy is n ot alw ays n ecessar y. W h en th is is th e case it is recom m en ded to follow th e evolut ion of th e disease by sequ en t ial im agin g, in it ially ever y 4 to 6 m on t h s, to fu lly assess t h e laten cy of t h e con dit ion . In rare sit u at ion s a pat ien t m ay presen t w ith n eu rologic deteriorat ion in a set t ing of a p ossible p rim ar y t u m or. Th e su rgeon is faced w ith th e decision to st age th e t u m or an d ob t ain a d iagn osis or p roceed w it h n eu rologic d ecom pression , poten t ially ren dering th e patient incurable due to the tum or being violated. Each case is di eren t , an d con dit ion al probabilit ies m u st be con sidered an d in tegrated in a sh ared d ecision -m aking p rocess w ith t h e p at ien t . Often som e delay in th e t reat m en t can be ach ieved w ith th e u se of steroids.

■ Oncological Staging En n eking et al3 proposed in 1980 a system to st age th e biological beh avior of t um ors of bon e an d soft t issu e. Th is system w as later ap plied also to th e spin e in several review s. Th is system d ivides ben ign t u m ors in to t h ree st ages (S1, S2, an d S3) an d localized m align an t t u m ors in to fou r stages (IA, IB, IIA, an d IIB). Tw o fu rth er st ages in clude m et astat ic h igh -grade in t ra- an d ext racom part m en tal m align an t t um ors (IIIA an d IIIB). Th is classi cat ion is based on clin ical, im aging, an d h istological n dings. Each st age is related to overall progn osis an d, by n ecessit y, d irect ly lin ked to categor ies of su rgical p roced u res based on t h e con cept of “m argin .”3,4

Benign Tumors Th e rst st age of ben ign t u m ors (S1, laten t , in act ive; Fig. 1.1) in cludes asym ptom at ic lesion s, bord ered by a t r u e cap su le, w h ich is u su ally seen as a sclerot ic r im on p lain rad iogram s. Th ese t u m ors do n ot grow, or grow ver y slow ly. No t reat m en t is requ ired , u n less su rger y is n eeded for decom pression or st abilizat ion . Ben ign stage 2 t u m ors (S2, act ive; Fig. 1.2) grow slow ly, cau sing m ild sym ptom s. Th e t um or is bord ered by a t h in cap su le an d by a layer of react ive t issue, som et im es foun d on plain radiogram s as an en largem en t of th e t u m or ou tlin e. Th e bon e scan is p osit ive. An in t ralesion al

Fig. 1.1 Stage 1 benign tumor. The tumor is latent and inactive, and a capsule surrounds the lesion.

3

4

Chapter 1

Fig. 1.2 Stage 2 benign tumor. The tumor grows slowly. The capsule is thinner. The tumor is surrounded by reactive tissue, as expression of local reaction to tumor growth.

Fig. 1.3 Stage 3 benign tumor. The tumor grows faster and invades early the epidural space or the perivertebral region. The capsule is very thin, and the reactive zone is wider.

excision can usually be perform ed w ith a low rate of recurren ce. Based on ver y low qualit y eviden ce, th e in ciden ce of recu rren ces can be low ered fu rt h er by local adjuvan t s (cr yoth erapy, em bolizat ion , rad iat ion th erapy). Ben ign st age 3 t u m ors (S3, aggressive; Fig. 1.3) are rap id ly grow ing, w it h a cap su le t h at is ver y th in , discon t in u ed, or absen t . Th e t u m or in vades n eigh boring com p ar t m en t s, an d a w ide react ive hyp er vascu larized t issu e (p seu docap su le) is often fou n d, som et im es perm eated by n eoplast ic digit at ion s. Th e bon e scan is h igh ly p osit ive. In dist in ct m argin s are seen on plain X-rays, CT scan sh ow s ext racom p ar t m en t al exten sion , an d MRI clearly d e n es a p seu d o-

cap su le w ith sign i can t edem a an d react ion . In t ralesion al excision (cu ret tage), even if augm en ted by rad iat ion , can be associated w it h a sign i can t rate of local recu r ren ce. En -bloc excision w ith m argin al or w ide m argin s is t h e t reat m en t of ch oice.

Fig. 1.4 Stage IA malignant tumor: low grade, intracompartm ental. The pseudocapsule may include a small island of tumor.

Fig. 1.5 Stage IB malignant tumor: low grade, extracompartmental. The pseudocapsule may include a small island of tumor.

Malignant Tumors Low -grad e m align an t t u m ors are assign ed st age I, su bdivided in to IA (th e t u m or rem ain s in side th e ver tebra; Fig. 1.4) an d IB (t um or in vades paraver tebral com par t m en ts; Fig. 1.5). No t r u e cap su le is associated w it h t h ese lesion s, bu t a t h ick p seu d ocap su le of react ive

Evaluation and Decision Making

Fig. 1.6 Stage IIA malignant tumor: high grade, intracompartm ental. The pseudocapsule is fully invaded by the tum or.

t issu e perm eated by sm all m icroscop ic islan ds of t u m or is seen . A resect ion p erform ed along th e pseudocapsu le (m argin al) can leave residu al foci of act ive t um or; stereotact ic radiat ion or proton -beam th erapy can be added to red uce th e risk of recu rren ce. Th e t reat m en t of ch oice, if feasible, is a w ide en -bloc resect ion , an d carefu l m argin al dissect ion if n eu rologic t issues are involved. High -grade m align an cies are de n ed as IIA (Fig. 1.6) an d IIB (Fig. 1.7). Th e n eoplast ic grow th is so rapid th at th e h ost h as n o t im e to form a con t in u ou s react ive t issue. Th ere is con t in u ous seeding w ith n eoplast ic n odules (satellites). Moreover, th ese t um ors can h ave n eoplast ic n odules at som e dist an ce from th e m ain t u m or m ass (skip m et ast ases). Th ese m align an cies are gen erally seen on plain radiogram s as radiolucen t an d dest r uct ive an d in m any cases are associated w ith a p ath ological fract u re. Invasion of th e epidu ral space is rapid in stage B, particularly in sm all cell tum ors (Ew ing’s sarcom a, lym p h om as) ch aracterized by sem i u id t issu e, w h ich are able to occu py th e epidural space after in lt rat ing th e cor t ical border of th e ver tebra. Th e m argin of th e resect ion m ust be w ide (it is n ot possible to ach ieve a “radical” m argin , in th e spin e), an d courses of radiat ion an d ch em oth erapy (according to th e t um or t ype) m ust be con sidered for th e local con t rol an d for th e avoidan ce of distan t spread. Often in areas of an atom ic con st rain t s, m argin al m argin s (discu ssed below ) or plan n ed in tralesion al t ran s-

Fig. 1.7 Stage IIB malignant tumor: high grade, extracompartmental. The pseudocapsule is fully invaded by the tum or. A sm all island of tumor can be found outside the main tumor mass.

gression s m u st be accepted, or con siderat ion given to th e sacri ce of n eu rologic st r u ct ures. St ages IIIA an d IIIB d escr ibe t h e sam e lesion s as IIA an d IIB, bu t are associated w it h d ist an t m et ast asis, an d in th e m ajorit y of sit u at ion s w ould be con sidered in cu rable.

■ Margins and Surgical

Planning As st aging m u st be based on th e n at u ral h istor y, surgical plan n ing m ust be based on w h at th e p rocedu re ach ieves in relat ion to th e m argin of th e lesion . Th e n al path ological m argin ach ieved th rough surgical tech n ique is directly related to progn osis. Sten er an d Joh n sen 5 w ere th e rst (to our kn ow ledge) to apply ext rem it y on cological p rin cip les to th e sp in e; th ese con cept s w ere p u blish ed in exh au st ive repor t s of en-bloc resections in the spine. Roy-Cam ille and Tom it a later p op u larized th e tech n iqu es of en bloc resect ion . Many pap ers review ed series of p at ien ts an d p rovided t ips an d det ails for variou s tech n iqu es.1,6–9 Sin ce th e late 1980s Wein stein an d Borian i10 t ried to focu s on a un i ed classi cat ion th at prop osed envelop es of resection as Enneking et al3 had don e for th e extrem ities. By retrospectively review ing his large case series, Borian i w as able to establish sign i can t relat ion sh ip s bet w een m argin an d ou tcom e sp eci c to th e spin e.

5

6

Chapter 1

Fig. 1.8 The Weinstein-Boriani-Biagini (WBB) surgical staging system . On the transverse plane, the vertebra is divided into 12 radiating zones (numbered 1 to 12 in a clockwise order) and into vie layers (A to E from the prevertebral to the dural

involvement). (From Boriani S, Biagini R, De Iure F, et al. En bloc resections of bone tumors of the thoracolumbar spine. A preliminary report on 29 patients. Spine 1996;21:1927–1931. Reproduced with permission.)

Th e Wein stein -Borian i-Biagin i (W BB) st aging system (Fig. 1.8) h as been su bjected to several clin ical evalu at ion s an d recen tly su bm it ted to a st udy of it s reliabilit y an d validit y by an in tern at ion al m u lt idiscip lin ar y grou p of sp in e tum or experts,10 w hich foun d m oderate in terobser ver reliabilit y an d su bst an t ial in t raob ser ver reliabilit y. Th e W BB staging system focu ses on th e exten t an d locat ion of th e t u m or. In th e t ran sverse plan e, th e vertebra is divided in to 12 radiat ing zon es (n um bered 1 to 12 in a clockw ise order) an d in to ve layers from th e p rever tebral to th e dural involvem en t (A th rough E). Th e longit udin al exten t of th e t um or is recorded by iden t ifying th e speci c ver tebrae in volved . Th is system allow ed for a m ore rat ion al approach to surgical planning, provided that all e or ts are m ade to perform surger y along th e required m argin s.

d oing m u lt icen ter st u dies, w h ich are so im por t an t in th is ch ron ically u n d erp ow ered p op u lat ion . A com m on ter m in ology is requ ired becau se if an en -bloc resect ion an d a gross tot al excision are erron eously described w ith th e sam e term s, it is di cu lt , if n ot im possible, to com pare resu lts. Surp risingly, m any auth ors n eglect or sim ply ign ore th e issue of t r ue m argins. Without a clear description of surgical and h istopath ological m argin s, it is im possible to evaluate the outcom e of any surgical treatm en t. To stress the im portance of a com m on language, the term s proposed by Enneking et al3 and later applied to th e spin e 9 ser ve as exam p les.

Terminology Th e ap p licat ion of a com m on term in ology is m an dator y for exch anging in form at ion an d for

Radical Resection Th is term sh ou ld be ap p lied to th e en -bloc rem oval of th e t u m or togeth er w ith th e w h ole com p ar t m en t of origin . An exam p le w ou ld be a th igh am putat ion for a t um or arising in th e t ibia, or a scapu lectom y for a t u m or arising an d st ill con t ain ed in th e scapu la. Con sequ en tly, a radical resect ion in t h e spin e de n es an en bloc resect ion of a full vertebra, provided th at

Evaluation and Decision Making th e t um or is fu lly con n ed in side th e vertebra, in clu d in g t h e sp in al cord above an d below . Even in th is dram at ic case, if th e t um or h as grow n in to th e epidural space, th e term radical resect ion w ill n ot be appropriate, as th e epid ural space is to be con sidered an ext racom p ar t m en t al area exten ding from th e sku ll to th e sacrum .

th e t u m or in order to ach ieve a w ide m argin . Marginal m argin m ean s th at th e plan e of dissect ion h as been p erform ed along th e react ive zon e. Th is is ver y close to th e t um or an d th e possibilit y that satellites of t um or m ay be tran sgressed or left beh in d is real. Int ralesional m argin m eans that the tum or has been violated, and th e t u m or sp ecim en is n ot covered by h ealthy t issue. Tum or cells w ill be left beh in d.

Intralesional Excision In t ralesion al excision m ean s a piecem eal rem oval of th e t u m or (also called cu ret t age, gross tot al resect ion , an d debulking). It is an in t ralesion al procedu re; th e t u m or is violated an d any ch an ce of obtain ing t um or-free m argin s is lost . Th is procedu re is ap propriate for stage 2 ben ign t um ors an d for m etast ases. Microscopic an d possible m acroscopic t um or is alw ays left after t h ese p roced u res, an d local p rogression can be expected according to th e grow ing poten t ial of th e disease.

En-Bloc Resection En -bloc resect ion en t ails t h e rem oval of t h e t u m or as a sin gle in t act w h ole, fu lly en cased by a cu of h ealthy t issue (a circum feren t ial m argin of t u m or-free n orm al t issu e). For th e en -bloc resect ion to h ave ach ieved it s su rgical goal, t h e p at h ologist m u st con rm t h at t h e resected t issu e arou n d th e t u m or m ass (t h e so-called m argin ) is t u m or free. En -bloc resect ion s can be carr ied ou t w it h m argin al m argin s or even p lan n ed t ran sgression s, w ith th e n al m argin ou tcom e again determ in ed by th e path ologist . Th e e ect iven ess of th e m argin as a barrier dep en ds on it s qu alit y an d also on th e aggressiven ess of th e t u m or. Th e m argin is de n ed as “w ide” w h en t h e p lan e of dissect ion h as been perip h eral to th e react ive zon e th rough n orm al t issu e. Th e qualit y of th e m argin an d th ickn ess are relevan t but n ot w ell quan t i ed. A fascial barrier represents a w ide m argin, w hereas 1 cm of m uscle or can cellous bon e m ay be in adequ ate to rep resen t an ap p rop riate barrier. Th e m ajor issu e in t h e d ecision -m akin g p rocess of plan n ing en -bloc resect ion in th e spin e is th e fu n ct ion al im p licat ion s (for exam p le, n er ve root s) related to st r u ct u res being resected w ith

Palliative Procedure Th is term describes all th e su rgical p rocedu res generally directed tow ard a functional purpose, for exam p le to d ecom p ress n eu rologic st r u ct u res or st abilize t h e sp in e. Th is t yp e of p roced u re is gen erally p er for m ed for m et ast at ic d isease.

■ Surgical Techniques:

En-Bloc Resections Th e tech n iqu es of en -bloc resect ion for t u m ors of the vertebral body are w ell know n.6 The W BB st aging system 9 can be h elpful in stan dardizing th e su rgical plan n ing of en -bloc resect ion according to th e region of th e sp in e an d th e t u m or exten t an d locat ion w ith in th at region . Th e great variabilit y of th ese t w o p aram eters dict ates th at th e sam e su rgical p rocedu re can n ot be perform ed in all cases an d th at surgical plan n ing is usu ally di eren t for each case. Un der th e gu idan ce of W BB staging system , six d i eren t ap proach es can be st an dardized: on ly anterior; only posterior; an terior rst, posterior secon d; posterior rst , dou ble an terior secon d an d t h ird ; p oster ior rst , com bin ed an ter ior an d p osterior secon d; an terior rst , p osterior secon d, com bined an terior (con t ralateral) an d p osterior. Th ese six approach es can be com bin ed to create 10 di eren t t ypes of surger y. Detailed descript ion of h ow to apply th e W BB system is beyon d th e scope of t h is ch apter, but it becom es clear as on e plan s to do an en -bloc resect ion an d con siders all th e an atom ic an d fu n ct ion al issu es related to safely an d e ect ively rem oving a t u m or en bloc. Un iqu e to bon e t u m ors in th e sp in e is th e p roxim it y of fun ct ion ally relevan t vit al st ru c-

7

8

Chapter 1 t u res (e.g., sp in al cord , n er ve root s, large arter ies an d vein s) th at rep resen t th e m argin . In su ch cases sacri ce of th ese st r uct u res m ust be p rop osed for p erform ing on cologically ap prop riate surger y. Speci c tech n iqu es of en -bloc resect ion h ave been rep or ted , w it h t h e sacr ice of th e follow ing st r u ct u res: d u ra,11 cau da equ in a,12 cer vical n er ve root s,7 sp in al cord ,13 an d visceral organ s.14 Plan n ing of th e approp riate en -bloc resect ion m u st th erefore con sider th e m argin s to be ach ieved . Th e ap p roach or th e app roach es required an d th e t im ing can be decided according to th e W BB su rgical st aging system . Th e single an terior ap p roach en ables en bloc resect ion of sm all t u m ors of th e th oracic an d lu m bar ver tebral bodies. Th ey m u st be en closed in side sectors 8 to 5, arising in layer A

an d B, bu t n ot exten ding to layer C. In th is case a posterior approach is required to provide an appropriate m argin un der direct visual con t rol by en tering th e can al an d releasing th e d ura. Th e single posterior approach can be appropriate, as described by Roy- Cam ille an d Tom ita to rem ove as a w h ole a t um or arising in th e ver tebral body of a th oracic ver tebra. Criteria to ach ieve appropriate m argin s in clude sector 9 or 4 free from t um or. If th e t u m or grow s in layer D, th e m argin w ill be in t ralesion al du ring th e release from th e d u ra. If t h e t u m or grow s in layer A, th e m argin w ill be in t ralesion al du ring th e release from th e an terior st r u ct u res (Fig. 1.9). Th e single posterior approach w ith sagit tal osteotom y enables en-bloc resection of a tum or eccen t rically grow ing in th e th oracic an d lu m -

Fig. 1.9 Single posterior approach. It enables the removal by en-bloc resection of a tumor arising in the vertebral body of a thoracic vertebra. Criteria to achieve appropriate m argins include sector 9 or 4 free from tumor. If the tum or grows in layer D, the margin will be intralesional during the release from the dura. If the tum or grows in layer A, the margin will be intralesional during the release from the anterior structures. This is a t wo-step procedure, as shown by roman numerals on the gure: (I) Piece-

meal excision of the posterior arch not involved by the tumor. At least four sectors are required, starting from sector 4 or from sector 9. Release from the dura and section of the nerve root(s) involved by the tumor. (II) Blunt dissection of the anterior part of the vertebral body from the mediastinum, osteotomy, or diskectomy above and below the tumor, with full release from the dura and nalizing the resection.

Evaluation and Decision Making bar spin e (Fig. 1.10), p rovided th e body is n ot involved over sector 5 at left an d over sector 8 at righ t . At least th ree sectors posteriorly m u st n ot be involved by t h e t u m or (4 to 1–2 or 12–11 to 9). A com bination of an terior approach rst and p osterior app roach secon d en ables en -bloc resect ion of a t um or located in th e th oracic an d lum bar spin e (Fig. 1.11) an d in th e cer vical sp in e grow ing an teriorly (layer A). W h en th e t um or is grow ing an teriorly (layer A) an an terior approach m u st be perform ed as th e rst step to provide a w ide/m argin al m argin un der visual con t rol. If th e t um or exten sion is m ostly eccen t ric (i.e., involving sectors 5 to 11), a sag-

ittal osteotomy from posterior to anterior can be perform ed to com plete the resection (Fig. 1.12). Huge lu m bar t u m ors can also be rem oved by en -bloc resect ion via t w o su rgical st ages: th e rst st age p oster ior, an d th e secon d an an ter ior ap p roach follow ed by a n al p oster ior ap p roach (Fig. 1.13). Th is tech n iqu e w as d escribed by Roy- Cam ille for lu m bar t u m ors an d is associated w ith th e h igh est rate of m orbidit y an d com plicat ion s.15 Th e com bin at ion of an terior and posterior approaches is preferred w hen it is assum ed th at com p lex m an euvers w ill be n ecessar y to rem ove th e specim en . To resect a t um or of L5, th ree st ages are suggested: rst , an terior ap proach to t h e con t ralateral side of

Fig. 1.10 Single posterior approach with sagit tal osteotomy. A tum or excentrically growing in the lum bar spine can be removed en bloc by a single posterior approach provided that the body is not involved over sector 5 at left and over sector 8 at right. At least three sectors posteriorly must not be involved by the tumor (4 to 1–2 or 12–11 to 9). This is a four-step procedure, as shown by roman numerals on the gure: (I) Provide the appropriate margin over the tumor posteriorly growing by resecting inside the posterior m uscles covering the tumor mass if it is expanding in layer A. The release will proceed laterally until the lateral side of the

vertebral body. In the thoracic spine the pleura can be left on the tum or, and in the lumbar spine the posterior part of the psoas must be dissected, but the segmental vessels must be found and ligated. (II) Piecemeal excision of the posterior arch not involved by the tumor. This step includes the approach to the canal, release of the dura from the tumor (if the tumor grows in layer D, the margin will result intralesional), and section of the nerve root(s) involved by the tumor. (III) Displace carefully the dura and perform the osteotomy from posterior to anterior in sector 8 or 5. (IV) The specimen is removed.

9

10

Chapter 1

Fig. 1.11 Two approaches: anterior rst, posterior second in the thoracic and in the lum bar spine. When the tum or is growing anteriorly (layer A), an anterior approach m ust be performed as the rst step to provide a wide/m arginal m argin under visual control. This is a four-step procedure, as shown by roman numerals on the gure: (I) In case of tum ors mostly occupying the vertebral body, the anterior approach can be the rst step to release from mediastinum or retroperitoneal, eventually leaving involved structures as the margin. A sheet of Silastic or similar m aterial can be left as protection. (II)

Posterior approach: piecemeal excision of the posterior arch not involved by the tumor. At least three or four sectors are required, starting from sector 4 or from sector 9. (III) Release of the dura from the tum or, section of the nerve root(s) involved by the tumor, and then provide the appropriate m argin over the tumor posteriorly growing by resecting inside the posterior m uscles covering the tumor m ass if it is expanding in layer A. (IV) The specimen is removed by rotating around the dural sac.

the tum or to release the aorta/cava bifurcation; secon d, p osterior; th ird, an terior an d posterior approach . In plan n ing th e su rgical p roced u re th e cord vascularit y m ust be con sidered. During th ese procedu res, p ar t icu larly w h en th e resect ion is m u lt ilevel or th e t u m or p ar t icu larly h uge, th e fu n ct ion al in tegrit y of th e sp in al cord is at risk m ost ly d u e to t h e m an ip u lat ion of t h e cord du r ing m an eu vers to d eliver t h e t u m or. For exam p le, an an atom ic feat u re u n iqu e to th e h igh er th oracic spin e (T1 to T4) is easier release of th e ver tebrae from th e aor t a, as n o segm en tal ar teries are closely con n ected to th e ver te-

bral body.16 Below T5 th e in tercost al ar teries run aroun d th e ver tebral body an d keep th e aor ta closely related to bon e; th erefore, carefu l ret ract ion is n eeded w h en en -bloc resect ion of th e ver tebral body is p erform ed by th e p oster ior ap p roach . At t h e t h oracolu m bar ju n ct ion t h e d iap h ragm p illar can exten d to T12, m aking th e dissect ion p osteriorly m ore di cu lt . Som et im es th is is best approach ed an teriorly for a safer release. Th e role of t h e ar ter y of Adam kiew icz as a single, exclusive feeding for th e an terior sp in al arter y is controversial. It seem s reasonable that cord vascularit y is not dependent on one artery.

Evaluation and Decision Making

Fig. 1.12 This procedure can be planned in case the tumor is eccentrically growing. Anterior rst, posterior second in the thoracic and lum bar spine, and sagit tal osteotomy posterior to anterior. This is a ve-step procedure, as shown by roman numerals on the gure: (I) In the rst stage the anterior approach provides a wide/marginal m argin under visual control, releasing from mediastinum or from peritoneal, eventually leaving involved structures as the margin. Diskectomies or transversal grooves in vertebral bodies are performed to de ne the upper and lower margins. A sheet of Silastic or any other tissue can be left as protection. In the second stage, the posterior approach, piecem eal excision of the

posterior arch not involved by the tumor is performed. (II) At least three sectors are required, starting from sector 4 or from sector 9. (III) Provide the appropriate m argin over the tumor posteriorly growing by resecting inside the posterior m uscles covering the tumor m ass if it is expanding in layer A. (IV) Release of the dura from the tum or, section of the nerve roots crossing the tumor, and osteotomy posterior to anterior at some distance from the tumor in order to leave uninvolved bone as m argin. (V) The resected specim en can be removed once the upper and lower diskectomies or osteotomies are nalized.

Previou sly, w e perform ed angiograph ic st u d ies before su rger y to iden t ify th e rad icu lom edu llar y ar ter y feeding th e arter y of Adam kiew icz. In fou r cases t h e n er ve root w as sacr i ced w ith out any dam age to cord vascu larit y. Sin ce th en , th e role of such a st udy w as felt to be less crit ical an d did n ot a ect th e plan n ing. Tom it a and his group 16–18 had the sam e experience, and dem on st rated th at th e risk of cord isch em ia is m ostly related to th e n u m ber of con t igu ou s radicu lar ar teries sacri ced rath er th at to a single ar ter y. Fur th erm ore, th ere is som e eviden ce in th e literat u re of m u lt iple feeders to t h e an te-

rior spin al ar ter y. We n ow recom m en d taking n o m ore th an th ree n er ve root s bilaterally in th e th oracic sp in e, an d avoiding acu te sh or ten ing or dist ract ion du ring th e resect ion .19

■ Complications Th e m orbid it y associated w it h en -bloc resect ion s is h igh , often com bin ing t h e risks an d com p licat ion s of an terior sp in e su rger y w ith th ose of m ajor posterior su rger y.

11

12

Chapter 1

Fig. 1.13 Two stages: rst posterior approach, second contemporary anterior and posterior approaches. This technique is t ypical and more appropriate for lumbar tumors and is associated with the highest rate of morbidit y and complications. In thoracic spine it is less advisable. The rst steps are performed with the patient in the prone position: piecemeal excision of the posterior arch not involved by the tumor. This is a four-step procedure, as shown by rom an num erals on the gure: (I) At least three sectors are required, starting from sector 4 or from sector 9. (II) In cases of a tumor posteriorly growing and invading layer A, an appropriate margin must be provided by resecting inside the posterior m uscles covering the tumor mass. Then release the dura from the tum or

(if the tumor grows in layer D, the margin will result intralesional) and section the nerve root(s) crossing the tumor. Diskectomies or transversal grooves in vertebral bodies are performed to de ne the upper and lower margins. The second stage is perform ed with the patient in the lateral position. Anterolateral approach (thoracotomy, thoracoabdom inal, retroperitoneal) and reopening of the posterior approach. (III) To provide an appropriate margin over the tumor, it must remain covered by pleura or by psoas. Spiral wires are used to embolize the segmental arteries to facilitate the release of the aorta on the contralateral side. (IV) Once the upper and lower diskectom ies or osteotomies are nalized, the specimen is removed by combined maneuvers.

Previou s su rger y, rad iat ion , an d p rolonged com bin ed approach es in crease th e risk of com p licat ion s, w it h in fect ion bein g p ar t icu larly th reaten ing, due to th e w ide prolonged exposu re an d th e com p rom ised im m un e st at u s of m any of th ese pat ien ts.15 Tw o cases of late aort ic dissect ion occurred after radiat ion th erapy (doses of 42 an d 44 Gy). In both cases an an terior ver tebrectom y w as perform ed follow ing release of th e aor t a from th e t u m or 15 . Non un ion is a com m on late com plicat ion due to

th e h ost ile environ m en t in w h ich to ach ieve a solid bony fu sion . Mor t alit y is n ot n egligible, w ith a rate of 2.2%.15

■ Chapter Summary Obt ain in g a d iagn osis is t h e rst cr it ical step in th e m an agem en t of prim ar y t u m ors of th e spine. This is achieved through staging that cul-

Evaluation and Decision Making m in ates in a carefu lly p lan n ed CT-gu ided biopsy. To determ in e m an agem en t , th e clin ical, im aging, an d h istological in for m at ion is review ed by a m u lt idisciplin ar y team . Exp ert ise in the biological behavior of the di erent tum or t ypes is m an dator y in deciding th e m ost ap propriate surgical m argin and the surgical techn iqu e to obtain it . To facilit ate th is p rocess, th e En n eking st aging system is u sed. A su rgical st aging system (su ch as th e W BB) is also requ ired for p lan n ing th e su rgical procedu re in order to ach ieve su ch m argin s. Decision m aking sh ou ld in clu de in -depth discu ssion s w ith th e p at ien t to en su re th at all issu es, in clu ding probabilit y of local recu rren ce, su r vival, an d risk, are clearly u n derstood. Th e in t rin sic m orbidit y of en -bloc resect ion s an d th e an atom ic sacri ces required to ach ieve appropriate m argin s m ust be w eigh ed again st th e in creasing risk of local recurrence associated w ith less than opt im al in t ralesion al m argin s, but th is m ay be th e on ly opt ion in som e cases. The com bination of Enneking and W BB staging system s sh ou ld • p reven t h igh m orbid it y su rger y w h en n ot n ecessar y (for exam ple, in st age 2 ben ign or in m et astat ic t um ors); • ach ieve th e appropriate m argin by correctly perform ing en -bloc resect ion s in selected cases; an d • advan ce th e kn ow ledge of th ese t um ors by com p aring u n iform repor t s on ou tcom es of h om ogen eou sly t reated t um ors. It is im por t an t to con sider th e com plicat ion s in th e t reat m en t of p rim ar y t u m ors of th e sp in e, as som et im es m an agem en t of com plicat ion s is m ore di cu lt th an m an agem en t of th e t u m or itself. To th is en d, t h e rst ch an ce is n ot on ly the m ost im portan t, but m ay be the on ly chan ce to cure th e pat ien t . Secon d ch an ces are often too late. Th e abilit y to in tegrate th e best su rgical an d n on su rgical t reat m en t s can a ect n ot only the im m ediate care of the patient, but also, an d m ore im por t an tly, th e long-term su r vival. Tu m ors of th e spin e are u n ique an d require h igh ly skilled team s to deal w ith th e di cu lt decision s invoked by ever y case. Th ese cases are n ot for th e rout in e spin e surgeon or t radit ion al on cological su rgeon .

Acknow ledgment Th e au th ors ackn ow ledge th e ou t st an d ing con tribution of Carlo Piovan i for data collection and origin al draw ings.

Pearls The decision-making process for the treatment of prim ary bone tumors of the spine is based on a thorough understanding of the biology and the behavior of each tumor. This is achieved by diagnosis and oncological staging (Enneking), which dictates the oncologically appropriate treatment (including surgery, radiation therapy, chemotherapy, and other m edical treatm ents). The decision-making process of primary bone tumors is perform ed by a multidisciplinary team and includes the following steps: ◦ Clinical ndings and imaging provide the working diagnosis. ◦ Histology and oncological staging indicate biological behavior and direct the treatment. ◦ Surgery of prim ary tum ors has a curative intent. Computer tomography–guided trocar biopsy is the method of choice to provide an appropriate specim en and the least possible contam ination of the surrounding tissues. The biopsy should be done by the sam e center that will adm inister the treatm ent. Surgery of primary tumors has a curative purpose. Oncologically appropriate surgical treatment can substantially improve the prognosis and can be considered a lifesaving procedure. Shared decision making with the patient is im perative in all cases of primary tum or resection. The patient’s acceptance of functional loss to achieve bet ter local control and prognosis, versus a preference for saving function despite the increased risk of local recurrence and decreased survival, m ust be discussed. Intentional violation of oncological principles such as planned intralesional transgression to preserve vital structures or reduce m orbidit y1 can be considered and should be discussed with the patient and a multidisciplinary team. In this set ting the role of further adjuvant therapy becom es critically important. Pitfalls Treating a patient without a diagnosis and appropriate staging could lead to rendering a potentially curable patient incurable. Do not perform an ill-conceived open biopsy. Tumors submit ted to open biopsy had a ve to

13

14

Chapter 1

seven times higher rate of local recurrence, independent of diagnosis. Do not create unnecessarily com prom ising m argins, especially in low-grade m alignancies (chordoma, chondrosarcom a), and assume that adjuvant treatments will compensate for violation of oncological principles. In at tempting a “cure,” assum ing this is possible and the patient understands the risks and benet s of such a treatment, an appropriate margin m ust be achieved even if it requires m ajor vascular and neurologic sacri ces. In the decision-

making process, this should never be o ered in cases with m etastases or benign stage 1 and the majorit y of stage 2. The assumption that an en-bloc resection done with an intralesional or m arginal margin along the dura is associated with the same rate of local control (recurrence) as intralesional piecemeal excision (gross total resection) is wrong. A piecemeal approach is always intralesional and always associated with local recurrence unless the speci c tumor is sensitive to adjuvant therapies such as radiation therapy.

Refere nces Five Must-Read Refe rences 1. Fisher CG, Keyn an O, Boyd MC, Dvorak MF. Th e surgical m an agem en t of prim ar y t um ors of th e spin e: init ial result s of an ongoing prospect ive coh or t st udy. Spin e 2005;30:1899–1908 PubMed 2. Ch i JH, Byd on A, Hsieh P, W it h am T, Wolin sky JP, Gokaslan ZL. Ep id em iology an d dem ograp h ics for p r im ar y ver tebral t u m ors. Neu rosu rg Clin N Am 2008;19:1–4 Pu bMed 3. En n eking W F, Span ier SS, Goodm an MA. A system for th e surgical st aging of m u sculoskelet al sarcom a. Clin Or th op Relat Res 1980;153:106–120 PubMed 4. Sim on MA. Surgical m argin s. In : Sim on MA, Springeld D, eds. Surger y for Bon e an d Soft Tissue Tum ors. Ph iladelph ia: Lippincot t-Raven , 1998:77–92 5. Sten er B, Joh n sen OE. Com p lete rem oval of th ree vertebrae for gian t-cell t u m ou r. J Bon e Join t Su rg Br 1971;53:278–287 Pu bMed 6. Yao KC, Borian i S, Gokaslan ZL, Su ndaresan N. En bloc spon dylectom y for sp in al m et ast ases: a review of techniques. Neurosurg Focus 2003;15:E6–E12 PubMed 7. Rh in es LD, Fourn ey DR, Siadat i A, Suk I, Gokaslan ZL. En bloc resect ion of m ult ilevel cer vical ch ordom a w ith C-2 involvem en t . Case repor t an d descript ion of operat ive tech n ique. J Neurosurg Spin e 2005;2:199– 205 PubMed 8. Bailey CS, Fisher CG, Boyd MC, Dvorak MF. En bloc m arginal excision of a m ultilevel cer vical chordom a. Case report . J Neurosu rg Spin e 2006;4:409–414 PubMed 9. Currier BL, Papagelopoulos PJ, Krauss W E, Un n i KK, Yaszem ski MJ. Tot al en bloc spondylectom y of C5 ver tebra for ch ord om a. Spin e 2007;32:E294–E299 PubMed 10. Borian i S, Wein stein JN, Biagin i R. Spin e Update. A su rgical st aging system for th erapeut ic plan n ing of prim ar y bon e t u m ors of th e sp in e. A con t ribu t ion to a com m on term in ology. Spin e 1997;22:1036–1044 PubMed 11. Ch an P, Borian i S, Fourn ey DR, et al. An assessm en t of th e reliabilit y of th e En n eking an d Wein stein -Borian i-

Biagin i classi cat ion s for st aging of prim ar y spin al t um ors by th e Spin e On cology St udy Group. Spin e 2009;34:384–391 PubMed 12. Biagin i R, Casadei R, Borian i S, et al. En bloc ver tebrectom y an d du ral resect ion for ch ordom a: a case report . Spin e 2003;28:E368–E372 PubMed 13. Keyn an O, Fish er CG, Boyd MC, O’Conn ell JX, Dvorak MF. Ligat ion an d p ar t ial excision of th e cau da equ in a as p ar t of a w id e resect ion of ver tebral osteosarcom a: a case repor t and descript ion of surgical tech n ique. Spin e 2005;30:E97–E102 PubMed 14. Murakam i H, Tom it a K, Kaw ah ara N, Oda M, Yah at a T, Yam aguch i T. Com plete segm en t al resect ion of th e sp in e, in clu ding th e sp in al cord, for telangiect at ic osteosarcom a: a report of 2 cases. Spin e (Ph ila Pa 1976) 2006;31:E117–122 15. Dru sch el C, Disch AC, Melch er I, et al. Surgical m an agem en t of recurren t th oracolum bar spin al sarcom a w ith 4-level tot al en bloc spondylectom y: descrip t ion of tech n ique an d repor t of t w o cases. Eur Spine J 2012;21:1–9 Pu bMed 16. Boriani S, Ban diera S, Don th in en i R, et al. Morbidit y of en bloc resect ion s in th e spin e. Eur Spin e J 2010; 19:231–241 Pu bMed 17. Kaw ah ara N, Tom it a K, Baba H, et al. Cadaveric vascu lar an atom y for tot al en bloc spon dylectom y in m align an t vertebral t u m ors. Spine 1996;21:1401–1407 PubMed 18. Kato S, Kaw ah ara N, Tom it a K, Mu rakam i H, Dem ura S, Fujim aki Y. E ect s on spin al cord blood ow an d n eu rologic fu n ct ion secon d ar y to in ter r u pt ion of bilateral segm en t al ar teries w h ich su pply th e ar ter y of Adam kiew icz: an experim en t al st u dy using a dog m odel. Spin e 2008;33:1533–1541 Pu bMed 19. Kaw ah ara N, Tom it a K, Kobayash i T, Abdel-Wan is ME, Mu rakam i H, Akam aru T. In u en ce of acu te sh or tening on th e spinal cord: an experim ent al st udy. Spin e 2005;30:613–620 Pu bMed

2 Safety and E cacy of Surgery for Primary Tumors of the Spine Daryl R. Fourney, Charles G. Fisher, and Stefano Boriani

■ Introduction Th e rarit y of prim ar y spine t um ors and the slow grow t h rate of m any t yp es (e.g., ch ord om a, low -grad e ch on drosarcom a, gian t cell t u m or) m ake th em di cult to st u dy, especially w ith regard to th e resu lt s of su rger y an d oth er t reatm en t s. Th e over w h elm ing m ajorit y of st u dies are retrospective, and the few prospective studies are lim ited by a sh or t follow -u p t im e.1 It is d i cu lt to com pare older st u dies because of loose in terpret at ion of th e term in ology an d a lack of st an dardizat ion in staging an d m an agem en t . Even in in d ivid u al case series, p at ien t m an agem en t often var ies over t im e d u e to advan ces in diagn osis, su rgical tech n iqu e, an d adjuvan t th erap ies. All of th ese factors h ave com bin ed to yield a h eterogen eous group of p ublish ed repor t s,2–9 bu t h as n ot obscu red th e prin cip al fact th at an on cologically valid ap proach to th ese diseases lead s to im proved disease con t rol an d su r vival.10 Th e w ork of Wein stein , Bor ian i, an d Biagin i2,3,11,12 an d ot h ers 5,6 h as sh ow n t h at t h e p r in cip les of m u scu loskelet al on cology origin ally developed by En neking 13,14 can be ap plied to th e spin e. Th e p rin cip les of prim ar y t u m or m an agem en t are ou tlin ed in Table 2.1, an d th ese prin ciples are revisited th rough out th is book, w ith variat ion s in h ow th ey are ad dressed based on th e path ology, th e locat ion an d exten t of disease, an d pat ien t factors (e.g., m edical tness, exposure to previous therapies, an d t reat m en t preferen ce). It is im por tan t to

em p h asize th e exten sive n at u re of en -bloc resect ion su rger y, it s in h eren t m orbid it y, an d it s cost . In fact , th e h igh rate of in t raoperat ive an d early postoperat ive com plicat ion s are often w eigh ed again st th e probabilit y of sur vival an d local con t rol, begging th e qu est ion of w h eth er en -bloc su rger y is just i ed in th e spin e.1,5,7 E cacy is a broad concept that evaluates the overall valu e or usefuln ess of an in ter ven t ion . It includes an an alysis of clin ical outcom es an d resource ut ilizat ion . For prim ar y spin e t u m ors, th e m ain clin ical ou tcom e of in terest is disease cont rol an d, ideally, cu re. Th is ch apter assesses t h e e cacy of en -bloc or on cologically ap p ro p r iate su rger y for p r im ar y sp in e t u m ors to achieve disease con t rol, an d discusses th e potent ial for m orbidit y an d st rategies to p rom ote safet y.

■ Long -Term Results

of Surgery Recen tly, th e rst popu lat ion -based st udy on m alignant prim ar y osseous tum ors of th e spin e th at h ad su cien t p ow er to evalu ate th e role of su rgical resect ion on su r vival w as repor ted by Mu kh erjee et al.15 Th e Su r veillan ce, Epidem iology, an d En d Resu lt s (SEER) regist r y w as an alyzed for cases of h istologically con rm ed p rim ar y (n on m et astat ic) t u m or of th e m obile sp in e or p elvis over a 30-year period (1973– 2003). A m ajor lim it at ion of th is st u dy w as th at

16

Chapter 2 Table 2.1 Principles and Pitfalls that A ect the E cacy of Surgery for Primary Tumors Principle

Pitfall

High index of suspicion

• Diagnosis after intralesional resection with contam inated tumor margins • Biopsy tract not incorporated in later excision • Fine-needle aspiration lim its abilit y to assess tissue architecture • Lack of precision describing surgical approach or margins • Enneking inappropriate surgery associated with higher rates of local recurrence and m ortalit y • All aspects of care not fully appreciated (e.g., pathology, local and system ic staging, surgical feasibilit y, technical points to reduce surgical morbidit y, need for adjuvant and neoadjuvant therapy, soft tissue coverage, patient preferences, etc.) • Delayed diagnosis of recurrence many years after excision of slow-growing tumor t ypes (e.g., chordom a, low-grade chondrosarcoma)

Biopsy before treatment (percutaneous CT-guided core biopsy) Consistent terminology Staging with Enneking and WBB  classi cation Multidisciplinary management/ experienced centers

Long term follow-up

Abbreviations: CT, computed tom ography; WBB, Weinstein–Boriani–Biagini.

th e su rgical m argin s an d th e fu n ct ion al ou tcom e of surger y w ere n ot reported. Th e auth ors perform ed a m u lt ivariate an alysis, an d after adju st ing for age, radiat ion th erapy, an d exten t of local t um or invasion , surgical resect ion w as independently associated w ith signi cantly im proved su r vival for ch ordom a (h azard rat io [HR], 0.617; 95%con den ce in ter val [CI], 0.25– 0.98), ch on drosarcom a (HR, 0.153; 95% CI, 0.07–0.36], osteosarcom a (HR, 0.382; 95% CI, 0.21–0.69), an d Ew ing’s sarcom a (HR, 0.494; 95% CI, 0.26–0.96). Radiat ion th erapy w as associated w ith p rolonged su r vival for pat ien ts w ith osteosarcom a an d ch ordom as, bu t n ot for pat ien ts w ith Ew ing’s sarcom a an d ch on d rosarcom a.

■ Applying Enneking’s

Principles to the Spine A predom in an tly prospect ive coh or t st udy by Fish er et al6 of pat ien t s w ith prim ar y spin e t um ors from fou r spin e cen ters in Can ada t reated bet w een 1982 and 2008 w as an alyzed to determ in e w h eth er ap p lying En n eking’s prin cip les to th e surgical m an agem en t of prim ar y bon e tum ors of the spine signi cantly decreases local recu r ren ce or m or t alit y. Cases w ere d ivid ed

in to t w o grou p s based on su rgical m argin s: En n eking app rop riate (EA) an d En n eking in ap prop riate (EI). Th ere w ere 147 p at ien t s w ith a m edian follow -u p of 4 years (range, 2–7 years). Th ere w as a sign i can t ly h igh er r isk of local recurren ce after EI su rger y, an d th ere w as a st rong correlat ion bet w een th e rst local recurren ce an d m ortalit y w ith an odds rat io of 4.69 (p < 0.0001). EI su rger y w as associated w ith a h igh er risk of m or talit y (HR, 3.10; p = 0.0485) com pared w ith an EA ap proach . Th e m ost sign i can t lim it at ion of th is st u dy w as that th e EI cohort had m ore high-grade t um ors; h ow ever, th e EI grou p also h ad a sign i can tly h igh er rate of adjuvan t th erapy, w h ich sh ou ld bias tow ard th e n ull hypoth esis, th us st rength en ing th e con clusion s.

■ Feasibility of Obtaining

Acceptable Margins In a system at ic review pu blish ed in 2009, Yam azaki et al1 assessed th e early e cacy of en bloc resect ion tech n iques by determ in ing th e n u m ber of w id e/m argin al resect ion s versu s t h e n um ber of in t ralesion al resect ion s. Th e resu lts varied sign i can tly am ong st u dies. All of th e st u d ies in clu ded in th e system at ic review

Safet y and Efficacy of Surgery for Primary Tumors of the Spine rep or ted th e p ostop erat ive m argin s, bu t n ot th e expected m argin s based on th e preop erat ive su rgical staging. Th e on e except ion w as a prosp ect ive st u dy by Fish er et al.5 Am ong th e 26 p at ien t s w it h p r im ar y sp in e t u m ors repor ted by Fish er et al., p ost op erative m argin s w ere w ide in 15, m argin al in four (all plan n ed w ide), an d in t ralesion al in seven (t w o plan n ed w ide, on e plan n ed m argin al). Th u s, Wein stein – Borian i–Biagin i (W BB) staging accu rately pred icted t h e m argin s in 19 of 26 cases (73%). How ever, if t h e gen eral goal of su rger y w as at t ain m en t of a w ide or m argin al en -bloc resect ion (i.e., exclu ding th e fou r cases th at w ere expected to be in t ralesion al), th e W BB staging accu rately p redicted th e at tain m en t of w ide or m argin al en -bloc resect ion in 23 of 26 cases (88%). In exp erien ced h an ds, en -bloc resect ion w ith accept able m argin s is ach ievable if W BB st aging determ in es th at it is feasible.

■ Long -Term Results of

En-Bloc Resection for Chordoma and Chondrosarcoma Borian i et al2 review ed 52 con secut ive ch ordom as of th e m obile spin e over a 50-year period. Th e series inclu ded a ret rospect ive review of 15 cases t reated prior to 1991 an d a p rospect ive group of 37 cases t reated from 1991 to 2002. All p at ien t s w ere st aged an d su bjected to on e of ve p rotocols: (1) radiat ion th erapy w it h or w it h ou t p alliat ive su rger y (n = 10); (2) in t ralesion al ext racap su lar excision (n = 8); (3) in t ralesion al excision plus radiat ion th erapy (n = 16); (4) at tem pted en -bloc resect ion , in t ralesion al or con tam in ated, com bin ed w ith rad iat ion th erapy (n = 8); an d (5) en -bloc resect ion (n = 10). Ever y pat ien t t reated w ith radiat ion th erapy alon e or in t ralesion al excision h ad recurrence w ith in 2 years of surger y. In tralesional extracapsular excision plus radiat ion th erapy also h ad a ver y h igh recu rren ce rate (12 of 16, m ean 30 m on t h s) bu t t h ree p at ien t s w ere con t in u ou sly d isease-free (m ean , 52 m on th s) an d ve p at ien t s w ere alive w ith st able disease (m ean , 69 m on th s) at last fol-

low -u p . Tw elve of 18 pat ien ts w h o received en -bloc resect ion w ere con t in u ou sly diseasefree for an average of 8 years (range, 48–155 m on t h s). For th e six p at ien t s w h o recu r red , it is n ot able th at all of t h em h ad p reviou sly u n d ergon e su rger y or h ad d ocu m en ted con t am in ated t um or m argin s. Th e on ly t reat m en t protocol th at resu lted in greater th an 5 years of con t in u ou sly disease-free su r vival w as en -bloc resect ion w ith m argin al or w ide m argin s. Borian i et al3 also ret rosp ect ively review ed 32 cases of ch on drosarcom a w ith in th e m obile sp in e. Th e average follow -u p period w as 81 m on th s (range, 2–236 m on th s). Recurren ces occu rred in 3 of 14 pat ien t s t reated w ith en bloc resect ion , com p ared w ith 100% of 18 p at ien t s t reated w ith in t ralesion al curet tage. A system at ic review p u blish ed in 2009 by Bor ian i et al4 com p ared t h e e ect s of w id e/ m argin al (en -bloc) resect ion w it h t h e e ect s of in t ralesion al resect ion on local recu r ren ce an d su r vival for ch ordom as an d ch on drosarcom as of th e spin e. Th ey also evaluated th e effect s of radiat ion th erapy. Th e auth ors rep or ted sign i can tly decreased local recurren ce an d d eath for both t um or t ypes after en -bloc resect ion , as com pared w ith in t ralesion al resect ion (p < 0.0001). Th e odds rat io for local recurren ce w ith intralesional resection was 10.25. For chord om a, th e odds rat io for death if local recurren ce occu rred w as 15.03. Radiation therapy is generally recom m ended in th e even t of in t ralesion al or in com p lete resect ion of ch ordom a or ch on drosarcom a, but fairly h igh doses (60 to 65 Gy equivalen ts) are required.4 Su perior dose dist ribu t ion m ay be ach ieved w ith proton beam th erapy or stereotact ic body radiosu rger y (SBRT).16–18 Th e pu b lish ed series u sing th ese t reat m en ts are m ain ly lim ited to sku ll base an d cer vical sp in e lesion s, but w e expect m ore st udies ut ilizing SBRT for m obile spine chordom as and chondrosarcom as in th e com ing years.

■ Morbidity Rate of

En-Bloc Resection In a system atic evaluation of com plications after su rger y for p rim ar y sp in e t u m ors, Yam azaki et

17

18

Chapter 2 al1 foun d t rem en dou s variabilit y in th e reported rate of com plications. Most studies were ret rosp ect ive series an d th ey reported low er rates of com plicat ion s com pared to th e on e p rosp ect ive st udy.5 How ever, th e ret rosp ect ive st u dies h ad greater long-term follow -u p , an d th u s late com plicat ion s, su ch as in st ru m en t at ion failu re an d t u m or recu rren ce, w ere m ore frequ en t ly d escr ibed .2,5,7,19 On ly t w o st u d ies u sed previously publish ed criteria 20 to classify com plicat ion s.7,19 Borian i et al21 ret rospect ively assessed m orbidit y in 134 patien ts w h o un der w en t en -bloc resect ion for spine t u m ors bet w een 1990 an d 2007, in clu ding 90 w ith prim ar y t um ors an d 44 w ith m et astatic t u m ors. Th e rate of com plicat ion s w as 35.1%. Th e risk of m ajor com p licat ion s w as m ore th an t w ice as h igh in pat ien ts w ith con tam in ated t um or m argin s (odds rat io [OR], 2.52; 95% CI, 1.01–6.30; p = 0.048). A sep arately pu blish ed an alysis of th e sam e data sh ow ed a h igh er rate of com plicat ion s in p atients referred from another center after an open biopsy or surger y w h o subsequen tly developed recurren ce (72%) com pared w ith previou sly untreated patients (20%).19 On m ultivariate analysis, th e in depen den t p redictors of m ajor com plications were m ultisegm ental tum or (OR, 1.95; 95%CI, 1.07–3.56; p = 0.03) an d use of a sim ultan eous com bin ed an terior-posterior approach (OR, 3.79; 95% CI, 1.09–13.17; p = 0.036).21 For sacral t u m ors, Fou rn ey et al7 fou n d a correlat ion bet w een h igh er level of sacrectom y and ner ve root sacri ce, w ith an increased com plicat ion rate an d longer inp at ien t stay. Th e in ciden ce of w ou n d com p licat ion s h as been redu ced by th e u se of sp ecialized ap closu re tech n iques, an d th is is perh aps best exem plied after sacrectom y, w h ere soft t issue defect s m ay be par t icularly large.

■ Mortality Rate of

En-Bloc Resection In th e system at ic review by Yam azaki et al,1 m or t alit y rates varied from 0 to 7.7%. Th e m ost com m on cau se of death related to su rger y w as resp irator y failu re. Th e m or talit y rate after en -

bloc resect ion in th e large series repor ted by Borian i et al21 w as 2.2%, occu rring in th ree pat ien t s: on e in traoperat ive death du e to inju r y of th e ven a cava, on e postoperat ive pu lm on ar y em bolism , an d on e late dissect ion of th e w all of th e aor ta 8 m on th s after surger y.

■ Safety Promotion

Strategies Th e m orbidit y of su rgical resect ion for spin e t u m ors derives from several factors, bu t th ere are m any e ect ive preven t ion st rategies (Table 2.2). Preoperat ive angiograph ic em bolizat ion is a w orth w h ile con siderat ion , especially for h igh ly vascularized lesion s such as gian t cell t um or. Su cien t vascu lar access, in clu ding th e p lacem en t of large-bore in t raven ou s cath eters, is n ecessar y in order to adm in ister large volum es of uids an d blood. Cen t ral ven ous pressure m on itoring is recom m en ded for m ost cases of en -bloc resect ion . We h ave gen erally avoided u se of t h e cell saver in p r im ar y t u m ors becau se of th e p oten t ial for t u m or dissem in at ion . Tran exam ic acid app ears to be a safe, e ect ive, low -cost m eth od to redu ce blood loss du ring an d after surger y. After in du ct ion of an esth esia, th e loading dose is 2 g in 100 m L for ad ults an d 30 m g/kg for ch ildren adm in istered over 20 m in u tes, follow ed im m ed iately by a m ain ten an ce dose of 1 g in 100 m L in fu sed at a rate of 100 m g/h for adu lt s an d 1 m g/kg/h for ch ildren .22 Th e length of th ese procedu res, t h e degree of tissue resection, and the postoperative period of convalescence required are all risk factors for deep ven ou s th rom bosis an d p u lm on ar y em bolism . We routinely utilize sequential pneum ocom p ression stockin gs in t h e p er iop erat ive p eriod an d prophylact ic low m olecular w eigh t h eparin postoperat ively. In t raoperat ive com plicat ion s are related to t h e m an ip u lat ion of vit al st r u ct u res. Th e r isk of inju r y in creases in th e case of reoperat ion or p reviou sly ir radiat ion t issu es, d u e to brou s scar an d t issu e fragilit y. If a d u ral tear is en countered, im m ediate sut uring w ith a m uscu-

Safet y and Efficacy of Surgery for Primary Tumors of the Spine Table 2.2 Factors that A ect Morbidity After En-Bloc Resection for Primary Spine Tumors Factor

Safety Promotion Strategy

Complex anatomy due to large tum or growth/extension

• Familiarit y with anatomy • Experience with en-bloc resection of the spine/ paraspinal structures • Preoperative embolization • Tranexamic acid • Careful surgical technique • Careful tissue handling with m icrosurgical techniques • Multimodalit y intraoperative neurophysiological monitoring • Prevent recurrence through careful staging and planning  of  the rst  resection •   Plan  appropriate  soft  tissue  coverage  ( ap  if  necessary) •   Avoid  operating  through  previously  irradiated elds • Circumferential reconstruction to enable a stable fusion

Excessive hemorrhage from the tumor or epidural vein Manipulation  or  sacri ce  of  vascular  or  nervous structures Epidural brosis  due  to  previous  surgery Wound dehiscence or infection Long-term hardware failure

lar coverage is recom m en ded. If a w atert igh t closu re can n ot be obt ain ed, cerebrospin al u id (CSF) d rain age is an opt ion bu t car r ies it s ow n r isks (e.g., overd rain age, p n eu m ocep h alu s, m en ingit is). Man ipulat ion of th e spin al cord, especially in th e th oracic sp in e, sh ou ld be avoided. Carefu l m icrosu rgical tech n iqu e is requ ired w h en h an dling n er vou s t issue. We recom m en d th e use of m ultim odalit y intraoperative neurophysiological m on itor in g (som atosen sor y evoked p oten t ials, m otor evoked p oten t ials, elect rom yograp hy). Late in st rum en tat ion failure h as been repor ted in abou t 7%of cases.19 En -bloc resect ion en tails sign i can t in stabilit y, and given th e goal of long-term disease-free sur vival, circum feren t ial recon st r u ct ion an d long-segm en t st abilizat ion is recom m en ded. Most im p or tan tly, m et icu lou s plan n ing an d th ough t m u st go in to biological xat ion . With long-term sur vival an t icipated, fusion m ust be ach ieved, an d th is can be di cu lt in a biologically h ost ile environ m en t . Vascu larized st r u t graft s an d augm en ted soft t issu e coverage are often n ecessar y in large resect ion s. Fin ally, t h e ben e t of exp er ien ce can n ot be u n derst ated. Th e su rgical an atom y is often com plex due to t um or grow th adjacen t to th e sp in al cord an d crit ical parasp in al st ru ct u res. Th e en t ire t reat m en t , from biopsy to resect ion to long-term follow -u p, sh ould ideally be per-

form ed at th e sam e cen ter, un der th e gu idan ce of a m u lt id isciplin ar y team led by an exp erien ced spin e on cology su rgeon .

■ Avoiding Problems Open Biopsy Poorly plan n ed biopsies in crease th e risk of local recu rren ce by t u m or dissem in at ion along fascial plan es an d th e biopsy t ract .8 In a series of pat ien t s w ith ch ordom a of th e m obile spin e an d pelvis, Bergh et al9 fou n d th at p erform an ce of an invasive diagn ostic procedure out side th e in dex cen ter an d in adequate surgical m argin s w ere associated w ith t u m or recu rren ce an d t u m or-related death . In p at ien t s w ith prim ar y t u m ors of th e sacr u m , Fou rn ey et al7 n oted a 78% con t in u ou sly disease-free su r vival for p at ien t s un dergoing both biopsy an d surger y at th e in dex in st it ut ion , com pared w ith 55% for th ose pat ien t s w h o un der w en t a prior proced ure at an oth er in st it u t ion . In cision al biop sy or in t ralesion al resect ion sign i can tly in creases th e risk of local recurren ce; t h erefore, t ran scu t an eou s com p u ted tom ography–gu ided t rocar biop sy is recom m en d ed .1 W h en t h ere is a su sp icion of a p r im ar y t u m or, t h e su rgeon w h o p er for m s t h e de n it ive su rger y sh ou ld id eally be t h e on e to p er for m or d irect t h e biop sy p roced u re.

19

20

Chapter 2 Per m an en t m arking of th e biopsy site is recom m en d ed so t h at it can be id en t i ed an d in clu ded in th e excision . Alth ough n e-n eedle aspiration provides cytom orphological features th at m ay yield a diagn osis, a t rocar n eedle m ay im prove accu racy by an alysis of all h istological feat u res of th e t issu e.1,23,24

Inconsistent Terminology and Staging Historically, th e assessm en t an d m an agem en t of prim ar y spin e t u m ors h ave follow ed gen eral prin cip les of n eu rosu rger y an d or th op edics; h ow ever, th e reluct an ce of spin e surgeon s to adopt En n eking’s prin ciples of m usculoskeletal on cology h as largely been overcom e in recen t years. Neverth eless, it is st ill com m on to receive con su lt at ion s regarding con t am in ated m argin s after a poorly conceived excisional biopsy or intralesion al resect ion . It is im p or t an t to em p h asize th e correct u se of term in ology an d staging m eth od s in h eren t to th e p lan n ing an d execu t ion of on cologically sou n d p rocedures.11,14

Underestimating the Technical Challenges and Functional Sacri ces of Surgery Th e m an agem en t of p rim ar y t u m ors of th e spin e is ch allenging. Many are slow grow ing an d m ay be exten sive by th e t im e of presen t at ion (e.g., ch ordom a, ch on drosarcom a), involve com p lex an atom y, an d on ly par t ially resp on d to adjuvant therapies.2,3,5–8,19,21 Local recurrence of ch ordom a in par t icu lar is di cu lt to p alliate as it often leads to m u lt iple aggressive recu rren ces, debilitat ion , an d even tual d eath .2,8,10,19 W h en staging determ in es th at en -bloc excision is th e proced u re of ch oice, su rgical plan n ing n eeds to t ake in to accoun t n ot on ly th e fu n ct ion al sacri ces n ecessar y to ach ieve th e on cological goals, bu t also th e in h eren t m orbidit y of surger y an d even t ual h ealth -related qu alit y of life (HRQOL). On ly on e st u dy h as addressed th is issu e an d fou n d th at at a m ean follow -u p of 3½ years after en -bloc resect ion for p r im ar y bon e t u m ors th e Sh or t Form (SF)-

36 m en t al com p on en t score w as 51 (essen t ially n orm al) an d t h e p hysical com p on en t sum m ar y 38.5

■ Chapter Summary Th e r isk of com p licat ion s an d recu r ren ce is h igh est after revision su rger y. Th erefore, t h e rst t reat m en t is a m ajor determ in an t of ou tcom e.21 W h en th ere is a su spicion of prim ar y sp in e t um or, referral to an exp erien ced cen ter is recom m en ded, becau se th e su rgeon w h o p erform s th e de n it ive excision sh ou ld ideally perform (or at least direct) the biopsy.1 This en sures th at unrelated function al tissue is not con tam inated w ith tum or. Contam ination w ould n ecessit ate th e t issue’s in clusion in subsequ en t en -bloc resect ion —som eth ing th at w ou ld n ot occu r w ith prop erly plan n ed biop sy. Local an d system ic st aging u sing Enn eking’s approach identi es patients w ho m ay be cured, or at least h ave a bet ter ch an ce at long-term disease con t rol an d red u ced m or t alit y w it h en -bloc resect ion . On cologically sou n d m argin s are likely to be ach ieved w h en W BB staging d eterm in es th at it is feasible.1 How ever, th e adverse-even t p ro le of th ese surgeries is ver y h igh , even at experien ced cen ters. Th erefore, w e recom m en d th at on ly exp erien ced, m u lt id isciplin ar y team s perform th em . Th e n eed for careful pat ien t select ion in clu ding detailed p reoperat ive cou n seling can n ot be overst ated.

Pearls Enneking’s principles of oncological staging are valid with respect to prim ary spine tumors. Adequate margins are feasible depending on the surgical staging (WBB system ). En-bloc resection has a high rate of complications and therefore should only be performed at experienced centers. Multidisciplinary tumor boards are necessary to plan and carry out appropriate treatment. Biopsy is ideally planned or perform ed by the surgical oncologist who will ultim ately perform the en-bloc resection so that the biopsy tract can be incorporated in the resected specimen.

Safet y and Efficacy of Surgery for Primary Tumors of the Spine Pitfalls A high index of suspicion for primary tumors is necessary to prevent incisional biopsy or intralesional resection that could make oncologically sound m argins impossible to achieve. Inconsistent term inology (e.g., “gross total resection,” “radical resection”)

Inadequate staging Inadequate surgical planning (e.g., need for soft tissue coverage) Underestimating the technical challenges and functional consequences of surgery

Refere nces Five Must-Read Refe rences 1. Yam azaki T, McLough lin GS, Patel S, Rh in es LD, Fourn ey DR. Feasibilit y an d safet y of en bloc resect ion for p r im ar y sp in e t u m ors: a system at ic review by t h e Spin e On cology St udy Group. Spine 2009;34(22, Su ppl):S31–S38 PubMed 2. Borian i S, Ban diera S, Biagin i R, et al. Chordom a of th e m obile spin e: ft y years of experien ce. Spin e 2006; 31:493–503 PubMed 3. Borian i S, De Iure F, Ban diera S, et al. Ch on drosarcom a of th e m obile spin e: repor t on 22 cases. Spine 2000;25:804–812 PubMed 4. Borian i S, Saravanja D, Yam ada Y, Varga PP, Biagin i R, Fish er CG. Ch allenges of local recu r ren ce an d cu re in low grade m align ant t um ors of the spin e. Spin e 2009;34(22, Su pp l):S48–S57 Pu bMed 5. Fisher CG, Keyn an O, Boyd MC, Dvorak MF. Th e su rgical m anagem en t of prim ar y t um ors of th e spin e: in it ial result s of an ongoing prospect ive coh or t st udy. Spin e 2005;30:1899–1908 PubMed 6. Fisher CG, Saravanja DD, Dvorak MF, et al. Surgical m an agem en t of prim ar y bon e t u m ors of th e spin e: validation of an approach to enhance cure and reduce local recurrence. Spine 2011;36:830–836 PubMed 7. Fourn ey DR, Rh in es LD, Hen t sch el SJ, et al. En bloc resect ion of prim ar y sacral t um ors: classi cat ion of su rgical approach es an d outcom e. J Neurosurg Spine 2005;3:111–122 PubMed 8. Su ndaresan N, Rosen G, Borian i S. Prim ar y m align an t t um ors of th e spin e. Or th op Clin Nor th Am 2009; 40:21–36, v v. PubMed 9. Bergh P, Kin dblom LG, Gu n terberg B, Rem ot t i F, Ryd W, Meis-Kin dblom JM. Progn ost ic factors in ch ordom a of th e sacr u m an d m obile spin e: a st u dy of 39 pat ien t s. Can cer 2000;88:2122–2134 Pu bMed 10. Fourn ey DR, Gokaslan ZL. Cu rren t m an agem en t of sacral chordom a. Neurosurg Focus 2003;15:E9 PubMed 11. Borian i S, Wein stein JN, Biagin i R. Prim ar y bon e t um ors of th e spin e. Term in ology an d surgical st aging. Spin e 1997;22:1036–1044 Pu bMed 12. Ch an P, Borian i S, Fourn ey DR, et al. An assessm en t of th e reliabilit y of the En n eking an d Wein stein Borian i-Biagin i classi cat ion s for st aging of prim ar y

spin al t um ors by th e Spin e On cology St udy Group. Spin e 2009;34:384–391 PubMed 13. En n eking W F, Span ier SS, Goodm an MA. A system for th e su rgical st aging of m u scu loskelet al sarcom a. Clin Or th op Relat Res 1980;153:106–120 Pu bMed 14. Enneking W F. A system of staging m usculoskeletal neoplasm s. Clin Orthop Relat Res 1986;204:9–24 PubMed 15. Mukh erjee D, Ch aich an a KL, Parker SL, Gokaslan ZL, McGir t MJ. Associat ion of su rgical resect ion an d su rvival in p at ien t s w ith m align an t prim ar y osseou s sp in al n eoplasm s from th e Su r veillan ce, Ep idem iology, an d En d Result s (SEER) dat abase. Eur Spin e J 2013;22:1375–1382 Pu bMed 16. Gw ak HS, Yoo HJ, Youn SM, et al. Hypofract ion ated stereot act ic rad iat ion t h erapy for sku ll base an d upper cer vical ch ordom a an d ch on drosarcom a: prelim in ar y resu lt s. Stereot act Fu n ct Neu rosu rg 2005; 83:233–243 Pu bMed 17. Hen derson FC, McCool K, Seigle J, Jean W, Har ter W, Gagn on GJ. Treat m en t of ch ordom as w ith CyberKn ife: Georgetow n Un iversit y experience an d t reat m en t recom m en dat ion s. Neurosurger y 2009;64(2, Suppl):A44–A53 Pu bMed 18. Park L, Delan ey TF, Liebsch NJ, et al. Sacral ch ordom as: im pact of h igh -dose proton /ph oton -beam radiat ion th erapy com bin ed w ith or w ith ou t su rger y for prim ar y versu s recurren t t um or. In t J Radiat On col Biol Phys 2006;65:1514–1521 Pu bMed 19. Ban diera S, Borian i S, Don th inen i R, Am en dola L, Cappuccio M, Gasbarrin i A. Com plicat ion s of en bloc resect ion s in th e spin e. Or th op Clin Nor th Am 2009;40:125–131, vii vii. Pu bMed 20. McDon nell MF, Glassm an SD, Dim ar JR II, Pun o RM, Joh n son JR. Periop erat ive com p licat ion s of an terior procedu res on th e spin e. J Bon e Join t Surg Am 1996; 78:839–847 PubMed 21. Boriani S, Ban diera S, Don th in en i R, et al. Morbidit y of en bloc resect ion s in th e spin e. Eur Spin e J 2010; 19:231–241 Pu bMed 22. Elw at idy S, Jam joom Z, Elgam al E, Zakaria A, Tu rkist an i A, El-Daw latly A. E cacy an d safet y of prophylact ic large dose of t ran exam ic acid in spin e su rger y:

21

22

Chapter 2 a p rospect ive, ran dom ized , d ou ble-blin d, p lacebocon t rolled st udy. Spin e 2008;33:2577–2580 PubMed 23. Lis E, Bilsky MH, Pisin ski L, et al. Percut an eous CTgu ided biopsy of osseous lesion of th e spin e in pat ien t s w ith kn ow n or suspected m alignan cy. AJNR Am J Neu rorad iol 2004;25:1583–1588 PubMed

24. Pierot L, Bou lin A. Percu t an eou s biop sy of t h e t h oracic an d lu m bar sp in e: t ran sp ed icu lar ap p roach un der uoroscopic gu idan ce. AJNR Am J Neuroradiol 1999;20:23–25 Pu bMed

3 Interventional Options for Primary Tumors of the Spine Sudhir Kathuria

■ Introduction Minim ally invasive im age-guided inter vent ions h ave u n dergon e a rapid evolut ion in th e last t w o decades. Most of th is evolu t ion h as been tech n ologically driven an d h as in cluded advan ces in real-t im e use of h igh -resolut ion im aging du ring in ter ven t ion s; n ew an d im p roved em bolic agen t s, bon e cem en t s, an d cath eters; an d gu idew ire tech n ology. Percu t an eou s tech n iques are being in creasingly ut ilized in both diagn ostic and therapeutic interventions of prim ar y sp in e t u m ors. Th is ch apter discu sses th e in ter ven t ion al tech n iqu es of sp in e biop sy, vertebral augm en tat ion , an d preoperat ive em bolizat ion , w ith an em ph asis on h ow th ey apply to prim ar y t u m ors of th e sp in e.

■ Image -Guided

Spine Biopsy Th e biopsy represen t s th e n al p h ase of th e local w orku p for a sp in al lesion . If p r im ar y sp in al t um or is in th e di eren t ial diagn osis, then the physician doing th e biopsy should discuss the plan w ith the spine surgeon coordinating th e pat ien t care to en sure th at th e biopsy t ract w ill n ot com prom ise de n it ive m an agem en t . Ideally, th e biop sy sh ou ld be perform ed

in th e sam e cen ter th at w ill provide th e de n itive treatm en t, and should involve consultation w ith th e m u sculoskeletal path ologist to con sid er any sp ecial p recau t ion s. Percu t an eou s im age-guid ed spin e biopsy is n ow com m on ly perform ed to diagn ose a su sp ected n eoplast ic process. Com pared w ith open su rgical biop sy, it is less invasive an d m ore cost-e ect ive, w ith an overall low er risk of com plicat ion s.1 Determ in ing w h eth er a lesion is ben ign or m align an t an d it s speci c h istological t ype an d grade is vit al for subsequ en t m an agem en t an d t reatm en t p lan n ing. Th e d ecision to p er for m a sp in e biop sy sh ou ld be m ade after a th orough an alysis of risks an d ben e t s. Th e gen eral risks of p ercu t an eou s spin e biopsy in clude bleeding, injur y to vascu lar or n eural st ru ct ures, an d in fection . In cer t ain t um ors, such as ch ordom a, disease can spread along th e n eedle t ract due to t um or im plan t at ion . In creasing u se of im age gu idan ce and coaxial needle technique has decreased the in ciden ce of th ese com p licat ion s. Certain com plicat ion s are sp eci c to th e an atom ic st ru ct u res th at are in p roxim it y or along th e p ath of n eed le p lacem en t du ring th e biopsy procedu re. Pn eu m oth orax can occu r du ring biopsy of th oracic spin e or lesion s adjacen t to the lung or pleu ra. St roke du e to inju r y to th e carot id or ver tebral ar ter y is a con cern w h en p erform ing cer vical spin e biopsy.2

24

Chapter 3

■ Patient Preparation Most spin e biop sies can be p erform ed on an ou t p at ien t basis u sin g local an est h esia an d m od erate sedat ion . Gen eral an est h esia m ay be n ecessar y in certain sit uat ion s, su ch as cervical spine biopsy w here patient m otion is risky, t ran soral C2 lesion biop sy w h ere in t u bat ion is requ ired to avoid asp irat ion , an d som e u n derlying m edical con dit ion s. Th e pat ien t is gen erally posit ion ed pron e for th oracic or lum bosacral spin e biopsy an d su pin e for cer vical spin e biopsy, except for certain lesion s located w ith in th e posterior elem en t s for w h ich th e pron e posit ion m ay be m ore ap prop riate. In cer t ain in st an ces, for exam ple, w h en a pat ien t h as a colostom y or gast rostom y t u be an d is u n able to lie com p letely pron e, biop sy can be p erform ed in th e lateral decu bit u s or sem i-pron e posit ion . St rict asep t ic tech n iqu e sh ou ld be u sed d u ring th is p rocedu re to m in im ize any iat rogen ic in fect ion s.

n eedle locat ion in relat ion to th e target lesion an d im p or t an t n or m al st r u ct u res in t h e vicin it y. Th is red u ces t h e risk an d sign i can t ly sh or ten s th e procedure t im e. CT u oroscopy is ext rem ely h elpfu l in ap proach ing lesion s in an atom ically di cult areas.3 Magn et ic reson an ce (MR)-gu ided biop sy procedu res can be safely p erform ed in clin ical pract ice. Alth ough CT u oroscopy cu rren tly is th e m odalit y of ch oice, MR gu idan ce provid es distin ct advan t ages in cer t ain sit u at ion s. Most cyst ic lesion s an d cer tain bony lesion s su ch as h em angiom as are bet ter detected on MRI. A m ajor advan t age of an MR-gu ided p rocedu re is th e lack of radiat ion exp osu re to both p at ien t s an d physician s, w h ich is especially im por tan t in children and pregnant wom en. Current m ajor lim itat ion s of th is m odalit y are longer p rocedu re t im e, h igh er cost , an d lim ited availabilit y of m ach in e an d in st ru m en t s. How ever, developm en t of n ew MRI gu idan ce m eth ods an d devices is rapid ly evolving.

■ Image Guidance

■ Biopsy Needle Systems

Th e im aging m odalit ies available for gu idan ce in clu de X-ray uoroscopy, com p u ted tom ography (CT) uoroscopy, m agn et ic reson an ce im aging (MRI); th ey can be u sed singly or in com bin at ion . Th e ch oice of m odalit y is determ ined by its availabilit y, the size and location of th e suspected lesion , an d operator preferen ce. X-ray uoroscopy h as the advantage of being able to visu alize an d n avigate t h e n eed le in real t im e using a t w o-dim en sion al project ion form at . Many sim ple biopsy procedures can be safely perform ed using th is system . How ever, it lacks t h e sp at ial an d soft t issu e con t rast resolu t ion . I p refer CT u oroscopy in m ost cases as it provides excellen t spat ial an d con t rast resolu t ion for p recise lesion localizat ion an d iden t i cation of im portant inter vening st ruct ures. This en ables th e op erator to select a safe biop sy n eedle t rajector y to access th e target lesion . CT uoroscopy provides a prom pt display of con t in u ou sly u p dated im ages th at sh ow th e exact

Th ere are several com m ercial biop sy n eed le system s available. Th e select ion d ep en d s on lesion locat ion an d w h eth er th e lesion is soft t issu e, cyst ic, or osseou s in n at u re. Asp irat ion biopsy of a cyst ic or soft t issu e lesion can be perform ed w ith 25-, 22-, or 20-gauge n eedles sp ecially design ed to aspirate cellular m aterial. Core biopsies are perform ed using soft t issue cut t ing n eedles (16- or 18-gauge) or t reph in e an d beveled t ip bon e biopsy n eedles. We recom m en d u sing a coaxial n eedle tech n iqu e w h ere on ly a sin gle gu id in g n eed le/ can nula pass is m ade from the skin to the lesion. This technique requires only a single biopsy tract w ith in th e soft t issue, th us reducing procedure t im e an d risk of soft t issu e inju r y associated w ith additional passes. The guiding needle provides access for several biopsy n eedle passes in side th e lesion . Th is tech n iqu e also red u ces th e risk of in lt rat ion of n orm al t issu es w ith poten t ial t um or cells from th e biopsy n eedle.2

and Technique

Interventional Options for Primary Tumors of the Spine

■ Biopsy Approaches for

Spine Tumors Th e ap propriate biop sy ap proach dep en d s on lesion size, locat ion , an d in ter ven ing an atom ic st r u ct u res. Th orough kn ow ledge of th e an atom y is essen t ial for plan n ing a safe route of access for n eedle placem en t . In th e cer vical spin e, th e crit ical an atom ic st r u ct u res in clu de th e vertebral and carotid arter y, jugular vein, esoph agu s, t rach ea, ph ar yn x an d hyp op h ar yn x, thyroid glan d, segm en t al n er ves, an d sp in al cord. In th e th oracic spin e an atom ic kn ow ledge of th e pleu ra, lu ng, aor t a, radicu lar ar teries, an d n eurologic st r uct u res is crit ical. In th e lum bar sp in e, t h e con u s an d exit in g root s, kid n eys, bow el, an d m ajor blood vessels m u st all be con sidered.4

Posterior Approach A p oster ior ap p roach is gen erally u sed for sacral, lu m bar, t h oracic, an d p oster ior cer vical lesion s. Th e p at ien t is p laced pron e or in th e lateral decu bit u s p osit ion . Th e biop sy n eed le can en ter t h e ver tebral body via a t ran sp edicular (Fig. 3.1) or p arap edicu lar ap p roach . Gen erally t h e t reat ing su rgeon w ou ld p refer

25

th e biopsy t ract to be as close to th e m id lin e as possible an d to be m arked w ith a su t u re or dye.

Anterior Approach Du e to sm aller p ed icle size, an an ter ior ap p roach is often u sed for m ost cer vical sp in e lesion s. Th e n eed le is in ser ted adjacen t to t h e m edial border of the sternocleidom astoid m uscle an d advan ced bet w een th e carot id sh eath an d n eck air w ays (Fig. 3.2). Sp ecial at ten t ion sh ou ld be p aid to avoid t h e esop h agu s an d hyp oph ar yn x. Th is procedure sh ould be don e w ith CT uoroscopy guidan ce, an d in term itten t CT im ages sh ould be obtain ed to ch eck th e n eedle trajector y.

Transoral Approach Th e t ran soral app roach can be ver y u sefu l for accessing C2 ver tebral body lesion s. It can also be u sed to access lesion s at adjacen t levels in clu ding C1 an d C3. Th is ap proach requ ires gen eral an esth esia, as in t u bat ion is n ecessar y to avoid any asp irat ion (Fig. 3.3). Prop hylact ic an t ibiot ics sh ould be used w ith th is approach , as it is d i cu lt to create a ster ile eld alon g the posterior phar yngeal w all though w hich th e biopsy n eedle is in serted. Th is is a relat ively

a

b

Fig. 3.1a,b Posterior approach. (a) Computed tomography (CT) scan of a patient in the prone position shows a well-def ned lytic lesion (arrow) in

the T1 vertebral body. (b) CT- uoroscopy–guided transpedicular biopsy needle placement from the posterior approach inside the lesion.

26

Chapter 3

a

b

Fig. 3.2a,b Anterolateral approach. (a) Initial planning CT scan of a patient with an abnorm al C5 lesion with a prevertebral component. The m easurem ent bar shows the width of vascular compartment at the level of the planned needle path. (b) CT- uoroscopy im age shows a biopsy

needle inserted through the medial part of the sternocleidomastoid m uscle and advanced bet ween the airway and the carotid sheath inside the lesion. Note the use of the measurement bar to avoid needle injury to the carotid artery.

a

b

Fig. 3.3a,b Transoral approach. (a) Close-up view of the intubation tube and transoral biopsy needle inserted through an open mouth. Note the gauze roll

used to keep the m outh open. (b) CT- uoroscopy image shows the biopsy needle tip inside the lesion.

Interventional Options for Primary Tumors of the Spine safe approach , as n o im portan t st ruct ures lie bet w een th e posterior ph ar yngeal w all an d th e bon e.5

follow ed by injection of the bon e cem ent in th e cavit y.

Patient Selection Postprocedure Care Direct com pression is applied at th e pun ct ure site for abou t 5 to 10 m in u tes to ach ieve h em ost asis. Th is is follow ed by placem en t of sterile dressing on th e p u n ct u re site an d adjacen t skin . Th e pat ien t sh ould be kept in th e recover y area for 1 to 2 h ou rs to w atch for any bleeding; to assess pain , w eakn ess, an d dizzin ess; an d to m on itor th e gen eral vit al sign s. Th e p at ien t is disch arged w ith a m edicat ion p rescript ion an d a con tact n u m ber to reach a physician if any com p licat ion s arise.

■ Vertebral Augmentation

for Spine Tumors and Pathological Fractures Ver tebral augm en tat ion is a m in im ally invasive procedu re prim arily u t ilized for t reat ing back pain associated w ith a ver tebral com pression fract ures. Th is procedu re w as rst in t roduced as a tech n ique for th e t reat m en t of sym ptom at ic ver tebral h em angiom a in 1987.6 Sin ce th at t im e, it h as becom e a valu able an d frequ en tly used th erap eut ic opt ion in th e m an agem en t of back pain caused by osteoporot ic an d path ological fract u res. Th e fu n dam en t al goal of a ver tebral augm en t at ion p roced u re is to st abilize an d im prove th e com p ressive st rength of th e ver tebral body th rough th e safe inject ion of stabilizing m aterial. This can be achieved by both vertebroplast y and kyphoplast y. Vertebroplast y involves th e inject ion of sp ecial bon e cem en t in side th e fract ured ver tebra u sing a n eedle under im age gu idan ce, gen erally X-ray or CT- u oroscopy. Th ere are several bon e cem en t s available, bu t th e m ost com m on ly u sed is acr ylic polym er polym ethylm ethacrylate (PMMA). Kyphoplast y, in com p arison w ith ver tebroplast y, involves an addit ion al step of creat ing a cavit y in side th e diseased ver tebral body by in at ing a balloon

Malign an t lesion s can sp read to th e sp in al colu m n by h em atogen ous, lym ph at ic, perin eu ral, or direct exten sion .7 Pain arises from cort ical erosion , fract u re, or im pingem en t on th e n er ve roots or spinal cord. There are various treatm ent opt ion s th at depen d on th e un derlying n at ure of t h e d isease an d in clu d e su rger y, m ed ical th erapy, radiat ion th erapy, or a com bin at ion of th em . Ver tebral augm en t at ion is n ow p laying an in creasing role in th is pat ien t p opu lat ion . Th e decision regarding th erapy sh ou ld be m ade by a m u lt idiscip lin ar y team t aking in to con siderat ion th e n at u re an d exten t of th e disease, th e m edical con dit ion of th e pat ien t , th e respon se to prior th erapy, an d th e p at ien t’s life expectancy. The goal of vertebral augm entation in p ath ological fract ures is to im prove pain an d st r u ct u ral st abilit y of th e bon e. Radiat ion th erapy is e ect ive in m any cases, but if th ere is spin al in st abilit y presen t , pain relief w ill n ot occur w ith radiation alone as there is no im m ediate bone strengthening.8 Vertebral augm entat ion can be a good th erapeut ic opt ion by itself or can be com bin ed w ith radiat ion an d surgical th erapy. Absolute contraindicat ions for vertebral augm en t at ion in clu d e ongoing local or system ic in fect ion , com p ressive m yelop at hy secon d ar y to ret rop u lsed bon e fragm en t or t u m or, an d any un correctable coagu lopathy. Relat ive con t rain dicat ion s un ique to path ological fract u res d epen d on th e operator expert ise an d th e tools available. Severely com pressed ver tebral bodies w ith epidural t um or exten sion are tech n ically di cu lt to t reat an d m ay be m ade w orse w ith in ject ion of PMMA. Som e of th ese cases can be t reated by an experien ced operator u sing a plasm a-m ediated tum or ablation system to rst create a cavit y w ith in th e t u m or, follow ed by ju diciou s PMMA cem en t inject ion .9 Th e use of CT- u oroscopy w ith excellen t sp at ial an d con trast resolution can be ver y useful in such cases, as th ey requ ire p recise n eedle placem en t an d close m on itoring of PMMA cem en t inject ion .

27

28

Chapter 3

Preprocedural Imaging Cross-sect ion al im agin g, in clu d in g MRI an d CT, is ext rem ely useful for iden t ifying th e locat ion an d exten t of th e disease. It also p rovides u sefu l in form at ion abou t any can al or n eu ral com p rom ise. CT in par t icu lar is ver y h elp fu l in dem on st rat ing osseou s dest r u ct ion th at can be a poten t ial route of cem en t ext ravasat ion an d for assessing t ran sp edicu lar access.

Tools, Technique, and Special Considerations A variety of disposable vertebroplast y and kyph oplast y n eedles are available. Needles range in sizes from 8 to 13 gauge. Sm aller size 15-gauge n eedles are also available an d m ore suitable for cer vical spin e. Several PMMA-based stabilizing agen ts w ith di eren t ch aracterist ics are com m ercially available th at can be u sed for ver tebral augm en t at ion . Th e basic prin ciples an d tech n iques of safe needle placem ent described earlier in the spine biopsy sect ion also apply to ver tebral augm en tat ion . A posterolateral approach is gen erally preferred for lum bar and th oracic spine lesion s. Special at tent ion should be paid to avoid breaking m edial an d in ferior borders of th e pedicles. Breach of th ese cort ical w alls can resu lt in n eedle en t r y in to th e spin al can al or n eu ral foram en , w ith p oten t ial inju r y to th e sp in al cord or to n er ves. For cer vical sp in e lesion s, an terolateral approach is gen erally m ore suitable. For lesion s w ith in th e C2 ver tebral body, a t ran soral rou te m ay be m ore favorable. Th e n eedle is advan ced using con t in uous or in term it ten t u oroscopy u n t il th e t ip is p laced in desired locat ion , gen erally in th e an terior par t of th e t u m or. Most inject ion devices th at are available are self-con tain ed system s, w ith a reser voir in to w h ich th e PMMA cem en t is lled. Th is device is t h en con n ected w ith th e n eedle an d th e PMMA can be injected in a con t rolled fash ion by rotat ion or t rigger m ech an ism (Fig. 3.4). PMMA sh ou ld be injected in sm all bolu ses w ith close at ten t ion to any leakage beyon d th e con n es of th e ver tebral body, especially into pulm onar y veins, epidural veins,

Fig. 3.4 CT uoroscopy image obtained during vertebroplast y treatment for a painful hem angioma. Note the white dense PMMA cement material (arrow) inside the lesion.

the cent ral canal, or th e neural foram en. A sm all leak th rough en d -plate fract u res m ay be accept able, but a large am oun t of leakage w ith in th e disk sp ace m ay in crease th e fu t u re risk of adjacen t fract ure.10 If th e PMMA preferen t ially ow s to u n d esirable areas, th e n eedle can be reposit ion ed, an d the m aterial already injected is allow ed to h ard en . A coa xial tech n iqu e is ver y h elpful in such in stan ces, as th e cem en tinject ing n eedle can be t aken ou t w h ile keep ing th e access n eedle in place. Th is allow s th e cem en t already injected in th e body to h arden faster, as it is exposed to a body tem perat u re th at is h igh er th an room tem perat ure. Inject ion is th en resum ed in few m in utes, w ith exp ect at ion of PMMA cem en t n ow lling in th e d esirable area. If the PMMA con t in u es to ow in u n desirable areas, fu rth er inject ion sh ou ld be stopp ed . An oth er u sefu l tool t h at can allow safe in ject ion of PMMA is plasm a-m ediated ablat ion . Th is can be esp ecially h elp fu l in t reat ing lesion s w it h p oster ior cor t ical d isr u pt ion an d large epidu ral exten sion s. Such lesion s carr y a h igh er risk of poten t ial ext raosseou s exten sion of PMMA cem en t. The bipolar plasm a-m ediated radiofrequen cy-based device can be advan ced th rough th e ver tebroplast y n eedle an d placed in side th e lesion . A p lasm a eld is th en created th at h as su cien t en ergy to dissolve soft t issue

Interventional Options for Primary Tumors of the Spine in to m olecu les at relat ively low tem p erat u re (40° to 70°C). Th is m olecu lar d issociat ion con verts the ablated tissue into gases that leave th e body via th e can n ula.11 Using th is process, a cavit y can be created inside th e desired port ion of th e lesion w ith preferen t ial ow of sub sequ en tly injected PMMA m aterial in to th is void du e to low resistan ce. Post procedure CT im aging sh ould be rout in ely obtain ed in path ological fract ures. Th is p rovid es u sefu l baselin e in for m at ion abou t t h e dist ribu t ion of PMMA an d dem on st rates any su sp ected or u n su sp ected com p licat ion s in clu d in g cem en t leakage, ch an ges in t u m or p osit ion , p roced u re-related h em atom a, an d fract u re.

Complications Th e com plicat ion rates in variou s m ajor series are 1.3% for osteoporosis, 2.5% for h em angiom as, an d 10% for n eop last ic disease.12 Th e p rim ar y cau se of a sym ptom at ic com p licat ion is th e leakage of cem en t beyon d th e con n es of ver tebral body. Th e cem en t can leak in to epidu ral an d p araver tebral ven ou s p lexu s or in adjacen t sp aces via exist in g fract u re lin es or cor t ical dest r u ct ion . More often th an n ot , sm all PMMA leakages are asym ptom at ic. How ever, PMMA m ater ial located w it h in t h e ep id u ral vein s or foram in al vein s can cau se sp in al cord or n er ve root com pression . A sign i cant n eurologic com p rom ise m ay requ ire im m ediate su rgical in ter ven t ion . Migrat ion of sm all am oun ts of PMMA to t h e p u lm on ar y vein s is gen erally n ot clin ically sign i can t , bu t sym ptom at ic p u lm on ar y em bolus an d death h ave been repor ted .12 Leakage w ith in th e disk sp ace can in crease th e fut u re risk of adjacen t level fract u res. Oth er com p licat ion s in clu de vascu lar or neural injuries from im proper n eedle tech nique, iat rogen ic fract u re, in fect ion , h em atom a, an d pn eu m oth ora x.10 Com plication s are best avoided by aw areness of th e factors th at con t ribute to th eir occurrence. Follow ing m et icu lou s n eedle p lacem en t tech n ique, close m on itoring, an d avoiding excessive am oun ts of PMMA inject ion especially in pathological fract ures are strongly suggested.

■ Preoperative and

Therapeutic Embolization Su rger y of hyp er vascu lar sp in e t u m ors can be associated w ith sign i can t blood loss an d in creased t ran sfu sion requ irem en ts. Excessive bleed ing d u ring su rger y also p rolon gs op erat ive t im e an d in creases com p lexit y. Sp in al t u m or em bolizat ion is p erform ed before th e su rger y to redu ce in t raoperat ive bleeding. By redu cing in t raop erat ive bleeding, it im proves visualizat ion of th e operat ive eld an d reduces operat ion t im e.13 Tum or em bolizat ion can also be u sed as p alliat ive t h erapy for cer t ain u n resect able t u m ors. It red u ces t h e m ass e ect on sp in al cord or n er ves by cau sing t u m or n ecrosis an d sh r in kage. How ever, on e sh ou ld be aware of th e possibilit y of im m ediate postem bolizat ion w orsening of th e m ass e ect due to edem a th at requires ju diciou s u se of sh ort term steroids. Curat ive em bolizat ion is occasion ally u sed in pat ien ts w ith cer tain ben ign prim ar y t um ors, such as aneur ysm al bone cysts an d at yp ical h em angiom as.14

Preprocedural Imaging An u n derstan ding of n orm al vascu lar an atom y, abn orm al t um or feeding vessels, an d poten tially dangerous an astom oses is essen tial in perform ing safe an d e ect ive em bolizat ion of th e vascular spin al t um ors. Preprocedure crosssect ion al im aging is n ecessar y for app rop riate plan n ing of spin al angiograp hy an d em bolizat ion . Con t rast-en h an ced CT an d MRI are ver y h elpfu l in determ in ing th e locat ion an d full exten t of th e t u m or.

Spinal Angiography High -qu alit y digit al su bt ract ion angiography is essen t ial for th e iden t i cat ion of th e t u m or feeding vessels an d for localizing radiculom edu llar y arteries poten t ially con t ribut ing to th e ar ter y of Adam kiew icz. Th is procedure gen erally can be perform ed un der m oderate sedat ion . Gen eral an esth esia is u sed for ch ild ren or u n cooperat ive pat ien ts.

29

30

Chapter 3 An ar terial vascular sh eath is rst placed w ith in th e com m on fem oral ar ter y th at provid es vascu lar access for t h e p roced u re. Dep en d ing on t h e t u m or level an d in ten d ed th erapy, cer tain vessels n eed to be select ively cath eterized to bet ter evalu ate th e blood su p ply an d per t in en t ow dyn am ics. For th e cer vical an d u pp er th oracic spin e tum ors, the vertebral arter y, thyrocer vical trunk, costocer vical t r u n k, an d su p erior in tercost al ar teries are select ively cath eterized an d evaluated. In ad dit ion , extern al an d in tern al carot id ar teries are evaluated for any blood supply an d poten t ially dangerou s an astom osis du ring em bolizat ion . For th e low er th oracic an d lum bar spin e lesion s, bilateral segm en tal arteries at th e level of tum or as well as t wo adjacent segm ents above an d below are evalu ated . In ad d it ion , in ter n al iliac, iliolu m bar, lateral, an d m ed ian sacral arteries are evaluated for low er lum bar an d sacral lesion s.14 The cath eter is selectively placed inside each vessel un der st u dy, 3 to 5 m L of n on ion ic iodin ated con t rast is injected by h an d, an d real-t im e digit al su bt ract ion im ages are obt ain ed. Tu m or blush w ith th e origin of t um or-feeding vessels an d any associated radiculom edullar y ar teries

are evaluated. It is crit ical to iden t ify an terior an d posterior spin al ar teries an d th eir con t rib u tors to avoid u n in ten t ion al em bolizat ion of th ese im por t an t ar teries th at can cau se sp in al cord st roke. A con e-beam volu m e st u dy th at can n ow be obt ain ed w ith C-arm digit al angiography system can provide m ore detailed an atom ic in form at ion abou t th e t u m or-feeding ar teries an d poten t ial con t ributors to th e spin al ar teries.

Embolization Procedure Th e goal of t u m or em bolizat ion is to select ively occlude th e t u m or-feeding ar teries by delivering em bolic agen t . It is even m ore im p or tan t n ot to em bolize any vessel th at can jeopardize n orm al fun ct ion . Th e t um or-feeding ar teries are select ively cath eterized using a 3-Fren ch (F) m icrocath eter advan ced th rough a 5F or 6F diagn ost ic cath eter in coaxial fash ion . A h igh qualit y m agni ed preem bolization angiogram is obtained to further delineate the tum or-feeding arter y and to con rm that there are no branches con t ribu t ing to th e spin al ar ter y (Fig. 3.5). Th is preem bolizat ion angiogram also p rovides im p or t an t in for m at ion abou t t h e rate at w h ich

a

b

Fig. 3.5a,b Preoperative tumor embolization of a hypervascular spine tum or. (a) Selective angiogram of the radiculomedullary artery with a dense tumor blush (arrows). (b) Postembolization angiogram of

the same vessel demonstrates occlusion of the tumor feeding arteries. Note the lack of contrast in the areas of tumor blush present in the preem bolization angiogram (arrows).

Interventional Options for Primary Tumors of the Spine th e em bolic m ater ial can be injected w it h ou t any re u x in to th e n or m al circu lat ion . If t h e t u m or-feeding vessels are too sm all to be select ively cath eterized, m icrocoils can be p laced in th e segm en t al ar ter y beyon d th e origin of tum or vessels. This provides a preferential blood ow in to th e t um or vessels. Th e em bolizat ion agen t is t h en d elivered p roxim al to t h e or igin of t u m or vessels, an d t h e p referen t ial blood ow carries th e em bolic agen t in to th e t um or vessels. Th e com m on ly u sed agen t s for em bolizat ion in clu de p olyvinyl alcoh ol (PVA) p ar t icles, t risacr yl m icrosph eres, liqu id N-bu t yl cyan oacr ylate (NBCA), ethyl-vinyl alcoh ol cop olym er (EVOH) onyx, Gelfoam pledgets, and m icrocoils. For opt im al t u m or em bolizat ion , th e em bolic agen t should pen etrate the t um or capillar y bed an d cause perm an en t vascu lar occlusion . Coils by th em selves are n ot e ect ive, as th ey occlude on ly th e proxim al t um or vessels, an d t um or can rap id ly recr u it n ew blood su p p ly from in tersegm en t al collaterals. How ever, coils can p lay a u sefu l role by p rovid in g p referen t ial ow aw ay from un involved dist al vessels an d dangerou s in tersegm en tal an astom oses.13 Liqu id agen t s such as alcoh ol, NBCA, an d onyx can provide rapid an d perm an en t em bolizat ion , bu t th eir u se requ ires experien ce an d great tech n ical expert ise. Eth an ol cau ses extensive t um or n ecrosis du e to in t im al sclerosis an d in ten se in am m ator y respon se. Select ive cath eter izat ion an d carefu l, slow in fu sion is requ ired to avoid any re u x an d com plicat ion s w ith th is agen t . NBCA rapidly polym erizes in to a solid m aterial as it com es in con t act w ith blood or salin e. Th is requires fast , con t in uou s, an d p recise inject ion for opt im al em bolizat ion .15 In con t rast , onyx can be injected slow ly and provides deeper pen etration, w ith th e abilit y to occlu d e m u lt ip le vessels from a single cath eterizat ion . Part iculate agen t s are th e m ost com m only used and are ver y e ective for tum or em bolizat ion .16 Th e ideal agen t sh ou ld be n on biodegradable w ith a size an d sh ape th at en able it to ach ieve distal p en et rat ion in to t u m or vessels and cause perm anent occlusion. Sm aller particles have bet ter pen etration into the t um or vessels. How ever, ver y sm all p ar t icles, especially of sizes less th an 100 µm , h ave a h igh er

risk of cau sing sp in al cord isch em ia an d soft t issue n ecrosis. Gen erally, par t icles of sizes 300 to 500 µm are preferred an d can provide adequ ate em bolizat ion w ith a low er risk of isch em ic com plicat ion s.17 Em bolizat ion can be risky if t u m or-feeding vessels are arising from th e sam e or an adjacen t radicu lom edu llar y ar ter y th at is also sup p lyin g t h e sp in al ar ter y. If th ere is su cien t dist an ce bet w een t h e t w o an d t h e t u m orfeed ing vessels are dist al to th e origin of th e radiculom edu llar y ar ter y, on e can ju diciously at tem pt em bolizat ion by select ively cath eterizing th e feeding ar ter y. Th e p rocedu re sh ou ld be im m ediately term in ated on any suspicion of re u x of em bolizat ion m aterial in to th e radicu lom edu llar y ar ter y or any ch ange in n eu rologic st at u s of th e p at ien t. Em bolizat ion of cer vical sp in e t u m ors is m ore tech n ically ch allenging becau se of com plex ar terial an atom y. Select ive cath eterizat ion of t u m or-feed in g ar ter ies is cr it ical to avoid brain an d sp in al cord st roke. If on e is u n able to select ively cath eter ize t h e t u m or-feed ing vessels, a tem p orar y balloon can be in ated dist ally in t h e ver tebral or carot id ar ter y to avoid any in t racran ial re u x of th e em bolizat ion m aterial. If th ere is t um or en casem en t of th e ver tebral or carot id ar ter y, p erm an en t occlu sion of th e ar ter y u sing coils or det ach able balloon en ables en -bloc surgical resect ion . A balloon test occlu sion is p erform ed before proceeding w ith perm an en t occlu sion to determ in e if th e p at ien t can tolerate su ch a procedure. Du ring th e balloon test occlusion , th e m ain ar ter y is tem porarily occlu ded by in at ing a balloon in side th e lu m en follow ed by app ropriate n eu rologic testing to evaluate its e ect . Th e paren t ar ter y can be sacri ced w ith relat ively low risk if th e patien t does n ot develop any n eu rologic de cits during th e tem porar y occlusion of th e arter y, in clu d ing cer t ain p rovocat ive m easu res su ch as red u cing blood p ressu re d u r ing t h e tem p orar y occlusion .18 A com plete n eurologic exam in at ion sh ou ld be perform ed before an d im m ediately after th e em bolizat ion procedu re to iden t ify any com p licat ion s. After em bolizat ion , p at ien t s are adm it ted to th e in ten sive care un it for close

31

32

Chapter 3 m on itor in g of t h eir n eu rologic st at u s. Post em bolizat ion n ecrosis an d in am m ator y resp on se can cau se t u m or sw elling an d sp in al cord com pression. Short-term steroids are generally u sed to m in im ize t h is. Som e p at ien t s experien ce th e self-lim it ing p ostem bolizat ion syn d rom e, ch aracterized by n au sea, vom it ing, low -grad e fever, an d pain . It can gen erally be m an aged w ith con ser vat ive th erapy. Th e em bolizat ion is ideally perform ed from 24 to 72 h ours before th e surgical resect ion to allow t im e for m a xim al th rom bosis of th e occlu ded vessels an d to p reven t any recan alizat ion or recr u it m en t of n ew blood su p p ly via collaterals. In som e cases w h ere th ere is cord com p ression , it m ay be p ru den t to do su rger y on th e sam e day, after th e em bolizat ion , to avoid w orsen ing of com pression due to t um or n ecrosis an d sw elling.19 Many st u dies h ave sh ow n th at in t raop erat ive blood loss is sign i can tly less in p at ien t s w ho underwent preoperative em bolization than in th ose w h o did n ot . Addit ion al ben e t s in clude sh or ter operat ing t im e an d a redu ced n eed for blood t ran sfusion . Th ere is eviden ce th at a com plete em bolizat ion result s in sign i can tly less in t raop erat ive blood loss th an an in com p lete em bolizat ion . Com p lete em bolizat ion can gen erally be ach ieved in 50 to 86% of patien ts w ith sp in al t u m ors.20 Com m on cau ses of in com plete em bolizat ion include th e presen ce of t h e adjacen t radicu lom edu llar y ar ter y su pplying th e sp in al ar ter y, di cult an atom y preclu ding select ive cath eterizat ion of feeding ar teries, an d dissect ion of th e feeding ar teries from t h e gu id ew ire or cat h eter, lim it in g t h e access for em bolizat ion . Em bolizat ion of a spin al t um or can occasion ally be u sed as th e p rim ar y th erap eu t ic t reat m en t . Cer t ain ben ign t u m ors su ch as aggressive h em angiom as an d an eu r ysm al bon e cyst s m ay respon d to serial em bolizat ion . Palliat ive em bolizat ion can be u sed in pat ien ts w ith un resectable spin al t u m ors. Th e m ost catast roph ic com plicat ion of spinal tum or em bolization is spinal cord and brain st roke. The oth er possible com plication s include parest h esias, t ran sien t paraparesis, an d cord com p ression du e to t u m or n ecrosis an d edem a.

A h igh -qualit y diagn ost ic angiogram an d carefu l evalu at ion of a p reem bolizat ion angiogram is crit ical to iden t ify an d avoid em bolizat ion of radicu lom edu llar y ar teries con t ribu t ing to spinal arter y. Frequent angiogram s sh ould be obtain ed during th e em bolizat ion procedure to iden t ify any n ew collateral path w ays th at w ere n ot visible on th e preem bolizat ion angiogram .

■ Chapter Summary Im age-guided biopsy h as em erged as a safe, effect ive, an d accurate tool for th e de n it ive diagn osis of spin e t u m ors. In m any in st it u t ion s, it h as becom e th e rst test in establish ing th e de n it ive diagn osis an d h as bet ter p at ien t toleran ce, low er m orbidit y, an d low er cost th an conven t ion al open surgical biopsy. Ver tebroplast y an d kyp h op last y are percu t an eou s vertebral augm en tat ion procedu res available for th e t reat m en t of p ain fu l t u m or an d associated p at h ological fract u res. Th e p r im ar y goal of th ese p rocedu res is to redu ce p ain an d im p rove st r u ct urally stabilit y. Th e clin ical outcom e of a ver tebral augm en t at ion p rocedu re p erform ed for spin e t um ors depen ds n ot on ly on th e skills of t h e su rgeon bu t also on t h e ap p rop r iate u t ilizat ion of available tools an d inject able m aterials. Th e judiciou s com bin ed use of plasm am ed iated ablat ion w it h ver tebrop last y an d th orough kn ow ledge of th e cem en t p rop er t ies im proves th e safet y an d e cacy of th ese procedu res. Preoperat ive em bolizat ion of vascu lar t u m ors of th e sp in e sign i can tly redu ces in t raoperat ive blood loss. A detailed un derstan ding of th e vascular an atom y of th e spin e is essen t ial for safe an d e ect ive em bolizat ion . Th e recent developm ent of new er em bolic agents and im proved im aging equ ipm en t en able a m ore com plete devascu larizat ion of t um ors. As w ith any surgical procedure, in ter ven t ion al procedu res are n ot w ith ou t th eir com plicat ion s. A th orough kn ow ledge of an atom y an d w ise select ion of th e tools is crucial to th e success of the procedure. In tegrat ion of im ageguided in ter ven t ion s in to a m ult idisciplin ar y team of experien ced surgeon s, radiat ion an d

Interventional Options for Primary Tumors of the Spine m edical on cologist s, an d p at h ologist s is im por t an t for en suring opt im al pat ien t care.

Pearls Im age-guided needle biopsy has becom e the def nitive test in assessing spine tumors and has lower morbidit y, bet ter patient tolerance, and lower cost than open surgical biopsy. Knowledge of anatomy and available tools is crucial to the success of the procedure. Technique should be discussed with the treating surgeon. Proper technique and selection of the im aging guidance system are often key to obtaining adequate diagnostic biopsy samples. Coaxial biopsy technique in appropriate anatomic planes should be followed to avoid further spread of disease. The goal of vertebral augm entation procedure in pathological fractures is to improve pain and structural stabilit y of the bone through the safe injection of stabilizing material. Plasm a-mediated tumor ablation is a useful tool in tum ors with large epidural extension to create

a cavit y in the tum or that allows for judicious use of PMMA injection. Preoperative embolization of hypervascular tumors result s in signif cant reduced intraoperative blood loss, shorter operating time, and less transfusion requirement. Pitfalls Complications are best avoided by awareness of the factors that contribute to their occurrence. Following m eticulous needle placem ent technique, close m onitoring, and avoiding an excessive am ount of cement, especially in pathological fractures, are strongly recom mended. A high-qualit y diagnostic spinal angiogram and careful evaluation of the preembolization angiogram are critical to identify and avoid em bolization of any branches contributing to spinal artery. Frequent angiogram s should be obtained during particle em bolization procedure to identify any new collateral pathways that may not be visible on preem bolization angiogram.

Refere nces Five Must-Read Refe rences 1. Korn blum MB, Wesolow ski DP, Fisch grun d JS, Herkow it z HN. Com p u ted tom ograp hy-gu id ed biop sy of th e sp in e. A review of 103 p at ien t s. Sp in e 1998;23: 81–85 PubMed 2. Ya e D, Greenberg G, Leit n er J, Gipstein R, Sh apiro M, Bach ar GN. CT-gu ided percut an eous biopsy of th oracic an d lum bar spin e: A n ew coaxial tech n ique. AJNR Am J Neuroradiol 2003;24:2111–2113 PubMed 3. Rim on di E, St aals EL, Erran i C, et al. Percu t an eou s CTguided biopsy of th e spin e: result s of 430 biopsies. Eur Spin e J 2008;17:975–981 PubMed 4. Tam pieri D, Weill A, Melan son D, Eth ier R. Percut an eous aspirat ion biopsy in cer vical spin e lyt ic lesion s. In dicat ion s an d tech n iqu e. Neuroradiology 1991;33: 43–47 PubMed 5. Gu pt a S, Hen n ingsen JA, Wallace MJ, et al. Percu t an eou s biop sy of h ead an d n eck lesion s w it h CT gu idance: various approach es an d relevan t an atom ic an d tech n ical con siderat ion s. Radiograp h ics 2007; 27:371–390 PubMed 6. Jen sen ME, Evan s AJ, Math is JM, Kallm es DF, Cloft HJ, Dion JE. Percutaneous polym ethylm ethacr ylate vertebroplast y in the treatm ent of osteoporotic vertebral body com pression fract u res: tech n ical aspect s. AJNR Am J Neuroradiol 1997;18:1897–1904 Pu bMed

7. Barr JD, Bar r MS, Lem ley TJ, McCan n RM. Percu t an eous ver tebroplast y for pain relief an d spin al st abilizat ion . Spin e 2000;25:923–928 PubMed 8. Janjan NA. Radiat ion for bon e m et ast ases: conven t ion al tech n iques an d th e role of system ic radioph arm aceut icals. Can cer 1997;80(8, Suppl):1628–1645 PubMed 9. Georgy BA, Won g W. Plasm a-m ed iated rad iofrequ en cy ablat ion assisted p ercu t an eou s cem en t in ject ion for t reat ing advan ced m align an t ver tebral com pression fract ures. AJNR Am J Neuroradiol 2007; 28:700–705 Pu bMed 10. Lin EP, Ekholm S, Hiw at ash i A, Westesson PL. Vertebroplast y: cem en t leakage in to th e disc in creases th e risk of new fract ure of adjacen t ver tebral body. AJNR Am J Neuroradiol 2004;25:175–180 PubMed 11. Pearce JA. Elect rosurger y. New York: Wiley Medical, 1986 12. Murphy KJ, Deram on d H. Percut an eous vertebroplast y in ben ign an d m align an t disease. Neu roim aging Clin N Am 2000;10:535–545 Pu bMed 13. Berkefeld J, Scale D, Kirch n er J, Heinrich T, Kollath J. Hyper vascular spin al t um ors: in uen ce of th e em bolization technique on perioperative hem orrhage. AJNR Am J Neuroradiol 1999;20:757–763 PubMed

33

34

Chapter 3 14. Ozkan E, Gupt a S. Em bolizat ion of spinal t um ors: vascular an atom y, in dicat ion s, and tech n ique. Tech Vasc In ter v Radiol 2011;14:129–140 Pu bMed 15. Mindea SA, Eddlem an CS, Hage ZA, Batjer HH, On dra SL, Bendok BR. En dovascular em bolizat ion of a recurren t cer vical gian t cell n eoplasm using N-bu t yl 2-cyan oacr ylate. J Clin Neu rosci 2009;16:452–454 PubMed 16. Gore P, Th eodore N, Brasilien se L, et al. Th e ut ilit y of onyx for preoperat ive em bolizat ion of cran ial an d sp in al t u m ors. Neu rosu rger y 2008;62:1204–1211, discussion 1211–1212 PubMed 17. Ben dszus M, Klein R, Burger R, Warm uth -Met z M, Hofm an n E, Solym osi L. E cacy of t risacr yl gelat in m icrosph eres versus polyvinyl alcoh ol par t icles in

the preoperative em bolization of m eningiom as. AJNR Am J Neu rorad iol 2000;21:255–261 PubMed 18. Bern stein A, Lasjun ias P, TerBrugge KG. Tu m ors of th e spin al colum n an d spin al cord. In : Surgical Neuroan giography, vol 2. Berlin : Springer, 2004:874–877 19. Kai Y, Ham ada J, Morioka M, Yan o S, Todaka T, Ush io Y. Appropriate inter val bet w een em bolizat ion an d su rger y in pat ien t s w ith m en ingiom a. AJNR Am J Neuroradiol 2002;23:139–142 Pu bMed 20. Prabh u VC, Bilsky MH, Jam bh ekar K, et al. Resu lt s of p reop erat ive em bolizat ion for m et ast at ic sp in al n eop lasm s. J Neu rosu rg 2003;98(2, Su p p l):156–164 PubMed

4 Radiation Therapy for Primary Bone Tumors Brent Y. Kimball and Yoshiya (Josh) Yamada

■ Introduction Pr im ar y m align an t bon e t u m ors of t h e sp in e p resen t sign i can t t reat m en t ch allenges w it h regard to ach ievin g local t u m or con t rol w h ile preser ving n eu rologic fu n ct ion . Th e m ost com m on prim ar y t u m ors involving th e spin al colu m n are ch ord om as, ch on d rosarcom as, an d osteogen ic sarcom as. Sign i can t advan ces in both su rger y an d radiat ion are rede n ing th eir roles in t h e t reat m en t of t h ese lesion s. Advan ced su rgical tech n iqu es for en -bloc resect ion of p r im ar y sp in e t u m ors im p roves t h e abilit y to ach ieve m argin al or w id e cu rat ive resect ion s 1–3 ; h ow ever, m any t u m ors are n ot am en able to en -bloc resect ion . Conven t ion ally fract ion ated radioth erapy is frequ en tly u sed as an adjuvan t to surger y. In m ost cases, conven t ion al external beam radiat ion is n ot an e ect ive prim ar y th erapy for m any prim ar y spin e t u m ors. Th e p oor lon g-term ou tcom es after conven t ion al radiat ion th erapy is likely related to th e relat ively low d ose of radiat ion th at can be given n ear th e sp in al cord. Con cern abou t sp in al cord toleran ce h as t yp ically lim ited radiat ion doses to less th an 54 Gy in st an dard fract ion s.4,5 Low -dose areas w ith in th e irradiated eld are t h e m ost likely to resu lt in d isease recu rren ce.6 Th ree rad iat ion tech n iqu es are cu r ren t ly being u sed to in crease th e d ose to th e t u m or w hile sp aring n orm al t issu e toleran ce in an attem pt to t reat prim ar y spin e t um ors: par t icle

beam t reat m en t , su ch as p roton beam t h erapy; brachyth erapy; an d h igh -dose con form al photon therapy, such as im age-guided intensit ym od u lated rad iat ion t h erapy (IMRT) or stereot act ic radiosu rger y (SRS). Alt h ough class 1 eviden ce is lacking, m oun t ing research dem on st rates th e con t ribu t ion of h igh -dose radiat ion to ach ieve local t um or con t rol for variou s spin e t u m or path ologies. Recen tly, th e abilit y to deliver SRS u sing im age-gu ided IMRT h as p roven u sefu l in th e n eoadjuvan t an d postoperat ive set t ing.

■ Surgery Th e role of en -bloc resect ion to ach ieve a m argin al or w ide resect ion is w ell est ablish ed for ext rem it y sarcom as.7–9 Over th e past 10 years, sp in e su rgeon s h ave developed tech n iqu es for en -bloc resect ion of prim ar y spin e t u m ors th at im proves cu re rates com pared w ith piecem eal, curet tage techniques. These techn iques are safe for pat ien t s w ith isolated ver tebral body or p osterior elem en t disease. Borian i et al10 pu blish ed a su rgical series regard ing en -bloc sp on dylectom y for low -grad e ch on d rosarcom a of th e m obile spin e in 22 pat ien ts. Of th e 12 patients w ho underwent en-bloc excision, 9 (75%) m ain t ain ed local con t rol at a m edian follow -u p of 81 m on t h s (ran ge, 2 to 236 m on t h s). Of th e th ree recu r ren ces, t w o h ad con t am in ated

36

Chapter 4 m argin s at su rger y from ep idu ral disease. Th e rem ain ing 10 p at ien ts u n der w en t cu ret tage resect ion an d by de n it ion h ad in t ralesion al resect ion s w ith p osit ive h istological m argin s. At a m edian follow -u p of 36 m on th s, th e recu rren ce rate w as 100%, an d 80%died of disease.10 Of n ote, in th is grou p , th ree pat ien t s received conventional-dose external beam radiation and n o pat ien ts received h igh -dose con form al radiat ion w ith proton beam s or im age-guided IMRT. Alth ough cu rat ive resect ion s are feasible in som e p at ien ts, m any p at ien t s p resen t w ith factors th at m ay preclude en -bloc resect ion th at ach ieves n egat ive h istological m argin s. En -bloc resect ion of th e isolated vertebral body or posterior elem ents is technically feasible. How ever, accord ing to t h e Wein stein –Bor ian i–Biagin i (W BB) classi cat ion ,11 p at ien t s w h o p resen t w ith epidural t um or, m ultilevel large paraspin al m asses, or circum ferential bone disease are not can didates for m argin al or w ide excision s (i.e., en bloc w ith n egat ive m argin s). Th e feasibilit y of ach ieving a w ide or m argin al excision is lim ited by th e risk of n eu rologic or adjacen t st ru ct ure injur y. For exam p le, resect ing th e d u ra en bloc w it h a sp ecim en cou ld p ossibly p rovid e a m argin on ep id u ral t u m or. Un for t u n ately, th e loss of sp in al u id bu ering th e spin al cord in creases th e probabilit y of inju r y to th e spin al cord an d com plicat ion s of cerebrospin al uid (CSF) leak. If t u m or is sp illed du ring resect ion , th ere seem s to be a h igher probabilit y of in t rad ural seeding as w ell.12 In a series of 59 sp in e sarcom as review ed using th e radiograph ic criteria est ablish ed by th e W BB classi cat ion , ap proxim ately 15% of patien ts w ere can didates for en -bloc excision s th at cou ld ach ieve a w ide or m arginal m argins. As noted in extrem it y sarcom as, on ce t h e t u m or is violated, th e risk of recurren ce sign i can tly in creases.7 How ever, th e t reat m en t of sarcom as at oth er sites h as dem on st rated th e u t ilit y of adjuvan t h igh -dose radiat ion in th e set t ing of m icroscopic or gross residual disease post resect ion .13

■ Radiation Therapy Rad iat ion t h erapy is an ext rem ely im p or t an t m odalit y in th e t reat m en t of p rim ar y sp in e t u -

m ors. How ever, th e relat ive radioresist an ce of th ese t u m ors requ ires doses w ell above spin al cord toleran ce for du rable local t u m or con t rol. From t h e p arad igm of ext rem it y sarcom as, radiat ion doses of greater th an 60 Gy in 2- Gy fract ion s are requ ired for con t rol of posit ive m icroscop ic m argin s an d greater th an 70 Gy for gross residual disease.14 Traditional concepts of spinal cord tolerance using conventional radiat ion tech n iqu es establish es th e TD5/5 (th e tolerance dose at w h ich th ere is a 5% p robabilit y of a com p licat ion w ith in 5 years) at 50 Gy in 1.2- to 2-Gy fract ion s above w h ich th ere ap pears to be a signi cantly increased risk of developing radiat ion m yelit is.15 Toxicit y to th e spin al cord m ay also be associated w ith th e length of cord ir rad iated . In ad d it ion to t h e sp in al cord , toxicit ies to p arasp in al st ru ct u res, su ch as th e bow el, t h e kid n eys, an d th e esop h agu s, n eed to be con sid ered . Un for t u n ately, t h e rad iat ion toleran ces of t h ese organ s ran ge from 23 to 60 Gy. Th u s, tech n ologies an d tech n iqu es th at m in im ize th e rad iat ion d ose to cr it ical organ s an d allow h igh d ose to th e t u m or are essen t ial for m ean in gfu l rad iat ion t reat m en t of sp in e m align an cies.

Proton Beam Advances in radiation technology, including the in t rodu ct ion of h adron s (h igh -dose proton s or ch arged part icles, in clu ding carbon ion s, h eliu m , or n eon ), h ave led to h igh er doses of radiat ion being delivered to th e target volu m e w ith redu ced t issue inju r y an d im proved radiobiological e ect .16 Th ere are m any im p or t an t sim ilarit ies as w ell as di eren ces bet w een IMRT an d proton beam radioth erapy. Th e rat ion ale for u sing p roton beam s is th e excellen t d ose d ist ribut ion at th e t u m or t arget an d virt ually n o exit dose beyon d th e t arget volu m e. Th e Bragg peak ph en om en on ch aracterist ic of proton beam radiat ion resu lt s in an ext rem ely steep dose fallo th at can be m easured over a cou rse of m illim eters (Fig. 4.1). Th e biological im pact of proton beam t reatm en t an d IMRT is relat ively m in or. Th e con version factor com m on ly used to com pare th e e ect iven ess of proton beam t reat m en t w ith photon radiotherapy is 1.1. Thus, 1 Gy of proton

Radiation Therapy for Primary Bone Tumors

Fig. 4.1 The Bragg peak e ect for 250-MeV protons. Note that there is no dose at a depth beyond the  Bragg peak.

radiat ion is biologically equ ivalen t to 1.1 Gy of radiat ion from a cobalt sou rce, or 1.1 cobalt gray equivalents (CGE). The advantage of proton beam t reat m en t does n ot lie in an en h an ced e ect of p roton s again st t u m or cells. Rath er, th e advan t age lies prim arily in t h e lack of exit dose after th e Bragg peak (Fig. 4.2). If th e Bragg peak occu rs ju st before th e sp in al cord, th en the spinal cord receives m inim al radiation dose, in st ark con t rast to p h oton radiat ion , for w h ich exit dose is sign i can t . If m an ipu lated p rop erly, proton beam s can deliver h igh er doses to a t um or in close proxim it y to th e spin al cord w h ile exposing th e spin al cord to m in im al radiat ion d oses. IMRT alw ays delivers h igh er radiat ion dose to sp in al cord adjacen t to a spin al t u m or, becau se t h ere is n o Bragg p eak e ect w it h p h oton s, an d som e p h oton s w ill p ass th rough th e t u m or in to th e sp in al cord . IMRT can deliver radiat ion d ose to th e t u m or u p to w h at is con sidered th e cord toleran ce. In m ost cases, doses of u p to 74 Gy can be given to a vertebral body t um or w h ile lim iting th e sp in al cord to 50 to 54 Gy. Th u s, IMRT is able to deliver sim ilarly h igh doses of radiat ion to th e t um or, but th e spin al cord doses are poten t ially low er w ith th e use of proton beam s.17 Proton beam radiat ion h as a dist in ct disadvan tage in pat ien t s w ith im plan ted spin al xat ion h ardw are. Th e greater den sit y of m etallic im -

plan t s can a ect th e range of proton s u p to 10 m m an d th e dose greater th an 10%in th e h igh dose region .18 W h en th e spin al cord m ay on ly be a few m illim eters aw ay from th e h igh -dose region , th e un cert ain t ies associated w ith t it an iu m im plan t s are u n accept able. Mon te Carlo sim u lat ion tech n iqu es are n ecessar y to obt ain accu rate d osim et r y in t h e set t in g of sp in al in st r u m en t at ion . On t h e ot h er h an d , p h oton rad iat ion an d IMRT d osim et r y are m in im ally im pacted by surgical h ardw are. Heavy ion s su ch as carbon ion s provide biological, in addition to physical, advantages com pared w ith photons or protons, in term s of their h igh relat ive biological e ect iven ess an d red uced oxygen -en h an cem en t rat io in th e t u m or region .19 Hypofract ion ated carbon ion th erapy is a prom ising t reat m en t approach for large sacral chordom as.20 Th ere h as been con sid erable exp er ien ce w it h t h e t reat m en t of n eu raxis t u m ors u sing proton beam s to deliver fract ion ated th erapy. Proton beam radioth erapy’s m ain lim it at ion is th e h igh cost of bu ilding an d m ain t ain ing su ch facilit ies. Th e cost/ben e t rat io of proton beam t reat m en t is ext rem ely con t roversial, bu t th e cost of proton th erapy is exp ected to decrease rapidly. Curren tly it is approxim ately t w ice as expen sive as IMRT.21 Excellen t resu lts for uveal m elan om a h ave been rep or ted u sin g p roton

37

38

Chapter 4

Fig. 4.2 Axial proton beam dose distribution for a sacral lesion entering posteriorly. Note the lack of exit  dose into the pelvis.

beam th erapy in a hypofract ion ated m an n er (m edian dose 70 CGE in ve fract ion s),22 but dat a repor t ing th e u se of single-fract ion radiat ion w ith p roton beam th erapy for th e m an agem en t of t u m ors of th e n eu raxis is lacking. Proton -beam th erapy com bin ed w ith w ide en -bloc excision is th e accepted treatm ent standard in t h e m an agem en t of ch ord om as at a num ber of cancer centers, especially in patients w ith a prim ar y t um or as opposed to disease recurren ce.16 Hug 23 rep or ted rad iat ion resu lt s from 47 p at ien t s w ith p rim ar y or recu rren t osteogen ic an d ch on drogen ic t u m ors t reated w ith com bin ed proton /p h oton th erapy. Th e 5-year local con t rol rates w ere 53% for ch ordom a, 50% for osteogen ic sarcom a, an d 100% for ch on drosarcom a.23 Of t h e six failu res, ve w ere in eld recurrences an d on e out of eld. A t ren d w as

n oted tow ard im proved local t um or con t rol in pat ients receiving greater than 77 CGE. The failu res w ere seen p rim arily in p at ien ts w ith less th an 77 CGE delivered to th e t u m or volum e. Au st in et al6 repor ted a case series of 141 p at ien t s t reated for ch ord om a an d ch on d rosarcom a of th e skull base an d cer vical spin e, u sing m ixed proton /ph oton beam th erapy at a m edian of 69 CGE. At a m edian follow -up of 3 years, 26 failures w ere n oted. Of th ese failures, 23%received th e prescribed tum oral dose. How ever, 77% of failures occu rred in areas th at received less th an th e prescribed t um oral dose, th e m ajorit y of w h ich w ere in region s con st rain ed by n orm al t issu e toleran ce. All t u m ors th at failed in th e h igh -dose region h ad volu m es greater th an 75 cc. Pat ien ts w ith cer vical spin e disease h ad a h igh er rate of recu rren ce (10 of 26) and larger tum ors (average volum e of 102 cc)

Radiation Therapy for Primary Bone Tumors than th ose w ith base of skull disease (16 of 115) w ith an average volu m e of 63 cc. Th ese st udies u n derscores th e ch allenges of delivering u n iform it y of prescribed dose an d th e t reat m en t of large-volum e t um ors in close proxim it y to dose-lim it ing n eural st r u ct u res.6

Intensity-Modulated Radiation Therapy Total doses in excess of 70 Gy can be rout in ely adm in istered to paraspin al t u m ors using ph oton IMRT tech n iques w h ile m ain tain ing spin al cord doses at th e sam e level ach ieved w ith p roton beam s (Fig. 4.3). Typically, ve to seven rad iat ion beam s are u t ilized arou n d a sin gle isocen ter. An inverse t reat m en t plan n ing algorith m is u sed to assign t arget doses to t arget volu m es, an d pen alt ies are given to excessive

doses to avoidan ce organ s at risk. Mu lt iple iterat ion s of th e n u m ber of beam s, beam angles, and beam energies are utilized to arrive at a best possible solu t ion th at m eet s th e dose goals for th e t arget volu m es an d dose-sen sit ive n earby organ s. Each in dividual beam in ten sit y is m odu lated du ring t reat m en t in a w ay th at th e su m tot al of all th e beam s results in a dose cloud th at con form s to th e th ree-dim en sion al ch aracterist ics of th e t arget volum e an d t ypically requ ires ver y h igh dose gradien ts n ear th e spin al can al an d esoph agus, in th e case of cer vical an d th oracic lesion s. Kidn eys an d bow el loops are com m on avoidan ce st r uct u res in th e lum bar spin e. At Mem orial Sloan -Ket tering Can cer Cen ter (MSKCC), w e h ave rep or ted t h e u se of h igh d ose IMRT for t h e m an agem en t of p rim ar y m align an cies of th e m obile spin e. Th e act u arial

Fig. 4.3 Dose distribution for a T11 osteosarcom a that is circum ferential around the spinal canal. The  prescribed dose was 72 Gy and the maximum dose to the spinal cord was 54 Gy.

39

40

Chapter 4 local control was found to be 75%, w ith follow-up exten ding to a m axim u m of 40 m on th s. No p at ien t experien ced m yelit is. Th e m edian dose w as 6,600 cGy (range, 5,400–7,200 cGy).24 We h ave t reated seven ch ordom as an d ve chon drosarcom as, t w o of w h ich w ere h igh grad e an d th e rem ain ing in term ediate or low grade. At a m edian follow -up of 18 m on th s, th e local con t rol rate for ch ord om as w as 86% an d for ch on d rosarcom as w as 80%, w it h t h e single failu re occu rring in a h igh -grade p at ien t . Th e follow -u p is too sh or t to draw m ean ingfu l con clu sion s, bu t fract ion ated th erapy u sing im ageguided IMRT is possible at doses sim ilar to th ose repor ted for proton beam th erapy.25

Brachytherapy Brachyth erapy h as also been u sed to t reat p rim ar y t u m ors. Iodin e 125 h as com m on ly been u sed to t reat posit ive m icroscop ic disease or m inim al residual gross disease follow ing tum or resect ion . Good resu lt s h ave been rep or ted in th e t reat m en t of p arasp in al t u m ors an d epidural disease.26,27 Rogers et al27 reported a series of 25 pat ien t s w h o w ere im p lan ted in t raop erat ively w ith 125 I seeds. Tw en t y-t w o (88%) failed prior extern al beam radiat ion th erapy. At a m ean follow -u p of 19.2 m on th s, fou r pat ien t s dem on st rated local failu re an d th e 3-year act u arial con t rol rate w as 72.9%. No radiat ion toxicit y w as seen in th is st u dy. How ever, rad iat ion m yelitis h as been reported 34 m onths follow ing the routine adm inistrat ion of 131 I for m etastatic thyroid can cer. Th e level of m yelit is en t ailed sign i can t epidu ral disease th at w as n early circum feren t ial, likely result ing in a h igh su rface dose to th e spin al cord over a sh or t p eriod of t im e.28 DeLan ey et al29 h ave developed ap p licators for t h e deliver y of h igh -d ose rad iat ion u sing ir id iu m 192 an d yt t r iu m 90.29 90 Y is a p u re B-em it ter an d id eal for d eliver in g h igh -d ose radiat ion (e.g., 7.5 to 15 Gy) to th e dura w ith ou t toxicit y to th e sp in al cord . Th e 90 Y d ose p en et ran ce is 29% at 2 m m an d 9% at 4 m m . Th is provides an adequate m argin on th e spin al cord, if th e gross t u m or h as been resected in th e absen ce of a spin al u id leak. In th is series, seven of eigh t p at ien t s w ere t reated for sar-

com a, six of w h om w ere con t rolled at a m edian of 24 m on th s. No radiat ion m yelit is h as been seen to date. 90 Y m ay u lt im ately p rove to be excellen t as an adjuvan t th erapy for radioresistant tum ors w ith epidural disease w hen com bin ed w ith im age-guided IMRT or proton beam th erapy. Im age-gu ided IMRT uses dose p ain ting to lessen th e dose at th e spin al cord m argin in order to spare spin al cord toleran ce, poten t ially u n derdosing th is area. 90 Y m ay im prove th e dose dist ribu t ion for epidu ral disease an d facilitate t reat m en t plan n ing an d deliver y of im age-guid ed IMRT. Th e lim itat ion of 90 Y is th at th e applicators are custom m ade for each p at ien t an d th e isotop e h as a ver y sh ort h alflife. If th e ep idural t um or resect ion is m ore exten sive th an predicted by m agn et ic reson an ce im agin g (MRI) or t h e case is d elayed several days, th e plaque m ay be w asted. Mem orial Sloan -Ket tering Can cer Cen ter is p ion eer in g t h e u se of a sim ilar d u ral p laqu e ap p licat ion of in t raop erat ive brachyt h erapy for th e t reat m en t of m et ast at ic an d p rim ar y sp in e t u m ors. In th is t ype of brachyth erapy a t h in p iece of silicon e is coated w it h rad ioact ive p h osp h or u s ( 32 P). 32 P h as a t w o w eek h alf life an d is sligh tly less pen et rat ing th an 90 YT, both im p or t an t advan t ages for 32 P. Th e plaqu e is tem p orarily in serted directly onto dura con t am in ated w ith t u m or cells du ring su rger y an d rem oved before th e operat ion is over (Fig. 4.4). Sim ilar to th e 192 Ir an d 90 Y used by Delan ey et al,29 th is allow s deliver y of h igh doses to th e cord surface (25 Gy) w h ile sparing th e n earby spin al cord to doses t ypically less th an 2 Gy. High -d ose-rate (HDR) brachyt h erapy can also be delivered to spin e t u m ors u sing sm all catheters w h ile th e pat ien t is un der an esth esia. As w ith oth er form s of brachyth erapy, th is ap p licat ion m ay assist in th e dosim et ric plan n ing ch allenges of radioth erapy.30 32 P h as been used successfully for years as a local t reat m en t for oth er t ypes of t um ors, but its safet y an d ut ilit y in th e m an agem en t of prim ar y spin al axis t um ors is st ill un der invest igat ion at MSKCC.31

Stereotactic Radiation Stereotact ic spin e radiosu rger y delivers an en t ire cou rse of radiat ion in on e to ve fract ion s.

Radiation Therapy for Primary Bone Tumors

Fig. 4.4 The

30

P source is placed directly on the dural surface after epidural decompression.

Hence, the dose of radiation per fraction is m uch h igh er th an w ith conven t ion al radioth erapy, an d ext rem e accuracy an d precision is cr ucial. IMRT tech n iqu es are t ypically used to gen erate th e n ecessar y con form it y to lim it doses w ith in safe levels to n earby organ s at r isk. Im ageguided technology, such as on-board cone beam com p u ted tom ograp hy (CT) scan n ers, w h ich are in tegrated in to th e radiat ion deliver y platform , provide th ree-dim en sion al im aging an d spat ial data to en sure th at th e radiat ion is given exactly to th e righ t place in n ear real t im e. Th e role for stereot act ic radiosu rger y in th e t reatm ent of prim ary tum ors is yet to be determ ined, an d experien ce is lim ited com pared to proton beam th erapy. Th eoret ically, h igh -dose singlefract ion th erapy m ay im p rove local t u m or con t rol com pared to st an dard fract ion th erapy. From a biological st an dpoin t , t u m or h istologies w ith low α /β rat ios, su ch as sarcom as, respon d bet ter to larger fract ion sizes. Preclin ical eviden ce suggests th at single-fract ion th erapy greater th an 8 to 10 Gy results in apoptosis of tum or cells based on the acid sphingom yelinase

path w ay.30 Microvascular dam age in th e t um or from h igh -dose fract ion s likely con t ribu tes to cell death . Th e radiobiological ben e t m ay be m ost p ron oun ced in t u m or t ypes t radit ion ally con sidered resist an t to st an dard fract ion ated radiat ion th erapy. A con st r u ct called th e biological e ect ive dose (BED) (Table 4.1) h as becom e a w idely accepted w ay to com pare the potency of d i eren t radiat ion dose sch edu les.32 Alth ough th e assu m pt ion s m ade for BED calcu lat ion s based on lin ear qu adrat ic form alism are n ot likely to be accu rate at h igh d oses per fract ion , su ch as in radiosurger y,33 a com parison of differen t d ose sch ed u les d em on st rates t h e sign i can t advan t age of in creasing t h e d ose of radiat ion per fract ion for radioresist an t t u m ors (α /β = 2) com pared w ith relat ively sen sit ive t um ors (α /β = 10). Th us in creasing th e fract ion size is a form of biological dose escalation. Based on th is sim pli ed an alysis, a cou rse of p roton beam rad ioth erapy giving 76 CGy in 2- Gy fract ion s w ou ld give th e BED of 152 Gy for radioresistan t t u m ors.

41

42

Chapter 4 Table 4.1 Biologic Ef ective Dose (BED) Schedules

α /β = 2 α /β = 10

24 Gy/12

24 Gy/8

24 Gy/6

24 Gy/4

240 Gy/2

24 Gy/1

48 Gy 29 Gy

72 Gy 31 Gy

96 Gy 34 Gy

120 Gy  38 Gy

168 Gy  53 Gy

312 Gy 81.6 Gy

Note: The schedules demonstrate the signi cant advantage of increasing the dose of radiation per fraction for radioresistant tumors (α /β = 2) compared to relatively sensitive tum ors (α /β = 10).

In ord er for su ch h igh biological d oses of radiat ion to be adm in istered safely, it is crit ical t h at rad iat ion be d elivered in a ver y con form al m an n er w ith h igh accu racy an d precision , to preven t th e exposu re of dose-sen sit ive n orm al t issu es to h igh doses. Im age-gu ided tech n iqu es 34 cou p led w it h IMRT h ave p rovid ed th e avenue for such t reat m en t . Th e ext rem ely h igh local con t rol rates an d ver y low toxicit y associated w ith m et astat ic ren al cell carcin om a sp in e lesion s t reated w it h h igh -d ose singlefract ion radiosurger y appears to suppor t th is paradigm .35 Stereotact ic radiosu rger y is cu rren tly being used as n eoadjuvan t th erapy at MSKCC. For t u m ors su ch as ch ordom a, ch on drosarcom a, an d lip osarcom a, n eoadjuvan t radiat ion m ay decrease th e possibilit y of t um or dissem in at ion from intralesion al resection. This m ay also bene t pat ien t s w h o u lt im ately u n d ergo en -bloc resect ion w ith n egat ive h istological resect ion s in w h ich t u m or d issem in at ion is som et im es fou n d . Alt h ough it is feasible to u se im agegu id ed IMRT to give h igh -d ose rad iat ion in st an dard fract ion at ion , m any p rim ar y t u m ors of t h e sp in e m ay resp on d best to h igh -d ose hypofract ion ated or single-fract ion radiat ion . How ever, w h en adm in istering su ch h igh -dose th erapy, ext rem e care an d cau t ion are m an dator y to m in im ize com plicat ion s. From a surgical persp ect ive, th e advan t age of u sing h igh ly con form al deliver y of radiat ion is th e decreased soft t issu e dam age an d con sequ en t decreased risk of wound com plications. Patients m ay safely u n d ergo su rger y w it h in 2 w eeks of im agegu ided IMRT. Th e tech n iqu e for spin e rad iosu rger y 36 ut ilizes an inverse opt im ized t reat m en t plan n ing. Typically, ve to seven radiat ion beam s are focused arou n d a single isocen ter. Each beam ’s in ten sit y is m odu lated w ith m u lt ileaf collim a-

t ion using a sliding w in dow tech n ique sim ilar to th at for IMRT. All t arget volu m es are gen erated based on International Com m ission on Radiat ion Un its an d Measurem en ts (ICRU) report 50 an d on th e gu idelin es set by th e In tern at ion al Stereot act ic Radiosurger y Con sort iu m .37 Th e gross t u m or volu m e (GTV) is de n ed as gross disease visu alized on MRI an d CT im aging. Th e clin ical target volum e (CTV) w as den ed as th e GTV as w ell as areas of p oten t ial m icroscop ic spread. Th e plan n ing t arget volum e (PTV) in cluded th e CTV w ith a 2- to 3-m m m argin for set u p u n cer t ain t y an d p oten t ial p at ien t m ot ion . In cases of ep idu ral disease, th e CTV an d PTV overlap in th e region of th e spin al can al in order to spare th e spin al cord. As par t of th e plan n ing process, all volum es are review ed by a team of radiat ion on cologist s and neurosurgeons. Sim ilarly, w ith anterior extraosseou s disease arou n d th e esoph agu s, CTV an d PTV overlap GTV to lim it esoph ageal dose. Th e spin al cord , as de n ed on a sim u lat ion CT m yelogram , is con st rain ed to a m axim u m dose of 14 Gy to a single voxel, an d n o t reat m en t plan sh ou ld exceed th is level. Th e esop h agu s is con st rain ed to 14.5 Gy to 2.5 cc of esop h agu s, alth ough th is con st rain t can be exceeded at th e discret ion of th e t reat ing p hysician . Th e cau da equin a is lim ited to m axim um dose of 18 Gy. Bow el is con st rain ed to allow n o m ore t h an 5 cc to receive m ore th an 16 Gy. In cases of sign i can t bow el overlapping t arget volum es in th e sacr u m , salin e m ay be in fu sed bet w een t u m or an d bow el du ring t reat m en t to tem p orarily displace bow el aw ay from th e an terior asp ect of th e t arget volu m e.38

Sarcoma and Radiation Treatment Conventional radiation treatm ent of sarcom as of th e spin e requ ires p rolonged t reat m en t cou rses

Radiation Therapy for Primary Bone Tumors an d resu lt s in p oor local con t rol. Hyp ofractionated and single-fraction im age-guided SRS (IG-SRS) m ay provide a m ore e ect ive m ean s of m an aging th ese radioresistan t lesion s. Pr im ar y sarcom as, alt h ough u n com m on , car r y a h igh m or t alit y bu rd en —n early 50%.39 Met ast at ic d isease in th e sp in e is far m ore com m on , accou n t in g for u p to 70% of all m et ast ases to bon e an d a ect ing up to 10% of all can cer p at ien t s.40,41 Aggressive su rgical an d radiotherapeutic approaches are often required to p reven t t u m or-associated n eurologic dam age, est ablish local con t rol, an d poten t ially im prove overall su r vival.40,42 Rad iat ion t h erapy is essen t ial in sp in al lesion s to opt im ize local t u m or con t rol.43,44 Conven t ion al extern al beam radiat ion th erapy (C-EBRT) provides lim ited con t rol, as spin al sarcom as are con sidered radioresistan t .42,45,46 Yet , SRS m ay p rodu ce local con t rol rates greater th an 90% w ith low risk of inju r y to n earby organ s an d th e spin al cord.40,47–49 High -dose sin gle or hyp ofract ion ated IG-SRS h as also been im plem en ted in th e p ostop erat ive set t ing w ith sim ilarly good con t rol rates.50–52 At MSKCC bet w een 2005 an d 2012, 147 prim ar y (26.5%) an d m et ast at ic (73.5%) sp in al sarcom a lesion s w ere t reated, 49 w ith hyp ofract ion ated SRS (m ean 24 Gy in th ree fract ions) an d 98 w ith single-fract ion SRS (24 Gy). Med ian follow -u p w as 14.1 m on t h s (ran ge, 1–80.7 m on t h s); for p at ien t s w it h p r im ar y lesion s th e m edian follow -up w as 19.1 m on th s and for m etastatic lesions th e m edian follow -up w as 12.5 m on th s. Of th e 147 lesion s, 18 (12.2%) w ere previously irradiated; 49 (33.3%) of the lesions were treated w ith IG-SRS postoperatively, an d 12 (8.2%) w ere t reated w ith n eoadjuvan t SRS. Overall local progression -free sur vival at 12 m on th s w as 92.1% (95% con den ce in ter val [CI], 86.9–97.3%); overall su r vival at 12 m on th s w as 70.7%(95%CI, 61.7–79.7%). Patients treated w ith single-fract ion IG-SRS h ad a sign i can t ly bet ter local p rogression -free su r vival th an p at ient s t reated w ith hypofract ion ated IG-SRS, 88.2%(95%CI, 80.2–96.2%) versus 68.2%(95%CI, 47.8–88.6%) at 24 m on th s (log-ran k p = 0.005). Sixt y-on e lesion s in 58 pat ien ts w ere t reated w ith n eoadjuvan t or adjuvan t IG-SRS (i.e., radiat ion t h erapy before or after su rgical m an age-

m ent); 49 lesions were irradiated postoperatively and 12 preoperat ively. Overall sur vival and local progression -free sur vival at 1 year w ere 70.1% (95% CI, 57.5–82.7%) an d 88.1% (95% CI, 78.9– 97.3%), respect ively. Th is coh or t in clu ded both m etast at ic an d prim ar y pat ien ts; for th e m etastat ic su bset (n = 32 pat ien ts, 35 lesion s), th e overall su r vival an d local progression -free su rvival at 1 year w ere 61.2% (95% CI, 43.6–78.8%) an d 86.5% (95% CI, 73.9–99.1%), resp ect ively; for p at ien t s w it h p r im ar y lesion s (n = 26 p at ien t s, 26 lesion s), overall su r vival an d local progression -free sur vival at 1 year w ere 81.8% (95% CI, 65.4–98.2%) an d 90.2% (95% CI, 77– 100%), respect ively. Overall sur vival an d local progression-free survival were not signi cantly di eren t (log-ran k p = 0.123 an d 0.442, resp ect ively) for prim ar y versus m etast at ic lesion s. Tim ing of th e radiat ion w ith respect to surger y did n ot h ave a sign i can t e ect on local progression -free sur vival (log-ran k p = 0.170); h ow ever, pat ien ts t reated w ith preoperat ive IG-SRS did h ave im p roved overall su r vival (logran k p = 0.040). Pat ien ts t reated w ith surger y and single-fraction IG-SRS had signi cantly better overall an d local progression -free su r vival th an pat ien ts t reated w ith hypofract ion ated IG-SRS (log-ran k p = 0.030 an d 0.013, respect ively). Eigh teen lesion s in 17 p at ien t s w ere treated w ith salvage IG-SRS (i.e., radiation therapy w ith or w ith ou t su rgical m an agem en t for a previously t reated lesion ); 15 (83.3%) w ere t reated in th e postop erat ive set t ing. Overall su r vival an d local p rogression -free su r vival at 1 year w ere 42%(95%CI, 14.6–69.4%) an d 100%, respect ively. Th is coh or t in cluded both m etast at ic an d p rim ar y p at ien t s; for th e m et ast at ic su bset (n = 9 pat ien t s, 10 lesion s), overall su rvival an d local progression -free su r vival at 1 year w ere 34.6% (95% CI, 0–71.2%) an d 100%, resp ect ively; for pat ien ts w ith prim ar y lesion s (n = 8 pat ien ts, 8 lesion s), overall sur vival an d local progression -free su r vival at 1 year w ere 50%(95%CI, 9.2–90.8%) an d 100%, resp ect ively. Overall su r vival an d local p rogression -free su r vival w ere n ot sign i can tly di eren t (logran k p = 0.197 an d 1, resp ect ively) for p rim ar y versu s m et ast at ic lesion s, w h ich rein forces th e u t ilit y of SRS in th e t reat m en t of prim ar y sp in al sarcom a.

43

44

Chapter 4 Based on th is st u dy, hypofract ion ated an d single-fract ion IG-SRS p rovides excellen t local con t rol for p rim ar y/recu rren t an d m et ast at ic sarcom as of th e spin e. Single-fract ion IG-SRS seem s to be sign i can tly m ore e ect ive th an hyp ofract ion ated t reat m en t an d sh ou ld be a con sid erat ion for rst -lin e t h erapy in t h e m an agem en t of sarcom as of th e spin e; a p rospect ive t rial is n eeded to dem on st rate a com parable outcom e. DeLan ey et al53 repor t a series of 50 pat ien t s w ith sp in al sarcom as, of w h ich 86% (43/50) w ere eith er ch ordom a or ch on drosarcom a. Th e oth er seven pat ient s w ere single path ologies, including osteosarcom a, Ew ing’s sarcom a, giant cell sarcom a, an giosarcom a, lip osarcom a, an d m align an t p erip h eral n er ve sh eat h t u m or spin dle/roun d cell—further illust rating the challenges in di eren tiat ing th e m any sarcom atous h istologies. DeLan ey et al dem on st rated 5-year act u arial local con t rol of 78%. Th ey delivered radiat ion using a “sh rin king elds” tech n ique: 50.4 CGE (Gy RBE [relat ive biological e ect iven ess]) to su bclin ical disease, 70.2 Gy RBE to m icroscopic disease in the t um or bed, and 77.4 Gy RBE to gross disease at 1.8 Gy RBE qd. Th e sp in al cord d ose w as lim ited to 63/54 Gy RBE to surface/center. Intraoperative boost doses of 7.5 to 10 Gy cou ld be given by du ral plaqu e. Th is st udy fou n d ext rem ely h igh local con t rol in 34 of 36 (94%) p rim ar y t u m ors an d in 23 of 23 (100%) prim ar y ch ordom as. Th is rein forces th e im port an ce of com bin ed radiat ion an d su rgical th erapy at th e t im e of prim ar y p resen t at ion .54 At MSKCC, Yam ada et al rep or ted a series of 14 p at ien t s w it h p r im ar y sp in e/p arasp in al sarcom as t reated w it h m u lt ifract ion ated stereot act ic an d im age-gu id ed IMRT cou p led w ith n on invasive body fram es.24 Previously un irrad iated pat ien t s received a m edian prescribed dose of 70 Gy (range, 59.4–70 Gy) w ith a m ed ian p lan n ing t arget volu m e receiving th e prescribed dose of 90%. Th e m edian dose m a xim u m to th e cord w as 68% of th e p rescribed dose for previou sly u n irradiated p at ien t s. Of th e prim ar y lesion s, 81% exh ibited local con t rol w ith 2 to 30 m on th s of follow -u p. No cases of radiat ion -in duced m yelopathy w ere en cou n tered .

■ Chordoma Ch ordom a is a rare an d relen tless disease of th e sku ll, spin e, an d sacru m .55 Sim ilar to th e m ajorit y of m esen chym al t u m ors, it is n ot con sidered sen sitive to radiat ion t reat m en t . In t ralesion al or m argin al resect ion s in th e sp in e an d sacru m h ave a h igh probabilit y of recurren ce 3 ; th u s, en -bloc resect ion to ach ieve w ide m argin s h as been advocated as th e on ly cure for ch ordom as.1 Despite m ajor su rgical advan ces, tot al en -bloc resect ion m ay on ly be ach ieved in app roxim ately 50% of sacral ch ord om as, w ith m uch low er rates for lesions of the spine or skull base; therefore, recurrence is com m on w ithout en -bloc resect ion .56,57 In par t th is m ay be du e to t h e lack of e ect ive adju van t t h erap ies.55 Radiot h erapy h as often been u sed as adjuvan t as w ell as prim ar y t reat m en t for ch ordom as. Good control rates w ere achieved in som e series after en -bloc excision of prim ar y ch ordom as in com bin at ion w ith preop erat ive an d postop erat ive h igh -d ose radiat ion ,53 yet recurren ce rates in excess of 50% h ave been dem on st rated in oth er series w ith long-term follow -u p.58 Because the tolerance dose of the spinal cord, brain stem , cran ial n er ves, an d rect u m is m uch low er th an e ect ive doses to t reat ch ordom as, deliver y of h igh doses is lim ited. For exam ple, treatm ent of the sacrococcygeal region w ith high doses of p h oton radiat ion th erapy (45–80 Gy) can be ach ieved becau se th e region is less su scept ible t h an t h e cer vical sp in e, w h ere m yelopathy du e to radiat ion inju r y is com m on .59 Radiat ion doses > 70 Gy are requ ired for gross disease at conventional fractionation schem as.60 Treat m ent w ith conventional radiation therapy at doses of 40 to 60 Gy h as led to poor resu lts, w ith 5-year local con t rol of on ly 10 to 40%.61–64 Th e best repor ted local con t rol rate w ith ph oton s based on m od ern , fract ion ated stereotact ic radiat ion th erapy for lesion s in th e skull base is a 5-year act u arial rate of approxim ately 50%.58 Over th e years, radical en -bloc surgical resect ion an d radiat ion dose escalat ion h ave been fur th er explored to im prove sur vival. Pearlm an an d Friedm an 65 suggested a possible dose–response e ect, although it has been

Radiation Therapy for Primary Bone Tumors est ablish ed th at in creasing radiat ion doses in excess of 70 Gy w ill lead to long-term tum or con t rol.9,66 Im age-guided tech n ology coupled w ith in tensit y-m odulated ph oton radiat ion h as m ade it possible to deliver ver y h igh dose radioth erapy to the spine and sacrum in a single fraction. A series at MSKCC sh ow ed th at single-fract ion sp in e rad iosu rger y can p rovid e long-term local con t rol in greater t h an 90% of cases for m et ast at ic t u m ors con sid ered resist an t to conven t ion al radiat ion .48 Single-session h igh dose radiosu rger y for sku ll base ch ordom as h as sh ow n prom ise.34,67 Recen tly, Yam ada et al55 review ed th eir exp er ien ce w it h sin gle-fract ion h igh -d ose SRS for ch ord om a of t h e sp in e an d sacr u m , an d d em on st rated good t u m or con t rol w ith low t reat m en t-related m orbidit y. Th is series rep or ted 24 pat ien t s w ith ch ordom a of th e sacru m an d m obile sp in e after being t reated w ith h igh -d ose single-fract ion SRS (m ed ian d ose, 24 Gy). Tw en t y-on e prim ar y an d th ree m etastatic tum ors were treated. Seven patients were t reated for postoperat ive t um or recurren ce. In seven patients SRS was adm inistered as planned adjuvan t th erapy, an d in 13 pat ien ts SRS w as adm in istered as neoadjuvant therapy. Although su rger y w as p lan n ed in m ost cases w h ere single-fract ion SRS w as delivered as th e in it ial t reat m en t , th e lack of sym ptom at ic an d radiograph ic progression after SRS resulted in several p at ien t s refu sin g su rger y. Th u s, on ly six of 13 pat ien t s w h o un der w en t n eoadjuvan t SRS proceeded to su rger y. Th is coh or t p rovides in for m at ion abou t local t u m or con t rol w it h sin gle-fract ion SRS as t h e sole t reat m en t of n ew ly diagn osed ch ord om as. Th e overall m ed ian follow -u p w as 24 m on t h s. Of t h e 24 p atients, 23 (95%) dem onstrated stable or reduced t u m or burden based on serial MRI. On e pat ien t h ad rad iograp h ic p rogression of t u m or 11 m on t h s after SRS. Com plicat ion s w ere lim ited to on e pat ien t w h o developed sciat ic n eu rop athy an d on e w ith vocal cord paralysis. Th e m ost opt im ist ic results in th e t reat m en t of ch ordom as h ave been rep or ted by p roton th erapy for th e skull base, in gen eral com bin ing m axim al surgical t um or resect ion w ith ad-

juvan t h igh -d ose par t icle th erapy. Th e prim ar y advan tage of a proton beam is th e proton s’ physical feat u re of th e Bragg peak, w h ich provides excellen t con form it y of th e irradiat ion eld. Ut ilizat ion of th is m odalit y h as been pion eered at Massach u set ts Gen eral Hospital in collaborat ion w ith th e Har vard Cyclot ron Lab orator y. A tech n iqu e called spot scan n ing w as develop ed at th e Pau l Sch errer In st it u te an d rep or ted on 26 pat ien ts w ith ext racran ial ch ordom as an d a m edian follow -u p p eriod of 35 m on th s after p roton th erapy. In th is st udy, a 3-year disease-free su r vival of 77% w as docu m en ted .68 Most of th e dat a on ch ordom as are t aken from sku ll bases rep or ts. A review of th is literature dem onstrates local con trol rates ranging from 67% at 3 years to 98% at 5 years w ith d oses ranging from 50 to 83 Gy RBE.69–73 Becau se ch ordom as are rare an d becau se it is di cult to ran dom ize pat ien ts to t reat m en t s other th an st an dard of care, it is u n likely th at ph ase 3 t rials w ill be don e to com p are di eren t t ypes of radiat ion , m aking th e di eren ces in clin ical e ect di cu lt to in terpret .16 Th e u se of radioth erapy as prim ar y or adjuvan t t reat m en t for ch ordom a is st ill debated. In th e past , stan dalon e radioth erapy w as proven to be in e ect ive w h en cou p led w ith debu lking or p alliat ive th erapy.74 It is clear th at w ith th e adjun ct of single-fract ion SRS an d m od ern proton beam schem as, th e role of adjuvant, neoadjuvant, an d stand-alone radiotherapy w ill continue to evolve in th e m anagem en t of spin al ch ordom as.

■ Chondrosarcoma Ch on drosarcom a is th e th ird m ost com m on t ype of prim ar y bon e t u m or (after m yelom a an d osteosarcom a) an d accou n ts for 20 to 27% of all prim ar y m align an t osseous t um ors.75 Typically th ese lesion s are low grade an d can arise de n ovo or from a preexist ing car t ilage lesion su ch as an osteoch on drom a or en ch on drom a. Less th an 10% of all ch on d rosarcom as occu r in t h e sp in e. Du e to t h eir biology, low p ercen t age of d ivid in g cells, an d p oor vascu lar it y, t h ese t u m ors h ave a ten d en cy to be

45

46

Chapter 4 ch em o - an d radioth erapy resist an t . How ever, ch on drosarcom as grow slow ly an d rarely m etastasize.76 Surgical resection rem ains the m ainst ay of t reat m en t of ch on d rosarcom as. Th e exten t of surgical resect ion an d adjuvan t th erapy is dependent on the clinical an d histological ch aracterist ics of th e lesion s. Alth ough w ide, en -bloc excision is ideal for in term ediate- an d h igh -grad e ch on d rosarcom a, less aggressive ap p roach es m ay be accept able for low -grad e ch on drosarcom a. Due to their sim ilar m anagem en t challenges, ch ordom a an d ch on drosarcom a t reat m en t ser ies are often com bin ed . As w it h ch ord om a m an agem en t , radiat ion of ch on drosarcom a is con sidered after in com plete resect ion , aim ing at m axim al local con t rol (cu rat ive) an d in sit uat ion s w h ere resect ion is n ot feasible or w ou ld cause un accept able m orbidit y (palliat ive). For cu rat ive in ten t ion s, doses > 60 Gy are required to ach ieve local con t rol.76 Du e to th e lim itat ion s of conven t ion al radioth erapy, p roton an d ph oton radioth erapy in ch on drosarcom a are ut ilized. As w ith ch ord om as, m ost of th e dat a on ch on drosarcom a are taken from skull base rep or ts. Th e literat ure dem on st rates local con t rol rates ranging from 94% at 3 years to 73% at 5 years w ith doses ranging from 50 to 83 Gy RBE.69–73 Sim ilar to ch ord om a t reat m en t , t h e im p or t an ce of an escalated d ose is em p h asized . Au st in et al6 repor ted 26 local relapses in 141 pat ien ts t reated by p roton irradiat ion for ch ordom a or ch on drosarcom a of th e cran ial base or cer vical spin e. Am ong 26 relapses, on ly six w ere rep or ted in an area t h at h ad received 70 Gy or m ore versu s 15 relap ses in areas receiving a low er dose. Often th e sh ielding of crit ical an atom ic st ru ct u res adjacen t to t u m or bou n daries is th e cau se of u n derdosing an d local failu re.77 Th e sku ll base literat u re rein forces th e relat ion sh ip bet w een dose in h om ogen eit y an d risk of local failu re.78 Noël et al77 dem on st rated 3-year local con t rol rates of 69% in ch ordom a an d 91%in ch on drosarcom as. Their results em ph asize th e im p or tan ce of adequ ate coverage of t h e t arget volu m e at an “e ect ive” d ose. An oth er im por tan t n ding by Noël et al is th at th e m in im u m dose correlates to local con t rol, a variable th at is determ in ed by dose con st rain ts

of n orm al st r u ct u res. St ill m ost au t h ors of rep ort s in th e sku ll base literat ure believe th at su rger y is t h e m ain st ay of t reat m en t for th is pathology.79,80 Published data dem onstrate that local con t rol is depen den t on th e size of th e residual t um or at th e t im e of radiat ion .70 Rad iot h erapy w it h carbon ion s or ot h er charged particles is anoth er at tractive radiat ion m odalit y. Th e physical advan tages of proton s are com bin ed w ith a h igh er radiobiological act ivit y. Sch u lz-Er t n er et al81 reported th e e ect iven ess an d toxicit y of carbon ion radioth erapy in ch on drosarcom as of th e sku ll base.

■ Chapter Summary From available dat a, en -bloc resect ion ach ieving n egat ive h istological m argin s m ay h ave a good rate of local t um or con t rol. For pat ien ts w h o are n ot can didates for m argin al or w ide resect ion s, adjuvan t radiat ion adds a m easure of t um or con t rol. Proton beam , in t raoperat ive radioact ive im p lan t s, an d h igh -dose con form al ph oton th erapy all p lay a role in local t u m or con t rol an d possibly cu re. Advan ces in stereotact ic plan n ing are overcom ing m any of th e t issu e toleran ce issu es w ith h igh -dose ph oton radiosurger y. Th e success of p roton th erapy dep en ds on dosim et ric factors, m ain ly th ose related to dose h om ogen eit y w ith in th e t arget , rath er th an on th e p rescribed dose. How ever, high -qualit y surgical resection rem ains critical, because it can dram at ically im prove an atom ic relations bet ween tum or and adjacent anatom ic st r u ct ures, w h ich is a p rogn ost ic factor of dose h om ogen eit y.77 Pearls Understand the importance of multidisciplinary  managem ent for optim al treatm ent of prim ary spine tumors. Image-guided radiotherapy, including stereotactic radiosurgery, provides a level of precision of  treatm ent delivery that enables m uch higher  doses of radiation to be given safely to prim ary  spine tumors. Hypofractionation is a strategy to increase the  biological e ectiveness of radiation and may be

Radiation Therapy for Primary Bone Tumors    an e ective treatm ent strategy for prim ary spine  tumors. Intraoperative dural brachytherapy is a method  of delivering high-dose radiation to the dural  margin without signi cantly increasing the dose  of radiation received by the spinal cord. Proton beam radiation and other charged particle  radiation will continue to play an important role in  the management of primary spine tumors.

Pitfall Patients should be referred to experienced centers for evaluation where m ultidisciplinary assessment is available and su ciently high-dose  radiation therapy can be provided.

Refere nces Five Must-Read Refe rences 1. Borian i S, Ban diera S, Biagin i R, et al. Ch ordom a of th e m obile spine: ft y years of experien ce. Spin e 2006; 31:493–503 PubMed 2. Tom it a K, Kaw ah ara N, Baba H, Tsuch iya H, Fujit a T, Toribat ake Y. Tot al en bloc spon dylectom y. A n ew su rgical tech n ique for prim ar y m align an t ver tebral t um ors. Spin e 1997;22:324–333 Pu bMed 3. Yao KC, Borian i S, Gokaslan ZL, Su n daresan N. En bloc spon dylectom y for sp in al m et ast ases: a review of tech n iques. Neurosurg Focus 2003;15:E6 PubMed 4. Fu ller DB, Bloom JG. Radioth erapy for ch ordom a. In t J Radiat On col Biol Phys 1988;15:331–339 Pu bMed 5. Kroch ak R, Har w ood AR, Cu m m ings BJ, Qu ir t IC. Results of radical radiation for chondrosarcom a of bone. Radioth er On col 1983;1:109–115 PubMed 6. Aust in JP, Urie MM, Card en osa G, Mu n zen rider JE. Probable causes of recurren ce in pat ien t s w ith ch ordom a an d ch on drosarcom a of th e base of skull an d cer vical spin e. In t J Radiat On col Biol Phys 1993;25: 439–444 Pu bMed 7. Bell RS, O’Sullivan B, Liu FF, et al. Th e surgical m argin in soft t issue sarcom a. Ch ir Organi Mov 1990;75(1, Suppl)126–130 PubMed 8. Pisters PW, Leung DH, Woodr u J, Sh i W, Bren n an MF. An alysis of p rogn ost ic factors in 1,041 pat ien t s w ith localized soft t issue sarcom as of th e ext rem it ies. J Clin On col 1996;14:1679–1689 Pu bMed 9. Tan abe KK, Pollock RE, Ellis LM, Murphy A, Sh erm an N, Rom sdah l MM. In u en ce of su rgical m argin s on ou tcom e in pat ien t s w ith preoperat ively irradiated ext rem it y soft t issue sarcom as. Can cer 1994;73: 1652–1659 Pu bMed 10. Borian i S, De Iure F, Ban diera S, et al. Ch on drosarcom a of th e m obile spine: repor t on 22 cases. Spine 2000;25:804–812 PubMed 11. Borian i S, Wein stein JN, Biagin i R. Prim ar y bon e t um ors of th e spin e. Term in ology an d surgical st aging. Spin e 1997;22:1036–1044 Pu bMed 12. Bilsky MH, Bolan d PJ, Pan ageas KS, Woodr u JM, Bren n an MF, Healey JH. In t ralesion al resect ion of prim ar y an d m et ast at ic sarcom a involving th e spin e:

outcom e an alysis of 59 pat ien t s. Neurosurger y 2001; 49:1277–1286, discussion 1286–1287 Pu bMed 13. O’Su llivan B, Davis AM, Turcot te R, et al. Preop erat ive versu s postoperat ive radioth erapy in soft-t issu e sarcom a of th e lim bs: a ran dom ised t rial. Lan cet 2002; 359:2235–2241 PubMed 14. DeLan ey TF, Tro m ov AV, Engelsm an M, Suit HD. Advan ced-tech n ology radiat ion th erapy in th e m an agem en t of bon e and soft t issue sarcom as. Can cer Con t r 2005;12:27–35 PubMed 15. Em am i B, Lym an J, Brow n A, et al. Toleran ce of n orm al t issu e to therapeut ic irradiat ion . In t J Rad iat Oncol Biol Phys 1991;21:109–122 PubMed 16. Walcot t BP, Nah ed BV, Mohyeldin A, Coum an s JV, Kah le KT, Ferreira MJ. Ch ordom a: cu rren t con cept s, m an agem en t , an d fut ure direct ion s. Lan cet On col 2012;13:e69– e76 PubMed 17. Pan ah an deh H, Spadea M, Oh K, Seco J. Dosim et ric an alysis of proton passive-scat tering stereot act ic body radioth erapy (SBRT) of t reated spin e lesion s versus ph oton SBRT. Med Phys 2011;38:3694 18. Verbu rg JM, Seco J. Dosim et ric accuracy of proton th erapy for ch ordom a pat ien t s w ith t it an iu m im plan t s. Med Phys 2013;40:071727 PubMed 19. Tobias CA, Blakely EA, Alpen EL, et al. Molecu lar an d cellu lar radiobiology of heavy ions. In t J Radiat On col Biol Phys 1982;8:2109–2120 Pu bMed 20. Im ai R, Kam ada T, Sugah ara S, Tsuji H, Tsujii H. Carbon ion radioth erapy for sacral ch ordom a. Br J Radiol 2011;84(Spec No 1):S48–S54 PubMed 21. Moh an R, Bortfeld T. Proton th erapy: clin ical gain s th rough cu rren t an d fu t u re t reat m en t p rogram s. Fron t Radiat Th er On col 2011;43:440–464 Pu bMed 22. Au st in -Seym ou r M, Mu n zen r id er JE, Goitein M, et al. Progress in low -LET h eavy p ar t icle t h erapy: in t racran ial an d p aracran ial t u m ors an d u veal m elan om as. Rad iat Res Su p p l 1985;8(Su p p l):S219–S226 Pu bMed 23. Hug EB. Review of skull base ch ordom as: progn ost ic factors an d long-term resu lt s of p roton -beam radioth erapy. Neurosurg Focus 2001;10:E11 PubMed

47

48

Chapter 4 24. Yam ada Y, Lovelock DM, Yen ice KM, et al. Mu lt ifract ion ated im age-guided an d stereot act ic in tensit ym od u lated rad iot h erapy of p arasp in al t u m ors: a p relim in ar y report . In t J Radiat On col Biol Phys 2005; 62:53–61 Pu bMed 25. Terezakis SA, Lovelock DM, Bilsky MH, Hun t MA, Zatcky J, Yam ada Y. Im age-guided intensit y-m odulated photon radiotherapy using m ultifractionated regim en to parasp in al ch ordom as an d rare sarcom as. In t J Radiat On col Biol Phys 2007;69:1502–1508 Pu bMed 26. Ch eng EY, Ozerdem oglu RA, Tran sfeldt EE, Th om pson RC Jr. Lu m bosacral ch ordom a. Progn ost ic factors an d t reat m en t . Spin e 1999;24:1639–1645 Pu bMed 27. Rogers CL, Th eodore N, Dickm an CA, et al. Surger y an d perm an en t 125I seed paraspin al brachytherapy for m align an t t u m ors w ith sp in al cord com p ression . Int J Radiat Oncol Biol Phys 2002;54:505–513 PubMed 28. Mu rakam i H, Kaw ah ara N, Yah at a T, Yokoyam a K, Kom ai K, Tom ita K. Radiat ion m yelopathy after radioact ive iodin e th erapy for spin e m et ast asis. Br J Radiol 2006;79:e45– e49 Pu bMed 29. DeLan ey TF, Ch en GT, Mau ceri TC, et al. In t raop erat ive dural irradiat ion by custom ized 192 iridiu m an d 90 yt t riu m brachyth erapy p laqu es. In t J Radiat On col Biol Phys 2003;57:239–245 PubMed 30. Garcia-Barros M, Paris F, Cordon -Cardo C, et al. Tum or respon se to radioth erapy regulated by en doth elial cell apoptosis. Scien ce 2003;300:1155–1159 PubMed 31. Mem orial Sloan Ket tering Cancer Center. h t tp://w w w .m skcc.org/cancer-care/adult/spine-t um ors/radiat ion -th erapy. Accessed October 1, 2013 32. Hall EJ, Giaccia AJ. Radiobiology for th e Radiologist , 6th ed. Ph iladelphia: Lippin cot t William s & Wilkin s, 2006 33. Flickinger JC, Kondziolka D, Lunsford LD. Radiobiological an alysis of tissue respon ses follow ing radiosurgery. Tech nol Cancer Res Treat 2003;2:87–92 PubMed 34. Mar t in JJ, Niranjan A, Kon dziolka D, Flickinger JC, Lozan n e KA, Lu n sford LD. Radiosurger y for ch ordom as an d ch on drosarcom as of th e sku ll base. J Neu rosu rg 2007;107:758–764 Pu bMed 35. Gerszten PC, Bu r ton SA, Ozh asoglu C, et al. Stereo t act ic rad iosu rger y for sp in al m et ast ases from ren al cell carcin om a. J Neu rosu rg Sp in e 2005;3:288–295 Pu bMed 36. Lovelock DM, Hua C, Wang P, et al. Accu rate set u p of parasp in al p at ien t s u sing a n on invasive pat ien t im m obilizat ion cradle an d por t al im aging. Med Phys 2005;32:2606–2614 Pu bMed 37. Cox BW, Sprat t DE, Lovelock M, et al. In tern at ion al Spine Radiosurgery Consortium consensus guidelines for t arget volu m e de n it ion in sp in al stereot act ic rad iosu rger y. In t J Rad iat On col Biol Phys 2012;83: e597–e605 PubMed 38. Folker t MR, Bilsky MH, Cox BW, et al. Ou tcom es for hypofract ion ated an d single-fract ion im age-gu ided radiat ion th erapy for prim ar y and m et ast at ic sarco-

m as of th e spin e. Con n ect ive Tissue Oncology Societ y 17th An n u al Meet ing, Pragu e, Novem ber 17, 2012 39. Siegel R, Naish adh am D, Jem al A. Can cer st at ist ics, 2012. CA Can cer J Clin 2012;62:10–29 40. Gerszten PC, Men del E, Yam ada Y. Radioth erapy an d radiosurger y for m et ast at ic spin e disease: w hat are th e opt ion s, in dicat ion s, an d outcom es? Spin e 2009; 34(22, Su p pl):S78–S92 PubMed 41. Ch ao ST, Koyfm an SA, Woody N, et al. Recursive part it ion ing an alysis index is predict ive for overall survival in pat ien t s u n dergoing sp in e stereot act ic body radiat ion th erapy for spin al m et ast ases. In t J Radiat On col Biol Phys 2012;82:1738–1743 Pu bMed 42. Bilsky MH, Gerszten P, Laufer I, Yam ada Y. Radiat ion for p r im ar y sp in e t u m ors. Neu rosu rg Clin N Am 2008;19:119–123 PubMed 43. Patch ell RA, Tibbs PA, Regine W F, et al. Direct decom pressive surgical resect ion in th e t reat m en t of spin al cord com pression caused by m et ast at ic can cer: a random ised trial. Lancet 2005;366:643–648 PubMed 44. Hug EB, Fit zek MM, Liebsch NJ, Mun zen rider JE. Locally ch allen ging osteo- an d ch on drogen ic t u m ors of th e axial skeleton : result s of com bin ed proton an d p h oton rad iat ion t h erapy u sin g t h ree-d im en sion al t reat m en t p lan n ing. In t J Rad iat On col Biol Phys 1995;31:467–476 Pu bMed 45. Mer im sky O, Kollen d er Y, Bokstein F, et al. Rad io t h erapy for spinal cord com pression in pat ien t s w ith soft-tissue sarcom a. Int J Radiat Oncol Biol Phys 2004; 58:1468–1473 PubMed 46. Boriani S, Saravanja D, Yam ada Y, Varga PP, Biagin i R, Fisher CG. Ch allenges of local recurren ce and cure in low grade m align an t t um ors of th e spin e. Spin e 2009;34(22, Su pp l):S48–S57 Pu bMed 47. Bilsky MH, Yam ada Y, Yen ice KM, et al. Inten sit ym odulated stereot act ic radioth erapy of paraspin al t um ors: a prelim in ar y repor t . Neu rosurger y 2004; 54:823–830, discu ssion 830–831 Pu bMed 48. Yam ada Y, Bilsky MH, Lovelock DM, et al. High -dose, single-fract ion im age-guided in ten sit y-m odulated radioth erapy for m et ast at ic spin al lesion s. In t J Radiat On col Biol Phys 2008;71:484–490 Pu bMed 49. Levin e AM, Colem an C, Horasek S. Stereot act ic radiosu rger y for th e t reat m en t of prim ar y sarcom as an d sarcom a m etastases of the spine. Neurosurger y 2009; 64(2, Suppl):A54–A59 PubMed 50. Rock JP, Ryu S, Sh ukair y MS, et al. Postoperat ive radiosurger y for m align an t spin al t um ors. Neurosurger y 2006;58:891–898, discu ssion 891–898 Pu bMed 51. Moulding HD, Elder JB, Lis E, et al. Local disease con t rol after decom pressive surger y an d adjuvan t h igh dose single-fract ion radiosurger y for spine m etastases. J Neu rosu rg Spin e 2010;13:87–93 Pu bMed 52. Sah gal A, Bilsky M, Ch ang EL, et al. Stereot act ic body radioth erapy for spin al m et ast ases: cu rren t st at us, w ith a focus on it s applicat ion in th e postoperat ive patient. J Neurosurg Spine 2011;14:151–166 PubMed

Radiation Therapy for Primary Bone Tumors 53. DeLan ey TF, Liebsch NJ, Pedlow FX, et al. Ph ase II study of high-dose photon/proton radiotherapy in the m an agem en t of spin e sarcom as. Int J Radiat On col Biol Phys 2009;74:732–739 PubMed 54. Ch ang UK, Ch o W I, Lee DH, et al. Stereot act ic radiosu rger y for prim ar y an d m et ast at ic sarcom as involving the spine. J Neurooncol 2012;107:551–557 PubMed 55. Yam ada Y, Laufer I, Cox BW, et al. Prelim in ar y resu lt s of h igh -dose single-fract ion radioth erapy for the m an agem en t of chordom as of th e spin e an d sacrum . Neurosurgery 2013;73:673–680, discussion 680 [Epub ah ead of p rin t] Pu bMed 56. Tzort zidis F, Elah i F, Wrigh t D, Nat arajan SK, Sekh ar LN. Pat ien t ou tcom e at long-term follow -u p after aggressive m icrosurgical resect ion of cran ial base ch ordom as. Neu rosu rger y 2006;59:230–237, discu ssion 230–237 Pu bMed 57. Fuchs B, Dickey ID, Yaszem ski MJ, Inw ards CY, Sim FH. Operative m anagem ent of sacral chordom a. J Bone Join t Surg Am 2005;87:2211–2216 Pu bMed 58. Debus J, Sch ulz-Er t n er D, Sch ad L, et al. Stereot act ic fract ion ated radioth erapy for ch ordom as an d chon drosarcom as of th e sku ll base. In t J Rad iat On col Biol Phys 2000;47(3):591–596 Pu bMed 59. Casali PG, St acch iot t i S, Sangalli C, Olm i P, Gron ch i A. Chordom a. Curr Opin Oncol 2007;19:367–370 PubMed 60. St aab A, Ru t z HP, Ares C, et al. Spot-scan n ing-based proton therapy for ext racran ial ch ordom a. In t J Radiat On col Biol Phys 2011;81:e489–e496 PubMed 61. Forsyth PA, Cascino TL, Sh aw EG, et al. In t racran ial chordom as: a clinicopathological and prognostic study of 51 cases. J Neu rosu rg 1993;78:741–747 Pu bMed 62. Cat ton C, O’Sullivan B, Bell R, et al. Ch ordom a: longterm follow -up after radical ph oton irradiat ion . Radioth er Oncol 1996;41:67–72 PubMed 63. Cu m m ings BJ, Hodson DI, Bu sh RS. Ch ordom a: th e result s of m egavolt age radiat ion th erapy. In t J Rad iat On col Biol Phys 1983;9:633–642 PubMed 64. Rom ero J, Card en es H, la Torre A, et al. Ch ordom a: resu lt s of rad iat ion t h erapy in eigh teen p at ien t s. Radioth er On col 1993;29:27–32 PubMed 65. Pearlm an AW, Friedm an M. Radical rad iat ion th erapy of ch ordom a. Am J Roen tgen ol Radiu m Th er Nucl Med 1970;108:332–341 Pu bMed 66. Tai PT, Craigh ead P, Bagdon F. Opt im izat ion of radioth erapy for p at ien t s w ith cran ial ch ordom a. A review of dose-respon se rat ios for ph oton tech n iqu es. Can cer 1995;75:749–756 PubMed 67. Krish n an S, Foote RL, Brow n PD, Pollock BE, Lin k MJ, Garces YI. Radiosu rger y for cran ial base ch ordom as an d ch on drosarcom as. Neu rosurger y 2005;56:777– 784, discu ssion 777–784 PubMed

68. Rut z HP, Weber DC, Sugah ara S, et al. Ext racran ial chordom a: outcom e in patients treated w ith functionpreser ving surger y followed by spot-scanning proton beam irradiat ion . In t J Radiat On col Biol Phys 2007; 67:512–520 PubMed 69. Fuji H, Nakasu Y, Ish ida Y, et al. Feasibilit y of proton beam th erapy for ch ordom a an d ch on drosarcom a of th e sku ll base. Skull Base 2011;21:201–206 Pu bMed 70. Noël G, Habran d JL, Mam m ar H, et al. Com bin at ion of ph oton an d proton radiat ion th erapy for ch ordom as an d ch on drosarcom as of th e skull base: th e Cen t re de Proton th érapie D’Orsay experien ce. In t J Radiat On col Biol Phys 2001;51:392–398 PubMed 71. Mu n zen r ider JE, Liebsch NJ. Proton t h erapy for t u m ors of t h e sku ll base. St rah len t h er On kol 1999; 175(Su ppl 2):57–63 PubMed 72. Au st in -Seym ou r M, Mu n zen rid er J, Goitein M, et al. Fract ion ated proton radiat ion th erapy of ch ordom a an d low -grade ch on d rosarcom a of th e base of th e skull. J Neu rosu rg 1989;70:13–17 PubMed 73. Hug EB, Loredo LN, Slater JD, et al. Proton radiat ion th erapy for ch ord om as an d ch on drosarcom as of th e sku ll base. J Neu rosu rg 1999;91:432–439 Pu bMed 74. Borian i S, Chevalley F, Weinstein JN, et al. Ch ord om a of th e sp in e above th e sacr u m . Treat m en t an d ou t com e in 21 cases. Spin e 1996;21:1569–1577 Pu bMed 75. Björn sson J, McLeod RA, Un n i KK, Ilst rup DM, Pritchard DJ. Prim ary chondrosarcom a of long bones and lim b girdles. Cancer 1998;83:2105–2119 PubMed 76. Gelderblom H, Hogendoorn PC, Dijkst ra SD, et al. Th e clin ical approach tow ards ch on drosarcom a. On cologist 2008;13:320–329 PubMed 77. Noël G, Feuvret L, Ferran d R, Boisserie G, Mazeron JJ, Habran d JL. Radioth erap eu t ic factors in th e m an agem en t of cer vical-basal ch ordom as an d chondrosarcom as. Neurosurger y 2004;55:1252–1260, discussion 1260–1262 PubMed 78. Terah ara A, Niem ierko A, Goitein M, et al. An alysis of th e relat ion ship bet w een t um or dose inh om ogen eit y an d local con t rol in pat ien t s w ith skull base ch ordom a. In t J Radiat On col Biol Phys 1999;45:351–358 PubMed 79. al-Meft y O, Borba LA. Sku ll base ch ordom as: a m an agem en t ch allenge. J Neurosurg 1997;86:182–189 PubMed 80. Sen CN, Sekh ar LN, Sch ram m VL, Jan ecka IP. Ch ordom a an d ch on drosarcom a of th e cran ial base: an 8-year experien ce. Neurosurger y 1989;25:931–940, discu ssion 940–941 PubMed 81. Sch u lz-Er t n er D, Nikogh osyan A, Hof H, et al. Carbon ion radioth erapy of skull base ch on drosarcom as. In t J Radiat On col Biol Phys 2007;67:171–177 Pu bMed

49

5 Medical Oncology Principles for the Spine Oncology Surgeon Scott H. Okuno, Steven I. Robinson, and Shreyaskumar Patel

■ Introduction Prim ar y t u m ors of th e sp in e are rare an d requ ire a m u lt idiscip lin ar y ap p roach to provide th e p at ien t w ith th e best ou tcom e. Th e m em bers of th e team bring th eir exper t ise an d th e u n d erlyin g p r in cip les of t h eir d iscip lin e. Med ical on cologist s h ave a solid u n d erst an d ing of t h e n at u ral h istor y an d available system ic opt ion s for each t u m or. Addit ion ally, th ey are often able to develop a long-stan ding relat ion sh ip w ith th e p at ien t an d th e fam ily, th us providing th e m u lt id iscip lin ar y team w ith in sigh t on th e goals an d expect at ion s of th e pat ien t , as w ell as th eir toleran ce for t reat m en t. Th is ch apter d iscu sses ve p r in cip les of m ed ical on cology th at w e con sider key for all spin e on cology su rgeon s w h en caring for pat ien t s w ith p rim ar y an d m etast at ic t u m ors of th e spin e: 1. 2. 3. 4. 5.

Tissu e is th e issu e St aging Th e role of adjuvan t th erapy Un derstan ding th e toxicit y of th erapy Understanding the preferences of the patient

Th ese p rin ciples can be ap p lied to th e p at ien t w ith localized or m etastat ic disease. Th is ch ap ter also d iscu sses com m on p at ien t scen ar ios th at illu st rate an d rein force t h ese p r in cip les. We d o n ot cover each of t h e t u m ors in d ept h , as th ey are discussed in oth er ch apters in th is m asters series.

It is vital, w h en dealing w ith a pat ien t w ith can cer, th at w e h ave th e correct diagn osis. Ob tain ing appropriate t issue facilitates n ot on ly m aking th e diagn osis bu t also perform ing addit ion al test ing on th e t u m or sam ple to h elp determ in e p rogn osis an d t reat m en t opt ion s. Advan ces in im aging, w ith presen t-day op t ion s in clu d ing m agn et ic reson an ce im aging (MRI), com p u ted tom ograp hy (CT), an d p osit ron em ission tom ograp hy (PET) scan s, h ave en abled m ore accu rate st aging. Un derst an ding both disease biology and im aging m odalit y lim it at ion s is essen t ial in guiding th e ch oice of th e appropriate im aging st udies to properly stage pat ien ts. Adjuvan t or n eoadjuvan t system ic th erapy has im proved the outcom e of patients w ith prim ar y bon e t u m ors su ch as Ew in g’s sarcom a an d osteogen ic sarcom a. Th e role of system ic th erapy in th e adjuvan t set t ing w ith oth er prim ar y sp in e t u m ors is less clear. How ever, th ere are circum stan ces w h ere th e use of adjuvan t system ic opt ions m ay positively im pact th e surgical app roach . It is di cult to keep abreast of all th e evolving th erapeu t ic opt ion s an d th eir associated side e ect s. Th ese side e ect s, in p ar t icu lar, can h ave im pact on th e su rger y an d th e post operat ive course. Th e m edical on cologist can be a valuable resource for in form at ion about th ese issu es. Ult im ately, ach ievin g t h e p at ien t ’s goals is cen t ral to t h e care p lan . As caregivers of p a-

Medical Oncology Principles for the Spine Oncology Surgeon t ien t s w ith both cu rable an d in cu rable can cers, w e n eed to p rovid e ou r p at ien t s w it h in form at ion abou t t h e p oten t ial ben e t s an d sid e e ect s of any t reat m en t w e recom m en d, allow ing th em to be an act ive part icipan t in th e decision -m aking process.

■ Tissue Is the Issue Th is rst p rin ciple of m edical on cology can n ot be un derest im ated. Not on ly is t issue n eeded to m ake th e diagn osis, but h ow w e obt ain th e t issu e an d w h at addit ion al test ing can be don e on th at specim en h as revolut ion ized th e w ay w e t reat pat ien t s w ith can cer. In th e past m ost of th e biopsies th at w ere n eeded to m ake th e diagn osis w ere don e w ith an op en ap proach (open biopsy). The open procedure enabled the su rgeon to obt ain an adequ ate t issu e sam ple to m ake th e d iagn osis an d do any requ isite an cillar y test ing. Addit ion ally, th e su rgeon w as able to con t rol any bleed ing t h at m igh t h ave occu r red d u r in g th e biop sy p roced u re. Th e op en biop sy often requ ired t h e u se of an esth esia an d sut ures. Th is added to th e expen se an d recover y t im e. With th e advan cem en t of crosssectional im aging, interventional radiologists are n ow able to perform percutan eou s CT-gu ided biopsies on alm ost any t issue in any locat ion . Th ese t issu e cores are for th e m ost p ar t en ough to obtain a diagn osis an d appropriate an cillar y test ing. Th is h as redu ced both recover y t im e an d cost . Th e diagn ost ic accuracy for spin e CTguided biopsies can be h as h igh as 97% w ith a low com p licat ion rate (see text box).1–3

p er for m ad d it ion al an cillar y test ing to h elp d eterm in e progn osis an d th erapeu t ic opt ion s. Alth ough th ese addit ion al test s at presen t h ave lim ited valu e for p r im ar y bon e t u m ors, t h ey d o h ave sign i can t im pact for n onprim ar y bon e t um ors for w h ich spin e on cologist are frequ en tly con su lted. We h igh ligh t t w o com m on t u m ors, lu ng can cer an d breast can cer, for w h ich addit ion al an cillar y test ing is im por tan t in th e m an agem en t of th e pat ien t (Figs. 5.1 an d 5.2). In th e past , pat ien ts diagn osed w ith lu ng can cer w ere eith er classi ed as sm all cell lu ng carcin om a (SCLC) or n on –sm all cell lu ng carcinom a (NSCLC). Th e NSCLC w as further classi ed as squ am ous cell carcin om a, aden ocarcin om a, large cell, or n ot oth er w ise sp eci ed (NOS). Th e lu m ping of all NSCLCs togeth er despite th e differen t h istologies w as largely d u e to t h e fact th at ou r th erap ies w ere n ot th at sp eci c for t h e di eren t h istologies. With th e in crease in th erap eu t ic opt ion s an d im p roved su bclassi cat ion of aden ocarcin om as, w e n ow kn ow th at all NSCLCs are n ot th e sam e an d th eir t reat-

Percutaneous CT–Guided Biopsy of the Spine : Accuracy in Diagnosis and Complication Rates Accuracy ◦ 71–97% Complications ◦ 0–21% ◦ Pulmonary, neurologic, infectious, and bleed

Not on ly is t issu e t h e issu e to m ake t h e d iagn osis, bu t it also p rovid es t u m or sam p le to

Fig. 5.1 Simple lung cancer classi cation.

51

52

Chapter 5

Fig. 5.2 Simple breast cancer classi cation.

m en t an d p rogn osis often di er. For in st an ce, squ am ou s cell carcin om a, w h ich accou n t s for 20% of NSCLCs, occurs m ain ly in sm okers, does n ot h ave m utat ion s in th e epiderm al grow th factor receptor, an d is less respon sive to th e chem otherapy agen t pem etrexed. W h en treated w ith a plat in u m -based doublet , pat ien t s w ith squ am ou s h istology w h o received cisp lat in an d pem et rexed h ad a 3-m on th decrease in th eir m edian overall su r vival com p ared to n on – squam ous cell h istology, bu t h ad an alm ost 2-m on th im proved m edian sur vival w h en th ey

received cisp lat in in com bin at ion w it h gem cit abin e.4,5 Th us t reat m en t decision s can be bet ter t ailored for pat ien t s w ith squ am ous cell carcin om a. For pat ien ts w ith aden ocarcin om a NSCLC in addit ion to th e h istology, w e n ow n eed to kn ow th e m olecu lar ch aracterizat ion of th e aden ocarcin om a to bet ter t ailor t reat m en t an d p rovide prognosis (Table 5.1). The tissue sam ples of pat ients w ith adenocarcin om a NSCLC routinely un dergo gen ot yp ing for ep iderm al grow th factor receptor (EGFR) m ut at ion , ALK (an aplast ic

Table 5.1 Adenocarcinoma Non–Small Cell Lung Carcinoma (NSCLC): Subclassi cation by Molecular Testing and Implications Adenocarcinoma NSCLC

Subclassi cation by Molecular Testing (%Incidence)

EGFR m utation ALK translocation RAS mutation

15 4 20–25

Abbreviation: EGFR, epidermal growth factor receptor.

Prognosis

Targeted Therapy

Favorable Favorable In m etastatic disease m ay help predict response or improve response to certain chemo therapy agents

Erlotinib and ge tinib Crizotinib None at this tim e

Medical Oncology Principles for the Spine Oncology Surgeon lym ph om a kin ase) t ran slocat ion , an d Ras (rat sarcom a) m utations. The EGFR m utat ions occur in approxim ately 15%of adenocarcinom a NSCLC h istology an d gen erally occur in n on sm okers. Th ese t um ors are sen sit ive to EGFR oral t yrosine kin ase in h ibitors such as erlot in ib or ge t in ib. an d pat ien ts h ave a m ore favorable progn osis. Th e ALK t ran slo cat ion sim ilarly occurs in n on sm okers, bu t in a sm aller n u m ber of p at ien ts, occu rring in less th an 4% of pat ien ts w ith adenocarcinom a NSCLC. These t um ors are sensitive to a n ew t yrosin e kin ase in h ibitor, crizot in ib, w ith im proved progn osis. Th e KRAS m ut at ion occu rs in about 25% of aden ocarcin om a NSCLC an d is fou n d in m ain ly sm okers. Th e p rogn ost ic value of KRAS is lim ited, bu t th erapeu t ic opt ion s an d fu r th er st u dies are u n der w ay to determ in e if at tacking th e dow n st ream target of KRAS m igh t be h elp fu l. Th u s, in ap proach ing patients w ith adenocarcinom a NSCLC, additional m olecu lar test ing can h elp w ith both p rogn osis an d t reat m en t opt ion s. An oth er com m on can cer th at m et ast asizes to bon e is breast carcin om a. Th e t w o m ain h istological classi cat ion s of breast can cer are lobu lar an d du ctal. Th e vast m ajorit y of breast can cer is du ct al (> 75%) an d, just like NSCLC, can be furth er classi ed based on additional ancillar y tum or testing. The three m ost com m on an cillar y tests are est rogen receptor (ER), progesteron e receptor (PR), an d Her2Neu (HER2). Kn ow in g t h e st at u s of t h ese t h ree test s, t h e on cologist can h elp provide th e pat ien t w ith in form at ion abou t th eir p rogn osis an d direct th erapeut ic opt ion s. Pat ien t s w ith t u m ors th at are ER an d PR posit ive are sen sit ive to h orm on al th erapy an d h ave a sign i can tly favorable progn osis com pared w ith pat ien t s w h ose t u m ors are ER an d PR n egat ive. Th e HER2 on cogen e is a m em ber of th e EGFR fam ily an d is also h elpful in p rogn osis an d tailoring t reat m en t . Tum ors th at are HER2/n eu posit ive are sen sitive to dr ugs th at target th e HER2 receptor. Trastuzum ab and pertuzum ab are intravenously adm in istered m on oclon al an t ibodies th at target HER2/n eu , an d lap at in ib is orally ad m in istered an d t arget s bot h EGFR an d HER2. Th e addit ion of th ese dr ugs to stan dard t reat m en t regim en im p roves th e overall su r vival of p at ien t s w ith m etastat ic breast can cer.

Pat ien t s w h ose t um ors are n egat ive for ER, PR, and HER2/neu are classi ed as having triple n egat ive breast can cer an d, in th e m etast at ic set t ing, h ave decreased su r vival. Hen ce, t issu e is th e issu e, h aving im p act on th e diagn osis, p rogn osis, an d t reat m en t decision m aking. Adequ ate t issu e can be safely an d accurately obtained by CT-guided biopsy to run th e n ecessar y an cillar y test s.

■ Staging Th e secon d p r in cip le in m ed ical on cology is st aging—delin eat ing t h e exten t of th e t u m or. Historically, clin ician s h ad to rely on physical exam in at ion n dings or crude radiology im aging to get an idea of th e exact locat ion of th e t u m or an d it s exten t w ith in th e body. Advan ces in radiology techniques have allowed us to stage th e pat ien t m ore accurately, bu t for cer tain t um ors w e st ill requ ire ad dit ion al test ing. Moreover, w e n eed to un derst an d th e pros an d con s of each im aging m odalit y. Th e appropriate st aging tests depen d on th e t ype of tum or (tissue is the issue—the rst principle of m edical on cology) an d un derst an ding of th e expected pat tern of m etastases. For th e t w o m ost com m on prim ar y solid bon e t u m ors of th e spin e, Ew ing’s sarcom a an d osteogen ic sarcom a, as w ell th e com m on h em atology prim ar y bon e t u m ors, lym p h om a or m u lt ip le m yelom a of t h e sp in e, st aging st u d ies d i er (Table 5.2). Ew ing’s sarcom a (or th e Ew ing fam ily of t u m ors) is often evalu ated in it ially w ith a p lain X-ray of th e a ected bon e. For su sp iciou s lesion s, CT im agin g can h elp to d eter m in e t h e extent of cortical destruct ion, w hereas MRI can better help to search for intram edullar y disease (bon e m et ast ases) an d soft t issue exten sion . Becau se Ew ing’s sarcom a h as a predilect ion to m et ast asize to th e lu ng, a CT scan of th e ch est is an in tegral p ar t of th e staging p rocess. Tech n et iu m bon e scan of th e en t ire skeleton com pletes th e im aging evalu at ion for th e presen ce of syn ch ron ous bon e m et astases. Alth ough u orod eoxyglu cose (FDG)-PET is often u sed for staging, there are lim itations and bene ts com -

53

54

Chapter 5 Table 5.2 Staging Studies Di erences Betw een Ew ing’s Sarcoma and Osteogenic Sarcoma Staging Studies

Ew ing’s Sarcoma

Osteogenic Sarcoma

CT MRI

Chest Of the a ected lesion including the  whole bone Can be done, but FDG PET/CT gaining favor Can be done if all bones are included and a diagnostic CT chest is done

Chest Of the a ected lesion including the  whole bone Preferred

Technetium bone scan FDG PET/CT

Bone marrow biopsy

Bilateral bone m arrow

Lim ited by lack of sensitivit y to pulmonary metastasis and detection of small bone lesions Not indicated

Abbreviations: CT, computed tomography; MRI, m agnetic resonance im aging; FDG-PET,  uorodeoxyglucose–positron  em ission tom ography.

pared w ith th e CT ch est an d bon e scan . Many of th e FDG-PET/CT scan s are low resolut ion , w h ich decreases th eir abilit y to detect pulm on ar y m etastases. In addit ion , rout in e FDG-PET/ CT scan s do n ot im age all th e bon e in th e body, an d u n less t h ere is collaborat ion w it h t h e rad iologist to cover all t h e bon es, t h e p at ien t m igh t n ot be com p letely st aging. FDG-PET im aging, h ow ever, is bet ter t h an bon e scan s in d etect in g sm all bon e lesion s. Bon e m ar row biopsies (bilateral) h ave been h istorically perform ed for th e st aging of Ew ing’s sarcom a because stan dard st aging st udies w ere n ot speci c or sen sit ive en ough to detect involvem en t of th e bon e m arrow. Th e use of FDG-PET in addit ion to a diagn ost ic CT ch est an d screen ing MRI of th e en t ire sp in e are becom ing com m on st aging st u dies for Ew ing’s sarcom a. Addit ion al blood testing, including a com plete blood coun t (CBC), creat in in e, an d liver fu n ct ion test s, is also rou t in ely d on e, an d fu r t h er evalu at ion w ou ld be n eeded if th ese tests are sign i can tly abn orm al. In con t radist in ct ion to Ew ing’s sarcom a, osteogen ic sarcom a rarely m et astasizes to bon e m arrow, so bilateral bon e m arrow biop sies are n ot par t of th e rout in e st aging st udies. Th e CT ch est an d bon e scan , as w ith Ew ing’s sarcom a, are st an dard st aging st u d ies for osteogen ic sarcom a given it s p red ilect ion to m et ast asize to th e lu ng an d bon e. Th e role of FDG-PET for osteogen ic staging is lim ited by its lack of detection of sm all pulm onary m etastases and less sen sit ivit y in detect ing bon e m et astases com pared w ith tech n et iu m bon e scan s. Th e role of

FDG-PET in m ore accurate assessm en t of respon se to preoperat ive th erapy is evolving. Ap prop riate im aging of th e local d isease in clu ding th e en t ire a ected an d adjacen t bon es w ith an MRI is p ar t of th e st aging, as skip lesion s can be detected. Th e h em atologic p rim ar y bon e t u m ors lym p h om a an d m u lt ip le m yelom a h ave d i eren t st agin g st u d ies. Lym p h om a can be classi ed broad ly in to Hodgkin’s lym ph om a (HL) an d n on Hodgkin’s lym ph om a (NHL). Staging st u dies of HL in clu de FDG-PET/CT, bu t bon e m arrow biopsies are u n n ecessar y. For NHL, th e st an dard st aging is a CT of th e ch est , abd om en , an d p elvis. In add it ion , becau se th e bon e m arrow is involved 30 to 50% of th e t im e, bilateral bon e m arrow biop sies are n ecessar y to com p lete th e st aging. Th e u se of FDG-PET is on ly for th e aggressive h istologies of NHL (t issu e is th e issue), su ch as d i u se large-cell NHL. Cerebrosp in al u id an alysis is ad dit ion ally in clu d ed in t h e st agin g of cen t ral n er vou s system (CNS) lym p h om a. Sim ilar to Ew ing’s sarcom a an d osteo genic sarcom a, routine CBC, creatinine, and liver fu n ct ion tests are obt ain ed, bu t ser u m p rotein elect rop h oresis (SPEP) an d β 2 -m icroglobu lin (in d olen t lym p h om a) are also obt ain ed for lym ph om a. Plasm acytom a/m ultiple m yelom a is another com m on h em atologic m align an cy of th e spin e, an d ap p rop r iate st aging is h elpfu l in gu id ing t reat m en t opt ion s. Solitar y plasm acytom a can occu r in th e bon e, an d st aging is n eeded to exclu d e m u lt ip le m yelom a. St agin g st u d ies in clu d e m et ast at ic bon e su r vey an d FDG-PET or

Medical Oncology Principles for the Spine Oncology Surgeon MRI of th e en t ire sp in e an d pelvis. In addit ion , laborator y st u dies, in clu ding CBC, creat in in e, ch em ist r y grou p, SPEP (ser u m p rotein elect rop h oresis), SIEP (ser u m im m u n olect rop h oresis), quant itative im m unoglobulins, urinary PEP (protein elect roph oresis), an d a un ilateral bon e m arrow biop sy sh ou ld be obt ain ed. If th e im aging st u dies do n ot sh ow any ad dit ion al lyt ic lesion s an d th e criteria for m u ltip le m yelom a are n ot m et on th e lab tests or bon e m arrow biopsy, th en t reat m en t is prim arily radiat ion . Th e ben e t s of addit ion al system ic th erapy are u n clear, but u p to 50%or m ore p rogress to m u lt iple m yelom a an d 11% h ave local relapse, so long-term follow -up is n eeded. Staging is essential in determ ining the extent of t um or involvem en t an d u lt im ately ch oice of m an agem en t . Each t u m or varies in it s predilect ion for th e site of m etastasis as w ell as sen sit ivit y for detect ion on th e varied im aging m odalit ies. Th u s th e p rocess of st aging is t issu e speci c.

■ Adjuvant Therapy Th e th ird m edical on cology prin cip le th at all spin e surgeon s sh ou ld kn ow is th e value of adjuvan t th erapy or n eoadjuvan t th erapy. Adjuvan t th erapy is referred to as th e addit ion al th erapy after th e p rim ar y lesion h as been com pletely rem oved . W h en th erapy is u sed p rior to resect ion of th e t u m or, t h is is refer red to as n eoadju van t t reat m en t . Th e m ain reason s on cologist s u se n eoadjuvan t/adjuvan t th erapy are to im prove local con t rol, preferably sh rin k/ calcify th e t u m or to im prove su rgical m orbidit y, an d to con t rol any ad d it ion al m icrom et ast at ic disease th at m igh t be p resen t bu t n ot detected by conventional m ethods. The value of adjuvan t/n eoadjuvan t th erapy is m ost eviden t for prim ar y t um ors of th e spin e: Ew ing’s sarcom a and osteogenic sarcom a. Ew ing’s sarcom a is a system ic disease. Even th ough th e t u m or appears to be localized to on e bon e by appropriate staging st u dies, w ith out ch em oth erapy, th e risk of recu rren t an d m et ast at ic disease is over 80%.6 With th e ad dit ion of adjuvan t/n eoadjuvan t system ic ch em oth erapy, th e cu re rate

can ap proach 80%. Osteogen ic sarcom a h as a sim ilar 80% rate of m etast at ic disease w ith ou t adjuvan t/n eoadjuvan t system ic th erapy. With th e addit ion of ch em oth erapy, th e cu re rates are arou n d 60 to 70%.7 Th e select ion of adjuvan t versu s n eoadju vant ch em oth erapy is often based on h istorical dat a, bu t th ere are ad dit ion al reason s to select on e approach over th e oth er. In prin ciple, th e use of neoadjuvant therapy w ill enable a reduct ion in t u m or size, th u s p oten t ially facilitat ing su rgical resect ion w it h ap p rop r iate m argin s. Th is can be esp ecially valu able for large t u m ors t h at are “bord erlin e resect able” or w h ere cr it ical n eu rologic or vascular st ruct ures can be spared. Neoadjuvan t th erapy also en ables th e n at ural biology of th e can cer to declare it self. Th is is esp ecially h elpfu l w h en t h e on cology team an d pat ien t are con tem plat ing an aggressive life-ch anging su rger y, an d addit ion al t im e prior to su rger y is n eeded to see if m etast at ic disease w ill develop an d obviate th e n eed for a life-changing curative surgery. In this situation, n eoadjuvan t ch em oth erapy can be u sed to h elp determ in e if th e t u m or is respon sive to ch em oth erapy an d th e likelih ood of con t rolling th e m icro-m et ast at ic disease. An oth er ben e t of n eoadjuvan t ch em oth erapy is it s progn ost ic value in estim ating the percent of necrosis after ch em oth erapy (p ath ological resp on se). For osteogen ic sarcom a, after t w o to fou r cycles of neoadjuvant chem oth erapy, tum ors at th e tim e of resect ion t h at h ave ≥ 90% n ecrosis h ave a favorable ou tcom e com pared w ith th ose w ith < 90%n ecrosis.8,9 W h eth er adjuvan t in ten si ed ch em oth erapy for t u m ors w ith < 90% n ecrosis im proves overall su r vival is n ot kn ow n , an d th e result s of th e com pleted EURAMOS st u dy are eagerly aw aited.10 Th ere are t im es w h en n eoadjuvan t th erapy, although preferred, is n ot adm in istered. Such in ciden tal discoveries of m align an cy can occu r w ith excision s of presum ed ben ign t um ors, or rem oval of lesion s fou n d during repairs of fract u res. In th ese sit u at ion s, th ere is eviden ce for both Ew ing’s sarcom a an d osteogen ic sarcom a that w hen all the norm ally plann ed chem otherapy is given postoperat ively, overall su r vival is n ot com prom ised.11,12 An oth er scen ario w h ere n eoadjuvan t th erapy is w ith h eld is in p at ien t s

55

56

Chapter 5 w h ose clin ical stat u s is in su cien t to w ith st an d or tolerate system ic ch em oth erapy u n t il th e prim ar y t u m or is rem oved. Th e on e caveat is th at p at ien t s w ith Ew ing’s sarcom a can h ave sign i can t system ic sym ptom s an d app ear n ot to be ch em oth erapy can didates, bu t because Ew ing’s sarcom a is ver y sen sit ive to chem otherapy, these patients can have dram atic clinical response, so all e orts to adm inister neoadjuvant ch em oth erapy sh ou ld be t ried . Oth er reason s to go directly to surger y w ith out n eoadjuvan t ch em oth erapy are if th e t u m or is cau sing sign i can t clin ical com prom ise, or if a m in or in crease in th e t um or size could be catast roph ic, p reclu d ing cu rat ive resect ion . In t h ese sit u at ion s, a solid u n d erst an d ing of t h e likelih ood of sh rin kage of th e t u m or w ith n eoadjuvan t ch em oth erapy is n eeded to ju st ify d elaying su rger y. For in st an ce, prim ar y osteogen ic sarcom a of th e spin e h istorically does n ot sh rin k w ith n eoadjuvan t ch em oth erapy. On th e oth er h an d , Ew in g’s sarcom a h as a greater ch an ce of resp on ding both in t u m or size an d percen t n ecrosis and thus justify neoadjuvant chem otherapy, w ith a d elay in su rger y. As w ith all th ese sit u at ion s, act ive collaborat ion bet w een th e m em bers of th e m u lt idiscip lin ar y team is critical in h elping to m ake th e best decision for th e pat ien t . Adjuvant th erapy is also routinely given after neoadjuvant chem oth erapy and surger y to continue to cont rol any additional m icrom etastatic d isease. As n oted above, in ten si cat ion of adju van t th erapy for poor resp on ders (< 90% n ecrosis) for osteogen ic sarcom a is n ot kn ow n to im pact su r vival, but add it ion al th erapy sh ould be adm in istered. Sign i can t delay in resum p t ion s of system ic ch em oth erapy on an adjuvan t basis can com prom ise overall sur vival. Good com m unication and coordination w ith the m edical on cologist for t im ely resu m pt ion of ch em oth erapy is im p or t an t w h en th e pat ien t h as appropriately recovered . Un for t un ately, n ot all p at ien t s are able or w illing to resu m e ch em oth erapy after su rger y. Deter m in in g t h is p r ior to su rger y cou ld h elp gu id e t h e su rgeon on h ow close t h e m argin s can be or h ow m u ch fu n ct ion sh ou ld be com p rom ised if overall su r vival w ill be red u ced w it h ou t adjuvan t ch em ot h erapy. Alth ough by

de n it ion it is n ot con sid ered adjuvan t th erapy if given w h en gross disease rem ain s or th ere are posit ive resect ion m argin s (R2 an d R1 resect ion , respect ively), kn ow ing th e h istology of th e t u m or cou ld en able ap propriate R2 or R1 resect ion follow ed by postoperat ive th erapy, especially if th e t u m or is sen sit ive to both ch em oth erapy an d radiat ion . Th e d ilem m a is th at th e on cology team can n ot alw ays predict w h o w ill be un able to un dergo adjuvan t th erapy or postop erat ive th erapy. For th e prim ar y bon e t um ors of th e sp in e (Ew ing’s sarcom a an d osteogen ic sarcom a), th e m ain ap p roach is to ad m in ister bot h n eoad ju van t an d adju van t system ic t h erapy w it h R0 resect ion w h en p ossible.

■ Toxicity of Systemic

Therapy Th e fou r th m edical on cology prin cip le is to u n derstan d th e toxicit ies of th erapy. Th e im portan ce of th is prin ciple for th e on cological spin e su rgeon is in determ in ing th e opt im al t im ing for surger y in relat ion to ch em oth erapy an d to appropriately m an age th e pat ien t in th e postoperat ive period. Th e vast m ajorit y of ch em oth erapy d r ugs u sed to t reat Ew ing’s sarcom a an d osteogen ic sarcom a of t h e sp in e an d bon e are cytotoxic agen t s t h at can a ect t h e bon e m ar row as w ell as t h e kid n ey an d liver. Most of th e toxicit y is eviden t by th e low cou n t s for CBC, elevated creat in in e for th e kidn ey, an d elevated liver fun ct ion test s for th e liver. On e h as to be aw are th at despite n orm al coun ts at th e t im e of ch em oth erapy adm in ist rat ion , th ere is a kn ow n n adir period for th ese drugs. Recover y can be as qu ick as 2 w eeks bu t can be delayed du e to cu m u lat ive toxicit y. Som e of th ese dr ugs can a ect th e kid n ey’s abilit y to m ain t ain elect rolyte balan ce (Fan con i’s syn drom e) desp ite a n orm al creat in in e. For m et ast at ic disease from oth er solid t u m ors, th ere are m any n ew agen t s th at t arget th e vascu lar en doth elial grow th factor an d predisp ose pat ien ts to bleed an d clot . Th e agen t bevacizum ab sh ould be stopped up to 4 w eeks prior to elect ive su rger y to redu ce th e risk of

Medical Oncology Principles for the Spine Oncology Surgeon bleeding.13 In som e pat ien t s th ere is n o t im e to d elay su rger y, an d m an aging th e bleeding w ill h ave to be don e exp ect an tly. Coordin at ion an d collaborat ion w ith th e t reat ing m edical on cologist to select an appropriate t im e for surgery is ideal. Th ere are m any n ew orally ad m in istered t yrosin e kin ase in h ibitors to vascular en doth elial grow th factor (VEGF) th at h ave sign i can tly sh or t h alf-lives com pared w ith bevacizu m ab, an d m ost clin ical t rials recom m en d th at th ese agen ts be stopped 1 w eek prior to su rger y, bu t th ere are im m erging dat a th at som e of th ese d rugs can be stop ped as sh ort as 1 day prior to surger y.14 Th e com plicat ion rate after discon t in u at ion of a VEGF in h ibitor is likely also related to th e locat ion an d h istology of th e t u m or, and continued caution should be practiced w hen operat ing after stop p ing a VEGF in h ibitor. Cer t ain d r ug exp osu res are im p or t an t in m an aging th e p eriop erat ive p eriod. Th e best kn ow n is th e prior exposure of th e pat ien t to bleom ycin. Th is drug accum ulates in th e lung an d can cause lung dam age w h en a h igh con cen t rat ion of oxygen is u sed perioperat ively. Th e n ew er t argeted agen t s h ave a h ost of adverse even t s th at are n ot t yp ically seen w ith stan dard cytotoxic agen ts. These adverse events sh ou ld be n oted an d m an aged accordingly (see text box).15

Common Adverse Events w ith Agents that Target VEGF Hypertension Thrombotic risk Hem orrhage Wound complications including  stulas Reversible posterior leukoencephalopathy syndrome Cardiomyopathy Proteinuria

■ Patient’s Preferences Th e n al m ed ical on cology p r in cip le is argu ably t h e m ost im p or t an t an d by n o m ean s u n iqu e to on cology: determ in ing th e p at ien t’s p referen ces. Th e m ed ical on cologist is often

able to establish a relat ion sh ip w ith th e pat ien t and the fam ily. Through initial consultation and follow -u p visits, th e m edical on cologist is able to get to kn ow t h e p at ien t s an d u n d erst an d th eir w ish es an d desires. Th is o ers th e on cology team valuable in sigh t th at can h elp direct t reat m en t opt ion s. Th e on ly p itfall is th at th e m edical on cologist does n ot h ave exper t ise regarding th e surgical opt ion s an d all th e n uan ces of th e expected ou tcom es an d lim itat ion s of t h e var iou s su rgical p roced u res. Th e su rgeon’s inpu t is th erefore vit al in th is regard, an d th e best overall outcom e requ ires an in teract ive m u lt idisciplin ar y approach . For p rim ar y spin e t um ors, especially if th ere is know n or expected spinal cord or nerve dam age, the surgeon is the m ost appropriate m em ber of th e on cology team to explain th e opt ion s to th e pat ien t an d to solicit th e pat ien t’s input on t h e p oten t ial d egree of loss of fu n ct ion or fu n ct ion al im p airm en t t h at is accept able to the pat ient. Many patien ts w ith m etastatic solid t u m ors to th e sp in e h ave u n dergon e several toxic th erapies an d are unw illing to u n dergo fu r th er t reat m en t; th ey feel ready for su p por tive care alon e. It t akes t im e for th e on cology team to un derst an d th e pat ien t’s preferen ces. Th e on cology team feels h elp less w h en addit ion al th erapy can be given , su ch as m ore ch em oth erapy, m ore radiat ion , an d m ore surger y, bu t t h e p at ien t says n o m ore t h erapy. It is im p or t an t to t ake t h e t im e to u n d erst an d t h e w ish es of th e p at ien t .

■ Chapter Summary Th e m ed ical on cology p r in cip les d iscu ssed in th is ch apter are useful for any provider w h o treats patient w ith a m alignancy. The rst principle, t issue is th e issue, em ph asizes th at in order to appropriately t reat th e t u m or w e n eed to kn ow th e speci c t ype of t um or, as t reatm en t s are t ailored for sp eci c t u m ors. On ce w e kn ow w h at t ype of t um or w e are dealing w ith , w e th en n eed to kn ow th e exten t of th e t u m or w ith ap p rop riate st aging (p rin cip le 2). Once w e know the t ype and extent of the tum or, w e th en can decide th e value of any adjuvan t

57

58

Chapter 5 or n eoadjuvan t th erapy (prin ciple 3) to im prove ou tcom es for th e sp eci c t u m or. Pat ien t s w h o are on ch em oth erapy often ben e t from su rgical in ter ven t ion s, an d a good u n derstan ding of toxicit ies of th erapy (prin cip le 4) w ill h elp th e surgeon plan th e t im ing of th e surger y an d appropriately m an age th e pat ien t postoperat ively. In th e en d, ou r role is to h elp th e pat ien t w h ose fears are real. We n eed to sp en d th e t im e to listen th eir w ish es an d con cern s (prin ciple 5) in order h elp gu ide th em to m ake th e righ t decision .

Pearls Tissue is the issue. An accurate diagnosis is essential to guide appropriate m anagement.

Staging. Before embarking on a long journey, you  need to know the extent of disease. The role of adjuvant therapy is to improve both  local and distant disease control. Tumor rem oval  is only one essential part of the treatment plan. Understand the risks and bene ts of systemic  therapy. When is the right tim e to operate without comprom ising the oncological outcom e? Understand the preferences of the patient; take the time to listen. Pitfalls Not all medical oncologists have the necessary experience and expertise. Get to know your  oncologist. Do not rush into surgery. Do not failure to understand the natural history of the tumor.

Refere nces Five Mu st-Read Referen ces 1. Rim on di E, Rossi G, Bartalena T, et al. Percutan eou s CT-gu ided biop sy of th e m u scu loskelet al system : results of 2027 cases. Eur J Radiol 2011;77:34–42 10.1016/j.ejrad.2010.06.055 Pu bMed 2. Rim on di E, St aals EL, Erran i C, et al. Percu t an eou s CTgu ided biopsy of th e spin e: result s of 430 biopsies. Eur Spin e J 2008;17:975–981 10.1007/s00586-0080678-x PubMed 3. Ya e D, Green berg G, Leit n er J, Gip stein R, Sh ap iro M, Bach ar GN. CT-gu ided percut an eous biopsy of th oracic an d lum bar spin e: A n ew coaxial tech n ique. AJNR Am J Neuroradiol 2003;24:2111–2113 PubMed 4. Scagliot t i GV, Parikh P, von Paw el J, et al. Ph ase III st u dy com p ar ing cisp lat in p lu s gem cit abin e w it h cisplat in plus pem et rexed in ch em otherapy-n aive pat ien t s w ith advan ced-st age n on -sm all-cell lu ng can cer. J Clin On col 2008;26:3543–3551 10.1200/ JCO.2007.15.0375 PubMed 5. Syrigos KN, Van steen kiste J, Parikh P, et al. Progn ost ic an d p redict ive factors in a ran dom ized p h ase III t rial com p ar in g cisp lat in -p em et rexed versu s cisp lat in gem cit abin e in advan ced n on -sm all-cell lu n g can cer. An n On col 2010;21:556–561 10.1093/an n on c/ m dp392 Pu bMed 6. Wom er RB, West DC, Krailo MD, et al. Random ized con trolled trial of inter val-com pressed ch em oth erapy for the treatm en t of localized Ew ing sarcom a: a report from the Children’s Oncology Group. J Clin Oncol 2012; 30:4148–4154 10.1200/JCO.2011.41.5703 PubMed

7. Bacci G, Picci P, Ferrari S, et al. Prim ar y ch em oth erapy an d delayed surger y for n on m et ast at ic osteosarcom a of th e ext rem it ies. Resu lt s in 164 pat ien t s preop erat ively t reated w ith h igh doses of m eth ot rexate follow ed by cisplat in an d doxor ubicin . Can cer 1993;72:3227–3238 PubMed 8. Huvos AG, Rosen G, Marcove RC. Prim ar y osteogen ic sarcom a: path ologic aspect s in 20 pat ien t s after t reat m en t w ith ch em oth erapy en bloc resect ion , an d prosth et ic bon e rep lacem en t . Arch Path ol Lab Med 1977;101:14–18 Pu bMed 9. Bacci G, Mercuri M, Longhi A, et al. Grade of ch em oth erapy-in duced n ecrosis as a predictor of local an d system ic control in 881 patients w ith n on -m etastatic osteosarcom a of th e ext rem it ies t reated w ith n eoadjuvan t ch em oth erapy in a single in st it u t ion . Eu r J Can cer 2005;41:2079–2085 10.1016/j.ejca.2005.03. 036 PubMed 10. Bacci G, Forn i C, Ferrari S, et al. Neoadjuvan t ch em oth erapy for osteosarcom a of th e ext rem it y: in ten si cat ion of preoperat ive t reat m en t does n ot in crease th e rate of good h istologic resp on se to th e p rim ar y t u m or or im prove th e n al outcom e. J Pediat r Hem atol On col 2003;25:845–853 PubMed 11. Bacci G, Picci P, Gitelis S, Borgh i A, Cam pan acci M. Th e t reat m en t of localized Ew ing’s sarcom a: th e experien ce at th e Ist it u to Or toped ico Rizzoli in 163 cases t reated w ith and w ith out adjuvan t ch em oth erapy. Can cer 1982;49:1561–1570 PubMed

Medical Oncology Principles for the Spine Oncology Surgeon 12. Goorin AM, Sch w ar t zen t ruber DJ, Devidas M, et al; Pediat ric On cology Group. Presurgical ch em oth erapy com pared w ith im m ediate surger y an d adjuvan t chem otherapy for n onm etastatic osteosarcom a: Pediatric On cology Grou p St u dy POG-8651. J Clin On col 2003; 21:1574–1580 10.1200/JCO.2003.08.165 Pu bMed 13. Gordon CR, Rojavin Y, Patel M, et al. A review on bevacizu m ab an d su rgical w ou n d h ealing: an im p or t an t w ar n ing to all su rgeon s. An n Plast Su rg 2009;62: 707–709 10.1097/SAP.0b013e3181828141 Pu bMed

14. Margulis V, Mat in SF, Tan n ir N, et al. Su rgical m orbidit y associated w ith adm in ist rat ion of t argeted m olecu lar th erap ies before cytoredu ct ive n eph rectom y or resect ion of locally recurren t ren al cell carcin om a. J Urol 2008;180:94–98 10.1016/j.juro.2008.03.047 PubMed 15. Ch en HX, Cleck JN. Adverse e ect s of an t ican cer agen t s t h at t arget t h e VEGF p at h w ay. Nat Rev Clin On col 2009;6:465–477 10.1038/n rclin on c.2009.94 PubMed

59

6 Spinal Osteoid Osteoma and Osteoblastoma Mélissa Nadeau and Christian P. DiPaola

■ Introduction Osteoid osteom a an d osteoblastom a are solitar y lesion s th at com prise th e osteoblast ic ben ign prim ar y t u m ors of th e spin e. Despite th e fact th at th ey are h istologically ext rem ely sim ilar, osteoid osteom as are not know n to progress to osteoblastom as. Th eir dist in ct ion is largely based on t h eir resp ect ive size an d d i ering biological beh avior. Osteoid osteom as are t ypically less th an 2 cm in diam eter, w hereas osteoblastom as are larger th an 2 cm . With regard to th eir biological beh avior, osteoid osteom as r u n a predictable ben ign course w ith an e ect ive response to treatm ent.1,2 Conversely, osteoblastom as exh ibit a sp ect ru m of beh aviors from relat ively laten t to ext rem ely aggressive, w ith occasion al m align an t t ran sform at ion to low grade osteogen ic sarcom as or “m align an t osteoblastom as” 12 to 25% of th e t im e.3 Both th e likelihood of local recurrence and the treatm ent of osteoid osteom as an d osteoblastom as di er d ram at ically. Osteoblastom as are on e of th e m ost ch allenging ben ign spin e t um ors to t reat becau se of a h igh p rep on deran ce for local recurrence, di cult y in predicting their behavior, an d th e lim ited t reat m en t opt ion s. Osteoid osteom as are four t im es m ore com m on th an osteoblastom as overall.4 Osteoblastom as m ake up 10 to 25%of all prim ar y osseous spin e t um ors.2 Both t ypes of lesion are seen m ain ly in t h e long bon es; on ly 10 to 20% of osteoid osteom as an d 40% of osteoblastom as

occu r in th e sp in e.1,4–6 Osteoid osteom as are m ore prevalen t in th e lum bar spin e, w h ereas osteoblastom as do n ot h ave a p redilect ion for any p ar t icular spin al region . Both lesion s are seen m ore com m on ly in m ales th an fem ales (2–4.5:1) in th eir late teen s to early 20s.5,7–10 Th e m ost com m on age of presen t at ion is in th e secon d to d ecad e of life. Eight y percen t to 100% of patients present w ith n eck or back pain , w h ich is usu ally localized 4 ; 25 to 62% of pat ien t s w ith an osteoid osteom a or an osteoblastom a have a scoliosis at the tim e of presen tat ion (m ore com m on w ith osteoid osteom a),7,8,11 and 45%of cer vical cases present w ith tor t icollis.9 Th ese n dings are both th e resu lt of an in am m ator y e ect on paraspin al m u scles, causing asym m et ric m uscle spasm s.11 In t h e im m at u re t h oracic an d lu m bar sp in e, th is leads to grow th in h ibit ion of th e ver tebral ep iphysis an d a rotat ion al deform it y, resu lt ing in rapidly progressive cu r ve at a h igh risk of becom ing st r u ct u ral.12 Th e vast m ajorit y of sp in al osteoid osteom as an d osteoblastom as a ect th e p osterior elem en t s.4,10,13 Du e to t h eir larger size, osteoblastom as m ay exten d in to th e an terior ver tebral body an d th e spin al can al, th us p oten t ially causing n eurologic sym ptom s.1,2,14 Becau se of t h eir d i eren t beh aviors, p rogn osis, an d t reat m en t s, it is cr it ical to fu lly in vest igate t h ese lesion s w it h t h e ap p rop r iate im aging an d biop sy in ord er to con r m t h e accurate diagn osis of osteoid osteom a, osteoblastom a, or osteosarcom a.

Spinal Osteoid Osteoma and Osteoblastoma

■ Staging and Terminology Ben ign m u sculoskelet al n eop lasm s are classied according to th e En n eking st aging system , w h ich h as been sh ow n to be reliable an d valid in th e sp in e.15 Th is system h elps th e clin ician assess t um or beh avior an d h elps guide t reatm en t . En n eking’s su rgical p rin ciples dict ate w h ich t reat m en t is m ost appropriate for each st age (Fig. 6.1). It is essen t ial for t h e on cological sp in e su rgeon to com pletely assim ilate t h e or igin al d e n it ion s of th e ter m s as d escr ibed by En n eking.16,17 Ter m s u sed to d escr ibe m argin s in clu d e intralesional (som e t u m or left beh in d), m arginal (dissect ion along th e t u m or cap su le or react ive zon e), or w ide (cu of h ealt hy t issu e arou n d th e t u m or is rem oved), an d th ey m ust be dist inguish ed from th e di eren t surgical tech n iques used to ach ieve th ese m argin s. Th ese t ypically in clude th e piecem eal tech n iqu e, w h ere th e t u m or is rem oved in several d i eren t pieces (e.g., cu ret t age), or en bloc, in w h ich th e t u m or is rem oved in on e p iece, regardless of th e su rgical m argin s. An oth er im -

61

por t an t dist in ct ion is th at of th e su rgical an d h istological m argin s. Th e surgical m argin s describe th e su rgeon’s plan n ed an d gross im p ression of th e exten t of resect ion in t raop erat ively. In con t rast , h istological m argin s are obt ain ed w h en th e in t raoperat ive sam ple is sen t to th e path ology lab, an d a m icroscopic an alysis ser ves as th e n al verdict for surgeon s regarding th eir su ccess at ach ieving th e plan n ed su rgical m argin s (in t ralesion al, m argin al, or w ide).16,17

■ Osteoid Osteoma Osteoid osteom as accou n t for ap proxim ately 10% of prim ar y bon e t um ors.18 Th ey are bon eform ing t u m ors w ith lim ited grow th poten t ial t h at ran ge from 15 to 20 m m in d iam eter. Pat ien t s t yp ically p resen t w it h con st an t or ep isod ic p ain t h at in creases at n igh t or w it h physical activit y. The pain is caused by the presence of ner ve endings w ithin the tum or, w hich are stim ulated by vascular pressure an d the prod uct ion of p rostaglan din s.14 Sym ptom s classi-

a

b

Fig. 6.1a,b A 40-year-old man presented with an insidious onset of left-sided neck pain radiating to his left scapular region, slowly progressive over 10 m onths. His pain was bothersome at night, and relieved with nonsteroidal anti-in amm atories. He denied any neurologic symptoms. Axial (a) and sagit tal (b) computed tomography (CT) scans showed a well-circumscribed lesion with m ineraliza-

tion in the nidus, radiolucent rim (representing the portion of the nidus that is not mineralized), and sclerotic rim. It m easured 8 mm in diam eter. Clinical and radiographic features were consistent with an osteoid osteoma, and the patient was therefore treated with radiofrequency ablation, resulting in a complete resolution of his symptom s.

62

Chapter 6 cally respond to nonsteroidal anti-in am m atory drugs (NSAIDs), at least in it ially. Osteoid osteom a is th e m ost com m on cau se of pain ful scoliosis in adolescen t s.18

Imaging On p lain radiograph s, osteoid osteom as m ay sh ow an osteosclerot ic lesion , w ith or w ith ou t a visible radiolucen t n idus, but th ey are often m issed becau se of th eir locat ion in th e p osterior elem en t s an d relat ively sm all size (< 2 cm ). On com p u ted tom ography (CT) scan , th ey are ch aracter ized by sclerosis su r rou n d in g a rad iolu cen t n idu s, an d periosteal bon e react ion is often seen . A cen t ral region of m in eralizat ion m ay be presen t . Magn et ic reson an ce im aging (MRI) is best used to aid in preoperat ive plan n ing to determ in e proxim it y to n eurologic st r u ct u res.19 On T1- an d T2-w eigh ted im ages, the calci cation w ithin the nidus and surrounding sclerosis are both of low sign al in ten sit y. En h an cem ent of th e vascu lar n idus w ith gadolin iu m m ay be seen . If n eith er of th ese im aging m odalit ies depict a lesion , th e m ost sen sit ive tool in th e diagn osis of osteoid osteom a w ou ld th en be a tech n et iu m bon e scan , w h ich sh ow s intense radionucleotide uptake at the nidus, and a less in ten se larger area of u pt ake su rrou n ding it .20,21 If th ere is any doubt radiograph ically an d clin ically th at th is m ay be som eth ing oth er th an an osteoid osteom a, a biopsy is in dicated. Th is is fur th er discussed in th e osteoblastom a sect ion of th is ch apter.

Treatment Treat m en t opt ion s for osteoid osteom a con sist of con ser vat ive t reat m en t w it h NSAIDs, su rgical excision (m ost com m on ly in t ralesion al cu ret t age or en -bloc excision ), or p ercu t an eous in ter ven t ion s. In adults, con ser vat ive m an agem en t is th e rst lin e of t reat m en t; if th is fails, or if avoidan ce of long-term use of an t iin am m atories is p referred, su rgical excision is in dicated. In you nger p at ien t s, surgical excision is gen erally recom m ended early, to prevent a scoliot ic deform it y. In cases w h ere sp in al deform it y is already presen t , resect ion of th e lesion can resu lt in reversal of th is deform it y if

th e lesion is resected w ith in 15 m on th s of it s on set .7,18 Conven t ion ally, th e su rgical excision of osteoid osteom a h as been con sidered accept able t reat m en t for cu re.7,9,22–24 More recen tly, less invasive tech n iques, such as (ILP) an d percu t an eou s radiofrequ en cy ablat ion (RFA), in it ially em ployed in th e appen dicular skeleton , h ave begu n to gain favor for u se in th e spin e. Th ese tech n iqu es h ave several advan t ages: th ey can be p erform ed u n der local an esth esia, th erefore sim u lt an eou s n eu rologic exam in at ion is p ossible; th ey cau se m in im al su rgical t raum a, an d in t urn pat ien ts experien ce less postoperat ive p ain an d h ave a decreased risk of im p airm en t an d in fect ion ; an d th ey are associated w ith a sh or ter recover y period, sh orter h ospital stay, an d th u s low er cost .14,25–27 On th e ot h er h an d, th e risk of causing local n eural elem en t injur y w ith th ese tech n iqu es h as been a sign i can t con cern in relat ion to th eir use in th e spin e. Speci cally regarding RFA, h eat ing a n eedle t ip to 90°C for 4 to 6 m in utes in th e n idus of a lesion raises con cern s about th erm al inju r y to n earby n eu ral elem en t s. How ever, an ex-vivo st udy h as sh ow n th at th ere is n o tem perat ure in crease w ith in th e sp in al can al w h en an in t act cor tex sep arates n eu ral st r u ct u res from t h e n eed le t ip , suggest in g an in su lat ing e ect of cor t ical bon e.28 Fu r t h er m ore, several recen t st u d ies h ave d em on st rated th e e cacy an d safet y of RFA w h en u sed in t h e sp in al colu m n .29,30 Mar tel et al 29 t reated 10 p at ien t s w it h 4 to 6 m in utes of RFA aim ed at th e n idu s of osteoid osteom as. Th e dist an ce bet w een th e n idu s an d adjacen t n eu ral t issu e w as 2 to 12 m m (m ean , 5 m m ). Th ere w ere n o com plicat ion s reported. Tw o pat ien t s (20%) required a secon d sim ilar RFA t reat m en t d u e to recu rren ce of p ain at 2 m on t h s. Rep eat t reat m en t resulted in long-term sym ptom resolution. Vand ersch u eren et al30 t reated 24 p at ien t s w ith sp in al osteom a w ith RFA. Th ey rep or ted a su ccess rate of 79%(19/24) after a single t reatm ent , an d 96% (27/28) after a secon d t reat m en t . Th ere w ere n o com plicat ion s, an d all pat ien ts w ere disch arged from th e h ospit al on th e sam e day as th e procedure. Th ey con cluded th at CTgu ided RFA sh ou ld be th e t reat m en t of ch oice in lesions located at least 2 m m aw ay from n eu -

Spinal Osteoid Osteoma and Osteoblastoma ral t issu e, w h ereas su rgical excision is advised for lesion s adjacen t (< 2 m m ) to n eu ral st ru ct u res. Several oth er rep or ts of a sm aller n u m ber of pat ien ts w ith osteoid osteom a in th e lam in a, t ran sverse p rocess,31 lum bar ver tebral body,27,32,33 pedicle,28 an d spin ou s processes 34 su ccessfu lly t reated w it h RFA su p p or t t h is as a t reat m en t opt ion for sp in al osteoid osteom a lesion s as w ell.

■ Osteoblastoma Osteoblastom as are ben ign bon e-form ing n eoplasm s sim ilar to osteoid osteom as. How ever, th ey h ave an u n lim ited grow th p oten t ial, being larger th an 2 cm by de n it ion , an d are clin ically m ore aggressive. Pat ien t s w ith th is t yp e of lesion presen t w ith dull back pain or w ith sym ptom s of n eu rologic com p ression . In fact , 25 to 50%of pat ien ts presen t w ith a n eurologic de cit.7,35 In contrast to osteoid osteom a, sym p tom s t ypically are not di eren t at nigh t and respon d poorly to NSAIDs. Scoliosis is also know n to occur in spin al osteoblastom a, bu t less frequen tly th an w ith osteoid osteom a. Osteoblastom as can gen erally be catego rized in to t w o t ypes: act ive an d aggressive. Th e dist in ct ion bet w een th e t w o is crit ical, as it sign i can tly im pact s t reat m en t an d ou tcom e. Un for t u n ately, t h is dist in ct ion is n ot easily m ade or w ell est ablish ed in th e literat u re, an d th erefore on e m u st u se a con stellat ion of clin ical an d im aging n dings. Invest igators sh ou ld use th e Enn eking st aging system (Table 6.1) as a basis to categorize lesion s; stage 2 lesion s are labeled as act ive, an d stage 3 as aggressive.1,2

Stage 2 lesions have well-de ned borders, com bin ed osteolyt ic an d sclerot ic feat u res (often resem bling osteoid osteom a w h ere a sclerot ic ring surroun ds a lyt ic n idus), an d n o soft t issue involvem en t . St age 3 lesion s are m ore rapidly grow ing, usually exceeding 4 cm in diam eter, and have a m ore dest ructive appearance on im aging. Th eir m argin s are poorly de n ed, th ey are p rim arily lyt ic, an d th ey erode t h e cor tex to invade th e sp in al can al an d su rroun ding soft t issues. Som e are ver y expan sile, resem bling an aneur ysm al bony cyst.2 The clinical course of both di er as w ell, w ith act ive lesion s cau sing slow ly progressive pain w ith or w ith ou t sp in al deform it y secon dar y to m u scle sp asm , an d aggressive lesion s cau sing a faster progression of sym ptom s w ith p ossible early n eu rologic com plain t s. An oth er di eren t iat ing feat u re m ay be found histologically: aggressive osteoblastom as com p r ise large ep it h elioid osteoblast s, ch aracterized by abun dan t eosin oph ilic cytoplasm t w ice th e size of conven t ion al osteoblasts.36

Imaging On plain rad iograp h s, 50% of osteoblastom as are lyt ic, 30%are sclerot ic, an d 20%are m ixed 8 ; 55% are fou n d en t irely in t h e p oster ior elem en t s, 42% involve th e p osterior elem en t s an d ver tebral body, an d 3% involve th e ver tebral body exclusively.36 Depen ding on w h eth er th ey are act ive or aggressive, th ey h ave a d i eren t radiological appearan ce as described in th e p reviou s paragraph . Un like osteoid osteom as, osteoblastom as are t yp ically readily visible on p lain radiograph s. CT scan n ing is opt im al p reoperat ively to precisely locate th e t u m or an d

Table 6.1 Enneking Staging System for Benign Primary Spinal Neoplasms Stage

Description

Margin for Control

1 2 3

Latent: usually asymptom atic Active: locally symptomatic Aggressive: symptomatic with local invasion and/or destruction of tissues with potential for metastasis

Intracapsular Marginal, or intracapsular + adjuvant therapy Wide, or marginal + adjuvant therapy

Source: Adapted from Enneking WF. A system of staging musculoskeletal neoplasms. Clin Orthop Relat Res 1986;204:9–24.

63

64

Chapter 6 de n e th e exten t of bony involvem en t , w h ereas MRI is h elp fu l in iden t ifying in t ra- an d ext raosseous react ive ch anges, in ltrat ion in to su rrou n ding soft t issu es or sp in al can al, an d n eurologic com pression . Th e appearan ce of osteoblastom a on MRI, how ever, is generally non speci c an d m ay overest im ate th e size of th e lesion due to local in am m at ion an d edem a.18

Histology Th e di eren t ial diagn osis of osteoblastom a inclu des an eu r ysm al bon e cyst , gian t cell t u m or, osteosarcom a, Ew in g’s sarcom a, car t ilagin ou s t u m ors (en ch on d rom a, osteoch on d rom a, an d ch on d rosarcom a), an d osteom yelit is. Th erefore, a biopsy is in dicated prior to un der t aking t reat m en t . Histological exam in at ion of osteoblastom as is ch aracterized by a brovascu lar st rom a an d a n id u s con t ain ing osteoblast s t h at p rod u ce osteoid t issu e an d w oven bon e. Th is is sim ilar in osteoid osteom as, bu t , in ad dit ion , osteoblastom as com m only have large vascular spaces and reactive giant cells.10,13 Aggressive (stage 3) osteoblastom as are m ore cellu lar, w ith sw ollen , plu m p osteoblast s,13 an d m ay h ave a m ult ifoci grow th p at tern m im icking perm eat ion . Th ese feat ures are also presen t in osteosarcom a; th eir foci are m icroscopically iden t ical to th ose of aggressive osteoblastom as. How ever, t ru e perm eat ion of su rrou n d ing t issu es an d lack of “m aturation” at the edges of the tum or are characterist ic of osteosarcom as an d can be used to di eren t iate osteosarcom a from aggressive osteoblastom a. Th is can easily be m issed , an d th e review of th e h istology by an experien ced m u sculoskeletal on cologist is recom m en ded.37

Treatment Osteoblastom as are m an aged su rgically because of the recalcitran t nature of the pain they cau se, as w ell as du e to th eir locally aggressive beh avior; th eir larger size leads to poten t ially sign i can t bony dest ru ct ion , deform it y, in stabilit y, or n eu rologic com p ression .18 Treat m en t opt ion s con sist of in t ralesion al curet t age w ith bon e graft ing or cem en t at ion , or en -bloc resect ion . Th e form er is recom m en ded for act ive

(En n eking 2) osteoblastom as, an d th e lat ter for aggressive (En n eking 3) osteoblastom as; t h e spin e’s ch allenging st r uct u ral an d n eurologic an atom y m akes it ext rem ely di cu lt to follow th ese gu idelin es.2 En -bloc resect ion w ith m argin al or w ide m argin s is m ore invasive an d im p oses greater m orbidit y on p at ien t s, bu t is ju st i ed by th e h igh recurren ce rate an d possibilit y for m align an t t ran sfor m at ion of aggressive osteoblastom as w h en t reated w ith a m ore con ser vat ive approach .1,2 Sch ajow icz an d Lem os 38 t reated a series of aggressive osteoblastom as an d recu rren ce rates w ere 100%(4/4) for th e on es t reated w ith in t ralesion al curet t age, 20% (1/5) for th e on es t reated w ith excision , an d 0%(0/3) for th e on es t reated en bloc. In a clin ically based system at ic review don e in collaborat ion w ith th e Sp in e On cology St u dy Grou p ,2 aggressive osteoblastom as w ere repor ted to recur at a rate of 50% w h en t reated w ith subtotal resect ion , com p ared w ith 10 to 15% in less aggressive lesion s. Desp ite w eak eviden ce, th e con sen su s exp er t opin ion recom m en dat ion is to t reat aggressive lesion s w ith m argin al or w ide en -bloc excision , an d n on aggressive lesion s w ith in t ralesion al cu ret t age.2 An oth er factor th at w as foun d to a ect th e rate of recu rren ce is w h eth er or n ot th e lesion s h ad been previou sly t reated w ith in t ralesion al curet tage (and therefore were recurrent lesions) or h ad u n dergon e an open biop sy. Borian i et al1 sh ow ed th at p at ien ts w h o h ad eith er an op en biopsy or curet t age h ad a recurren ce rate of 67% (2/3 pat ien t s) after en -bloc resect ion , an d 75% (3/4 pat ien ts) after a secon d in t ralesion al curet tage. This com pares w ith a recurrence rate of 0% (0/10 pat ien ts) after en -bloc resect ion , an d 7.1% (2/30 p at ien t s) after a secon d in t ralesion al cu ret t age in p at ien t s w h ose lesion s w ere intact (no previous procedure at th at site). Lu cas et al,36 in a review of 306 cases, fou n d th at th ere w as recurren ce after en -bloc on ly w h en th e p rocedu re w as p erform ed th rough a p r ior resect ion cavit y. Th is h igh ligh t s t h e im por tan ce of th e correct diagn osis in it ially, in cluding th e di eren t iat ion of aggressive versus act ive lesion s, an d appropriate t reat m en t based on t h is diagn osis, at th e t im e of in it ial t reat m en t .

Spinal Osteoid Osteoma and Osteoblastoma

Adjuvant Therapies: Radiation and Chemotherapy Th e role of rad iat ion an d ch em ot h erapy for in com plete surgical resect ion of osteoblastom a is n ot w ell d e n ed . W h en aggressive lesion s are n ot com p letely excisable du e to an atom ic (e.g., involvem en t of sign i can t n eu ral elem en t s) or m edical con siderat ion s, th ey h ave a h igh er likelih ood of recurren ce. If in t ralesion al resect ion is th e on ly opt ion , p ostoperat ive radiat ion th erapy is th e m ost com m on ly u sed adjuvan t th erapy an d is con sidered reason able. Radioth erapy is also u sed in recu rren t lesion s, m ost often after rep eat in t ralesion al cu ret t age. Th e ben e t of th is m odalit y, h ow ever, is st ill ver y con t roversial, w ith th e m ajorit y of cases sh ow ing n o advan t age an d a m in orit y of cases dem on st rat ing som e ben e t .6,39 Marsh et al6 con clu ded, based on a review of 197 osteoblastom a cases, th at radiat ion th erapy does n ot alter th e course of th e disease an d appears to be con t rain dicated. In con t rast , Borian i et al1 review ed a series of 30 osteoblastom a cases an d fou n d t h at recu r ren ces occu r red in ve of 22 p at ien t s t reated w it h in t ralesion al cu ret t age alon e, com p ared w ith n o recu r ren ces occu rring in th e 15 p at ien t s t reated w ith both curet tage an d radiat ion (at 2-year follow -up).

65

Harrop et al,2 in th eir system at ic review, con cluded th at the eviden ce w as again w eak, but th at radioth erapy in recurren t lesion s or in com pletely resected aggressive osteoblastom as sh ould be con sidered as a t reat m en t opt ion . The use of adjuvan t chem otherapy is lim ited to an ecd ot al cases, an d evid en ce to su p p or t it s use is even w eaker th an for radioth erapy.2 Th ere are a lim ited n u m ber of case repor t s an d an ecdotes of recurren t osteoblastom as th at un der w en t reexcision or radiat ion .40–42 In all cases n o fu r th er recu r ren ce occu r red w h en th e ch em oth erapy w as ut ilized; h ow ever, th is su ccess w as often at t ribu table to th e surgical m an agem en t on ly, w ith n o dem on st rated addit ion al ben e t of ch em oth erapy. Har rop et al2 con clu d ed in t h eir system at ic review t h at t h ere is a lim ited role for ch em oth erapy in recu rren t aggressive osteoblastom a. Furth er research is requ ired to clarify th e role of both th ese adjuvan t t reat m en ts in th e m an agem en t of osteoblastom a.

■ Clinical Cases Pert in en t clin ical cases are presen ted in Figs. 6.1, 6.2, 6.3, 6.4.

a

b

Fig. 6.2a–e A 17-year-old boy presented with a history of progressive midthoracic pain, radicular circumferential chest wall pain, and numbness from his feet gradually extending up to his abdom en. Examination revealed a sensory level at T6 and m ild

spasticit y in his lower extremities. Sagit tal (a) and axial (b) CT scans showed a large lytic lesion encom passing the bilateral lam inas, left-sided transverse process and lateral mass, and alm ost half of the vertebral body. (continued on next page)

66

Chapter 6 Fig. 6.2a–e (continued ) (c) Soft tissue CT scan windows showed extension of the lesion into the spinal canal and compressing the cord. Biopsy con rm ed the diagnosis of osteoblastoma. (d,e) Embolization of the tumor preceded an en-bloc resection with placem ent of an anterior cage, and a posterior instrum ented fusion was perform ed. Pathology of the operative specim en revealed negative tum or m argins. CT and magnetic resonance imaging (MRI) scans at 2 years showed no evidence of tumor recurrence.

c

d

e

■ Chapter Summary Osteoid osteom a an d osteoblastom a com p rise th e osteoblast ic ben ign p rim ar y t u m ors of th e sp in e. Bot h ten d to occu r in you ng m ales an d in th e posterior spin al elem en ts. Dist in ct ion is based on t h eir resp ect ive size an d d i eren t biological beh avior. Osteoid osteom as ten d to be sym ptom atologically w ell con t rolled w it h NSAIDs. In tralesional resection t ypically resu lt s

in d e n it ive cu re. Less invasive tech n iqu es, m ost n ot ably RFA, is becom ing m ore accepted as a t reat m en t opt ion in spin al osteoid osteom as an d h as yielded excellen t resu lt s in th e literat u re. Osteoblastom as are divided in to act ive an d aggressive lesions. The Enneking staging system , radiograp h ic app earan ce, an d h istology on biopsy h elp determ in e w h eth er a par t icular lesion is active or aggressive. Intralesional excision

Spinal Osteoid Osteoma and Osteoblastoma

67

a

b

c

d

Fig. 6.3a–d A 21-year-old man presented with numbness and weakness in his legs. On exam he had a sensory level at T9, increased tone in his lower extremities, upper motor neuron signs, and pyramidal weakness. (a,b) Postgadolinium MRI showed an enhancing lesion at T10 and T11 involving the right posterior elem ents as well as the vertebral bodies. An epidural soft tissue mass was also evident and involving the right T9-10, T10-11,

and T1-2 neuroforamina, as well as the spinal canal, where it caused cord compression. A CT scan (no images available) further delineated the extent of bony destruction. A CT-guided biopsy con rmed the diagnosis of osteoblastoma. (c,d) An en-bloc resection of the T10-T11 osteoblastoma was performed and stabilized with a bular strut graft at the right T10 level and posterior instrum ented fusion from T7 to L1.

68

Chapter 6

a

b

c

d

Fig. 6.4a–d A 24-year-old woman presented with a 1.5-year history of coccygeal region pain, and occasional left but tock radiating pain. Examination revealed tenderness to palpation over the sacral and coccygeal regions, but was otherwise normal. (a,b) CT scan showed an expansile lytic lesion within the fourth sacral vertebral body, expanding the left foramen as well. (c,d) Postgadolinium MRI showed

an enhancing lesion involving the S4 and S5 vertebral bodies and posterior elements, with soft tissue extension into the canal and presacrally. A CT-guided biopsy con rmed the diagnosis of osteoblastoma. An en-bloc excision was performed from the S2-3 level down. The right S4 nerve root was sacri ced due to tum or involvem ent.

h as proven to be e ect ive in act ive lesion s, w h ereas en -bloc resect ion is recom m en ded for aggressive lesion s, du e to th eir relat ively h igh rate of recurren ce w h en t reated in t ralesion ally. Lesion s th at represen t recurren ces or th at w ere par t ially resected (du e to an atom ic con st rain ts) can be t reated w ith radiat ion th erapy, in addi-

t ion to su rgical t reat m en t, in an at tem pt to m in im ize th e rate of recurren ce. Osteoblastom as can becom e locally aggressive an d can even becom e m align an t . Th e differen t biological beh aviors of osteoid osteom a, osteoblastom a, an d osteosarcom a are of th e u t m ost im por tan ce in t reat m en t plan n ing, as is

Spinal Osteoid Osteoma and Osteoblastoma recogn izing th e aggressiven ess of th e lesion . Th is ch apter described th e t yp ical clin ical presen t at ion an d im aging feat u res of bot h t yp es of lesion s, as w ell as th e advisable t reat m en t for each .

Pearls Osteoid osteoma of the spine has never been reported to convert to osteoblastoma. Osteoid osteom a can be treated with minim ally invasive techniques such as RFA. Osteoid osteoma can be treated with intralesional surgical resection if nonoperative measures fail. Not all osteoblastomas are equal; di erentiating stage 2 active from stage 3 aggressive is critical because it in uences treatm ent and prognosis. Aggressive osteoblastom as are generally large, lytic, and destructive, with broad zones of transition and soft tissue expansion. Histologically they comprise large epithelioid osteoblasts, characterized by abundant eosinophilic cytoplasm . Enneking stage 3 osteoblastomas have high recurrence rates when treated with subtotal resection. Therefore, the surgical goal should be a complete resection of the lesion with marginal or wide margins using an en-bloc surgical technique when anatomically possible. Where this is not feasible and positive margins m ust be left, radio-

therapy is a reasonable consideration and m ay help decrease the rate of local recurrence. Pitfalls Recognize the presentations of osteoid osteom a and osteoblastoma. They are a comm on cause of painful scoliosis in teenagers. Osteoblastomas t ypically occur in patients in their late teens (in m ales m ore com monly than females). However, there are reports of children developing these conditions at a younger age. Because these lesions can becom e locally aggressive and lead to spinal deformit y or neurologic de cits, the clinician should be prepared to recognize these tum ors and to pursue the appropriate investigations (im aging, biopsy). If standard im aging m odalities (X-ray, CT scan, MRI scan) fail to identify a lesion, a technetium bone scan is the most sensitive tool for ruling out an osteoid osteom a. Avoid intralesional resection of Enneking stage 3 osteoblastomas if anatom ically feasible. Do not underestimate the potential for malignant degeneration of osteoblastom a. Clinical behavior, tim ing of progression, neurologic involvement, and invasiveness all justify an aggressive treatm ent plan. Initial treatment has a signi cant impact on the recurrence rate.

Refere nces Five Must-Read Refe rences 1. Boriani S, Am endola L, Bandiera S, et al. Staging and treatm ent of osteoblastom a in the m obile spine: a review of 51 cases. Eur Spine J 2012;21:2003–2010 PubMed 2. Harrop JS, Sch m idt MH, Borian i S, Sh a rey CI. Aggressive “ben ign ” prim ar y spin e n eoplasm s: osteoblastom a, aneur ysm al bone cyst, and giant cell t um or. Spin e 2009;34(22, Su ppl):S39–S47 Pu bMed 3. Nish ida K, Doit a M, Kaw ah ara N, Tom it a K, Ku rosaka M. Tot al en bloc sp on dylectom y in th e t reat m en t of aggressive osteoblastom a of th e th oracic sp in e. Orth opedics 2008;31:403 Pu bMed 4. Bu r n SC, An sorge O, Zeller R, Drake JM. Man agem en t of osteoblastom a an d osteoid osteom a of th e sp in e in ch ildh ood. J Neurosurg Pediatr 2009;4:434– 438 PubMed 5. Jackson RP, Reckling FW, Man t s FA. Osteoid osteom a and osteoblastom a. Sim ilar histologic lesions w ith differen t n at u ral h istories. Clin Or th op Relat Res 1977; 128:303–313 PubMed 6. Marsh BW, Bon glio M, Brady LP, En n eking W F. Ben ign osteoblastom a: range of m an ifest at ion s. J Bon e Join t Su rg Am 1975;57:1–9 PubMed

7. Pet t in e KA, Klassen RA. Osteoid-osteom a an d osteoblastom a of th e spin e. J Bon e Join t Surg Am 1986; 68:354–361 Pu bMed 8. Nem oto O, Moser RP Jr, Van Dam BE, Aoki J, Gilkey FW. Osteoblastom a of th e spin e. A review of 75 cases. Spin e 1990;15:1272–1280 Pu bMed 9. Raskas DS, Grazian o GP, Herzen berg JE, Heid elberger KP, Hen singer RN. Osteoid osteom a an d osteoblastom a of th e spine. J Spin al Disord 1992;5:204– 211 Pu bMed 10. Gasbarrin i A, Cappuccio M, Ban diera S, Am en dola L, van Urk P, Borian i S. Osteoid osteom a of th e m obile spin e: surgical ou tcom es in 81 pat ien t s. Spin e 2011;36:2089–2093 Pu bMed 11. Saifuddin A, W h ite J, Sh erazi Z, Sh aikh MI, Nat ali C, Ran sford AO. Osteoid osteom a an d osteoblastom a of th e spin e. Factors associated w ith the presen ce of scoliosis. Spin e 1998;23:47–53 Pu bMed 12. Ran sford AO, Pozo JL, Hu t ton PA, Kir w an EO. Th e beh aviour pat tern of th e scoliosis associated w ith osteoid osteom a or osteoblastom a of th e spine. J Bon e Join t Su rg Br 1984;66:16–20 Pu bMed

69

70

Chapter 6 13. Zileli M, Cagli S, Basdem ir G, Ersah in Y. Osteoid osteom as an d osteoblastom as of th e spin e. Neurosu rg Focus 2003;15:E5 PubMed 14. Beauch am p CP, Dun can CP, Dzus AK, Morton KS. Osteoblastom a: experien ce w ith 23 pat ien t s. Can J Surg 1992;35:199–202 PubMed 15. Ch an P, Borian i S, Fou rn ey DR, et al. An assessm en t of th e reliabilit y of the Enn eking an d Wein stein -BorianiBiagin i classi cat ion s for st aging of prim ar y spin al t um ors by th e Spin e On cology St udy Group. Spin e 2009;34:384–391 PubMed 16. Fisher CG, Saravanja DD, Dvorak MF, et al. Surgical m an agem en t of prim ar y bon e t u m ors of th e spin e: validation of an approach to enhance cure and reduce local recu rren ce. Sp in e 2011;36:830–836 Pu bMed 17. En n eking W F. A system of st aging m u scu loskelet al n eoplasm s. Clin Or t h op Relat Res 1986;204:9–24 Pu bMed 18. Th akur NA, Dan iels AH, Sch iller J, et al. Ben ign t um ors of th e spin e. J Am Acad Or th op Surg 2012;20: 715–724 PubMed 19. Hosalkar HS, Garg S, Moroz L, Pollack A, Dorm an s JP. Th e diagn ost ic accu racy of MRI versu s CT im aging for osteoid osteom a in ch ildren . Clin Or th op Relat Res 2005;433:171–177 Pu bMed 20. Azouz EM, Kozlow ski K, Marton D, Sprague P, Zerh oun i A, Asselah F. Osteoid osteom a an d osteoblastom a of th e spin e in ch ildren. Report of 22 cases w ith brief literat ure review. Pediat r Radiol 1986;16:25–31 PubMed 21. Iyer RS, Chapm an T, Ch ew FS. Pediat ric bon e im aging: diagn ost ic im aging of osteoid osteom a. AJR Am J Roen tgen ol 2012;198:1039–1052 PubMed 22. Fet t HC Sr, Russo VP. Osteoid osteom a of a cer vical ver tebra; rep or t of a case. J Bon e Join t Su rg Am 1959; 41-A:948–950 Pu bMed 23. Savin i R, Mart ucci E, Prosperi P, Gusella A, Di Silvest re M. Osteoid osteom a of th e sp in e. It al J Or th op Trau m atol 1988;14:233–238 PubMed 24. Herm an n G, Abdelw ah ab IF, Casden A, Mosesson R, Klein MJ. Osteoid osteom a of a cer vical vertebral body. Br J Radiol 1999;72:1120–1123 Pu bMed 25. Kan er T, Sasan i M, Okten oglu T, Ayd in S, Ozer AF. Osteoid osteom a an d osteoblastom a of th e cer vical spin e: th e cause of un usual persisten t n eck pain . Pain Physician 2010;13:549–554 PubMed 26. Saccom an n i B. Osteoid osteom a an d osteoblastom a of th e spin e: a review of th e literat ure. Curr Rev Musculoskelet Med 2009;2:65–67 Pu bMed 27. Ost i OL, Sebben R. High -frequen cy radio-w ave ablat ion of osteoid osteom a in th e lum bar spine. Eur Spin e J 1998;7:422–425 Pu bMed

28. Dupuy DE, Hong R, Oliver B, Goldberg SN. Radiofrequen cy ablat ion of spin al t um ors: tem perat ure dist ribut ion in th e spin al can al. AJR Am J Roen tgenol 2000;175:1263–1266 PubMed 29. Mar tel J, Bu en o A, Nieto-Morales ML, Or t iz EJ. Osteoid osteom a of t h e sp in e: CT-gu id ed m on op olar radiofrequ en cy ablat ion . Eu r J Radiol 2009;71:564–569 PubMed 30. Van dersch ueren GM, Oberm an n W R, Dijkst ra SP, Tam in iau AH, Bloem JL, van Erkel AR. Radiofrequ en cy ablat ion of sp in al osteoid osteom a: clin ical ou tcom e. Spin e 2009;34:901–904 Pu bMed 31. Cové JA, Tam in iau AH, Oberm an n W R, Van dersch ueren GM. Osteoid osteom a of th e spin e t reated w ith percut an eous com pu ted tom ography-guided th erm ocoagulat ion . Spine 2000;25:1283–1286 Pu bMed 32. Lin d n er NJ, Ozaki T, Roed l R, Gosh eger G, W in kelm an n W, Wör t ler K. Percu t an eou s rad iofrequ en cy ablat ion in osteoid osteom a. J Bon e Join t Su rg Br 2001;83:391–396 Pu bMed 33. Woertler K, Vest ring T, Boet t n er F, Win kelm an n W, Hein del W, Lindn er N. Osteoid osteom a: CT-guided percu t an eou s radiofrequ en cy ablat ion an d follow -u p in 47 pat ien t s. J Vasc Inter v Radiol 2001;12:717–722 Pu bMed 34. Laus M, Albisin n i U, Alfon so C, Zappoli FA. Osteoid osteom a of t h e cer vical sp in e: su rgical t reat m en t or p ercu t an eou s rad iofrequ en cy coagu lat ion ? Eu r Spin e J 2007;16:2078–2082 PubMed 35. Cerase A, Priolo F. Skelet al ben ign bon e-form ing lesions. Eur J Radiol 1998;27(Suppl 1):S91–S97 PubMed 36. Lu cas DR, Unn i KK, McLeod RA, O’Conn or MI, Sim FH. Osteoblastom a: clin icop ath ologic st u dy of 306 cases. Hu m Pathol 1994;25:117–134 Pu bMed 37. Bertoni F, Un n i KK, McLeod RA, Dah lin DC. Osteosarcom a resem bling osteoblastom a. Can cer 1985;55: 416–426 Pu bMed 38. Schajow icz F, Lem os C. Malign an t osteoblastom a. J Bon e Joint Surg Br 1976;58:202–211 PubMed 39. Tonai M, Cam pbell CJ, Ahn GH, Schiller AL, Mankin HJ. Osteoblastom a: classi cat ion an d rep or t of 16 p at ien t s. Clin Or th op Relat Res 1982;167:222–235 Pu bMed 40. Beyer W F, Kü h n H. Can an osteoblastom a becom e m align ant? Virchow s Arch A Path ol Anat Histopathol 1985;408:297–305 Pu bMed 41. Berberoglu S, Ogu z A, Aribal E, At aoglu O. Osteoblastom a respon se to radioth erapy and ch em oth erapy. Med Ped iat r On col 1997;28:305–309 Pu bMed 42. Cam it t a B, Wells R, Segura A, Un n i KK, Murray K, Dun n D. Osteoblastom a respon se to ch em oth erapy. Can cer 1991;68:999–1003 PubMed

7 Aneurysmal Bone Cyst and Giant Cell Tumor Stefano Boriani, Stefano Bandiera, Riccardo Ghermandi, Simone Colangeli, Luca Amendola, and Alessandro Gasbarrini

■ Introduction Many bon e t u m ors in clu d e gian t cells in th eir cytoarch itect u re. Am ong t h em , an eu r ysm al bon e cysts (ABCs) an d gian t cell t u m ors (GCTs) are th e m ost frequen tly obser ved in th e spin e. Som et im es th ese t u m ors are di cu lt to di eren t iate from each oth er du e to sim ilar clin ical n dings, im aging n dings, an d debat able h istological pat tern . Th erefore, th is just i es th e in clusion of th ese t u m ors in th e sam e ch apter.

■ Aneurysmal Bone Cyst An eu r ysm al bon e cyst s w ere rst d escr ibed by Ja e an d Lich ten stein 1 in 1942, w h en it w as di eren t iated from h em an giom as an d ot h er t u m ors con t ain in g gian t cells. Accord in g to th e World Health Organizat ion (W HO), ABC is de n ed as a dest ruct ive, expan sile, ben ign n eop lasm of bon e, com p osed of m u lt ilocu lated blood- lled cyst ic spaces sep arated by con n ective tissue septa containing osteoclast-t ype giant cells, broblast s, an d react ive w oven bon e.2 Th e ABCs rep resen t abou t 1.0 to 1.4% of all p r im ar y bon e t u m ors. Th e m et a-ep ip hyseal region s of long bon es are m ost likely to be a ected, esp ecially th e fem u r an d t ibia.1 Most cases of ABC are iden t i ed in th e rst t w o decades of life, an d predom in an tly bet w een 10 an d 20 years of age. It rarely occu rs after 30

years of age, even th ough it h as been described in pat ien t s older th an 50 years old. Ap proxim ately 10 to 30% of ABCs involve th e m obile sp in e, m ost com m on ly in th e th oracic an d lu m bar spin e. Th e m ost com m on sym ptom is local pain . Sw elling is frequen t w h en an ABC arises from su per cial bon es, bu t th is rarely occu rs in th e spin e. ABCs m ostly arise from th e posterior elem en t s, but frequ en tly invade th e pedicles, th e ep idural space, an d th e ver tebral body, often resu lt ing in path ological fract u res an d n eu rologic com prom ise. Radiograph s t yp ically sh ow an exp an sile osteolyt ic cavit y w ith st ran ds of bon e form ing a bu bbly appearan ce. Th e frequ en t ly obser ved “level” im ages are caused by th e dou ble den sit y of th e cyst’s con ten t s (blood an d m em bran es) an d m ay be con sidered path ogn om on ic of ABC. Th e cor tex is often eggsh ell th in an d blow n ou t (Fig. 7.1). Th e lesion is often ext rem ely vascular, bu t th is feat u re is n ot often angiograp h ically ap preciated even th ough it is grossly apparen t , m ain ly due to th e large n u m ber of ver y th in feeding ar teries. Sp in al cord or n er ve root com p ression associated w it h n eu rologic d e cit s can be t h e p rom in en t in it ial sym ptom , as som e lesion s can presen t w ith ver y rapid grow th an d local aggressiven ess. Secon dar y ABC-like ch anges are seen in 10% of GCTs (Fig. 7.2), an d less frequ en t ly in osteoblastom as, ch on droblastom as, an d brou s dysplasias that have undergone hem orrhagic cystic ch anges. Th ese h em or rh agic areas can be ob -

72

Chapter 7

a

b

Fig. 7.1a,b An aneurysm al bone cyst at levels T5 and T6 in a 41-year-old man. (a) Typical magnetic resonance imaging (MRI) pat tern demonstrating well-de ned erosion of the posterior elements and part of the vertebral body and expanding into the soft tissue. There are multiple cystic images with double content of blood and m embranes, invasion of the canal, and compression of the cord. The patient reported severe pain on the Visual Analogue Scale as well as lower limb weakness. (b) Five selective arterial embolizations (SAEs) were perform ed, with no surgical treatment. A computed tomography (CT) scan performed 2 years after treatment demonstrates shrinking and ossi cation of the cyst and complete bone remodeling. There is no soft tissue extension and no canal encroachment. Pain receded after the rst SAE, and the patient gradually recovered his muscle strength. At the nal follow-up he had no pain and norm al muscle strength, and was able to perform his norm al activities.

Aneurysmal Bone Cyst and Giant Cell Tumor served also in prim ary m alignant tum ors (osteosarcom a, angiosarcom a) and even in m etastases. Treat m en t an d p rogn osis in th ese cases are sim ilar to th ose of a solid t u m or, an d accu racy of h istological diagn osis is m an dator y for ap propriate m an agem en t . Conversely, ABCs can contain variable am ount of solid t issue, w h ich is predom in an t in th e socalled solid ABC t u m or varian t . Su ch m ixt u re of solid an d cyst ic pat tern m akes it som et im es ext rem ely di cu lt to di eren t iate GCTs from ABCs, both gian t cell–rich t um ors. Th e u lt im ate di eren t iat ion can be don e via im m u n oh istoch em ical an d m olecu lar biology st u dies.

73

Th e ABCs som et im es h ave a self-lim it ing cou rse, evolving to sp on t an eou s h ealing ch aracterized by disappearan ce of th e dou ble con ten t and ending in ossi cation of the lytic areas. Tu rbu len t cou rses, on th e oth er h an d, can be obser ved w ith h uge m asses an d m assive bon e erosion . Pat h ological fract u res or sp in al cord or n er ve com p ression can follow, som et im es requ iring em ergen cy su rger y. It is com m on ly accepted th at in t ralesion al excision (cu ret t age) is t h e best t reat m en t for p r im ar y ABCs. Th is p roced u re is frequ en t ly risky an d som et im es com p licated by sign i can t in terop erat ive h em orrh age, w h ich can be

a

c

b

Fig. 7.2a–e A 26-year-old m an presented in 1985 with a giant cell tumor (GCT) and a huge aneurysm al bone cyst (ABC). (a) The huge ABC-like mass is expanding into the abdom en. (b) A cauda equina is compressed by the epidural tum or growth, which

was Enneking stage 3. An SAE was performed, which succeeded in reducing the mass, enabling a double-approach gross total excision, followed by conventional radiation therapy. (c) A CT scan 15 years later demonstrated no local recurrence.

74

Chapter 7

e

Fig. 7.2a–e (continued ) (d) A standard radiogram 15 years later. (e) A CT scan sagit tal reconstruction demonstrated loss of lordosis and fusion. The radiographic result is poor sagit tal alignment, but the patient does not complain about pain and participates in sports activities, such as running. At a 25-year follow-up phone contact, the patient reported light back pain, full activit y, and no tumor recurrence.

d

con t rolled by fu ll rem oval of t h e cyst -lin in g w all at t h e t im e of su rger y. In th e sp in e, fu ll su rgical rem oval of th e con ten t of an ABC can be ch allenging an d t im e-con sum ing, an d m ay exp ose th e p at ien t to p rofu se bleed ing. Many st u dies h ave at tem pted to redu ce su rgical m orbidit y an d at th e sam e t im e ach ieve fu ll local con t rol, w ith resu lt s in dicat ing th at p reop erat ive select ive em bolizat ion (w it h p olyvinyl alcoh ol or absorbable gelat in or cyan oacr ylate) can sign i can tly redu ce bleeding an d becom e itself a t reat m en t opt ion .3,4 Recu rren ce rates for ABC in th e sp in e h ave been repor ted in th e range of 10 to 25%. ABC is con sidered radiosen sit ive,5 an d radioth erapy can be u sed as sin gle t reat m en t or adju van t t h erapy, bu t rad iat ion can also p rod u ce ad verse e ects on grow th in ch ildren w ith consequent late deform ities. Additionally, it m ay h ave

delayed e ect s on th e sp in al cord (radiat ion in du ced m yelop athy) an d in crease th e risk for sarcom as. In som e st u d ies, m ore aggressive su rger y such as en -bloc resection h as been advocated as a possible opt ion ,6 con sidering th at an ext ralesion al surger y is associated w ith less blood loss an d w ith th e h igh est local con t rol rate. Th e required excision is frequen tly so dest r uct ive th at in st r um en ted fusion m ust be con sidered in th e operat ive p lan n ing.

■ Giant Cell Tumor According to th e W HO classi cat ion , GCT is a ben ign bu t locally aggressive prim ar y t um or, com p osed of proliferat ion of m on on u clear cells

Aneurysmal Bone Cyst and Giant Cell Tumor am ong w hich num erous m acrophages and large osteoclast-like gian t cells are scat tered.2 GCTs com prise 4 to 8% of all prim ar y bon e t um ors, an d it m ost com m on ly occu rs bet w een t h e secon d an d fou r th d ecade of life, w ith a sligh t fem ale p redom in an ce. Gian t cell t um or is m ost com m on ly foun d in th e ju xta-ar t icular m et aphysis of long bon es 7 an d occu rs rarely in t h e sp in e; t h e in cid en ce in t h e m obile sp in e ranges from 1.4 to 9.4%. If th e sacru m is in clu ded as p ar t of th e sp in e th e in ciden ce rises to 10%. GCT of th e spin e u su ally ar ises from t h e ver tebral bod ies, bu t m u lt icen t r ic sim u lt an eou s p resen t at ion h as been docu m en ted. Im aging pat tern s in clu de a m oth -eaten or irregu lar radiolu cen t erosive lesion (Fig. 7.3) w ith irregular th in sclerot ic react ion , som et im es exp an ding th e ver tebral con tou r. Tu m or exp an sion in th e soft t issu e or in th e can al can cau se d est ru ct ion of th e periph eral cor tex (Fig. 7.4). Secon dar y ABC-like ch anges are seen in 10% of GCTs (Fig. 7.2). Pain , m ostly occurring during th e n igh t , is th e m ost frequ en t sym ptom . Neu rologic sym p tom s due to foram en or spin al can al en croach m en t by th e t u m or m ass are less frequ en tly repor ted, as are path ological fract u res. How ever, th e erosive activit y of undetected GCT m ay ult im ately lead to p ath ological fract ures. Ever sin ce GCTs w ere iden t i ed, predict ing th eir beh avior h as been ch allenging.7 Various at tem pt s to develop staging system s for th e t um or h ave been u n successful.8 According to the Enneking staging system ,9 GCT can be found as st age 2 act ive or stage 3 aggressive ben ign lesions. Active lesions are characterized by fully erosive p at tern s t ypically occu rring in th e vertebral body, w ith w ell-de n ed borders. Stage 3 aggressive lesions have in distin ct m argins eroding th e cortex, expan ding in th e surroun ding soft t issu e, early invading th e epidu ral sp ace. Prolonged disease-free sur vivals h ave been repor ted after cu ret tage an d radioth erapy, alth ough som e p at ien t s m ay requ ire t w o or m ore additional procedures ow ing to local recurrence. On th e oth er h an d, en -bloc excision , w h en feasible, is curat ive. In terest ingly, GCTs can p roduce p u lm on ar y m et ast ases even w ith ou t h istological evid en ce

Fig. 7.3 A CT scan demonstrating a fully lytic change with thin osteosclerotic rim at the periphery in a 36-year-old m an. The margins are well de ned. No soft tissue mass is expanding outside the margin, and there is no epidural extension and no pathologic bone formation. This imaging is consistent with Enneking stage 2.

of m align an cy. Th ese m etast ases can be leth al in 25% of cases, bu t occasion ally m ay regress spon tan eou sly or be con t rolled by m argin al excision . In 20 p at ien t s w ith lu ng m et ast ases out of 671 pat ien ts w ith GCT obser ved at th e Mayo Clin ic, t w o p at ien t s died of th e disease. Select ive ar terial em bolizat ion (SAE) can decrease in t raop erat ive bleeding an d is th erefore m an dator y before in t ralesion al excision . Radiat ion t h erapy m u st be con sidered on ly as an adjuvan t after com plete in t ralesion al excision , bu t h igh -grad e sarcom as are est im ated to occu r in 5 to 15% of cases t reated by t h is m odalit y. Spin al GCTs h ave a con siderably poorer progn osis th an th ose in th e ap p en d icu lar skeleton , w ith recurren ce rates of up to 80% after t reat m en t . How ever, th is h igh rate is p ossibly m ore related to th e tech n ically dem an ding su rger y in th e spin e th an to som e in t rin sic feat ure of GCT of th e spin e. En -bloc resect ion is associated w it h t h e low est risk of local recu rren ce, bu t th is tech n ique en tails con siderable m orbidit y an d a n ot n egligible m or talit y. Con sidering th at th e on ly

75

76

Chapter 7

a

b

c

d

Aneurysmal Bone Cyst and Giant Cell Tumor su ccessfu l t reat m en t seem s to be en -bloc resect ion (w h ich is n ot alw ays feasible an d is burden ed by h igh m orbidit y an d m or talit y), th e p ossible role of local an d system ic adjuvan t th erap ies sh ou ld be invest igated . Cu r ren t ly, d en osu m ab (a m on oclon al an t ibody) is p art icularly p rom ising. Im aging n dings in clu de lyt ic ch anges both in ABCs an d in GCTs. No p ath ological bon e form at ion is detected. React ive bon e is frequ en tly obser ved as a sclerot ic border of variable th ickn ess arou n d th e t um ors, som et im es creat ing an e ect of sw ollen ver tebra. Th e t um or m ass is som et im es large, an d t h e erosive e ect on th e ver tebra can be exten sive. Magn et ic reson an ce im aging (MRI) best de n es th e t um or exten sion , th e am ou n t of solid com pon en t , an d th e cyst ic areas. Som e pat tern s of ABC im aging are path ogn om on ic. Mult iple bubble st ruct ures, each in clu ding dou ble con ten t (blood an d m em bran es, w h ose in ter face ch anges or ien t at ion according to th e gravit y), w ith a variable am oun t of solid radiolucen t t issu e is frequ en tly obser ved. St age 2 GCTs (so-called act ive according to th e En n eking st aging system ) h ave clearly de n ed m argin s, som et im es w ith an ossi ed bord er. St age 3 GCTs (also called “aggressive ben ign ”) h ave t ypically fuzzy borders, early ep idural in vasion, and radiolucent erosion of the vertebra. Even if im aging can be path ogn om on ic, h istological diagn osis sh ould alw ays be perform ed du e to th e p ossibilit y th at ABC is occu rring as a secon dar y lesion in GCT or even in a m align an t t u m or. Diagn osis sh ou ld alw ays be ach ieved by histological and im m unohistoch em ical studies. Biopsy sh ou ld be perform ed by t rocar u n d er com p u ted tom ograp hy (CT) scan con t rol. Th is tech n iqu e n ot on ly red u ces m orbid it y an d

t u m or con t am in at ion of su rrou n ding soft t issu es, bu t also en ables taking th e m ost rep resen tat ive sam p le for th e path ologist , as th e CT scan can iden t ify th e area of solid t issu e w h ere en ough m aterial for di eren t ial diagn osis w ill be foun d. Considering that both spontaneous healing10 an d progression to m align an cy 11 are reported in th e literat ure, as w ell as secondar y t um or, su ch as leu kem ia, occu r r in g after t reat m en t w it h radiat ion th erapy, th e decision -m aking p rocess for select ing th e m ost app ropriate t reat m en t sh ould in clude a ver y careful evaluat ion of th e risk–ben e t rat io of each possible t reat m en t . Opt ion s for t reat in g ABCs h ave in clu d ed sim p le in t ralesion al excision w it h or w it h ou t xat ion an d bon e graft ing, gross total excision , en -bloc resect ion ,6 em bolizat ion ,4 in t racyst ic inject ion s,12 radiat ion th erapy,5 or a com bin at ion of t h ese m et h od s. Th ere is n o evid en ce of a bet ter ou tcom e from any of th e p rop osed t reat m en t s, bu t several com p licat ion s w ere associated w ith m ore aggressive m odalit ies of t reat m en t , in cluding m assive bleeding, lim itat ion of grow th an d range of m otion in cases of ar th rodesis, secon dar y m align an cies after rad iat ion , an d even fatal even t s. Th e t reat m en t associated w ith th e low est recurren ce rate is en -bloc resect ion ,6 bu t th is t reat m en t p ossibly exp oses th e p at ien t to h igh su rgical m orbidit y an d th eoret ically m u st be con sidered as an over t reat m en t , con sidering th e ben ign h istological pat tern . En -bloc resect ion sh ou ld be p er for m ed on ly in ver y select cases, for exam p le in p oster iorly located lesion s th at are less tech n ically dem an ding an d th at exp ose th e p at ien t to less blood loss com pared w ith in t ralesion al excision .

Fig. 7.4a–d (opposite) An Enneking stage 3 GCT at level L3 in a 37-year-old wom an. (a) Lytic erosion of the vertebral body is seen, with soft tissue tum or expanding into the foramen, into the canal, and into the surrounding soft tissues. The patient reports back pain and weakness in the lower limbs. (b) An en-bloc resection was performed in 1994 by the double approach. The margins were de ned as

tumor free (marginal margin). Reconstruction was perform ed by allograft and anterior plate and posterior xation with the hardware available at that time. (c) Standard radiogram at 16-year follow-up. The posterior system has been replaced due to screw failure. (d) A CT scan at 19-year follow-up shows no evidence of local recurrence and fusion of the graft.

77

78

Chapter 7 Radiat ion th erapy h as also been u sed w ith good resu lt s, but due to th e risk of radiat ion in du ced sarcom a an d cord m yelop athy, th is procedu re sh ou ld be con sidered on ly w h en all oth er t reat m en t s fail.4,5 In th e literat ure, SAE w as rst u sed to redu ce th e in t raop erat ive bleeding, an d w as palliative or cu rat ive in ver y select cases.13 Th e rst case fu lly t reated w ith SAE w as rep or ted in 1990 by DeRosa et al.3 Later, th ere w ere several rep or t s of good result s in pat ien t s t reated on ly w it h SAE. In a recen t ly p u blish ed ser ies of seven cases p rosp ect ively su bm it ted to sequ en t ial em bolizat ion s,4 n o case required surger y an d all evolved to full h ealing (Fig. 7.5). Th e tech n iqu e in clu ded m icrocath eterizat ion an d in fusion of em bolic agen ts on ly in to th e feeding arteries via the m icrocatheter, thus sparing n orm al, un involved ar teries. Most of th ese cases, h ow ever, requ ired m u lt ip le session s— som et im es as m any as seven em bolizat ion s ever y 2 m on t h s. Th e longer t im e requ ired to perform SAE an d th e exposu re to radiat ion du e

to rep eated angiograp h ies an d CT scan n ing m u st be con sid ered in th e d ecision -m akin g process. On e of th e lim it at ion s of SAE is th e risk of em bolizat ion of th e ar ter y of Adam kiew icz. In case a feeding ar ter y is detected as a bran ch of th e ar ter y to be em bolized, th e procedu re m u st be abor ted an d conver ted to surger y. An exh aust ive literat u re review w as perform ed in 2009 on 482 ar t icles on ABC.14 On ly 94 ar t icles per t ain ed to th e spin e or ver tebrae, m ostly con sist ing of case rep or t s an d sm all case series. Based on su ch “ver y low qu alit y” literat u re, th erefore, a w eak recom m en dat ion suggested select ive ar terial em bolizat ion as a st an d -alon e m odalit y, w h ereas a st ron g recom m en dat ion cou ld be given for select ive arterial em bolizat ion as a preoperat ive adjuvan t opt ion , as it facilit ates resect ion by reducing in t raoperat ive blood loss. In con t rast , radiat ion an d ch em oth erapy w ere w eakly recom m en ded as addit ion al th erapy after in com plete resect ion an d recu rren ce.

a

b

Fig. 7.5a,b An ABC at level L3 in a 21-year-old m an. (a) Persistent and increasing anterior thigh pain suddenly evolved into severe m otor weakness of the quadriceps. A CT scan at presentation shows the extension of the tumor in the soft tissues. The tumor is fully lytic, with a well-de ned border in the vertebral body. (b) Sequential SAE was perform ed every 4 to 6 weeks. Pain disappeared after the

second SAE and the patient’s m otor strength gradually recovered. A total of seven embolizations were performed. A CT scan performed 42 m onths after the last treatment session demonstrates shrinkage of the mass, full ossi cation, and no instabilit y. The patient has had no recurrence, and enjoys full functioning.

Aneurysmal Bone Cyst and Giant Cell Tumor Th e beh avior of GCT is u np redict able based on cu r ren t ly available dat a. Rock 15 d em on st rated t h e u t ilit y of t h e En n ekin g st agin g 9 in predict ing local recu rren ces in th e ext rem it y, w ith En n eking st age 3 t um ors h aving a h igher local failure rate. A higher recurrence rate in patients under th e age of 30 has been frequently repor ted.16 Alth ough it is n ot clear from th e data, it is p ossible t h at t h e m ore aggressive beh avior n oted in p repu ber t al su bject s (Fig. 7.6) an d in pregn an t w om en m ay be con n ected w ith h orm on al in u en ces. How ever, gen d er is n ot con sidered as a risk factor for local recu rren ce. GCT can also occu r in p at ien t s w ith Paget’s disease, som et im es d evelop ing large an d h igh ly vascu larized m asses (Fig. 7.7) requiring accurate h istological diagn osis in ord er to r u le out secon dar y sarcom a. In t h e literat u re, GCTs in t h e ext rem it ies h ave been t reated su rgically w it h en -bloc excision or by in t ralesion al excision com bin ed w ith local adjuvan t s such as ph en ol, hydrogen peroxide, an d cr yosu rger y or w ith th e m argin s exten ded by th e u se of a h igh -sp eed bur.17 W h eth er in t ralesion al excision or an en bloc w ide excision is perform ed depen ds on th e locat ion of th e t u m or, it s size, an d th e presen ce or absen ce of soft t issu e involvem en t . A h igh er risk of recu rren ce h as been repor ted w ith intralesional procedures as com pared w ith w ide resect ion s, alth ough Gitelis et al17 repor ted sim ilar local con t rol rates w ith en -bloc procedures versus intralesional surgeries, w hich w ere m oreover associated w ith few er com plicat ion s an d bet ter fu n ct ion al ou tcom es. Th e recu rren ces of GCT of t h e lim bs are u su ally su ccessfu lly m an aged w ith fu r t h er su rger y, an d th is fu r th er su p por t s th e recom m en dat ion in favor of less aggressive su rger y. How ever, in GCT of th e sp in e, th e t reat m en t of th e recurren ce is n ot as safe an d e ect ive as in th e lim b, due to it being a m ore tech n ically ch allenging t reat m en t . In th ese cases, en -bloc resect ion m ay be a bet ter opt ion , as local con t rol is a prim ar y goal.18 A st udy involving 49 pat ien ts fou n d a rate of recu rren ce of 22% u p to 60 m on th s after in dex su rger y.16 En -bloc resect ion w as associated w ith bet ter local con -

t rol w ith En n eking stage 3 t u m ors (p = 0.01) an d in t ralesion al resect ion p rovided adequ ate con t rol of En n eking st age 2 t um ors. Alt h ough t h e m or t alit y an d m orbid it y of en -bloc resect ion in th e spin e are w ell kn ow n , th e risks of en -bloc resect ion m u st be w eigh ed again st th e risks of local recu rren ce an d p rogression of m align an cy. GCTs of th e sp in e m ay be life-th reaten ing, eviden ced by a recen t report of t w o pat ien ts w h o died from p rogression an d lung m et ast ases of th eir GCTs, an d both of th em h ad h ad a local recu rren ce.16 Th e exp ected su r vival after p u lm on ar y m et ast asis from GCT is arou n d 17%. It is also rem arkable th at 80% of th e p u lm on ar y m et ast ases are associated w ith a local recu rren ce. Radioth erapy is st ill con t roversial. Excellen t result s are reported after m egavolt age radiat ion th erapy for axial or in operable GCTs (w ith local con t rol rates as h igh as 77%). Ch akravar t i et al19 reported an 85% rate of progression -free sur vival in cases of in operable GCTs follow ing radiat ion t reat m en t , an d foun d n o cases of m alignant transform ation after a m edian follow -up t im e of 9.3 years. Later, Cau dell et al20 advised caut ion using radiat ion for local con t rol in previously t reated pat ien t s. Th ey repor ted a local control rate of 42%, contrasting w ith a local cont rol rate of 82% in prim arily radiated pat ien t s. Th e overall local con t rol rate w as 62% at 5 years. Miszczyk 21 reported on a series of GCTs t reated eith er postoperat ively w ith radiat ion or w ith radiat ion alon e. Th e st u dy rep or ted a local con t rol rate of 83% w h en radiat ion w as u sed as an adjuvan t an d 69% w h en u sed alon e. How ever, th ese di eren ces w ere n ot stat ist ically sign i can t . In an oth er st u dy, conven t ion al radiat ion did n ot im prove th e local con t rol rate after in t ralesion al excision com p ared w ith in t ralesion al excision alon e.16 Th e risk of p ost rad iat ion sarcom a is of p art icular con cern in pat ien ts w ith GCT. St u dies from th e Mayo Clin ic 22 reported a 17% rate of secon dar y sarcom a ar ising in p reviou sly ir radiated GCTs of t h e sp in e. In t h e p reviou sly reported review, on e p at ien t developed an osteogenic sarcom a 18 m onths after being treated w ith an in t ralesion al resect ion an d adjuvan t radiat ion th erapy (4,200 Gy).16 Non eth eless, as

79

80

Chapter 7

a

Fig. 7.6a,b A 12-year-old prepubertal girl was referred in 1992 with incomplete paraplegia that was formerly diagnosed as eosinophilic granulom a. (a) Early vertebral body collapse with soft tissue expansion in the canal. The canal encroachment is not consistent with the diagnosis of eosinophilic granuloma. The biopsy led to a diagnosis of GCT. An intralesional excision and xation was performed. Her neurologic symptoms recovered, but ve local recurrences followed, with t wo further episodes of m otor and sensory paraplegia, requiring progressively more aggressive surgeries, including doubleapproach gross total excisions. Several complications occurred, including cerebrospinal uid (CSF) leakage in the pleural cavit y and failure of the posterior implant. Radiation therapy and a course of chemotherapy were performed after the last excision. (b) Radiographic image at 21-year follow-up (19 years after the last surgery) shows no evidence of disease.

Fig. 7.7a–e (opposite) A 54-year-old m an presented in 2010 with a huge GCT evolving from Paget’s disease (diagnosed 10 years earlier) in a 54-year-old man. (a) An MRI demonstrates a huge soft tissue mass expanding into the retroperitoneal and epidural space, with fuzzy margins. It is categorized as Enneking stage 3. (b) The mass extends over three vertebrae. (c) A CT-guided biopsy nds a

b

benign GCT tumor. An SAE and a double-approach gross total excision are performed. A recurrence occurred after 8 m onths and was treated by a posterior-approach intralesional excision. Then denosumab was started in January 2011. (d,e) Scan at 3-year follow-up (July 2013). There has been no local recurrence, and the patient is still on Denosumab.

Aneurysmal Bone Cyst and Giant Cell Tumor

81

a

b

c

e

d

82

Chapter 7 m et h od s of rad iat ion d eliver y h ave evolved , th e risk of secon dar y sarcom a h as dim in ish ed an d w ill probably be even low er in th e fut ure, gran t ing a relevan t role to radioth erapy for in operable lesion s. Em bolizat ion as a prim ar y th erapy h as been recen tly prop osed also for GCT, bu t on ly a few cases of sacral GCT h ave been rep or ted yet . Gian t cell t u m or is rich in osteoclast-like gian t cells an d con t ain s m on on uclear st rom al cells t h at exp ress RANK ligan d (RANKL), a key m ediator of osteoclast form at ion an d act ivat ion . It is p ossible t h at th e recr u it m en t of osteoclast-like gian t cells is related to st rom alcell exp ression of RANKL an d t h at t h e gian t cells are resp on sible for t h e aggressive osteolyt ic act ivit y of t h e t u m or. Th u s, t h e m ost at t ract ive act u al p e rsp e ct ive con ce r n s t h e t h e rap e u t ic role of d e n osu m ab, a h u m an m on oclon al an t ibody th at speci cally in h ibit s RANKL, thereby in hibiting osteoclast-m ediated bone destruction . Th e prelim inar y results show a su bst an t ial red u ct ion of t h e p rop or t ion of gian t cell an d a sign i can t in crease of m at ure bon e.22 The satisfactory results initially obtained w ith den osu m ab in th e t reat m en t of GCT an d th e clear im m u n oh istoch em ical sim ilarit y an d relat ion sh ip bet w een GCT an d ABC ju st ify th e hypoth esis th at den osum ab m ay also h ave posit ive e ect s on ABC.23 An exh aust ive literat ure review 14 w as perform ed in 2009, n ding 3060 ar t icles on GCT. On ly 178 ar t icles per t ain ed to th e spin e or vertebrae, m ostly con sist ing of case repor t s an d sm all case series. On ly fou r ar t icles w ere iden t i ed w ith case series larger th an 10 pat ien ts. Th ese abst ract s w ere all ver y low qu alit y literat u re, an d n o ran dom ized or p rospect ive literat u re w as available. A st rong recom m en dat ion (on ver y low qualit y eviden ce) can be m ade for en -bloc resect ion as t h e best t reat m en t of GCT, w h en feasible based on staging, for both prim ar y (En n eking st age 3) an d recu rren t GCT of th e th oracic an d lu m bar sp in e. A w eak recom m en dat ion can be m ade for serial clin ical an d radiograp h ic obser vat ion for residual tum or as the best strategy in m anaging in com pletely resected GCT. A w eak recom m en dat ion (based on ver y low qu alit y eviden ce)

can be m ade for radiat ion th erapy as a t reatm en t opt ion for recu rren t GCT.

■ Problems to Avoid Diagnosis Planning the Treatment Without a Careful Histological and Immunohistochemical Diagnosis Not w ith st an ding th e path ogn om on ic elem en t s rep orted describing im aging st udies, h istology is m an dator y before p lan n in g t reat m en t . A radiograph ically p resum ed ABC u n resp on sive to em bolizat ion cou ld be a GCT on h istological st u dies. A t rocar biop sy from a cyst ic area in a GCT can m islead th e path ologist in to doing a h istological an alysis, result ing in th e diagn osis of an ABC—or an ABC-like area w ith in a GCT. More rarely, ABC-like areas can be fou n d even in m align an t t u m ors, associated w ith an even w orse progn osis. Misdiagn osis or late diagn osis can th reaten th e life of th e pat ien t .

Treatment Performing an Intralesional Excision of ABC and GCT Without Preoperative Embolization Th is sit u at ion sh ou ld be absolu tely avoided, as both t um ors are h igh ly vascularized. Em bolizat ion is proven to sign i can tly m in im ize th e risk of in t raoperat ive u n con t rolled bleed ing, w h ich can be dangerous an d life-th reaten ing. Excessive bleeding m ay also p reven t com plete tum or resection particularly in the spine, w here direct pressure is n ot alw ays possible.

Intralesional Partial Excision of a GCT Th is tech n iqu e exp oses th e pat ien t to a h igh risk of local recu rren ce, ranging from 20 to 30% (Fig. 7.6). Alth ough th e t reat m en t of recurren ces in th e ext rem it ies is relat ively safe, th e t reat m en t of a recu rren ce in th e spin e is di cult , less e ect ive, an d en tails h igh er m orbidit y (ep id u ral scar, d u ra alm ost im p ossible to

Aneurysmal Bone Cyst and Giant Cell Tumor release). Th e rate of fu r t h er recu r ren ce is n ot n egligible.

En-Bloc Resection of ABC Th is p rocedu re exposes th e p at ien t to u n n ecessar y m orbidit y. Th e rate of recu rren ce follow ing en-bloc resection of an ABC is negligible, but m or t alit y an d m orbidit y of en -bloc resect ion sh ou ld alw ays be con sidered.

Not Considering New est Options Th e ABCs can su ccessfu lly be t reated by sequ en t ial em bolizat ion . Alth ough exposu re of th e p at ien t to radiat ion for m u lt iple angiograph ies an d frequ en t CT scan s is n ot n egligible, it seem s less dangerou s t h an in t ralesion al excision , p ar t icu larly if con sidering th e risk of local recu rren ce an d possible secon dar y dam ages to th e spin al cord an d th e m orbidit y of su rger y in previou sly t reated p at ien ts. Denosum ab (a m onoclonal hum an antibody, RANKL in h ibitor) can be a viable opt ion for recurren t GCT an d ABC even th ough long-term resu lt s are st ill u n clear. It s role cou ld be also relevan t as a preop erat ive adjuvan t to redu ce th e t u m or m ass an d to allow easier an d m ore e ect ive excision .

■ Chapter Summary An eur ysm al bon e cysts an d gian t cell t um ors are gian t cell–rich t um ors th at m ay involve th e spin e. Both presen t as lyt ic lesion s on X-rays an d CT scan s an d can evolve in to large dest ru ct ive m asses. Cyst ic areas lled w ith dou ble con ten t (blood an d m em bran es) are t ypical of ABCs an d possibly obser ved in about 10%GCTs. ABC-like areas can addit ion ally be obser ved in a n um ber of ben ign an d m align an t con dit ion s (“secon dar y” ABC). Di eren t iat ion bet w een th ese t w o t u m ors is som et im es di cu lt on im aging st u dies, an d som et im es h istology can n ot clearly di eren t iate th em . Th e de n it ive diagn osis is m ade via im m u n oh istoch em ist r y. Select ive ar terial em bolizat ion is a valid tech n ique for th e t reat m en t of ABCs, even if in m ost of th e cases m u lt ip le p rocedu res are re-

quired. It sh ould be con sidered th e rst opt ion becau se it h as th e best cost–ben e t rat io. It is indicated in intact an eur ysm al bon e cysts, w h en diagn osis is cert ain , w h en it is tech n ically feasible an d safe (n o feeding arter y to th e cord), an d w h en n o p at h ological fract u re or n eu rologic involvem en t is fou n d . If em bolizat ion fails, ot h er opt ion s for t reat m en t (m ost ly in t ralesion al excision ) w ould st ill be available. On cological an d surgical staging are h elpful for th e deciding on th e t reat m en t of a GCT of the spine. Staging is based on CT and MRI scans, on ce a h istological d iagn osis is ach ieved by t rocar biop sy u n der CT scan con t rol. En -bloc resect ion sh ou ld be con sidered for En n eking stage 3 GCTs of th e spin e. Th e ch oice of en -bloc resect ion m ust be balan ced w ith th e in h eren t risks of the procedure, including m orbidit y, m ort alit y, an d fu n ct ion al loss possibly required by th e sacri ce of an atom ic st r uct ures represen ting th e m argin . In cases of requ ired in ten t ion al t ran sgression or of acciden tal t ran sgression , adjuvan t th erap ies sh ou ld be con sidered, st ar ting w ith radiation therapy. Intralesional excision of En n eking stage 2 t u m ors u sually provides adequ ate local con t rol, an d p at ien t s sh ou ld be follow ed for at least 5 years becau se local relapse can occur late. Denosum ab (a hum an m onoclonal antibody) can be an in terest ing opt ion for GCT an d ABC, bu t th ere are con cern s abou t th e progression of th e t um ors after stopping th e t reat m en t . Its role could be also relevan t as a preoperat ive adjuvan t to redu ce th e t u m or m ass an d to allow easier an d m ore e ect ive excision .

Pearls The outcome of ABCs and GCTs can be heavily a ected by a late or incorrect diagnosis or by inadequate treatm ent. Although 75% of ABCs occur in the rst t wo decades of life, 85% of GCTs occur in patients older than 19. ABCs most comm only arise from the posterior vertebral elements, whereas GCTs are t ypically a tumor of the vertebral body. ABCs are characterized by a double- lled cystic pat tern, whereas GCTs are m ostly solid. Both occur m ore frequently in females.

83

84

Chapter 7 Pitfalls Planning the treatment without a careful histological and imm unohistochem ical diagnosis Performing intralesional excision of ABCs or GCTs without preoperative embolization Intralesional partial excision of a GCT En-bloc resection of ABCs Not considering newest options Not considering newest options

Acknow ledgment Th e au th ors are in debted to Carlo Piovan i for design , arch ive research , an d im age an d editorial assistan ce.

Refere nces Five Must-Read Refe rences 1. Ja e HL, Lichten stein L. Solit ar y un icam eral bon e cyst s: w ith em ph asis on th e roen tgen p ict u re, th e path ologic app earan ce an d th e p ath ogen esis. Arch Su rg 1942;44:1004–1025 2. Fletcher CDM, Bridge JA. Hogen doorn PCW, Merten s F. W HO Classi cat ion of Tum ours of Soft Tissue an d Bon e. Lyon , Fran ce: In tern at ional Agen cy for Research on Can cer (IARC), 2013 3. DeRosa GP, Grazian o GP, Scot t J. Ar terial em bolizat ion of an eur ysm al bon e cyst of the lum bar spin e. A report of t w o cases. J Bon e Join t Surg Am 1990; 72:777–780 PubMed 4. Am en dola L, Sim on et t i L, Sim oes CE, Ban diera S, De Iu re F, Borian i S. An eu r ysm al bon e cyst of t h e m obile spin e: th e th erap eu t ic role of em bolizat ion . Eu r Spin e J 2013;22:533–541 PubMed 5. Maeda M, Tateish i H, Takaiw a H, Kin osh it a G, Hat an o N, Nakan o K. High -en ergy, low -dose radiat ion th erapy for an eur ysm al bon e cyst . Report of a case. Clin Orth op Relat Res 1989;243:200–203 Pu bMed 6. Zileli M, Isik HS, Ogut FE, Is M, Cagli S, Calli C. An eur ysm al bon e cyst s of the spin e. Eur Spin e J 2013;22: 593–601 Pu bMed 7. Ja e HL, Lich ten stein L, Port is RB. Gian t cell t um or of bone: it s path ologic appearan ce, grading, supposed varian t an d t reat m en t . Arch Patol 1940;30:993– 1031 8. Cam pan acci M, Giu nt i A, Olm i R. Gian t cell t um or of bon e: a st u dy on 209 cases w ith long term follow -u p . In 130. It al J Or th op Trau m atol 1975;1:249–277 9. En n eking W F. A system of st aging m u sculoskelet al neoplasm s. Clin Orth op Relat Res 1986;204:9–24 PubMed 10. Malgh em J, Maldagu e B, Esselin ckx W, Noel H, De Nayer P, Vin cen t A. Spon t an eous healing of an eur ysm al bon e cyst s. A repor t of th ree cases. J Bon e Join t Su rg Br 1989;71:645–650 PubMed 11. Kyriakos M, Hardy D. Malign an t t ran sform at ion of an eu r ysm al bon e cyst , w ith an an alysis of th e literat ure. Can cer 1991;68:1770–1780 PubMed 12. Adam sbaum C, Mascard E, Guinebretière JM, Kalifa G, Dubousset J. In t ralesion al Eth ibloc inject ion s in pri-

m ar y an eur ysm al bon e cyst s: an e cien t an d safe treatm en t. Skeletal Radiol 2003;32:559–566 PubMed 13. Koci TM, Meh ringer CM, Yam agat a N, Ch iang F. An eu r ysm al bon e cyst of th e th oracic spin e: evolut ion after p ar t icu late em bolizat ion . AJNR Am J Neu roradiol 1995;16(4, Su p pl):857–860 Pu bMed 14. Harrop JS, Schm idt MH, Borian i S, Sh a rey CI. Aggressive “ben ign ” prim ar y spine n eoplasm s: osteoblastom a, aneur ysm al bone cyst, and giant cell t um or. Spin e 2009;34(22, Su ppl):S39–S47 Pu bMed 15. Rock M. Curet t age of gian t cell t um or of bon e. Factors in uen cing local recurren ces an d m et ast asis. Ch ir Organ i Mov 1990;75(1, Su pp l):204–205 Pu bMed 16. Boriani S, Ban diera S, Casadei R, et al. Gian t cell t u m or of th e m obile sp in e: a review of 49 cases. Spin e 2012;37:E37–E45 Pu bMed 17. Gitelis S, Mallin BA, Piasecki P, Tu rn er F. In t ralesion al excision com p ared w ith en bloc resect ion for gian tcell t um ors of bone. J Bon e Join t Surg Am 1993;75: 1648–1655 Pu bMed 18. Sten er B, Joh n sen OE. Com p lete rem oval of th ree vertebrae for gian t-cell t u m ou r. J Bon e Join t Su rg Br 1971;53:278–287 Pu bMed 19. Ch akravart i A, Spiro IJ, Hug EB, Man kin HJ, E rd JT, Su it HD. Megavolt age radiat ion th erapy for axial an d inoperable giant-cell t um or of bon e. J Bon e Join t Surg Am . 1999 Nov;81(11):1566-1573 20. Caudell JJ, Ballo MT, Zagars GK, et al. Radioth erapy in th e m an agem en t of gian t cell t um or of bon e. In t J Radiat On col Biol Phys 2003;57:158–165 21. Miszczyk L, Wydm an ski J, Spin del J. E cacy of radioth erapy for gian t cell t u m or of bon e: given eith er postoperat ively or as sole t reat m en t . Int J Radiat On col Biol Phys 2001;49:1239–1242 22. Th om as D, Hen sh aw R, Sku bit z K, et al. Den osu m ab in pat ien t s w ith gian t-cell t um our of bon e: an open label, ph ase 2 st udy. Lan cet On col 2010;11:275–280 PubMed 23. Lange T, Steh ling C, Froh lich B, et al. Den osum ab: a poten t ial n ew an d in n ovat ive t reat m en t opt ion for an eu r ysm al bon e cyst s. Eu r Sp in e J 2013;22:1417– 1422

8 Chordoma Jackson Sui, Patricia L. Zadnik, Mari L. Groves, and Ziya L. Gokaslan

■ Introduction Ch ordom as are a rare, ben ign t u m or arising from rem n an t n otoch ordal cells. Th ese lesion s occu r m ost com m on ly in th e clivu s an d sacral spin e, at th e em br yological en d poin ts of th e n otoch ord. Ch ordom as are m ost often locally invasive an d aggressive; h ow ever, m et astases to th e lungs an d oth er sites are n ot un com m on . Furth er, t um ors in th e sacral spin e m ay grow to large sizes w ith in th e pelvis an d abdom en , disru pt ing th e bow el an d bladder, an d m otor an d sexu al fu n ct ion . Alt h ough som e d r ug th erapies are un der invest igat ion , ch ordom as are classically resist an t to ch em oth erapy an d conven t ion al rad iat ion , an d en -bloc su rgical excision or gross tot al resect ion w it h su bsequ en t radioth erapy o ers t h e best ou tcom es for p at ien t s.1–4 Ch ordom as occu r m ost frequ en tly in m iddleaged m en . According to th e Su r veillan ce, Epidem iology, an d En d Resu lts (SEER) dat abase, w hich review ed th e outcom es of 414 ch ordom a pat ien ts, th e average p at ien t age is 59.9 years, w ith a sligh t m ajorit y (62.8%) of p at ien t s being m ale.5 Ch ordom as are rare in th e pediatric p opu lat ion . Classically, p at ien t s w ith local disease d o w ell after su rger y, w ith a m edian su r vival of 75 m on th s rep or ted in recen t literat u re.5,6 Pat ien ts w ith m et astat ic disease h ave sign i can t ly decreased su r vival (m edian 24 m onth s), com pared to a prolonged m edian su rvival for prim ar y sp in al ch ordom a p at ien t s.7

Based on n at ural h istor y data, 5- an d 10year su r vival rates for ch ordom a pat ien t s are 62 to 67% an d 40 to 45%, respect ively.5 How ever, for pat ien t s t reated in m odern case series w it h en -bloc resect ion an d adju van t rad io th erapy, 5-year overall su r vival (OS) h as been rep orted even h igh er for spin al ch ordom as.8–11 Th is ch apter review s th e diagn osis an d t reat m en t of spin al ch ordom a, as w ell as th e h istological an d gen et ic ch aracterist ics th at facilitate diagn osis an d are d riving n ovel th erap eu t ics. Cu rren t literat u re an d classes of eviden ce sup por t ing th e u se of en -bloc resect ion an d adju van t th erapy are review ed .

■ Clinical Presentation Clin ically, pat ien ts w ith ch ordom a m ost often presen t w ith p ain . For p at ien t s w ith cer vical sp in e ch ord om a, p ain on n eck exion or ro t at ion is m ost com m on . Myelopathy m ay be presen t if th ere is spin al cord com pression . Th oracic an d lu m bar ch ord om a p at ien t s classically h ave local m ech an ical pain . Am ong all ch ord om as, sacral ch ord om as are t h e m ost com m on follow ed by lu m bar ch ordom as. For pat ien ts w ith sacral ch ordom a, bow el or bladd er dysfu n ct ion , sad d le an est h esia, an d foot d rop or ot h er gait abn or m alit ies com m on ly occu r, accom pan ied by local sacral pain . Neurologic com prom ise is less com m on for t u m ors

86

Chapter 8 in th is region . Ch ord om as in th e sacral sp in e grow slow ly, gradu ally com p ressing th e ab dom in al viscera, an d pat ien t s m ay com plain of ch ron ic u rin ar y reten t ion , con st ip at ion , or p ain th at h as progressively w orsen ed over several m onths or years. For patients w ith large tum ors, a m ass m ay be palpable aroun d th e but tock. Furth er, sciat ic t ype pain is n ot u n com m on , radiat ing dow n th e back of th e leg to th e bot tom of th e foot . In a m iddle-aged p at ien t presen t ing w ith sacral pain an d sym ptom s of cauda equin a, com p u ted tom ograp hy (CT) an d m agn et ic reson an ce im aging (MRI) sh ou ld be perform ed to invest igate th e presen ce of a m ass. On im aging, a ch ordom a m ay h ave a h eterogen eously calcied app earan ce on p lain radiograp h s. CT scan s are m ore e ect ive at capt uring bony rem odeling an d dest ru ct ion associated w ith ch ordom a grow t h , an d evid en ce of w iden ed foram in a or th in n ing of cor t ical bon e m ay be eviden t . Overall, MRI is th e preferred im aging m odalit y for exp loring soft t issu e exten sion associated w ith ch ord om a. Un like ot h er t u m ors of t h e sp in e, ch ordom as invade th e disk sp aces (Fig 8.1). On T1-w eigh ted im ages, ch ordom as appear hypo-

to isoin ten se to t h e ver tebral bod ies, an d m asses are likely to exten d to adjacen t levels an d in to th e paraspin al m usculat ure. Ch ordom as ap pear iso- to hyp erin ten se to th e bon e on T2-w eigh ted im ages (Fig. 8.1a) due to a h igh con ten t of m u cin or ch on droid m at rix, an d these lesion s m ay h ave m oderate to robust con t rast en h an cem en t . Fat-su p p ressed MRI m ay be h elp fu l to bet ter d elin eate t u m or m argin s (Fig. 8.1b).

Case Illustration A 57-year-old m an presen ted w ith a h istor y of th oracic ch ordom a exten ding from th e C7-T1 jun ct ion to th e T3-4 disk space (Fig 8.1). He w as in it ially in form ed th at th is w as un resectable due to its locat ion , an d h e w as t reated w ith t radit ion al extern al beam radiat ion th erapy an d stereotact ic radiosurger y; h ow ever, h e con t in ued to h ave progressive m yelopathy. Th e pat ien t w as th en o ered staged surger y for en -bloc resect ion of th e t u m or. A posterior in cision w as m ade from C1 to T9 an d su bp eriosteal d issect ion w as com p leted from C2 to T8. Exten sive dissect ion w as m ade ben eath th e

a

b

Fig 8.1a,b A heterogeneously hyperintense tumor was noted on T2-weighted (a) and fat-suppressed (b) imaging to extend from the C7-T1 junction to

the T4 disk space with signif cant paraspinal extension. The lesion was proven to be a chordoma on biopsy.

Chordoma scap u la bilaterally to visu alize 7 cm from t h e m id lin e. Cer vical lateral m ass screw s w ere p laced at C3- C6, w ith a rod in ser ted across th e spin ous process of C2 (Fig 8.2a). Th oracic p edicle screw s w ere placed at T5-T8 an d t h e ribs w ere dissected circum feren t ially an d disart icu lated at T2-T4 bilaterally. Care w as taken to leave th e t u m or in tact . Bilateral total lam in ectom y of C7-T4 w as perform ed, as w ell as rem oval of th e C7 lateral m ass an d th e C7-T1 join t . Th e T1-T3 n er ve roots w ere ligated an d

87

t ran sected. Addit ion al dissect ion of th e esoph agu s an d vessels w as m ade an d t h ese st r u ct u res w ere p rotected , an d a Tom it a saw w as in t roduced at th e T3-T4 disk space (Fig. 8.2a). Tw o con tou red , t ap ered rod s w ere p laced bilaterally w ith con n ectors. For th e secon d st age of th e p rocedu re th e follow ing day, an an terior cer vical in cision w as m ade an d exten ded th rough th e cer vicoth oracic jun ct ion . Th e C7-T1 disk space w as exposed an d th e disk excised, discon n ect ing th e

a

b

Fig 8.2a,b A computed tomography (CT) reconstruction of instrumentation following cervicothoracic chordom a resection. (a) Cervical lateral mass screws were placed at C3-C6, with a rod inserted across the spinous process of C2. Bilateral total laminectomies were performed at C7-T4. Thoracic pedicle screws were placed at T5-T8. A Tom ita saw

can be seen at the T3-T4 disk space. Two contoured, tapered rods were placed bilaterally with connectors. (b) Postoperative view of instrumentation with the cage visible anteriorly. The T3-T4 disk space was transected with the Tomita saw, and a C7-T1 discectomy facilitated en-bloc removal of the mass. A cage was packed with allograft bone from C7-T4.

88

Chapter 8 cer vical an d th oracic spin es. Th e pat ien t w as then placed in th e left lateral decubit us position w ith h is h ead in a May eld th ree poin t xator. Th e p osterior m idlin e in cision w as op en ed an d th e lat issim u s dorsi m u scle an d scap u la w ere re ected to visu alize th e ch est w all. Th oracic su rger y assisted w ith th e ch est w all exp osu re, w it h a t h oracotom y bet w een t h e fou r t h an d ft h r ibs, to visu alize t h e lu n g. Th e lu n g w as ret racted cau dally an d th e p arietal pleu ra w as in cised to access th e w ires of th e Tom it a saw. Th e T3-T4 disk sp ace w as th en t ran sected. Th e t u m or w as d issected from t h e su r rou n d in g st r u ct u res w it h ou t violat in g t h e t u m or cap su le, an d th e t u m or w as rem oved en bloc. Recon st r u ct ion w ith a t it an iu m m esh cage p acked w ith dem in eralized bon e m at rix from C7-T4 w as perform ed (Fig. 8.2b). Th e posterior fu sion w as com p leted u sing decor t icated lam in ar bon e from T4-T8, an d t w o ch est t u bes w ere p laced . Plast ic su rger y assisted w it h t h e closu re of t h e ch est w all an d m u scu lat u re. Th e pat ien t rem ain ed in in ten sive care for 4 days, an d w as disch arged to inpat ien t reh abilitat ion 2 w eeks after surger y. At 1 year after su rger y, th e p at ien t h ad ret u rn ed to fu ll-t im e w ork. He h ad 5/5 st rength in all ext rem it ies, w it h n or m al gait an d n e m otor skills in his hands. There w as no evidence of residual t um or. He is alive 2½ years after su rger y.

■ Diagnosis Biop sy of th e lesion sh ou ld be con ducted u n der CT guidan ce to avoid open biopsy an d con tam in at ion of t u m or m argin s. Th is con siderat ion m u st be balan ced w ith t h e p ath ologist’s n eed for t issu e. At th e t im e of biopsy, th e skin in cision sh ou ld be carefu lly m arked an d excised en bloc in subsequ en t surgical procedures to preven t t u m or seed ing along th e su rgical t ract . If a pat ien t presen ts w ith a susp ected ch ordom a, referral to a ter t iar y academ ic cen ter is recom m en ded for th e biop sy to en su re th e ap propriate p ath ological diagn osis an d t reat m en t . Ch ordom as m ay presen t in a variet y of h istological subt ypes, an d due to th e rarit y of th is

lesion , path ologists m ay h ave t rou ble dist in gu ish in g ch ord om a from ot h er t u m or t yp es. Conven t ion al ch ord om a is ch aracterized by cells w ith a lobulated arch itect ure an d abun dan t m yxoid m at rix. On h istop ath ological exam in at ion , ch ordom a t issu e is often arranged in a lobu lar arrangem en t kn ow n as physalip h orous w ith in t racytoplasm ic “bubbles.”1 Tu m ors are m ore aggressive if h istological t um or n ecrosis or Ki-67–posit ive stain ing is foun d. An aplast ic feat u res, in cluding m arked n u clear at yp ia or a sarcom atoid p at ter n an d solid arch itect u re, cor relate w it h h igh p roliferat ion , m et ast asis, an d a poor progn osis. Tu m ors w ith in creased am oun t s of ch on droid m at rix are kn ow n as ch on droid ch ordom a. Cellular based lesion s are at yp ical ch ordom as an d th ose w ith h igh grad e sp in dle cell di eren t iat ion are dedi eren t iated ch ordom a, w h ich is often h igh ly aggressive an d p resen t s w it h in creased r isk of m etastasis. With dedi erent iated chordom a, the lesion m ay transform to a high-grade lesion, t ypically brou s h ist iocytom a, brosarcom a, or, less often , osteosarcom a or rh abdom yosarcom a. Outcom e data for pat ien t s w ith dedi eren t iated ch ordom a are lim ited to case repor ts an d sm all coh or t st u dies. A case st u dy of fou r p at ien t s w it h d ed i eren t iated ch ord om a rep orted su r vival of 63 an d 76 m on th s for t w o p at ien ts follow ing su rgical resect ion , an d t w o d eath s at 7 an d 9 m on th s.12 Th e gen et ics of ch ordom a h ave been st u died exten sively in recen t years. Th e t ran scrip t ion factor T (brachyu r y) is overexpressed in ch ordom as. In an im al m odels, m em bers of ou r group gen erated a n ovel ch ordom a cell lin e, JCH7, from a 61-year-old w om an w ith sacral ch ordom a, an d silen ced th e brachyury gen e using sh RNA (sm all or sh ort h airp in RNA).13 Th is resulted in a 20 to 25% decrease in cell viabilit y an d cell size, as w ell as an arrest of cell grow th an d loss of th e ch aracterist ic physaliferous cytoplasm . In a study using w hole-exom e and Sanger sequencing of brachyur y in patien ts w ith chordom as and ancestry-m atched controls, a com m on single n u cleot ide p olym orp h ism (SNP) w as associated w ith ch ordom a risk.14 In a review of ch ordom a gen et ics an d th erapeut ic t arget s, ou r grou p id en t i ed ot h er ch rom o som al abn orm alit ies, in clu ding delet ion s in 1p

Chordoma an d 9p th at are associated w ith un con t rolled cellu lar proliferat ion .15 Th ese gen es m ay ser ve as target s for n ovel th erap ies in th e fu t u re.

■ En-Bloc Resection En -bloc resect ion refers to a su rgical tech n iqu e of rem oving a t um or in on e piece w ith ou t violat ing th e t u m or m argin s; h ow ever, it is th e n al m argin s repor ted by th e path ologist th at are of progn ost ic sign i can ce. Th e m argin s can be in t ralesion al, m argin al, or w id e. Classically, a w ide surgical m argin involves en -bloc resect ion of t h e t u m or w it h ou t d isr u pt ion of t h e t u m or cap su le an d sacr i ce of adjacen t soft t issue. Alth ough su rroun ding t issues m ay ap p ear m acroscop ically u n involved , locally in vasive disease m ay be presen t at a cellular or m icroscopic level. Th u s, w ide m argin s p rovide th e m ost th orough resect ion of t h e lesion . Ot h er con sid erat ion s, su ch as t h e p roxim it y of t h e sp in al cord an d bow el, can lim it t h e feasibilit y of w id e m argin al resect ion for sp in al ch ord om as, u n less fu n ct ion al im p air m en t is accepted .16 Margin al resect ion involves th e en -bloc rem oval of th e t u m or w ith dissect ion along th e t um or capsule, w h ich r un s th e risk of satellite t um or cells being left beh in d. In con t rast , in t ralesion al resect ion involves cu ret tage

an d piecem eal resect ion of th e t um or. Violat ion of th e t u m or capsu le in creases th e risk of local t u m or sp illage an d seed ing alon g su rgical m argin s, an d alm ost alw ays leaves residu al t u m or beh in d. En -bloc resect ion , w ith m argin al or w ide m argin s, u n equ ivocally provides bet ter longterm su r vival. Th e Spin e On cology St udy Group (SOSG) issu ed a st rong recom m en dat ion based on m oderate-qualit y eviden ce an d con sen sus exp er t op in ion th at en -bloc resect ion sh ou ld be u n der t aken for surgical t reat m en t of spin al ch ordom a.2 Th is w as th e resu lt of a system at ic review an d am bisp ect ive m u lt icen ter coh or t st u dy by t h e SOSG.2 How ever, t h e lan gu age used in th e ch ordom a literat ure is in con sisten t . Th e term s gross total, radical, an d en-bloc resect ion are often u sed in terch angeably, w ith con siderable variabilit y based on th e su rgeon’s p refer red ter m in ology. In a literat u re review of peer-review ed, PubMed-in d exed, English langu age ar t icles from 1996 to 2011 w ith fu ll text available, repor t ing dat a for en -bloc resect ion for ch ord om a, am ong st u d ies w it h 94% or m ore pat ients undergoing en -bloc resect ion , th e m ean disease-free sur vival w as 70 to 84.2 m on th s (Table 8.1). Surgical resect ion of th e spin al t um or also varies by t um or location . Speci cally, ch ordom as in the cer vical spine present a unique challenge du e to com plex an d fun ct ion ally crit ical

Table 8.1 Literature Review of Reported Mean Disease -Free Survival

Source Hsieh et al, 2011 Cloyd et al, 2009 Ahmed et al, 2009 Barrenechea et al, 2007 Boriani et al, 2006 Fuchs et al, 2005 Boriani et al, 1996

Sample Size

Male (%)

En Bloc (%)

5 2 18 7

60 50 56 57

100 100 94 0

52 52 29

71 65 45

19 40 97

Margin Type (W, M, I) a W and M M W,M, and I I W, M, I, no resection W, M, and I W, M, and I

Mean Follow -Up (Months)

Mean Disease -Free Survival (Months)

54.7 21 132 59

84.2 NA 70 NA

NA 94 30

NA 70 NA

Note: Am ong studies with 94% or m ore patients undergoing en-bloc resection, m ean disease-free survival was 70 to 84.2 months. Inclusion criteria for this review included Pubm ed-indexed articles in English from 1996 to 2011, with full text available. Articles were reviewed for data reporting outcom es following en-bloc resection of chordom a. a Margin t ype: W, wide; M, m arginal; I, intralesional.

89

90

Chapter 8 an atom y. In esop h agu s, t rach ea, carot id , an d ver tebral ar teries, C5 th rough T1 lim it th e possibilit y of t r u ly w id e su rgical m argin s. Locat ion at th e cran iocer vical ju n ct ion , or exten sion of a skull base ch ordom a in to th e cer vical spin e, is associated w ith w orse outcom e.17 W h en en bloc resect ion is n ot possible, in t ralesion al resect ion w ith su bsequ en t stereot actic rad iat ion has been reported to provide im proved survival. Th is m eth od h as sh ow n a 46% con t rol rate at 5 years for t u m ors of th e cer vical spin e.18–20 In recen t years, advan ces in surgical tech n iqu es h ave en abled m ore aggressive resect ion in pat ien t s w it h cer vical sp in e ch ord om a, often th rough m u lt ist age operat ion s.17,21 Th is h as led to m ean a disease-free su r vival of 84.2 m on th s for t u m ors of th e cer vical spin e follow ing en bloc resect ion in on e coh or t st u dy.10 Prop er su rgical p lan n ing th rough im aging h as en abled th e m ajorit y of ch ordom as of th e th oracolu m bar sp in e to be rem oved en bloc w ith disease-free m argin s. Using m u lt iple su rgical approach es is com m on , w ith a posterior approach being u sed in n early ever y th oracolu m bar ch ordom a su rger y. Th e addit ion of an an terior approach m ay be perform ed as dictated by a t um or grow ing ven t rally beyon d th e vertebra. Th is h as been sh ow n to im p rove disease-free sur vival.8

disk sp ace. Tw o days later, a vascu lar su rgeon assisted in an an terior t ran sabdom in al an d retroperiton eal exposu re of th e L3-L4, L4-L5, an d L5-S1 in tersp aces, w ith m obilizat ion of th e ab dom in al aor t a. Th e Tom it a saw s w ere advan ced an d th e t u m or w as rem oved en bloc from th e eld; h ow ever, p ath ological review of th e specim en dem on st rated t u m or w ith in 1 m m of th e an terior m argin s of th e specim en . An terior reconstruction from L2-S1 w as perform ed using a dist ractable t itan iu m case an d a t ran sver tebral screw in th e L2 vertebral body, w ith t itan ium cablin g to t h e p oster ior rod /screw con st r u ct (Fig 8.3c). Postop erat ively, t h e p at ien t exp er ien ced transfusion-related acute lung injury (TRALI) an d rem ain ed in t ubated follow ing th e secon d posterior op erat ion . Sh e w as disch arged to in pat ien t reh abilit at ion w ith 5/5 m otor st rength in all ext rem it ies. Sh e received p ostoperat ive proton beam radiat ion. Tw o years after surger y, loosen ing of th e iliac screw s w as n oted an d sh e u n der w en t revision for replacem en t of bilateral iliac screw s an d ad dit ion of allograft bular st ru ts for posterior ar th rod esis. Sh e h ad n o eviden ce of recurren ce of disease at 2 years of clin ical follow -u p, an d is alive 2½ years after surger y.

Case Illustration

■ Sacral Chordoma

A 45-year-old w om an presen ted w ith back pain after a fall. Im aging revealed a lesion at L4 (Fig. 8.3a,b), an d a percu t an eou s biop sy w as con sisten t w ith ch ordom a. Sh e w as o ered a st aged procedu re for en -bloc resect ion of th e t u m or. St age on e involved a posterior ap proach w ith L2, L3, an d L5 lam in ectom ies an d en -bloc rem oval of t h e p oster ior elem en t s of L4 u sin g a Tom it a saw . Poster ior in st r u m en t at ion w as placed from T11 to th e pelvis, w ith p edicle screw s at T10-L1 an d S1, an d bilateral iliac screw s. A Silast ic sh eet w as p laced bet w een th e t h ecal sac an d in st r u m en t at ion , an d t h e w ou n d w as closed in a m ult ilayered fash ion . Th e n ext day, th e iliopsoas an d lu m bar p lexu s w as m obilized an d dissected, from th e posterior ap proach , an d Tom it a saw s w ere placed posterior to th e L2-L3 disk sp ace an d th e L5-S1

Sacral ch ordom as pose u n iqu e ch allenges for su rgeon s, as th ey can grow to m assive sizes. Poten t ial problem s in clu de m assive blood loss, soft t issu e d efect s, an d a h igh r isk of w ou n d in fect ion . Procedures are posterior or staged anterior and posterior. Lum bopelvic reconstruct ion is rarely requ ired after t h e p roced u re u n less a tot al or h igh sacrectom y is requ ired ; recon st ru ct ion adds con siderable com plexit y to th e case. Th e best available eviden ce suggests th at w ide m argin s sh ould be ach ieved du ring th e resect ion .3,8,9,22 Sacrectom y level is d ict ated by th e n er ve roots involved w ith th e t um or. Fourney an d colleagues,9 correlating the level of resect ion w ith th e postoperat ive m otor, bow el, an d blad der d e cit s, p roposed a classi cat ion sch em e. Neurologic de cits can lead to

Chordoma

91

a

b

c

Fig. 8.3a–c Sagit tal T2-weighted fat-suppressed (a) and T1-weighted (b) magnetic resonance imaging (MRI) demonstrating a lumbar chordoma at L4 with ventral extension from L3-L5. There is a loss of vertebral body height at L4. (c) Postoperative sagit tal CT demonstrates anterior/posterior instrumentation.

92

Chapter 8 profoun d m orbidit y, an d, w h en possible, even u n ilateral preser vat ion can perm it sat isfactor y fu n ct ion . Alth ough p reser vat ion of bilateral L5 n er ve root s is n ecessar y for sat isfactor y am bu lat ion , u n ilateral p reser vat ion of S2-S5 p erm its sat isfactor y bow el con t rol an d possible m ain ten an ce of bladder an d sexual fun ct ion . Bilateral preser vation of S2 w ith unilateral S3 results in alm ost n or m al bow el/blad d er an d sexu al fu n ct ion . For p at ien t s w it h com p lete d e cit s prior to surger y, it is un likely th at su rger y w ill restore th e de cit . Surgical resect ion of sacral ch ordom a classically involves a st aged an terior an d p osterior resect ion , alth ough m ore aggressive p osterioron ly p rocedu res are n ow being d on e. In a large case series described at ou r in st it u t ion , a posterior-only approach w as safely used for en -bloc sacrectom y to lim it th e m orbidit y associated w ith anterior approaches.23 The trade o , h ow ever, is large soft t issu e d efect s. W h en an an ter ior ap p roach is d on e rst , a p ed icled ap reconstruction w ith a vertical rectus abdom inis m u scu locu t an eou s (VRAM) ap is sim ple an d readily available on ce th e secon d-stage p osterior app roach is com p leted. With lon e posterior ap p roach es on ly, local glu teal m u scle aps along w ith an acellu lar derm al m at rix to preven t bow l h ern iat ion are available. Som et im es th e local glu teal m u scle is n ot en ough an d free aps m ust be used. Th ese w ou n ds com m on ly becom e infected, and rates up to 40%have been rep or ted.22 Pat ien t s sh ou ld be in form ed of th is likely com plicat ion . Exten sive lum bopelvic recon st r uct ion is required w h en m ore th an h alf of th e sacroiliac join t is resected. A fem oral st ru t or t ran siliac rod, lum bar posterior instrum entation, and iliac or S2 screw s w ith cross-con n ectors sh ou ld be u sed to e ect ively rejoin th e lum bar spin e an d p elvis follow ing total sacrectom y. Resect ion of th e cau dal sacru m up to th e low er h alf of th e sacroiliac joint m ay be perform ed w ithout com p rom ising stabilit y.

Case Illustration A 66-year-old m an presen ted w ith sacral pain , loss of bow el an d bladder con t rol, righ t-sided foot drop, an d a palpable m ass in th e bu t tocks.

Im aging revealed a m assive t u m or, an d biopsy w as con sisten t w it h ch ord om a (Fig. 8.4a,b). On qu est ion ing, th e p at ien t w as aw are of th e t u m or for several years. A m u lt id iscip lin ar y, st aged su rgical ap p roach w as o ered to t h e p at ien t , con sist ing of, on th e rst day of th e op erat ion , m obilizat ion of t h e iliac vessels from t h e lu m bosacral ju n ct ion an d ligat ion of th e in ter n al iliac vessels by t h e vascu lar team , a colostom y by th e gast roin test in al su rger y service, an d a p ar t ial L5 corpectom y. On th e secon d day, h ar vest ing of th e leftsided rect u s abdom in is p ed icle ap (fed by th e in ferior ep igast ric vessels) w as com pleted by th e p last ic su rger y team , an d a p osterior ap proach w as initiated by the neurosurgical team . A T-sh aped in cision w as m ade, exten ding bet w een th e greater t roch an ters an d vert ically to exp ose th e lu m bar sp in e to L3. Th e sacral n er ve roots were ligated and posterior lam inectom ies w ere com p leted at L4 an d L5, as an exten sive dissect ion w as m ade to rem ove all soft t issu e from th e t u m or. Th e L5 sp on dylectom y w as com pleted w ith sparing of th e L5 n er ve root , an d th e t um or w as rem oved en bloc w ith th e en t ire sacr u m . Pedicle screw in st ru m en t at ion follow ed in L2-L5, follow ed by iliac screw p lacem en t , an d rod p lacem en t bet w een th e ilia (Fig. 8.4c– e). A con tou red rod w as placed ver t ically an d con n ected to th e iliac rod w ith J-sh ap ed con n ectors. Th en a fem u r sh aft w as cut in h alves an d w as t ted in to th e defect exten ding from on e ilium to th e oth er side. Th is w as w ired in p lace u sing a t itan iu m w ire. Th en , th e dorsal asp ect of th e spin al colu m n w as decort icated, exten ding from L2 all th e w ay dow n to th e pelvis bilaterally. An allograft bon e graft w as m ixed w it h ost iu m an d d em in eralized bon e m at rix, an d th e bon e graft w as packed t igh t ly, exten d in g L2 all th e w ay d ow n th e p elvis bilaterally an d d ow n to th e fem u r sh aft (Fig. 8.4c– e). Th e pat ien t experien ced a drop in oxygen sat urat ion an d blood pressure in th e operat ing room ; h ow ever, h e w as successfully resu scitated an d th e plast ic surger y team com pleted th e com p lex w ou n d closure th e follow ing day. He con t in u ed to im p rove an d w as disch arged to inpat ien t reh abilitat ion 1 m on th after surger y. He is alive 2 years after surger y.

Chordoma

93

a

b

c

d

e

Fig. 8.4a–e Axial T1-weighted (a) and T2-weighted fat-suppressed (b) MRI of a massive sacral chordoma. The patient presented with sacral pain from a known lesion of several years. (c,d) Lumbopelvic reconstruction on postoperative CT scan. (c) Fem oral strut with iliac screws and rods. (d) En-bloc sacrectomy is evident, seen as a complete absence of the bony sacrum. (e) CT scout image illustrating the full extent of lumbopelvic reconstruction, including bilateral iliac screws.

94

Chapter 8

■ Adjuvant Therapy Radiation Therapy Classically, ch ordom a h as been con sidered a radioresistant lesion, requiring either high-dose fract ion ated ph oton or proton th erapy; h ow ever, n o class I eviden ce exist s to guide th e adm in ist rat ion of radioth erapy. Th e SOSG issued a w eak recom m en dat ion based on low -qualit y eviden ce th at radiat ion th erapy of 60- to 65- Gy equ ivalen t s or h igh er sh ou ld be adm in istered follow ing in com plete or in t ralesion al resect ion of ch ordom a.2 Th is w as th e result of a system at ic review an d am bispect ive m u lt icen ter coh or t st u dy by th e SOSG.2 In th e literat ure, h igh -dose con form al con ven t ion al rad ioth erapy h as been u sed in case repor t s an d coh or t st u dies to deliver directed th erapy for ch ordom a follow ing in com p lete su rgical resect ion or p rior to su rgical resect ion to reduce th e violat ion of con t am in ated t um or m argin s 11,19 ; h ow ever, in ciden t al radiat ion to th e spin al cord an d paraspin al st r uct u res is a serious con cern . Stereot act ic radiosurger y h as been p roposed in som e sm all case series for p at ien t s w it h sp in al ch ord om a an d in larger coh or t s w ith sku ll base ch ord om a, an d h as d em on st rated safet y an d e cacy w ith a 5 year OS of 74.3%.24 Altern at ively, proton th erapy h as em erged as yet an oth er th erapeut ic opt ion for ch ordom a.11,25 In con t rast to ph oton th erapy, p roton s h ave lim ited scat ter an d can t arget t u m ors located in deep t issu es w ith ou t deposit ing radiat ion to st ru ct u res beh in d or deep to th e t u m or. Th is is par t icu larly advan t ageou s for ch ordom as arising w ith in th e skull base or ver tebral colu m n th at abu t th e delicate st r u ct u res of th e n er vou s system . Radiat ion as an adjuvan t th erapy follow ing su rgical resection for ch ordom a h as h istorically dem on st rated im p roved p rogression -free sur vival (PFS) an d OS.19,26 Radiotherapy as a salvage th erapy follow ing m u lt ip le resect ion s is d em on st rably less e ective. In a ret rospect ive an alysis of 27 p at ien t s w ith sacral ch ordom a w h o received su rger y an d p roton ± p h oton radiat ion , pat ien ts w ith prim ar y t um ors h ad a 92.9% 5-year overall su r vival, w h ereas pat ien t s w ith recu rren t t u m ors h ad a 66.7%5-year OS.11

Not ably, radiat ion doses w ere h igh in th is coh or t , ≥ 53 Gy in 24-42 fract ion s. In th e literat u re, m any st u dies of h igh -dose p roton /ph oton th erapy as a d e n it ive single th erapy are lim ited to skull base or cran iocervical jun ct ion ch ordom as, as th ese lesion s are less am en able to en -bloc resect ion . In con t rast , few st u dies h ave been p u blish ed on t h e ou tcom es for proton /ph oton th erapy alon e for th e t reat m en t of spin al ch ordom a. In a ret rospect ive review of 24 p at ien ts w ith p rim ar y u n resect able ch ordom a of th e spin e w h o did n ot u n d ergo su rgical resect ion , local PFS an d OS w ere 79.8%an d 78.1%at 5 years, respect ively.25 In t h is st u dy, t h e m ean dose of p roton /p h oton th erapy w as 77.4 Gy RBE (Gy RBE is de n ed as th e p roton d ose u n it , equ al to Gy × t h e relat ive biological e ect iven ess).25 In a sm all coh or t of six p at ien t s w h o received h igh -d ose p roton or ph oton radiat ion on ly for prim ar y or recu rren t ch ord om a, fou r p at ien t s received h igh -d ose radiat ion at ≥73 Gy equ ivalen t s, an d th ree of th ese pat ien t s h ad local con t rol at 2.9, 4.9, an d 7.6 years.11

Chemotherapy In a review of th e curren t th erapeut ic opt ion s for ch ordom a, m em bers of our grou p su m m arized th e u se of ch em oth erapy as an adjuvan t for ch ordom a.15 No class I eviden ce curren tly exists to drive the treatm ent of chordom a; how ever, regim en s repor ted in sm all case-coh or t studies include anthracycline, cisplatin, alkylatin g agen t s, an d cam ptot h ecin . In a p h ase II p rospect ive t rial of im at in ib, a t yrosin e kin ase in h ibitor used for ch ron ic m yelogen ou s leukem ia, 56 p at ien t s w ith p latelet-derived grow th factor (PDGF)-posit ive ch ordom a w ere given 800 m g of im atin ib per day un t il t um or progression .27 Overall t um or respon se rate, as den ed by th e Response Evaluat ion Criteria In Solid Tum or (RECIST), w as recorded. A 64%clin ical ben e t rate w as fou n d, in clu ding p at ien t s w ith com plete respon se, p ar t ial resp on se, or st able disease for ≥ 6 m on th s.27 Recen t t rials of th e epiderm al grow th factor receptor (EGFR)in h ibitor erlot in ib h ave sh ow n a good respon se w ith a 12-m onth uninterrupted use of the drug in a single pat ien t ,28 an d oth er EGFR in h ibitors

Chordoma h ave been repor ted in a lim ited n um ber of case rep or ts. Sim ilarly, on e grou p rep or ted resp on se in recu r ren t d isease w ith p u lm on ar y m et ast ases w ith th e com bin ed th erapy cet u xim ab/ ge t in ib over a 9 m on th follow u p period.29

■ Chapter Summary Pat ien t s w ith sp in al ch ordom a h ave th e poten tial for excellent long-term survival w hen given th e p rop er care. Treat m en t st ar t s at t h e rst clin ical en cou n ter, w it h refer ral to a ter t iar y care cen ter w it h h igh -volu m e exp er ien ce in ch ordom a m an agem en t . Op en biop sy is n ot recom m en d ed, an d CT gu ided or n e n eedle aspirat ion is appropriate w ith m arking of th e biopsy t ract for later in clusion in th e en -bloc resect ion . Path ological con rm at ion of p hysaliferous cells, as w ell as preoperat ive im aging hyperin ten sit y on T2 an d fat-suppressed im ages, in a t u m or crossing t h e d isk sp ace or invad ing in to t h e p arasp in al m u scu lat u re st rongly suggest s a diagn osis of ch ordom a. Th e SOSG issued a st rong recom m en dat ion for en bloc resect ion an d a w eak recom m en dat ion for h igh -dose radioth erapy follow ing in com plete or in t ralesion al resect ion . A recen t ph ase II t rial of im at in ib dem on st rated m odest clin ical ben e t based on th e RECIST. Su rgical resect ion and postoperative adjuvant therapy are dictated by th e level of th e t u m or, an d cer vical an d sacral t u m ors pose un ique ch allenges for enbloc resection and postoperative reconstruction. Gen et ic st udies involving an im al an d h um an

dat a h ave isolated th e brachyury gen e as a possible fut u re target for th erapy. In th e fu t u re, ran dom ized con t rolled prospect ive clin ical t rials w ill be h elpfu l to determ in e th e ideal t reatm en t s for p at ien t s w ith ch ord om a w ith in th e spin al colu m n .

Pearls The Spine Oncology Study Group (SOSG) issued a strong recomm endation on moderate-qualit y evidence that en-bloc resection should be attempted in spinal chordoma. The SOSG o ered a weak recom mendation based on low-qualit y evidence that radiation therapy greater than 60 to 65 Gy equivalents is indicated following incomplete resection. Chemotherapy recom mendations are based on mostly class II and IV evidence. One prospective, phase II trial reported a moderate clinical benef t based on the RECIST using im atinib in a subset of chordom a. For patients with sacral chordom a, en-bloc resection will result in a large soft tissue defect, and coordination of care with a plastic surgeon is recomm ended. Pitfalls Open biopsy seeds the biopsy site with tum or and precludes later en-bloc resection. Surgical planning involves excision of the biopsy tract. En-bloc total or high sacrectomy can result in massive blood loss. Removal of greater than 50% of the sacroiliac joint requires lum bopelvic reconstruction. Wound infection rates are high following sacrectomy for sacral chordoma, and patients should be warned of this likely complication.

Refere nces Five Must-Read Refe rences 1. Bjorn sson J, Wold LE, Ebersold MJ, Law s ER. Ch ordom a of th e m obile spin e. A clin icopath ologic an alysis of 40 pat ien t s. Can cer 1993;71:735–740 Pu bMed 2. Borian i S, Saravanja D, Yam ada Y, Varga PP, Biagin i R, Fisher CG. Ch allenges of local recurren ce and cure in low grade m align an t t um ors of th e spin e. Spin e 2009;34(22, Su pp l):S48–S57 Pu bMed 3. Sciubba DM, Ch eng JJ, Pet teys RJ, Weber KL, Frassica DA, Gokaslan ZL. Ch ordom a of th e sacrum an d verte-

bral bodies. J Am Acad Or th op Su rg 2009;17:708– 717 PubMed 4. Walcot t BP, Nah ed BV, Mohyeldin A, Coum an s JV, Kah le KT, Ferreira MJ. Ch ordom a: current con cept s, m an agem en t , an d fut ure direct ion s. Lan cet On col 2012;13:e69– e76 PubMed 5. Mukh erjee D, Ch aich an a KL, Gokaslan ZL, Aaron son O, Ch eng JS, McGir t MJ. Sur vival of pat ien t s w ith m alignan t prim ar y osseous spinal n eoplasm s: resu lt s

95

96

Chapter 8 from th e Su r veillan ce, Epidem iology, an d End Resu lt s (SEER) dat abase from 1973 to 2003. J Neu rosurg Spin e 2011;14:143–150 Pu bMed 6. McMaster ML, Goldstein AM, Brom ley CM, Ish ibe N, Parr y DM. Ch ord om a: in ciden ce an d su r vival p at tern s in th e Un ited St ates, 1973-1995. Can cer Cau ses Con t rol 2001;12:1–11 Pu bMed 7. Mu kh erjee D, Ch aich an a KL, Ad ogw a O, et al. Associat ion of exten t of local t u m or invasion an d su r vival in pat ient s w ith m align an t prim ar y osseou s spin al neoplasm s from th e sur veillan ce, epidem iology, an d en d result s (SEER) dat abase. World Neurosurg 2011; 76:580–585 PubMed 8. Borian i S, Ban diera S, Biagin i R, et al. Chordom a of th e m obile spin e: ft y years of experien ce. Spin e 2006; 31:493–503 PubMed 9. Fourn ey DR, Rh in es LD, Hen t sch el SJ, et al. En bloc resect ion of prim ar y sacral t um ors: classi cat ion of su rgical approach es an d outcom e. J Neu rosurg Spine 2005;3:111–122 PubMed 10. Hsieh PC, Gallia GL, Sciubba DM, et al. En bloc excisions of ch ordom as in th e cer vical spin e: review of ve con secut ive cases w ith m ore th an 4-year follow -up. Spin e 2011;36:E1581–E1587 PubMed 11. Park L, Delaney TF, Liebsch NJ, et al. Sacral ch ordom as: Im pact of h igh -dose proton /ph oton -beam radiat ion th erapy com bin ed w ith or w ith ou t su rger y for prim ar y versu s recu rren t t u m or. In t J Radiat On col Biol Phys 2006;65:1514–1521 Pu bMed 12. Han na SA, Tirabosco R, Am in A, et al. Dedi eren t iated ch ordom a: a rep or t of fou r cases arising “d e novo.” J Bon e Join t Surg Br 2008;90:652–656 PubMed 13. Hsu W, Mohyeldin A, Sh ah SR, et al. Gen erat ion of ch ordom a cell lin e JHC7 an d th e iden t i cat ion of Brachyur y as a n ovel m olecular t arget . J Neurosurg 2011;115:760–769 PubMed 14. Pillay N, Plagn ol V, Tarp ey PS, et al. A com m on singlenu cleot ide varian t in T is st rongly associated w ith ch ordom a. Nat Gen et 2012;44:1185–1187 Pu bMed 15. Bydon M, Papadim it riou K, With am T, et al. Novel th erap eu t ic t arget s in ch ordom a. Exp er t Op in Th er Target s 2012;16:1139–1143 Pu bMed 16. Borian i S, Wein stein JN, Biagin i R. Prim ar y bon e t um ors of th e spin e. Term in ology an d surgical st aging. Spin e 1997;22:1036–1044 Pu bMed 17. Yasuda M, Bresson D, Ch ibbaro S, et al. Ch ordom as of th e skull base an d cer vical spin e: clin ical outcom es associated w ith a m ultim odal surgical resection com bin ed w it h p roton -beam rad iat ion in 40 p at ien t s. Neu rosu rg Rev 2012;35:171–182, d iscu ssion 182– 183 Pu bMed

18. Barrenech ea IJ, Perin NI, Trian a A, Lesser J, Cost an t in o P, Sen C. Su rgical m an agem en t of ch ord om as of th e cer vical spin e. J Neu rosu rg Spin e 2007;6:398– 406 PubMed 19. Boriani S, Chevalley F, Weinstein JN, et al. Ch ordom a of th e spin e above the sacr um . Treat m en t an d outcom e in 21 cases. Spin e 1996;21:1569–1577 Pu bMed 20. Noël G, Habran d JL, Jau ret E, et al. Radiat ion th erapy for ch ord om a an d ch on d rosarcom a of th e sku ll base an d th e cer vical spin e. Progn ost ic factors an d pattern s of failu re. St rah len ther On kol 2003;179:241– 248 PubMed 21. Cloyd JM, Ch ou D, Deviren V, Am es CP. En bloc resect ion of prim ar y t um ors of the cer vical spin e: repor t of t w o cases an d system at ic review of th e literat ure. Spin e J 2009;9:928–935 Pu bMed 22. Ruggieri P, Angelin i A, Ussia G, Mon t alt i M, Mercuri M. Su rgical m argin s an d local con t rol in resect ion of sacral ch ordom as. Clin Or th op Relat Res 2010;468: 2939–2947 Pu bMed 23. Clarke MJ, Dasen brock H, Bydon A, et al. Posterioronly approach for en bloc sacrectom y: clinical out com es in 36 con secu t ive pat ient s. Neurosu rger y 2012;71:357–364, d iscu ssion 364 Pu bMed 24. Hen derson FC, McCool K, Seigle J, Jean W, Har ter W, Gagn on GJ. Treat m en t of ch ordom as w ith CyberKn ife: Georgetow n Un iversit y experien ce an d t reat m en t recom m en dat ion s. Neurosurger y 2009;64(2, Suppl):A44–A53 Pu bMed 25. Ch en YL, Liebsch N, Kobayash i W, et al. De n it ive high -dose photon /proton radiotherapy for unresected m obile spin e an d sacral ch ordom as. Spin e 2013;38: E930–E936 Pu bMed 26. Th ieblem on t C, Biron P, Roch er F, et al. Progn ost ic factors in ch ordom a: role of p ostop erat ive radioth erapy. Eur J Can cer 1995;31A:2255–2259 Pu bMed 27. St acch iot t i S, Casali PG, Lo Vullo S, et al. Chordom a of th e m obile sp in e an d sacr u m : a ret rosp ect ive an alysis of a series of pat ien t s surgically t reated at t w o referral cen ters. An n Surg On col 2010;17:211–219 PubMed 28. Laun ay SG, Ch et aille B, Medin a F, et al. E cacy of epiderm al grow th factor receptor t arget ing in advan ced ch ordom a: case repor t an d literat ure review. BMC Can cer 2011;11:423–2407 Pu bMed 29. Hof H, Welzel T, Debus J. E ect iven ess of cet u xim ab/ ge t in ib in th e th erapy of a sacral ch ordom a. On kologie 2006;29:572–574 PubMed

9 Chondrosarcoma Stefano Boriani, Stefano Bandiera, Riccardo Ghermandi, Simone Colangeli, Luca Amendola, and Alessandro Gasbarrini

■ Introduction Ch on drosarcom a (CHS) is th e th ird m ost com m on p rim ar y m align an t bon e t u m or after osteosarcom a an d Ew ing’s sarcom a. Accord in g to th e World Health Organ izat ion’s (W HO) accepted de n it ion ,1 th e term chondrosarcom a refers to “a grou p of locally aggressive or m alignan t t um ors produ cing cartilagin ous m atrix, w ith di eren t m orph ological feat u res an d clinical beh avior.” Th is fam ily of t u m ors in clu des several su bt yp es: cen t ral, p er ip h eral, m esen chym al, clear-cell, an d dedi eren t iated . Central CHS arises centrally in bone and then expan ds, in ltrat ing th e t rabeculae an d being p ar t ially cou n tered by som e th in react ive bon e (Fig. 9.1). It th en erodes th e cor tex an d invades th e epidu ral space (Fig. 9.2) an d th e su rrou n ding soft t issu es. Perip h eral CHS t akes th e form of car t ilagin ou s an d m yxoid m asses th at slow ly grow an d invade the surroun ding soft t issues (Fig. 9.3). Mesen chym al CHS is rare, com prising 0.7% of m align an t t u m ors obser ved at t h e Mayo Clin ic.2 Usually, bu t n ot alw ays, low -grade CHS in clu des car t ilage an d u n di eren t iated sm all roun d cells, som et im es bearing st riking sim ilarit y to th ose in Ew ing’s sarcom a. For t u n ately, th ey can be d i eren t iated by areas of ch on droid m at r ix, w h ich d o n ot occu r in Ew ing’s sarcom a. Mesen chym al CHS in th e spin e is ext rem ely rare.3

Clear-cell CHS is rarely observed in the spine. It is t ypically arranged in pat tern s of car t ilage lobu les. Ben ign gian t cells are u su ally fou n d th rough ou t t h e t u m or, eith er in sm all clu sters or in dividu ally. Th is t um or is described as expan sile an d dest r u ct ive, often w ith soft t issu e exten sion an d lack of m in eralizat ion . It s cou rse is m ore aggressive an d it is associated w ith a w orse p rogn osis if it is u n d erest im ated an d su bm it ted for in t ralesion al excision , w h ereas en -bloc resect ion is m ore su ccessfu l.4 Clearcell CHS, h ow ever, can recur as long as 10 to 15 years after resect ion , an d it can m etast asize. Dedi eren t iated CHS is a h igh ly vascu larized, h igh -grade sarcom a th at is ext rem ely rare in the spine.5 Its cartilage com ponent has a low grade ap pearan ce in con t rast to th e h igh -grade car t ilagin ou s p or t ion s t h at are often fou n d in osteosarcom as. Th is CHS var ian t requ ires aggressive surger y (en -bloc resect ion ) com bin ed w ith ch em oth erapy. Histological grading of all CHSs range from 1 to 3; a grade 4 CHS corresp on ds to dedi eren t iated CHS. Th e m axim u m in ciden ce of CHS is fou n d in pat ien ts 30 to 70 years of age. Rarely does CHS occur before th e age of 20 years, an d on ly except ion ally is it fou n d before p u ber t y. It occu rs ap p roxim ately t w ice as often in m en as in w om en . Sp in e CHSs rep resen t on ly 10% of all CHS. Most cases of CHS observed in the spine are cent ral an d p eriph eral, both p ossibly arising from

98

Chapter 9

a

c

b

Fig. 9.1a–c A central chondrosarcoma (CHS) in a 54-year-old man. (a) A computed tomography (CT) scan demonstrates a lytic area arising in the vertebral body of T10, fully encased in the vertebral body, with well-de ned margins, osteosclerotic reaction, and scat tered rounded calci cations inside the tum or m ass. (b) T1-weighted magnetic resonance im aging (MRI) demonstrates that the tum or has low intense signal. The tumor border is well de ned. (c) T2-weighted MRI dem onstrates that the tumor has a high-intensit y signal and is surrounded by high-intensit y cancellous bone signal related to edematous reaction.

Chondrosarcoma

a

99

b

d

c

Fig. 9.2a–h A central CHS invading the canal, earlier than might be expected, in a 42-year-old wom an. (a) CT scan dem onstrates that a thin osteosclerotic border is limiting the tumor anterolaterally. There is a well-de ned limit to the cancellous bone. Scattered calci cations are seen within the tumor mass. (b,c) MRI dem onstrates bet ter de nition of the epidural extension (d) CT-guided trocar biopsy. (continued on next page)

100

Chapter 9

e

f

g

Fig. 9.2a–h (continued) (e) On the resected specimen the epidural extension is fully contained by the longitudinal ligament. (f) Reconstruction after the en-bloc resection. A carbon- ber cage is lled with chips of autogenous graft. (g,h) At 5-year follow-up, there is no local recurrence. Bone ingrowth of the graft is seen inside the carbon- ber cage, with fusion to the contiguous end plates.

h

Chondrosarcoma

Fig. 9.3 Peripheral CHS arising from the posterior elements of C4 without any evidence of preexisting osteochondrom a (exostosis) in a 28-year-old m an. A huge radiolucent soft tissue mass is seen, with scat tered round calci cations.

a precursor lesion such as an en ch on drom a, osteoch on drom a, or exostosis. Th ese cases are referred to as “secon dar y” CHS (Fig. 9.4). Th e CHSs in th e spin e are t ypically slow grow ing t u m ors, an d diagn osis is u su ally m ade later in th eir clin ical cou rse. Pain is th e m ost frequen t com plain t of pat ien t s w ith cen t ral CHS in th e m obile spin e, as it provokes a dull, aching pain unrelated to activit y and frequently n oct urn al. Sacral CHSs are m ost com m on ly in ciden tal t um ors, ow ing to th e large area w h ere th e t u m or can develop before cau sing sym p tom s of com pression (Fig. 9.5). W h en th e posterior vertebral arches are involved, a xed, hard, an d n on ten der m ass can be foun d (Figs. 9.3 an d 9.6). Ver y rarely does a spin al CHS in it ially m an ifest w ith sp in al cord or n er ve root com pression (Fig. 9.6). Radiograph ically, cen t ral CHS of th e sp in e is prim arily a lyt ic lesion th at arises cen t rally in th e ver tebral body (w ith or w ith ou t a p erip h eral p er iosteal react ion ), an d som et im es sclerot ic m argin s can also be visu alized . In ot h er cases, h eavily calci ed lobu lated m asses arise from th e cor t ical m argin s. Later on , lyt ic bon e d est r u ct ion w ith ou t m ot t led calci cat ion can be seen, either as a soft tissue m ass invading the ep idu ral space (Fig. 9.2) or as a soft t issu e m ass w ith occulat ing calci cat ion s. As th e t um or

Fig. 9.4 Peripheral CHS secondary to exostosis of the spinous process of L5 in a 23-year-old woman. The image demonstrates the implant of the exostosis, t ypically characterized by continuit y of the cortical bone and ingrowth of cancellous bone from the host bone. The cartilaginous cap is larger than 2 cm , thus con rm ing the diagnosis of CHS.

grow s in size, th e am oun t of react ion of su rrounding bone evolves into m assive perm eative dest ruct ion of th e un derlying bon e. Com p u t e d t om ograp hy (CT) scan w it h m agnetic resonance im aging (MRI) best dem on st rates th e extent of the lesion an d the relat ion sh ip w ith th e su rrou n ding an atom ic st ru ct u res, an d it evalu ates spin al cord com pression (Fig. 9.2b,c).

Fig. 9.5 Peripheral CHS secondary to exostosis of L5, growing in the abdomen without provoking pain, in a 36-year-old man. A painless mass was noted on physical examination.

101

102

Chapter 9

a

b

Fig. 9.6a,b A T4 CHS in a 59-year-old man. The patient gradually developed a gait impairment due to pyramidal symptom s. He experienced no pain; thus, the symptoms were underestimated until

imaging was performed. (a,b) A huge chondroid low-vascularized m ass is seen, arranged in lobules arising from and replacing the posterior elem ents, with severe cord compression.

Th e m ost di cu lt di eren t ial diagn osis is bet w een low -grad e CHS an d ben ign , act ive, car t ilagin ou s lesion s. Classically, t h e im agin g of p er ip h eral CHS can be d ist in gu ish ed from an osteoch on drom a (exostosis) by th e th ickn ess of th e car t ilagin ous cap. Th e size lim it bet ween osteochondrom a (exostosis) and CHS has been establish ed at 2 cm ; a cap th at is th icker th an 2 cm is con sidered p at h ogn om on ic of CHS (Fig. 9.4). Prim ary peripheral CHSs can be additionally dist ingu ish ed from osteoch on drom a (exostosis) becau se CHSs con sist of calci ed car t ilage, w h ereas m ost of th e m ass of an osteoch on drom a (exostosis) con sist s of p u re bon e. Fu rth erm ore, CHS can be dist ingu ish ed from oth er m align an t bon e t u m ors of th e spin e becau se it frequen tly arises from th e posterior elem en ts, in con t rast to osteosarcom a, Ew ing’s sarcom a, m u lt ip le m yelom a, an d th e m ajorit y of bon e m et ast ases. Cen t ral CHS is radiograph ically less dist in gu ish able. It is ch aracter ized by rad iolu cen t p er m eat ive dest r u ct ion of t h e ver tebral body (sim ilar to chordom a) m ixed w ith calci ed areas.

Th e calci cat ions in the low -grade t um ors have th e ch aracterist ic ringed p at tern . In h igh -grade lesion s th e calci cat ion s are n ot th e dom in an t radiograp h ic p at tern , an d th ere are few radiological clu es rem in iscen t of th e car t ilagin ou s h istogen esis of th e t um or: poorly de n ed m argin s, perm eat ive dest ruct ion , an d in con sisten t bone reaction. Furtherm ore, high-grade lesions th at h ave lit tle or n o calci cat ion s m ay sh ow u n expectedly as “cold” lesion s on isotope scan , m u ch like m yelom a. Angiograp hy is even less valuable th an isotope scan because of th e lit tle vascularit y of CHSs. These negative ndings are helpful in determ ining the diagn osis w hen they are associated w ith aggressive im ages on the MRI. Th e m an agem en t of CHSs of t h e lim bs is clearly est ablish ed an d com m on ly accepted . Th ere is exten sive eviden ce, dat ing back m any years, th at th e progn osis is dep en d en t on th e h istological grade of m align an cy an d on th e su rgical t reat m en t .6–9 CHSs h istorically h ave been con sidered resistan t to m ost protocols of radiat ion th erapy an d ch em oth erapy. According to Cam panacci et al,9 intralesional excision can be con sidered a valid opt ion in

Chondrosarcoma grade 1 CHS of th e ap pen dicular skeleton , resu lt ing in a low rate of resect able recu rren ce versu s m ore e ect ive bu t sign i can tly m ore aggressive en -bloc resect ion . How ever, in th e spin e local adjuvan t s su ch as ph en ol an d cr yosurger y are less com m on ly u sed. Recu rren t CHS is frequ en tly associated w ith m alignant progression, worsening the progn osis. Local recu r ren ces in th e sp in e are debilit at ing an d aggressive, som et im es p rovoking ser iou s n eu rologic d e cit s an d lead ing to d ifcu lt p alliat ion an d even t u ally to d eat h after revision su rger y, ow ing to its h igh m orbidit y. Th u s, it w ou ld be feasible to say th at w ith CHSs of th e spin e, th ere is a direct correlat ion bet w een th e su ccess of th e rst su rger y an d th e u lt im ate p rogn osis. Th e role of en -bloc resect ion in th e t reatm en t of CHS h as been accepted sin ce th e pion eering w ork by Sten er,10 w h o also described h ow to plan th ese surgical procedures based on th e t u m or exten sion an d th e local an atom y. Becau se of th e di cult ies associated w ith en bloc surger y in th e spin e, an d becau se spin e su rgeon s u se variou s on cological criteria (Fig. 9.7), lesion s of th e ver tebral colu m n h ave h ad a poor p rogn osis in dep en den t of th e h istological grade.11–14 An exh aust ive literat u re review based on th e best available eviden ce an d clin ical exp ert ise con clu ded th at th e opt im al su rgical m an agem en t for CHS of th e sp in e is en -bloc su rgical resect ion w ith w ide or m argin al m argin s,11 as eviden ced by a d ecrease in local recu rren ce, longer disease-free su r vival, an d, u lt im ately, a redu ct ion of death related to disease (st rong recom m en dat ion , m oderate-qu alit y eviden ce). Fu r th erm ore, radiat ion is n ot in dicated for th e p rim ar y t reat m en t of ch on drosarcom a of th e sp in e (st rong recom m en dat ion , ver y low qualit y eviden ce). How ever, radiat ion th erapy of at least 60 to 65 Gy equ ivalen ts is in d icated as an adjuvan t t reat m en t w h en th ere h as been in com plete resect ion or an in t ralesion al m argin (w eak recom m en dation, low -qualit y evidence). Becau se CHSs grow slow ly, local recurren ce an d m et ast ases m ay occu r m ore th an 10 years after rem oval of th e t u m or; t h e 5-year su r vival is n ot a sign i can t criterion for cu re. All rep orts w ith sh or ter follow -u ps h ave a lesser validit y

103

a

b

Fig. 9.7a–f A C5–C7 recurrent CHS in multiple exostoses in a 56-year-old woman, who submit ted to ve previous surgeries that were planned to remove piecemeal the exostosis without any concern about histological diagnosis or the criteria of surgical oncology, aim ing only to preserve function. (a) Conventional radiogram demonstrates ossi cations in the soft tissues. (b) MRI demonstrates a peripheral CHS involving the right lateral vertebral body, including the vertebral artery and the foram en, and expanding into the soft tissues. (continued on next page)

104

Chapter 9

c

d

e

Fig. 9.7a–f (continued ) (c) The specim en after en-bloc resection, con rmed by a pathologist to be a marginal margin tum or-free resection. The sacri ce of the C5-C6-C7 right nerve roots was required for achieving such an oncological result. The sagit tal section of the vertebral body and the canal are seen. (d,e) Radiogram s of the resected specimen. (f) Full circumferential reconstruction after sagit tal en-bloc resection of three vertebral bodies.

f

Chondrosarcoma w hen com paring paradigm s of treatm ent. Moreover, due to th e frequen t un derest im at ion of CHS, th e overall rate for cure is less th an th at for osteosarcom a at 5 years.

■ Technique Diagnosis To carefu lly determ in e th e variet y of CHS (cen t ral, p erip h eral [p rim ar y or secon dar y], m esen chym al, clear cell, or dedi eren t iated) an d its grade of m align an cy, th e p ath ologist n eed s an adequate am oun t of appropriate an d sign ifican t t issue. Th e su rgeon or th e in ter ven t ion al radiologist m u st be able to “read” th e im aging an d n d th e m ost represen tat ive specim en for th is pu rp ose. CT-gu ided biopsy sh ou ld be perfor m ed by a t rocar t h ick en ough (12 or 16 gauge) to provide an adequate am oun t of m aterial. Th e t rocar m ust be directed to an area of car t ilage, avoiding n ecrot ic areas or m at ure bon e. Th is is par t icularly relevan t w h en th e im aging st udies suggest a secon dar y CHS. All t issu e p ossibly belonging to th e prim it ive ben ign lesion sh ould be avoided, an d it w ou ld be best to perform th e biopsy from th e car t ilagin ous cap, possibly taking out a carrot of all its th ickn ess. Th e m ost aggressive an d in lt rat ive portion of th e t um or is the best location for t aking a biop sy sp ecim en . Grossly, CHSs are m ostly m yxoid an d di uen t. Th ey are signi cantly less vascularized an d pron e to bleeding th an oth er m align an cies, bu t th e t u m or h as an in t r in sic p ressu re t h at exp els th e soft , viscou s, jelly-like m aterial, w h ich quickly ow s in to the surroun ding spaces. Open biopsy can spread t um or in ext racom par t m en tal areas, such as th e epidural space, dram at ically red u cing th e e ect iven ess of a resect ion even if it is perform ed w ith an appropriate m argin . Even m in or con t am in at ion of th e soft t issu es by spilling of th e m yxoid con ten t s of th e t u m or can w orsen th e p rogn osis, an d n ot even a t rocar biopsy can preven t it . En -bloc resect ion w ith ou t biop sy can be con sidered, as suggested by Sten er,10 in selected cases w h ere im aging is path ogn om on ic an d en -bloc resec-

t ion is feasible. How ever, th is is rarely feasible even for h igh ly exp erien ced t u m or surgeon s.

Treatment Som e prognostic factors have been signi cantly related to p oor local con t rol an d disease su rvival. Th e outcom e of CHS seem s to be a ected by the size of the lesion, the histological grade,12 th e in cision al biop sy an d in adequ ate su rgical m argin s,1 p rim ar y su rger y ou t of a t u m or cen ter, and older patient age.11–15 On the other hand, en -bloc resect ion w ith t u m or-free m argin s is associated w ith th e best results,13–16 w h ereas m ost pat ien t s w ith recurren t t um ors die of th e d isease.15,16 A large m ult icen ter st udy repor ted n dings sim ilar to th ose of th e system at ic review,15 an d p rovided m oderate-qu alit y eviden ce supporting en -bloc resect ion in th e m an agem en t of CHSs of th e spin e. Th ere w as a st at ist ically sign i can t in crease in m or t alit y in pat ien t s w h o developed a local recurrence. Mortalit y was also d irectly related to th e m argin s, w ith a sign i can t in crease in m or t alit y in p at ien t s u n dergoing in t ralesion al su rger y. On ce it is accepted th at en -bloc resect ion is th e best opt ion for th e t reat m en t of CHS of th e sp in e, as in CHS in th e lim bs, t w o m ajor issu es arise: (1) w h en p lan n ing an en -bloc resect ion , th e decision -m aking process sh ou ld con sider th e possibilit y of fun ct ion al loss related to th e exten sion of t h e resect ion , an d (2) a st rategy sh ou ld be est ablish ed in cases w h ere en -bloc resect ion is n ot feasible or is n ot able to ach ieve fu ll t um or-free m argin s.

■ Common Problems Margins and Anatomic Sacrif ces On e of th e ch allenges in t reat ing spin al t u m ors is th e close p roxim it y of th e t u m or m argin s to im por t an t st ru ct u res su ch as th e sp in al cord, n er ve roots, an d large ar teries an d vein s. It is im por t an t for th e su rgeon to in form th e p at ien t th at it m ay be n ecessar y to sacri ce som e of these structures w hen at tem pting an en-bloc

105

106

Chapter 9 resect ion in ord er to m in im ize th e risk of recu rren ce an d im p rove th e p rogn osis. Th ere are speci c su rgical tech n iques for t um or resect ion in th e du ra m ater,17 th e cau da equ in a, th e aor ta,18 an d even th e sp in al cord. In case of ext racom par t m en tal CHS in th e sp in e, an en -bloc resect ion could in clu de n eigh boring st ru ct u res su ch as th e th oracic w all, w h ich cou ld represen t th e on cological m argin . En -bloc resect ion of cer vical CHSs is seldom possible an d tech n ically dem an ding; m ost of th e t im e on e or m ore cer vical n er ve root s n eed to be sacri ced (Fig. 9.7) w ith consequent funct ion al loss. In cases of sacral CHSs, w h en th e t u m or grow s above th e level of S3, en -bloc resect ion h as been associated w ith loss of sp h in cter con t rol.16

Inappropriate Oncological Excision Th e su rgeon m ay decide again st doing an en bloc resect ion , or th e pat ien t m ay refuse perm ission , for several reason s: • Tu m or-free m argin s can n ot be ach ieved because of invasion of both pedicles or in cases of large ext racom par t m en t al t u m ors, or for oth er reason s. • Th e surger y is tech n ically dem an ding (for exam ple, a h igh cer vical spin e t u m or). • Th e su rger y-related m orbidit y is deem ed to be excessive (for exam ple, sacral t um ors in pat ien ts over 80 years of age). • Th e su rger y en t ails excessive fu n ct ion al loss (for exam p le, sacral t u m ors in you ng pat ien ts). In th e decision -m aking process, th e t reatm en t team sh ou ld con sid er an d d iscu ss t h e h igh er m orbid it y t h at m ay be associated w it h t h e low est r isk of local recu r ren ce, th e n eed fu r th er su rgeries an d t reat m en t s, an d th e p ossible progression of m align an cy th at even t u ally m ay lead to th e p at ien t’s death . Perform ing eith er a gross total excision or an en -bloc excision w ith in t ralesion al m argin s, for in ten t ion al or acciden tal m argin violat ion , w ill leave residual t um or requiring local adju van t s. Th e im pact th at th e addit ion of radiat ion or local adjuvan t th erapy h as on local recurren ce an d su r vival sh ou ld be carefu lly con sidered. Altern at ive t reat m en t s sh ou ld be sough t

w h en en -bloc resect ion is n ot feasible or w h en th e p at ien t refu ses to give p erm ission . Radiat ion th erapy for CHS, w h eth er as a p rim ar y or adjun ct th erapy, h as ch anged sign i can tly in recen t years, as it h as been suggested th at doses h igh er th an 50 Gy m ay be ben e cial.19

Incorrect Diagnosis Th e m ost frequ en t m isdiagn oses en t ail con fu sing low -grade CHS w ith enchondrom a or osteoch on drom a, or con fusing dedi eren t iated CHS w ith osteosarcom a. Th e solut ion is to perform a t rocar biopsy un der CT scan con t rol, to obtain a representative specim en for pathological analysis, an d to avoid areas of n ecrosis as w ell as th e periph eral par t of th e t u m or. W h en a car t ilagin ou s cap is n oted, it is advisable to obtain a fu ll sam ple of th e tot al h eigh t .

Incorrect Biopsy Technique Op en biop sy of a CHS w ill con t am in ate all ad jacen t t issu es. W h en th is occu rs in th e spin e, w h ere in cluding all th e con tam in ated t issues in th e resected sp ecim en is di cu lt or im possible, it w ill negatively a ect the prognosis. Even w orse is to perform an in t ralesion al excision or a cord decom pression w ith out rst h aving don e biop sy. A full im aging w orku p sh ould alw ays be perform ed; a careful an alysis of a radiograp h ic d et ail can raise th e su spicion of a CHS. A decom pression perform ed in a CHS presum ing th e diagnosis of m etastasis can im prove n eurologic fu n ct ion , but it exposes th e pat ien t to t um or progression an d can preclude doing th e app rop riate surger y. Th is can resu lt in recu rren ce, p rogression of m align an cy, an d even t u ally death . In cases w h ere th e diagn osis or treatm ent is in doubt , the patient should alw ays be referred to an experien ced t um or cen ter. A biop sy p erform ed in an in exp er ien ced cen ter or an in ap p ropr iate in t ralesion al excision can h ave a n egat ive im pact on th e progn osis.

Undertreatment Perform ing an in t ralesion al excision in a case w h ere en -bloc excision is feasible exposes th e pat ien t to a p oorer local an d system ic p rogn osis. Th u s, su rgeon s sh ou ld obtain ap p rop riate

Chondrosarcoma on cological t rain ing before t reat ing prim ar y spin e t um ors. Th e criteria for en -bloc resect ion h ave been proposed based on th e Wein stein – Bor ian i–Biagin i (W BB) st agin g system ,13 bu t th ey are also discu ssed in corn erston e p ap ers by Roy Cam ille, Tom it a, an d, p rim arily, Sten er,10 w h o brillian tly described h ow to plan en -bloc resect ion s in th e spin e. Pat ien t s m ay deem th e fun ct ion al sacri ces en t ailed in th is t yp e of m ajor su rger y to be u n acceptable. But it w ould be a m istake to recom m en d an in t ralesion al su rger y to th ese p at ien t s if an en -bloc resect ion is feasible, as it is th e best stan dard of care.

Errors in the Decision-Making Process Th e op p osite m ist ake w ou ld be to perform en bloc resect ion in a grade 1 CHS of th e sacru m w ith out rst h aving a can did discussion w ith th e p at ien t . In th is case, a m ajor sacri ce of gen itourin ar y fun ct ion m u st be carefully explain ed to th e p at ien t . Pat ien ts u n derstan dably m ay refu se to accept su rger y th an en t ails su ch a sacri ce, even in th e case of a late recu rren ce after adjuvan t h igh -dose radioth erapy of an in t ralesion al m argin to save fu n ct ion (in ten t ion al t ran sgression ).

Incongruous Selection of Nonsurgical Treatments Conven t ion al radioth erapy as an adjuvan t an d ch em ot h erapy in low -grad e CHS are n otor iously in e ect ive. New radioth erapy tech n iques an d experim en tal m edical on cology protocols are op en ing n ew p ath w ays. A m u lt idisciplinary approach is m andator y to o er patients a w ider sp ect ru m of th erap eu t ic opt ion s.

■ Chapter Summary Ch on d rosarcom a is t h e t h ird m ost frequ en t m align an t p rim ar y bon e t u m or, after osteosarcom a an d Ew ing’s sarcom a, bu t it is th e m ost frequ en t m align an t t u m or in t h e sp in e (10% of all CHSs occu r in t h e sp in e). CHSs occu r m ain ly in adu lt s, p ar t icu larly elderly m en . Ra-

diograph ically, CHSs present as scat tering round ossi cat ion s w ith in a lyt ic area or as a m ass exp an ding th e soft t issue. Th e m an agem en t of CHS of th e spin e is ch allenging. It is a slow -grow ing t u m or, often exten sive at p resen t at ion , th at involves com plex an atom y. It respon ds par t ially to adjuvan t th erap ies an d ten ds to recu r if it is in adequ ately m an aged. Th e con dit ion ten ds to be leth al after a long d isease cou rse. Progression of m align an cy or d ed i eren t iat ion also h as been d escribed in long-su r viving p at ien t s an d is m ostly obser ved in recu rren t t um ors. Th e cou rse of th e disease d ep en ds on th e aggressiven ess of th e t um or, on its size, an d on th e feasibilit y of p erform ing an en -bloc resect ion . A literat u re review suggests th at en -bloc su rger y, as p ion eered in CHS by Sten er 10 in 1971, can give th e pat ien t th e best ch an ce of sur vival an d t h e low est rate of local recurrence, w h en it su cceeds in rem oving th e w h ole t u m or m ass togeth er w ith a con t in uous, even thin (m argin al m argin ), sh ell of h ealthy t issu e, an d w hen its expected m orbidit y out weighs the risk of d isease-related m orbid it y or d eat h . Carefu l p reop erat ive p lan n in g for su rgical t u m or rem oval an d sp in e recon st r u ct ion is m an dator y an d m u st be based on on cological an d su rgical staging. In t ralesion al excision (p erform ed because of th e an atom ic exten sion or size of th e CHS, or becau se th e p at ien t refu sed to u n dergo an en bloc p rocedu re) an d en -bloc resect ion w ith in t ralesion al m argin (because th e t um or m ass cam e apart , w ith con sequen t con tam in at ion ) seem to be e ect ive on ly in slow ing th e t u m or grow th an d do n ot preven t even a late (3 years or m ore) local recurrence. No chem otherapy protocol h as been validated to im prove th e con trol of th e disease in th ese cases, bu t in t ralesional excision can be com bin ed w ith m egavolt age or p roton -beam th erapy or oth er radiat ion th erapy tech n iqu es to deliver h igh doses (> 70 Gy) to residu al t u m ors. Recen t t rials u t ilizing n ew tech n ologies are p rom ising, as th ey redu ce th e su rger y-related m orbid it y, p ar t icu larly w h en an atom ic relevan t st r u ct u res can be resected , allow ing th e sam e local an d system ic con t rol t h at u n t il n ow h as been available on ly from su rgical p roced u res.

107

108

Chapter 9 Pearls

Pitfalls

Diagnosis Open biopsy can spread tumor in extracompartmental areas. Computer tom ography–guided biopsy should be performed with a trocar from an area of cartilage (from the cartilaginous cap of a peripheral CHS) and should avoid necrotic areas or mature bone, possibly taking out a carrot of all its thickness to provide an adequate am ount of material.

Incorrect Diagnosis The most frequent m isdiagnoses are low-grade CHS versus enchondrom a or osteochondroma, and dedi erentiated CHS versus osteosarcom a.

Treatment En-bloc resection with tum or-free m argins is associated with the best results. Prognosis is a ected by m any factors: histological grade, size of the lesion, increasing patient age, incisional biopsy, primary surgery out of a tumor center, but the most important is inadequate surgical margin. To achieve appropriate margins, dem anding surgery that entails morbidit y may be necessary, and functional sacri ces may be required. Intentional or incidental transgression of oncological principles may require adjuvant treatment. The role of radiation therapies is still unclear.

Acknow ledgment Th e au th ors are in debted to Carlo Piovan i for design , arch ive research , an d im age an d editorial assistan ce.

Incorrect Biopsy Technique Open biopsy of a CHS will contaminate all adjacent tissues. Intralesional excision performed for cord decom pression without previous biopsy exposes the patient to a high risk of local recurrence. Undertreatment Perform ing an intralesional excision in a case where en-bloc excision is feasible exposes the patient to a worse local and system ic prognosis. Inaccurate Decision-Making Process Perform ing en-bloc resection in a grade 1 CHS of the sacrum without accurate discussion with the patient and careful explanation of the loss of genitourinary function would be a mistake. Incongruous Selection of Nonsurgical Treatments Conventional radiotherapy and chem otherapy in low-grade CHS are not e ective. New radiotherapy protocols (including high doses of radiations and accelerated particles) and experimental medical oncology protocols are opening new pathways.

Refere nces Five Must-Read Refe rences 1. Fletcher CDM, Bridge JA, Hogen doorn PCW, Merten s F. W HO Classi cat ion of Tum ours of Soft Tissue an d Bon e. Lyon, France: In ternat ional Agency for Research on Can cer (IARC), 2013 2. Un n i KK, Inw ards CY. Dah lin’s Bon e Tum ors, 6th ed. Ph iladelph ia: Lippincot t William s & Wilkin s, 2010 3. Mat su da Y, Sakayam a K, Sugaw ara Y, et al. Mesen chym al ch on drosarcom a t reated w ith tot al en bloc sp on dylectom y for 2 con secu t ive lu m bar ver tebrae resulted in con t in uous disease-free sur vival for m ore th an 5 years: case rep or t . Sp in e 2006;31:E231–E236 PubMed 4. Hsu W, McCarthy E, Gokaslan ZL, Wolin sky JP. Clearcell ch on drosarcom a of th e lum bar spin e: case report an d review of th e literat u re. Neurosurger y 2011;68: E1160–E1164, discussion 1164 Pu bMed 5. Lit t rell LA, Wenger DE, Wold LE, et al. Radiograph ic, CT, and MR im aging features of dedi erentiated chon -

6.

7.

8.

9.

drosarcom as: a ret rosp ect ive review of 174 d e n ovo cases. Radiograph ics 2004;24:1397–1409 PubMed En n eking W F. Musculoskelet al Tum or Surger y, vols I an d II. Edin bu rgh , New York: Ch u rch ill Livingston e, 1983 Ozaki T, Lin dn er N, Hillm an n A, Röd l R, Blasius S, W in kelm an n W. In u en ce of in t ralesion al su rger y on t reat m en t ou tcom e of ch on d rosarcom a. Can cer 1996;77:1292–1297 PubMed Man kin HJ, Can t ley KP, Lip p iello L, Sch iller AL, Cam p bell CJ. Th e biology of h u m an ch on drosarcom a. I. Descript ion of t h e cases, grad ing, an d bioch em ical an alyses. J Bon e Join t Su rg Am 1980;62:160–176 PubMed Cam pan acci DA, Scoccian t i G, Fran ch i A, et al. Su rgical t reat m en t of cen t ral grade 1 ch on drosarcom a of th e ap pen dicu lar skeleton . J Or th op Trau m atol 2013; 14:101–107 Pu bMed

Chondrosarcoma 10. Sten er B. Tot al sp on dylectom y in ch on drosarcom a arising from th e seven th th oracic ver tebra. J Bon e Join t Surg Br 1971;53:288–295 Pu bMed 11. Borian i S, Saravanja D, Yam ada Y, Varga PP, Biagin i R, Fisher CG. Ch allenges of local recurren ce and cure in low grade m align an t t um ors of th e spin e. Spin e 2009;34(22, Su pp l):S48–S57 Pu bMed 12. Bergh P, Gu n terberg B, Meis-Kin dblom JM, Kin dblom LG. Progn ost ic factors an d outcom e of pelvic, sacral, an d sp in al ch on drosarcom as: a cen ter-based st u dy of 69 cases. Cancer 2001;91:1201–1212 PubMed 13. Borian i S, De Iu re F, Ban diera S, et al. Ch on drosarcom a of th e m obile spin e: repor t on 22 cases. Spine 2000;25:804–812 PubMed 14. Schoen feld AJ, Horn icek FJ, Pedlow FX, et al. Ch on drosarcom a of th e m obile sp in e: a review of 21 cases t reated at a single cen ter. Spin e 2012;37:119–126 PubMed 15. Fish er CG, Keyn an O, Boyd MC, Dvorak MF. Th e su rgical m anagem en t of prim ar y t um ors of th e spin e: in it ial result s of an ongoing prospect ive coh or t st udy. Spin e 2005;30:1899–1908 PubMed

16. Hsieh PC, Xu R, Sciu bba DM, et al. Lon g-ter m clin ical ou tcom es follow ing en bloc resect ion s for sacral ch ordom as an d ch on drosarcom as: a series of t w en t y con secu t ive p at ien t s. Sp in e 2009;34:2233–2239 10.1097/BRS.0b013e3181b61b90 Pu bMed 17. Krepler P, Win dhager R, Tom a CD, Kit z K, Kot z R. Dura resect ion in com bin at ion w ith en bloc sp on dylectom y for prim ar y m align an t t u m ors of th e spin e. Spin e 2003;28:E334–E338 PubMed 18. Gösling T, Pich lm aier MA, Länger F, Kret tek C, Hü fn er T. Tw o-st age m u lt ilevel en bloc sp on dylectom y w ith resect ion an d replacem en t of th e aort a. Eur Spin e J 2013;22(Su pp l 3):S363–S368 10.1007/s00586-0122471-0 Pu bMed 19. Fow eraker KL, Bur ton KE, Mayn ard SE, et al. High dose radioth erapy in th e m an agem en t of ch ordom a an d ch on drosarcom a of th e sku ll base an d cer vical spin e: Par t 1—clin ical outcom es. Clin On col (R Coll Radiol) 2007;19:509–516 PubMed

109

10 Osteogenic Sarcoma and Ew ing’s Sarcoma of the Spine Derek G. Ju, Patricia L. Zadnik, and Daniel M. Sciubba

■ Introduction Osteogen ic sarcom a an d Ew ing’s sarcom a are m align an t t u m ors th at gen erally arise in th e appen dicular skeleton , bu t in rare cases occur as prim ar y t u m ors of th e spin al colum n . Togeth er w ith ch on drosarcom a, th ese th ree sarcom as accou n t for 90% of all p rim ar y sarcom as of th e bon e.1 Classic osteogen ic sarcom a an d Ew ing’s sarcom a are by de n it ion h igh -grade

Diagnostic and Management Strategies De nitive diagnosis is needed in order to start patients on neoadjuvant chem otherapy prior to de nitive local control for Ewing’s sarcoma and osteosarcoma. Fine- or core-needle biopsy are appropriate methods if a speci c diagnosis is suspected and a lim ited am ount of tissue is required for diagnosis.3 If an open biopsy is required, the incision should be permanently marked to facilitate a wide excision at another tim e if necessary.3 When m etastatic disease is present, consultation with a m edical oncologist and radiation oncolo gist is helpful to determ ine if the patient is a surgical candidate. If there is direct soft tissue extension with bowel involvement, consultation with a general surgeon may further facilitate surgical planning. En-bloc procedures are designed to achieve wide margins, and the three main techniques in the spine are spondylectomy, sagit tal resection, and resection of the posterior arch.

t u m ors, an d, left u n t reated, th ey are associated w ith aggressive local and m etastatic spread. The goals of m an agem en t are to con t rol localized disease an d p reven t m et astat ic sp read, w h ich m ay be accom p lish ed by u t ilizing a com bin at ion of ch em ot h erapy, radiat ion , an d su rger y. Su rgical con t rol of m align an t t u m ors in t h e sp in al colu m n p resen t s a com p lex ch allenge du e to an atom ic con st rain t s th at m ake it di cu lt to obt ain a w ide m argin of excision .2

Sagit tal resections involve a wedge resection of the vertebral body with excision of the posterior elem ents, and are ideal in situations when tum or is con ned to an eccentric portion of the vertebral body, pedicle, or transverse process.3 En-bloc resection of the posterior arch is performed only when the tum or is located in the posterior arch without pedicle involvem ent. Spondylectomy involves the rem oval of the entire tum or in one piece together with portions of the posterior elem ents.4 This is the procedure of choice when the tum or is centrally located and involves no more than one pedicle.3 Previous data from our group has dem onstrated success with posterior-only approaches for enbloc resection of sacral tumors.5 If more than half of the sacroiliac joint is removed for tumor resection, hardware reconstruction with a transiliac bar or femoral allograft is required. Lesions not involving the sacroiliac joint can be removed with a low sacral amputation without any hardware reconstruction.

Osteogenic Sarcom a and Ewing’s Sarcom a of the Spine Th is ch apter review s t h e ep id em iology, clin ical p resen t at ion , rad iograp h ic n d in gs, path ology, and recom m ended m ultidisciplinar y m an agem en t opt ion s of th e m ost com m on sarcom as a ect ing th e bon e: osteogen ic sarcom a an d Ew ing’s sarcom a. Ch on drosarcom a of th e spine is discussed in Ch apter 9.

or sacrum .11 Nonetheless, these cases have been d ocum en ted to accou n t for as m u ch as h alf of th e spin al osteogen ic sarcom a cases in som e t um or databases.8 An oth er im p or tan t risk factor for osteogen ic sarcom a is p r ior rad iat ion exp osu re, w it h t h ose su bjected to h igh d oses at a you nger age h aving an elevated risk.12

Clinical Manifestation ■ Osteogenic Sarcoma

(Osteosarcoma) Epidemiology Osteogen ic sarcom a is a p rim ar y m align an t tum or of connective tissue origin, characterized by bon e or osteoid m atrix form ation w ithin the t u m or m ass. Th ere are m any t u m or varian t s w ith di eren t m orph ologies, w ith th e m ost com m on t ype being classic osteoblast ic osteogen ic sarcom a. It is th e m ost frequen t m align an t con dit ion of th e bon e, accoun t ing for 35% of prim ar y bon e m align an cies.6 With in th e sp in e, osteogen ic sarcom a is th e m ost com m on sarcom a an d m akes u p 3 to 15% of all p rim ar y sp in al t u m ors.7,8 Th e p eak in ciden ce of osteogen ic sarcom a in th e spin e occurs in th e four th decade of life.9,10 In con t rast , m ost pat ien ts w ith osteogen ic sarcom a in th e long bon es are u n der th e age of 30. Prim ar y osteogen ic sarcom a rarely arises from th e cer vical sp in e an d occu rs at a p rop ort ion ally h igh er frequen cy in th e th oracic spin e and sacrum , w ith the sacrum being th e prim ary site in 68 to 75%of pat ien t s.7,10 According to an an alysis of 430 pat ien ts w ith spin al osteogen ic sarcom a from th e Su r veillan ce, Ep idem iology, and End Results (SEER) database m aintained by th e Nat ion al Can cer In st it u te, th e in ciden ce is essen t ially equ al bet w een th e sexes (1.2 to 1), an d spin al t um ors are rare in people of Asian , Nat ive Am erican , an d African descen t .10 Oth er con dit ion s associated w ith in creased risk of develop ing osteogen ic sarcom a in clu de Paget ’s d isease of bon e, en ch on d rom atosis, h ered it ar y m u lt ip le exostoses, an d brou s dysplasia.4 Degen erat ion in to a pagetoid osteosarcom a occu rs in less th an 1% of pat ien ts, an d of th ose cases on ly 6% develop in th e ver tebrae

Th e m ost frequ en t p resen t ing sym ptom is axial or radicular pain .2,7 Neurologic de cits are th e secon d m ost frequen t p resen t ing sym ptom an d can occur in over 50%of pat ien ts.7,13 Bow el an d bladder sym ptom s are less com m on , an d u sually presen t later as th e t um or progresses.2 In th e Cooperat ive Osteosarcom a St udy Group t rial, th e m edian delay in diagn osis from th e on set of sym ptom s w as 5 m on th s.7 According to the SEER database, 28%of patients diagnosed w ith osteogen ic sarcom a of the spin e presented w ith gross m et astat ic disease.10

Radiographic Features Th e role of com p u ted tom ograp hy (CT) an d m agnetic resonance im aging (MRI) has becom e in creasingly im p or tan t in th e diagn osis an d st aging of spin al osteogen ic sarcom a. Conven t ion al plain radiograp h s are u su ally obtain ed prior to on cological referral, an d are u sefu l in determ in ing th e aggressiven ess of a bon e lesion . Plain radiograp h s t yp ically reveal a den se, solid, an d sm ooth ap pearan ce of osteogen ic sarcom a.9 How ever, th e n dings m ay be variable, ranging from a osteosclerot ic appearan ce in 20%of pat ien ts to a seem ingly n orm al radiograph.2,7 For this reason, CT is the recom m ended m odalit y for diagn ost ic im aging of osteogen ic sarcom a, an d scan s sh ow sign i can t m at rix calci cation in 80% of cases.14 On MRI, th e solid, n on m in eralized p or t ion s of th e t um or are hypoin ten se on T1-w eigh ted sequ en ces an d hyperin ten se on T2-w eigh ted sequen ces 9 (Fig. 10.1). Th ere m ay be focal region s of hypoin ten sit y on all pulse sequen ces, an d u id– u id levels are com m on in th e telangiectat ic subt ype.15 In con t rast , m in eralized t um ors appear isoin ten se to bon e on T1- an d T2-w eigh ted sequ en ces 4 (Fig. 10.1). Edem a in

111

112

Chapter 10

a

b

Fig. 10.1a–c Sagit tal T1 (a), T2 (b), and fatsuppressed (c) m agnetic resonance imaging (MRI) scans demonstrating large sarcoma of the left hemipelvis extending into the left retroperitoneum, with extension into the left psoas, iliacus, and iliopsoas muscles. The lesion is heterogeneously enhancing secondary to inconsistent mineralization of the tumor.

c

th e su rrou n ding t issu es is often seen , bu t is n ot speci c for osteogen ic sarcom a.9 Th e m ajorit y of osteogen ic sarcom a pat ien t s w ho presen t w ith m et ast at ic disease h ave detect able lung m et ast ases, an d a CT of th e ch est , abdom en , an d pelvis can assist w ith t um or st aging.16 Ch est CT is th e m ost sen sit ive im aging tech n iqu e in detect ing lung m et astases, an d sh ou ld be u sed w h en available.16 A tech n et ium -99m m ethylen e diph osph on ate ( 99m Tc) bon e scan m ay be n ecessar y to iden t ify addit ion al bon e m et astases an d skip lesion s.

Histopathology Grossly, osteogen ic sarcom a exh ibit s a red , gr it t y, gran u lar qu alit y d u e to osteoid for m at ion .4 Foci of h em or rh age or n ecrosis are com m on h istological n d ings.4 Th e m ajor it y of osteogen ic sarcom as are by de n it ion h igh grade an d h ave a m edullar y origin .17 On a cellu lar level, osteogen ic sarcom a is ch aracterized by spin dle cells w ith n u clear p leom orph ism .18 Th e h istological hallm ark of osteogenic sarcom a is th e produ ct ion of bon e or osteoid m at rix

Osteogenic Sarcom a and Ewing’s Sarcom a of the Spine w ith in th e t um or, an d is th e key to diagn osis.9 Convent ional osteogenic sarcom a cells produce di eren t t yp es of ext racellu lar m at rix, h en ce th e division in to osteoblast ic, ch on droblast ic, an d broblast ic su bt ypes.18 How ever, m ost t u m ors do n ot sor t n eatly in to a su bcategor y an d presen t w ith a m ixed h istology, an d th ere is n o sign i can t d i eren ce in p rogn osis or t reat m en t bet w een th e subt ypes.17 Oth er rare bu t n ot able h istological t yp es in clude telangiect at ic, sm all-cell, an d ep ith elioid osteogen ic sarcom as. Telangiect at ic osteogen ic sarcom a sh ares feat u res w ith an eur ysm al bon e cyst s, an d th e lesion is com p osed of m u lt ip le blood- lled sin u soids.17 With th e applicat ion of m odern -era ch em oth erapy agen t s, th e p rogn osis for telangiect at ic osteogen ic sarcom a h as im proved an d n ow carries a p rogn osis sim ilar to t h at of t h e conven t ion al t yp e.18 Sm all-cell osteogen ic sarcom a m ay be con fu sed w it h Ew ing’s sarcom a du e to it s rou n d, hyp erch rom at ic n uclei an d CD99 posit ivit y.17 Fur th er, t ran slocat ion s m im icking Ew ing’s sarcom a bet w een ch rom osom es 11 an d 22 h ave also been obser ved 17 (see Histop ath ology, in th e Ew ing’s Sarcom a sect ion of th e ch apter, below ). Fin ally, epith elioid osteogen ic sarcom as are poorly differen t iated an d m ay resem ble carcin om as.17

Management Cu r ren t ly, t h ere are n o st an dard on cological st agin g scales u sed for osteogen ic sarcom a. Ap p licat ion of bon e t u m or st aging system s, su ch as th ose from th e Musculoskelet al Tu m or Societ y an d th e Am erican Join t Com m it tee on Can cer are lim ited in e cacy du e to th e h igh grad e n at u re of osteogen ic sarcom a an d it s lack of lym ph n ode involvem en t .1 Fu r th er, th e un iqu e an atom ic con st rain t s for pat ien ts w ith osteogen ic sarcom a of th e spin e com pared w ith th at of th e ext rem it ies m ay lim it th e ap plicat ion of a gen eralized grading sch em e. Recen tly, a grading scale for m align an t osseous spin al n eoplasm s, in cluding osteogen ic sarcom as, h as been p rop osed in ord er to en h an ce r isk st rat icat ion of t reat m en t can didates.19 Th e m ost im por t an t progn ost ic variables are pat ien t age, m et ast at ic st at u s, an d exten t of local t u m or invasion .19

Th e gold st an dard of t reat m en t for p at ien t s w ith localized osteogen ic sarcom a involves a m u lt idiscip lin ar y app roach , con sist ing of n eoadjuvan t ch em oth erapy w ith w ide local surgical excision . Th e lan d m ark dat a for m u lt im odal t reat m en t of osteogen ic sarcom a of th e lim bs w as provided by Lin k et al,20 in w h ich pat ien t s u n dergoing lim b resect ion w ere ran dom ly assign ed to adjuvan t ch em otherapy or obser vat ion on ly. Th e relapse-free su r vival at 2 years w as st at ist ically sign i can t bet w een th e t w o coh or t s, w ith 17% of th e cont rol group ach ieving relapse-free su r vival com pared to 66% in t h e adju van t ch em oth erapy grou p . Gh erlin zon i et al21 perform ed a ret rospect ive review on 355 p at ien t s w it h osteogen ic sarcom a of th e lim bs, an d foun d th at th e in ciden ce of local recurren ce w as related to surgical m argin an d to t u m or n ecrosis in d u ced by n eoadju van t ch em oth erapy. Ch em otherapy h as also been sh ow n to provide im proved local con t rol an d sur vival for p at ien ts w ith osteogen ic sarcom a of th e spin e, alth ough h igh -qu alit y st u dies are lacking.7,13,22 Cu rren t n eoadjuvan t ch em oth erapy protocols use a com bin at ion of m eth ot rexate, cisplat in , doxoru bicin , ifosfam ide, BCD (bleom ycin , cytoxan , dact in om ycin ), etoposide, an d m uram yl t rip ept ide.6 Sh ives et al13 evalu ated 27 pat ien ts w ith spinal osteogenic sarcom a and found a longer survival in patients w ith adjuvant therapy, alth ough n o st at ist ical analysis w as perform ed. Re ect ing th ese resu lts, th e 2009 con sen su s recom m en dat ion of th e Spin e On cology St u dy Group (SOSG) st ated th at n eoadjuvan t ch em otherapy o ers signi cant im provem en ts in local control an d long-term survival for spin al osteogen ic sarcom a.6 Th e on ly e ect ive su rgical t reat m en t of localized sp in al osteogen ic sarcom a is tot al sp on dylectom y or w ide local excision w ith vertebral colum n recon st ru ct ion 4 (Fig. 10.2), Alth ough th e st rongest eviden ce for aggressive surger y derives from t reat m en t of pat ien t s w ith lim b t u m ors, as eviden ced by Lin k et al20 an d Gh erlin zon i et al,21 th ere is eviden ce of spin al osteogen ic sarcom a as w ell. DeLan ey et al23 ret rospect ively review ed 41 pat ien t s w ith osteogenic sarcom a, including eight patients w ho h ad a spin al t um or. Th e auth ors foun d a h igh er

113

114

Chapter 10

Fig. 10.2 Coronal computed tomography (CT) demonstrating an L4-5 vertebrectomy and hemipelvectomy with hardware reconstruction.

local con t rol rate for pat ien ts u n dergoing gross tot al resect ion w ith n egat ive versus posit ive m argin s (78% vs 68%), alth ough th e di eren ce w as n ot st at ist ically sign i can t . Ozaki an d colleagues 7 evalu ated 22 pat ien ts w it h osteogen ic sarcom a of t h e a xial sp in e. Th ere w as a sign i can t sur vival di eren ce bet w een ve pat ien ts w h o u n der w en t w ide or m argin al surger y an d 17 pat ien ts w h o did n ot , leading th e au th ors to con clude th at at least m argin al excision is w arran ted in poten t ially resect able t u m ors. Likew ise, Su n d aresan et al22 review ed t h e resu lt s of 24 p at ien t s w it h sp in al osteosarcom a w h o u n der w en t aggressive t reat m en t (aggressive resect ion , radiat ion , an d ch em ot h erapy) com p ared w it h p at ien t s w h o u n d er w en t less aggressive t reat m en t (lim ited resect ion an d radiat ion ). Th e au th ors found th at pat ien t s w ith a m ore aggressive t reat m en t h ad an im proved su r vival, alth ough n o stat ist ical an alysis w as don e due to th e low sam ple size. More recen tly, Mu kh erjee an d colleagu es 24 exam in ed 158 p at ien t s in th e SEER dat abase

w ith localized spin al osteogen ic sarcom a an d foun d a th reefold sur vival (h azard rat io [HR], 0.382; 95% con den ce in ter val [CI], 0.21–0.69) for pat ien t s w ith su rgical resect ion over biopsy alon e. Th e dram at ic en h an cem en t in sur vival w as in dep en d en t of ot h er var iables su ch as t u m or locat ion or p at ien t age; h ow ever, t h e exten t of resect ion an d su rgical m argin s w ere n ot available in th e SEER database. Th e SOSG st rongly recom m en ds th e u se of en -bloc resect ion of osteogen ic sarcom a w ith w ide m argin s, as it provides im proved local con t rol an d poten t ially im proved su r vival.6 Th e m orbidit y an d m or talit y of en -bloc procedures are con siderable, an d th ey sh ou ld be p erform ed on ly at experien ced cen ters w ith m u lt idisciplin ar y team s. Th e overall 5-year su r vival rate in all p at ien t s w h o h ave spin al osteogen ic sarcom a is 18%.10 Th e est im ated m edian sur vival in pat ien t s w h o presen t w ith localized disease an d receive aggressive t reat m en t is 18 m on th s.10 An est im ated 7% of pat ien ts w ith spin al osteogen ic sarcom as presen t w ith m etastat ic disease, an d h ave an ap proxim ate m ed ian su r vival of 7 m on th s.10 Radiat ion th erapy m ay p lay a role in palliat ive care for th ose w ith un resectable or m etastatic disease; how ever, osteogenic sarcom as are radioresistant an d do not respond w ell to st an dard doses.4 An im p rovem en t in su r vival rates h as also been associated w it h th e use of radiat ion w h en com bin ed w ith su rger y.23,24 Th is m ay be due to th e e cacy of radiat ion t reat m en t in t reat ing m icroscop ic or m in im al residual disease after surger y.

■ Ew ing’s Sarcoma Epidemiology Ew ing’s sarcom a rep resen t s a fam ily of n eoplasm s origin at ing from a precursor neural cell, in clu ding classic Ew ing’s sarcom a an d prim it ive n eu roectod erm al t u m or.4,25 Ew in g’s sarcom a is t h e secon d m ost com m on can cer of bon e in ad olescen t s an d you ng ad u lt s, after osteogen ic sarcom a, w ith an in ciden ce of 2.1 per 1 m illion ch ildren in th e Un ited States.26

Osteogenic Sarcom a and Ewing’s Sarcom a of the Spine Th is sarcom a of th e sp in e ten ds to p resen t at a you ng age, w ith a p eak in cid en ce in th e secon d decade an d an age range sp an n ing th e rst th rough th ird decades.9 Th ere is a sligh t m ale p redom in an ce in th e sp in e (1.8 to 1). Th e con d it ion is exceedingly rare in peop le of Asian , Nat ive Am erican , or African an cest r y.10 Th e cause of Ew ing’s sarcom a is n ot clear, and is m ost likely due to spon tan eous gen et ic t ran slocat ion s rath er th an fam ilial or environ m en t al factors.25 Fu r t h er, u n like osteosarcom a, th ere is n o associated r isk from p r ior exp osu re to radiat ion .4 Th e sp in e is an u n com m on site, w it h on ly 3 to 15% of Ew ing’s sarcom a occu rring in th e sp in e.2 How ever, Ew ing’s sarcom a is th e m ost com m on m align an t ver tebral t um or foun d in ch ild ren .27 It can a ect any segm en t of t h e sp in e, as w ell as or igin ate w it h in t h e p aravertebral m u scles an d invad e in to t h e ep id u ral sp ace.4 With in th e a xial skeleton , Ew ing’s sarcom a is m ost com m on ly fou n d in th e sacr u m an d lum bar spin e, w ith com m on involvem en t of the sacral ala and posterior elem ents, respectively.3 Within the sacrum , the tum or m ay grow to a large size before th e on set of pain . Th erefore, th e in sidiou s m align an cy m ay rem ain u n d etected for a prolonged period of t im e.4

Clinical Manifestation On p resen t at ion , th e p rim ar y sym ptom of spin al Ew ing’s sarcom a is localized pain w ith variable in ten sit y, w h ich occu rs in vir t u ally all pat ien ts.2 Th is pain is often m istaken for n orm al grow ing pain s or sport s-related inju ries in t h e ad olescen t p op u lat ion .25 Neu rologic d e cit s are often seen , occu rr ing in 40 to 60% of p at ien t s.2 Bow el an d blad d er sym ptom s are rare an d ten d to occu r later in th e disease progression .2 System ic sym ptom s su ch as fever an d in dicators of a ch ron ic in am m ator y st ate m ay also occu r an d lead to a m ist aken d iagn osis of in fect ion .4,25 Th ese fairly n on sp eci c sym ptom s com m on ly lead to a delay in th e establish m en t of diagn osis, allow ing th e m align an t t u m or to progress. According to th e SEER dat abase, 34%of 430 pat ien t s w ith Ew ing’s sarcom a of th e sp in e p resen t w ith gross m et ast at ic disease.10

Radiographic Features Th e m ost im portant diagn ostic im aging m odalities for evaluation of potential Ew ing’s sarcom a are CT an d MRI. Un like osteogen ic sarcom a, p lain rad iograp h s are rarely u sefu l in d iagn osing Ew ing’s sarcom a, as th is disease process cau ses a “m ot h -eaten ” d est r u ct ion of bon e rat h er t h an com p lete osteolysis.2,15 Rad io grap h ic im ages of early disease can range from a seem ingly n egat ive radiograph to subtle n dings such as osteolysis an d h azin ess of th e en d plates.28 Com puted tom ography is a useful com plem entary m odalit y to evaluate the bony elem ents a ected by Ew ing’s sarcom a. An osteolytic m ass is detected in app roxim ately 90%of cases, w ith rare varian ts th at m ay presen t w ith a m ixed lytic/sclerot ic or purely sclerot ic m orphology.28 Th e m ajor it y of t u m ors in t h e m obile sp in e involve th e posterior elem en ts, w h ereas m ost t u m ors in t h e sacr u m involve t h e sacral ala 28 (Fig. 10.3). A CT of t h e ch est , abd om en , an d pelvis is recom m ended for surgical staging. Th e m ost com m on sites for m et ast at ic disease are th e lu ngs, bon e, an d bon e m arrow. Fu rt h er, a 99m Tc bon e scan d em on st rates in ten se t racer u ptake in Ew ing’s sarcom a.9,16 Th e recom m en ded m odalit y to evalu ate Ew ing’s sarcom a is MRI w ith gadolin iu m en h an cem en t , w h ich en ables su cien t evalu at ion of th e soft t issu e m ass an d it s su rrou n ding an atom y (Fig. 10.4). Th e t u m or bu lk is isoin ten se to bon e m arrow on T1-w eigh ted sequ en ces an d isoin ten se to hyper in ten se to bon e m ar row on T2-w eigh ted sequ en ces.9 Th e cor tex of th e bon e is u su ally p reser ved, an d pat ien t s m ay p resen t w ith a p ath ological fract u re if th e lesion occu rs w ith in th e ver tebral colu m n .15

Histopathology Grossly, Ew in g’s sarcom a ap p ears as a grayw h ite t u m or w it h areas of h em or rh age an d necrosis.3 Evidence of necrosis an d at ypical h istological features have been found to have w orse progn ost ic valu e in pat ien ts w ith prim ar y Ew ing’s sarcom a of th e ext rem it ies.4 On a m icroscopic level, Ew ing’s sarcom a con sists of sm all,

115

116

Chapter 10

b

Fig. 10.3a,b Preoperative axial (a) and sagit tal (b) CT scan dem onstrating osteolytic destruction of the left-side sacral ala. T1- and T2-weighted MRI scans (not shown) demonstrated tumor growth and extension.

a

rou n d cells w it h oval, u n ifor m n u clei.9,18 Th e cellular boun daries are in dist in ct , an d give th e appearan ce of a syn cyt ium w ith m u lt iple n u clei.9 Occasion ally presen t are m itot ic gu res, w ith ap optot ic an d kar yopykn ot ic cells.3 Th e di eren t ial diagn osis for sm all, rou n d cell t um ors in clu des lym p h om a of bon e, m et ast at ic neuroblastom a, em bryonal rhabdomyosarcom a, sm all cell osteosarcom a, and osteom yelitis; thus, diagn osis of Ew ing’s sarcom a is di cu lt on ligh t m icroscopy alon e.1,4 De n it ive diagn osis is aid ed by th e u se of im m un oh istoch em ist r y, elect ron m icroscopy, an d cytogen et ic m eth ods. Viable t issue for h istopath ology, sterile t issu e for cytogen et ic, an d a sm all sam ple for elect ron m icroscopy sh ou ld be t aken d u ring biop sy for p at h ological evalu at ion . On im m u n oh istoch em ist r y, st ron g expression of th e in tegral m em bran e glycop rotein CD99 (MIC2) is ch aracterist ic of Ew ing’s sarcom a, bu t CD99 is n ot a sp eci c m arker for Ew ing’s sarcom a.29 Vim en t in p osit ivit y is an oth er sen sit ive bu t n ot sp eci c m arker for Ew ing’s sarcom a.29 Gen et ically, Ew ing’s sarcom a, like prim it ive n euroectoderm al t um or, is ch aracterized by a recip rocal t ran slocat ion bet w een th e EW S gen e on ch rom osom e 22 an d a m em ber of th e ETS

fam ily of transcription factors.29 Approxim ately 85% of th e t u m ors con t ain a t(11;22)(q24;q12) t ran slocat ion result ing in a EWS-FLI1 fu sion p rotein , an d 10% of t u m ors con tain a t(21;22) (q22;q12) tran slocation resulting in a EWS-ERG t ran script .29 Th ese ch im eric t ran script ion factors ret ain their poten t abilit y to in duce t ran script ion of variou s t arget gen es requ ired for t u m or grow th . Notably, th e t ype 1 EWS-FLI1 fu sion p rotein is recogn ized to be an in dicator of favorable p rogn osis d u e to it s less p oten t t ran sact ivat ion cap abilit ies.25 Add it ion ally, u p to 20% of Ew in g’s sarcom as exh ibit ot h er gen et ic abn or m alit ies su ch as m u t at ion s of t h e t um or suppressor gen es p53 an d ink4A, w h ich are associated w ith a w orse progn osis.29

Management Cu r ren t ly, t h ere are n o u n iversally accepted on cological st agin g system s for Ew in g’s sarcom a. St agin g system s for p r im ar y bon e t u m ors are available from t h e Mu scu loskelet al Tu m or Societ y an d th e Am erican Join t Com m it tee on Can cer, bu t th ey are n ot com p letely applicable an d m ay h ave lim ited ut ilit y in Ew ing’s sarcom a.25 Alth ough th e m ost im por t an t p rogn ost ic var iables—p resen ce of m et ast at ic

Osteogenic Sarcom a and Ewing’s Sarcom a of the Spine

a

117

b

Fig. 10.4a,b (a) Gadolinium -enhanced sagit tal T1-weighted MRI demonstrating a lesion that is hypointense to the bone m arrow occupying S1-S3. There is involvem ent of the anterior and posterior

elements at these levels. (b) T2-weighted gadolinium -enhanced sagit tal T2-weighted MRI dem onstrates a heterogeneously enhancing lesion in S1-S3.

d isease an d prim ar y t um or size—are in clu ded in th ese st aging system s, th e n odal st at u s an d grade are irrelevan t becau se Ew ing’s sarcom a rarely h as lym ph n ode involvem en t an d is h igh grade by de nition.25,30 Further, the staging system s do n ot t ake in to accoun t th e locat ion of th e prim ar y t u m or, w h ich can h ave sign i can t progn ost ic valu e. For p at ien t s w ith n o evid en ce of m etastases at diagn osis, age less th an 15 years an d t u m or less th an 8 cm in size w ere fou n d to be in depen den t p redictors of im p roved su r vival based on a large coh or t of p at ien t s w it h Ew in g’s sarcom a of th e bon e from variou s p rim ar y locat ion s.31 Sp in al locat ion is associated w ith a w orse progn osis, du e to th e an atom ic lim itat ion s an d d i cu lt y in obt ain in g local con t rol in th e axial skeleton . For p at ien t s w ith m et ast at ic disease, th ose w ith isolated pulm on ar y m et ast asis an d skip m et astases h ave a bet ter prognosis than other sites of m etastatic spread.25 Th e t im e from diagn osis to recu rren ce is a sign i can t progn ost ic factor; recurren ce w ith in 2 years of d iagn osis is associated w ith a decreased 5-year sur vival.31 As m en tioned earlier,

certain histological or genetic characteristics of the pat ient’s t um or m ay confer som e prognostic valu e as w ell. Regardless of t um or locat ion or exten t , system ic ch em oth erapy is th e m ain stay of t reatm en t for all p at ien t s w it h Ew in g’s sarcom a u n less con t rain d icated .25 Th e classic ch em otherapeut ic agents used in Ew ing’s sarcom a are vin crist in e, act in om ycin D, cycloph osp h am ide, an d doxor ubicin (VACD). Th e developm en t of e ect ive m u lt iagen t th erapy for Ew ing’s sarcom a has been a m ajor treatm ent breakthrough, con t ribu t ing to an in crease in th e 5-year su rvival from 5 to 10% in th e 1980s to th e curren t 5-year sur vival of 65 to 70%.32–34 Th e valu e of ch em oth erapy w as illu st rated in 1981 by th e rst In tergrou p Ew in g’s Sarcom a Grou p t h at docum ented th at th e addition of doxorubicin to VAC con ferred an im p roved su r vival over VAC therapy alone.34 Th e VACD regim en w as further est ablish ed as st an dard care of Ew ing’s sarcom a, as at least eigh t clin ical t rials from 1973 to 2001 using VACD dem on st rated im proved sur vival.25,33 Later, second-generation regim ens in corporated ifosfam id e an d etoposide to t h e

118

Chapter 10 stan dard VACD regim en , w h ich im proved even tfree sur vival for pat ien ts w ith localized disease.33 Th rough ou t th ese periods th e role of ch em oth erapy h as been su pp or ted for Ew ing’s sarcom a in all t u m or locat ion s, in clu d in g t h e sp in e.32,35,36 Du e to t h e p oten t ial for ch em ot h erapy-in du ced card iac, ren al, an d h ep at ic toxicit y, evaluat ion of th ese organ s sh ou ld be a crit ical par t of th e pret reat m en t assessm en t . Th e t reat m en t of p at ien t s w ith localized Ew ing’s sarcom a of th e sp in e begin s w ith n eoadjuvan t ch em oth erapy, w h ich is design ed to facilitate fut ure local in ter ven t ion s by sh rin king the soft tissue m ass.25 For patients w ho presen t w ith eviden ce of spin al cord com pression but are neurologically stable, an addit ional bene t of ch em oth erapy m ay be relief of spin al cord com prom ise.37 In 2009 th e SOSG st rongly recom m en d ed t h e use of n eoadjuvan t ch em oth erapy for p at ien t s w ith Ew ing’s sarcom a of th e sp in e du e to sign i can t im p rovem en t s in local control and long-term survival.6 Follow ing chem otherapy, restaging of the tum or should be p erform ed prior to plan n ing of de n it ive local con t rol via rad iat ion th erapy or surger y. Radiat ion th erapy for local con t rol h as historically been th e p referred t reat m en t du e to th e radiosen sit ivit y of Ew ing’s sarcom a. How ever, th is approach is used less frequen tly in pract ice n ow

du e to con cern over in du cing secon dar y m alignancies, adverse e ects on bone grow th, and the im provem en t of advan ced surgical tech n iqu es for w ide resect ion s in th e spin e an d sacrum .25,37 Radioth erapy sh ou ld be used in cases of u n resect able t um ors or after an in t ralesion al or m argin al su rgical resect ion . As t h e m ajor it y of Ew ing’s sarcom a pat ien t s are pediat ric pat ien t s, th e eld of radiat ion m u st carefu lly ou tlin e th e ver tebral body of in terest in order to preven t asym m et ric grow th .29 Su rger y for local con t rol of Ew ing’s sarcom a sh ou ld be con sidered in all cases in w h ich th e prim ar y t u m or can be com p letely rem oved in an en -bloc fashion , and subsequen t reconstruction w ith h ardw are sh ou ld be p erform ed (Fig. 10.5). Bacci et al38 in 2006 review ed th e ou tcom es of 512 pat ien t s w ith Ew ing’s sarcom a over a 20-year span at a single in st it ut ion , com paring p at ien ts w h o u n der w en t su rger y alon e or surger y plus radiat ion w ith pat ien ts t reated w ith radiation alone. A signi cant im provem ent in local con t rol (88.8% vs 80.2%, resp ect ively; p < 0.009) an d 5-year disease-free su r vival (63.8% vs 47.6%, respect ively; p < 0.0007) w as foun d in pat ien t s treated w ith su rger y, a n ding th at w as esp ecially pron ou n ced in p at ien t s w h o w ere foun d to have adequate surgical m argins. Generally, aggressive extralesional surgical

b

a

Fig. 10.5a,b (a) Coronal CT scan demonstrating femur allograft and hardware reconstruction. (b) An anteroposterior (AP) radiograph dem onstrating the extent of the surgical reconstruction.

Osteogenic Sarcom a and Ewing’s Sarcom a of the Spine resect ion p rovides an im provem en t in overall su r vival an d local con t rol.35,38–40 How ever, it is im por t an t to n ote th at st at ist ically sign i can t im provem en t s in th ese st udies w ere fou n d in p at ien t s w ith t um ors in th e lim bs, n ot th e axial sp in e, as h igh -qu alit y st u dies sup port ing su rgical t reat m en t of Ew ing’s sarcom a in th e sp in e are lacking. Th e SOSG issu ed a w eak recom m en dat ion in 2006 in dicat ing that en -bloc resection provides im proved local control, but not im proved overall su r vival for spin al Ew ing’s sarcom a.6 Th e recom m en dat ion also su pp or ts th e alter n at ive u se of rad iot h erapy for local con t rol alon e or to su p plem en t an in com plete resect ion . More recen tly, Mu kh erjee et al24 st u died 182 pat ien ts w ith spin al Ew ing’s sarcom a an d foun d a m ore th an t w ofold sur vival advan tage (HR, 0.494; 95% CI, 0.26–0.96) for p at ien t s t reated w ith su rgical resect ion over biop sy alon e. Gen erally, th e 5-year even t-free su r vival for pat ien ts w ith localized spin al Ew ing’s sarcom a is rep or ted to be 50 to 60% w ith aggressive m u lt im odalit y t reat m en t .4 Based on dat a from th e SEER regist r y, th e m edian su r vival of p at ien t s w ith localized sp in al t u m ors w as 90 m on th s.10 Patients w ith m etastatic or recurrent Ew ing’s sarcom a h ave a poor progn osis an d rem ain a th erap eu t ic ch allenge, an d t h e 5-year su r vival rem ain s ap p roxim ately 25 to 30%w ith a m edian sur vival of 20 m on th s.10 Th e use of h igh -dose ch em oth erapy an d au tologou s bon e m arrow t ran splan tat ion h as n ot im proved su rvival rates, an d is act ually associated w ith an 8%in ciden ce of t reatm ent-related leukem ia and m yelodysplastic syndrom es.25 Patients w ith recurren t disease fare even w orse, w ith a 5-year su r vival of less th an 20%.37 Alth ough th e rare solit ar y lesion m ay be su rgically cu red , t h e m ajorit y of p at ien t s m ay on ly be t reated w ith salvage ch em oth erapy regim en s.

■ Clinical Considerations W h en presen ted w ith a pat ien t w ith suspected Ew ing’s sarcom a or osteogen ic sarcom a, rst perform a biopsy and involve m edical oncology

early in t h e t reat m en t p lan . En -bloc resect ion o ers t h e best ch an ce for p rolon ged su r vival in p at ien t s w it h osteogen ic sarcom a an d to a lesser exten t for Ew ing’s sarcom a p at ien t s; h ow ever, th ere is n o de n it ive cure for eith er disease. Neoadjuvan t ch em oth erapy can im prove local con t rol an d su r vival for Ew ing’s an d osteosarcom a pat ien t s, but due to th e poten t ial for ch em oth erapy-in duced cardiac, ren al, an d h ep at ic toxicit y, evaluat ion of th ese organ s sh ou ld be a crit ical par t of th e p ret reat m en t assessm en t . Radiat ion m ay ser ve a role for pat ien t s w h o refuse surger y, or in cases of in t ralesion al resect ion .

Prognosis The Spine Oncology Study Group (SOSG) states that neoadjuvant chemotherapy o ers signi cant improvements in local control and long-term survival for spinal osteogenic sarcoma. The SOSG strongly recomm ends the use of enbloc resection of osteogenic sarcom a with wide margins, as it improves local control and potentially improves survival. The overall 5-year survival rate in all patients who have spinal osteogenic sarcoma is 18%. The estimated median survival in patients who present with localized disease and receive aggressive treatment is 18 m onths. The 5-year survival of Ewing’s sarcoma has increased from 5 to 10%in the 1980s to the current 5-year survival of 65 to 70%. Based on data from the SEER registry, the m edian survival of patients with localized spinal tum ors is 90 m onths. Patients with metastatic or recurrent Ewing’s sarcoma have a poor prognosis; the 5-year survival remains approximately 25 to 30% with a m edian survival of 20 months. The SOSG strongly recom mends the use of neoadjuvant chemotherapy for patients with Ewing’s sarcom a of the spine due to signi cant improvements in local control and long-term survival. The SOSG issued a weak recomm endation promoting en-bloc resection for improved local control, but not improved overall survival for spinal Ewing’s sarcoma. The recom mendation also supports the use of radiotherapy for local control alone or to supplem ent an incomplete resection. Osteosarcoma is radioresistant; Ewing’s sarcom a is radiosensitive, but radiation exposure may increase the risk of secondary m alignancies and should be used with extreme care in the pediatric population.

119

120

Chapter 10

■ Case Examples Osteosarcoma A 76-year-old m an presen ted w ith pain in h is left h ip an d left low er ext rem it y w eakn ess, w ith left foot drop. MRI dem on st rated a large t u m or (Fig. 10.1), an d a biopsy of th e lesion w as con sisten t w ith ch on droblast ic osteosarcom a. He w as given t h ree cycles of ch em o th erapy w it h cisp lat in an d d oxor u bicin , bu t p rogressed in h is d isease. A st aged su rgical ap proach w as o ered for en -bloc resect ion of th e m ass. For th e rst , posterior st age of th e op erat ion , L3-S4 w as exp osed. Th e t u m or w as im m ediately eviden t p osterior to th e sacr u m on th e left side; t h u s, t h e m u scu lat u re w as left in t act on th e left side to m ain tain a n egat ive m argin . Th e left side p araspin ou s m u scle an d fascia w as t ran sected at th e L3-L4 level. Pedicle screw s w ere p laced at L3-S1 on th e righ t side an d L3 on th e left , w ith t w o righ t-sided iliac screw s. Righ t-sided h em ilam in ectom ies w ere perform ed at L4-S3, exten ding bilaterally at S4. Th e left sid e L4-S3 n er ve roots w ere iden t i ed, ligated, and cut. Then a m idline osteotom y was m ade from L4 to S4, extending laterally through L4 to gen erate a h em iver tebrectom y at L4 an d L5 (Fig. 10.2). Du e to in t raoperat ive blood loss, th e p at ien t requ ired p latelets an d pressors an d rem ain ed in th e in ten sive care u n it for several days p rior to th e secon d stage of h is su rger y. He d evelop ed exten sive bilateral d eep vein th rom boses (DVTs) requ ir ing an in fer ior ven a cava (IVC) lter placem en t . For t h e secon d st age of t h e p roced u re, a rad ical left h em ip elvectom y w as com p leted by t h e or t h op ed ic su rger y team , follow ed by com plex w ou n d closu re by th e plast ic su rger y team . Th e secon d stage of th e procedure w as com p licated by a left com m on iliac vein inju r y w ith direct repair an d ligat ion of th e left in tern al iliac ar ter y. Th e p last ic su rger y team assisted in com p lex w ou n d closu re. Th e su rger y w as furth er com plicated by a suspected in t raoperat ive air em bolism . Th e pat ien t’s postop erat ive cou rse w as com p licated by bilateral su bdu ral h em atom as, d eep w ou n d in fect ion requ iring irrigat ion an d debridem en t , Entero-

coccus sepsis, p ercu tan eou s en doscop ic gast rostom y (PEG) t u be placem en t , an d respirator y failure requ iring t rach eostom y. After a lengthy h osp it al st ay, h e w as d isch arged to inp at ien t reh abilit at ion . On e year after su rger y, h e is doing w ell w ith n o eviden ce of recu rren t or m et ast at ic disease an d com plain t s of m in im al pain , w ith st rong left an d righ t qu adriceps fu n ct ion bu t loss of left low er ext rem it y fu n ct ion below th e kn ee.

Ew ing’s Sarcoma A 26-year-old w om an h as a h istor y of Ew ing’s sarcom a in th e th oracic spin e an d sacrum . Her disease rem ain ed st able w ith radiat ion th erapy and m ultiple cycles of chem otherapy for 2 years; h ow ever, in t im e sh e developed w orsen ing left low er ext rem it y p ain , n u m bn ess, an d left foot drop . Im aging dem on st rated grow th of th e sacral t um or, bu t a par t ial respon se of h er th oracic lesion to m edical m an agem en t (Fig. 10.4). Surger y w as o ered . A posterior p rocedu re w as perform ed for en -bloc resect ion of th e t u m or. A T-sh ap ed in cision w as m ade from L2 to th e coccyx, exten ding from th e m idlin e to th e left p osterior iliac crest. Periosteal dissection w as perform ed from L3 to S5. L4-L5 lam in ectom ies an d left-sided S1-S5 lam in ectom ies w ere p erform ed, as w ell as a left-sided L4-L5 facetectom y. The left-sided L4-S5 n er ve roots w ere iden t i ed, ligated, an d resected. A h igh -speed drill w as u sed to create an osteotom y th rough th e m idlin e of th e sacrum , exten ding to th e L5-S1 disk space. Th e L5-S1 disk w as excised an d th e left-side lateral sacroiliac join t w as det ach ed from th e ilium u sing a h igh -sp eed drill. Th e left h em isacrum an d t um or w ere ret racted dorsally, w ith carefu l an terior dissect ion of th e iliac vessels an d th e rect u m aw ay from th e t um or m ass. Th e left side of th e pelvis w as th en recon st r u cted u sing a fem u r sh aft , so t h at an ap p ropriate fem u r allograft w as cu t to ll in th e defect , exten ding from th e m edial aspect of th e rem ain ing ilium to th e S1 ver tebral body, su periorly m aking con t act w ith th e L5 en d p late. Th e fem u r sh aft w as secu red in p lace u sin g recon st r uct ion screw s th rough th e graft in to th e S1 ver tebral body an d th e rem ain ing iliu m .

Osteogenic Sarcom a and Ewing’s Sarcom a of the Spine Th en th e p edicle screw s w ere in ser ted at th e L3, L4, an d L5 levels bilaterally. Tw o p elvic screw s w ere in serted on th e left side an d t w o on th e righ t side. Th ese screw s w ere th en con n ected to th e lum bar screw s w ith several rod at tachm ents an d m ultiple close con nectors. The plast ic su rger y ser vice w as called to assist in closu re. Th e patient’s postoperative course w as com plicated by m u lt ip le w ou n d w ash ou ts an d revision s for h er w ou n d closu re, as w ell as a cerebrospinal uid leak requiring a lum bar drain and, later, a ventriculoperitoneal shunt. She also exp erien ced ch ron ic u rin ar y t ract in fect ion s an d urin ar y reten t ion requiring a suprapubic t u be. Six m on th s after h er su rgical p rocedu re, sh e ret u rn ed to h er n at ive cou n t r y an d w as lost to follow -u p.

■ Chapter Summary Osteogen ic sarcom a an d Ew ing’s sarcom a of th e spin e are rare. As in m ost sp in al prim ar y t u m ors, p resen t ing sym ptom s are often n on speci c and m ay delay diagnosis and treatm ent. Curren tly, pat ien t s w ith osteogen ic sarcom a an d Ew ing’s sarcom a are t reated w ith m ult ip le m odalit ies, in clu ding su rger y, radiat ion , an d ch em oth erapy. For both t u m ors, n eoadjuvan t ch em oth erapy follow ed by a w id e su rgical excision is recom m en ded for pat ien ts w h o h ave localized d isease an d resect able t u m ors. Th e progn osis for pat ien ts p resen t ing w ith m etast at ic disease is grim , an d cu rren t t reat m en t s are gen erally lim ited to palliat ive care.

Pearls Osteogenic sarcom a and Ewing’s sarcom a of the spine are high-grade tumors, with the goals of managem ent being to control localized disease and prevent metastatic spread. Osteogenic sarcom a is the m ost comm on sarcoma of the spine and makes up 3 to 15% of all primary spinal tum ors, with the majorit y occurring in the fourth decade of life. Neoadjuvant chem otherapy is strongly recom mended in the m anagem ent of both osteogenic sarcom a and Ewing’s sarcom a due to signi cant improvements in local control and long-term survival. Pitfalls As in other prim ary spinal tumors, symptoms are often vague and ill-de ned, which can lead to a delay in diagnosis, with a signi cant num ber of patients presenting with gross metastatic disease. Ewing’s sarcoma is the m ost comm on m alignant vertebral tum or found in children, with a peak incidence in the second decade. En-bloc resection o ers the best chance for prolonged survival in patients with osteogenic sarcoma and to a lesser extent in patients with Ewing’s sarcoma; however, there is no de nitive cure for either disease. The m orbidit y and mortalit y of en-bloc procedures are considerable, and these procedures should be performed only at experienced centers with multidisciplinary teams. Avoid the following: ◦ Intralesional resection or excisional biopsies for tum or resection ◦ Surgical resection without neoadjuvant chemotherapy ◦ Forget ting to restage Ewing’s sarcoma after neoadjuvant chemotherapy. ◦ Proceeding with surgical resection without prior staging.

Refere nces Five Must-Read Refe rences 1. Gebh ardt MC, Spring eld D, Ne JR. Sarcom as of bon e. In : Abelo MD, ed. Abelo ’s Clin ical On cology. Ph iladelphia: Churchill Livingstone, 2008:1945–2008 2. Wang VY, Pot t s M, Ch ou D. Sarcom a an d th e spinal colum n. Neurosurg Clin N Am 2008;19:71–80 PubMed 3. Kim HJ, McLaw h orn AS, Goldstein MJ, Bolan d PJ. Malign an t osseou s t u m ors of th e p ed iat ric sp in e. J Am Acad Orth op Su rg 2012;20:646–656 PubMed 4. Su ndaresan N, Rosen G, Borian i S. Prim ar y m align an t t um ors of th e spin e. Or th op Clin Nor th Am 2009; 40:21–36, v Pu bMed

5. Clarke MJ, Dasen brock H, Bydon A, et al. Posterioron ly ap p roach for en bloc sacrectom y: clin ical ou t com es in 36 con secu t ive p at ien t s. Neu rosu rger y 2012;71:357–364, discu ssion 364 PubMed 6. Sciubba DM, Okun o SH, Dekutoski MB, Gokaslan ZL. Ew ing an d osteogen ic sarcom a: eviden ce for m ult idiscip lin ar y m an agem en t . Sp in e 2009;34(22, Su pp l): S58–S68 PubMed 7. Ozaki T, Flege S, Liljen qvist U, et al. Osteosarcom a of the spine: experience of the Cooperative Osteosarcom a St u dy Grou p . Can cer 2002;94:1069–1077 Pu bMed

121

122

Chapter 10 8. Kelley SP, Ash ford RU, Rao AS, Dickson RA. Prim ar y bon e t um ours of th e spin e: a 42-year sur vey from th e Leeds Region al Bon e Tum our Regist r y. Eur Spin e J 2007;16:405–409 Pu bMed 9. Rop p er AE, Cah ill KS, Han n a JW, McCar t hy EF, Gokaslan ZL, Ch i JH. Pr im ar y ver tebral t u m ors: a review of epidem iologic, h istological an d im aging n dings, part II: locally aggressive an d m align an t t um ors. Neurosurger y 2012;70:211–219, discussion 219 Pu bMed 10. Mukh erjee D, Ch aich ana KL, Gokaslan ZL, Aaron son O, Ch eng JS, McGirt MJ. Sur vival of pat ient s w ith m align an t prim ar y osseous spinal n eoplasm s: resu lt s from th e Su r veillance, Epidem iology, an d End Resu lt s (SEER) dat abase from 1973 to 2003. J Neu rosurg Spin e 2011;14:143–150 Pu bMed 11. Un n i KK, Inw ard s CY. Dah lin’s Bon e Tu m ors: Gen eral Aspect s an d Dat a on 10,165 Cases. Ph iladelp h ia: Lip pin cot t William s & Wilkin s, 2009 12. Hayden JB, Hoang BH. Osteosarcom a: basic scien ce and clin ical im plications. Orthop Clin North Am 2006; 37:1–7 Pu bMed 13. Sh ives TC, Dah lin DC, Sim FH, Pritch ard DJ, Earle JD. Osteosarcom a of the spin e. J Bon e Joint Surg Am 1986;68:660–668 PubMed 14. Rodallec MH, Feydy A, Larousserie F, et al. Diagn ost ic im aging of solit ar y t um ors of th e spin e: w h at to do an d say. Radiograph ics 2008;28:1019–1041 Pu bMed 15. Sciubba DM, Wasserm an BA, Gokaslan ZL. Tum ors of th e sp in e. In : Kh an n a AJ, ed . MRI for Or th op aedic Su rgeon s. New York: Th iem e, 2010:316–337 16. Meyer JS, Nadel HR, Marin a N, et al. Im aging gu idelin es for children w ith Ew ing sarcom a an d osteosarcom a: a repor t from the Ch ildren’s Oncology Group Bon e Tum or Com m it tee. Pediat r Blood Can cer 2008; 51:163–170 PubMed 17. Klein MJ, Siegal GP. Osteosarcom a: an atom ic and h istologic varian t s. Am J Clin Path ol 2006;125:555–581 PubMed 18. Un n i KK, Inw ards CY. Tu m ors of th e osteoart icular system . In : Fletch er CM, ed. Diagn ost ic Histopath ology of Tum ors. New York: Churchill Livingstone, 2007: 1527–1592 19. McGirt MJ, Gokaslan ZL, Chaich an a KL. Preoperat ive grading scale to predict sur vival in pat ien t s un dergoing resect ion of m align an t prim ar y osseous spin al neoplasm s. Spin e J 2011;11:190–196 Pu bMed 20. Lin k MP, Goorin AM, Miser AW, et al. The e ect of adjuvan t ch em oth erapy on relapse-free su r vival in pat ient s w ith osteosarcom a of th e ext rem it y. N Engl J Med 1986;314:1600–1606 PubMed 21. Gh erlin zon i F, Picci P, Bacci G, Cam pan acci D. Lim b sparing versus am put at ion in osteosarcom a. Correlat ion bet w een local con t rol, su rgical m argin s an d tum or n ecrosis: Istit uto Rizzoli experien ce. Ann Oncol 1992;3(Su p p l 2):S23–S27 PubMed

22. Su ndaresan N, Rosen G, Huvos AG, Krol G. Com bin ed t reat m en t of osteosarcom a of th e spin e. Neurosurger y 1988;23:714–719 Pu bMed 23. DeLan ey TF, Liebsch NJ, Pedlow FX, et al. Ph ase II study of high-dose photon/proton radiotherapy in the m an agem en t of spin e sarcom as. Int J Radiat On col Biol Phys 2009;74:732–739 Pu bMed 24. Mu kh erjee D, Ch aich an a KL, Parker SL, Gokaslan ZL, McGir t MJ. Associat ion of su rgical resect ion an d su rvival in p at ien t s w it h m align an t p r im ar y osseou s sp in al n eop lasm s from t h e Su r veillan ce, Ep id em iology, an d En d Result s (SEER) Dat abase. Eur Spin e J 2013;22:1375–1382 PubMed 25. Lu dw ig JA. Ew ing sarcom a: h istorical perspect ives, curren t st ate-of-th e-ar t , an d oppor t un it ies for t argeted th erapy in th e fut ure. Cu rr Opin On col 2008; 20:412–418 PubMed 26. Arn dt CA, Crist W M. Com m on m uscu loskelet al t um ors of ch ildh ood an d adolescen ce. N Engl J Med 1999;341:342–352 PubMed 27. Ch i JH, Byd on A, Hsieh P, W it h am T, Wolin sky JP, Gokaslan ZL. Epidem iology an d dem ograph ics for prim ar y ver tebral t u m ors. Neu rosu rg Clin N Am 2008;19:1–4 Pu bMed 28. Ilaslan H, Sun daram M, Un n i KK, Dekutoski MB. Prim ar y Ew ing’s sarcom a of th e ver tebral colum n . Skelet al Radiol 2004;33:506–513 PubMed 29. Bern stein M, Kovar H, Paulussen M, et al. Ew ing’s sarcom a fam ily of t um ors: curren t m an agem en t . On cologist 2006;11:503–519 Pu bMed 30. Mukh erjee D, Ch aich an a KL, Adogw a O, et al. Associat ion of exten t of local t u m or invasion an d su r vival in pat ien t s w ith m align an t prim ar y osseous spin al n eoplasm s from th e sur veillan ce, epidem iology, an d end result s (SEER) dat abase. World Neurosurg 2011; 76:580–585 PubMed 31. Cot terill SJ, Ahren s S, Paulussen M, et al. Progn ost ic factors in Ew ing’s t um or of bon e: an alysis of 975 pat ien t s from th e Eu ropean In tergroup Cooperat ive Ew ing’s Sarcom a St udy Group. J Clin On col 2000;18: 3108–3114 PubMed 32. Evan s RG, Nesbit ME, Geh an EA, et al. Mu lt im odal th erapy for th e m an agem ent of localized Ew ing’s sarcom a of pelvic an d sacral bones: a repor t from th e secon d in tergroup st udy. J Clin On col 1991;9:1173– 1180 PubMed 33. Grier HE, Krailo MD, Tarbell NJ, et al. Add it ion of ifosfam id e an d etop osid e to st an dard ch em ot h erapy for Ew in g’s sarcom a an d p r im it ive n eu roectod erm al t u m or of bon e. N Engl J Med 2003;348:694–701 PubMed 34. Nesbit ME Jr, Geh an EA, Burgert EO Jr, et al. Mult im odal th erapy for th e m an agem en t of prim ar y, non m et ast at ic Ew ing’s sarcom a of bon e: a long-term follow -up of th e First In tergroup st u dy. J Clin On col 1990;8:1664–1674 PubMed

Osteogenic Sarcom a and Ewing’s Sarcom a of the Spine 35. Sluga M, W in d h ager R, Lang S, et al. A long-term review of th e t reat m en t of pat ient s w ith Ew ing’s sarcom a in on e in st it u t ion . Eur J Su rg On col 2001;27: 569–573 Pu bMed 36. Paulin o AC, Nguyen TX, Mai W Y. An an alysis of prim ar y site cont rol an d late e ect s according to local con t rol m odalit y in n on -m et ast at ic Ew ing sarcom a. Pediat r Blood Can cer 2007;48:423–429 Pu bMed 37. Marco RA, Gen t r y JB, Rh in es LD, et al. Ew ing’s sarcom a of th e m obile spin e. Spin e 2005;30:769–773 PubMed 38. Bacci G, Longh i A, Briccoli A, Ber ton i F, Versari M, Picci P. Th e role of su rgical m argin s in t reat m en t of

Ew in g’s sarcom a fam ily t u m ors: exp er ien ce of a sin gle in st it u t ion w ith 512 p at ien t s t reated w ith ad juvan t and n eoadjuvan t ch em otherapy. In t J Radiat On col Biol Phys 2006;65:766–772 Pu bMed 39. Talac R, Yaszem ski MJ, Currier BL, et al. Relat ion sh ip bet w een su rgical m argin s an d local recu rren ce in sarcom as of th e spin e. Clin Or th op Relat Res 2002; 397:127–132 PubMed 40. Schw ab J, Gasbarrin i A, Ban diera S, et al. Osteosarcom a of th e m obile spin e. Spine 2012;37:E381– E386 PubMed

123

11 Margins in Spine Tumor Resection: How Much Is Enough? Is Planned Transgression Okay? Ilya Laufer, Mari L. Groves, and Jean-Paul Wolinsky

■ Introduction Cure, or long-term local con t rol, rep resen t s th e prim ar y goal in th e t reat m en t of prim ar y t u m ors. Variou s st aging system s w ere developed in order to facilit ate p rogn ost icat ion of th e probabilit y of cure an d sur vival an d to h elp direct t reat m en t d ecision s. Physician s m ost com m on ly u se th e En n eking st aging system w h en discu ssing p rim ar y m u scu loskelet al t u m ors.1 Because prim ary extradural tum ors of the spine rep resen t a su bset of m uscu loskelet al t u m ors, th e En n eking system an d p rin ciples th at w ere in it ially d evelop ed in t h e con text of ap p en d icu lar t u m ors h ave been ap p lied to sp in al t u m ors. Alth ough sp in al an d ap p en d icu lar t u m ors m ay be h istologically id en t ical, t h e an atom ic ch allenges en cou n tered in th e sp in e d ist ingu ish th e sp in al t u m ors from th eir ap p en dicu lar coun terpar ts. Th e Wein stein -Borian i-Biagin i (W BB) surgical system and the Tom ita surgical classi cation of vertebral t um ors w ere developed in order to facilit ate th e ap plicat ion of th e En n eking con cept s to th e spin e an d to describe th e sp eci c ver tebral elem en ts th at h arbor th e t u m or an d t h e exten t of in t ra- an d ext raosseou s exten sion .2,3 Based on th e circu m feren t ial exten t of th e t u m or an d th e am ou n t of p araspin al an d epidural extension, the system s help predict th e m argin s th at m ay be ach ieved during su rger y. Th e su rrou n ding st r u ct u res, t issu e plan es, an d associated n eu ral elem en ts ch ange along th e

cou rse of t h e sp in al colu m n ; t h erefore, t h e safet y an d feasibilit y of at tem pt ing a resect ion of a clear surroun ding m argin or en -bloc excision also m u st be con sid ered in t h e con text of th e sp eci c t u m or level. Th u s, alth ough am pu t at ion of a large p or t ion of th e sacru m w ith su rrou n ding m u scu loskeletal st r u ct u res an d sacral n er ve roots m ay result in m in im al st ruct u ral an d n eu rologic com prom ise, h igh er am pu t at ion w ou ld gen erally be associated w ith profound neurologic de cits and require extensive recon st ru ct ion . How m u ch m argin sh ou ld th ere be, an d is a m argin n ecessar y, are frequen t qu est ion s th at arise w ith regard to spin al t um ors. To an sw er th is, an u n derst an ding of th e de n it ion of a m argin is critical. Surgical m argins in spinal surger y are de n ed by conven t ion s adopted from th e or th opedic ext rem it y literat u re. Bu t th e term s in use are frequen tly m isused, w h ich h as con t am in ated th e spin al t u m or su rgical literat u re. In addit ion to u n derstan ding w h at th e de n it ion of a m argin is, correlat ing th e t yp e of m argin th at is required w ith th e path ology being t reated is crit ical to delivering th e appropriate t reat m en t to a speci c pat ien t .

■ Margin Def nition Margin s can be de n ed as p osit ive, m argin al, w ide, and radical.1 A positive m argin m eans that

Margins in Spine Tumor Resection residu al t u m or rem ain s w ith in th e p at ien t . A m argin al m argin in dicates t h at t u m or com es u p to th e m argin of th e specim en , bu t it is con tain ed w ith in th e capsule of th e t u m or. A w ide m argin in dicates th at a t u m or h as been excised an d at least 1 cm of h ealthy t issue h as been excised su rrou n ding th e t um or. A radical resect ion m ean s th at th e t u m or is rem oved an d th e en t ire com p ar t m en t w ith in w h ich th e t u m or resides is also rem oved . Th e m argin gives n o in form at ion on h ow a su rgical resect ion w as undertaken; it only describes the resultant surgical bed. Often , in spin al oncology, unfort unately, on ly m argin al m argin s are su rgically feasible. Resect ion of sacral ch ord om as, even w ith n er ve sacri ce, an d resect ion of soft t issues in cluding skin , fat , an d m uscle, an d plan n ing an osteotom y rost ral to th e exten t of t um or usually st ill results in a m arginal m argin. It is rare that sacral ch ordom as do n ot h ave ven t ral exten sion of th e t u m or in to th e p elvis beh in d th e rect u m . In m ost cases, th e t u m or cap su le is p reser ved at th is m argin , bu t th e resect ion m argin is at th e m esorect um . With ou t resect ion of th e rect u m , sacral ch ordom a resect ion s are m argin al at th e ven t ral m argin or th e m esorect u m (Fig. 11.1).

Pat ien t s w it h m obile sp in e osseou s t u m ors t yp ically presen t w ith epidu ral exten sion . In th is sit u at ion , w ith ou t du ral resect ion , th e best m argin w ill be a m argin al m argin at th e t u m or cap sule along th e dura (Fig. 11.2). Wide m argin resect ion s are possible w ith spin al t um ors in rare sit uat ion s. Som e pat ien ts m ay presen t w ith sm all t u m ors com p letely con t ain ed w ith in th e ver tebral body (Fig. 11.3), an d, in th ese sit uat ion s, com plete resect ion of th e ver tebral body w it h t h e t u m or can resu lt in a w ide m argin . Radical resect ion s of th e sp in e are tech n ically p ossible bu t are n ot perform ed. A radical resect ion requires resect ion of th e t u m or an d the entire com partm ent w ithin w hich the tum or lies. In an ext rem it y t u m or, th is w ou ld involve resect ion of th e t u m or, th e bon e from w h ich it arises, an d th e m uscles in ser t ing in to th e bon e th rough th eir d ist al in ser t ion s. An exam ple w ou ld be a t u m or arising in th e t ibia, w ith a plan n ed resect ion being an above-th e-kn ee am putat ion , above th e m uscular in ser t ion s of the m uscles surrounding the tum or. In the spine, a con ceivable exam ple w ould be a sacral t u m or, contained w ithin the sacrum , w ith no vent ral extension through the sacrum , and no extension

b

a

c

Fig. 11.1a–d A large sacral chordoma with ventral extraosseous extension. (a) The postoperative surgical eld demonstrates the proximit y of the tumor to the intrapelvic organs. (b) Posterior view of the specim en demonstrates the wide m argin of

d

healthy tissue, which can be obtained. On the other hand, the anterior (c) and lateral (d) views of the specimen show a marginal margin at the site of ventral extraosseous extension of the tumor. (Courtesy of Jean-Paul Wolinsky, MD.)

125

126

Chapter 11 a

b

c

Fig. 11.2a–c (a) Sagit tal T2-weighted magnetic resonance imaging (MRI) of a T10 chordom a with epidural extension. Superior (b) and posterior (c) views of the vertebrectomy specimen demonstrate

that only a marginal margin was obtained due to the dorsal extension of the tumor into the epidural space, with tum or displacing the posterior longitudinal ligament. (Courtesy of Jean-Paul Wolinsky, MD.)

Fig. 11.3 A malignant peripheral nerve sheath tumor entirely con ned to the vertebral body without extraosseous extension. This tumor was resected with a wide margin of bone. (Courtesy of Jean-Paul Wolinsky, MD.)

of t h e t u m or to t h e d u ra, w h ere a h em icorp orectom y is perform ed. Th is operat ion w ou ld m ost likely n ever be p rop osed for su ch t u m or arch itect ure, irresp ect ive of th e p ath ology, as th e m orbidit y of th e op erat ion w ou ld be u n accept able. With exten sion of th e t um or to th e du ra, a radical resect ion , even w ith a h em icorporectom y w ou ld n ot be p ossible, as resect ion of th e du ral “com par t m en t”, w ould requ ire resect ion of th e du ra th rough ou t th e en t ire sp in e an d cran ium . Th e Spin e On cology St udy Grou p evaluated the reliabilit y of th e application of the Enneking an d W BB st aging system s in th e spin e using a series of prim ar y t u m or case exam ples.4 Moderate interobserver reliabilit y was dem onstrated in th e u se of th e En n eking st aging system , an En n eking-recom m en ded su rgical m argin , th e W BB layers, an d th e W BB-recom m en d ed resect ion . Th e in t raobser ver reliabilit y w as n earperfect for th e En n eking staging system an d su bstan t ial for th e W BB system . Th e En n eking tum or extent determ ination proved particularly ch allenging, w ith on ly sligh t to fair in terob ser ver reliabilit y. A m u lt icen ter coh or t st u dy exam in ed t h e cor relat ion bet w een t h e abilit y to ach ieve an En n eking-recom m en ded m argin an d th e ou tcom e of su rger y for p r im ar y osseou s t u m ors of th e sp in e.5 Pat ien ts h arboring t u m ors th at w ere am en able to resect ion w ith th e m argin th at is recom m en ded by th e En n eking staging system h ad sign i can tly low er local recu rren ce

Margins in Spine Tumor Resection rates com pared w ith pat ien t s in w h om an op t im al m argin cou ld n ot be ach ieved. How ever, we m ust em phasize that the resectabilit y of the t u m or w it h an app ropriate m argin is largely an in t rin sic ch aracterist ic associated w ith th e in dividu al h istology, grade, locat ion , an d size. Th is is su p por ted by th e obser vat ion t h at in th e st udy coh ort, th e Enneking-recom m ended m argin cou ld n ot be ach ieved in a sign i can t m ajorit y of h igh -grade t um ors an d t um ors located in th e cer vical an d th oracic sp in e, bu t cou ld be ach ieved in a sign i can t m ajor it y of ben ign t u m ors an d t u m ors located in th e sacru m . Fu rth erm ore, th e risk of blood loss of m ore th an 5,000 m L an d th e risk of in fect ion w ere sign i can tly h igh er in pat ien t s w h o u n der w en t su rger y for En n eking-recom m en ded m argin s.

■ Surgical Technique

Terminology Furth er com plicat ing th e literat ure is th e m isu se of th e term s spondylectom y, en-bloc spondylectom y, total en-bloc spondylectom y (TES), int ralesional, an d en-bloc resect ion. Spondylectom y m ean s rem oving th e en t ire segm en t of th e spin e (th e vertebral body, pedicles, su perior an d in ferior ar t icu lat ing processes, pars, t ran sverse p rocesses, lam in a, an d spin ou s p rocess) (Fig. 11.4). It is a surgical tech n ique, but does n ot specify if th e t u m or is rem oved in an en -bloc or in t ralesion al fash ion . An en t ire spin al segm en t can be rem oved “piecem eal” an d resu lt in a spon dylectom y, bu t th is sh ou ld n ot be con fu sed w ith an en -bloc resect ion . Th e term s en-bloc spondylectom y an d total en-bloc spondylectom y (TES) are frequently used in t h e p r im ar y t u m or literat u re,6 bu t in t h e m obile sp in e t h ey are seld om , if ever, p erform ed. An en -bloc spon dylectom y w ould requ ire resect ion of th e sp in al segm en t w ith ou t violat ing any port ion of it . Becau se each sp in al segm en t of th e m obile sp in e con tain s th e sp in al cord or th e cau da equin a, rem oval of th e ver tebra w ith ou t open ing of th e sp in al can al ring w ou ld requ ire t ran sect ion of th ese n eu ral elem en ts w ith th e sp ecim en . Th is does occu r in t h e sacr u m , w h en sacral am p u t at ion s are

Fig. 11.4 Spondylectomy specimen. (Courtesy of Jean-Paul Wolinsky, MD.)

perform ed, as th e sacri ce of n er ves roots in th is region is u su ally u n der t aken to ach ieve a m ean ingful resect ion .7 Th e ter m int ralesional resect ion sh ou ld be u sed anyt im e t h e t u m or cap su le is violated an d t h e t u m or is en tered . Even if t h e in ten t of th e t u m or resect ion is to rem ove th e en t ire sp ecim en in on e p iece, if th e sp ecim en d em on st rates th at th e m argin w as violated on path ological review , t h en t h is is an in t ralesion al resect ion . If a su rger y p lan in cor p orates t h e n eed for a p lan n ed t ran sgression of th e t u m or 8 (Fig. 11.5) in order to ach ieve a resect ion , th is is, by de n it ion , an in t ralesion al resect ion desp ite com p lete rem oval of th e t u m or. Th e term en bloc is u sed if th e t um or is rem oved in on e p iece w ith ou t violat ion of th e t u m or m argin . Cer tain t u m or path ologies requ ire en -bloc resect ion s w ith n egat ive m argin s to ach ieve local t um or con t rol. In cert ain circum stances, an en-bloc resection m ight be used to con t rol blood loss in vascu lar t u m ors even if th e t um or p ath ology does n ot requ ire an en bloc resect ion for t um or con t rol. Likew ise, an

127

128

Chapter 11

a

b

c

Fig. 11.5a–c (a) A computed tomography (CT) scan of a T10 solitary breast metastasis with invasion of the vertebral body, epidural space, pedicle, and the costovertebral joint. Horizontal (b) and lateral (c) radiographs of the specimen showing the

intralesional resection of the tumor with intentional transgression of the pedicle with tumor invasion. 10 (Reproduced with permission from Lippincot t Williams & Wilkins.)

en -bloc resect ion m ay be em ployed to ach ieve a com p lete resect ion if t h e t u m or an d su rroun ding an atom y h ave becom e distor ted, an d th e in terp ret at ion of t u m or m argin s w ou ld be com p licated by an in t ralesion al resect ion .

de n it ive diagn osis, w h ich u su ally requ ires a biopsy. Th e m argin th at is n eeded to con t rol th e t u m or is also dict ated by th e t u m or path ology. Many t im es, th e plan n ed surgical m argin th at is requ ired h inges on t h e available adju van t th erapies for a given p ath ology. Fu r th erm ore, alth ough en -bloc excision m ay be th e desired m eth od of t u m or rem oval for cer tain t u m ors, som e eviden ce suggest s th at an opt im al n al m argin m ay be th e m ore m ean ingfu l goal an d p redictor of local con t rol.9 Cer t ain t um or p ath ologies h ave a ver y good resp on se to adjuvan t th erapies su ch as ch em ot h erapy an d rad iat ion t h erapy, an d t h erefore

■ Treatment Options Th e t yp e of su rger y in dicated, en bloc versu s in t ralesion al, is determ in ed by th e p ath ology being t reated (Table 11.1). Th u s, p rior to doing th e resect ion , it is m ost im p or t an t to obt ain a

Table 11.1 Histology-Specif c Recommended Minimal Margin Pathology (Primary Bone Tumors)

Minimal Margin Needed

Plasmacytoma/m ultiple myeloma Eosinophilic granuloma/Langerhans cell histiocytosis Ewing’s sarcoma

Biopsy Biopsy

Chordoma Osteogenic sarcoma Chondrosarcoma (m obile spine and pelvis) Chondrosarcoma (skull base) Giant cell tumor Aneurysmal bone cyst Osteoid osteomas Osteoblastoma Hemangioma

Biopsy, en-bloc resection in rare instance of residual tumor after chemotherapy/radiation En-bloc resection En-bloc resection En-bloc resection Intralesional resection En-bloc resection Intratumoral chem otherapy vs intralesional resection Intralesional resection En-bloc resection if path unclear Intralesional resection

Margins in Spine Tumor Resection th e su rgical role m ay be m in or in th ose in st an ces. Exam p les of su ch path ologies in clu de plasm acytom a, m ultiple m yelom a, eosinophilic gran u lom a (Lan gerh an s h ist iocytosis), an d Ew ing’s sarcom a. Advan ces in ch em oth erapy h ave ch an ged t h e ap p roach to Ew in g’s sarcom a from a su rgical lesion to on e th at is n ow t reated p rim arily w ith ch em oth erapy an d radiat ion th erapy. On e im plicat ion of th is n ew ap proach occu rs in rare sit u at ion s w h ere th ere is on ly a solitar y residual focu s of t u m or; in th ese sit u at ion s, a n egat ive m argin resect ion is advocated by som e au th ors to t r y to eradicate any gross d isease in Ew ing’s sarcom a.10 Th ere are cer t ain t u m ors w h ere an en -bloc resect ion w ith a n egat ive m argin is crit ical to achieving long-term con t rol of th e disease. Th e textbook/quin tessen t ial represen tat ion of such a t u m or is a ch ord om a. It is clear from t h e su rgical literat u re th at piecem eal resect ion of ch ordom as h as a h igh propen sit y for local recu rren ce.11 In addit ion , it is also clear th at if an en -bloc resect ion is plan n ed, bu t th e n al p at h ological m argin is con t am in ated , t h ese p at ien t s h ave a sign i can tly w orse progn osis an d h igh er likelih ood of recu rren ce. Th ere are p robably t w o reason s for th is n ding: (1) ch ord om as h ave a h igh p ropen sit y to seed th e su rrou n d in g t issu e if sp illed , an d (2) adju van t ch em oth erapy an d radiat ion th erapy h ave ver y lim ited e ect iven ess in th e t reat m en t of ch ord om as. It is un clear h ow w ide a n egat ive m argin n eeds to be for resect ion of ch ordom as, as n o st u dies h ave suggested an an sw er. In ou r experien ce in th e sacru m , local recu rren ces n ever occu r at th e m argin of th e rost ral or lateral osteotom ies. Th ey do n ot occur at th e ven t ral m argin at th e rect um , w h ich is frequen tly th e m ost m argin al m argin . If th ere is a local recurren ce, it is in th e m u scle, u su ally along th e piriform is m u scle. Of n ote, th is is u sually th e w idest m argin th at is possible, an d it is un clear w hy th e t um or w ould recur in th is area. On e possibilit y is th at u n detected m icroscopic t u m or exists in th ese m u scles already, an d th ese cells are already beyon d th e su rgical m argin . In an at tem pt to decrease th e ch an ce of recurrence in this region, w ider m argin s have been t aken by sect ion ing th ese m u scles at th eir in sert ion s in th e greater t roch an ter (Fig. 11.6).

Fig. 11.6 A sacrectomy specimen of a chordoma with resection of piriformis muscles. The muscles are sectioned at the at tachment of the piriformis to the greater trochanter. (Courtesy of Jean-Paul Wolinsky, MD.)

Altern at ively, som e cen ters are delivering preoperat ive radiat ion to th ese region s in an attem pt to reduce the chance of recurrence in this region .12 Bu t too few pat ien ts h ave u n dergon e th is m ore exten ded resect ion or p reop erat ive radiat ion t reat m en ts to kn ow if th is app roach h as a p osit ive im pact . Osteogen ic sarcom a is an oth er t u m or th at ten ds to beh ave in a m align an t an d locally aggressive fash ion , bu t it is u n clear if th e t u m or act ually seeds th e local environ m en t . It is clear th at if th e m argin s for resect ion are posit ive, th en th e overall p rogn osis for th e pat ien t is poor. In gen eral th e progn osis for osteogen ic sarcom a, even w ith n egat ive m argin s, is also poor.13 En-bloc resection for osteogenic sarcom a m ay h ave im p act on local con t rol bu t n ot on su r vival. Th is m ost likely is n ot a con sequ en ce of th e surgical m argin , but h as m ore to do w ith th e beh avior of th e t u m or, esp ecially it s aggressiven ess an d it s propen sit y for m et ast at ic spread. With out a good respon se to ch em oth erapy, pat ien ts w ith osteogen ic sarcom a do poorly.14 Neoadjuvan t ch em oth erapy is u su ally em ployed w ith th e t reat m en t of th ese t um ors, an d th e percen t n ecrosis of th e t um or w ith in a surgical specim en has overall predictive value.15 An in abilit y to ach ieve a n egat ive m argin m ay be m ore a m arker th at th e t u m or is ver y aggressive an d th at m icro-m etastat ic d isease h as already occu rred.

129

130

Chapter 11 Ch on drosarcom a ap pears to beh ave di eren tly in di eren t p ar t s of th e body. Histologically, ch on drosarcom as of th e sku ll base, spin e, an d pelvis m ay appear iden t ical, but th ey do n ot beh ave in th e sam e m an n er.16 Skull base ch on drosarcom as, given th eir location , can n ot be resected in an en -bloc fash ion w ith n egat ive m argin s. Th e m orbidit y of an en -bloc resect ion p reclu des th is opt ion in th e sku ll base. Th ese t um ors in th e skull base are usually resected in a p iecem eal, su btot al fash ion . Adjuvan t radiat ion th erapy in th e form of proton beam irradiat ion can ach ieve local con t rol rates u p to 98% at 10 years.17 Th is h as n ot been th e exp erien ce in th e m obile sp in e an d pelvis. In th ese locat ion s, su btotal resect ion , even w ith adjuvan t radiat ion th erapy, is associated w ith ver y h igh recu r ren ce rates at 1 year.18,19 It is n ot clear, th ough , if t h ese t u m ors absolu tely n eed an en -bloc resect ion or if t h e t u m or m argin can be violated as long as u lt im ately th e resect ion bed is n egat ive. It is ou r pract ice to resect th ese t u m ors in an en bloc-fash ion , if possible. If th e t u m or arch itect u re preclu des doing so, su ch as a m obile spin e ch on drosarcom a th at circum feren t ially en cases th e spin al cord (Fig. 11.7), th en t h e p lan is to t ran sgress th e t u m or so as to resect it en t irely. Th e con cept of a p lan n ed transgression of the t um or is supported by w hat

a

Fig. 11.7a,b A CT scan (a) and MRI (b) of a clear cell chondrosarcoma with circum ferential encasement of the spinal cord. Excision of this tumor

h as been seen w ith pelvic ch on drosarcom as. Reoperat ion of ch on drosarcom as w ith residu al or locally recurren t t um ors can h ave favorable long-ter m d isease-free su r vival if com p lete resect ion of th e t um or w ith n egat ive m argin s can be ach ieved. Gian t cell t u m ors are t h ough t to be low grad e bu t to h ave som e m align an t p oten t ial. Th ese t u m ors h ave been n oted to m et ast asize to th e lung an d h ave dem on st rated th e poten tial for m align ant transform at ion.20 It is unclear if th is m align an t t ran sform at ion is a result of adjuvan t radiat ion th erapy or is d e n ovo t ran sform at ion of th e t u m or. Gian t cell t u m ors presen t ing in th e ext rem it ies are u sually t reated w ith in t ralesion al resect ion an d p ostcu ret t age t reat m en t of th e surgical bed w ith a caust ic subst an ce su ch as p h en ol.20 With th is t reatm en t , th ere h as been reason able con t rol w ith lim ited recu rren ces. In th e spin e, ou r experien ce w ith in t ralesion al resect ion of gian t cell t u m ors is th at it en tails a recurren ce rate of u p to 90%. Th e reason for th is high er recurren ce rate for spin al com p ared w ith ext rem it y gian t cell t u m ors m ay lie w ith th e fact th at cau st ic su bstan ces su ch as ph en ol can n ot be u sed in th e su rgical bed, especially in th e p resen ce of sen sit ive n eu ral t issu e. Tu m or recu rren ces in th e spin e can be ver y di cu lt to t reat , an d th ey

b

requires transgression of the tumor margin in order to preserve the integrit y of the spinal cord. (Courtesy of Jean-Paul Wolinsky, MD.)

Margins in Spine Tumor Resection can resu lt in t u m or resolu t ion ,24 an d th erefore t u m ors th at resp on d to th is t reatm en t m ay n ot n eed su rgical resect ion . Osteoid osteom as are sm all t um ors w ith out m alignant potential. Surgical resection is aim ed at gross t u m or resect ion for pain con t rol, bu t en -bloc resect ion is n ot n eeded.25 Th e larger var ian t of osteoid osteom as, osteoblastom as, are also ben ign t u m ors, bu t given th eir size, sam pling error can be an issu e, an d care m u st be t aken to d ist in gu ish t h em from osteosarcom as.26 If di eren t iat ion of an osteoblastom a from an osteosarcom a can n ot be m ade p reop erat ively based on im aging an d biopsy, th en aggressive m an agem en t w ith an en -bloc resect ion is indicated (Fig. 11.8). Hem angiom as are ben ign t u m ors w ith out m align ant poten t ial. Th ey ten d to be con tain ed w ith in th e ver tebral body, but in rare cases can grow beyon d th e an terior colu m n an d cause n eu rologic com p rom ise.27 Histologically, th ese at ypical h em angiom as are n o di eren t from other h em angiom as. Treatm ent is aim ed at n eu-

resu lt in sign i can t m orbid it y; t h erefore, w e advocate en -bloc resect ion w it h a m argin al m argin of th ese t u m ors if th e m orbidit y of th e operat ion is acceptable. As our un derstan ding of th ese t um ors is evolving, th is t reat m en t paradigm m ay ch ange. Th ere h as been som e evidence that adm inistration of denosum ab results in gian t cell t u m or grow th arrest an d decrease in size th rough RANK-L (receptor act ivator of n uclear factor kappa-B ligan d) in h ibit ion an d o ers a prom ising m edical therapy in the treatm en t of th ese t u m ors.21 An eur ysm al bon e cyst s (ABCs) are vascu lar low -grade t u m ors. Th ey can occasion ally h arbor oth er t um ors w ith in th em .22 On cologically, t h ese t u m ors can be t reated w it h an in t ralesion al resect ion .23 How ever, su btot al in t ralesion al resect ion s can resu lt in a recu r ren ce. Given th e vascularit y of th ese t um ors, preoperat ive em bolizat ion is often em p loyed an d en -bloc tech n iques m ay be used to decrease operat ive blood loss. Treat m en t w ith in t rat u m oral injection of calcitonin an d corticosteroids

a

b

Fig. 11.8a,b (a) A three-dim ensional CT reconstruction of a giant sacral osteoblastoma. (b) A sagit tal view of the surgical specimen after en-bloc resection of the tumor. (Courtesy of Jean-Paul Wolinsky, MD.)

131

132

Chapter 11 rologic p rotect ion an d t u m or con t rol. To p revent recurrences, com plete resection should be u n dert aken . En -bloc resect ion is n ot n ecessar y, but, as w ith ABCs, en-bloc resection m ay be em ployed to decrease blood loss during resect ion.

■ Chapter Summary Su rger y represen t s th e m ain t reat m en t m odalit y in th e t reat m en t of prim ar y sp in al osseou s t u m ors. Th e h istology of th e t u m or rem ain s th e st rongest determ in an t of su r vival an d local con t rol. Th e su rgical m argin in th e spin e ranges from in t ralesion al to w ide, w ith w ider m argin s in dicated for m ore aggressive t um ors. Th e abilit y to ach ieve a w ide m argin in th e t reat m en t of spin al t um ors is often lim ited by th e con tact of t h e t u m or w it h n eu ral elem en t s, an d t h e goal of w id e m argin m u st be balan ced w it h the functional de cit that m ay result from such surger y. Furtherm ore, the spinal cord often lim its th e surgeon s’ abilit y to ach ieve an en -bloc resect ion , n ecessitat ing t u m or t ran sgression . Su rgical dat a in d icate th at failu re to ach ieve th e in d icated m argin in p rim ar y t u m or resect ion is associated w ith in creased risk of t u m or recu rren ce. Alth ough su rgical tech n iqu e plays an im por t an t role in ach ievem en t of th e desired m argin, tum or biology an d stage often act as th e ult im ate determ in an t of th e abilit y to carr y ou t an en -bloc resect ion or th e in dicated m argin . A clear u n derst an ding of th e desired su rgical m argin an d of th e an atom ic ch allenges of th e spin al colum n en ables surgeon s to devise a su rgical p lan th at m in im izes th e risk of local recu rren ce an d su rgical m orbidit y. The behavior of prim ary spinal tum ors ranges from ben ign to h igh ly m align an t . Alth ough asym ptom at ic ben ign t u m ors m ay be m on itored, sym ptom at ic an d m align an t t u m ors requ ire t reat m en t w ith th e goal of cu re or long-term local control. The spinal colum n presents unique an atom ic ch allenges, an d th e on cological surgical prin ciples u sed in th e t reat m en t of ap pen dicu lar m u scu loskeletal t u m ors h ave been

successfully app lied to spin al t u m ors. Th e propen sit y of m align an t t u m ors to form satellite an d skip lesion s ou t sid e of th e clear rad iograph ic an d h istological t um or boun daries requ ires t u m or excision w ith a w ide m argin of h ealthy t issue surroun ding th e t um or an d th e u se of radiat ion an d ph arm acological t h erapy in order to increase the probabilit y of local cont rol. Gross excision of low -grade an d ben ign t u m ors m ay p rovide cu re w ith ou t th e u se of adjuvan t th erapy. Th e d escript ion of th e su rgical m argin m u st be dist ingu ish ed from th e descript ion of th e su rgical tech n iqu e u sed to rem ove th e t u m or. Th e t u m or m ay be rem oved in a p iecem eal fash ion or en -bloc fash ion ; th e lat ter refers to excision of t h e w h ole sp ecim en w it h ou t vio lat ion of t h e t u m or cap su le. Th e descr ipt ion of th e m argin an d th e t yp e of t u m or excision p rovid e com p lem en t ar y in for m at ion abou t th e m an n er in w h ich th e t um or w as rem oved. Alth ough th e h istology of th e t um or gen erally serves as the prim ary determ inant of prognosis, in m ost in st an ces failu re to obtain a h istologyappropriate t um or m argin h as been sh ow n to be an adverse progn ost ic in dicator associated w ith local recurren ce. Th erefore, ever y e or t m u st be m ade to p erform an opt im al on cological resect ion w h ile balan cing t h e m orbidit y of t r ying to ach ieve th e in dicated m argin .

Pearls Margins: intralesional, marginal, wide, radical Surgical plan: piecem eal, en bloc Surgical technique: spondylectomy Biopsy: critical for determining the margin requirement, the surgical plan, and the surgical technique Pitfalls Failure to understand the de nition of term s for literature descriptors or to underst and tum or outcomes Failure to identify pathology before form ulating the treatment plan

Margins in Spine Tumor Resection Refere nces Five Must-Read Refe rences 1. En n eking W F. A system of st aging m usculoskelet al neoplasm s. Clin Orth op Relat Res 1986;204:9–24 PubMed 2. Borian i S, Wein stein JN, Biagin i R. Prim ar y bon e t um ors of th e spin e. Term in ology an d surgical st aging. Spin e 1997;22:1036–1044 Pu bMed 3. Tom it a K, Kaw ah ara N, Baba H, Tsuchiya H, Nagat a S, Toribat ake Y. Tot al en bloc spon dylectom y for solit ar y spin al m et ast ases. In t Orth op 1994;18:291–298 PubMed 4. Ch an P, Borian i S, Fou rn ey DR, et al. An assessm en t of th e reliabilit y of th e En n eking an d Wein stein -Borian iBiagin i classi cat ion s for st aging of prim ar y spin al t um ors by th e Spin e On cology St udy Group. Spin e 2009;34:384–391 PubMed 5. Fisher CG, Saravanja DD, Dvorak MF, et al. Surgical m an agem en t of prim ar y bon e t u m ors of th e spin e: validation of an approach to enhance cure and reduce local recu rren ce. Sp in e 2011;36:830–836 Pu bMed 6. Tom it a K, Kaw ah ara N, Baba H, Tsuch iya H, Fujit a T, Toribat ake Y. Tot al en bloc spon dylectom y. A n ew su rgical tech n ique for prim ar y m align an t ver tebral t um ors. Spin e 1997;22:324–333 Pu bMed 7. Fou rn ey DR, Rh in es LD, Hen t sch el SJ, et al. En bloc resect ion of prim ar y sacral t um ors: classi cat ion of su rgical approach es an d outcom e. J Neurosurg Spine 2005;3:111–122 PubMed 8. Tom it a K, Kaw ah ara N, Kobayash i T, Yosh ida A, Murakam i H, Akam aru T. Surgical st rategy for spin al m et ast ases. Sp in e 2001;26:298–306 PubMed 9. Virkus W W, Marsh all D, En n eking W F, Scarborough MT. Th e e ect of con t am in ated su rgical m argin s revisited. Clin Orth op Relat Res 2002;397:89–94 PubMed 10. Marco RA, Gen t r y JB, Rh in es LD, et al. Ew ing’s sarcom a of th e m obile spin e. Spin e 2005;30:769–773 PubMed 11. Borian i S, Ban diera S, Biagin i R, et al. Chordom a of th e m obile spin e: ft y years of experien ce. Spin e 2006; 31:493–503 PubMed 12. Wagn er TD, Kobayash i W, Dean S, et al. Com bin at ion sh or t-course preoperat ive irradiat ion , surgical resect ion , an d red u ced - eld h igh -d ose p ostop erat ive irradiat ion in th e t reat m en t of t u m ors involving th e bon e. In t J Radiat On col Biol Phys 2009;73:259–266 PubMed 13. Schoen feld AJ, Horn icek FJ, Pedlow FX, et al. Osteosarcom a of th e spin e: experien ce in 26 pat ien t s t reated at th e Massach uset t s Gen eral Hospit al. Spin e J 2010;10:708–714 Pu bMed 14. Rosen G, Marcove RC, Huvos AG, et al. Prim ar y osteogen ic sarcom a: eigh t-year experien ce w ith adjuvan t ch em ot h erapy. J Can cer Res Clin On col 1983; 106(Su p p l):55–67 PubMed

15. Bacci G, Mercuri M, Longh i A, et al. Grade of ch em oth erapy-in du ced n ecrosis as a p redictor of local an d system ic control in 881 patien ts w ith non-m etastatic osteosarcom a of th e ext rem it ies t reated w ith n eoadjuvan t chem oth erapy in a single in st it u t ion . Eur J Can cer 2005;41:2079–2085 Pu bMed 16. An dreou D, Ruppin S, Feh lberg S, Pin k D, Wern er M, Tu n n PU. Su r vival an d p rogn ost ic factors in ch on drosarcom a: result s in 115 pat ien t s w ith long-term follow -up. Act a Or th op 2011;82:749–755 PubMed 17. Rosen berg AE, Nielsen GP, Keel SB, et al. Ch on d rosarcom a of th e base of th e sku ll: a clin icopath ologic st u dy of 200 cases w ith em p h asis on it s dist in ct ion from ch ordom a. Am J Surg Path ol 1999;23:1370– 1378 PubMed 18. Bergh P, Gun terberg B, Meis-Kin dblom JM, Kin dblom LG. Progn ost ic factors an d outcom e of pelvic, sacral, an d spin al ch on drosarcom as: a cen ter-based st u dy of 69 cases. Cancer 2001;91:1201–1212 PubMed 19. York JE, Berk RH, Fu ller GN, et al. Ch on d rosarcom a of th e spin e: 1954 to 1997. J Neurosurg 1999;90(1, Suppl):73–78 PubMed 20. Erran i C, Ruggieri P, Asen zio MA, et al. Gian t cell t um or of th e ext rem it y: a review of 349 cases from a single in st it u t ion . Can cer Treat Rev 2010;36:1–7 PubMed 21. Th om as D, Hen sh aw R, Skubit z K, et al. Den osum ab in pat ien t s w ith gian t-cell t um our of bon e: an open label, ph ase 2 st udy. Lan cet On col 2010;11:275–280 PubMed 22. Fang Z, Chen M. Ch on droblastom a associated w ith an eu r ysm al cyst of th e n avicu lar bon e: a case rep or t . World J Su rg On col 2013;11:50 Pu bMed 23. Zenon os G, Jam il O, Govern ale LS, Jern igan S, Hedequist D, Proctor MR. Surgical t reat m ent for prim ar y spin al an eu r ysm al bon e cyst s: exp erien ce from Ch ildren’s Hospit al Boston . J Neu rosu rg Ped iat r 2012;9: 305–315 PubMed 24. Oh ash i M, Ito T, Hiran o T, En do N. Percut an eous in t ralesion al inject ion of calciton in an d m ethylprednisolon e for t reat m en t of an an eur ysm al bone cyst at C-2. J Neurosurg Pediat r 2008;2:365–369 PubMed 25. Gasbarrin i A, Cappuccio M, Ban diera S, Am endola L, van Urk P, Borian i S. Osteoid osteom a of th e m obile spin e: surgical outcom es in 81 pat ien t s. Spin e 2011;36:2089–2093 PubMed 26. Borian i S, Am en dola L, Bandiera S, et al. St aging an d t reat m en t of osteoblastom a in th e m obile spin e: a review of 51 cases. Eur Spin e J 2012;21:2003–2010 PubMed 27. Jan kow ski R, Now ak S, Zukiel R, Szym aś J, Sokół B. Su rgical t reat m en t of sym ptom at ic ver tebral h aem angiom as. Neurol Neuroch ir Pol 2011;45:577–582 PubMed

133

12 Principles Behind Determining the Right Approach Nicolas Dea and Row an Schouten

■ Introduction

■ Enneking Classi cation

Prim ar y t u m ors of t h e sp in e are rare. As a resu lt , clin ical exp er t ise h as been con cen t rated in selected quatern ar y referral cen ters. In recen t years th ese cen ters h ave collaborated an d com bin ed the best available evidence w ith their collect ive exper t op in ion to en h an ce th e p u b lish ed literat u re an d fu r th er de n e th e essen t ials of su ccessfu l su rgical care. Surger y plays a key role in th e m an agem en t of m ost prim ar y sp in al t u m ors. Historically, in ferior resu lt s an d h igh recurren ce rates w ere re ect ive of a p oor u n derst an ding of th e fu n dam en t als in tegral to th e su ccessfu l op erat ive m an agem en t of th ese rare en t it ies, cou p led w ith th e tech n ical com plexit ies of th eir surgical ap proach . More recen tly, tech n ological advan ces an d tech n ical descript ion s of a n um ber of surgical approach es h ave dem on st rated th e feasibilit y of follow ing eviden ce-based on cological care principles. Despite this, and because of th e poten t ial for severe an d un accept able n eurologic de cits an d th e in t rin sic st abilit y fu n ct ion of th e sp in al colu m n , su rger y rem ain s ch allenging. Th is ch apter discu sses th e key prin ciples to con sider w h en select ing a su rgical ap proach an d illu st rates th em w ith represen tat ive cases.

Origin ally described in th e 1980s for p rim ar y t u m ors of th e ap pen dicu lar skeleton , th e En n eking classi cat ion rem ain s th e fun dam en tal foun dat ion for approach ing prim ar y m esen chym al tum ors, including those originating from th e sp in e.1 Th is system classi es p rim ar y t u m ors based on th ree factors: a grade of t h eir biological aggressiveness (G), its local extent (T), an d th e presen ce or absen ce of m etast asis (M). Histological, radiological, an d clin ical feat u res correlate to id en t ify ben ign (G0), low grade m align an t (G1), or h igh -grade m align an t (G2) lesions. Ben ign (G0) lesions are th en fu rth er su bcategorized as being laten t (st age 1), act ive (stage 2), or aggressive (stage 3) based on ch aracter ist ics of t h e t u m or h ost m argin . St age 1 ben ign lesion s are slow grow ing or st at ic, an d ch aracter ized by m at u re brou s t issue or cort ical bon e en capsulat ion . Stage 2 ben ign lesion s grow steadily an d are bordered by a th in capsule surroun ded by an area of react ive t issue. St age 3 ben ign t um ors exten d rapidly, usually preceded by a th ick p seudocap su le of react ive t issu e w ith a pen et rated or absen t cap sule. Th e locat ion of th e t um or is categorized as intracapsular (T0), intracom partm ental (T1), or

Principles Behind Determining the Right Approach Table 12.1 Enneking Staging System Stage Benign 1 2 3 Malignant 1A 1B 2A 2B 3

Grade

Local extension

Metastases

G0 G0 G0

Intracompartm ental (T0) Intracompartm ental (T0) Intra- or extracompartm ental (T1/T2)

No (M0) No (M0) No (M0)

Low (G1) Low (G1) High (G2) High (G2) Any

Intracompartm ental (T1) Extracompartmental (T2) Intracompartm ental (T1) Extracompartmental (T2) Any

No (M0) No (M0) No (M0) No (M0) Yes (M1)

Source: Adapted from Enneking WF. A system of staging musculoskeletal neoplasms. Clin Orthop Relat Res 1986;204:9–24.

ext racom p ar t m en t al (T2). Absen ce of m et ast ases is d en oted as M0 an d t h e p resen ce of d istan t m et astasis as M1. Th ese th ree factors com bin e to create th e En n eking st age (Table 12.1). For each st age, a speci c resect ion m argin is recom m en ded (Table 12.2). Adh ering to an d resp ect ing th e En n eking prin cip les for t u m ors of th e app en dicu lar skeleton h as been sh ow n to result in a low er rate of

Table 12.2 Modi ed Articulation of Enneking Stages w ith Surgical Margins Enneking Stages 1 2 3 1a 1b 2a 2b 3a 3b

Margin for Control No management unless for decompression or stabilization Intralesional excision ± local adjuvants Marginal en-bloc excision Wide en-bloc excision Wide en-bloc excision Wide en-bloc excision + e ective adjuvants Wide en-bloc excision + e ective adjuvants Palliative Palliative

Source: From Chan P, Boriani S, Fourney DR, et al. An assessm ent of the reliabilit y of the Enneking and Weinstein-Boriani-Biagini classi cations for staging of primary spinal tum ors by the Spine Oncology Study Group. Spine 2009;34:385. Reproduced with perm ission.)

recurren ce an d in im proved sur vival. During recen t decades, advan cem en ts in im aging an d su rgical tech n ology along w ith sp in al on cology su bspecializat ion h as p erm it ted th e safe ap plicat ion of th e En n eking p rin ciples for t u m ors of th e spin al colu m n . As h as been dem on st rated in th e appen dicu lar skeleton , th is st an dardized approach h as been sh ow n to ach ieve acceptable m orbidit y, m or talit y, an d h ealth -related qu alit y of life ou tcom es for p rim ar y sp in e t u m ors.2–4 In a m u lt icen ter am bispect ive coh or t an alysis of 147 p at ien t s w it h p r im ar y sp in e t u m ors, Fish er et al 5 d em on st rated t h at t h e adopt ion of En n eking p rin cip les resu lted in a signi cant reduction in rates of local recurrence an d im proved life expect an cy. St u dies assessing th e recurren ce risk an d t u m or-free sur vival follow ing su rgical resect ion of sp in al colu m n ch ordom as 6,7 an d ch on drosarcom as 8 h ave also con rm ed th e im por t an ce of adh ering to En n eking fu n dam en tals. For spinal colum n tum ors, the Enneking classi cat ion does n ot t ake in to accou n t th e presen ce of a con t in u ou s ep idu ral com par t m en t , th e n eu rologic im p licat ion s, an d th e n eed for restoring spinal stabilit y.9 Furth erm ore, th e size of th e t um or, w h ich h as been foun d to be correlated w ith adverse p rogn osis,6 is n ot con sidered. Th e system w as origin ally based on th e n at ural h istor y of m esen chym al t um ors an d thus is not applicable to tum ors originating from bon e m arrow, th e ret iculoen doth elial system , or m et ast at ic carcin om as.

135

136

Chapter 12

■ Weinstein-Boriani-Biagini

Classi cation Th e Wein stein -Borian i-Biagin i (W BB) classi cation incorporates the En neking principles into th e st aging an d su rgical ap p roach of p rim ar y t u m ors of th e sp in e by addressing th e u n iqu e com plexit y of lesions in this anatom ic set ting.10 Th e p ion eering w ork of th ese au th ors sign i can tly assist s su rgical plan n ing by est ablish ing feasibilit y criteria an d st rategies for ach ieving on cological resect ion of th ese t um ors w h ile sparing th e n eurologic elem en ts based on th e locat ion of th e lesion . In th is su rgical st aging system , ver tebral bod ies are divided in to 12 equ al radiat ing zon es in th e axial p lan e (Fig. 12.1) n u m bered clockw ise from t h e left sid e of th e sp in ou s process (1) to th e righ t side of th e sp in ou s p rocess (12). Th e t u m or is fu r t h er divid ed in to ve con cen t ric layers cen tered about the dural sac: A, extraosseous soft tissues; B, in t raosseou s sup er cial; C, in t raosseou s deep ; D, (ext raosseou s ext radu ral; an d E, ext raosseous in t radu ral. Fin ally, th e longit u din al exten t of th e t u m or is recorded as th e n u m ber

Fig. 12.1 Weinstein-Boriani-Biagini (WBB) surgical staging of spine tum ors. (From Boriani S, Biagini R, De Iure F, et al. En bloc resections of bone tumors of

of ver tebral segm en t s involved. Based on th e W BB st ages, Borian i et al4 proposed in dicat ion s for surgical procedures based on th eir experien ce w ith 29 pat ien ts. For t u m ors of th e spin al colu m n , m oderate in terobser ver reliabilit y an d su bst an t ial n earperfect in t raobser ver reliabilit y for both th e En n eking an d W BB classi cat ion , in term s of st aging an d gu idan ce for t reat m en t , h ave been rep orted.11

■ Surgical Margins St an dardized term in ology, w h ich h as been a sou rce of con fu sion in th e literat u re, is fu n dam en t al for th e st u dy of p rim ar y sp in al colu m n t u m ors. Cu ret tage or p iecem eal resect ion refers to deliberate in t ralesion n al resect ion . En bloc resect ion , on th e oth er h an d, sign i es an at tem pt to resect a t u m or in on e piece. En -bloc resect ion on it s ow n h as n o clear sign i cat ion w ithout an appropriate histological description of t h e resect ion m argin s (in t ralesion al, m argin al, w ide, or radical). An in t ralesion al m argin

the thoracolum bar spine. A preliminary report on 29 patients. Spine 1996;21:1927–1931. Reproduced with permission.)

Principles Behind Determining the Right Approach den otes th at th e p lan e of dissect ion h as t ran sgressed in to th e lesion . A m argin al m argin refers to dissect ion w ith in th e react ive zon e or pseudocapsule surrounding the t um or, w h ereas a w ide resection m argin sign i es the dissection h as occurred beyon d th e react ive zon e th rough norm al tissue. In spine surger y, radical m argins, w h ich refer to ext racom par t m en tal resect ion , are rarely feasible, as th e epidural space is con sidered a con t in u ou s com p ar t m en t exten ding from occiput to the sacrum . The on ly theoretical scen ario is a stage I/IIA t u m or tot ally con n ed to the vertebra (w ithout epidural disease) w here a com plete resect ion in cluding th e spinal cord is perform ed. Achieving w ide resection m argin s in t u m ors involving th e epidu ral com part m en t w ould im ply dissect ion of du ral t issue, w h ich would signi cantly com plicate the approach and in crease th e com p licat ion rate. Con sequ en tly w h en th ere is epidural t um or, m argin al resect ion at th e du ra is often p erform ed. Th e cost– ben e t rat io of ach ieving a w ide resect ion at th e d ura h as n ot been fu lly evalu ated. W h en preser vat ion of th e u n involved n eu ral elem en t s can n ot be ach ieved w ith out com prom ising th e su rgical m argin s, th e su rgeon , conjoin tly w ith th e pat ien t , h as to decide on a com p rom ise. Creat ing a n eu rologic de cit can be ju st i ed to ach ieve a p oten t ial cu re in a p at ien t w ith a long life expect an cy. Th e altern at ive is to com prom ise th e En n eking surgical prin cip les an d p erform a plan n ed t u m or t ran sgression to en able deliver y of th e t u m or w ith out neural elem ent sacri ce. Com prom ising the surgical m argins in a planned, predictable fashion h as th e poten t ial to cau se less t u m or spillage an d a low er rate of local recu rren ce th an a deliberate intralesionn al resection. It cann ot be overstated th at th ese decision s m ust in clude th e p at ien t’s person al preferen ces follow ing a t h orough d iscu ssion of opt ion s. Ch apter 11 d iscu sses t h e issu es of su rgical m argin s an d p lan n ed t ran sgression in det ail.

■ Determining the Diagnosis Th e ap proach to p rim ar y sp in al t u m ors begin s w ith determ in ing th e appropriate h istological

diagn osis. Becau se of th e rarit y of th ese t um ors, a p ath ologist exp erien ced w ith diagn osing th ese con dit ion s sh ou ld be con su lted. Th e d i eren t ial d iagn osis in clu d es p r im ar y m esen chym al t u m ors, prim ar y n on m esen chym al tum or (hem atological or reticuloendothelial system m alignancies), m etastatic disease, pyogen ic an d n onpyogen ic spin al in fect ion s, in am m ator y processes, as w ell as th e m ore com m on t rau m at ic an d degen erat ive p ath ologies. Th e m u lt it u de of p rim ar y t u m or p ath ologies an d their varying intrinsic biological behaviors m ake determ in ing an accu rate diagn osis of th e u p m ost im p or t an ce. Ew in g’s sarcom as, for exam p le, can resp on d st r ikingly to n eoadju van t th erap ies an d can th u s be con sidered for m edical th erapy on ly. How t issu e is obt ain ed to facilit ate a h isto logical d iagn osis can h ave a sign i can t in u en ce on p at ien t ou tcom es an d t h e su rgical st rategy su bsequ en tly selected. A system at ic review add ressed th is issu e.3 A com puted tom ography (CT)-guided trocar biopsy appears to be the best oncological w ay of safely determ ining a diagn osis, as in cision al or open biopsies h ave been associated w ith high er recurren ce rates an d low er disease-free sur vival. W h en a prim ar y spin al t u m or is su sp ected, an on cological spin e su rgeon sh ould be involved early, as th e biop sy t ract n eeds to be m arked to en su re its in clu sion during th e d e n itive resect ion .

■ Multidisciplinary Teams Assem bling a m u lt id iscip lin ar y team for preop erat ive d ecision m akin g an d p lan n in g is fu n dam en t al to ach ievin g opt im al p rim ar y spin e t um or m an agem en t . Each case sh ould be discu ssed in rou n ds at ten ded by an on cological spin e surgeon , radiat ion on cologist , m edical oncologist, pathologist, and radiologist. A group review of th e local an d system ic w orku p, h istological diagn osis, an d staging w ill en able con sen sus con clusion s to be reach ed on key issu es in clu d ing t h e role an d feasibilit y of su rger y, an d t h e su it abilit y of n eoadjuvan t an d adju van t th erapies. If a n eoadjuvan t t reat m en t is deem ed n ecessar y, restaging is recom m en ded

137

138

Chapter 12 after com plet ion , before con tem plat ing surgical t reat m en t . W hen resection is indicated, various surgical su bspecialt ies can play sign i can t su p p or t ing roles to th e spin e surgeon . W h en con sid erable vascular m anipulation or reconstruction is necessar y, a vascu lar su rgeon is in dispen sable. Plast ic su rgeon s can assist w ith th e closu re of precariou s w ou n ds or p erform m u scu locu tan eou s ap coverage w h en n ecessar y. W h en abdom in al or p elvic con ten t s an d t h e t u m or are closely ap p roxim ated, a gen eral su rgeon can assist in th e dissect ion an d create d erivat ion colostom ies w h en requ ired. Com p lex d issect ion s of th e n eck can be facilit ated by h ead an d n eck surgeon s, w h ereas dissect ion w ith in th e ch est cavit y can be assisted by t h oracic su rgeon s.

■ Surgical Planning Becau se each prim ar y sp in e t u m or h as u n ique characteristics and localizations, surgeons m ust be exible about determ ining the right approach. Each case sh ou ld be app roach ed in dividu ally. Th e ch oice of su rgical m argin s according to th e En n eking prin cip les is th e rst step in p lan n ing su rgical in ter ven t ion (Table 12.2). With advancem ents in surgical expertise, im aging and technology are im portant adjuncts in determ in ing the surgical approach to any spinal t um or. We believe th ere sh ou ld n ot be any p reset ap proach es based on localizat ion . En -bloc resect ion for t u m ors of th e spin al colu m n w as rst described in th e 1960s. Lièvre et al12 described a t w o-st age en -bloc resect ion of an L4 gian t cell t u m or. Th e rst st age con sisted of resect ion of th e posterior sp in al elem en ts, follow ed by an an terior vertebral body resect ion 2 w eeks later. In th e follow ing years, Sten er described a sim ilar ap p roach for a gian t cell t um or of T11-L1 13 an d later rep or ted on h is exten sive experien ce (1968–1981) w ith com p lete rem oval of vertebrae for ext irp at ion of t um ors in 23 con secutive pat ien ts w ith a m in im um of 7-year follow -u p .14 Th e basic prin ciples used in h is case series are st ill u sed today.

Sten er p ion eered th e fu n dam en tal prin cip le of con sidering th e spin al colu m n to be like a ring th rough w h ich th e n eu ral elem en t s p ass. To ach ieve a resect ion w h ile adh ering to En n eking’s prin ciple an d preser ving n eu rologic fu n ct ion , a t u m or-free w in dow h as to be created in th at “ring” th rough w h ich th e sp in al cord can be delivered du ring t u m or rem oval. Th e w in d ow can be created by piecem eal resect ion becau se it involves n on t u m oral t issu e. A secon d fun dam en tal prin ciple is to h ave access to th e n er ve root at th e dural m argin via a clear p lan e of dissect ion bet w een t h e t u m or an d th e dura w h en am put at ion of a root is n ecessar y. Th is is often w h ere on ly a m argin al resect ion can be accom plish ed, bu t it is essen t ial for th e deliver y of th e t um or. An oth er area at risk of oncological contam ination is the pedicle. Th e localizat ion of t h e t u m or w ith in th e “ring” w ill dict ate th e feasibilit y, approach , an d m eth od by w h ich th e t u m or is delivered during en-bloc resection. Boriani et al10 described th ree com m on scen arios for th oracolum bar lesion s. First , an en -bloc resect ion can be perform ed w ith an app rop riate m argin if th e t u m or is localized in zon es 4 to 8 or 5 to 9 of th e W BB st aging system . In oth er w ords, th e posterior elem en ts an d at least on e pedicle h ave to be free of t u m or to be able to ach ieve on cological resect ion w h ile safely delivering th e spin al cord. Th is can be ach ieved via a single posterior or via a staged posteroanterior approach .15 Secon d, a sagit tal resect ion can be accom plish ed w h en th e t um or is con n ed to zon es 3 to 5 or 8 to 10. Again , th is can be com pleted th rough a single posterior or an anteroposterior approach. Th ird, w h en th e t um or is con n ed to th e posterior elem en t s on ly (zon es 10 to 3), en -bloc resect ion can be accom plish ed using a posterior ap proach on ly (Table 12.3). Tu m ors located in t h e cer vical sp in e p ose a sign i can t ch allen ge becau se of t h e m any u n ique an atom ic feat ures of th is region . Th e close proxim it y of th e ver tebral ar ter y, th e peculiar bony architecture, and th e functional im p ort an ce of th e cer vical roots all m ake en -bloc resect ion of a cer vical spin e t um or dem an ding. Alth ough th e feasibilit y of en -bloc resect ion of p rim ar y t u m ors of th e cer vical sp in e h as been

Principles Behind Determining the Right Approach Table 12.3 Articulation of Weinstein-BorianiBiagini (WBB) Stages w ith Surgical Procedures Radiating Zone

Procedure

4–8 or 5–9

Vertebrectomy (double approach) Sagit tal resection (double approach) Posterior arch resection (posterior approach)

2–5 or 7–11 10–3

Source: From Chan P, Boriani S, Fourney DR, et al. An assessm ent of the reliabilit y of the Enneking and Weinstein-Boriani-Biagini classi cations for staging of primary spinal tum ors by the Spine Oncology Study Group. Spine 2009;34:385. Reproduced with perm ission.)

dem on st rated, becau se of th ese an atom ic issu es sign i can t m orbidit y an d com p licat ion s are frequen t .16–19 Am ong oth ers, prolonged in t ubat ion , acu te respirator y dist ress syn drom e, d eh iscen ce of t h e p oster ior p h ar yn geal w all, p rolon ged dysp h agia, an d h ardw are failu re requ ir ing revision h ave all been rep or ted .17,19 In th e decision -m aking p rocess th ese adverse even t s h ave to be p lot ted again st th e dism al n at u ral h istor y of th ese t u m ors an d h ave to be d iscu ssed can didly w ith th e p at ien t . Th e con sequ en ces of u n ilateral ver tebral ar ter y resect ion are d ep en d en t on t h e var iable an atom y of t h e ver tebrobasilar system .20 An en d ovascu lar balloon -occlu sion test , cerebral blood- ow st u dies, an d in t raoperat ive tem porar y occlusion w ith n eurom onitoring are st rategies for assessin g w h et h er sacr i ce can be p er form ed w it h ou t sign i can t isch em ic com p rom ise. Alth ough th e resect ion of cer vical n er ve root s frequ en tly carries sign i can t m orbid it y, in selected cases it can be tolerated. Bailey et al19 repor ted n o clin ically sign i can t con sequen ces follow ing un ilateral C2 to C4 n er ve root sacri ce despite poten t ial h em idiaph ragm p aralysis. Tu m ors located in th e sacr um are associated w ith a di eren t set of su rgical con siderat ion s. From a neurologic perspective, sacral roots often n eed to be sacri ced to ach ieve a t ru e en -bloc on cological resect ion , pu t t ing bow el, bladd er,

an d sexual fu n ct ion at risk. Todd et al21 sh ow ed th at preser vat ion of bow el an d bladder con t in en ce after m ajor sacral resect ion occurs in th e m ajorit y of pat ien t s follow ing u n ilateral sacral root resect ion or if at least on e S3 root is preser ved in th e case of bilateral sacral resect ion . Oth er con siderat ion s in clu de th e close p roxim it y of m ajor vessels an teriorly an d th e risk of sign i can t blood loss. Th e close p roxim it y of th e an u s an d t h e risk of bow el perforat ion also in crease t h e r isk of su rgical-site in fect ion . Der ivat ion colostom ies, especially w h en loss of bow el fun ct ion is an t icipated, h ave been u sed to m itigate this occurrence. Wide resection also result s in large soft t issu e defects. Rot at ion al gluteal or t ran spelvic vert ical rect us abdom in is m yocu tan eou s (VRAM) aps can be u sed to obliterate th is dead space an d im prove w oun d h ealing.22 Lastly, h igh sacral resect ion s can in terfere w ith spin op elvic stabilit y, n ecessitat ing com p lex recon st ru ct ive tech n iqu es (see Ch ap ter 13). At our institution, preoperative angiography an d em bolizat ion is rout in ely at tem pted. Th is en h an ces kn ow ledge of th e vascu lar an atom y an d h elps w ith operat ive plan n ing an d th e m an agem en t of in t raop erat ive blood loss. Even t h ough t u m oral bleed in g sh ou ld n ot be a con cern if t u m or violat ion is avoided, em bolizat ion of involved radicular ar teries in th e th oracolu m bar spin e an d th e vertebral arter y in th e cer vical spin e, as w ell as m ajor t u m or feeders, h as been e ect ive in redu cing blood loss– related com plicat ion s.

■ Illustrative Cases Case 1 History and Examination A 50-year-old Asian m an p resen ted w ith an 18-m on th h istor y of p rogressive r igh t -sid ed clu m sin ess, w eakn ess, an d n u m bn ess, st ar t ing in it ially in h is righ t arm th en p rogressing to in volve h is righ t leg. Exam in at ion dem on st rated bilateral u pper ext rem it y altered sen sat ion as w ell as hyperre exia an d gait dist urban ce con -

139

140

Chapter 12 sisten t w ith m yelop athy bu t w ith ou t object ive m otor de cit .19

Staging Local staging, w ith CT an d m agn et ic reson an ce im aging (MRI), revealed a large C1–3 m ass expan d ing in to th e su rrou n ding soft t issu es an d spin al can al, causing m arked righ t-sided spin al cord com p ression an d en casing th e righ t C1,

C2, an d C3 n er ve root s (Fig. 12.2). Th e C2 vertebral body (VB) w as prim arily involved w ith exten sion in to th e base of th e den s. Th e righ t ver tebral ar ter y (VA) w as deviated p osterolat erally by t h e t u m or, bu t rem ain ed p aten t on p reop erat ive angiograp hy. An occlu sion test w as tolerated w it h ou t n eu rologic sequ elae. System ic w orku p revealed t h at t h is w as an isolated lesion . A CT-guided core-n eedle biopsy w as perform ed ju st lateral to th e m idlin e an d

a

c

b

Fig. 12.2a–c Case 1: preoperative images. (a) Sagit tal T2-weighted magnetic resonance imaging (MRI) showing the large tumor centered on C2. (b) Coronal postgadolinium image. (c) Axial T2-weighted im age showing the characteristic features of a chordoma as well as spinal cord displacement by the tumor.

Principles Behind Determining the Right Approach perpen dicu lar to th e skin to en able in clu sion of th e biop sy t ract in th e su bsequ en t resect ion . Ch ordom a w as con rm ed h istologically. Th e case w as subjected to m ultidisciplinar y review. Th e rst step in surgical plan n ing is to classify th e t u m or an d ch oose ap p rop riate resect ion m argin s. Using th e En n eking classi cat ion , th is is a low -grade m align an t t u m or w ith ext racom p ar t m en tal sp read (stage 1B), requ iring a w ide en -bloc resect ion to ach ieve cure. Neoadjuvan t th erapy w as n ot in dicated. Surgical st aging, u sing th e W BB classi cat ion , localized th e t u m or to areas 1 to 6, A to D. Su rgical resect ion w as con sidered feasible becau se th e t w o ten ets of en -bloc resect ion w ere sat is ed: rst , a w in dow could be created in th e spin e ring to deliver th e t u m or from th e sp in al cord ; an d secon d, a p lan e cou ld be d evelop ed from th e spin al cord to reach th e origin of th e roots on th e t u m or sid e, allow ing t h em to be divid ed an d t aken w it h t h e t u m or sp ecim en . A d et ailed p reop erat ive su rgical p lan w as com p leted . Wide m argin s w ere plan n ed, except at th e level of th e du ra, w h ere m argin al resect ion w as an t icipated.

Surgical Execution (Fig. 12.3) An am bit ious 3-day su rgical plan w as devised an d p erform ed. On th e rst day, su rgical in st r u m en t at ion from th e occipu t to C6 w as com p leted. Th e biopsy t ract w as in corporated in th e resect ion an d th e righ t-sided VA w as ligated at th e C1- C2 jun ct ion proxim ally an d at the C3-C4 level distally. A left-sided C2-C4 lam inectom y facilitated visualization of the thecal sac an d exit ing n er ve root s. On th e left side, an osteotom y w as th en perform ed from above th e C2 p edicle to th e C3- C4 disk space. Th e cau dal osteotom y th rough th e C3- C4 disk space w as th en com p leted from left to righ t . Th e righ t C2C3-C4 n er ve root s w ere sacri ced at th eir du ral origin . On th e secon d day, w ith th e pat ien t placed in th e righ t lateral p osit ion , th e p osterior in cision w as reop en ed an d, w ith th e assist an ce of a h ead an d n eck surgeon , an exten ded an terolateral ap proach w as u n d er t aken , facilit at ing visualizat ion of th e an terolateral asp ects of th e VB from th e sku ll base to C4. Th e left-sided os-

teotom ies t h at w ere accom p lish ed t h e day before w ere t h en com p leted an teriorly u n d er direct visu alizat ion . Th e an terior arch of C1 w as th en resected on th e left to facilitate visualizat ion of th e odon toid process to com plete th e osteotom y at t h at level. Fu r th er t u m or m obilization w as then achieved on the third day to perm it th e t u m or to be rot ated an d delivered en bloc th rough th e an terior w ou n d. Recon st r u ct ion w as th en com p leted. Fin al h istological an alysis con rm ed a ch ordom a w ith w ide m argins th roughout the specim en except at th e du ra, w h ere, as an t icip ated, m argin al resect ion w as n oted.

Discussion Th is case illu st rates t h e feasibilit y of a t r u e en -bloc resect ion of a p r im ar y t u m or in t h e cer vical sp in e. Alth ough it is an exh au st ive procedu re, con su m ing con siderable t im e an d resou rces, an d associated w ith sign i can t m orbidit y, th is eviden ce-based approach is w ell foun ded to preven t a poten t ially cu rable disease from becom ing in cu rable by an in t ralesion al resect ion .

Case 2 History and Examination A 30-year-old m an presen ted w ith an 8-m on th history of righ t sacroiliac an d leg pain . The right leg sym ptom s were pluriradicular in nature but predom in an tly involved th e S1 dist ribu t ion . Physical exam in at ion revealed 3/5 w eakn ess in th e righ t L4-L5-S1 m yotom es.

Staging Local st aging revealed an exten sive lesion , cen tered on th e righ t sacroiliac join t (Fig. 12.4). Th e t u m or exten ded th rough th e greater sciat ic foram en an d w as im m ediately adjacen t to th e an terior asp ect of th e righ t sacru m at th e S2-S3 level. The patient w as system ically cleared of any t u m oral act ivit y. Tissu e obtain ed w ith CT-guided, core-n eedle biopsy w as h istopath ologically an alyzed an d a ch on drosarcom a w as diagn osed. According to th e En n eking classi cat ion , for cu rat ive in ten t th is t u m or requ ires

141

142

Chapter 12

Fig. 12.3 Case 1: artist’s illustration of the procedure. (From Bailey CS, Fisher CG, Boyd MC, Dvorak MFS. En bloc marginal excision of a multilevel

cervical chordoma. Case report. J Neurosurg Spine 2006;4:409–414. Reproduced with permission.)

Principles Behind Determining the Right Approach

a

143

b

Fig. 12.4a,b Case 2: preoperative images. (a) Axial T2-weighted MRI showing the extensive tum or centered on the right sacroiliac joint. (b) Sagit tal postgadolinium image.

a w ide en -bloc resect ion . How ever, th ere are several ch allenges to ach ieving th is resect ion , an d sign i can t associated m orbidit y is predicted . Th e close p roxim it y of th e righ t-sided sacral root s w arran t th eir com plete sacri ce along w ith th at of th e righ t sciat ic n er ve, w ith th e exp ected associated n eu rologic d e cit s discu ssed above, in clu ding th e risk of im p aired bow el, bladder, an d sexual fun ct ion ing. Given th e exten t of bony resect ion n ecessar y to ach ieve w ide m argin s, recon st ruct ion w ou ld also be ch allenging. How ever, follow ing applicat ion of th e on cological p rin cip les discu ssed in th is ch apter, su rgical resect ion w as deem ed possible. Th e diagn osis, proposed su rger y, poten t ial com p licat ion s, an d exp ected d e cit s w ere discussed w ith th e pat ien t on m ult iple preop erat ive occasion s. Becau se of h is age an d th e ch an ce of a cu re, th e decision w as m ade to u n dert ake su rgical resect ion .

Surgical Execution A com plex surgical plan w as developed by a m u lt idiscip lin ar y su rgical team , con sist ing of a gen eral su rgeon , a p last ic su rgeon , an or t h o p ed ic t rau m a su rgeon , an d t w o on cological sp in e su rgeon s. Th e rst st age con sisted of a p oster ior ap p roach an d p oster ior release of th e t um or. In st r um en t at ion w as carried from L3 to th e p elvis. Th e righ t-sided L5 to S5 n er ve root w ere cu t . On th e secon d day, w ith th e pa-

t ien t p laced in th e lateral decu bit u s p osit ion , sim ultaneous anterior and posterior approaches w ere execu ted . After a m et icu lou s an terior retroperiton eal release, th e osteotom y cuts in it iated during th e rst st age w ere com pleted. On ce freed th e t um or w as delivered en bloc posteriorly. Com plex recon st ru ct ion w as th en perform ed ut ilizing a vascu larized bu la graft , an d in st ru m en t at ion bridging L3 to th e isch ial t u berosit y on t h e righ t , an d L3 to t h e ileu m on th e left (Fig. 12.5). Su rgical m argin s w ere all n egat ive. The patient’s postoperative course w as com plicated by a deep w oun d in fect ion requiring m ult ip le debridem en ts an d closure w ith a free latissim us dorsi transfer. At 7-year follow -up, he is am bulating, has preser ved bow el and bladder fun ct ion , an d n o sign of t um or recu rren ce.

Discussion of Both Cases Th ese cases illu st rate th e basic p rin ciples leading to de n it ive su rgical m an agem en t of p rim ar y sp in al colu m n t u m ors. W h en , according to Enn eking’s evidence-based classi cation system , an en -bloc resect ion is in dicated an d its feasibilit y is suppor ted by th e W BB surgical classi cation, surgical execution is possible. The exp ected com p licat ion s of each sp eci c ap proach h ave to be an t icipated . Detailed su rgical p lan n ing is cen t ral to th e su ccess of th ese dem an ding procedu res.

144

Chapter 12

a

b

Fig. 12.5a,b Case 2: postoperative images. (a) Computed tomography coronal reconstruction showing solid bony union. (b) Anteroposterior X-ray showing the complex reconstruction performed.

Each t u m or is u n iqu e. Carefu l con siderat ion h as to be given to th eir in dividual an atom y an d t u m or ch aracter ist ics. W h en resect ion is on cologically possible, w e n ow h ave th e su rgical exp er t ise, tech n iqu es, an d recon st ru ct ion capabilit y for any app roach to ach ieve th is goal in any locat ion of th e spin e. It is our experien ce th at m et icu lou s p reoperat ive p lan n ing u sing an atom ic draw ings an d th e creat ion of a step by-step surgical plan helps ensure achievem ent of evid en ce-based su rgical goals even in ext rem ely ch allenging scen arios. Sp in e su rgeon s dealing w ith p rim ar y sp in al tum ors are urged to adopt recognized term inology and classi cation system s to standardize th e t reat m en t an d en su re con sisten t report ing in th e pu blish ed literat u re, w ith th e u lt im ate aim of im proving pat ien t care.

■ Chapter Summary Cen ters w ith th e n ecessar y su rgical experien ce an d subspecialt y suppor t are best ser ved to

m an age p rim ar y sp in e t u m ors. Carefu l st aging an d surgical plan n ing are m an dator y for th ese challenging high-risk procedures. Respecting the resect ion m argin s proposed in th e En n eking classi cat ion h as been correlated w ith low er recurren ce rates an d in creased t um or-free survival. A fun dam en tal con cept is to con sider th e spin al colum n as a ring th rough w h ich th e n eu ral elem en ts pass. To be able to ach ieve an en bloc resect ion w h ile adh ering to En n eking’s prin cip les, a t u m or-free w in dow h as to be created in th at “ring” to en able th e t u m or to be rem oved w h ile p reser vin g sp in al cord in tegr it y. Alt h ough t h ese p r in cip les d eter m in e t h e feasibilit y of on cological resect ion , th ere are several available su rgical dissect ion s, aided by tech n ological advan ces an d subspecialt y surgical su p por t , th at en able th e speci c su rgical st rategy to be in dividu ally t ailored for each tum or. The m orbidit y of each approach requires carefu l con siderat ion . Alth ough in ict ing a n eu rologic de cit is som et im es an accept able com p rom ise to opt im ize th e ch an ce of a cu re, at all t im es a p at ien t ’s p erson al p referen ces an d opin ion s, guided by an experien ced m ult i-

Principles Behind Determining the Right Approach disciplinar y team , should rem ain cen tral to this often com plex decision -m aking process. Pearls Primary spinal tum or resection is best performed in an experienced center with multidisciplinary support. Adhering to the Enneking principles results in lower rates of recurrence and improved survival. A tumor-free window has to be created in the spinal “ring” to allow both an en-bloc resection and neurologic preservation. The localization of the tumor within the ring will dictate the feasibilit y of en-bloc resection.

The surgical approach should be customized to the tumor’s characteristics. Appreciating the morbidit y associated with each approach is integral to the surgical decision-m aking process. Patients need to be fully inform ed and their preferences considered in all surgical decisions. Pitfalls The importance of an accurate histological diagnosis should not be understated. Obtaining tissue for analysis using a suboptim al technique can profoundly a ect patients’ outcomes and their future surgical options.

Refere nces Five Must-Read Refe rences 1. En n eking W F. A system of st aging m usculoskelet al neoplasm s. Clin Orth op Relat Res 1986;204:9–24 PubMed 2. Fisher CG, Keyn an O, Boyd MC, Dvorak MF. Th e surgical m an agem ent of prim ar y t um ors of th e spin e: in it ial result s of an ongoing prospect ive coh or t st udy. Spin e 2005;30:1899–1908 PubMed 3. Yam azaki T, McLough lin GS, Patel S, Rh in es LD, Fourney DR. Feasibilit y an d safet y of en bloc resect ion for prim ar y sp in e t u m ors: a system at ic review by th e Spin e On cology St udy Group. Spin e 2009;34(22, Su ppl):S31–S38 PubMed 4. Borian i S, Biagin i R, De Iure F, et al. En bloc resect ion s of bon e t u m ors of th e th oracolum bar spine. A prelim in ar y report on 29 pat ien t s. Spin e 1996;21:1927– 1931 PubMed 5. Fisher CG, Saravanja DD, Dvorak MF, et al. Surgical m an agem en t of prim ar y bon e t u m ors of th e spin e: validation of an approach to enhance cure and reduce local recu rren ce. Sp in e 2011;36:830–836 10.1097/ BRS.0b013e3181e502e5 PubMed 6. Bergh P, Kin dblom LG, Gu n terberg B, Rem ot t i F, Ryd W, Meis-Kin dblom JM. Progn ost ic factors in ch ordom a of th e sacr u m an d m obile spin e: a st u dy of 39 pat ien t s. Can cer 2000;88:2122–2134 Pu bMed 7. Borian i S, Ch evalley F, Weinstein JN, et al. Ch ordom a of th e spin e above th e sacr um . Treat m en t an d outcom e in 21 cases. Spin e 1996;21:1569–1577 Pu bMed 8. Borian i S, Saravanja D, Yam ada Y, Varga PP, Biagin i R, Fish er CG. Ch allen ges of local recu r ren ce an d cu re in low grade m align an t t u m ors of th e sp in e. Sp in e 2009;34(22, Su ppl):S48–S57 PubMed 9. Jaw ad MU, Scu lly SP. In brief: classi cat ion s in brief: En n eking classi cat ion : ben ign an d m align ant t um ors of th e m usculoskelet al system . Clin Orth op Relat Res 2010;468:2000–2002 PubMed

10. Boriani S, Wein stein JN, Biagin i R. Prim ar y bon e t um ors of th e spin e. Term in ology an d surgical st aging. Spin e 1997;22:1036–1044 Pu bMed 11. Ch an P, Borian i S, Fourn ey DR, et al. An assessm en t of th e reliabilit y of th e En n eking an d Wein stein -Borian iBiagin i classi cat ion s for st aging of prim ar y spin al t um ors by th e Spin e On cology St udy Group. Spin e 2009;34:384–391 PubMed 12. Lièvre JA, Darcy M, Pradat P, et al. [Gian t cell t um or of th e lum bar spin e; tot al spon dylectom y in 2 st ates]. Rev Rh u m Mal Osteoar t ic 1968;35:125–130 PubMed 13. Sten er B, Joh n sen OE. Com p lete rem oval of t h ree ver tebrae for gian t-cell t um ou r. J Bon e Join t Surg Br 1971;53:278–287 PubMed 14. Sten er B. Com plete rem oval of ver tebrae for ext irpat ion of t u m ors. A 20-year exp erien ce. Clin Or th op Relat Res 1989;245:72–82 Pu bMed 15. Tom it a K, Kaw ah ara N, Baba H, Tsuchiya H, Nagat a S, Toribat ake Y. Tot al en bloc spon dylectom y for solit ar y spin al m et ast ases. In t Orth op 1994;18:291–298 PubMed 16. Coh en ZR, Fou rn ey DR, Marco RA, Rh in es LD, Gokaslan ZL. Tot al cer vical sp on dylectom y for p rim ar y osteogen ic sarcom a. Case repor t an d descript ion of operat ive tech n ique. J Neurosu rg 2002;97(3, Suppl): 386–392 Pu bMed 17. Rh in es LD, Fou rn ey DR, Siadat i A, Su k I, Gokaslan ZL. En bloc resect ion of m u lt ilevel cer vical ch ordom a w ith C-2 involvem en t . Case repor t an d descript ion of operat ive tech n ique. J Neurosurg Spin e 2005;2:199– 205 PubMed 18. Fujit a T, Kaw ah ara N, Mat sum oto T, Tom it a K. Ch ordom a in th e cer vical spin e m an aged w ith en bloc excision . Spin e 1999;24:1848–1851 PubMed 19. Bailey CS, Fish er CG, Boyd MC, Dvorak MFS. En bloc m argin al excision of a m u lt ilevel cer vical ch or-

145

146

Chapter 12 dom a. Case rep or t . J Neu rosu rg Sp in e 2006;4:409– 414 PubMed 20. Hosh in o Y, Kurokaw a T, Nakam ura K, et al. A rep or t on the safet y of u n ilateral ver tebral ar ter y ligat ion during cer vical spin e surger y. Spin e 1996;21:1454– 1457 Pu bMed 21. Todd LT Jr, Yaszem ski MJ, Currier BL, Fu ch s B, Kim CW, Sim FH. Bow el an d blad der fu n ct ion after m ajor

sacral resect ion . Clin Orth op Relat Res 2002;397: 36–39 PubMed 22. Sciubba DM, Ch i JH, Rhin es LD, Gokaslan ZL. Chordom a of th e sp in al colu m n . Neurosu rg Clin N Am 2008;19:5–15 Pu bMed

13 Spinopelvic Reconstruction/ Fixation and Fusion Peter Paul Varga

■ Introduction Th e t u m orou s p ath ologies of th e lu m bosacral ju n ct ion requ ir in g su rgical in ter ven t ion are ver y ch allenging even for th e m ost experien ced spin e surgeon s. Tum ors of th is region often grow in sidiously w ith ou t speci c sym ptom s, reach ing a large size th at m akes th eir com p lete rem oval a com p lex feat . Du e to th e localizat ion , special issu es sh ould be con sidered w h en in vestigating all aspects of the local anatom ic and biom ech an ical proper t ies irrespect ive of th e par t icu lar on cological app roach . Ach ievem en t of th e opt im al on cosurgical resect ion m argin s often en t ails p oten t ial fu n ct ion al losses, resu lt ing in psych ological issues th at sh ould be con sid ered d u r in g t h e p lan n in g p rocess. Th e pat ien t sh ou ld be p rep ared accordingly by th e su rgeon an d psych ologist s for th is tech n ically dem an d ing su rger y. In m et ast at ic lesion s (w h ich are m ore com m on in th e sacr u m an d in t h e en t ire sp in e th an are th e prim ar y t u m ors), surger y is alw ays palliative. The m ain goal of the palliative procedure is fun ct ion al im provem en t , alth ough th e literat u re suggest s th at debu lking of th e m etast at ic m ass can be on cologically ben e cial an d can provide longer su r vival.1 Prim ar y m align an t sp in al t u m ors are rare con d it ion s, an d th e sacru m is th e m ost com m on locat ion .2,3 Th e m ain aim of su rgical t reatm en t in t h ese cases is to be cu rat ive. En -bloc

resect ion of a p rim ar y m align an t sacral t u m or w it h w id e on cological m argin s cou ld sign i can tly in u en ce th e biom ech an ics of th e sp i n opelvic com plex, w h ich in t u rn can result in severe deteriorat ion of th e am bu lator y st at u s, w orsen ing t h e p ostop erat ive m orbid it y an d m or t alit y. Th e procedures t arget ing th e spin opelvic soft t issu e an d bony recon st ru ct ion after the resection signi cantly lengthen th e surger y an d in crease t h e r isk for in t ra- an d p ostop erat ive com plicat ion s. Due to m ult idisciplin ar y coop erat ion , an d p ar t icu larly to th e develop m en t of th e in ten sive care an d to an est h esiology advancem ents, as w ell as th e spinal im plant tech n ology, it is possible to perform th ese exten sive surgeries even rout in ely in com preh en sive on cosu rgical cen ters, w h ere t h e gen eral m ed ical an d tech n ical environ m en t fosters a h igh level of spin al surgical experien ce. Th is ch apter discusses th e special con siderat ion s of sp in opelvic stabilizat ion after sacral t u m or resect ion .

■ Biomechanical

Considerations Th e body w eigh t from t h e a xial skeleton is t ran sferred to th e low er lim bs via th e sacroiliac (SI) join ts. Resect ion of th e sacr um , involving par t ially or totally th e u n i or bilateral SI join ts,

148

Chapter 13 can a ect th e biom ech an ical in tegr it y of t h e pelvic ring, resu lt ing in dist u rban ces in th e physiological m ovem en t of th e p at ien t p ar t icu larly in st an ding an d w alking. Th ere is n o con sen sus am ong sacral t um or exp er t s regarding th e n eed for st abilizat ion an d bony recon st ru ct ion after resect ion of a sacral t um or. Som e auth ors reported that spin opelvic stabilization w as not necessary after total sacrectom y.4,5 In th eir opin ion , based on th eir clin ical obser vat ion s, th e rem ain ing m uscles an d th e developing scar t issu e at th e site of th e sp in opelvic defect of th e body w all can som ew h at st abilize th e spin e as a sling du ring th e postoperative course. Detailed analysis of funct ion al outcom es after long-term follow -up is st ill un clear. In con t rast , biom ech an ical st udies an d case ser ies of p at ien t s u n d ergoing tot al sacrectom y sh ow t h e ben e t s of bony recon st r u ct ion su p p or ted by sp in op elvic st abilizat ion .6–16 In cases of tot al sacrectom y an d in certain cases of high sacrectom y, w here the cranial resection plain is above the S I/II junction un i or bilaterally, sp in op elvic st abilit y sign i can tly d eter iorates.17–19 In cases of h igh sacrectom y, if th e resect ion is don e above th e S1 foram in a, th e rem ain ing sacral bon e is p oten t ially n ot st rong en ough to t ran sfer th e loading from t h e sp in e to t h e p elvic r in g, yield in g a h igh er r isk of su bsequ en t fat igu e fract u res of th e sacral rem ain s.18 Th e im m ed iate advan t age of st abilizat ion is t h at it en ables st ar t in g reh abilit at ion im m ed iately after th e su rger y, star t ing w ith in bed exercises an d, a few days later, progressing to w alkin g. Th e early m obilizat ion of th ese p at ien t s h elp s p reven t t h e p ostop erat ive com p licat ion s related to long term in act ivit y, su ch as th rom boem bolism , p n eu m on ia, an d p ressu re sores. Th e p sych ological e ect of th e early m obilizat ion is also con sid erable. Based on th ese results, an d also suppor ted by our ow n clinical experien ce, spin opelvic stabilizat ion is st rongly advised after destabilizing sacrectom y su rgeries. Th ere are n o gold-st an dard tech n iqu es for spinopelvic stabilization. In vitro studies showed th e e cacy of di eren t st abilizing (i.e., m ot ion reducing) tech n iqu es, bu t th ere is n o eviden ce su ppor t ing th e absolu te advan tage of th e m ost

rigid xat ion con st ru ct . It is exp ected th at th e opt im al spin opelvic xat ion m eth od sh ould be sh ock absorbing an d a fu sion p rom ot ing. Th ere is n o dou bt th at th e develop ing bony fu sion bet w een th e lu m bar sp in e an d th e p elvic bon es w ill p rovid e d e n it ive, lon g ter m biom ech an ical stabilit y. Rigid con st r uct s w ith st ruct ured bon e graft s seem to be e ect ive, bu t t h ere are n o available st u d ies (or rad iological im ages p u blish ed) in t h e literat u re w ith long-ter m follow -up. Th e total rigidit y of th e im plan t con st r u ct can absolu tely ease th e graft–h ost su rface from loading, w h ereas a n on rigid con st ruct ion can enable cyclic loading during w alking, p rom ot ing th e in corporat ion of th e bon e graft . At th is t im e th e opt im al rigidit y of th e di eren t stabilizing con st r u cts an d th eir e ect on graft in corporat ion are open quest ion s. Sagit tal balan ce h as been reported recen tly to be a global biom ech an ical factor in u en cing th e degen erat ive sp in al d isord ers as w ell as th e progression of deform it ies. In th e n at ural cou rse of sagit t al balan ce an d in t h e develop m en t of im balan ce, spin op elvic param eters su ch as p elvic t ilt , sacral slop e, an d p elvic in cid en ce seem to h ave som e p red ict ive valu e. Total sacrectom y an d spin opelvic recon st ruct ion in u en ce th e sp in op elvic p aram eters per se,20 an d th ere can be long-term con sequen ces of the alteration of sagit tal alignm en t, especially if an im balan ced sit uat ion is xed by th e spin o p elvic in st ru m en t at ion . Proper sagit t al balan ce is also assu m ed to play a role in am bulator y abilit y an d in th e develop m en t of im plan t failu res or n on un ion .

■ Spinopelvic Stabilization After Total Sacrectomy Sp in op elvic xat ion (SPF) proced u res join th e lu m bar sp in e to th e pelvis (Fig. 13.1). Th e Galveston L-rod technique w ith pedicle screw s and iliac rods system w as am ong th e rst p u blish ed tech n iques in 1984.21 Th e origin al Galveston tech n ique (GT) perform ed th e SPF w ith th e im plantation of the L rods into the iliac crests, but th ese d irectly im p lan ted rods com m on ly loosen ed. A m odi ed Galveston tech n iqu e (m GT)

Spinopelvic Reconstruction/Fixation and Fusion

a

b

c

d

e

f

Fig. 13.1a–f Spinopelvic  xation (SPF) methods. (a) Galveston technique (GT). (b–d) Modi ed Galveston  technique (mGT). (e) Four-rod reconstruction (FRR). (f) Closed loop technique (CLT).

w as in t rodu ced in 2000 in w h ich th e rods w ere also xed to iliac screw s in th e caudal en d of th e con st ru ct . Th is procedu re h as been w id ely used, and further biom echanical studies showed the positive e ect of the use of dual iliac screw s, in depen den t of th eir orien tat ion .22,23 Th e four rod (FRR) or double rod double iliac screw system is a fu r th er m odi cat ion of th e m GT.24 In th is case, th e addit ion al lu m bar rods in crease th e rigidit y of th e con st ru ct bu t can carr y a h igh er risk of w ou n d h ealing problem s d ue to th e excessive m et al im p lan t s.25 Th e closed loop tech n ique (CLT) for SPF w as pu b lish ed in 2009.15 With th is tech n iqu e, on e sin gle U-sh aped rod is used to con n ect all pedicle an d iliac screw s (bilaterally in th e low er double position ), providing a bet ter load dist ribu t ion along th e con st ru ct (Fig. 13.2). Th is m eth od applies th e ph ilosophy of n on rigid xat ion to avoid th e st ress-sh ielding ph en om en on an d to prom ote th e bony fu sion bet w een th e lu m bar ver tebral body an d th e p elvis. By th is m eth od, safe an d st ron g bony con n ect ion s d evelop w ith in th e rst 24 m on th s after th e surger y (Fig. 13.2g-i).

Posterior pelvic ring xation (PPRF) is an addit ion al dim en sion of spin opelvic stabilizat ion (Fig. 13.3). Th is m eth od uses par t of th e w h ole con st r u ct to con n ect th e t w o iliac bon es w ith each oth er. On e pu blish ed tech n iqu e is th e t riangular fram e recon st r uct ion (TFR), in w h ich on e h or izon t al rod is u sed to at t ach t h e L5 ver tebral body to th e t w o iliac w ings an d th e low er p ar t of th e ilia are also con n ected w ith an oth er h orizon tal rod. Th is tech n ique en ables good load d ist ribu t ion across im p lan t s, bu t excessive st ress in th e iliac bon es cou ld en t ail an in creased risk of im plan t loosen ing an d failu re.13 An ot h er tech n iqu e w as p u blish ed by au t h ors from Joh n s Hop kin s Hosp it al (JHH) in 2005.26 In t h is tech n iqu e, an m GT w it h a t ran siliac bar in ser ted t h rough t h e iliac crest s is com p leted w it h a h or izon t al p elvic xat ion screw -rod im p lan t an d a h or izon t al fem oral allograft placed bet w een th e t w o iliac crest s. The di erent parts of the construct are at tached w ith variou s con n ectors. Anterior spin al colum n xation (ASCF) is th e th ird possible dim en sion of spin opelvic stabilizat ion (Fig. 13.4), w h ich con sist s of adding an

149

150

Chapter 13

a

b

d

e

g

c

f

h

Fig. 13.2a–i A giant sacral chordom a in a 42-yearold man. Sagit tal (a) and axial (b) magnetic  resonance imaging (MRI). (c) Three-dimensional  computed tomography (CT) reconstruction.  (d,e) After an extended total sacrectomy, the 

i

spinopelvic  xation was performed with the closed  loop technique. (f) Morselized bone graft was  placed bet ween the decorticated lumbar laminas  and iliac bones. (g–i) Two years after the surgery,  complete spinopelvic bony fusion can be observed.

Spinopelvic Reconstruction/Fixation and Fusion

a

151

b

Fig. 13.3 Posterior pelvic ring  xation (PPRF) methods. (a) Triangular fram e reconstruction (TFR). (b) Johns Hopkins Hospital (JHH) method.

an terior suppor t to th e lu m bar spin e. Kaw ah ara et al9 u sed a t w o ver t ically in serted p edicle screw s at t ach ed to a sacral rod con n ect ing th e t w o ilia to directly con n ect th e L5 vertebral body to the spinopelvic xation construct. Their n ovel techn ique is a com bin at ion of SPF, ASCF, an d PPRF tech n iques using a m in im al am oun t of im plan ts. Dickey et al7 rep or ted th e bilateral bular graft (BFFR) tech niqu e, w h ere t w o bu lar graft s placed bet w een th e L5 ver tebral body an d th e bilateral iliop ect in eal area provide a good an terior su pp or t st rengt h en ing th e ASCF aspect of th e con struct. It can be com bined w ith any SPF an d SPF+PPRF tech n iqu es.

a

b

Th ere is a w id e var iet y of sp in op elvic st abilizat ion m eth od s rep or ted in t h e literat u re, bu t th e n um ber of cases an d th e details of th e rep or ted long-term radiological an d clin ical resu lts are lim ited or in su cien t . Th e goals of an opt im al spin opelvic stabilizat ion tech n ique after total sacrectom ies are n ot on ly to restore th e spin opelvic st abilit y an d to prom ote th e bony fu sion , but also to suppor t th e di eren t su rgical tech n iqu es for t h e soft t issu e recon st r u ct ion . W it h t h e p rop er st abilizat ion tech n ique, decreasing th e size of th e large cavit y after th e rem oval of th e t u m orou s sacrum can be ach ieved by pulling caudally th e

c

Fig. 13.4a–c Anterior spinal column  xation (ASCF) combined with SPF and PPRF. (a) Novel reconstruction (NR). (b) Bilateral  bular graft technique (BFFR). (c) Modi ed bilateral  bular graft technique (mBFFR). 

152

Chapter 13 rem aining part of the lum bar spine and approxim ating the resection surfaces of the iliac crests. Th e im p lan t con st r u ct ion s cou ld also p rovide the possibilit y of anchoring the soft tissue aps. Th e volu m e of th e im plan t is also a sign i can t factor; th e larger th e am oun t of im plan t , th e h igh er th e risk of w ou n d h ealing problem s.25

■ Spinopelvic Fusion Only a m assive and strong bone bridge bet w een th e lu m bar sp in e an d th e p elvis can p rovide long-term biom ech an ical stabilit y after tot al sacrectom y. If t r u e bony fu sion is n ot ach ieved, th e im p lan t w ill cer t ain ly fail, even if it does n ot cau se any sym ptom s. But th e gap bet w een th e h ost su rfaces can be ver y large. Fu sion can be exp ected on ly 1 to 2 years after th e su rger y, an d delayed fusion or developm en t of pseudarth rosis are com m on com p licat ion s. On ly a few pu blicat ion s h ave rep or ted an an alysis of th e fusion process. In som e con st ru ct s th e st r u ct u ral graft h as an im por t an t m ech an ical role too (see Ch apter 14). Th e BFFR m eth od ach ieves th e ASCF w ith th e im plan t at ion of t w o bu lar graft s bet w een th e base of th e cau dal ver tebral body an d th e pelvic bon e.7 Th e graft s are placed along th e force t ran sm ission lin e bet w een t h e lu m bar sp in e an d th e h ip join t s, creat ing a t riangu lar construct. The proxim al and distal docking sites of the graft are under cyclic com pression , w hich can facilitate bon e h ealing. Th e JHH m eth od rep resen t s an ot h er solu t ion .26 In t h is case, a h or izon t al st r u ct u ral fem u r allograft is u sed to con n ect th e t w o iliac w ings. Th e allograft is in serted in to th e h ost site to bridge th e large defect after tot al sacrectom y an d to ach ieve a bony PPRF. Non st ruct ural bon e graft is used to bridge th e gap bet w een th e cau dal ver tebral body an d th e iliac w in gs p oster iorly. To create a w ellbleeding can cellou s h ost surface, w e decor t icate th e facing bony su rfaces w ith a h igh -sp eed drill. Th e graft is placed in to th e p osterior load t ran sm ission lin es, w h ich facilitate bony fu sion (Fig. 13.2f); h ow ever, redu ct ion of th e gap w ith th e app roxim at ion of th e bony su rfaces is also

advised. Th e am oun t of disease-free local autograft is lim ited in su ch p roced u res, an d can cellou s an d m orselized allograft bon e ch ip s as w ell as syn t h et ic bon e graft s are to be u sed in ord er to h ave su cien t graft m ater ial. By p lacing a layer of Gelfoam , w e provide a soft but safe an terior an d posterior w all aroun d th e m orselized bony graft m ass. In our clinical experience, the fusion process w ith routinely used autologous m orselized graft from th e rest of th e posterior iliac crests takes about 12 to 24 m on th s. We did n ot obser ve any di eren ce in th e t im ing of fu sion com p aring th e au tologou s graft s an d th e syn th et ic osteoin duct ive product s. By th e progression of th e fu sion p rocess, t h e n ew bony m ass d evelop s in th e lin e of th e load t ran sfer. Th at par t of th e bony graft m ass, w h ich is ou t of t h is lin e, slow ly disapp ears, an d after th e secon d year it is n ot even detectable on com p u ted tom ography (CT) (Fig. 13.2). According to the general biom echanical principles of osteosynthesis, the absolute reduction of m ot ion w ith spin opelvic xat ion is presum ably an adverse sit u at ion for th e developm en t of the bony fusion.27 A thin, but so far unknow n, equilibriu m bet w een th e rigidit y an d th e m icrom ovem en t in th e con st r u ct facilit at ing th e biological process of the bony rem odeling w ould be th e opt im al solut ion for sp in op elvic xat ion , but th ere are few clin ical or experim en tal reports.

■ Problems to Avoid Sp in op elvic xat ion is a dem an ding an d ch allenging su rgical tech n iqu e. Nu m erou s st u dies em ph asize th e im por tan ce of a p rop er su rgical tech n iqu e in total sacrectom y, bu t th e sp in o pelvic recon st ru ct ion alon e also requ ires h igh su rgical skill an d detailed p reop erat ive p lan n ing. Oth er w ise, th e on cologically appropriate resect ion of th e sacral t um or can n ot provide an opt im al surgical an d fu n ct ion al outcom e because of th e deteriorated biom ech an ics an d th e subsequ en t im plan t failu res. Spinopelvic reconstruction m ust be planned in detail preoperatively, w ith consideration given

Spinopelvic Reconstruction/Fixation and Fusion to t h e m argin s of t h e bony resect ion an d to th e speci c pat ien t’s ch aracterist ics. Appropriate an d th orough p lan n ing w ill resu lt in few er in t ra- an d postoperat ive com plicat ion s. Im p rovisat ion sh ou ld be avoided du ring th e su rger y. Becau se of t h e com p lex an atom ical, biom echanical, and technical consideration, these surgeries can be safely perform ed on ly after a long learning curve. Both the surgeon and the w hole team (anesthesiologist, residents, nurses) should be exp er ien ced in t h is p rocedu re. Th ere are few st u dies available on t h is p roced u re, an d th ey lack long-ter m follow -u p of large coh orts, but th e biom ech an ical speci cation s th at th ese st u dies discu ss m u st n ot be u n derest im ated . Th e xat ion con st ru ct h as to bear ver y large loading forces; m oreover, th e con st ru ct ion sh ou ld absorb som e of th is load. Oth erw ise, local sh earing st ress th at is gen erated aroun d th e im plan ts can cause dest r uct ion of th e ver tebral/p elvic bon e, p ar t icu larly in p at ien t s w ith osteop orosis. Ach ieving an opt im al st ress d ist r ibu t ion en t ails som e im p or t an t con siderat ion s, su ch as th e length of th e im p lan t at th e lu m bar spin e. Th e n u m ber of th e lu m bar segm en t s involved is h igh ly depen den t on th e pat ien t’s speci c m easu rem en ts an d bon e qu alit y. For exam ple, w e m ust take in to accoun t th e long lever arm s of a t all p erson , an d w e h ave to exten d th e stabilizat ion in pat ien ts w ith large body w eigh t . In our experien ce, th e in st rum en tat ion of th e th ree low est segm en t s p rovid es th e ap prop riate st abilit y in m ost cases, bu t th e p at ien t’s sp eci c requ irem en t s m ay en t ail lengt h en ing th e p osterior st abilizat ion u p to th e th oracic spin e. For cran ial load sh aring, th e FFR tech n ique can be a good solut ion . In pat ien t s w ith osteop orosis, t h e st an dard m et h od s of augm en t at ion of th e t ran sp edicu lar screw s can be perform ed . Th e n u m ber of iliac screw s required can be u n derest im ated, an d th eir posit ion ing is sim ilarly im p or tan t . Mu rakam i et al13 publish ed in vit ro dat a regarding th e st ress dist ribu t ion in an m GT con st ru ct , dem on st rat ing th at th e iliac screw is on e of th e m ost vu ln erable p oin t s of t h e sp in op elvic st abilizin g con st r u ct becau se m a xim um st ress can be obser ved aroun d it .13 Th e dissip at ion of th e st ress is h igh ly recom -

m en ded by using m ore iliac screw s posit ion ed for bet ter st ress d ist r ibu t ion . Yu et al23 com p ared t h e biom ech an ical p rop er t ies of fou r d i eren t iliac screw tech n iques in vit ro. Th ey fou n d th at bilateral dou ble iliac screw s in th e low er p osit ion opt im ally m an ages t h e local st resses an d p rovid es bet ter com p ressive an d torsion al st i n ess com p ared w it h sin gle iliac or dou ble cran ially posit ion ed screw s. Th e postoperat ive sagit tal spin opelvic align m en t ach ieved by th e xat ion is also an im p ort an t con sid erat ion . Most pat ien ts lose som e m otor act ivit y of th e pelvic an d low er ext rem it y m u scles th at are in n er vated by th e n er ves resected du ring th e en -bloc sacrectom y. If th e p lan e of t h e n er ve resect ion is m ore cran ial, t h e p ostop erat ive bony sagit t al balan ce is of greater im por t an ce. Due to th e im por tant role of th e glu teal m u scles in h ip join t st abilizat ion in th e erect p osit ion , th e loss of bilateral L5 roots presen ts a great challenge for the patient’s reh abilitat ion . If th e bony st abilizat ion xes th e n ew spin opelvic jun ct ion in an un balan ced con dit ion , th e pat ien t can n ot w alk or stan d w ith ou t extern al su pp or t . Pat ien t s w ith good p ostoperat ive spin opelvic balan ce can learn to w alk w ith ou t cru tch es w ith in 6 m on th s. In ou r clin ical experien ce on ly elderly pat ien ts n eed aid in w alking or st an ding; you nger p at ien t s are stabilized in good (or tolerable) balan ce and d o n ot n eed any sup port (Fig. 13.5). Du e to th e in creasing st resses at th e bon e– m et al in terfaces, an un balan ced postoperat ive spin opelvic align m ent increases the risk of early im plan t failure (i.e., rod or screw breakage). It is often di cu lt to predict preoperat ively th e ap p rop r iate p osit ion in g of t h e lu m bar sp in e to t h e p elvis, an d it is often d i cu lt d u r in g su rger y, too. In th e fu t ure, th is aspect of th e su rger y w ill be m ore easily m an aged u sin g advan ced t h ree-d im en sion al recon st r u ct ion com put at ion al tools com bin ed w ith n ew in t raoperat ive n avigat ion al tech n iqu es. With th e con t in uing developm en t of tech n ology, spi n opelvic recon st ru ct ion w ill be design ed an d tested in virtual sim ulation using m odern n ite elem en t m odels. Th ese advan ces w ill dem on st rate th e opt im al surgical procedure to apply to th e speci c patient, and w ill sim ulate the pat ien t’s long term m ech an obiological funct ion -

153

154

Chapter 13

a

b

c

Fig. 13.5a–c Frontal (a) and sagit tal (b) alignment 2 months after spinopelvic  xation with the closed loop  technique. (c) EOS reconstruction study of the case represented in Fig. 13.2.

ing, h igh ligh t ing th e p oten t ial bony fu sion or th e risk of im plan t loosen ing. On e of th e key factors in th e success of spin opelvic bony recon st r uct ion is th e fusion process. Delayed fu sion or p seu dar th rosis en t ails th e r isk of failu re. Fat igu e fract u res can occu r in any st age. Becau se t h e on cological outcom e of th ese pat ien t s is m ore favorable, th ey requ ire a st able sp in opelvic ju n ct ion . In our clin ical experien ce th e use of m orselized

autologous bon e graft w ith n on rigid stabilizat ion (closed loop tech n iqu e) p rovides an ap prop riate solu t ion . Th e bony bridge bet w een th e last lu m bar ver tebral body an d th e posterior iliac crests develops at the proper site of the load transfer. Bet ween the lum bar segm ents (to decrease th e su rgical m orbidit y), w e p erform a posterior (in terlam in ar) or p osterolateral (in ter t ran sverse) tech n ique in stead of th e in terbody procedures.

Spinopelvic Reconstruction/Fixation and Fusion

■ Chapter Summary Fixat ion an d bony recon st ru ct ion follow ing sacral t um or resect ion s plays an im portan t role in th e early an d safe m obilizat ion of th e pat ien t after su rger y, decreasing th e risk of perioperat ive com p licat ion s su ch as th rom boem bolism , p n eu m on ia, an d pressu re sores. St rong an d safe bony fu sion develop ing bet w een th e lu m bar sp in e an d th e p elvis p rovides a good base for st an ding an d w alking, en abling bet ter fu n ct ion al resu lt s an d qu alit y of life for th is ver y special grou p of p at ien ts. Spinopelvic stabilization is a dem anding and di cu lt p rocedu re to perform , requ iring su rgical exp erien ce, p rop er su rgical p lan n ing, an d appropriate preparat ion of th e pat ien t . Exten sive kn ow ledge of sp in op elvic biom ech an ics, surgical an atom y, spinal stabilizat ion, and bone graft ing tech n iqu es is essen t ial. Th ere is n o gold st an dard in t h is eld , alth ough t h ere is an in creasing n u m ber of biom echanical studies and clinical papers published in th e literat ure. Th e lack of long-term dat a on

fu sion rates an d fu n ct ion al resu lt s en cou rages research ers to collect an d an alyze clin ical an d exp erim en tal data from m u lt icen ter sources. With th e developm en t of th e com puterized three dim ensional tools, a better understanding can be ach ieved of th e biom ech an ical issu es of th is su rgical tech n iqu e an d of th e speci c p atient requirem ents during th e plan ning process of th ese ch allenging su rgeries.

Pearls Spinopelvic bony fusion can provide a good longterm  functional outcome after total sacrectomy. The role of biom echanics and proper sagit tal balance cannot be underestimated. Pitfalls There is no gold-standard spinopelvic stabilization  technique.  This surgery should not be perform ed without  su cient preoperative planning.

Refere nces Five Must-Read Refe rences 1. Ibrah im A, Crockard A, An ton iet t i P, et al. Does spin al su rger y im p rove t h e qu alit y of life for t h ose w it h ext radural (spin al) osseou s m et ast ases? An intern at ion al m ult icen ter prospect ive obser vat ion al st udy of 223 pat ien t s. Invited subm ission from th e Join t Sect ion Meet ing on Disorders of th e Spin e an d Periph eral Ner ves, March 2007. J Neurosurg Spin e 2008;8:271–278 PubMed 2. Kelley SP, Ash ford RU, Rao AS, Dickson RA. Prim ar y bon e t um ou rs of th e spin e: a 42-year su r vey from th e Leeds Region al Bon e Tum our Regist r y. Eur Spin e J 2007;16:405–409 Pu bMed 3. Mukh erjee D, Ch aich an a KL, Gokaslan ZL, Aaron son O, Ch eng JS, McGir t MJ. Su r vival of p at ien t s w ith m align ant prim ar y osseous spin al n eoplasm s: resu lt s from th e Sur veillan ce, Epidem iology, an d En d Resu lt s (SEER) dat abase from 1973 to 2003. J Neurosu rg Spin e 2011;14:143–150 PubMed 4. Guo Y, Yadav R. Im proving fun ct ion after tot al sacrectom y by u sing a lu m bar-sacral corset . Am J Phys Med Reh abil 2002;81:72–76 PubMed 5. Ruggieri P, Angelin i A, Ussia G, Mon t alt i M, Mercuri M. Su rgical m argin s an d local con t rol in resect ion of sacral ch ordom as. Clin Or th op Relat Res 2010; 468:2939–2947 Pu bMed

6. Clarke MJ, Dasen brock H, Bydon A, et al. Posterior only approach for en bloc sacrectom y: clinical out com es in 36 con secu t ive pat ient s. Neurosu rger y 2012;71:357–364, d iscu ssion 364 Pu bMed 7. Dickey ID, Hugate RR, Fu ch s B, Yaszem ski MJ, Sim FH. Recon st ruct ion after tot al sacrectom y. Clin Or th op Relat Res 2005;438:42–50 8. Fou rn ey DR, Rh in es LD, Hen t sch el SJ, et al. En bloc resect ion of prim ar y sacral t um ors: classi cat ion of su rgical approach es an d outcom e. J Neurosurg Spine 2005;3:111–122 PubMed 9. Kaw ah ara N, Mu rakam i H, Yosh ida A, Sakam oto J, Oda J, Tom it a K. Recon st ruct ion after tot al sacrectom y using a n ew in st rum en t at ion tech n ique: a biom ech an ical com parison . Spin e 2003;28:1567–1572 PubMed 10. Kelly BP, Sh en FH, Sch w ab JS, Arlet V, Diangelo DJ. Biom ech an ical test ing of a n ovel four-rod tech n ique for lum bo-pelvic reconst ruct ion.Spine 2008;33:E400– E406 Pu bMed 11. Li G, Fu D, Ch en K, et al. Surgical st rategy for th e m an agem en t of sacral gian t cell t u m ors: a 32 case series. Spin e J 2012;12:484–491 Pu bMed 12. Min dea SA, Ch in th aku nt a S, Moldavsky M, Gudipally M, Kh alil S. Biom ech an ical com parison of spi-

155

156

Chapter 13 nopelvic recon st ruct ion techn iques in th e set t ing of total sacrectomy. Spine 2012;37:E1622–E1627 PubMed 13. Mu rakam i H, Kaw ah ara N, Tom ita K, Sakam oto J, Oda J. Biom ech an ical evalu at ion of recon st r u cted lu m bosacral spin e after tot al sacrectom y. J Or th op Sci 2002;7:658–664 PubMed 14. Quraishi NA, Wolin sky JP, Bydon A, With am T, Gokaslan ZL. Gian t d est ru ct ive m yxop ap illar y ep en dym om as of th e sacrum . J Neurosurg Spin e 2010;12: 154–159 Pu bMed 15. Varga PP, Bors I, Lazar y A. Sacral t um ors an d m an agem en t . Orth op Clin Nor th Am 2009;40:105–123, vii PubMed 16. Zhu R, Cheng LM, Yu Y, Zander T, Chen B, Rohlm ann A. Com p ar ison of fou r recon st r u ct ion m et h od s after tot al sacrectom y: a n ite elem en t st u dy. Clin Biom ech (Bristol, Avon ) 2012;27:771–776 Pu bMed 17. Gun terberg B, Rom an us B, Stener B. Pelvic st rength after m ajor am p u t at ion of th e sacru m . An exp erim en tal st udy. Act a Or th op Scan d 1976;47:635–642 PubMed 18. Hugate RR Jr, Dickey ID, Ph im olsarn t i R, Yaszem ski MJ, Sim FH. Mech an ical e ect s of p ar t ial sacrectom y: w hen is reconstruct ion necessar y? Clin Orthop Relat Res 2006;450:82–88 Pu bMed 19. Yu B, Zh eng Z, Zh uang X, et al. Biom ech an ical e ect s of t ran sverse par t ial sacrectom y on th e sacroiliac joint s: an in vit ro h um an cadaveric invest igat ion of th e borderlin e of sacroiliac joint in st abilit y. Spin e 2009;34:1370–1375 Pu bMed

20. Got tfried ON, Om eis I, Meh t a VA, Solakoglu C, Gokaslan ZL, Wolin sky JP. Sacral t um or resect ion an d th e im p act on pelvic in cid en ce. J Neu rosu rg Sp in e 2011;14:78–84 PubMed 21. Allen BL Jr, Fergu son RL. Th e Galveston tech n ique of pelvic xation w ith L rod instrum entation of the spine. Spin e 1984;9:388–394 Pu bMed 22. Yu BS, Zh uang XM, Li ZM, et al. Biom ech an ical e ect s of the extent of sacrectomy on the stabilit y of lum boiliac reconstruction using iliac screw techniques: W h at level of sacrectom y requires the bilateral dual iliac screw techn ique? Clin Biom ech (Bristol, Avon ) 2010; 25:867–872 Pu bMed 23. Yu BS, Zh uang XM, Zh eng ZM, Li ZM, Wang TP, Lu W W. Biom ech an ical advan t ages of dual over sin gle iliac screw s in lum bo iliac xat ion con st ruct . Eur Spin e J 2010;19:1121–1128 Pu bMed 24. Sh en FH, Harper M, Foster WC, Marks I, Arlet V. A n ovel “four-rod tech n ique” for lum bo-pelvic recon st ru ct ion : th eor y an d tech n ical con siderat ion s. Sp in e 2006;31:1395–1401 Pu bMed 25. Ch ang DW, Friel MT, Youssef AA. Reconstructive strategies in soft tissue recon struction after resection of spinal neoplasm s. Spine 2007;32:1101–1106 PubMed 26. Gallia GL, Haqu e R, Garon zik I, et al. Sp in al p elvic recon st r uct ion after tot al sacrectom y for en bloc resect ion of a gian t sacral ch ordom a. Tech n ical n ote. J Neurosurg Spin e 2005;3:501–506 Pu bMed 27. Rü edi TP, Buckley RE, Moran CG. AO Prin cip les of Fract ure Man agem en t . New York: Thiem e, 2007

14 Structural Graft Selection Y. Raja Rampersaud

■ Introduction Resect ion of p r im ar y sp in al or p arasp in al t u m ors w it h d irect sp in e invasion t yp ically requ ires com p lex an d at t im es in n ovat ive recon st r u ct ion tech n iqu es. Fu r t h er m ore, for p r im ar y t u m ors for w h ich cu rat ive in ten t is t yp ically t h e m ain object ive, long-ter m d u rabilit y is p aram ou n t to recon st r u ct ion goals. Recon st r u ct ion of sin gle or m u lt ip le ver tebral body en -bloc resect ion s involving m ore th an on e th ird to on e h alf of th e ver tebra(e) an d un ilateral or bilateral p osterior colu m n resect ion t ypically involve extensive instrum entation and st r u ct u ral graft s or graft s to ll th e osseou s spin al defect(s). Speci c to st ru ct ural graft selection, the goal is to provide durable structural su p por t across th e defect(s) an d ach ieve biological in tegrat ion w ith th e h ost bon e (i.e., fu sion ). Th is ch apter discu sses factors th at sh ou ld be con sidered for st ruct ural graft select ion an d th e opt ion s available for recon st r u ct in g sp in al d efect s follow in g p r im ar y t u m or resect ion . Du e to th e rarit y of p rim ar y sp in al t u m ors, m ost relat ively large case series describing recon st r u ct ion an d graft select ion tech n iqu es are su rgeon /cen ter speci c, u sing a variet y of graft select ion an d su rgical tech n iqu es th at are often u n iqu e to each pat ien t an d an atom ic defect requ ir in g recon st r u ct ion .1–3 Th e m ajor it y of repor t s u n derst an dably focus on th e on cological ou tcom e an d tech n ical asp ect s of each case;

h ow ever, ver y few provide any long-term outcom e in form at ion speci c to th e graft select ion or recon st ruct ion du rabilit y.1,2 In addit ion , due to clin ical and technological advancem ents, the reconstruction techniques and graft choices h ave evolved w it h in d i eren t cen ters t h at p er form a h igh er volum e of th ese cases.1–3 Con sequ en tly, th ere is gross variabilit y in graft select ion an d recon st ru ct ion tech n iqu es from su rgeon to su rgeon an d from cen ter to cen ter. Also, th ere are n um erou s case rep or t s describ ing in n ovat ive an d ver y com p lex recon st ru ct ion for ver y large t u m ors th at pu sh th e lim it s of resect ion an d recon st r uct ion exper t ise. Alth ough cru cial to th e u n derst an ding of rare con d it ion s or procedu res, th ese case report s do n ot provide us w ith gen eralizable in form at ion regarding structural graft selection and are often lim ited by relat ively sh or t-term follow -u p.

■ Survey of Practitioners To p rovid e a con tem p orar y an d broad rep resen t at ion of d ecision m aking in t h ese com p lex recon st ruct ion s, I an d m y colleagues C.P. DiPaola an d C.G. Fish er recen tly con du cted an in tern at ion al Web -based su r vey to elicit exper t op in ion regarding decision m aking abou t th e t ype of bon e graft ing ch oice for recon st r u ct ion follow ing resect ion of prim ar y sp in al t u -

158

Chapter 14 m ors. Thirty-one (61%) of the 51 m em bers of AO Sp in e Nor th Am erica–Sp in e Net On cology an d AO In tern at ion al On cology Kn ow ledge For um resp on ded. Th e resp on den ts w ere equ ally divided bet ween orthopedic surgeons and neurosu rgeon s. Th e m ajorit y of resp on den ts (71%) h ave been in clin ical pract ice for over 10 years. Th e average n u m ber of p rim ar y spin e t u m or cases p erform ed per year by th e resp on den t s in th is grou p w as 16, w ith a range of 2 to 100. Com parat ively, th e average n u m ber of m etast at ic cases p er year for th is grou p w as 35, w ith a range of 7 to 150. Th e respon se for each sur vey quest ion varied from 28 to 31 / 31 resp on den ts. For p rim ar y t u m ors, th e m ajorit y (93%) of resp on den ts attem pt to gen erate a fu sion , w ith 63% of th em n ot ing th at th ey use bon e graft or bon e graft su bst it u tes ever y t im e an d 30% u sing th em alm ost ever y t im e. Th e preferred ch oice of bon e graft m aterial was equally split bet ween m orselized iliac crest (39%) an d allograft (39%). Th e use of local bone alone w as the preferred choice of on ly 16% of respon den ts. Ceram ics (e.g., cal-

ciu m p h osph ates) w ere n ot p referred by any respon den t , an d o -label u se of bon e m orp h ogen et ic protein (BMP) w as preferred by on ly t w o (6.5%) respondents. The m ajorit y of responden t s t yp ically u sed com bin at ion s of m orselized iliac crest , allograft , an d local bon e. W h en an terior st r u ct u ral su p p or t w as required, st ru ct ural allograft w as preferred by 47% of respon den t s, follow ed closely by a p refabricated prosth et ic replacem en t device (i.e., a “cage”) preferred by 40%. Th e use of vascularized bu la or rib w as p referred by on ly th ree (10%) respon den ts, an d polym ethylm eth acr ylate (PMMA) w as p referred by n o on e (Table 14.1). In th e even t of a 360-degree spon dylectomy, the respondents overw helm ingly reported rout in e an terior colu m n an d bilateral p osterior colu m n recon st ru ct ion , w ith 42% doing so ever y t im e an d 42% doing so alm ost ever y t im e. An terior colu m n recon st ru ct ion on ly an d prim ar y spin al sh or ten ing (i.e., h ost bon e-on bon e, for a single-level resect ion ) w ere occasion ally/som et im es perform ed by 30%an d 19% of resp on den ts, respect ively.

Table 14.1 Structural Graft Selection: Summary of International Survey

Graft Choice*

Preferred by Respondents (%)

Structural iliac crest

3.3%

Structural allograft

46.7%

Prefabricated prosthetic replacem ent (i.e., cage)

40%

Bone cement

Vascularized bula or rib

0%

10%

Author/ Editor Consensus Will consider for 1- to 2-level cervical reconstruction, or 1-level thoracic, particularly with associated radiation. Donor morbidit y must be discussed with the patient. Preferred if available. Provides excellent integration, enables accurate tumor surveillance and lower cost than prefabricated prosthesis, but has lim ited availabilit y and is susceptible to fracture if unprotected by instrumentation. Preferred. Greater exibilit y regarding sizes, shape (can also get custom or modular implants), and availabilit y compared with allograft. Metal cages create imaging artifact, which may delay detection of local recurrence. Will consider in patients with short life-expectancy (palliative/intralesional procedure) and/or very low likelihood of fusion (e.g., high-dose postoperative radiation will be required). Will utilize when there is a high risk of nonunion, in a revision case, or to supplement or provide soft tissue reconstruction (i.e., myo-osseous or osteocutaneous ap).

*Survey question: “Considering only prim ary spine tum or cases, when anterior reconstruction with structural support is necessary, what is your preference for structural support?”

Structural Graft Selection

■ General Factors to Consider

in Graft Selection As n oted above, th ere is often sign i can t caseto-case variabilit y even from th e sam e surgeon or cen ter. Th is is due to a variet y of key factors th at n eed to be con sidered w h en deciding on an opt im al graft select ion . Th ese factors can be broadly categorized in to pat ien t , local, t um or, an d h ealth system factors. Pat ien t factors in clu de, bu t are n ot lim ited to, age, m edical com orbidit y, an d life exp ect an cy. Local factors can in clude th e an atom ic locat ion of th e lesion (discussed below ), th e circum feren t ial an d lon git udin al degree of th e defect , th e bon e qu alit y, th e adequ acy of soft t issu e coverage, an d p rean d postoperat ive radiat ion . In addit ion , w h en con sidering au tologou s graft s su ch as an iliac crest or a vascularized osseou s or com posite graft, the m orbidity and functional consequences from th e don or site also n eed to be con sidered an d w eigh ed again st altern at ives to th e local recon st ru ct ion n eeds. Tu m or factors in clu de h istology, perioperat ive ch em oth erapy, abilit y to ach ieve n egat ive m argin s, an d th e n eed for

lifet im e local su r veillan ce. Health system factors in clude, but are n ot lim ited to, available m ultiteam expertise (e.g., for m icrovascular aps or resect ion an d recon st r uct ion of associated vascu lar or visceral involvem en t), graft availabilit y, an d cost . In our su r vey w e also asked th e exper ts to ran k, in order of im port an ce, th e factors th at in uen ce th eir decision regarding graft select ion . Th e top ve factors ran ked as ext rem ely im port an t in graft select ion w ere pat ien t’s life exp ectan cy, ach ieving a clear m argin , t u m or h istology, bon e qu alit y, an d adequ acy of soft t issue coverage (Table 14.2). Th e top th ree factors ranked as m oderately im portant w ere graft availabilit y, grafts th at en able local t u m or surveillan ce, an d p re- or postop erat ive rad iat ion th erapy. A detailed discussion of each factor is beyon d th e scope of th is ch apter, but th e follow ing discu ssion provides a brief persp ect ive on th e key issu es for each of th e top factors. A sh or t life expect an cy, n ot ach ieving a clear m argin , an d t u m or h istology t yp ically relate to a h igh er likelih ood of persisten t or recurren t t u m or an d th e probable n eed for ongoing adjunctive treatm ent (e.g., postoperative radiation)

Table 14.2 Factors that In uence Bone Graft Selection: Summary of International Survey Decision Making Factors* Extremely important: Life expectancy Likelihood of achieving a clear margin Tumor histology Bone qualit y Adequacy of soft tissue coverage Moderately important: Graft availabilit y Enable local tumor surveillance Radiation therapy pre- or postoperative

Percent of Respondents 63.3% 51.5% 43.3% 43.3% 43.3% 56.7% 53.3% 46.7%

*Survey question: “Considering only prim ary spine tum or cases, rate the importance of each factor that m ay a ect bone graft selection in prim ary spine tum or reconstructions.” Factors included patient age, tumor histology, likelihood of achieving a clear m argin, life expectancy, anatom ic location, history of preoperative chem otherapy, radiation therapy pre- or postoperative, bone qualit y, adequacy of soft tissue coverage, enable local tum or surveillance, graft cost, and graft availabilit y.

159

160

Chapter 14 or fu t u re su rger y.4 In th ese scen arios, a m ore in er t graft su ch as PMMA th at w ill be resistan t to biological e ect s of local t reat m en t an d t u m or invasion sh ou ld be con sid ered . Poor bon e qu alit y m ay n ecessitate con sid erat ion of tech n iqu es th at u t ilize PMMA to augm en t or in corp orate adjacen t segm en t s in to th e an terior st r uct ural graft an d in crease th e required n u m ber of posterior xat ion levels.3 Adequ ate soft t issu e coverage is requ ired to m ain t ain or en able local vascularizat ion an d healing as w ell as redu ce th e likelih ood of w oun d breakd ow n an d in fect ion . In th is scen ario a vascular graft w ould be given h igh con siderat ion if adequate coverage w ere n ot ach ievable by oth er m ean s.2 Graft availabilit y an d n eed for adequate t um or sur veillan ce are self-eviden t . Th e n egat ive e ect s of rad iat ion on soft t issu e h ealing are w ell kn ow n . How ever, preop erat ive rad iat ion alon e m ay n ot sign i can tly adversely a ect local bony h ealing, bu t it cer t ain ly resu lt s in local osteopen ia.5 Acu te p ostoperat ive rad iat ion (p lan n ed or d u e to in t raop erat ive t u m or con t am in at ion or u n exp ected p osit ive m argin ) does h ave a p rofou n d e ect on bony h ealing of an terior in terbody st rut graft s.5 In t h e scen ar io w h ere t h e n eed for acu te p ost op erat ive radiat ion is likely or is im posed due to p osit ive m argin s, th e u se a vascu lar graft or PMMA, depen ding on oth er factors p reviously d iscu ssed, sh ou ld be con sidered .

■ Speci c Considerations

Based On Anatomic Site As n oted above, th e recom m en d ed st r u ct u ral graft ch oices (Table 14.1) in m ost region s of the spin e are t ypically allograft or prefabricated cages (p acked w it h allo - or au tograft bon e) th at are ap p rop r iately sized w it h resp ect to d iam eter (preferably con t act ing t h e ap op hyseal rim of th e en dp late) an d length . St r u ct u ral bon e allograft h as th e advan tage of direct biological fu sion (Fig. 14.1) w h ere graft u n ion w it h t h e h ost is p ossible.6 Allograft also en ables un im peded im aging for assessm en t of fu sion an d t u m or su r veillan ce. For st r u ct u ral allograft s, t h e opt ion s are as follow : for t h e

cer vical spin e—radiu s, bular, or u ln a; for th e th oracic sp in e—t ibia or h u m eru s (u pp er th oracic); an d for th e lum bar sp in e—t ibia or fem ur. Perh aps th e greatest cu rren t lim itat ion of st ruct ural allograft pertains to its relative availabilit y. Du e to t h e p ossible r isk of d isease t ran sm ission as a resu lt of p oor region al p rocurem en t an d sterilizat ion processes or cu lt u ral beliefs, allograft is n ot available in m any cou n t r ies. Sp eci c to t h e recon st r u ct ion of segm en tal defects, th ere is an associated risk of fract u re, p ar t icu larly for irradiated or freezedried allograft .6 How ever, due to th e load sh aring an d torsional stabilit y th at is o ered by spin al in st r um en tat ion , th is is an un com m on occurrence in spinal cases. Com paratively, prefabricated prostheses often referred to as “cages” o er th e advan tage of greater availabilit y an d m ore size an d sh ape opt ion s. Fu r th erm ore, for com plex defects w ith transitional anatom y such as at th e cer vicoth oracic jun ct ion , con toured or cu stom opt ion s are available. Most prefabricated prostheses do not directly in tegrate w ith the h ost , an d con sequen t ly n eed to be packed w ith m orselized bon e (auto- or allograft) to en able fusion . For prim ar y t um or recon st ruct ion , m etallic cages h ave th e dist in ct disadvan t age of m etal ar t ifact on m agn et ic reson an ce im aging (MRI) an d to a lesser exten t on com p u ted tom ograp hy (CT), w h ich m ay d elay t h e d etect ion of local d isease recu r ren ce on t u m or su r veillan ce (t h e clin ical sequ elae of w h ich can be d ebated). Fu r t h er m ore, t h e cost of cages, p ar t icu larly con tou red or cu stom devices, m ay be p roh ibit ive in cer t ain h ealt h care system s. Recon st ru ct ion of com p lete u n ior bilateral posterior colum n defect s is also variable from case to case, but appropriately sized load-sh aring st r ut graft(s), t ypically alloor autograft , are preferable. At our cen ter, w e h ave h ad excellen t su ccess w ith using rib graft s for th is pu rpose.7 In cases w h ere th ere is a h igh risk of failu re (revision s or postoperat ive radiat ion ), a vascularized rib can be con sidered. Ju n ct ion al region s of t h e sp in e t yp ically p resen t w it h fu r t h er u n iqu e an atom ic an d m echanical challenges that w arran t discussion. Du e to t h e frequ en t n eed for m assive resect ion s, recon st ru ction of lum bar-pelvic defects t ypically represents th e m ost variable an d dif-

Structural Graft Selection

161

Planned Resection Margin

L2

L2

a

b

L2

c

d

Fig. 14.1a–d Coronal magnetic resonance imaging (MRI) (a) and axial computed tomography (CT) of a large retroperitoneal primitive neuroectodermal tumor (PNET) in a 46-year-old man. (b) The planned resection is demonstrated on the axial CT. The patient is 6 years postresection, working on his farm, and is disease free at last assessment. (c) A

5-year axial CT dem onstrating the allograft femur and aortic stent anteriorly and the posterolateral bony elements that were retained to provide vas0 cularized bone in the posterior column. (d) Sagit tal CT at 5 years demonstrating excellent incorporation at the host–allograft junctions.

cu lt recon st r u ct ion s, w ith th e greatest risk of com p licat ion . Th is top ic is d iscu ssed in Ch ap ter 13. Due to th e st raigh t alignm en t , st ru ct u ral graft select ion at th e th oracolum bar ju n ct ion does n ot p resen t any addit ion al speci c issu es beyon d th ose already discussed. On th e oth er hand, the transition al anatom y (rapid change in size an d con tou r) of th e cer vicoth oracic ju n c-

t ion (CTJ) often requ ires th e u se of con tou red cages or unique allograft step -cuts, particularly for m u lt ilevel resect ion th at cross th e CTJ. Th e occipitocervical junction (OCJ) is associated w ith unique access and recon struction challenges for m arginal or w ide en-bloc resections.8 This is par t icu larly so for resect ion s th at require h igh anterior access and the possible resection of the

162

Chapter 14 posterior p h ar yngeal w all, w h erein adequ ate soft t issu e coverage is crit ical to avoid in fect ion an d en able osseous h ealing. In th e lat ter scen ario, a m yo-osseous vascularized bu lar st rut graft (i.e., w ith at tach ed m u scle an d fascia) is recom m en d ed to p rovide st r u ct u ral su pp or t (clivus to cer vical spin e) an d posterior ph ar yn -

geal w all recon st r uct ion (Fig. 14.2). Altern at ively, if th e posterior p h ar yngeal w all is in t act , t h en p oster ior recon st r u ct ion w it h cages or au tograp h st r u t s from t h e occip it al con dyles to th e cer vical lateral m asses can be u t ilized du e to th e sagit tal ver tebral axis being p osterior in th e upper cer vical spin e.9 Th e term in al

a

b Solid fusion, with graft -host hyypert rophy at 3years post -op

c

d

Fig. 14.2a–d Axial (a) and sagit tal (b) MRI scans of a C1-C4 chordoma in a 59-year-old mam. The chordoma entailed bilateral vertebral artery involvem ent at C2. Management involved a m ultistage en-bloc R0 resection that included the posterior pharyngeal wall, unilateral vertebral artery bypass, myo-osseous vascularized bular strut graft, posterior central rib graft, and circumferential

instrumentation. Patient is disease free at 4 years with persistent dysphonia and dysphagia. (c) Postoperative sagit tal CT scan at 3 years demonstrating solid integration of both grafts, with hypertrophy at the term inal ends of the vascularized bular graft. (d) Lateral radiograph at 3 years dem onstrating the reconstruction.

Structural Graft Selection en ds of th e cages can be com p ressed in to th e con dyle an d cer vical lateral m ass to p rovid e im m ediate stabilit y of th e graft .

■ Reconstruction Results Despite th e ext rem ely h igh -risk environ m en t for osseou s h ealing, n on un ion follow ing en bloc resect ion is less com m on th an on e w ould exp ect . In t h e largest p u blish ed ser ies of en bloc resect ion s (n = 134, in clu d in g n = 90 for p r im ar y t u m ors), at a m ed ian follow -u p of 47 m on th s Borian i et al1 rep or ted 7% in st ru m en t at ion failu re u sin g a p oster ior p ed icle screw xat ion con n ected w it h a m od u lar an terior carbon ber system p acked w it h au togen ou s graft or bon e subst it utes. Th ey at t ributed th e failu re to sh or t segm en t p osterior xat ion . In a m u ch sm aller clin ical series, Matsum oto et al10 reported on late instrum entation (anterior m esh cage an d p edicle screw xat ion ) failu res follow ing en -bloc resect ion s in 15 p at ien t s; seven of th e resect ion s w ere for prim ar y t um ors. At a m ean follow -u p of 41.5 m on th s, an 40%overall in st r u m en tat ion failu re rate (rod or cage fract u re) w as rep or ted, w ith m ost failu res occu rring after 2 years. In th e seven p rim ar y t u m or tot al en -bloc spon dylectom ies (TESs), t w o failures (29%) occu rred . Cage su bsiden ce (≥ 5m m ), preop erat ive radiat ion , an d th e n u m ber of levels of p osterior in st ru m en tat ion (≤ 4 levels) w ere reported to be sign i can tly associated w ith late in st r u m en t at ion failure. Th e pitfall of sh or t-segm en t posterior xat ion is su p p or ted by th e biom ech an ical st u dy from Disch et al.11 Using a on e-level TES m odel, th e au th ors dem on st rated th at it is th e n u m ber of levels of p osterior xat ion th at predom in an tly determ in es st abilit y even w ith an terior xat ion . An terior-alon e or sh or t-segm en t (on e level above an d below ) circu m feren t ial xat ion is n ot recom m en ded follow ing TES. For a on e-level TES, posterior pedicle screw xat ion t w o adjacen t segm en t s or m ore above an d below w ith an terior colum n st ru ct ural suppor t is recom m en ded. In another interm ediate-term (m ean follow up of 41.5 m on ths) prosp ect ive case series of

26 en -bloc resect ion s for p r im ar y t u m ors, Fish er et al 2 rep or ted a revision rate of 10%, w it h t w o of 20 su r vivin g p at ien t s requ ir in g revision in str um en tat ion for n on un ion (on e each in th e th oracic an d th e lu m bar sp in es). Th e au th ors u sed m u lt ilevel p osterior xat ion in all cases w ith a variet y of an terior st r uct ural graft ch oices. In a ser ies of h igh ly com p lex en -bloc resect ion s for cer vical ch ord om a, Hsieh et al8 rep or ted late n on u n ion an d in st r u m en t at ion failure in th ree of th e ve cases recon st r ucted w ith in st ru m en ted an terior cages an d post rod-screw xat ion . Th e m ean clin ical follow -u p w as 4.5 years. Un for t u n ately, t h e au t h ors d id n ot com m en t on t h e sp eci c m od e(s) of failu re in th ese cases. In m y p erson al series of t w o su ch cases, on e requ iring C1- C3 en -bloc resect ion follow ing an in com plete in t ralesion resect ion perform ed at an oth er cen ter, an d th e oth er a C1- C4 en -bloc resect ion , a vascu larized m yoosseous bu lar graft h as successfully in corporated in both cases (Fig. 14.2). Th ese repor t s dem on st rate th e crit ical im por t an ce of st ru ct u ral graft select ion an d reconstruction techniques that incorporate sound biom ech an ical prin ciples an d long-term goals to ach ieve fusion . Revision surgeries in th ese cases w ith already sign i can tly com p rom ised h ost t issue can be daun t ing, an d th e best st rategy is opt im izat ion of t h e in d ex p roced u re. Sou n d kn ow ledge of w h at h as been t ried an d resulted in failu re or success is p aram oun t to sou n d decision m aking for st ru ct u ral graft select ion an d recon st ru ct ion follow ing en -bloc resect ion for prim ar y t u m ors.

■ Chapter Summary Alt h ough t h e literat u re on sp in al in st r u m en t at ion an d fu sion tech n iqu es is exten sive, t h e literature pertaining to reconstruction of struct u ral defect s follow ing en -bloc p rim ar y t u m or resect ion of th e sp in e is relat ively lim ited. Du e to th e rarit y of prim ar y spin al t u m ors, case series describing recon st ru ct ion an d graft select ion tech n iqu es are su rgeon /cen ter an d pat ien t speci c, an d th ey focu s on th e on cological an d

163

164

Chapter 14 tech n ical aspects of th e cases. Ver y few repor ts p rovid e long-ter m dat a on t h e d u rabilit y of th e graft select ion an d recon st r uct ion . Fur th erm ore, th ere are n u m erou s case rep or t s t h at d escr ibe ver y sp eci c an d com p lex recon st r u ct ion s t yp ically for ver y large m u lt ilevel resect ion s. Con sequ en tly, n o case-sp eci c (i.e., sign i can t h eterogen eit y w ill alw ays exist bet w een cases) evid en ce-based best p ract ices can be clearly st ated on th is topic. Th e u lt im ate goal of spinal fusion follow ing en-bloc resection of a prim ar y tum or represents a signi cant ch allenge du e to a variet y of p oten t ially u n favorable factors th at n egat ively a ect osseous h ealing; th ese in clude gross in stabilit y of th e spin e, poor vascularit y due to w ide dissect ion of th e surroun ding soft t issues, an d a h ost ile local environ m en t secon dar y to radiat ion or ch em oth erapy.

Pearls The overall goal of structural graft selection should be to provide durable structural support across the spinal defect(s) and enable biological integra-

tion (directly or indirectly) with the host bone (i.e., fusion). General key factors in structural graft selection fall into four broad categories: patient, local, tumor, and health system. To successfully achieve access, soft tissue coverage, biom echanically sound xation and longterm structural graft integration, unique technical requirem ents based on the anatomic location of the defect, particularly at junctional segments, need to be carefully considered and planned for each case. Pitfalls Thinking of only the short-term rather than the long-term goals will almost always be a regret table decision in these cases. Not all centers have the necessary m ultidisciplinary expertise and equipm ent required to achieve the end goals of reconstruction in these cases. Ensure that all goals will be m et and contingency plans are identi ed and executable before proceeding with the index procedure. Regardless of appropriate graft selection, inadequate xation will lead to a poor outcom e. Avoid anterior-alone or anterior and short-segment posterior (i.e., one level above and below) xation in these t ypes of cases.

Refere nces Five Must-Read Refe rences 1. Borian i S, Ban diera S, Don th inen i R, et al. Morbidit y of en bloc resect ion s in th e spin e. Eur Spin e J 2010; 19:231–241 Pu bMed 2. Fisher CG, Keyn an O, Boyd MC, Dvorak MF. Th e surgical m an agem ent of prim ar y t u m ors of th e spin e: in it ial result s of an ongoing prospect ive coh or t st udy. Spin e 2005;30:1899–1908 PubMed 3. Fourn ey DR, Abi-Said D, Rh in es LD, et al. Sim ultan eous an terior-posterior approach to the th oracic an d lum bar spin e for th e radical resect ion of t um ors follow ed by recon st ruct ion an d st abilizat ion . J Neurosurg 2001;94(2, Suppl):232–244 PubMed 4. Fish er CG, Saravanja DD, Dvorak MF, et al. Su rgical m an agem en t of p r im ar y bon e t u m ors of t h e sp in e: validat ion of an ap p roach to en h an ce cu re an d red u ce local recu r ren ce. Sp in e 2011;36:830– 836 Pu bMed 5. Em er y SE, Brazin ski MS, Koka A, Ben su san JS, Steven son S. Th e biological an d biom ech an ical e ect s of irrad iat ion on an ter ior sp in al bon e graft s in a can in e m odel. J Bon e Join t Su rg Am 1994;76:540–548 PubMed

6. Delloye C, Cornu O, Druez V, Barbier O. Bone allografts: W hat they can o er and w hat they cannot. J Bone Join t Su rg Br 2007;89:574–579 Review PubMed 7. Lew is SJ, Ku lkarn i AG, Ram persaud YR, et al. Posterior colum n recon st r uct ion w ith autologous rib graft after en bloc t u m or excision . Spin e 2012;37:346– 350 PubMed 8. Hsieh PC, Gallia GL, Sciubba DM, et al. En bloc excisions of chordom as in the cervical spine: review of ve con secut ive cases w ith m ore th an 4-year follow -u p. Spin e 2011;36:E1581–E1587 PubMed 9. Scheer JK, Tang J, Eguizabal J, et al. Opt im al recon st ru ct ion tech n iqu e after C-2 cor pectom y an d spon dylectom y: a biom ech an ical an alysis. J Neurosurg Spin e 2010;12:517–524 PubMed 10. Mat su m oto M, Wat an abe K, Tsuji T, et al. Late in st ru m en t at ion failure after tot al en bloc spon dylectom y. J Neurosurg Spin e 2011;15:320–327 PubMed 11. Disch AC, Sch aser KD, Melch er I, Lu zzat i A, Feraboli F, Schm oelz W. En bloc spon dylectom y recon st ruct ion s in a biom ech an ical in -vit ro st u dy. Eur Spin e J 2008; 17:715–725 Pu bMed

15 Wound Closure Techniques Justin M. Sacks and Justin M. Broyles

■ Introduction Radical on cological resect ion is w arran ted for aggressive ben ign t u m ors an d for low -grade an d h igh -grade m align an t t u m ors of th e spin e. Advan ces in n eu rosu rgical tech n iqu e, in st r u m en t at ion for sp in al st abilizat ion , an d adju van t t h erapy allow for a greater n u m ber of su rgical can d idates th an ever before.1 How ever, su rgical w ou n d s t h at con t ain h ardw are an d vit al n eu ral st r u ct u res can h ave d evast at in g con sequ en ces for p at ien t s if t h e w ou n d s becom e exp osed or in fected. Com plex, com posite w oun ds of th e spin e are m an aged w ith im m ediate soft t issue recon st r u ct ion u sin g m u scle or fascial ap s.2 Th is is part icularly in dicated for those pat ien t s iden t i ed as being at h igh risk for w ou n d-h ealing com p licat ion s, su ch as th ose w ith prior operations, infections, potential for planned placem ent of spin al in st r um en tat ion , or m edical com orbidit ies th at predispose to com plicat ion s.3 Th e on cological sp in e su rgeon sh ou ld w ork in t an d em w ith th e p last ic an d recon st r u ct ive su rgeon to opt im ize both th e ablat ive an d th e recon st ru ct ive su rgical proced u re. Recon st r uct ive opt ion s range from p rim ar y closure to local m uscle aps to m ore com plex m et h od s of recon st r u ct ion su ch as m icrovascular free tissue t ransfer. Defects in regions that h ave been exp osed to p rior irradiat ion or su r-

ger y, or h ave a paucit y of local t issue require m ore com p lex t yp es of recon st r u ct ive in terven t ion s. Using a com bin at ion of th ese tech n iqu es, th e goal of th e recon st r u ct ive su rgeon , in tan dem w ith th e n eurosurgical team , is to obliterate dead sp ace an d p rovide coverage of vascu larized soft t issu e over crit ical areas in an e or t to preven t in fect ion an d restore fun ct ion to th ese p at ien t s.

■ Preoperative Evaluation Th e p hysiological stat u s of th e pat ien t m ust be con sidered an d balan ced w ith th e overall recon st r u ct ive plan . Th e overall p rogn osis of th e pat ien t m u st also be taken in to accou n t as w ell. Recon st r u ct ion to im p rove a p at ien t’s qu alit y of life is often con sidered even in a palliat ive scen ario.4

Medical Comorbidities Th e p at ien t’s m edical h istor y sh ou ld be th orough ly review ed to st rat ify th e relat ive risk of in fect ion . Eviden ce of m aln u t rit ion , sm oking, diabetes m ellit u s, an d p erip h eral vascu lar disease sh ou ld be evalu ated, as th ese con dit ion s can h ave deleteriou s e ect s on w ou n d h ealing as w ell as ap su r vival.5

166

Chapter 15

Radiation Therapy Radiation therapy induces tissue injur y through ch anges in th e m icrocircu lat ion of th e defect an d its surrou n ding areas, leading to decreased perfusion an d im paired w ound healing. Indeed, spin al surger y pat ien ts w h o un dergo adjuvan t rad iat ion h ave u p to th ree t im es th e rate of w oun d healing com plication s as com pared w ith th eir n on radiated cou n terpar t s.6 For th is reason , th e recon st r u ct ive su rgeon sh ould be cogn izant of the tim ing, dosage, an d location of any p rior or plan n ed radiat ion . Addit ion ally, th e p last ic an d recon st ru ct ive su rgeon sh ou ld u t ilize t issu e out side th e eld of radiat ion for ap reconstruction. This w ill take the form of a pedicled or free t issue t ran sfer. A variable am ou n t of skin , adip ose, fascia, m u scle, an d bon e can be rot ated or t ran sp lan ted on a vascu lar p edicle com p osed of a speci c ar ter y an d vein .

Timing of Reconstruction Th e t im ing of sp in al soft an d h ard t issu e recon st r u ct ion is m ost often dict ated by th e st at u s of th e t u m or an d th e surgical m argin s. Cer tain pat ien t factors su ch as advan ced age, m u lt ip le com orbidities, or the need for locoregional cont rol w ith adjuvan t radioth erapy m u st also be con sidered in th e t im ing. Prim ar y recon st ru ct ion carries a sign i can tly decreased rate of w ou n d h ealing com plicat ion s an d is preferred to delayed reconstruction.7 Delayed recon struction , alth ough n ot opt im al, is occasion ally u n avoidable for d efects w ith exten sive soft t issu e de cits or in cases of p atien t in stabilit y. If a pat ien t requ ires delayed recon st r u ct ion , a n egat ive p ressu re closu re device is th e preferred tem p or izing m easu re u n t il d e n it ive recon st ru ct ion can be perform ed at a later date. Negat ive pressure closure devices over exposed n eural st ruct ures in th e region of th e spin al cord can create both cerebrosp in al u id leaks an d p ain .

Imaging Preoperat ive im aging w ith com p u ted tom ograp hy (CT) an d/or m agn et ic reson an ce im aging (MRI) sh ould be perform ed to evaluate th e in -

tegrit y of th e surroun ding soft t issu e, vascular an atom y, an d any previou sly p laced h ardw are. Often , local t issu e, su ch as th e p araspin al m u sculat ure, w ill be sough t to obliterate surgical defects; h ow ever, if th is t issu e is n ot available, th e recon st ru ct ive su rgeon w ill h ave to plan in advan ce to u t ilize region al or dist an t t issu e outside of th e zon e of injur y to recon st ruct th e d efect . Kn ow ledge of t h e associated vascu lature helps w ith preoperative planning and w ith th e even t u al in t raoperat ive decision s.

■ Reconstructive Surgical

Tenets Th e core prin cip le u n derlying recon st r u ct ive algorith m s used by plast ic surgeon s is to progress from sim p le to m ore com p lex recon st r u ct ion s on t h e basis of th e sp eci c w ou n d requirem en t s. Th e u n derlying goal for t h e reconstruction is to close a w ound prim arily w ith local t issu e th at w ill be ten sion -free an d obliterate dead sp ace. W h en prim ar y closu re is n ot feasible, local t issue aps of t issue can be used. Local t issu e ap s en able su rgeon s to reconst r u ct soft t issu e d efect s w ith sim ilar t issu e from an adjacent location. “Random ” local aps com prise adjacent skin and subcutaneous tissue an d are based on a subderm al plexus vascu lar supply. By de n it ion , a ran dom ap does n ot h ave a dist in ct , speci c blood su pply. In con t rast , axial-p at tern aps are based on sp eci c blood vessels. Axial aps can be fasciocutaneous (deep m uscle fascia w ith overlying skin), myocutan eou s (m uscle w ith skin ), or osteocut an eou s (bon e w ith overlying skin ); th ese aps en able recon st ru ct ive surgeon s to repair defect s w ith t issue th at is sim ilar to th e resected t issu e. Microvascular free t issue t ran sfer involves h ar vest ing a t issue con st ruct an d its n am ed blood supply from a distan t region of th e body an d placing it in to a defect . Vascu lar an astom osis bet w een th e ap ’s don or vessels an d th e pat ien t’s recip ien t vessels is perform ed u n der m agn i cat ion p rovided by a su rgical m icroscope. Th e decision to u se a p ar t icu lar ap is based on th e requ irem en ts for replacing m issing skin , ad ipose t issu e, fascia, m u scle, or bon e.

Wound Closure Techniques Th e p rim ar y advan t age of m icrovascu lar free t issu e t ran sfer is th at t issu e of a qu alit y sim ilar to that of the resected tissue can be m oved from a rem ote part of the body, thereby enabling op t im al aesth et ic an d fu n ct ion al ou tcom es. Th is also en ables irradiated or in fected t issue to be rem oved an d replaced w ith soft , pliable, an d vascularized t issue. Th e draw backs of free t issu e t ran sfer are related to don or-site m orbidit y an d th e poten t ial for longer operat ive t im es. How ever, w ith carefu l coordin at ion , h ar vests of vascularized osteocut an eous aps (e.g., free bu la ap ) can be perform ed sim u lt an eou sly or in con cer t w ith on cological resect ion s.

■ Anatomic Location and

Axial Pattern Flap Availability Sp in al soft t issu e defect s can be classi ed according to a sch em e suggested by Casas an d Lew is 8 in w h ich w ou n ds are divided in to th e upp er th ird, m iddle th ird, an d low er th ird. Th e upp er th ird of th e sp in e in clu des all cer vical ver tebrae u p to T7. Th e m iddle th ird of th e spin e includes T1 to T12. Th e low er th ird of th e spin e in cludes ver tebrae L1 to S5.8

Upper Third Defect s of th e u p per th ird of th e sp in e, d e n ed by th e sp in ou s p rocesses of C3 t h rough C7, are am ong th e m ost com m on ly en coun tered by th e plast ic su rgeon an d are am en dable to a m u lt it u de of axial pat tern ap s. Th e m ajorit y of th ese defects can be recon st ructed using bilateral paraspin al ap s. Addit ion ally, a u n ilateral t rapezius m uscle ap can be used w hen th e parasp in al m u scles h ave been com prom ised. A lat issim u s d orsi m u scle w ith or w ith ou t a skin padd le can be h ar vested for larger, m ore exten sive defect s.8

Middle Third Th e m idd le th ird of th e spin e is d elin eated by th e sp in ou s processes of C7 to L1. Recon st r u c-

t ion of th is area is sim ilar to th at of th e u pp er th ird, w h ere bilateral p araspin ou s m u scle aps are able to provide coverage for a m ajorit y of defects. If th ese m u scle ap s are u n available du e to t rau m a, irradiat ion , or defect size, th e lat issim u s dorsi m u scle can p rovide du rable coverage eith er as a st an dard, axial pat tern ap or by ut ilizing reverse lat issim us m uscle ap.9

Low er Third Th e bou n daries of th e low er th ird of th e spin e are dictated by th e spin ous processes of L1 to S5. Du e to th e p roxim it y of th is region to th e gluteal an d posterior th igh region s, th ere is a m u lt it u de of p oten t ial ap don or sites available for ap recon st ruct ion . With in th is region , th e p araspin ou s m u scles are robu st an d readily available for coverage of m ost defects. If th ey are un available, oth er region al ap opt ion s are th e lat issim u s dorsi t u rn over m u scle ap , gluteus m uscle ap, an d pedicled ver t ical rect us abd om in is m yocu t an eou s ap (VRAM) based on th e deep in ferior epigast ric vessels. W h en u t ilized in th e low er th ird, th e pedicled VRAM is rou t in ely u sed to obliterate th e act u al sp ace created by a subtot al or tot al sacrectom y. In th ese clin ical sit u at ion s, th e bow el can h ern iate p osteriorly an d becom e en t angled in sp in al h ardw are or develop a bow el obst ru ct ion . In th e set t ing of a tot al sacrectom y, th e free bula bon e can be an astom osed to th e su perior blood su p ply of a p edicled VRAM or th rough a vein graft to th e in tercostal vessels.10

■ Adjuncts to Flap Surgery Negative Pressure Wound Therapy Negat ive p ressu re-assisted closu re can p rovide for tem porar y coverage in soft t issu e spin al defect s w h en de n it ive recon st ruct ion is delayed. W h en u t ilized approp riately, th is device can prom ote n eovascu larizat ion , decrease edem a, an d in crease local gran ulat ion t issue, as w ell as provide con t ract ile force at w ou n d edges.11 Alth ough th is m odalit y can be used to prepare t h e w ou n d bed for d e n it ive recon st r u ct ion w it h soft t issu e ap s in a d elayed fash ion , it

167

168

Chapter 15 can also be u sed to p rom ote h ealing by secon dar y in ten t ion in par t ial-th ickn ess defects.

Internal Tissue Expansion Tissu e expan sion is a process in w h ich an in atable prosth et ic im plan t w ith a silicon e sh ell is u sed to exp an d local an d region al t issu es so th at th ey can even t ually be advan ced in to th e w ou n d in a delayed fash ion . Th e in at able im plan t is in serted at t h e t im e of t um or ext irpat ion or du ring a secon d p rocedu re. At su bsequ en t o ce visit s, salin e is injected th rough an in tegrated or rem ote port to gradu ally expan d th e im p lan t . On ce t h e t issu e h as been su cien tly expan ded, it can be advan ced in to th e defect . Becau se t issu e expan sion takes t im e, the m ethod is not feasible for spinal reconstruct ion th at requ ires im m ediate coverage of spin al in st r u m en t at ion or n eu rovascu lar st ru ct u res. Risks of t issu e exp an sion in clu de in fect ion , ext rusion , an d ru pt u re of th e im plan t .12

External Tissue Expansion A recen t ly adapted alter n at ive to in ter n al t issue expansion is extern al tissue expansion . Th is tech n iqu e relies on t h e ap p licat ion of a con st an t , exter n al force to t h e u n d erlyin g skin , fascia, an d m u scle. Th ere are several adapted m ean s of exp an sion u sing di eren t ten sion ing devices th at are often secu red to th e periph er y of th e w oun d. As th e extern al forces are ap plied to th e overlying skin an d fascia, several cytoskelet al an d ext racellu lar ch anges occu r to produ ce a bioch em ical respon se th at resu lt s in an overall in crease of t issu e m ass an d size.13,14 Exam ples of extern al expan der system s in clude Jacob’s ladder, retention sutures, and com m ercially available extern al expan sion system s. Each of th ese tech n iqu es app lies extern al force an d ten sion to th e edge of th e defect to produ ce a gradu al closu re u sing au tologou s t issue. Th e disadvan tages of th is tech n ique in clude th e lengthy t im e requ ired for closure as w ell as p ain an d discom for t du ring th e expan sion p rocess. Th ese are n ot p rim ar y or secon dar y op t ions in spin al soft t issu e recon st ruct ion but m ust be con sidered in th e clin ical set t ing of th e p at ien t .

Biological Tissue Matrices Com m ercially available biological t issue m at rices (BTMs) cu rren tly com e from ve di eren t sou rces: h u m an derm is, p orcin e d erm is, p orcine sm all intestinal subm ucosa, bovine derm is, an d bovin e pericardium .15 Th ese BTMs can be u sed to recon st r uct soft t issue defect s w h ere th ere is a paucit y of available don or t issue. Alth ough each BTM is in h eren tly di eren t due to it s proprietar y p rocessing tech n iqu e, t h e cen t ral ten et th at all BTMs should be based on is th e provision of st rength an d early revascularizat ion capacit y in an e ort to provide soft t issue coverage an d resist in fect ion . With in spin al and sacral reconstruction, BTMs can be used in a m u lt it ude of capacit ies in cluding th e provision of du rable coverage over im p lan ted h ardw are as w ell as th e creat ion of pelvic diap h ragm s to p reven t visceral h ern iat ion in to low sacral defect s.16 Th ese defects, as m en t ion ed previou sly, arise from subtotal or total sacrectom y defects.

■ Characterization of Axial

Pattern Flaps Paraspinous Muscle Th e p araspin ou s m u scles are a m u lt ilayered grou p of m u scles th at lin e each side of th e spin al colum n an d h ave th e poten t ial to provide excellen t coverage of m oderately sized defect s at alm ost any level of th e sp in al colu m n . Although com posed of nin e paired m uscle groups, th e collect ive ap is referred to as th e p araspin ous m u scle ap an d is t ran sposed based on eit h er t h e m ed ial or lateral p er forat in g in tercostal vessels. The super cial layer is com posed of t h e sp len iu s cap it is an d t h e sp len iu s cer vicis m uscles, both of w h ich are foun d in th e cervical sp in e, or u pper th ird. Th e in term ediate layer of th e sp in al colu m n is com posed of th e iliocostalis, th e longissim u s, an d th e spin alis m u scles, an d th ese span th e en t ire length of the spinal colum n. Fin ally, the deepest layer inclu des th e t ran sversosp in al, sem isp in alis, m u lt i dus, an d rotatores m uscles. Th ese paired m idlin e m uscles are dissected su p er cially to reveal th e exten t of th e lateral

Wound Closure Techniques border. Th en th e dissect ion proceeds along th e m edial u n dersu rface of th e m u scle grou ps, an d th e parasp in ou s m u scles are elevated o of th e ribs, w ith care t aken to p reser ve th e lateral in tercostal perforators. After elevat ion , th e paraspinous m uscles are m anually advanced tow ard th e m idlin e an d su t u red togeth er.17 Th e advan tages of th e paraspin ous m uscle aps in clu de conven ien ce as w ell as a relat ive ease of dissect ion . Addit ion ally, th ere is often n o don or-site defect , as m any t im es th e ap is w ith in th e w oun d bed. How ever, because of this proxim it y to th e w oun d, t h e ap is often com p rom ised secon dar y to resect ion an d is n ot available for recon st ru ct ion (Fig. 15.1).

Trapezius Muscle Th e t rap eziu s ap can be h ar vested as a m yocutan eous or m uscle-on ly ap an d w as rst described in 1979.18 Th e ap is based o of th e d escen ding bran ch of th e t ran sverse cer vical ar ter y an d vein . In ad dit ion to th e dom in an t blood su pp ly from th e t ran sverse cer vical system , it also receives m in or con t ribu t ion s from th e dorsal scapu lar ar ter y an d perforat ing posterior in tercost al vessels. Th is ap is u sefu l for th e coverage of defect s in th e upper th ird, or cer vical region of th e spin al colu m n .

Th e advan t ages of th is ap in clude a relat ively at an d th in cu t an eou s por t ion th at can be u sefu l in sh allow d efect s. Th e d isadvan t ages in clude a relat ively lim ited arc of rot at ion an d sign i can t don or-site m orbidit y of upper ext rem it y w eakn ess. Addit ion ally, du e to th e proxim it y of th e m u scle to th e op erat ive eld, the m uscle m ay be dam aged or irradiated, leading to in abilit y to u se th e ap (Fig. 15.2).

Latissimus Dorsi Muscle Th e lat issim u s dorsi m u scle ap is a versat ile ap an d can be h ar vested eith er as a m uscle on ly or as a m yocut an eous ap. Th is ap can be u sed to recon st ru ct u pper, m iddle, an d low er areas of the spinal colum n and has a rem arkable arc of rotat ion .19 Th e lat issim us dorsi m uscle h as a du al blood su pply. Th e dom in an t vascu lar p edicle is derived from th e th oracodorsal ar ter y, an d th e ap also receives a sign i can t su p ply from th e segm en tal in tercostal vessels th at en ter th e m uscle at th e m idlin e. Th e t radit ion al lat issim us dorsi ap can be raised as an axial p at tern ap o of th e th oracodorsal ar ter y an d vein . Conversely, w h en th e lat issim us dorsi is d et ach ed at its in ser t ion in to th e h u m eru s, its vascu lar su pply becom es based on th e secon dar y in tercostal vessels, an d

Fig. 15.1 Bilateral paraspinous m uscle advancem ent  aps for exposed thoracic hardware.

169

170

Chapter 15 Fig. 15.2 Pedicled myocutaneous trapezius ap for cervical soft tissue defect.

th e resu lt ing ap is referred to as th e reverse lat issim u s dorsi ap.9 Th e advan tages of th e ap in clude a ver y large area of coverage w ith a w ide arc or rotat ion . Addit ion ally, th is ap is frequen tly out of th e zon e of inju r y, so it p rovides robust coverage of spinal wounds w ithout th e p oten t ial for radiat ion dam age. Poten t ial lim itations are the donor-site m orbidit y of harvest ing a large skin paddle an d th e n eed for skin graft coverage of th e back. Addit ion ally, ap h ar vest can lim it upper ext rem it y m ot ion in som e pat ien t s. Also, th e resu lt an t don or-site defect is p ron e to h em atom a an d serom a for-

m at ion . Large-caliber self-suct ion drain s are rout in ely placed an d rem oved w h en th e outp ut is less th an 20 m L of serous uid for 3 successive days (Fig. 15.3).

Gluteus Maximus Muscle Th e glu teu s m a xim u s m u scle is t ypically h arvested as a m u scle-on ly ap , bu t can be h arvested as a m yocu t an eou s ap if n eeded. Th e ap is based o of th e superior gluteal ar ter y an d h as a sh or t axis for rotat ion , ren dering it u sefu l on ly for defect s of th e low er th ird of th e

Fig. 15.3 Unilateral latissim us dorsi  ap for exposed thoracic hardware.

Wound Closure Techniques

Fig. 15.4 Unilateral pedicled superior gluteal artery perforator  ap and rhomboid fasciocutaneous  ap for  large sacral defect after oncological resection.

spin e. Th e superior h alf of th e m uscle is able to provide coverage for th e con t ralateral sacru m , an d th e in ferior h alf of th e m uscle is able to provide coverage for th e ipsilateral isch ium . Th e gluteu s m a xim us ap provides a robu st , relat ively large am ou n t of vascu larized m u scle an d fascia an d is able to obliterate th e largest of sacral defect s.20 Th e don or site of th e ap can be closed w ith relat ive ease using a V to Y advan cem en t closure. Because th e ap h as such a large m u scle, th e glu teu s m axim u s is p ron e to dener vation atrophy. Additionally, its proxim it y to the sciatic nerve represen ts a poten tial source of m orbidit y du ring dissect ion (Fig. 15.4).

Rectus Abdominis Muscle The pedicled VRAM ap is a versatile ap based on th e in ferior epigast ric system . It can be h arvested as eith er a m u scle-on ly or a m yocu t an eous ap an d is m ost useful for defects of th e low er th ird of th e sp in e an d th e p elvic region . Th e ap h as a robu st vascu larit y an d abu n dan t soft tissue bulk, especially in corpulen t patien ts. Th e ap is m ost u sefu l for recon st r u ct ing an d obliterat ing th e dead space th at en sues after sacrectom y or pelvic exen terat ion .21 Th e advan t age of th is ap is it s abilit y to p rovid e robu st , vascu lar ized t issu e in to t h e defect t h at is ou t sid e of t h e eld of rad iat ion . In th e set t ing of a su btotal or tot al sacrectom y

th is ap is of par t icu lar in terest as it both obliterates dead space in th e base of th e p elvis an d obt urates any sm all or large bow el segm en t th at m ay h ern iate to th e posterior aspect an d becom e lodged in spin al h ardw are or develop in to a bow el obst ru ct ion . Th e disadvan t age of th is ap is th at it n ecessit ates p erform ing th e su rger y from an an terior ap p roach , w h ich can lead to h ern iat ion , bu lge form at ion , an d w ou n d healing com plication s or injur y to the surrounding viscera (Fig. 15.5).

Fibular Bone Th e free bu lar ap is t h e ideal ch oice for t reatm en t of sacrectom y defect s th at con t ain vast am oun ts of dead space an d n o autologous bon e for st r uct u ral su ppor t , as it is a w ell-vascularized sou rce of both bon e an d soft t issu e. Th e ap is based o of th e perforat ing ar teries from th e p eron eal ar ter y. If addit ion al soft t issue is n eeded, it can be used as a ch im eric ap w ith th e soleu s m u scle. Th is ap can on ly be h arvested and used as a free ap, and therefore can be ut ilized in any region of th e spin e for osseous recon st r uct ion ; h ow ever, its m ain ut ilit y is in pelvic recon st ru ct ion an d ch ron ic osseou s nonunion s, as the bular bon e provides a stable bu t t ress to su ppor t surroun ding h ardw are.22 Th e rich vascu larit y of th is ap h as been sh ow n to p rom ote im p roved h ealing an d an

171

172

Chapter 15 Fig. 15.5 Pedicled vertical rectus ab dom inis  ap for subtotal sacral defect with  nonvascularized bone graft.

abilit y to overcom e osteom yelit is. Th e soft t issu e p or t ion of th e ap obliterates any residu al dead space. Addition ally, it can be har vested sim ultaneously w hile perform ing sacral debridem en t , e ect ively m in im izing op erat ive t im es. Th e disadvan t age of th is ap is t h at it en t ails a steep learn ing cu r ve for dissect ion an d ap h ar vest (Fig. 15.6).

■ Postoperative Care

su re on th e w ou n d closu re an d ap, am bu lat ion is recom m en ded on th e rst postop erat ive day if th ere h as been n o dural violat ion . But if a dural violat ion h as occurred, th en bed rest is recom m en ded for 3 days, du ring w h ich t im e th e p at ien t sh ou ld be t u r n ed ever y 2 h ou rs to decease th e in ciden ce of isch em ic ulcer creat ion as w ell as to o oad in cision al pressu re. Using th is st rategy, prolonged p ressu re on th e in cision is avoided an d th e poten t ial for subsequ en t ap n ecrosis is m it igated.

Ambulation

Drain Management

In an e or t to m it igate th e th rom bot ic e ects of surger y as w ell as poten t ially o oad pres-

W h en closing w ou n ds over an area of t um or ext irp at ion or h ardw are im plan t at ion , a vast

Fig. 15.6 Pedicled vertical rectus abdom inis  ap with free  bula  ap for total sacral defect.

Wound Closure Techniques am oun t of dead space is invariably created. It is crit ical for th e recon st ruct ive surgeon to m it igate th is dead space w ith a com bin at ion of vascularized soft t issue an d self-suct ion drain s. Closed-su ct ion drain s assist in th e elim in at ion of serom a and hem atom a form ation and should be left in place un t il each resp ect ive drain p rodu ces less th an 20 m L of exu date per day over a span of 3 con secu t ive days.

■ Complications Com plicat ion s in pat ien ts requiring soft t issue coverage of im p lan ted h ardw are are devast ating to th e pat ien t an d th e surgical team .2 Th e m ost com m only encountered com plications include serom a, hem atom a, w ound infection, and ap failure. In pat ients in w h om soft tissue uid collect ion s are su spected, im aging st udies such as CT or MRI are in dicated to evalu ate th e locat ion an d exten t of th e suspected collect ion as w ell as th e p oten t ial for h ardw are involvem en t . If th ere is any in dicat ion of in fect ion , cu lt u redirected, broad-spectrum antibiotics should be started, an d sh arp debridem en t of all n ecrot ic t issue sh ould be perform ed. Flap failu re, eith er p ar t ial or com plete, can occu r for a m yriad of reason s an d sh ou ld gen erate an op erat ive evalu at ion of th e ap to in terrogate th e potent ial for reversible problem s. If ap failu re persist s, th e pat ien t w ill likely requ ire fur th er operat ive m an agem en t to correct th e resu lt an t defect .

■ Chapter Summary Th e goal of recon st ru ct ive su rger y for can cer pat ien ts w h o h ave u n dergon e ext irpat ive sp in al su rger y is to restore form an d fun ct ion . Th e cen t ral ten et u n derlying th e recon st r uct ive algorith m is to progress from sim ple too m ore com p lex recon st ru ct ion s on th e basis of t h e speci c w ound requirem ents. Defects in regions th at h ave been exp osed to p rior irradiat ion or

surger y, or have a paucit y of local tissue, requ ire m ore com plex t ypes of reconstructive inter vent ion s. Using a com bin at ion of th ese tech n iqu es, th e recon st ru ct ive su rgeon aim s to obliterate dead sp ace an d provide coverage over crit ical areas in an e or t to preven t in fect ion an d restore form an d fun ct ion to th ese pat ien ts.

Pearls Com munication bet ween the plastic and neuro surgical team s is of the upmost importance when  planning surgery. Adequate comm unication en ables the surgical team  to provide full disclosure  to the patient with regard to potential donor site  morbidit y. Additionally, it enables determining  the appropriate preoperative imaging and evaluation of potential  ap donor sites. When closing wounds over the spinal column, it  is critical to identify structures that must be cov ered with vascularized tissue. Local m uscle  aps  based on axial pat tern blood supplies are opti m al to obliterate dead space and cover structural  grafts and hardware in the upper, m iddle, and  lower third of the spine. Obliterate all associated dead space in the spinal  wound with both vascularized tissue and closed suction drains. Preventing hematoma and seroma  formation is an important component of success ful wound closure over the spinal colum n. When reconstructing soft tissue defect s over  the spine, it is critical to m aintain a vascularized  wound bed free of devitalized tissue. Scar tissue,  devitalized adipose, and muscle potentially act as  a nidus for infection. Pitfalls When closing wounds over the spinal column, it  is critical to use closed suction drains in order to  eliminate serom a and hem atoma. Early rem oval  of drains leads to serom a formation and potential  infection. Failure to o load pressure on the midline spinal  wound results in pressure necrosis at the wound  site. This can subsequently lead to dehiscence and  infection of the spinal wound. At tempting closure of a spinal wound in a pri mary fashion without muscle  aps leads to higher  rates of wound dehiscence and infection. The  obliteration of dead space with muscle  aps de creases the formation of  uid collections.

173

174

Chapter 15 Refere nces Five Must–Read References 1. Ciol MA, Deyo RA, How ell E, Kreif S. An assessm en t of su rger y for spin al sten osis: t im e t ren ds, geograph ic variat ion s, com plicat ion s, an d reop erat ion s. J Am Geriat r Soc 1996;44:285–290 Pu bMed 2. St ah l RS, Bu rstein FD, Liep on is JV, Mu rp hy MJ, Piep m eier JM. Extensive w oun ds of th e spine: a com prehen sive approach to debridem en t an d recon st ruct ion . Plast Reconst r Su rg 1990;85:747–753 Pu bMed 3. Few JW, Marcu s JR, Lee MJ, On d ra S, Du m an ian GA. Treat m en t of h ost ile m idlin e back w ou n ds: an ext rem e approach . Plast Recon st r Surg 2000;105:2448– 2451 PubMed 4. Weigel B, Magh sudi M, Neum an n C, Kret schm er R, Mü ller FJ, Nerlich M. Su rgical m an agem en t of sym p tom atic spinal m etastases. Postoperative outcom e and qualit y of life. Spin e 1999;24:2240–2246 Pu bMed 5. Th algot t JS, Cotler HB, Sasso RC, LaRocca H, Gardn er V. Postoperat ive infect ion s in spin al im plan t s. Classi cation and analysis—a m ulticenter study. Spine 1991;16:981–984 Pu bMed 6. Gh ogaw ala Z, Man s eld FL, Borges LF. Spinal radiat ion before surgical decom pression adversely a ect s ou tcom es of surger y for sym ptom at ic m et ast at ic spinal cord com pression. Spine 2001;26:818–824 PubMed 7. Ch ang DW, Friel MT, You ssef AA. Recon st ruct ive st rategies in soft t issue recon st ruct ion after resect ion of spin al n eoplasm s. Spin e 2007;32:1101–1106 PubMed 8. Casas LA, Lew is VL Jr. A reliable approach to th e closure of large acqu ired m idlin e defect s of th e back. Plast Recon st r Su rg 1989;84:632–641 PubMed 9. Stevenson TR, Roh rich RJ, Pollock RA, Dingm an RO, Bost w ick J III. More exp erien ce w ith th e “reverse” lat issim us dorsi m usculocut an eous ap: precise locat ion of blood supply. Plast Recon st r Surg 1984;74: 237–243 Pu bMed 10. Gar vey PB, Clem en s MW, Rh in es LD, Sacks JM. Vert ical rect us abdom inis m usculocutaneous ow -through ap to a free bu la ap for tot al sacrectom y recon st ru ct ion . Microsu rger y 2013;33:32–38 PubMed 11. Mor ykw as MJ, Argen t a LC, Sh elton -Brow n EI, McGu ir t W. Vacu u m -assisted closu re: a n ew m et h od for w oun d con t rol an d t reat m en t: an im al st udies an d basic foun dat ion . An n Plast Su rg 1997;38:553– 562 Pu bMed

12. Cun h a MS, Nakam oto HA, Herson MR, Faes JC, Gem perli R, Ferreira MC. Tissu e exp an der com p licat ion s in plast ic surger y: a 10-year experien ce. Rev Hosp Clin Fac Med Sao Paulo 2002;57:93–97 Pu bMed 13. Lash een AE, Saad K, Raslan M. Extern al t issue expan sion in h ead an d neck recon st r uct ion . J Plast Recon st r Aesth et Surg 2009;62:e251–e254 PubMed 14. Baird R, Gh oloum S, Laberge JM, Pu ligandla P. Man agem en t of a gian t om p h alocele w ith an extern al skin closure system . J Pediat r Su rg 2010;45:E17–E20 PubMed 15. Broyles JM, Abt NB, Sacks JM, Butler CE. Bioprosth et ic t issue m at rices in com plex abdom in al w all recon st ruct ion . Plast ic an d Recon st r uct ive Su rger y–Global Open . 2013;1:e91 16. Korn JM, Con n olly MM, Walton RL. Single-st age sacral coccygectom y an d repair using h um an acellular d erm al m at rix (AlloDerm ) w ith bilateral glu teu s m axim us aps for h ern ia prophylaxis. Hern ia 2009; 13:329–332 Pu bMed 17. Hu lt m an CS, Jon es GE, Losken A, et al. Salvage of in fected sp in al h ardw are w it h p arasp in ou s m u scle ap s: an atom ic con siderat ion s w ith clin ical correlat ion . An n Plast Surg 2006;57:521–528 Pu bMed 18. Dem ergasso F, Piazza MV. Trapezius m yocut an eou s ap in recon st r u ct ive su rger y for h ead an d n eck can cer: an original tech n ique. Am J Surg 1979;138:533– 536 Pu bMed 19. Giessw ein P, Const an ce CG, Mackay DR, Man ders EK. Superch arged lat issim us dorsi m u scle ap for coverage of th e p roblem w ou n d in th e low er back. Plast Recon st r Su rg 1994;94:1060–1063 PubMed 20. Fur ukaw a H, Yam am oto Y, Igaw a HH, Sugih ara T. Gluteus m axim us adipom uscular t urnover or sliding ap in th e surgical t reat m en t of exten sive sacral ch ordom as. Plast Reconstr Surg 2000;105:1013–1016 PubMed 21. Glat t BS, Disa JJ, Meh rara BJ, Pusic AL, Bolan d P, Cordeiro PG. Recon st ruct ion of exten sive par t ial or tot al sacrectom y defect s w ith a t ransabdom in al vert ical rect us abdom in is m yocu t an eous ap. An n Plast Surg 2006;56:526–530, discu ssion 530–531 PubMed 22. Ch oudr y UH, Moran SL, Karacor Z. Fun ct ion al recon st ru ct ion of th e p elvic ring w ith sim u lt an eou s bilateral free bular aps follow ing tot al sacral resect ion . An n Plast Surg 2006;57:673–676 PubMed

16 Complications and Their Avoidance : How to Plan Primary Tumor Resection to Minimize Complications and Maximize Outcome Michael C. Oh, Vedat Deviren, and Christopher P. Ames

■ Introduction Alth ough prim ar y t um ors of th e spin e are rare, th eir t reat m en t requ ires som e of th e m ost dem an ding an d com p lex su rgical ap proach es an d can resu lt in sign i can t m orbidit y an d m or t alit y.1 Th e best ch an ce of cu re or m in im izat ion of recurren ce often requ ires tot al en -bloc resect ion for m align an t t u m ors, w h ereas som e ben ign t um ors m ay be am en able to in t ralesion al p iecem eal rem oval. Du e to tech n ical ch allenges required for w ide-m argin en-bloc resection w ith com plex recon st ru ct ion , com plicat ion rates rem ain h igh . Th ese com plicat ion s in clude con st ruct failure, align m en t failure, pseudarth rosis, in fect ion w ith or w ith ou t skin deh iscen ce, sp in al cord or nerve root injur y, cerebrospinal uid leak w ith pseu dom en ingocele, excessive bleeding requ iring t ran sfu sion s, large-vessel inju r y, an d in t raop erat ive con t am in at ion of m align an t tum ors. Large tum ors m ay involve nearby ner ve roots or dura, w ith plan n ed an d u nplan n ed risk of n eu rologic d e cit s; w id e-m argin en -bloc resect ion for m align an t p rim ar y t u m ors m ay require sacri ce of n eu ral elem en t s to ach ieve th e best ch an ce of recu rren ce-free cure. Dep en d in g on t h e locat ion of t h e t u m or, d i eren t vital an atom ic st ruct ures m ay be in -

volved, including vertebral arteries, th e t rach ea, esoph agus, aor ta, ven a cava, iliac vessels, ureters, and parts of the distal gastrointestinal tract. Oth er com m on postoperat ive com plicat ion s, su ch as deep ven ous th rom bosis, pu lm on ar y em bolism , urin ar y t ract in fect ion , pn eum on ia, dysp h agia, ileu s, an d m yocardial in farct ion m ay be caused by prolonged im m obilit y an d exten sive surgical resect ion s w ith long operat ive t im es. Becau se of t h e com p lexit y involved in resect ing p r im ar y sp in e t u m ors, carefu l p reop erat ive plan n ing is m an dator y, w h ich in clu des im agin g, biop sy, an d , p erh ap s m ost im p ort an tly, detailed p at ien t coun seling an d sh ared decision m aking. Pat ien ts sh ou ld be w ell in form ed of all poten t ial com p licat ion s, an d discussion s regarding th e risks an d ben e ts of th e poten t ial sacri ce of n eu ral elem en ts for th e best ch an ce of cu re sh ou ld be par t of ever y p reoperat ive coun seling. Th is ch apter review s th e literat ure related to com p licat ion s in su rgical resect ion of p rim ar y spin e t u m ors an d discusses h ow best to plan p rim ar y sp in e t um or resect ion to m in im ize com plications w hile m axim izing outcom e. We p resen t several cases from ou r ow n exp erien ce w ith prim ar y spin e t um ors to illu st rate

176

Chapter 16 som e of th e m ore com m on com p licat ion s as w ell as th e im port an ce of ju dicious preoperat ive plan n ing.

■ Preoperative Planning

and Imaging Planning of prim ary spine tum or resection starts w ith a th orough im aging evaluat ion . Stan ding scoliosis X-rays sh ou ld be u t ilized to assess th e region al an d global align m en t . Th is is esp ecially im por tan t for lum bar t um ors, w h ere th e p ostoperat ive p elvic in ciden ce (PI) sh ou ld m easu re w ith in 11 degrees of lu m bar lordosis. Th e location of th e in dividualized thoracic apex sh ou ld be n oted to en su re t h at th oracic in st r u m en tat ion exten d s beyon d th e ap ex w h en plan n ing th oracic resect ion s. Cer vicoth oracic kyp h osis sh ou ld be n oted w h en plan n ing cervical resections as well. Because prim ary tum or pat ien ts t reated w ith en -bloc resect ion s often h ave lon g su r vival, con for m in g to d efor m it y p r in cip les is im p or t an t to opt im ize fu n ct ion alit y an d pat ien t-repor ted sat isfact ion , an d to preven t con st ru ct failu re. Com pu ted tom ograp hy (CT) is u sed to evalu ate th e exten t of bony invasion s, but is also crit ical in evaluat ing th e bon e qualit ies of adjacen t ver tebral bodies for poten t ial in st rum en tat ion . Th e n um ber of involved ver tebral levels sh ou ld be carefu lly n oted. W h eth er th e t u m or is con n ed to th e vertebral body or exten ds in to p ed icles or lam in as sh ou ld be evalu ated an d grad ed by t h e Wein stein -Bor ian i-Biagin i (W BB) surgical st aging system .2 For exam p le, w h eth er on e or both pedicles are involved can ch ange su rgical plan n ing, as th e sequ en ce an d locat ion of osteotom ies for total en -bloc spon dylectom y di er depen ding on th e exten t of tum or invasion into pedicles and posterior bony st ruct ures.2,3 The deliver y corridor for the specim en sh ould be decided based on th e size an d locat ion of th e bu lk of th e m ass an d it s adh eren ce to su rrou n ding st ru ct u res su ch as viscera an d m ajor vessels. Magn et ic reson an ce im agin g (MRI), sp ecifically T1-w eigh ted scan s w it h gadolin iu m or T2-w eigh ted scan s, can provide addit ion al in -

form at ion regarding th e exten t of t u m or invasion . Involvem en t of n on -bony st ru ct u res, su ch as th e epidural space, n euroforam in a, or paravertebral tissues, should be carefully evaluated. Alth ough rou t in e p reop erat ive angiogram s are n ot n ecessar y in m ost cases, select ive ar terial em bolizat ion m ay provide good ou tcom es in p at ien t s w ith an eu r ysm al bon e cyst s,4 especially given th e n ding t h at su rger y can resu lt in 15 to 30% com plicat ion rates.5 Preoperat ive angiogram s in pat ien ts w ith large cer vical t um ors m ay be h elp fu l in evalu at ing th e p aten cy of ver tebral ar teries (VAs).6 Occlu sion tests can also be perform ed to determ in e if it is safe to sacri ce the arter y, and if sacri cing the VA w ill assist in t u m or resect ion w ith w ide m argin s. Coils sh ould be placed distal an d proxim al to sites of in ten ded VA ligat ion to preven t in adver ten t coil rem oval an d back bleeding at su rger y. An exam p le of su ch case is dep icted in (Fig. 16.1). In th is pat ien t w ith gian t cer vical ch ordom a sp an n ing fou r levels from C3 to C6, an angiogram w as rst perform ed to coil an d sacr i ce t h e r igh t VA, w h ich w as en cased by th e t u m or. Th is w as follow ed by in t raop erat ive ligat ion of th e righ t VA during th e rst port ion of th e procedu re via a p osterior app roach . Follow ing p oster ior osteotom ies an d ligat ion of n er ve root s, a Silast ic sh eet w as placed bet w een th e d u ra an d th e p osterior longit u din al ligam en t to provide an easier p lan e of dissect ion for th e en -bloc rem oval of t u m or from con sequen t an terior ap proach .

■ Preoperative Diagnosis

and Biopsy Prop er su rgical p lan n in g requ ires biop sy for path ological diagn osis. Alth ough sm all asym p tom atic ben ign tum ors m ay be obser ved, sym p tom at ic ben ign t um ors m ay require surger y for decom pression or st abilizat ion .2 For m align an t t u m ors, a biopsy sh eath sh ould be used w ith th e t ract carefully n oted an d m arked, as en bloc resect ion requ ires resect ion of th e biop sy t ract for th e best outcom e. Th e st udy by Fourn ey and colleagues 7 suggested th at surger y th at w as p erfor m ed at th e sam e cen ter w h ere t h e

Complications and Their Avoidance

b

c

d

e

a

f

Fig. 16.1a–g A four-level en-bloc resection of a cervical chordoma in a 60-year-old woman who presented with severe neck pain. She was found to have a giant chordoma centered at C4 and expanding the anterior aspects of C3 thru C6, as shown by sagit tal T2-weighted magnetic resonance imaging (MRI) (f). She underwent a four-level en-bloc resection by a t wo-stage approach, where the rst approach consisted of posterior osteotomies, ligation of right vertebral artery, and placement of a

g

Silastic sheet (a). (a–c) This was followed by anterior en-bloc tumor resection. The patient also underwent preoperative coil embolization of the right vertebral artery (b, arrowhead) and intraoperative surgical clipping (b, arrow). The right C4 and C5 nerve roots were sacri ced. Stabilization consisted of occiput to T3 posterior spinal fusion and instrumentation with a C3-C6 anterior cage and plate (d,e). (g) Postoperative T2-weighted MRI showed good decompression of the spinal cord.

177

178

Chapter 16 in it ial biopsy w as don e h ad bet ter outcom es. Alth ough n ot stat ist ically sign i can t , pat ien ts w h o received biop sy an d su rger y at t h e sam e refer ral cen ter w ere 78% d isease free at t h e n al follow -u p, w h ereas th ose w h ose biopsy w as don e at on e cen ter an d w h o w ere th en referred elsew h ere for surger y w ere 55% disease free at n al follow -up.7 Con sisten t w ith th ese resu lt s, p at ien t s t reated for sacral ch ordom as by biop sy an d su rger y at t h e sam e referral center h ad a 75% disease-free rate, w h ereas th ose in it ially t reated elsew h ere h ad a 42% diseasefree rate. Alth ough th is n d ing w as also n ot st at ist ically sign i can t , experts gen erally agree th at biopsy an d surger y sh ould be perform ed at t h e sam e in st it u t ion , p referably on e w it h exten sive exp erien ce.1,7 Follow ing com p lete im aging an d biop sy st udies, a system at ic su rgical st aging sh ou ld be don e to form u late con sisten t su rgical plan n ing. We recom m en d th e u se of th e W BB su rgical st aging system 2 to h elp guide m ore con sisten t su rgical p lan n ing an d for system at ic report ing of su rgeries an d th e exten t of t u m or invasion .

■ Surgical Nuances Extent of Resection Malign an t prim ar y sp in e t u m ors, su ch as ch ordom as, ch on drosarcom as, osteosarcom as, an d Ew ing’s sarcom as, requ ire w ide en -bloc resect ion w ith adh eren ce to on cological p rin ciples for best ou tcom es.1,2,8–10 For exam p le, en -bloc resect ion w it h w id e m argin s p rovid es good long-term local con t rol for 92.3% of gian t cell t um ors,11 78% of chordom as,8 and 82% of chond rosarcom as.9 By com parison , in t ralesion al resect ion p rovides local con t rol for 72.2%of gian t cell t um ors,11 22% of ch ord om as,8 an d 0% of ch on drosarcom as.9 Alth ough it is clear th at th e best ou tcom es are p rovid ed by en -bloc resect ion , m orbid it y rem ain s h igh , w it h an overall com p licat ion rate of 35.1% w it h 2.2% overall m or t alit y.1 Certain benign t um ors, n am ely giant cell t um ors (GCTs) w ith En n eking st age III2,11 or p oster iorly located an eu r ysm al bon e cyst s,4 also h ave been sh ow n to h ave t h e best ou tcom e

w ith en -bloc resect ion . On e su ch pat ien t in our exp erien ce w as a 29-year-old m an w h o w as previou sly t reated in a di eren t cou n t r y for a T12 GCT w ith lam in ectom y, par t ial an terior corp ectom y, an d an terior an d p osterior spin al in st ru m en tat ion (Fig. 16.2). A recu rren ce of th is t u m or, ju st lateral to t h e an terior cage, w as noted on follow -up MRI. We perform ed en -bloc spondylectomy via thoracotom y, w hich included th e recu rren t t u m or as w ell as th e an terior cage from th e previou s su rger y as on e piece. Alth ough com p lete sp on dylectom y for resect ion of prim ar y spin e t u m ors w as in it ially d escribed by Roy- Cam ille’s grou p 12 an d by Sten er,13 tot al en -bloc sp on dylectom y u sing a t h read-w ire saw w as rst d evelop ed by Tom ita.3,14 Th is tech n iqu e h as also been exten ded for t reat m en t of m et astat ic t u m ors for selected pat ien t s. Develop m en t of th ese n ew su rgical tech n iqu es an d d evices h as en abled p er for m ing en -bloc sp on dylectom ies in th e sp in e. Furth erm ore, th e use of a th read-w ire saw h as been sh ow n to m in im ize t um or con tam in at ion during en -bloc resect ion s.14 Fu t ure d evelop m en t of n ew d evices, su ch as t h e th read-w ire saw guide an d spin al cord protector device, m ay h elp reduce n eu rologic de cit s an d ot h er m ajor in t raoperat ive com p licat ion s involved w ith tech n ically ch allenging tot al en bloc spon dylectom ies. Som e tot al en -bloc sp on dylectom ies require m ultidisciplinary approaches w ith m ultist aged operat ion s (Fig. 16.3). In a 38-year-old w om an w ith a gian t ch ordom a sp an n ing from T5 th rough T8, w e p erform ed a th ree-st age en bloc resect ion w ith assist an ce from th oracic surgeon s to dissect an d m obilize th e aorta an d esoph agus. Th e pat ien t w as rst t reated preoperat ively w ith stereotact ic radiosurgery w ith Cyberkn ife (2,500 cGy), follow ed by posterior d ecom pression an d osteotom ies to release t h e p osterior com pon en t s. Th e secon d st age con sisted of a left th oracotom y for dissect ion of th e aor ta o th e t u m or m ass w ith an terior spin al osteotom ies at T5 an d T8-9. Th e n al st age con sisted of d issect ion an d m obilizat ion of t h e esop h agu s w ith en -bloc spon dylectom y of T5-T8 an d d eliver y of th e t u m or. Th e p at ien t is cu rren tly recu rren ce-free 18 m on th s after su rger y an d lives a ver y act ive life.

Complications and Their Avoidance

a

b

c

Fig. 16.2a–c En-bloc resection of a recurrent giant cell tumor with previous instrum entation in a 29-year-old man who was previously diagnosed with a T12 tumor and treated at another hospital with a lam inectomy, partial anterior corpectomy, and anterior and posterior spinal fusion and instrumentation. Biopsy at the time con rm ed giant cell tumor. (a) Preoperative MRI. (b) On follow-up MRI, recurrence was evident lateral to the anterior cage.

(c) Thus, we performed an en-bloc spondylectomy via a posterior osteotomy and then a thoracotomy with T10 rib resection, followed by T12 anterior spinal osteotomy, L2-L3 diskectomy, and en-bloc corpectomy of T12-L2. Careful planning allowed salvage of this patient with negative surgical margins. Finally, the spine was stabilized with cage reconstruction from T12 to L3 and anterior spinal fusion and instrumentation from T11 to L3.

Tumor Location

spine. There were eight com plications, both early an d late, in cluding graft ext r usion , graft don orsite in fect ion , in st ru m en t failu re an d revision , respirator y com plicat ion , diabetes in sipidus of u n kn ow n et iology, postoperat ive ep idural h em atom a, seven t h n er ve p alsy, an d on e p ost op erat ive d eat h . Of n ote, t h ere w as on e VA bleed ing d u r in g t u m or resect ion , w h ich w as rep aired w ith out com p licat ion s. Cloyd an d colleagues 10 perform ed a system at ic review of t h e literat ure to iden t ify pat ien t s w h o u n der w en t en -bloc resect ion for prim ar y cer vical spin e t u m ors. Eigh t p at ien ts h ad com plications. The m ost com m on com plication was dysphagia. Other com plicat ions included pneu m onia, respirator y distress, seizure, throm botic even t , an d w ou n d deh iscen ce. Of n ote, m ean operat ive t im e, estim ated blood loss, an d length of st ay w ere 18.6 h ou rs (range, 4.3–56 h ou rs), 2.9 L (range, 0.6–8.7 L), an d 34.6 days (range, 16–54 days), respect ively. From th is st udy, on e can con clude th at en -bloc resect ion of cer vical spin al t u m ors carries a h igh m orbidit y rate, an d th u s, pat ien t s sh ou ld be carefu lly selected and sh ould be w ell inform ed about the risks and

Decision m akin g regard in g t h e u se of d i eren t ap p roach es an d m u lt id iscip lin ar y team s, w h ich m ay in clu d e h ead an d n eck su rger y, th oracic surger y, gen eral su rger y, an d plast ic surger y, are also determ ined by t um or locat ion . Alt h ough t h e evalu at ion of t u m or exten sion in to su r rou n d ing st r u ct u res is a cr it ical step in preoperat ive plan n ing, th e t um or locat ion (cer vical, th oracic, lum bar, or sacral) presen ts in h eren t risks to di eren t an atom ic st r uct u res. Prim ar y t u m ors of th e cer vical sp in e p resen t risks to th e VAs, esop h agu s, t rach ea, cer vical n er ve roots, an d spin al cord. Th e aor ta, esoph agu s, an d ven a cava are at risk in th e th oracic spin e during su rger y. Late aor t ic dissect ion causing paraplegia and death has been reported follow ing th oracic t u m or surger y.1 Lastly, su rger y in th e lum bar an d sacral region s presen ts risks to t h e cau da equ in a, aor t a, iliac vessels, ureters, an d distal gast roin test in al tracts. In on e in st it u t ion al series by Zileli an d colleagu es,6 66 surgeries w ere p erform ed in 35 pat ien ts w ith p rim ar y t u m ors of th e cer vical

179

180

Chapter 16

a

b

c

d

f

e

g

Fig 16.3a–i Three-stage en-bloc resection of giant thoracic chordoma with dissection and mobilization of the aorta and esophagus in a 38-year-old woman who presented with back pain. (a,b) She was found to have a giant chordoma spanning T5 to T8 as shown on T2-weighted MRIs. (c) She underwent stereotactic radiosurgery with Cyberknife (2,500 cGy), followed by a three-stage en-bloc resection. The rst stage consisted of a posterior lam inectomy of T5-T9, transpedicular osteotomies of T6-T8, rhizotom ies of T6-T8, and rib resection at T6-T9.

h

i

The second stage consisted of a left thoracotomy for dissection of the aorta o of the tumor mass, with anterior spinal osteotomies at T5 and T8-T9. (d–i) The nal stage consisted of dissection of the esophagus o of the tumor with en-bloc spondylectomy of T5-T8 with delivery of tumor, cage reconstruction of T5-T9, and anterior spinal instrum entation of T5-T9. The patient tolerated the procedures without major complications and is fully ambulatory at last follow-up.

Complications and Their Avoidance ben e t s of su rger y. Moreover, d u e to vit al an atom ic st ru ct ures in th e cer vical spin e, w ide en bloc spon dylectom ies, alth ough ach ievable, are often di cult in the cer vical spine 6,10 (Fig. 16.1). An oth er locat ion t h at p resen t s special ch allen ges for t h e su rgeon is t h e sacr u m . Tu m ors in th is region can often grow to ver y large p rop or t ion s before becom ing sym ptom at ic. High sacrectom y leads to bow el an d bladd er in con tinence, raising risks of infection, m otor de cits (loss of p lan t ar exion ), an d sen sor y abn orm alit ies. Th ese com plicat ion s w ere foun d to be d epen den t on th e exten t of spin al n er ve resect ion an d th e level of th e sacrectom y.7,15–17 For exam ple, p reser ving th e S3 n er ve root w as th e best indicator for preser ving bow el and bladder fu n ct ion , as u n ilateral an d bilateral S3 n er ve root resect ion resu lted in a 37.5% an d 75% bow el an d bladder in con t in en ce rate, respect ively.16 How ever, for large t u m ors th at exten d above th e S3 n er ve root , in ten t ion al sacri ce of m ult iple n er ve root s m ay be w arran ted, w ith h igh sacrectom ies o ering th e best ch an ce of recu rren ce-free cu re. Bow el inju r y during su rger y is also a poten t ial m ajor com plicat ion ,15 as is p ostop erat ive ileu s. A d iver t ing colostom y m ay be of ben e t in red u cin g in fect ion r isk du e to fecal con t am in at ion in tot al sacrectom y pat ien ts.

■ Complications Hardw are Construct and Alignment Failure In cases w h ere th e spin al colu m n is destabilized, com p lex circu m feren t ial recon st r u ct ion w it h in st r u m en t at ion is requ ired to restore th e th ree colum n s of th e spin e (an terior, m iddle, an d posterior). On ce th ough t to be best ach ieved by con n ect ing an terior an d posterior im plan t s, advan ces in m odern im plan ts often do n ot requ ire th is con n ect ion . Au togen ou s bon e grafts or bon e subst it u tes sh ou ld be used to h elp for m solid fu sion s. Rib graft s m ay be p laced adjacen t to cages to facilit ate p r im ar y fu sion , an d m u lt ip le rod con st r u cts m ay also

h elp preven t rod fract ure related to delayed u n ion . Prop er st abilizat ion is cru cial for early m obilizat ion an d t im ely recover y of m u scle fu n ct ion , w ith th e abilit y to bear w eigh t an d thus decrease pressure on the extensive wounds. For prim ar y sacral t u m ors, sp ecial recon st r u ct ion tech n iqu es are requ ired follow ing en bloc sacrectom ies to restore th e con t in uit y of th e p elvic ring an d to est ablish bilateral u n ion s bet w een th e lu m bar spin e an d iliac bon es,18 w h ich are key steps in ach ieving m ech an ical su p por t for early m obilizat ion . Fu r th er st u dies are n eeded, h ow ever, to evaluate th e biom ech an ics of th e pelvic st r u ct u res in m ore det ail follow ing sacrectom y. In on e pat ien t , w e perform ed an en -bloc t um or resect ion for a gian t sacral ch ordom a involving th e low er sacrum below th e in ferior S2 segm en t (Fig. 16.4). Th is 65-year-old m an u n d er w en t a h igh sacrectom y below S1 w ith ou t in st r u m en t at ion . On follow -up pelvic X-ray an d CT scan s, h e w as foun d to h ave a m idsacral fract ure an d pelvic op en -book fract u re w it h d est r u ct ion of t h e an terior ram i of th e bilateral pelvis. Alth ough previou s st u dies h ave in dicated th at sacrectom ies below th e S1 n er ve root do n ot requ ire in st r u m en ted recon st r u ct ion for early m obilizat ion ,19 t h is case exem p li es t h e n eed for fu r th er evaluat ion of pelvic biom ech an ics follow ing sacrectom y. Con st r u ct failu re, w h eth er early or late, can en t ail p seu dar th rosis, rod fract u re, or m alp osit ion or m igrat ion of th e con st ru ct . Fig. 16.5 illu st rates a case of con st r u ct failu re d u e to p seu d oar t h rosis an d rod fract u res follow ing recu rren t th oracic ch on drosarcom a resect ion 5 years after en -bloc in dex surger y. Follow ing en -bloc resect ion of a T3-T4 ch on drosarcom a, th e p at ien t developed p seu dar th rosis an d rod fract ures, resulting in severe kyphosis w ith dorsal m igrat ion of th e cage in to th e spin al can al. Th is required reoperat ion to reposition the cage, w ith exten sion of th e previous fusion across th e cer vicoth oracic jun ct ion . In on e st u dy by Bor ian i’s grou p ,1 am on g 134 pat ien t s w h o un der w en t en -bloc resect ion of prim ar y spin e t u m ors, 11 pat ien ts h ad h ardw are failu re or loosen ing, t h ree w it h d efor m it y, and on e w ith a m alpositioned anterior

181

182

Chapter 16

a

c

b

e

d

f

Fig 16.4a–f Sacral and pelvic open-book fractures following a high, but not complete, sacrectomy in a 65-year-old man who presented with sacral and leg pain. (a) He was found to have a giant sacral chordoma (arrow) involving the lower sacrum up to the inferior S2 segment. Patient underwent high sacrectomy for en-bloc tumor resection. Although

S1 was left intact, a follow-up pelvic X-ray (b) and CT scans (c) revealed a m idsacral fracture (arrow) and a pelvic open-book fracture (d, arrow) with pelvis destruction of the bilateral anterior rami. (e,f) This patient was later instrum ented at another hospital.

cage du ring su rger y t h at requ ired reoperat ion . Alth ough reoperat ion w as m ostly due to t um or recu rren ce, h ardw are failu res also con t ribu ted sign i can t ly to t h e n u m ber of reop erat ion s. In an oth er repor t , th ere w ere eigh t repor ted com p licat ion s from 66 operat ion s for cer vical prim ar y sp in e t u m ors.6 Tw o of th e eigh t com plicat ion s (25%) w ere h ardw are-related (graft extrusion, and instrum en t failure an d revision ). Alth ough m alposit ion of th e con st r uct du ring surger y is rare, cont inued follow -up w ith X-rays is cru cial because, over t im e, som e pat ien t s develop p seu dar t h rosis, h ardw are failu re or loosen ing, m igrat ion of h ardw are, or deform it y. Becau se large t u m or resect ion from t h e sp in e in evit ably d est abilizes t h e sp in e, it is im por tan t to restore both sagit t al an d coron al

balan ce w h en recon st r u ct ing t h e sp in e follow ing resect ion . Th e sam e prin ciples used in deform it y su rger y ap ply h ere. We recom m en d pre- and postoperative standing scoliosis X-rays to evaluate th e pat ien t’s global spin al balan ce. Global align m en t can fail follow ing p rim ar y spin e t um or resect ion due to in fect ion , pseu dar t h rosis, an d h ardw are failu re, an d m ay requ ire revision w it h exten sion of fu sion to addit ion al levels (Fig. 16.6). On e p at ien t w h o u n der w en t en -bloc resect ion of an L3-L4 ch on drosarcom a did w ell in it ially, bu t th en star ted develop ing severe low back p ain w ith sagit t al an d coron al im balan ce an d a for w ard t ilt . A CT scan w as su sp iciou s for d eep in fect ion cen tered at t h e L4-L5 d isk sp ace, w it h p seu darth rosis an d lu cen cy arou n d th e allograft . Th e

Complications and Their Avoidance

a

b

d

e

c

Fig 16.5a–e Construct failure due to pseudarthrosis and rod fractures following recurrent thoracic chondrosarcom a resection 5 years after en-bloc index surgery. This is a 57-year-old m an with a history of a T3-T4 chondrosarcoma previously resected via a transpedicular T3-T4 corpectomy with anterior cage reconstruction and a T1-T7 posterior spinal fusion and instrumentation. Although this patient did well initially, he later presented with increasing bilateral leg pain and weakness. Workup

with X-ray (a,b) and CT (c) showed pseudarthrosis at T2-T3 (a,b, arrows) and rod fractures (c, arrow). This resulted in severe kyphosis with dorsal m igration of the cage into the spinal canal. The patient was returned to the operating room for a transpedicular corpectomy of T2 with repositioning of the cage. (d,e) Posterior spinal instrum entation was extended up to C5. At the nal follow-up, his symptoms improved with a stable construct.

pat ien t w as reop erated w ith exten sion of th e p oster ior sp in al fu sion an d in st r u m en t at ion d ow n to t h e p elvis w it h restorat ion of coron al an d sagit t al balan ce. Cu lt u res later t u r n ed p osit ive for Pseudom onas aerugin osa. Revision s n ear t h e cer vicot h oracic an d t h oracolu m bar ju n ct ion s m ay requ ire fu sion across th ese segm en t s, w h en align m en t fails du e to com p licat ion s (Fig. 16.5). We h ave also foun d th at fou r-rod p osterior recon st ru ct ion follow ing tot al spon dylectom y p rovides a m ore solid con st r u ct w ith less poten t ial for failu re (Figs. 16.3 an d 16.7).

Wound Infection Becau se of long operat ive t im es w ith com p lex recon st ru ct ion involved, in fect ion rates rem ain h igh , often requ ir in g reop erat ion for w ou n d w ash ou t or revision of in st r u m en t at ion (Fig. 16.6). Dead space closu re w ith m uscle aps or via paraspin al m obilizat ion is crit ical in p reven t ing serom a form at ion , w h ich p resen t s a h igh risk of delayed in fect ion . Non resorbable su t u res sh ou ld be u sed for fascial closu res in h igh -risk w oun ds, at th e cer vicoth oracic jun ct ion , an d in p reviou sly radiated areas. On e su ch

183

184

Chapter 16

a

c

e

b

f

d

Fig. 16.6a–f Alignment failure with sagit tal imbalance following en-bloc resection of an L3-L4 chondrosarcom a in a 48-year-old m an. (a,b) Anterior cage reconstruction and L1-L5 posterior spinal fusion and instrumentation. (c,d) He did well initially but then started developing severe low back pain with sagit tal and coronal im balance and a forward tilt. CT scan was suspicious for deep infection centered at the L4 (b, arrow) to L5 disk

space with pseudarthrosis and lucency around the allograft. (e,f) Patient was reoperated with extension of the posterior spinal fusion and instrum entation down to the pelvis with restoration of coronal and sagit tal balance via Smith-Peterson osteotomies at L4-L5 and L5-S1. The infected L4-L5 disk space was curet ted and then packed with autograft from the iliac crest. Cultures later returned positive for Pseudomonas aeruginosa.

case is a 76-year-old m an w h o u n d er w en t an en -bloc t u m or resect ion an d recon st r u ct ion for a C2 ch ordom a p er for m ed from a t ran sm an dibu lar-t ran sglossal app roach (Fig. 16.8). Follow ing su rger y, th e p at ien t u n der w en t p roton beam th erapy. Tw o m on th s later, h e p resen ted w ith p u ru len t , op en drain age from h is ph ar yngeal w ou n d w ith an exposed an terior cage th rough th e ph ar yngeal w ou n d . Th is p at ien t even t u ally requ ired reop erat ion w ith a fat graft w it h free ap closu re an d lon g-ter m an t ibiot ic th erapy. W h en at tem pt ing to p reven t postoperative phar yngeal erosion or w oun d in fect ion s after proton beam th erapy, a ret rop h a-

r yngeal approach , autograft rath er th an m et al im plan t s, an d soft t issue aps h ave all be em ployed. Desp ite th ese m easu res, h igh an terior cer vical recon st r uct ion s h ave h igh com plicat ion rates. In on e sem in al st u dy by Bor ian i’s grou p 1 looking at m orbid it y associated w it h en -bloc resect ion s of prim ar y spin e t um ors, th e d eep in fect ion rate w as foun d to be 5.4%, an d surgical debridem en t an d long-term m u lt iple an t ibiot ics w ere required. Th e rate of deep in fect ion w as h igh er (9.8%) in pat ien ts previously t reated w ith rad iat ion th erapy. Overall, prim ar y bon e t u m ors of t h e sp in e car r y a h igh er in cid en ce

Complications and Their Avoidance

a

b

c

d

e

f

g

Fig 16.7a–g Supplem entation of posterior spinal instrum entation with a four-rod system for increased stabilization in a 78-year-old woman who presented with severe back pain. (a,b) She was found to have a hyperintense lesion at T10 on T2-weighted MRI, consistent with chordoma. She

underwent a total en-bloc spondylectomy of T10 (c,d), followed by anterior cage reconstruction and a four-rod posterior spinal fusion construct (e –g). She tolerated the procedure without signi cant complications and remained neurologically intact.

of su rgical-site in fect ion com pared w ith oth er spin e t um or surgeries. Th e in fect ion rate for prim ar y bony spin al t u m ors w as repor ted to be 13.7%, w h ereas prim ar y n on -bony an d m etast at ic sp in al t u m ors car r ied an in cid en ce of su rgical-site in fect ion of 8.9% an d 9.5%, resp ect ively.20 Staphylococcus aureus w as th e m ost com m on organ ism , fou n d in a th ird of all su rgical-site in fect ion s. Major risk factors for in fect ion follow ing spin al t u m or su rger y w ere previou s spin al su rgeries, com p lex plast ic closures, increasing num ber of com orbidities, presen ce of a h ospital-acquired infect ion at the tim e of a p reviou s su rger y, p reviou s rad iat ion t h erapy, an d longer h osp it al st ay.20 In gen eral, com bined anterior/posterior approaches, w h ich

are com m only required for prim ary spinal tum or su rgeries, age greater th an 60 years, sm oking, diabetes, previou s surgical in fect ion , in creased body m ass in dex, an d alcoh ol abu se are all pred isp osing risk factors for in fect ion follow ing spin e surger y. Postsurgical infection and w ound dehiscen ce follow ing sacrectom y w arran t a special review because w oun d in fect ion rates are especially h igh given th e vicin it y close to th e large bow el, w ith in creased in ciden ce of bow el in con t inence follow ing sacrectom y. On e such case w as a 52-year-old w om an w h o u n der w en t en -bloc sacrectom y w ith reconst ruct ion for sacral ch ondrosarcom a (Fig. 16.9). Sh e presen ted 4 m onths later w ith a 3 cm × 1 cm w ou n d deh iscen ce an d

185

186

Chapter 16

a

c

b

d

e

Fig. 16.8a–e Pharyngeal complication following en-bloc tumor resection and reconstruction for C2 chordoma perform ed from a transmandibulartransglossal approach in a 76-year-old m an who presented with severe neck pain, and who underwent a posterior laminectomy and instrumentation with a biopsy con rming a chordom a at another hospital. Patient then started developing leg paresthesia 10 days after his initial surgery. (a,b) MRI showed collapse of the C2 vertebral body with retropulsion into spinal canal. Once he was transferred to our hospital, he underwent posterior removal of hardware, exposure and mobilization of

bilateral vertebral arteries at C2, osteotomy of C2 pedicles, and occiput to T2 posterior spinal fusion and instrumentation, followed by a pharyngeal en-bloc C2 tumor resection with anterior cage reconstruction via a transm andibular-transglossal approach (c,d). Following surgery, patient underwent proton beam therapy. (e) Two months later, he presented with purulent open drainage from his pharyngeal wound. There was an open wound with exposure of the anterior cage in the back of the throat, which required reoperation with fat graft and free ap closure.

Fig. 16.9a–k (opposite) Wound infection following en bloc sacrectomy for chondrosarcoma. (a–e) This is a 52-year-old woman who underwent en bloc sacrectomy with reconstruction for sacral chondrosarcoma, who presented 4 months following surgery with 3x1cm wound dehiscence and purulent drainage. (a, b) Preoperative T2-weighted MRI shows tumor in ltration throughout the sacrum as well as (c) paravertebral soft tissue dorsally and o midline. (d,e) She underwent instrumentation following sacrectomy. (f, g) T1-weighted MRI with

gadolinium shows enhancing uid collection, (f, g arrows), which was also noted on CT (h arrow). She required wound washout with removal of hardware, rotational ap closure, and long-term antibiotic therapy. Culture was positive for vancomycin-resistant enterococcus. (h, i) CT scans prior to hardware removal and (j, k) after hardware rem oval are shown. Despite removal of hardware, patient was able to walk 10 weeks following her initial surgery after prolonged recumbancy.

Complications and Their Avoidance

a

f

h

g

j

b

c

d

e

i

k

187

188

Chapter 16 pu r u len t d rain age, w h ich requ ired a w ou n d w ash out w ith rem oval of h ardw are, rotat ion al ap closu re, an d long-term an t ibiot ic th erapy. Cult u re w as p osit ive for van com ycin -resist an t en terococcu s. Despite rem oval of all h ardw are, th e p at ien t w as fu lly am bu lator y at 8 m on th s follow ing h er in it ial su rger y. In fect ion rates as h igh as 25 to 44% follow ing sacrectom ies h ave been rep or ted .15,17 In fect ion r isks are in creased w it h bow el in con t in en ce,16 greater exten t of resect ion (i.e., high sacrectom ies),16,17 and long surgery tim e.17 The m ost frequent organism s were Enterococcus (23%), Escherichia coli (20%), an d Pseudom onas aeruginosa (18%), an d approxim ately th ree out of four patients had m ultim icrobial infections.17 Alt h ough in fect ion s can be t reated w it h su rgical d ebr id em en t s an d long-ter m an t ibiot ic th erapy, alm ost h alf of pat ien ts requ ire m u lt ip le su rgical debridem en ts: 52.2% of pat ien t s w ere t reated w ith on e su rgical debridem en t , w h ereas 34.8% an d 13.0% of p at ien t s requ ired t w o an d th ree surgical debridem en ts, respect ively.17 Moreover, in fect ion s can length en th e h ospital stay, delay healing, and cause pseudarthrosis w ith h ardw are failu re (Fig. 16.6). Longterm w ou n d follow -u p is recom m en ded, given th at 30% of w oun d in fect ion s occur 4 w eeks to 6 m on th s after th e sacrectom y 17 ; th e rem ain ing 70% of in fect ion s occu rred w ith in 4 w eeks of su rger y. Th e prolonged u se of n egat ive pressu re drain s an d esp ecially in t raw ou n d van com ycin p ow der, w h ich h as sh ow n sign i can t ben e t in oth er areas of sp in al surger y, sh ould be st rongly con sidered.

■ Chapter Summary Careful preoperative surgical planning and prudent patient selection are absolute requirem ents

for m in im izing com plicat ion s an d m a xim izing ou tcom es of prim ar y spin e t um or su rger y. Th is p lan n ing st ar t s w ith a th orough im aging evalu at ion follow ed by a biop sy for p at h ological d iagn osis of th e t um or. Alth ough m align an t t u m ors m ay requ ire en -bloc resect ion w ith w ide m argin s for th e best ou tcom e, ben ign t u m ors m ay be am en able to in t ralesion al p iecem eal rem oval. How ever, t h e m orbidit y of en -bloc resect ion sh ou ld alw ays be w eigh ed again st p oten t ial n eu rologic de cit s an d ot h er m ajor com p licat ion s involved w ith exten sive w ide m argin resect ion . Som e su rgeries m ay requ ire m u lt idiscip lin ar y app roach es w ith su rgeries at large referral cen ters w ith exten sive exp erien ce for th e best outcom e.

Pearls Understand the di erence bet ween wide-m argin en-bloc resection and intralesional piecemeal resection and when each should be utilized. Utilize di erent imaging techniques (i.e., standing scoliosis X-rays, CT, and MRI) for thorough preoperative evaluation. Proper surgical planning requires biopsy for pathological diagnosis. Understand the di erent potential complications related to tum or location. Understand the di erent instrumentation techniques to prevent hardware failure, alignment failure, and pseudarthrosis. Pitfalls Wide-margin en-bloc resection requires the same principles of deformit y surgery to provide solid stabilization with preservation or restoration of global and local spinal balances. Complications such as hardware failure, wound infection with dehiscence, and tum or recurrence may be delayed, requiring long-term follow-up with serial imaging.

Refere nces Five Must-Read Refe rences 1. Bor ian i S, Ban d iera S, Don t h in en i R, et al. Morbidit y of en bloc resect ion s in the spin e. Eur Spin e J 2010;19:231–241 PubMed

2. Boriani S, Wein stein JN, Biagin i R. Prim ar y bon e t um ors of th e spin e. Term in ology an d surgical st aging. Spin e 1997;22:1036–1044 Pu bMed

Complications and Their Avoidance 3. Tom it a K, Kaw ah ara N, Baba H, Tsuchiya H, Fujit a T, Toribat ake Y. Tot al en bloc spon dylectom y. A n ew su rgical tech nique for prim ar y m align an t ver tebral t um ors. Spin e 1997;22:324–333 Pu bMed 4. Borian i S, De Iure F, Cam pan acci L, et al. An eur ysm al bon e cyst of th e m obile spin e: report on 41 cases. Spin e 2001;26:27–35 PubMed 5. Papagelopou los PJ, Cu rrier BL, Sh augh n essy W J, et al. Aneur ysm al bon e cyst of th e spine. Man agem en t an d outcom e. Spin e 1998;23:621–628 PubMed 6. Zileli M, Kilin çer C, Ersahin Y, Cagli S. Prim ar y t u m ors of the cer vical spin e: a ret rospect ive review of 35 surgically m an aged cases. Spin e J 2007;7:165–173 PubMed 7. Fourn ey DR, Rh in es LD, Hen t sch el SJ, et al. En bloc resect ion of prim ar y sacral t um ors: classi cat ion of su rgical approach es an d outcom e. J Neu rosu rg Sp in e 2005;3:111–122 PubMed 8. Borian i S, Ban diera S, Biagin i R, et al. Chordom a of th e m obile spin e: ft y years of experien ce. Spin e 2006; 31:493–503 PubMed 9. Borian i S, De Iure F, Ban diera S, et al. Ch on drosarcom a of th e m obile spine: repor t on 22 cases. Spine 2000;25:804–812 PubMed 10. Cloyd JM, Ch ou D, Deviren V, Am es CP. En bloc resect ion of prim ar y t um ors of th e cer vical spin e: report of t w o cases an d system at ic review of th e literat ure. Spin e J 2009;9:928–935 Pu bMed 11. Bor ian i S, Ban d iera S, Casad ei R, et al. Gian t cell t u m or of t h e m obile sp in e: a review of 49 cases. Spin e 2012;37:E37–E45 Pu bMed 12. Roy- Cam ille R, Saillan t G, Bisserié M, Ju d et T, Hautefor t E, Mam oudy P. [Tot al excision of th oracic ver te-

brae (au th or’s t ran sl)]. Rev Ch ir Or th op Rep ar Ap par Mot 1981;67:421–430 Pu bMed 13. Sten er B. Tot al spon dylectom y in chon drosarcom a arising from th e seven th th oracic vertebra. J Bon e Join t Su rg Br 1971;53:288–295 PubMed 14. Tom it a K, Kaw ah ara N, Baba H, Tsu ch iya H, Nagat a S, Toribat ake Y. Tot al en bloc spon dylectom y for solit ar y spin al m et ast ases. In t Orth op 1994;18:291–298 PubMed 15. Clarke MJ, Dasen brock H, Bydon A, et al. Posterioron ly approach for en bloc sacrectom y: clin ical outcom es in 36 consecutive patients. Neurosurgery 2012; 71:357–364, discu ssion 364 PubMed 16. Guo Y, Palm er JL, Sh en L, et al. Bow el an d bladder con t in en ce, w oun d h ealing, an d fu nct ion al ou tcom es in pat ien t s w ho un der w en t sacrectom y. J Neu rosu rg Spin e 2005;3:106–110 PubMed 17. Ruggieri P, Angelin i A, Pala E, Mercuri M. In fect ion s in surger y of prim ar y t um ors of th e sacru m . Spin e 2012;37:420–428 PubMed 18. Jackson RJ, Gokaslan ZL. Sp in al-p elvic xat ion in p at ien t s w ith lu m bosacral n eoplasm s. J Neurosu rg 2000;92(1, Su p p l):61–70 PubMed 19. Hugate RR Jr, Dickey ID, Ph im olsarn t i R, Yaszem ski MJ, Sim FH. Mech an ical e ect s of p ar t ial sacrectom y: w h en is recon st ru ct ion n ecessar y? Clin Or th op Relat Res 2006;450:82–88 Pu bMed 20. Om eis IA, Dh ir M, Sciubba DM, et al. Postoperat ive surgical site in fect ion s in pat ien t s u ndergoing spin al t um or surgery: incidence and risk factors. Spine 2011; 36):1410–1419 PubMed

189

Index

Page n um bers follow ed by f or t in dicate gu res an d t ables, resp ect ively. A Adjuvan t th erapy, 13, 55–56. See also under specif c t um ors de n it ion of, 55 t u m or resp on siven ess to, 128–129 Allografts, 158, 160, 161 An esth esia, 18 An eur ysm al bon e cyst-like ch anges, 71, 73–74, 75, 77, 82 An eur ysm al bon e cyst s, 71–74, 73–74f ch em oth erapy for, 78, 82, 83 clin ical presen t at ion of, 71 de n it ion of, 71 di eren t iated from gian t cell t u m ors, 73 em bolizat ion of, 32, 78, 78f, 83 en -bloc resect ion of, 74, 83, 131, 178 h istology of, 71, 72, 73 im aging of, 2, 72f, 77 in t ralesion al excision of, 73–74, 77, 82, 131 m edical t reat m en t of, 131 m isdiagn osis of, 82 radiat ion th erapy for, 74, 78 recu rren t , 74, 83 Angiograp hy during em bolizat ion , 33 pre-em bolizat ion , 29–31, 30f, 32, 33 preop erat ive, 11, 139, 176 Angiosarcom a, 71, 73 An terioposterior approach , to th oracolu m bar t um ors, 138 An terior approach , 12f, 92 An terior colum n recon st ru ct ion , 158 An terior rst , posterior secon d ap p roach , 7, 9, 10, 10f, 11, 11f An terior-on ly approach , 7 An terior spin al colu m n xat ion (ASCF), 149, 151, 151f, 152

An terolateral ap proach , 26f Aort ic dissect ion , 11, 179 Arter y of Adam kiew icz in en -bloc resect ion s, 10–11 in adverten t em bolizat ion of, 78 B Balloon -occlu sion test , 31, 139 Ben ign t u m ors, st aging of, 3–4, 3f, 4f Bevacizum ab, 56–57 Bilateral bu lar graft (BFFR) tech n iqu e, of sp in o pelvic xat ion , 151, 151f, 152 Biological e ect ive dose (BED), 41–42, 42f Biological t issu e m at rices (BTMs), 168 Biop sy of bon e m arrow, 54, 55 frozen sect ion , 2–3 im age-gu ided, 12–13, 19–20, 23–27, 137 com plicat ion rate of, 51t diagn ost ic accuracy of, 51, 51t locat ion for p erform an ce of, 176, 178 op en /in cision al, 13–14, 19, 51, 137 Biop sy site, in clu sion in t u m or excision s, 20, 176 Bleom ycin , toxicit y of, 57 Bon e cem en t , 27, 28–29, 158 Bon e graft s, 181 select ion of. See St ru ct u ral graft select ion Bon e m orp h ological protein (BMP), 158 Bon e scan s, 2, 4, 54, 62, 69, 112 Brachyth erapy, 40, 41, 41f Brachyury gen e, as ch ordom a th erap eu t ic t arget , 88, 95 Bragg peak e ect , 36, 37, 37f, 45 Breast can cer, 52, 52t, 53, 128f N-bu t yl cyan oacr ylate (NBCA), 31

192

Index C Cages, for spinal recon st ru ct ion , 158, 160, 162–163, 178, 179 carbon - ber, 100 recon st ruct ion /rep osit ion ing of, 180, 184, 186f rib graft s adjacen t to, 181 subsiden ce of, 163 Calciton in , as an eu r ysm al bon e cyst t reat m en t , 131 Carbon ion radioth erapy, 37, 46 Carot id ar ter y biopsy-related injur y to, 23, 26 occlusion of, 31 Cen t ral ven ous pressu re m on itoring, 18 Cerebrospin al uid leaks, 36 Cer vical n er ve root s, resect ion of, 139 Cer vical spin e, radiosen sit ivit y of, 44 Cer vical spin e t um ors em bolizat ion of, 31 en -bloc resect ion of, 138–141, 142, 179, 181 Cer vicoth oracic jun ct ion graft s for, 161 revision s near, 183 w oun d closu re at , 183 Cet u xim ab, 95 Ch em oth erapy. See also under specif c t um ors toxicit y of, 56–57, 119 t um or respon siven ess to, 128–129 Ch ildren , Ew ing’s sarcom a in , 114–115, 117, 118, 121 Ch on droblastom a, secon dar y an eu r ysm al bon e cyst-like ch anges in , 71 Ch on drosarcom a, 97–109 adjuvan t th erapy for, 103, 107, 108 biopsy of, 105 incorrect tech n iqu e in , 106, 108 cen t ral, 97, 98–99, 98–101f, 101, 102 cer vical, 38–39, 46, 103–104f, 106 ch em oresist an ce of, 45–46, 102 clear-cell, 97, 130f clin ical presen t at ion of, 101 dedi eren t iated, 97, 106 de n it ion of, 97 diagn osis of, 105 en -bloc resect ion of, 35–36, 103–104f, 178 an atom ic sacri ces in , 105–106 w ith con t am in ated su rgical m argin s, 17 inappropriate, 107 incom plete, 17 versus in t ralesion al excision , 106–107, 108 long-term resu lt s of, 16, 17 w ith ou t biopsy, 105 epidural exten sion of, 99–100 5-year sur vival rate, 103, 105 im aging of, 98–101f, 101–102 in ten sit y-m odu lated radioth erapy for, 40 in t ralesion al excision of, 17, 102–103 in appropriate, 106–107, 108 of th e lim bs, 102 lu m bar, align m en t failu re in , 182–183, 184

m align an t t ran sform at ion of, 103 m esen chym al, 97 m isd iagn osis of, 106, 108 of th e m obile spin e, 17, 130 path ology of, 105 pelvic, 130 periph eral, 2, 97, 101, 101f proton beam radiat ion th erapy for, 17, 46 proton /p h oton beam radiat ion th erapy for, 38–39 radiat ion th erapy for, 16, 17, 106, 107, 108 radioresist an ce of, 45–46, 102 recu rren t , 17, 35–36, 103, 130 reoperat ion on , 130 sacral, 101, 106 secon dar y, 97, 101, 105 of th e sku ll base, 38–39, 46, 130 stereot act ic radiosu rger y for, 17, 42, 44 Ch ordom as, 85–96 adjuvan t th erapy for, 94, 95, 129 w ith bilateral vertebral arter y involvem en t , 162f biopsy of, 88 biopsy-related sp read of, 23 cer vical, 85, 89–90 en -bloc resect ion of, 16, 139–141, 142f, 163, 177f proton /ph oton radiat ion th erapy for, 38–38 proton radiat ion th erapy for, 46 ver tebral arter y ligat ion in , 176, 177 cer vicoth oracic, 86–88, 87f clin ical presen t at ion of, 85–86 at cran iocer vical ju n ct ion , 90, 94 de n it ion of, 85 diagn osis of, 88–89 en -bloc resect ion of, 44, 86–90, 178 w ith adjuvan t rad iat ion th erapy, 44–45 con t am in ated su rgical m argin s in , 17, 129 in com plete, 17 long-term resu lts of, 17 epid u ral exten sion of, 126f gen et ics of, 88–89, 95 h istop ath ology of, 88 im aging of, 86 in ten sit y-m odulated radioth erapy for, 40 in t ralesion al excision of, 17, 90 lu m bar, 90, 91, 91f m etast at ic, 85 in m obile sp in e, 17 proton beam radiat ion th erapy for, 17, 46 proton /p h oton beam radiat ion th erapy for, 38 radiat ion th erapy for, 16, 17 radioresist an ce of, 94 recu rren t , 20, 44, 129 sacral, 90, 92–93 carbon ion th erapy for, 37 clin ical presen t at ion of, 85–86 sacrectom y of, 90, 92, 93, 93f, 129f, 150f, 181, 182 spin opelvic xat ion of, 150 surgical ap p roach to, 92

Index surgical m argin s for, 125 ven t ral ext raosseou s exten sion of, 125f salvage th erapy for, 94 of th e skull base, 90 adjuvan t ch em oth erapy for, 94–95 proton beam radiat ion th erapy for, 45, 46 proton /ph oton rad iat ion th erapy for, 38–39, 94 recu rren t of, 45 st aging of, 140–141, 140f stereotact ic radiosu rger y for, 17, 42, 44, 45, 178, 180 th oracic, 85, 86–88, 178, 180f th oracolum bar, 90 t um or seeding of, 129 ven t ral exten sion in to th e p elvis, 125 Clin ical t arget volum e (CTV), 42 Closed loop tech n ique, of spin opelvic xat ion, 149, 149f, 154f Com bin ed an terior-posterior app roach , 8, 9, 10, 11, 18, 185 Com plicat ion s, of prim ar y sp in al t um or su rger y, 11–12, 13, 17–18 h ardw are con st ruct an d align m en t failure, 181–183, 183f, 184f in t raoperat ive, 18–19 p reven t ive st rategies for, 18–19, 19t, 175–189 Com pu ted tom ography (CT), 2 of ben ign spin al t um ors, 4 p reop erat ive, 176 p rior to vertebral augm en t at ion , 28 for t um or st aging, 53–54 Com pu ted tom ography u oroscopy, 24, 25, 26, 27, 28, 28f Coron al balan ce, 182–183 Cor t icosteroids, as an eu r ysm al bon e cyst t reat m en t , 131 Craniocer vical ju n ct ion , ch ordom as of, 90, 94 Cu ret t age. See In t ralesion al excision Cysts, an eur ysm al bon e. See An eur ysm al bon e cysts D Dead space, 139, 172–173, 183 Debulking. See In t ralesion al excision Deep ven ous th rom bosis, 18 Den osum ab, 77, 82, 83, 131 Diagn osis, of prim ar y spin al t u m ors, 12–13, 50, 176, 178. See also Biopsy ancillar y test ing for, 51–53 Di eren t ial diagn osis, of prim ar y spin al t um ors, 137 Dural tears, 18–19 Dysph agia, postop erat ive, 179 Dysplasia, brous, 1, 71, 111 E Ed em a, 111–112 E cien cy, of prim ar y spin al t um or su rger y, 15–22, 16t Em bolism , pulm on ar y, 18, 29

Em bolizat ion . See also under specif c t um ors palliat ive, 32 preoperat ive, 18, 29–32, 139 th erapeu t ic, 29, 32, 72 En -bloc resect ion com bin ed w ith p roton beam radioth erapy, 38 con t rain d icat ion s to, 36 de n it ion of, 7, 61, 89, 127–128, 136 e ect on h ealth -related qu alit y of life, 20 gross tot al, 89 in st ru m en t at ion failu re rate in , 163 m ort alit y rate of, 18 w ith n egat ive su rgical m argin s, 129 n on un ion follow ing, 163 ou tcom e of, 35–36 prin cip les of, 138–139 radical, 6–7, 89 surgical m argin s of. See Margin s, su rgical tech n iques in , 7–11 vit al an atom ic st ruct u re sacri ces in , 7–8 En ch on d rom as, 97, 101, 106 En ch on d rom atosis, 111 En n eking st aging system , 13, 16, 61, 63, 63t, 66, 124, 126, 134–135, 135t, 143, 145 Epiderm al grow th factor receptor in h ibitors, 94–95 Epiderm al grow th factor receptor m u t at ion s, 52–53 Epidu ral t u m ors, 40, 99–100, 101, 126, 137 Erlot in ib, 53, 94–95 Ew ing’s sarcom a, 114–119 adjuvan t/n eoadjuvan t th erapy for, 55, 56, 118, 119, 121 biop sy of, 54 ch em oth erapy for, 56, 117–118, 119, 129 clin ical presen t at ion of, 115 d iagn osis of, 116 d i eren t ial d iagn osis of, 54t, 97, 102, 113 en -bloc resect ion of, 118–119, 118f, 178 epid em iology of, 114–115 5-year su r vival rate in , 117–118, 119 gen et ics of, 116, 117 h istopath ology of, 115–116 im aging of, 2, 115, 116, 116f, 117, 117f, 118 m an agem en t goals for, 110 m edical th erapy for, 137 m et ast at ic, 55, 110, 115, 116–117, 119 radiat ion th erapy for, 16, 118, 129 radiosen sit ivit y of, 119 recu rren t , 55, 119 sacral, 115 st aging of, 53–54, 54t, 116–117 Exostoses, 1, 97, 101, 101f, 102, 103f, 111 F Fibu lar bon e aps, 171–172 Fibu lar graft s, 158 Flaps adju n ct s to, 167–168 axial p at tern , 166, 167, 168–172 failu re of, 173

193

194

Index Fluoroscopy, 24, 25, 26, 27, 28 Four-rod recon st ru ct ion , 149, 149f, 183, 185 Fract u res osteoporot ic, 27, 29 path ological, 1, 27–29, 73, 75 rod, 181, 183 sacrectom y-related, 181, 182f vertebral augm en t at ion t reat m en t for, 27–29 G Galveston L-rod tech n iqu e, 148–149, 149f Ge t in ib, 53, 95 Gian t cell t um ors, 73, 79–83 an eur ysm al bon e cyst-like ch anges w ith in , 71, 73–74, 73–74f, 75, 77, 82 biopsy of, 82 ch em oth erapy for, 79, 82, 83 de n it ion of, 74–75 di eren t iated from an eu r ysm al bon e cysts, 73 em bolizat ion of, 75, 82 en -bloc resect ion of, 75–76f, 77, 79, 82, 83, 130–131, 178, 179 im aging of, 2, 73–74f, 75–76f, 76, 77 in t ralesion al excision of, 79, 80–81f, 82–83, 130 of th e lim bs, 79 m align an t t ran sform at ion of, 130 m edical t reat m en t for, 130 m isdiagn osis of, 82 pu lm on ar y m et ast ases from , 75, 79 radiat ion th erapy for, 79, 80, 82 recurren t , 79, 82–83, 83, 130–131, 178, 179f Gluteal arter y p erforator aps, 171f Gluteu s m axim us m uscle aps, 170–171 Granu lom as, eosin oph ilic, 80, 129 Gross tot al resect ion . See In t ralesion al excision Gross t u m or volu m e (GTV), 42 H Hadron s, 36 Hem angiom as, 1, 27–29, 28f, 32, 131–132 Hem atom as, surgical w ou n d-related, 173 Hem iver tebrectom y, 114f, 120 Hem orrh age bevacizum ab -related, 56–57 in t raoperat ive, 18, 19, 74 osteosarcom a-related, 112 Histocytosis, Langerh an s. See Gran u lom as, eosin oph ilic Histological diagn osis, of p rim ar y sp in al t u m ors, 137 Histological m argins, di eren t iated from su rgical m argin s, 61 Hyper vascular t um ors, preop erat ive em bolizat ion of, 18, 29–32

I Im aging. See also Com p u ted tom ograp hy (CT); Magn et ic reson an ce im aging (MRI); Posit ron em ission tom ograp hy (PET) im p ort an ce of, 50 p reoperat ive, 166, 176 Im at in ib, 94, 95 In fect ion s of su rgical sites, 11, 165, 173 as vertebral augm en t at ion con t rain dicat ion , 27 In form ed con sen t , 175 In st ru m en tat ion advan ces in , 178 failu re or revision of, 19, 163, 183–184 In ten sit y-m odu lated rad iat ion th erapy (IMRT), 36–38, 39–40, 41, 42 In tercost al arteries, 10 In ter ven t ion al opt ion s, for prim ar y sp in al t u m ors, 23–34 im age-gu id ed biop sy, 23–27, 32, 33 p reoperat ive an d th erap eu t ic em bolizat ion , 29–32 vertebral augm en t at ion , 27–29, 32, 33 In t ralesion al excision , 128. See also under specif c t um ors de n it ion of, 7, 61, 127, 136 t u m or recu rren ce associated w ith , 14, 19 Isch em ia, sp in al, 11, 31 J Joh n s Hopkin s Hosp it al (JHH) m eth od, of sp in opelvic xat ion , 149, 151, 152 K Kyp h op last y, 27, 32 Kyp h osis, cer vicoth oracic, 176 L Lat issim u s dorsi m u scle ap s, 169–170, 170f Liposarcom a, 42 Lu m bar-pelvic defect s, recon st ru ct ion of, 160–161 Lu m bar spin e, p reop erat ive im aging of, 176 Lu m bar sten osis, in ciden ce of, 1 Lu m bar t u m ors, su rgical ap proach es to, 9–10 Lu m bop elvic recon st ru ct ion , 92, 93 Lu m bosacral jun ct ion t u m ors, 147 Lu ng can cer, 51–53, 52t Lym p h om a, st aging of, 53, 54 M Magn et ic reson an ce u oroscopy, 24 Magn et ic reson an ce im aging (MRI), 2 of ben ign sp in al t u m ors, 4 preoperat ive, 176 prior to vertebral augm en t at ion , 28 for t u m or st aging, 53–54 Malign an t t u m ors, st aging of, 4–5, 4f, 5f

Index Margin s, surgical, 3, 5, 136–137. See also En n eking st aging system ; Wein stein -Borian i-Biagin i st aging system accu rate descript ion s of, 6 con tam in ated, 18, 20 de n it ion of, 124–127 determ in at ion of, 13, 14, 20 di eren t iated from h istological m argin s, 61 e ect iven ess of, 7 e ect on en -bloc resect ion e cacy, 16–17 h istology-speci c, 128 in t ralesion al, 7, 16–17, 61, 89, 132, 136–137 m argin al, 7, 16–17, 61, 89, 124, 125, 132, 137 n egat ive, 129 posit ive, 124–125 radical, 124, 125, 132, 137 relat ion sh ip to p rogn osis, 5 relat ion sh ip to surgical tech n ique, 127, 132 role in surgical plan n ing, 5–6 select ion of, 128 t ran sgression of as n eurologic de cit cause, 137, 144–145 plan n ed, 7, 137 u n n ecessarily com prom ising, 14 w ide, 7, 16–17, 61, 89, 124, 125, 132 Medical on cology prin ciples, for th e spin e on cology surgeon , 50–59 adjuvan t/n eoadjuvan t th erapy, 50, 55–56, 57–58 diagn osis, 50, 51–53, 57 pat ien t’s preferen ces, 50, 57, 58 st aging, 50, 53–55, 57, 58 toxicit y of system ic th erapy, 50, 56–57, 58 Melan om a, uveal, 37–38 Metal art ifact s, 160 Metal im plan t s, im plicat ion for radiat ion th erapy, 37 Metast ases in ciden ce of, 1 pain m an agem en t of, 27 pulm on ar y, 75, 79, 117 secon dar y an eur ysm al bon e cyst-like ch anges in , 71, 73 Microcoils, 31 Mu lt idisciplin ar y team s, 13, 137–138, 144–145, 179 Mu lt iple m yelom a, 53, 54 Myelit is, radiat ion , 40 Myelopathy, 27, 74, 85 Myo-osseous vascularized bular st ru t s, 162, 163 N Necrosis, 31, 32, 55, 112, 115, 173 Negat ive p ressure-assisted w ou n d closu re, 167–168 Neoadjuvan t th erapy, 137–138. See also under specif c t um ors de n it ion of, 55 w ith h olding of, 55–56 Neu rologic de cits ch ordom a-related, 85–86 Ew ing’s sarcom a-related, 115

osteosarcom a-related, 111 prim ar y spin al t u m or-related, 3 sacrectom y-related , 90, 92 surgical m argin t ran sgression -related, 137, 144–145 Non steroidal an t i-in am m ator y d rugs, as osteoid osteom a t reat m en t , 61–62, 66 O Obser vat ion , of asym ptom at ic t u m ors, 176 Occip itocer vical jun ct ion , graft s for, 161–162 Onyx, 31 Osteoblastom as act ive, 63, 64, 66, 68, 69 aggressive, 63, 64, 66, 68, 69 an eur ysm al bon e cyst-like ch anges in , 71 clin ical presen t at ion of, 60, 63 de n it ion of, 60 di eren t iated from osteosarcom as, 64, 131 gian t , 131f h istology of, 64 im aging of, 63–64, 65–68f m align an t degen erat ion of, 64, 69 recu rren t , 60 as scoliosis cau se, 2 Osteoch on drom as, 97, 101 di eren t iated from ch on drosarcom as, 102, 106 Osteoid osteom as clin ical presen t at ion of, 60, 61–62 de n it ion of, 60, 131 im aging of, 2, 61f, 62, 69 as scoliosis cau se, 2 t reat m en t for, 61–63, 66 Osteom yelit is, 171–172 Osteosarcom as, 111–114 adjuvan t th erapy for, 56, 113 ch em oth erapy for, 56, 113, 129 clin ical presen t at ion of, 111 de n it ion of, 111 di eren t iated from osteoblastom as, 64, 131 en -bloc resect ion of, 16, 64, 69, 113–114, 129, 178 epid em iology of, 111 5-year su r vival rate in , 114 h istop ath ology of, 112–113 im aging of, 111–112 in ten sit y-m odulated radioth erapy for, 30, 39f m an agem en t goals for, 110 m et ast at ic, 55, 110, 111, 112, 113 n eoadjuvan t th erapy for, 55, 113, 121, 129 palliat ive th erapy for, 114 progn osis for, 129 proton /p h oton beam th erapy for, 38 radiat ion th erapy for, 16, 114 radioresist an ce of, 119 recu rren t , 64, 69 sacral, 111 sagit t al, 8–9, 11

195

196

Index Osteosarcom as (cont inued ) secon dar y an eur ysm al bon e cyst-like ch anges in , 71, 73 st aging of, 53, 54, 113 tot al spon dylectom y of, 113 Osteotom y, 8–9, 9f, 11, 11f, 178, 180 Outcom e, factors a ect ing, 178–181 P Paget’s disease, 79, 80–81, 111 Palliat ive procedures, 7, 32, 114 Parasp in ous m u scle aps, 168–169, 169f Pat ien t s, involvem en t in t reat m en t decision m aking, 13, 57, 58 Pelvic in ciden ce, postop erat ive, 176 Pelvis, sacral ch ordom a exten sion in to, 125 Periph eral n er ve sh eath t u m ors, 126f Pert u zum ab, 53 Ph ar yngeal w all, recon st ru ct ion of, 161–162 Ph ar yngeal w oun ds, 184, 186 Physician -pat ien t relat ion sh ip , 13, 57, 58 Piecem eal excision . See In t ralesion al excision Piriform is m uscles, t um or recu rren ce in , 129, 129f Plan n ing t arget volum e (PTV), 42 Plasm acytom as, 54–55, 129 Plasm a-m ediated t u m or ablat ion , 28–29, 32, 33 Pn eum oth orax, 23 Polym ethylm ethylacr ylate (PMMA), 27, 158 indicat ion s for u se of, 159–160 inject ion tech n iqu es for, 28–29, 33 leakages of, 29 plasm a-m ediated t um or ablat ion of, 28–29, 32, 33 Polyvinyl alcoh ol (PVA) part icles, 31 Posit ron em ission tom ograp hy (PET), 2 u orodeoxyglucose (FDG), 53–55 Postem bolizat ion syn drom e, 32 Posterior approach , 12f, 25f to sacral ch ordom as, 92 to th oracolum bar t u m ors, 138 Posterior xat ion , sh ort-segm en t , 163 Posterior-on ly surgical ap p roach es, 7, 8, 10 Posterior pelvic ring xat ion (PPRF), 149, 151, 151f Posteroan terior ap proach , to th oracolu m bar t um ors, 138 Posterolateral approach , in vertebral augm en t at ion , 28 Preoperat ive plan n ing, 175. See also Surgical plan n ing Prim ar y spin al t u m ors clin ical presen t at ion of, 1–2 di eren t iated from m etast at ic t um ors, 1 im aging of, 1–2 inciden ce of, 1 Prim it ive n euroectoderm al t u m ors (PNETs), 161f Proton beam radiat ion th erapy, 36–39, 38f. See also under specif c t um ors com bin ed w ith t um or resect ion , 45

com parison w ith in ten sit y-m odu lated radioth erapy, 36–37, 40 as w oun d in fect ion risk factor, 184 Proton /p h oton beam rad iat ion th erapy, 38–39, 94 Pseu doar th rosis, 154, 181, 182, 183, 184, 188 Q Qu alit y of life, h ealth -related, 20 R Radiat ion th erapy, 35–49. See also under specif c t um ors adverse e ect s of, 36, 74–75, 111, 118, 119, 160, 166, 184 biological e ect ive d ose (BED) of, 41–42, 42t conven t ion al extern al beam , 43 e ect on su r vival t im e, 16 hyp ofract ion ated, 46–47 t u m or resp on siven ess to, 128–129 for vertebral fract u re-related pain , 27 Radiculom edullar y arter y angiography of, 29, 30, 30f in em bolizat ion , 29, 30, 30f, 31, 32 preoperat ive iden t i cat ion of, 11 Radiculop athy, cer vical, 1 Radiofrequ en cy ablat ion , 61, 61f, 62–63, 66 Radioresist an ce, of prim ar y spin al t u m ors, 36 Recon st ru ct ion procedu res spin opelvic. See Sp in opelvic recon st ru ct ion / xat ion an d fusion , after sacral t um or resect ion of surgical w ou n d s. See Wou n d closu re/ recon st ru ct ion tech n iqu es Rect u s abdom in al m u scle aps, 171, 172f Resp irator y failu re/d ist ress, p ostop erat ive, 18, 179 Resp on se Evaluat ion Criteria in Solid Tu m ors (RECIST), 94, 95 Ret rop h ar yngeal su rgical ap proach , 184 Revision su rger y, 163 Rh om boid fasciocu t an eou s aps, 171f Rib grafts, 158, 160, 162, 181 S Sacral t um ors. See also Ch ordom as, sacral; Sacrectom y; Sp in op elvic recon st ru ct ion / xat ion an d fu sion , after sacral t u m or resect ion bow el an d bladder con t in en ce preser vat ion in , 139 Ew ing’s sarcom a, 115 ap closu re tech n iqu es for, 18 osteosarcom as, 111 proton beam rad iat ion th erapy for, 38f radical resect ion of, 125–126 Sacrectom y, 90, 92, 93, 93f, 129f, 150f com plicat ions of, 18, 181, 182, 185–188 Sacrococcygeal region , rad iat ion toleran ce in , 44 Sacroiliac join t t u m ors, 141–143, 143f, 144, 144f Sacru m . See also Sacral t u m ors am p u t at ion s in , 127

Index Sagit t al balan ce, 148, 153, 154, 182–183, 184 Sarcom as. See also Ch on drom as; Ch on drosarcom as; Ew ing’s sarcom a ext rem it y, 36 h em ipelvic, 112f im aging of, 112f osteogen ic. See Osteosarcom a radiat ion -related , 74, 79, 82 radiat ion th erapy for, 42–44 Sciat ic pain , 86 Scoliosis, 1–2, 60, 62, 63, 69 Serom a, 173, 183 Single an terior approach , 7, 8 Single p osterior approach , 7, 8–9, 8f, 9f Sku ll base t um ors. See also Ch on d rom as, of th e sku ll base; Chon drosarcom as, of th e sku ll base stereotact ic radiosu rger y for, 44, 45 Spin al colum n , t u m or-free “ring” con cept of, 138, 144, 145 Spin al cord ch on drosarcom a en casem en t of, 130 radiat ion toleran ce of, 36 vascu larit y of, as surgical con siderat ion , 10–11 Spin al cord com pression , 32, 73, 101, 102, 102f, 118 Spin al cord injur y, 1, 36 Spin e On cology St udy Grou p (SOSG), 89, 94, 95, 113, 114, 119, 126 Spin opelvic recon st ruct ion / xat ion an d fu sion , after sacral t u m or resect ion , 19, 139, 147–156 after tot al sacrectom y, 148–152 align m en t failure in , 182–183, 184 biom ech an ical con siderat ion s in , 147–148, 152–153 h ardw are con st ruct failure in , 181–182, 183 problem s associated w ith , 152–154 spin opelvic fu sion , 152, 154 spin opelvic st abilizat ion p rocedures, 148–152 Spon dylectom y, 127f de n it ion of, 127 en -bloc, 127 360 °, 158 tot al en -bloc, 127, 163, 178 Spot scan n ing, 45 St aging, of prim ar y sp in al t u m ors, 3–5, 53–55. See also En n eking staging system ; Wein stein Borian i-Biagin i st aging system case st u dies of, 140–141 d e nit ion of, 53 in consisten t , 20 of m align an t t um ors, 4–5 t issue speci cit y of, 55 Stereotact ic radiosu rger y (SRS), 40–42 Cyberkn ife, 178, 180 hypofract ion ated, 43–44 for sarcom a, 43–44 single-fract ion , 43–44, 45 Steroid th erapy. See also Cor t icosteroids p rior to biopsy, 2–3

St roke, 23, 30, 32 St ru ct u ral graft select ion , 157–164 an atom ic-site con siderat ion s in , 160–163 gen eral factors in , 159–160 im p licat ion for recon st ru ct ion resu lt s, 163 pract it ion ers’ p referen ces regarding, 157–158, 158t, 159t Su bspecialist s, role in prim ar y spin al t u m or su rger y, 138 Su rgical ap proach es. See also specif c surgical approaches prin cip les for determ in at ion of, 134–146 case exam ples of, 139–144 h istological diagn osis, 137 m ult idiscip lin ar y team s, 137, 138, 144–145 surgical m argin s, 136–137 surgical p lan n ing, 137, 138, 144 Su rgical p lan n ing, 138–139, 144 surgical m argin s as con siderat ion in , 5–6, 7–8 Su rgical resect ion , of p rim ar y sp in al t u m ors. See also En -bloc resect ion ; In t ralesion al excision ; Sacrectom y long-term resu lt s of, 15–16 Su r veillan ce, Ep idem iology, an d En d Resu lt s (SEER) dat abase, 15–16, 85, 111, 114, 119 T Term in ology, of p rim ar y sp in al t u m or su rger y, 6–7, 20, 21, 127–128 Th oracic t um ors, surgical ap p roach es to, 9, 10 Th oracolum bar ju nct ion in en -bloc resect ion s, 10 graft s for, 161 revision s n ear, 183 Th oracolum bar t u m ors, en -bloc resect ion of, 138 Th oracotom y, in ch on d rom a resect ion , 87, 88 Tissu e exp an sion tech n iqu es, 168 “Tissu e is the issu e” prin ciple, 50, 51–53, 58 Tom it a saw s, 87, 87f, 88, 90, 178 Tom it a su rgical classi cat ion , of vertebral t u m ors, 124 Tran exam ic acid, 18 Tran sfu sion -related acu te lu ng inju r y (TRALI), 90 Tran sm an dibu lar-t ran sglossal app roach , 183–184, 186 Tran soral app roach , 25f Trapeziu s m u scle ap s, 169, 170f Trast u zu m ab, 53 Triangu lar fram e recon st ru ct ion (TFR), 149, 151 Trisacr yl m icrosph eres, 31 Tu m or seeding biopsy-related, 23 en -bloc resect ion -related, 36, 42 Tyrosin e kin ase in h ibitors, 57 V Vascu lar access, 18 Vascu lar en doth elial grow th factor (VEGF)in h ibitors, 57, 57t

197

198

Index Vertebrae, osteosarcom a of, 111 Vertebral arter y bilateral, 162f biopsy-related injur y to, 23, 26 em bolizat ion of, 139 ligat ion of, 176–177 occlusion test ing of, 31, 176 un ilateral resect ion of, 139 Vertebral augm en tat ion , 27–29, 28, 33 Vertebral bod ies ch on drosarcom a of, 102, 103–104 as in st rum en t at ion site, 176 t um or cont ain m en t w ith in , 125, 126 Vertebrectom y, of osteosarcom a, 114f Vertebrop last y, 27–29, 32 Vert ical rect us abdom in is m yocut an eous (VRAM) aps, 139 Vit al an atom ic st ru ct ures preser vat ion of, 13 sacri ce of, 7–8, 105–106

t u m or locat ion -related su rgical risk to, 175, 179, 181 W Wein stein -Borian i-Biagin i st aging system , 5–6, 7, 13, 16–17, 36, 124, 126, 136, 136f, 138, 139t, 143, 176 Wou n d closu re/recon st ru ct ion tech n iqu es, 165–174 adju n ct s to ap su rger y, 167–168 axial p at tern aps, 166, 167, 168–172 com plicat ions of, 173 for in fect ion p reven t ion , 183 m icrovascular free t issu e t ran sfer, 166–167 postoperat ive care for, 172–173 preoperat ive evalu at ion for, 165–16 Wou n d deh iscen ce, 179 X X-ray u oroscopy, 24 X-rays, of ben ign sp in al t u m ors, 4