ao spine masters series volume 4 adult spinal (bookzz.org)

7/23/2019 AO Spine Masters Series Volume 4 Adult Spinal (BookZZ.org)

http://slidepdf.com/reader/full/ao-spine-masters-series-volume-4-adult-spinal-bookzzorg 1/166



AOSpine Masters SeriesAdult Spinal Deformities



AOSpine Masters SeriesAdult Spinal Deformities

Series Editor:

Luiz Roberto Vialle, MD, PhDProfessor of Orthopedics, School of Medicine

Catholic University of Parana State

Spine Unit

Curitiba, Brazil

Guest Editors:

Law rence G. Lenke, MDJerome J. Gilden Distinguished Professor

Orthopaedic SurgeryProfessor, Neurological Surgery

Chief of Spinal Surgery

Director of the Advanced Deformity Fellowship

Washington University School of Medicine

St. Louis, Missouri

Kenne th M.C. Cheung, MBBS(UK), MD (HK), FRCS, FHKCOS, FHKAM(Orth)Head, Depart ment of Orthopae dics & Traumat ology

Jessie Ho Professor in Spine Surge ry

The University of Hong Kong

Queen Mary Hospita l

Pokfulam , Hong Kong

With 92 f gures

Thieme New York • Stut tgar t • Delh i • Rio d e Janeiro



Thiem e Med ical Publishers , Inc.

333 Seventh Ave.

New Yor k, NY 1000 1

Executive Editor: William Lamsback

Managing Editor: Sarah Land is

Director, Edito rial Services: Mar y Jo Casey

Editorial Assistant: Haley Paskalides

Product ion Editor: Barbara A. Cher nowInternational Production Director: Andreas Schabert

Vice President , Editorial an d E-Product Developm ent : Vera Spillner

International Marketing Director: Fiona Henderson

Int ern ational Sales Director: Louisa Tur rell

Director of Sales, Nort h Am erica: Mike Rosem an

Senior Vice President an d Chief Operating O cer: Sarah Vand erbilt

President: Brian D. Scanlan

Compositor: Carol Pierson, Chernow Editorial Services, Inc.

Library of Congress Cataloging-in-Publication Data

AOSpin e m aste rs se ries. v. 4, Adu lt sp inal d eform ities / ed itors , Luiz Robe rt o Vialle, Law ren ce G. Len ke,Ken net h M.C. Che un g.

p. ; cm.

Adu lt spinal deform ities

Includes bibliographical referen ces and index.

ISBN 978- 1-626 23-1 00-9 (alk. pap er) —ISBN 978-1 -626 23-10 1-6 (eISBN)

I. Vialle, Luiz Rober to , ed itor. II. Len ke, Law rence , 1960 – , ed itor. III. Che un g, Ken ne th M. C., ed itor.

IV. Title: Adult sp inal d eform ities.

[DNLM: 1. Spinal Diseases—sur ger y. 2. Orth ope dic Proced ure s—m eth od s. 3 . Spin e—sur ger y. WE 72 5]

RD768

617.4'71—dc23 2015 00197 9

Copyright ©2015 by Thieme Medical Publishers, Inc.

Impo rtant note: Medicine is an ever-changing science undergoing continual development. Research and

clinical experience are continually expand ing our knowledge, in par ticular our kn owledge of proper

treatm ent and d rug the rapy. Insofar as th is book ment ions any dosage or app lication, readers m ay rest

assured that th e authors, editors, and pub lishers have m ade every e ort to ensure th at such references are

in accordance w ith the state of knowle dge at the time o f production o f the book.

Nevert heless, th is does not involve, imply, or express any guaran tee or responsibility on the par t of th e

publ ish er s in re sp ect to an y d osage inst ruct ions a nd for m s o f ap plica t ions s tat ed in the b oo k. Every us er is

requested to exam ine carefully the ma nu facturers’ lea ets accompan ying each drug and to check, if

necessary in consultation w ith a physician or specialist, wheth er th e dosage schedu les ment ioned th erein or

the contraindications stated by the m anufacturers di er from th e statemen ts made in the present book.

Such examination is particularly importan t w ith drugs that a re eitherrarely used or have been n ewlyreleased on t he m arket. Every dosage schedu le or every form of application used is entirely at th e user ’s

own r isk and responsibility. The au thors an d pu blishers request every user t o report to th e publishers any

discrepan cies or inaccuracies noticed. If erro rs in this work are found afte r pub lication, errata w ill be post ed

at ww w.thieme.com on th e product description page.

Some of the produ ct nam es, patent s, and registered designs referred to in this book are in fact registered

tradem arks or proprietary nam es even th ough speci c reference to this fact is not always m ade in the text.

Therefore, the appearan ce of a nam e w ithout d esignation as proprietar y is not to be construe d as a

represent ation by the publisher that it is in th e public domain.

Printed in China by Everbest Print ing Ltd.

5 4 3 2 1

ISBN 978-1- 62623 -100- 9

Also available as an e-boo k:

eISBN 978-1-6 2623- 101-6



Volum e 1 Metastat ic Spina l Tum ors

Volum e 2 Prima ry Spinal Tum ors

Volum e 3 Cer vical Degen erat ive Cond itions

Volum e 4 Adu lt Spinal Deform ities

Volum e 5 Cer vical Spine Trau m a

Volum e 6 Thoracolumba r Spine Traum a

Volum e 7 SCI and Regeneration

Volum e 8 Back Pain

Volum e 9 Pediatric Spina l Deform ities

Volum e 10 Spinal Infect ion

AOSpine Masters Series

Luiz Robert o Vialle, MD, PhDSeries Editor



Series Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

Luiz Roberto Vialle

Guest Editors’ Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

Law rence G. Lenke and Kenneth M.C. Cheung

1 Preo pe rat ive Eva lu at io n and Op timizat io n for Su rge ry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1

Scott C. Wagner, Daniel G. Kang, Ronald A. Lehm an, Jr., and Law rence G. Lenke

2 Decision Making in Adu lt Deform ity Surgery:

Decom pression Ve rsu s Sh or t or Lon g Fu sion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Kenneth M.C. Cheung an d Jason P.Y. Cheung

3 Th e Use o f Ost eot omies for Rigid Sp in al De formit ie s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 8

Stephen J. Lew is and Sim on A. Harris

4 Indications and Techniques for Sacral-Pelvic Fixation in Adult Spinal Deform ity . . . . . . . . . 45

Kristen E. Jones, Robert A. Morgan , and Dav id W . Polly, Jr.

5 Instrumentation Strategies in Osteoporotic Spine: How to Prevent Failure? . . . . . . . . . . . . . 56

Ahm et Alanay an d Caglar Yilgor

6 The Inciden ce and Managem en t of Acute Neurologic Com plications

Follow in g Com p lex Ad ult Sp in al Defor m it y Su rge ry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 8

Joseph S. Butler and Law rence G. Lenke

7 Po st op erat ive Co ron al Deco mpensat io n in Ad u lt De formit y . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 8

Yong Qiu

8 Measu r in g Ou tco me and Va lu e in Ad u lt De formit y Su rge ry . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 5

Robert W aldrop an d Sigurd Berven

9 Ju n ct ion al Issu es Fo llow in g Ad ult Defo rmit y Su rge ry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 06

Han Jo Kim , Sravisht Iyer, and Christopher I. Shaf rey, Sr.

Contents



viii Contents

10 Biomechan ics and Mater ia l Science fo r Deformi ty Correc tion . . . . . . . . . . . . . . . . . . . . . . . . . 120

Manabu Ito, Yuichiro Abe, an d Rem el Alinga lan Salm ingo

11 Pseudarthrosis and Infect ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

Michael P. Kelly an d Sigurd Berven

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141



Spine care is advancing at a rapid pace. The

challenge for today’s spine care professional is

to quickly synthesize the best available evi-

dence and expert opinion in the man agem ent

of spine pathologies. The AOSpine Masters

Series p rovides just that —each volume in th e

series delivers pathology-focused expert opin-

ion on procedures, diagnosis, clinical wisdom,

and pitfalls, an d highlights today’s top rese arch

pap ers.To bring the value of its masters level edu-

cational courses and academ ic congresses to a

w ider au dience, AOSpine h as assem bled inter-

nationally recognized spine path ology leaders

to develop volumes in th is Masters Series as a

vehicle for shar ing their exper iences and expe r-

tise and providing links to the literatu re. Each

volume focuses on a current compelling and

somet imes cont roversial topic in spine care.

The unique and e cient format of the

Masters Series volumes quickly focuses the

attention of the reader on the core informa-

tion critical to understanding the topic, while

encouraging the rea der to look furth er into the

recomm ended literature.Through this approach, AOSpine is advanc-

ing spine care w orldwide.

Luiz Roberto Vialle, MD, PhD

Series Preface



Adu lt spin al deform ity (ASD) is a clinical prob-

lem of increasing prevalence, and thus physi-

cians and patients worldwide are aware of it.

With increasing longevity, the normal degen-

eration of the spine may lead to various ASD

prob lem s such as lum bar degenerat ive scolio -

sis with or without accompanying spinal ky-

phosis. In addit ion , ASD in clu des a sp ect rum

of preexistent childhood deformities, such as

scoliosis or kyphosis, that slowly progress tosymptomatic stages over adulthood. Clinical

manifestations may include progressive de-

form ity, poten tial spinal im balance, and spinal

stenosis, w ith resultant a xial or lower extre m -

ity symptomatology. Health-related quality-

of-life assessmen ts often d em onstrate severe

adverse e ects of ASD that can inter fere w ith

m any aspects of physical, emot ional, and psy-

chological well-being. Wh en clinical and rad io-

graphic scenarios warrant, surgical intervention,

ranging from simple decompressions to com-

plex total spine recon st ruct ion s, sh ou ld be con -

sidered in appropr iate patient s.

We have assembled a global panel of spe-

cialists to share w ith us the ir experience in the

m anagem ent of ASD, from evaluation to t reat-

ment, and including such issues as instru-

m ent ation and surgical techniques, as well as

prevent ing and m anaging complica t ions. Tho r-

ough patient evaluation, both m edical and sur -

gical, is warranted, with patient selection forindicated surgical intervent ion one of the m ain

keys to a successful outcome. Pertinent issues,

such as bone density evaluation and p reopera -

tive optimization, must be addressed w ith the

use of intraoperative adjuvants to ensure sta-

ble in ternal xat ion to the sp inal colum n in

pat ients requ ir ing st ab ilizat ion w ith or w ith-

out rea lignm ent . For pat ients w ith progressive

deformity producing segmental, regional, or

global malalignment, various corrective strat-

egies are discussed to safely realign the spinal

colum n u sing various forms of spinal osteoto-mies w ith adjuvant spinal instrum entation to

secure the spinal segments in their realigned

pos it ion. Spin al xat ion techniqu es are espe-

cially challenging whe n instr um ent ing the sac-

ropelvic unit in long constructs. The various

form s of osteotom ies utilized range from sim -

ple facet excis ions to ext rem ely co mplex thre e-

colum n osteotom ies such as pedicle subtr action

and vertebral colum n resection techniques that

are occasionally required for patient s w ith se-

vere deform ity w ith accompanying im balance.

Ensu ring n eu rologic safety du ring ASD surgery

is paramount, because these operations have

an early neurologic complication rate that is

not insigni cant and can lead to permanent

de cits. All of th ese essent ial pre ope rat ive and

intraope rative factors are discussed in det ail.

Even with initial surgical success, the long-

term success of surgery for ASD is controver-

sial. Variou s factors, such a s wo un d infection s,

pse udarthrosis, and adjacent segm ent pat hol-ogy, the most common being proximal junc-

tiona l kyph osis (PJK), can lead to d ete riorat ion

Guest Editors’ Preface



xii Guest Editors’ Preface

of the clinical outcomes over time. The dura-

bilit y of clin ica l outcom e m easures for these

pat ients is an im por tant focus along w ith the

nan cial implications for treating ASD patient s.

Thus, as in all areas of medicine, the value

prop os it ion o f tre at ing ASD pat ients, bot h w ithnonoperative and operative procedures, must

be as certained to just ify the w ide spect rum of

interventions available. As in all areas of sur-

gery, selecting the appropriate patient and per-

forming the least aggressive surgery to solve

the clinical problem w hile ensuring long-term

success is the opt imal app roach.

We hope this book will help spine surgeons

from around t he world navigate the often con-

troversial and complicated clinical issues in-

volved in th e m an agem en t of ASD pat ient s, so

that the outcome can be maximized and the

complications m inimized.

Law rence G. Lenke, MD

Kenneth M.C. Cheung, MBBS(UK), MD (HK),

FRCS, FHKCOS, FHKAM(Orth )



Yuichiro Abe, MD, PhD

Att en ding Spine Surgeon

Depar tm ent of Orth opaed ic Surgery

Eniwa Hospital

Eniw a, Jap an

Ahm et Alanay, MD

Professor

Departm ent of Orthoped ics andTraumatology

Facult y of Med icine

Acibadem University

Istan bu l, Turkey

Sigurd Berven , MD

Professor in Residen ce

Director of Spine Fellowsh ip an d Residen t

Edu cation Program


Univers ity o f Californ ia–San Fran cisco

San Fran cisco, California

Jose ph S. Butler, PhD, FRCS (Tr&Orth )

Clinical Fellow

Spina l Deform ity Unit

Royal Nation al Ort hop aed ic Hospita l

Stan m ore, Middlesex , United Kingdom

Jas on P.Y. Cheung, MBBS, MMedSc, FHKCOS,

FHKAM(Orth), FRCSEd(Orth )

Clinical Assistant Professor

Depar tm ent of Ort hopae dics & Traum atology

The University o f Hong Kong

Queen Mary Hospital


Kenneth M.C. Cheung, MBBS(UK), MD (HK),

FRCS, FHKCOS, FHKAM(Orth)Head, Depar tm ent of Orth opaed ics &

Traumatology

Jessie Ho Professor in Spine Surger y

The University o f Hong Kong

Queen Mary Hospital


Sim on A. Harris, MA, MB, BChir, FRCSC

Fellow

Departm ent of Orthoped ics

Toron to West ern Hospital, University of

Toronto

Toronto, Ontario, Canada

Manabu Ito , MD, PhD

Director

Cen ter for Spine a nd Spina l Cord Disorde rs

Nat ional Hosp it al Organizat ion Hok kaido

Medical Cen ter

Sapporo, Japan

Contributors



xiv Contributors

Sravish t Iyer, MD

Orthopaedic Surgery Resident

Hospital for Special Surgery

New York, New York

Krist en E. Jones , MDFellow

Depar tm ent s of Orth opaed ic Surgery and

Neuro su rgery

University of Minnesot a

Mineapolis, Minnesota

Daniel G. Kang , MD

Spin e Surgery Fellow

Depar tm ent of Orth opedic Surgery

Washington University

St. Louis, Missouri

Michael P. Kelly, MD, MSc

Assistant Professor of Ort hop ed ic Surger y

Assistant Professor o f Neurological Surgery

Depar tm ent of Orth opedic Surgery

Washingt on University School of Med icine

Sain t Louis, Missouri

Han Jo Kim , MD

Assistant Professor of Ort hop aed ic Surger yCo-Director of Edu cation

Spine Ser vice

Hospital for Special Surgery

New York, New York

Ronald A. Leh m an , Jr., MD

Professor of Ort hop aed ic Surgery

Professor of Neurological Surger y


BJC Instit ut e o f Health

St. Louis, Missouri

Law ren ce G. Lenke, MD

Jerome J. Gilden Distinguished Professor

Distinguished Professor of Orthopaedic Surgery

Professor of Neurological Surger y

Chief of Spin al Surger y

Directo r of the Advanced Deform ity

Fellowship


St. Lou is, Missour i

Step hen J. Lew is , MD, MSc, FRCSC

Associate Professor

University of Toront o

Depar tm ent of Surgery

Division of Ort hop aed ics

Toronto Western Hospital for SickChildren

Toronto, Ontario, Canada

Robert A. Morgan, MD

Assistant Professor

Orth opaed ic Surgeon


Minneapolis, Minnesota

David W. Po lly, Jr., MDProfessor and Chief

Spine Ser vice



Minneapolis, Minnesota

Yong Qiu , MD

Professor an d Direct or

Depar tm ent of Spine Surgery

Nan jing Drum Tow er Hosp it alMed ical Schoo l of Nanjing University

Nan jing, Jiangsu Province, China

Rem el Alingalan Salm ingo , PhD

Visiting Researcher

Biomedical Enginee ring

Tech nica l University of Den m ark (DTU)

Engineer

JJ X-Ray A/S

Tech nica l University of Den m ark (DTU)

Scion

Kongens Lyngby, Den m ark

Chris topher I. Shaf rey, Sr., MD

John A. Jan e Professor of Neuro logical

Surgery

Professor of Ort hop aed ic Surger y

Depar tm ent of Neurological Surgery

University of Virginia School o f

Medicine

Char lot tesville, Virgin ia VA



Contributors xv

Scot t C. Wagner, MD

Instr uctor of Surgery

Division of Surgery

Departm ent of Orthopaed ics

Uniform ed Services University of the Healt h

SciencesWalter Reed Nation al Militar y Medical Cen ter

Bethesda, Maryland

Robert Waldrop, MD

Fellow in Spine Surger y


Univers ity o f Californ ia–San Fran cisco

San Fran cisco, California

Caglar Yilgor, MD

Assistant Professor

Departm ent of Orthoped ics and

Traumatology

Facult y of Medicine

Acibadem UniversityIsta nb ul, Turkey



! Introduction

Adu lt spinal deform ity is an umbrella term en-

compassing various developm ent al, progres-

sive, or degen erative conditions t hat contribute

to an altered three-dimensional structure of

the h um an spine. There are t hree m ain types of

adult spinal deformity: t ype 1, de n ovo, or pr i-

m ary de generative scoliosis; type 2 , unt reated

adolescent idiopathic scoliosis that has pro-gressed into adu lthood; and t ype 3, secondary

scoliosis related to altered vertebral anatomy

due to previous surgery, trauma, or metabolic

bon e dise ase.1 A second ar y form of adult scoli-

osis is iatrogen ic imb alance cau sed by previous

spinal surger y.2 The m ost clinically imp orta nt

and m ost comm only encountered t ypes of adult

deformity are t ypes 1 and 3.1

Str uctu ral curves that develop in adulthood

(type 1) generally begin and then progress as

the intervertebral disks degenerate with nor-

mal aging. As disk degeneration progresses,

pos terior ele m ent incompet ence leads to axia l

rotation of the spinal motion segments, with

perm anent ro tator y deform it y in turn leading

to ligamentous laxity and eventual lateral lis-

thesis of the vertebral bodies.3 Destr uction of

the d iskoligam entou s comp lex and e nsuing de-

generation of the facet joints leads to ab norm al

m otion at each vertebral segment , subsequen tly

causing reactive chan ges such as osteophytosisat th e en d plates, facet joint hypert rophy/cysts,

and ligamentu m avum hypertrophy. In addi-

tion, the concavity of the m ajor and fractional

curve can cause foram inal narrowing, w hich is

often furthe r exacerbated by disk degeneration

and loss of foram inal height (up/dow n foram-

inal stenosis). These changes cause nar rowing

of the spinal cana l (cent ral and lateral recess)

and ne ural foramen ,1 and collectively contrib-

ute to t he clinical symptom s of adult scoliosis

or spinal deformity. Thus, understanding the

complex pathomechanics and anatomy of thisdegenerat ive p rocess is vital for spine surgeons

considering per form ing deform ity su rgery. As

the population ages and life expectancy in-

creases, the prevalence of degene rative a dult

spinal deformity w ill continu e to increase.2 In

fact, the impact on overall public health and

disability of the United States population by

adult degenerative scoliosis cannot be over-

stated, and there will likely be an increased

num ber of these pat ients electing surgical cor-

rection of their deformity and treatment of

their symptom s.2,4

! Epidemiology

New -on se t adult degenera t ive deform it ies are

considered in the context of a popu lation older

than 40 years of age, without a prior h istory of

adolescen t idiop ath ic scoliosis (AIS). Adu lt sco-liosis can b e asymp tom atic, and the inciden ce

of spinal curves of less tha n 10 d egrees m ay be

1

Preoperative Evaluation andOptimization for Surgery

Scott C. Wagner, Danie l G. Kang, Ronald A. Lehman, Jr.,

and Law rence G. Lenke



2 Chapte r 1

as h igh as 64%.5 In fact, 30%of elderly patien ts

w ithout a previous history of spinal deform ity

w ill develop new str uctu ral abnorm alities, w ith

m en and wom en a ected equally (in contrast to

adolescent idiopathic scoliosis, in which girls

are more com monly a ected than boys).3 Pa-tient s w ith progressive degenerative spinal de-

form ities typically present in the sixth de cade

w ith various symptom s, frequently including a

combination of back pain, radiculopathy, and

neurogenic claudication.3 Adult degenerative

deformities tend to progress up to 6 degrees

per year, ave raging 3 degre es per year, if left

untreated,3 and radiographic parameters that

p re dict a high risk for progression inclu de a

Cobb an gle greater th an 30 degrees, lateral olis-

thesis greater than 6 m m , and a large degree of

apical rotation.3 However, open su rgical spina l

deformity correction in adu lt patients is asso-

ciated w ith a com plication rate of up to 8 6%,

including a 7.8%rate of early wound infection,

and is typically associated w ith large am ount s

of int raoperat ive blood loss, deep w ound infec-

tion, and p ulmonary em bolism .4,6,7

Therefore, thorough preoperative evalua-

tion and opt imization is absolutely param ount

when considering surgical treatment of adultspinal deform ity, because this patient popu la-

tion is often elderly, with m ultiple associated

comorbidities, and at h igh r isk for m edical and

surgical complications.8 A multidisciplinary

app roach, including the p rima ry care p rovider,

an internist, an endocrinologist, a cardiologist,

as well as the t reating spine surgeon, should be

unde rtaken in the perioperative evaluation pro-

cess to minimize the potential medical risks

and maximize the functional bene ts.

! Clinical Evaluation

Initial Assessment

The initial assessment must include taking

a comprehensive history and performing a

thorough physical examination. A previous

diagnosis of spinal deform ity (e.g., adolescent

idiopathic scoliosis, kyphosis, congenital de-formity), a history of prior spine surgeries, as

well as any previous imaging studies demon-

strating progression of degenerative changes

and deformity w ill provide clinical cues to ap -

pro priat ely gu ide the rem ainder of th e w orku p.

Patients typically present with a combination

of various comp laint s, including uppe r or lower

back pain , ra diat ing low er ext rem it y pain orweakness, paresthesias/numbness, neurogenic

claudication, di culty with gait or upright

pos ture, and progress ion of their defor m it y.

Changes in body hab itus/postu re (par ticularly

chan ges in th e t of cloth ing), di culty w ith

gait or decreased walking distance tolerance,

and changes in the use of assistive devices are

elicited du ring the h istory-t aking process. Back

pain is t he m ost com m on pre sent ing symptom ,

and complaints of pain must be di eren tiated

with regard to axial versus radicular symp-

toms. Isolated low back pain may represent

par asp in al m uscle fat igu e or m echanica l in st a-

bilit y at the painful segm ent ,1 with increased

pain severit y oft en su ggest ing signi cant sag-

ittal and coronal imbalance.3 If radicular pain

is present in addition to axial pain, duration/

onset of symp tom s, exacerbating activities, and

laterality of the symptoms provide guidance

for p otent ial decomp ression.1,3 Radicular ex-

tremity pain can be caused by an acute diskherniation, localized foraminal or lateral re-

cess nerve root compression from osteophytes/

spondylotic chan ges, foram inal compression on

the concave side of the fractional curve, or t rac-

tion on the convex side of the deformity, or

m ay be related instead to single- or m ultilevel

central stenosis. Neurologic de cits are less

com m on in adult deform ities, but w hen pres-

ent are often related to segmental instability

causing foraminal compression or congenital

spinal stenosis, w hich is exacerbated by degen-

erative changes causing further central canal

stenosis.1 The operative approach should take

into consideration the extent and t ype of de-

compression and fusion constr uct, if any, that

is indicated based on the patient’s symptom-

atology, as well as any recent changes or pro -

gression of symptom s.1

The clinical exam ination includes assessment

of a shift in the trunk, and the relationship of

the h ead to th e pelvis in th e coronal and sagit-tal plane is noted. Asymm etr y of the shoulder

or pelvic girdles provides useful information



Preoperative Evaluat ion and Optimization for Surgery 3

w ith regard to th e severity of the deform ity, as

do pelvic obliquity and leg-lengt h discrep an cy.

Othe r subt le clues to severity an d pr ogression

of the deformity include skin creases around

the trun k/abdomen and standing posture (e.g.,

pelvic re t roversio n, h ip /k nee ex ion). Havingthe p atient perform forward an d lateral bend-

ing during the exam can provide important

prognost ic in for m at ion, as the r igid it y of the

curve can a ect the overall outcom e of nonop -

erative and subsequ ent op erative intervent ion.

Hip and knee exion contractu res should also

be assessed w ith t he pat ient lying in t he supine

posit ion on the exam inat ion table. Then, w ith

the patient lying in the prone position on the

table, the exibility of the curve without grav-

ity can be dete rm ined, and th e pat ient’s ability

to tolerate the prone position and overall phys-

ical condit ioning can be a ssessed. The p atien t’s

inability to t urn prone indepen dently may in-

dicate signi cant deconditioning and th at the

pat ient is a high -r isk su rgical candidat e). Neu-

rovascular examination includes overall gait

assessmen t, motor strength, deep ten don re-

exes, sensation and cranial nerve function,

as well as extremity pulse assessment.3 The

pat ien t sh ould also b e exa m in ed for long t ractsigns, as myelopathy m ay be a componen t of

severe thora cic deform ity, as well as to ensur e

that the patient does not have concomitant

cervical stenosis.

Radiographic Evaluation

Radiographic evaluation includes full-length

standing anteroposterior and lateral radio-

graphs of the spine, with the patient’s knees

and hips straight, as well as supine full-length

lms to provide information regarding any

spontaneous deformity reduction with gravity

forces removed. Cobb angle measu rem ent s and

radiographic determ ination of spinopelvic im-

balance provide crit ica l in form at ion, as t he d e-

gree of the curve and t he extent of imbalance

can necessitate discussion of operative inter-

vention at the time of initial evaluation. For

the purpose of preoperative planning, these

m easurem ent s are imp erative. Rotatory sublux-ation, the presence and location of osteophyto-

sis, and a ny ant eroposterior or lateral listhesis

are noted. Magnetic resonance imaging (MRI)

is routinely obtained , particularly in th e p res-

ence of radicular pain or neu rologic symptom s,

though it is not uncomm on for th ese older pa-

tients t o be u nable to und ergo an MRI for var-

ious reasons (e.g., presence of a pacemaker).Also, in the revision setting, previous spinal

instrum ent ation m ay cause signi cant im age

art ifact and di culty in MRI interpr etation. In

such patien ts, comp uted tom ography (CT) my-

elogram is obtained instead of MRI, and pro-

vides information regarding signi cant areas

of stenosis. We also rout inely obtain a CT scan

in adult spinal deform ity patients for preoper-

ative p lanning, wh ich e nables evaluation of the

extent of spondylotic changes and the levels/

areas of autofusion, helps deter m ine th e feasi-

bilit y an d sizing of sp inal xat ion poin ts, an d, in

the revision set ting, helps ana lyze th e location/

size of any previous decom pressions, the heal-

ing of previously fused regions, and the position

of previous spinal instru m ent ation. In ad dition,

at our institution, the CT scan is useful in pa-

tient s w ith complex deformity (e.g., congenital/

segmen tation abnorm alities, signi cant angu-

lar deformity, previous postsurgical changes)

through t he u se of a thre e-dim ensional acrylicm odel for p reoperat ive p lanning, and can also

be use d in t raop erat ively to iden t ify topo-

graphic landmarks and guide placement of

instrumentation.

Provo cative Testing

Selective nerve root/transforaminal cortico-

steroid injections can also be used to provide

diagnostic inform ation as w ell as a th erape ut ic

e ect.1 We use selective ner ve root/transforam-

inal injections in patients with a component

of radicular/lower extremity pain to help de-

term ine the speci c ner ve root causing symp -

tom s, provide tem porar y relief prior to su rgical

treatm ent , and ultim ately to localize the levels

in which decompression may result in symp-

tom relief. However, th e u tility of select ive ne rve

root/transforaminal injections remains unclear,

as the lack of response to t he injection m ay be

attributable to the injection technique or to poor pat ient re ca ll. We sp eci cally ask the pa-

tient about the immediate relief of symptoms,



4 Chapte r 1

w ithin 5 to 10 m inutes follow ing the injection,

as a criterion for a diagnostic injection (symp-

tom s likely arising from th at level of injection).

In contrast, an injection causing relief hours

or days later m ay be a function of the system ic

anti-in ammatory e ect following systemicabsorption of the corticosteroid. Similarly, in

our experience, epidural corticosteroid injec-

tions provide limited diagnostic information,

as the corticosteroid medication distributes

throu ghout m ultiple levels and is also absorbed

system ically. However, we o er ep idura l cort i-

costeroid injections for patients w ith signi cant

central or lateral recess stenosis to poten tially

provide tem porary r elief of sym ptom s an d im -

prove physica l fu nct ion to e nable pre op era t ive

optim ization of tn ess and m obility. We do not

routinely use facet blocks or diskography for

diagnostic assessment in the adult spinal de-

form ity patient.9 However, in patien ts w ith iso-

lated axial back pain an d ar thr itic facet changes

on imaging stud ies, facet blocks m ay be ut ilized.

Because the pain generator can be located at

any point in t he spine relative to t he a pex of

the curve, facet blocks are performed sequen-

tially at di erent levels to isolate speci cally

which motion segments are causing the pain,with subsequent relief of symptoms after in-

ject ion/a blat ion.1

! Nonoperative Manageme nt

A trial of nono perat ive m anagem ent is indicated

for almost all patients presenting with adult

spinal deformities, particularly curves of less

than 30 degrees, less than 2 mm of listhesis,

and if the constellation of symptom s is relatively

m inor. In contradistinction to th e t reatm ent al-

gorithm of adolescent idiopathic scoliosis, th ere

is no role for bra cing in adu lt spinal deform ity

pat ients3 because th e progression of the cur ve

is related to degenerative changes and me-

chanical instab ility, and not longitu dinal growt h

of th e axial skeleton . The refore, th e ben e t of

tem porar y pain relief is out weighed by the p o-

tential deconditioning of the paraspinal mus-cles an d by skin complications resulting from

brace trea tm ent in th is pat ient pop ulat ion.3,10

However, in rare cases in w hich the pain source

cannot be adequately localized, thoracolum-

bar or t horacolu mbosacral o rthoses (TLO/TLSO)

m ay be considere d for tem porar y stabilization

and pain relief.1 Low-impact core strengthen-

ing programs an d ph ysical therapy are utilizedto improve patient reserves as well as to stabi-

lize the surrounding musculature to provide

improved support to the spinal column.3 Non-

stero idal ant i-in am m ator y drugs (NSAIDs) are

used t o p rovide relief of axial and, occasionally,

radicular pain an d n eurogenic claudication. We

do not routinely provide n arcotic pain m edica-

tions for nonoperative treatment, and pain

management specialists are consulted to pro-

vide multimodal therapy with optimization of

non narcotic pain m edications (e.g., gabapent in,

pre gabalin ), althou gh som et im es sh or t p eriods

of narcotics or pain m edications may be n eces-

sary. Also if ope rat ive tre atm en t is decided, we

encourage reduction or complete discontinua-

tion of any narcotic pain m edications to avoid

di cult pain m anagement in the postoperative

period.

! Surgical Indications

Indications for surgery in th ese pat ients include

failure of nonoperative pain m anagement w ith

signi cant ly dim inished qua lity of life/fun ction ,

or progression of deformity/imbalance, with

correlation between radiographic and clinical

ndings. As previously mentioned, lumbar

curves greater than 30 degrees or with 6 m m

of listhesis in a ny plane a re considered for sur-

gery because the deformity is at high risk for

progressio n. Also , pat ients w ith annual defor-

mity progression greater than 10 degrees or

with increasing listhesis (lateral, anterior, or

poste rior ) greate r than 3 m m, and w hose sym p-

tom s are progressively worsening, are o ered

surgical stabilization. Ultimately, the decision

to proceed w ith surgical management is predi-

cated on several major factors, including the

pat ient’s symptom at ology, age, genera l m edi-

cal health, and t he p atient’s expectations w ithregard to the out come of such a signi cant pro-

cedure.1 If surgical options are to be pursu ed,




medical optimization of the patient and de-

tailed preope rative surgical planning are a bso-

lutely critical to promote the success of the

treatment plan.

! Optimization for Surgery

As previously m ent ioned, the present ing age of

pat ients w ith adult sp inal deform it ies is typ i-

cally between 60 and 70 years, and systemic

m edical com orbidities are com mon .1,3 Diabetes

and cardiac and vascular disease can signi -

cantly impact the surgical outcome, particu-

larly for a large reconst ru ctive procedu re, given

the potential for considerable intraoperative

blood loss and overall surgica l t im e.1,3 Postop-

eratively, elderly patients also require longer

rehabilitation, given t heir decreased cardiopul-

m onary reserves.1 Therefore, consultation w ith

the anesthesiologist and the patient’s primary

care provider is recommended to pursue an

interdisciplinary approach for stratifying the

pat ient’s perioperat ive m edica l r isks and op t i-

m izing medical com orbidities prior to proceed -

ing with surger y.Halpin et al11 and Sugrue et al 12 described

the ir high-r isk protocol for pat ients u nde rgoing

major spinal surgery: patients are considered

high risk if the surgeon an ticipates longer than

6 hour s of operative time, more th an six vert e-

bral levels w ill be included, o r that the pro ce-

dure will be staged, or if the patient presents

w ith signi cant m edical comorbidities. In these

authors’ protocols, all high-risk patients are

evaluated by a hospitalist an d anesth esiologist,

and various param eters are evaluated an d opti-

mized, including nutritional status, pulmonary

statu s, cardiac and renal function, and he patic

function.11,12 The case is then discussed at a

conferen ce for h igh-risk spine procedu res th at is

atten ded by all m anaging providers before op-

erative clearan ce is granted.11 At ou r institu tion,

the use of similar goal-directed, evidence- based

prot ocols to coor din at e the ca re of com plex

pat ien ts has im prove d outcom es and ove rall

pat ient sat isfact ion postop era t ively.12

Nut rit ional s tat us of the ad ult sp inal d efor-

m ity patient sh ould be assessed preope ratively.

This evaluation is typically accomplished by

measuring serum albumin, prealbumin, total

prot ein , a nd t ransferr in , w hich provide in for -

mation regarding patient protein reserves.13

Patients with albumin levels less than 3.5 g

per d eciliter h ave been s how n t o h ave a s ign i -cantly higher risk of complications and m or-

tality.14 Prealbumin levels below 11 mg per

deciliter require nutritional support, and be-

cause these levels are not a ected by hydration

status, prealbumin is the recommended mea-

surem ent tool for assessing nut ritional stat us.14

Any insu ciency in the nutr itional state iden -

ti ed preope ratively should be corrected prior

to surgery, consulting with a nutritionist if

necessary. The d uration of nut ritional support

is dependent on the severity of the malnour-

ished state and the patient’s general health,

bu t generally is 6 to 12 w eeks in order t o at tain

approp riate nutr itional optim ization, although

some patients may require a longer period.

Postoperative nutrition is an important aspect

for all patients following spinal deformity

surgery, particularly with complex spinal re-

construct ive procedures that enta il signi cant

m etabolic deman d. There is often a balance in

timing for the start of nutrition by mouth andreturn of bowel function (i.e., bowel sound,

atus, and bowel m ovem ent). Star ting an oral

diet too early may result in ileus or obstruc-

tion, wh ich can signi cantly increase th e pa-

tient ’s pain and limit early rehabilitat ion e orts,

whereas unnecessarily delaying the start of

nutrition may fail to meet metabolic require-

m ent s to optimize healing and rehabilitation in

the postoperat ive p eriod. Therefore, in certa in

cases, part icularly follow ing complex spin al re-

constructive procedures, we at tempt placem ent

of a small bowel feeding tube (SBFT) on post-

operative day 1, with the goal to begin tube

feed s by postop erat ive day 2. If th ere is di -

culty in placing th e SBFT distal to t he pylorus,

we begin parenteral nutrition support through

central access. We continue small bowel tube

feeds or parenteral nutrition support until

the patient is tolerating adequate nutrition by

mouth.

Perioperative blood management is an aspect of adult sp in al deform it y su rgery that

requires particular attention. Low preoperative



6 Chapte r 1

hem oglobin concentration and increased num -

ber of levels fuse d have been sh ow n to be

signi cant risk factors for allogeneic blood

transfusion at the time of surgery.15 The risks

associated with transfusion are myriad, and

include benign febrile reaction, infectious dis-ease transmission, and anaphylaxis. Therefore,

e orts to redu ce the potent ial need for tran sfu-

sion should be undertaken preoperatively. In

the absence of any contraindications, we rec-

omm end t hat patients w ith adult spinal defor-

mity take iron supplements for 2 to 4 weeks

prior to s urgery.14 The re is evidence to suggest

that preoperative recombinant human eryth-

ropoietin (rhEPO) adm inistrat ion in the p reop-

erative period can reduce the transfusion rate

without increasing complications.16 However,

at our institut ion this is not a com m on pract ice

given t he signi cant exp en se of rhEPO, an d, in

our experience, its lim ited e ectiveness in the

adu lt spinal deform ity patient . We t ypically use

other perioperative adjunctive measures and

blood m anage m ent st ra tegies, w hich inclu des

the use of intravenous anti brinolytics (e.g.,

tranexamic acid), cell saver, and topical hemo-

stat ic agent s (e.g., Surgi o, th rom bin), as well

as paying meticulous attention to hemostasisthroughout the procedure (including packing

o segmen ts w ith rolled surgical sponges to

redu ce blood loss wh en at tent ion is focused on

decompressing or instrum enting more cepha-

lad or caud ad spina l levels).

Open anter ior surgical deform ity correction

for severe spinal deform ity has been shown to

have detr imen tal e ects on postoperative pul-

monary function, particularly in older adult

pat ients or pat ients w ith preexist ing lung dis-

ease.17,18 Although pu lmon ary function testing

is not routinely performed preoperatively, we

typically evaluate patients with pulmonary

symptom s, di culty with or poor endurance

w ith daily activity an d am bulation, or comp lex

or severe thoracic deform ity (often with planned

three-colum n osteotomy). We use pulmonar y

function testing in th ese patient s to stratify the

risk of poten tial postoperative pulm onar y com -

plica t ions, and we o btain a p ulm on ary sp ecial-

ist consultation for per ioperative optim ization.Also, preope rat ive sm oking cessation is imp er-

ative for at least 8 w eeks prior to surgery.

Typically, major deformity correction and

fusion has been accomplished via combined

anterior/posterior approaches; th e an terior re-

lease with fusion is achieved via a thora cotomy

or thoracoabdominal approach, followed by

posterior inst rum entat ion, which provid es im - proved fusio n rat es and be t ter ove rall corre c-

tion.19 However, it is postulated t hat d isruption

of the thoracic cage during the anterior ap-

pro ach lead s to injury to t he r espirator y m ech -

anism.18 Because of this theor y, there h as been

interest in posterior-only management of se-

vere deform ities (e.g., via t hree -colum n osteot-

omies such as pedicle substraction osteotomy

[PSO] or vert ebra l colum n resect ion [VCR]) an d

in the theoret ical bene ts of obviating the an -

terior approach on pulmonary function. There

is some evidence that posterior-only surgery

can achieve sim ilar postoperat ive ra diographic

outcomes19; however, patien ts w ith such severe

deform ities often presen t w ith chronic restr ic-

tive lung disease, with m inim al potent ial for

improvemen t in lung function despite correc-

tion of the thoracic deformity, and a recent

study found that, in adult patients, utilization

of VCR for severe deformity correction did

not improve postoperative pulmonary func-tion.18 Preoperat ive pulm onar y funct ion test-

ing, therefore, may be worthwhile in patients

w ith signi cant thoracic deform ities and base-

line pulmonary disease to establish potential

reserves. Thus, it is imp orta nt to counsel older

pat ients w ith m ore severe deform it ies t hat de-

spite the correction a orded by the surgery,

wh ich m ay require extensive osteotom ies, pul-

m onar y function may not imp rove signi cantly

pos top erat ively.18

Hypovitaminosis D, although extremely

common, is often missed in the preoperative

setting, despite the potentially serious com-

p licat ions ar isin g from th is de cie ncy. It is

estimated that more than half of all general

m edicine inpatients are de cient in vitam in D,

though the prevalence in patients undergoing

spine surgery remains largely unexplored.20 A

recent stu dy from a single institu tion foun d an

overall vitamin D de ciency rate of 57%in pa-

tients undergoing spinal surgery of any kind,and t he rate for patient s with diagnosed spinal

deform ity w as 18 %; th is relatively low preva-




lence is likely att ributable to an increased rate

of vitam in D supp lemen tation in this cohor t.20

It is thus important to consider this diagnosis

and recom men d ade quate vitam in D intake for

pat ients w ith diagnosed sp inal deform it y, es-

pecially p re op erat ively, as calciu m m et abo lismis extremely important in the prevention of

osteoporosis.

Along with hypovitaminosis D, osteoporo-

sis is also very com m on in this patient p opula-

tion. The management of this serious disease

requires the cooperation of a mu ltidisciplinary

team. Postmenopausal women are at a high

risk for d evelopmen t an d p rogression of oste-

oporosis, which can lead to fragility fractures

and increased mortality; however, older men

may also present with osteoporosis, and any

clinical suspicion should prompt an initial

workup . The World Health Organization (W HO)

recomm ends that all peri- and postm enopausal

wom en u ndergo screening for low bone m in-

eral density (BMD),21 and dual-energy X-ray

absor pt iome tr y (DEXA) is th e gold stan dard for

assessment of BMD. We obtain DEXA BMD

measurements of the lumbar spine and hips

for all preop erat ive pat ient s, regardless of age

or gender, to identify osteoporotic patientswh o may require optimization/treatm ent w ith

consultation of an endocrinologist or primary

care provider prior to surger y. Postm enop ausal

women diagnosed with osteoporosis should

also receive 1,500 m g calcium and 400 IU vita-

m in D daily. The re also exist medical modalities

for optimization of BMD, including bisphos-

phon at es , p arat hyro id hor m on e (t eripara t ide),

estrogen modu lators or horm one replacement ,

and calcitonin. The use of these medications

should be m onitored in consultation with the

pat ient’s endocrinologist or prim ary care pro -

vider. Identifying patients with osteoporosis

pr ior to su rgery facilitat es t reat m ent and op t i-

m ization of their BMD, and can improve surgi-

cal outcom es by optim izing th e xation strengt h

of the sur gical instrum ent ation and ultim ately

improve bone healing/fusion.

Cardiopulmonary, nutritional, and bone-

quality assessments are vital in this patient

pop ulat ion. Com or bidit ies a re in tuit ively m orecommon in the adult deformity population

wh en compared with t he adolescent idiopathic

scoliosis population, and the presence an d se-

verity of these comorbidities guide the initial

management of the deformity. Although there

has been some evidence that osteopenia and

osteoporosis do not play a signi cant role in

the progression of adult spinal deformity,3 in pat ients ele ct ing to proceed w ith su rgical cor -

rection of scoliosis th e p resence of osteoporo-

sis can a ect the ability to obtain purchase in

the bony spine. In p atients over 50 years of age

und ergoing spine surger y of any type, the inci-

dence of osteoporosis has been reported to be

14.5%for me n and 51 .3%for wom en.22 Indeed,

osteoporosis is associated w ith rep orted fusion

rates a s low a s 56%, as well as iatrogen ic inst a-

bilit y and fract u re follow ing su rgery.23 Sur-

veys have found that most orthopedic spine

surgeons feel uncomfortable managing the

treatm ent of osteoporosis after it has been di-

agnosed24 ; therefore, prompt referral to pri-

mary care providers or endocrine specialists

for partial or complete management of osteo-

poros is prior to an y planned su rgical proce-

dure is recomm ended.

Lastly, psychosocial factors m ust be con sid-

ered. Mental health issues are common in the

older adult popu lation, and the p resence of de- pression , an xiety, psych osis, or o ther prem orbid

psych ological condit ions ca n adverse ly a ect

surgical outcomes and patient perception of

surgical success.11 These factors can be m an-

aged by e ectively utilizing a team of social

workers or case man agers and psychiatric sup -

por t , an d shou ld n ot be ove rlooked p rior to u n-

der taking major spinal deform ity surgery.

! Preventing Complications

Medical comp lications surrou nding ad ult spi-

nal deformity surgery can range from mild to

extrem ely severe, with an overall comp lication

rate ranging from 40 to 86%in patien ts und er-

going d eform ity surger y.25 Thorough atte nt ion

to the preoperative medical optimization pro-

cess can r educe t he inciden ce of postoperative

complications, and str ategies to m inimize suchcomp lications sho uld be judiciously em ployed.

The m ost comm on m inor complication in th e



8 Chapte r 1

pos top erat ive period is urinary t ract in fect ion

(UTI), with a re por ted rate of 9%.25 UTIs can b e

prevented pre - and in t ra op erat ively by appro -

priat e st erile technique during inse rt ion of the

catheter, unrestricted catheter drainage, early

removal, and, in some instances, instillation of be nign ba ct eria in the urinary t ra ct .25 Pulmo-

nary abnormalities, including atelectasis and

pneum on ia, are a lso very com m on in th is p op -

ulation. These complications can be p revented

in the preop erative setting via sm oking cessa-

tion at least 8 we eks prior to surgery, as noted

above, as well as approp riate use of bronchodi-

lators or pulm onar y rehabilitation p rotocols.25

Of course, many oth er intra- and p ostoperative

strategies exist to m inim ize th e nu m erous com-

plica t ions that m ay occu r, bu t t hese a re beyon d

the scope of this chapt er.

! Preoperative Planning

Levels of Treatment

Six levels of operative t reatm ent were described

by Silva an d Lenke3 in 2 010: I, decomp ressionalone; II, decompression and limited instru-

m ente d p osterior spinal fusion; III, decompres-

sion and lumbar curve instrumented fusion;

IV, decompression w ith an terior an d p osterior

spinal instr um en ted fusion; V, th oracic in-

strumentation and fusion extension; and VI,

inclusion of osteot om ies for speci c deform i-

ties. Each level represent s a u nique ap proach t o

surgical m anagem ent of adu lt spinal deform ity,

p re dica te d on the constellat ion of symptom s

reported by the patient, and d esigned to pro-

vide independent symptom management. For

pat ients w ith neurogenic claudicat ion alon e

secondar y to cent ral canal stenosis, level I treat-

m ent , which ent ails lim ited decomp ression, is

app ropriate. These patient s often presen t with

minimal back pain, and radiographic analysis

may reveal small osteophytes with less than

2 mm of subluxation. Additionally, these pa-

tient s should have no cosm etic or major defor-

mity complaints, and the coronal and sagittal balance m ust be w ithin re ason , as isolat ed cen-

tral decompression in the presence of curves

greater tha n 30 degrees (or with kyph osis) can

lead to worsening of the deformity.3 A rela-

tively large series found that coronal imbalance

greater than 4 cm correlated with decreased

overall patient-related ou tcome scores on the

Scoliosis Research Society-22 (SRS-22) scale

an d t he Oswestr y Disability Ind ex (ODI),26 andthus these param eters are extrem ely impor tant

in th e su rgical decision-making process. How-

ever, for the relatively well-balanced patient

with m ore than 2 m m of subluxation, the addi-

tion of posterior instr um ent ation at th e level of

the decomp ression imp roves stability and con-

stitutes level II of treatment. If such patients

also have comp laint s of signi cant lumb ar pain

associated with the lumbar deformity greater

than 30 degrees, but maintain global sagittal

and coronal alignme nt, the e ntire lumbar curve

m ust be included in the instrum ented region,

wh ich const itutes level III of treat m ent .3 Trans-

foraminal lumbar interbody fusion (TLIF) may

also be utilized as an adjunct when fusing to

the sacrum to imp rove xation and fusion at the

tran sitional lumbosacral jun ction.3

Loss of lumbar lordosis, often associated

with at-back syndrome in adult deform ity pa-

tient s, is often m anaged via an ant erior fusion

approach. Utilizing anterior fusion in additionto poste rior xation const itut es level IV an d

provid es bo th load sh ar ing t o re duce posterior

strain and a dditional ceph alocaudad foram inal

decompression.3 In addition to the aforemen-

tioned criteria, patients with additional sagit-

tal imbalance can be managed by expanding

the fusion proximal to t he thoracolum bar junc-

tion, which constitutes level V of treatm ent .3

It is also important that anterior osteophytes

be m inim al, an d sign i cant thoracic kyphosis

contraindicates this t reatmen t approach.3 Once

signi cant sagittal or coronal imbalance has

developed, spinal fusion without adjustment

of global alignm en t will be insu cient to

control symptom s. A recent retrospective stud y

exam ining the role of preoperative coronal and

sagittal balance found that postoperative cor-

rection of sagittal balance was the strongest

predictor of clin ical ou tcom es, whereas another

study has suggested that severe preoperative

coronal imbalance predicts worse functionalrecovery.7,26 Historically, patients with severe,

rigid spinal deform ities have been managed w ith

combined anter ior/posterior app roaches; how-




ever, there has been increased interest in the

use of complex three-column osteotomies to

enab le an all poster ior approach, however inter-

est in the u se of complex three-colum n osteoto-

mies to enable an all posterior approach; the

use of these osteotomies constitutes Level VIof surgical m anagem ent . These comp lex three-

colum n osteotom ies require highly experienced

surgeons and a specialized operating room team

to ensure optimal outcomes and the highest

level of safety, and even with this expertise

the re is still a 30 to 40% rate of complications

following th ese p rocedures.17

! Chapter Summary

Patients with adult spinal deformity represent

some of the m ost complex surgical candidates

in the population, and estimates suggest that

the number of patients electing to undergo

surgical correction will continue to increase.

Adu lt scoliosis comp rises a diverse sp ect ru m o f

disease, w ith m ultiple potent ial etiologies and

natural histories, and as such there is no one

single approach to management that can beapplied to a ll adult deform ity patient s. Radio-

graph ic, clinical, an d subject ive nd ings m ust

be assessed preop erat ively by a m ult id isci-

p lin ary team . Because these pat ients typ ica lly

prese nt aft er the sixt h decade of life , w it h

m ultiple associated m edical com orbidities, the

spine surgeon must be aware of the potential

for signi cant risk exposure in the pe riopera-

tive setting. A multidisciplinary approach to

pre op era tive evaluat ion m ust be employe d,

and t he p atient’s prim ary care provider, inter-

nist, endocrinologist, and cardiologist should

be act ively engaged in det erm in ing if the pa-

tient is appropriate for surgery and in prepar-

ing the pat ient for th e procedu re. If the patient

is not cur rent ly being evaluated for major med-

ical conditions common to this population,

such as restrictive lun g disease or osteoporosis,

the spine surgeon may be the rst provider to

initiate assessmen t and recomm end treatm ent.

The complexity of the t hree- dime nsional path-oanatomy and associated biomechanics that

can signi cantly a ect postoperat ive outcom es

must be understood and respected, and the

preop erat ive eva luat ion and op t im iza t ion for

surgery process must be tailored to each in-

dividual patient. With ap propriate patient se -

lection, understanding of all treatment options

and decision algorithms, as well as under-

standing the importance of a team approachto perioperative medical management, spine

surgeons can expect good results for their pa-

tients undergoing surgical treatment for adult

spinal deformit y.

Pearls

Back pain is the m ost com mon complaint in adu lt

spinal deformity patients.

If claudication symptoms are present in addition

to axial pain complaints, laterality of the pain

provides gu idance for poten tial de com pre ssion

and likely instrum ent ed fusion.

The rigidity of the curve can be assessed both

clinically and radiographically, and a ects overall

outcom es of nonop erative and subsequen t oper-

ative intervention.

Full-length standing anteroposterior and lateral

radiographs o f the spine are essent ial, and supine

full-length lms provide informat ion regard ing

any spontaneous deformity reduction related to

gravity. Indications for surgery in these patients include

failure o f nonope rative p ain manag ement , as well

as correlation between radiographic and clinical

ndings.

Consultation with the anesthesiologist and the

patient’s p rima ry ca re provider is recom men de d

to ensure an multidisciplinary approach for strat-

ifying the patient’s perioperative medical risks

and optimizing medical comorbidities prior to

pro ceed ing with surge ry

Preoperative pulmonary function should be

evaluated, as increased impairment or minimal

improvement in pulmonary function can be expe cted postop era tively, and patients should be

informed abou t this matt er.

Hypovitaminosis D and osteoporosis are ex-

tremely common in this patient population,

and, given the signi cant de trimenta l e ect on

fusion rates and potentially overall clinical out-

comes, should be managed in consultation with

an endocrinologist.

Operative candidates can be classi e d based on

severity and type of their symptoms, as well as

pre op era tive radiograph ic nd ings.

Consideration should be given to posterior-only

deformity correction t echniques, which m ay re-

duce m orbidit y associated with the ante rior tho -

racotomy or thoracoabdom inal approaches.



10 Chapte r 1

Pitfalls

Bracing is not rout inely utilized in adu lt spinal de-

formity patients and may result in decondition-

ing a nd skin comp lications.

Failure to identify and evaluate osteoporosisand subsequently failing to optimize BMD may

result in subopt imal xation and const ruct /fusion

failure.

Patients may ex the ir knees and hips, with sub-

sequent pelvic retroversion, to compensate for

xed sagitt al imba lance, and t he surgeo n should

ensure that radiographs are obt ained withou t

these compensatory mechanisms.

Narcotic pain me dicat ions should not be routine ly

prescribed preoperatively, and patient s with sig-

ni cant narcotic pain medication use preopera-

tively should be weaned to optimize postopera-

tive pain management.

Failure to identify preoperative nutritional de -

ciency may result in poor wound hea ling, di -

culty with rehabilitation, and prolonged fusion

healing. Complex three-column osteotomies require highly

experienced spine surge ons and a spe cialized op-

erating room team to ensure optimal outcomes

and the highest level of safety. Therefore, the

spine surgeo n should always consider every othe r

option or technique to obtain realignment and

opt imize balance (e.g., positioning, poste rior soft

tissue/ligament releases, facetectomies, poste-

rior column osteot omy) rather t han use a three-

column osteotomy.

References

Five Must- Read Reference s

1. Aebi M. The adu lt scoliosis. Eur Spine J 2005;14 :925 –

94 8 PubMed

2 . Mes n A, Len ke LG, Brid we ll KH, et a l. Does pre op -

erative narcotic use adversely a ect outcom es and

complications after spinal deformity surgery? A com-

par iso n of non nar cot ic- w ith nar cot ic- usin g grou ps.

Spine J 2014;14: 2819–28 25 PubMed

3. Silva FE, Len ke LG. Adu lt de gene rat ive scoliosis: eva-

luation and management. Neurosurg Focus 2010;

28:E1 PubMed

4. Mum m an en i PV, Sha rey CI, Len ke LG, et a l; Mini-

mally Invasive Surgery Section of the International

Spine Study Group. The minimally invasive spinal

deform ity surgery algorithm : a reproducible rational

framework for decision making in minimally in-

vasive spinal deformity surgery. Neurosurg Focus

2014;36:E6 PubMed

5. Schwab F, Dubey A, Gamez L, et al. Adult scoliosis:

p revalence, SF-36, an d nut r it ional par am et er s in an

elderly volunteer population. Spine 2005;30:1082–

1085 PubMed

6. Schwab FJ, Hawkinson N, Lafage V, et al; Interna-

tional Spine Study Group. Risk factors for m ajor per i-

operative complications in adult spinal deformity

surgery: a multi-center review of 953 consecutive

pat ien ts. Eur Spin e J 20 12 ;2 1: 26 03 –2 61 0 PubMed

7. Daubs MD, Lenke LG, Bridwell KH, et al. Does cor-

rection of preoperative coronal imbalance make a

di erence in outcom es of adult patients w ith defor-

m ity? Spine 2013;38:476–483 PubMed

8. Acost a FL Jr, McClen don J Jr, O’Sha ugh ne ssy BA, et a l.

Morbidity and mor tality after spinal deform ity sur-

gery in patients 75 years and older: complications

and predictive factors. J Neurosurg Spine 2011;15:

667–674 PubMed

9. Grubb SA, Lipscom b HJ, Suh PB. Resu lts o f sur gical

treatment of painful adult scoliosis. Spine 1994;19:

1619–1627 PubMed

10. van Dam BE. Nonope rative treat m ent of adult scolio-

sis. Orthop Clin North Am 1988;19: 347–351 PubMed

11. Halpin RJ, Sugr ue PA, Gould RW, et a l. Stan dar dizing

care for high-r isk patient s in spine surgery: the Nor-

thwestern high-risk spine protocol. Spine 2010;35:

2232–2238 PubMed

12. Sugr ue PA, Halp in RJ, Koski TR. Trea tm en t algor ith m s

and protocol practice in high-risk spine surgery.

Neu rosurg Clin N Am 20 13 ;2 4: 21 9– 23 0 PubMed

13. Klein JD, Hey LA, Yu CS, et al. Perioperative nutri-

tion and postoperative complications in patients

undergoing spinal surgery. Spine 1996;21:2676–

2682 PubMe d

14. Kelly MP, Hu SS. Nutrition and pain management in

the adult sp inal deformity p atient . Scoliosis Research

Society e-text. http://etext.srs.org/. Accessed August

30, 2014

15. Nut ta ll GA, Horlocker TT, San tr ach PJ, Oliver WC Jr,

Dekutoski MB, Bryan t S. Predictor s of blood t ran sfu-

sions in spinal instrumentation and fusion surgery.

Spine 2000;25:596–601 PubMed

16. Shap iro GS, Boach ie-Adjei O, Dhaw likar SH, Maier LS.

The use of Epoetin alfa in complex spine deformity

surger y. Spine 2002;2 7:2067–2 071 PubMed

17. Aue rb ach JD, Len ke LG, Brid we ll KH, et a l. Major

complications and comparison between 3-column

osteotomy te chniques in 105 consecutive spinal defor-

m ity procedures. Spine 2012;37:1198–1210 PubMed

18. Bumpa ss DB, Len ke LG, Brid we ll KH, et al. Pulm onar y

function improvemen t after vertebral column resec-

tion for severe spinal deformity. Spine 201 4;39:587 –

59 5 PubMed




19. Good CR, Len ke LG, Brid we ll KH, et a l. Can poste rio r-

only surgery pr ovide similar radiograph ic and clinical

results as combined an terior (thoracotomy/thoraco-

abdom inal)/posterior app roaches for adult scoliosis?

Spine 2010;35:210–218 PubMed

20 . Sto ker GE, Buch ow ski JM, Bridw ell KH, Len ke LG,

Riew KD, Zebala LP. Preoper ative vit am in D st atu s of

adults u nde rgoing surgical spinal fusion. Spine 20 13;

38:507–515 PubMed

21. Lane JM, Nydick M. Osteoporosis: current modes of

pre vent ion an d t re at m en t . J Am Acad Orthop Surg

1999;7:19–31 PubMed

22. Chin DK, Park JY, Yoon YS, et al. Preva len ce o f ost eo -

porosis in pa tien ts requir ing sp ine surger y: inciden ce

and signi cance of osteoporosis in spine disease. Os-

teoporos Int 2007;18:1219–1224 PubMed

23. Park SB, Chu ng CK. Stra tegies of spin al fusion on os-

teoporotic spine. J Korean Neurosurg Soc 2011;49:

317–322 PubMed

24 . Dipao la CP, Bible JE, Biswas D, Dipaola M, Grauer JN,

Rechtine GR. Survey of spine surgeons on attitudes

regarding osteoporosis and osteomalacia screening

and t reatm ent for fractures, fusion surgery, and p seu-

doar throsis. Spine J 2009;9:5 37–544 PubMed

25. Baron EM, Alber t TJ. Medical comp lications of su rgi-

cal treatment of adult spinal deformity and how

to avoid th em . Spine 2 006;31(19, Suppl):S106–S118

PubMed

26. Glassm an SD, Ber ven S, Bridw ell K, Horton W, Dima r

JR. Correlation of radiograph ic param eters and clini-

cal symptom s in adult scoliosis. Spine 2005 ;30:682–

68 8 PubMed



! Introduction

Degenerative scoliosis most comm only a ects

the lumbar spine in the elderly. It occurs as a

result of facet and disk degeneration, leading to

increased loads and resulting deformity. The

preva lence o f adult scoliosis in the elderly p op -

ulation m ay be as h igh as 6 8%,1 and this num-

ber w ill only in crea se as p eop le live lon ger a nd

wan t to m aintain th eir activity levels. Unt reatedscoliosis can lead t o pain, spinal osteoar th ritis,

worsening deformity, spinal stenosis with ra-

diculopathy, coronal and sagittal imbalance,

associated m uscle fatigue, and psychological ef-

fects from poor cosmesis and redu ced m obility.

The m anagem ent of adult deformity is con-

troversial, with a lack of high-quality evidence

to guide treatment, such as determining the

best candidat es for conse rvat ive versu s su rgi-

cal treat m ent , and t he best surgical procedu res

for speci c clinical scen arios. Nevert heless, pa-

tients tend to present w ith back or leg pain, and

concerns about deformity progression need to

be ad dressed. Thus, an understandin g of the pos-

sible causes is important before appropriate

recom m endations for treatmen t can be made.

! Diagnosis o f Back Pain

One of the most common presenting yet

di cult-to-discern comp laints of degenerat ive

scoliosis is back pain. The p ain m ay be stat ic or

m echanical, localized or regional, or associated

with buttock or leg pain, and there may even

be neurologic symptom s. It is im por tant to

elicit a thorough h istory docum ent ing the pain’s

severity, its aggravating and relieving factors,

and its funct ional lim itations that a ect work

or recreation or r educe t he patient ’s ability to

walk distances. This information helps eluci-

date the cause of the pain, and thus helps indeterm ining the appropriate treatment .

Axial back pain can be cause d by de genera-

tion of the intervertebral disk (diskogenic) or

disk height loss leading to segm ent al instability

(degenerative spondylolisthesis). There could

also be single- or m ultisegment facet joint de-

generation. All these ndings are part of the

degen erat ive cascade a s described by Kirkaldy-

Willis et al.2 Thus, a thorough clinical exam-

ination w ould include careful palpation of the

lumb ar spine, its mu sculature, and the sacroil-

iac joints, to look for areas of local tenderness

that would help pinpoint path ology. Additional

characteristic ndings include the presen ce of

an “instability catch” or the patient’s experi-

ence of a catching pain in the lower back wh ile

rising from a forwa rd-leaning posture, wh ich

requires supporting their weight by putting

their hands on their knees. The patient may

also have a “painful catch,” in w hich th e ra ised,

straightened leg is unable to move down butsuddenly drops due to a shar p pain in th e lower

back. Bot h sympto m s could point to the pres -

2

Decision Making in Adult DeformitySurgery: Decompression VersusShort or Long Fusion

Kenneth M.C. Cheung and Jason P.Y. Cheung



Decision Making in Adult Deformity Surgery 13

ence of spinal instability, likely from a degen-

erative spondylolisthesis.

Back pain can a lso result from postur al im-

balance in bot h the co ro nal and sagit tal planes.

This imbalance is often referred to as the “cone

of economy” as discussed by Jean Dubousset.3 The cone is projected from the feet u p, and so

the trunk is only within a narrow range. This

concept relates to the par t of the cone w here

the body can rem ain balanced without extern al

support and using m inim al e ort. The mu scular

e ort required in an upright posture is much

greater wh en th e cone is exceeded, and correc-

tion should be considered .

Coronal imbalance of the spine can lead to

tr uncal translation and r ib-on-p elvis impinge-

m ent . Sagitt al plane deform ities include di -

culties in stan ding upright, resulting in mu scular

fatigue a nd discom fort from comp ensating for

the global sagittal kyphosis. These global de-

form ities may furthe r stress th e sacroiliac and

hip joints and lead to buttock and groin pain.

Usually the location of the pain is quite accu-

rate in determining the problematic site, but

the sacroiliac and h ip joints a re comm on sites

of misdiagnoses of back pain and should be

thoroughly assessed by clinical examination.Shoulder or pelvic asym m etr y and shoulder or

rib prom inence are clues for coronal deform i-

ties. Although a forw ard-leaning post ure could

be re lat ed to m uscle fat igue cause d by sagit tal

imbalance, it could also be due to a xed ky-

phot ic deform it y of the sp ine it self or a re su lt

of hip extensor weakness. In addition, during

the gait assessment, patients may have wors-

ening kyphotic posture due to m uscle decom -

pensat ion associated w ith pro longed walking.

Radiological assessment of causes of back

pain wou ld require fu ll- lengt h st anding pos-

teroanterior and lateral radiographs of the

spine, which must include, at a minimum, C7

to th e h ip joint s, but ideally wou ld include C1

to the hip joints, so that balance parameters

can be easily m easured . Flexion an d exten sion

views are useful, and in our expe rience, stand -

ing exion and prone traction radiographs show

the maximum displacement of a spondylolis-

thesis and its maximum reduction.4,5 Addi-tional m agnetic resonan ce imaging (MRI) of the

lumbar spine is needed to assess neurologic

impingeme nt as well as to rule out other causes

of back pain. Sometimes, because of the se-

verity of the d eform ity, a comp uted tom ogra-

phy (CT) m yelogram could be a use fu l adju nct

to identify the exact location of nerve root

compression.Radiological instab ility is com monly de ned

by the degree of slip (Fig. 2.1), and t he change

in slip angle (Fig. 2.2 ) and disk height (Fig. 2 .3).

These radiographic features can be found on

standing lateral radiographs (degree of slip)

and dynam ic exion-exten sion lateral radio-

graphs (slip angle and disk height). Oblique

lms can be t aken to look for a pars d efect. For

measurement of the degree of slip, a line is

Fig. 2 .1 Measureme nt of the degree of slip. A line

is dropped from the poste rior border of the cranial

vertebrae to the caudal vertebrae. The distance

from this point to the posterior border of the caudal

vertebrae is divided by the total vertebral body

width of the caudal vertebrae. Grade 1 is de ned as

0 to 25%, grade 2 is "25 to 50%, grade 3 is "50 to

75%, and grade 4 is "75 t o 100%.



14 Chapte r 2

drawn from the posterior border of the cranial

vertebrae to t he caud al vertebr ae. The d istan ce

from this point to the posterior border of the

caudal vertebrae is divided by the total verte-

bral bo dy w idth of t he caudal verteb rae. Grad e

1 is de ne d as 0 to 25%, grad e 2 is "25 to 50%,

grade 3 is "50 to 75%, and grade 4 is "75 to 100 %.

Spond yloptosis is de ned as m ore than 100%

slip. A slip of greater th an 50%is un stable an d

associated with progression and lumbosacral

kyph osis. The slip angle of an L5-S1 spond ylo-

listhesis is measured by a line drawn perpen-

dicular to the posterior aspect of the sacrum

and a line drawn along the inferior end of theen d plate of L5. In th e cran ial segm en ts, the slip

angle is made by the superior end plate of the

caudal vertebra and the inferior end plate of

cranial vertebra. For measuring disk height, a

line is drawn from the midline inferior end

plat e of the cranial verteb ra to the upper end

plat e of the caudal verteb ra. A ra t io bet ween

this distance and the midline vertebral height

of the cranial vertebrae is compared on dy-

namic views. In these cases, fusion surgery

is indicated to prevent p rogression of the in-

stability, correct any segmental deformity,

and treat the axial back pain caused by spinal

instability.

Full-length stand ing coronal and sagitt al ra-

diographs are used for assessmen t of the over-all coronal and sagitt al balance using the cente r

sacral vertical line (Fig. 2.4) and C7 plum bline

Fig. 2 .2 Measureme nt of the slip angle. The angle

is made by the superior end plate o f the caudal

vertebrae and t he inferior end plate of the cranial

vertebrae. The slip angle of a L5-S1 spondylolisthesis

is measured by a line perpendicular to the posterior

aspect of sacrum and a line drawn along the inferior

end of the end p late of L5.

Fig. 2 .3 Measurem ent of the d isk height. A line is

dropped from the midline inferior end plate of the

cranial vertebrae to the upper end plate of the

caudal vertebrae. The ratio be twee n t his distance

and the midline vertebral height of the cranial

vertebrae is compared on dynamic views.




(Fig. 2 .5). The b isector of the cen ter sacral ver-

tical line is also useful for nd ing th e proximal

neutral vertebra. Sagittal balance is measured

by the C7 sagit tal p lum bline, an d lum ba r lor -

dosis is usually measured from the upper end

plat e of T12 to the end plat e of S1. Shoulderheight, apical verteb ral translation of the th o-

racic and lum bar curves, curve m agnitudes, and

exibility should be documented. Common

local deform ities seen on radiographs include

an L2-L3 apex deformity, lateral listhesis or

Fig. 2 .4 Measurem ent of cent ral sacral vert ical line.

Using the top of the iliac crest to contro l for tilting,

a vertical perpendicular line is drawn up from t he

center of S1. The proximal neutral vert ebra can be

bisected from this line, and in this gure it would

be L2.

Fig. 2 .5 Measureme nt of sagitt al C7 plumb line is

done by dropping a vert ical perpendicular line to

the horizontal from the C7 vertebral body and

comparing its horizontal position with the position

of the poste rosuperior corner of the S1 superior end

plate. Sagitt al imbalance is norm ally conside red to

be > 5 cm deviat ion from the S1 posterosupe rior

corner, and in this gure the re is a positive sagitt al

balance.



16 Chapte r 2

rotatory subluxation (Fig. 2.6), lum bar hypol-

ordo sis, an d shor t reciprocat ing curves. Latera l

listhesis or rotatory subluxation is measured

by t he h or izontal d ist an ce be tween the supero-

lateral corner of the caudal vertebra and the

inferolateral corner of cephalad vert ebra. Bend -

ing radiographs can di eren tiate sti curves

from exible curves, but is m ore impor tan t for

deciding on the instrumentation levels during

surgery. MRI is useful to assess disk degen era-

tion an d spinal sten osis. Deform ity correctionis required for these symptoms and may re-

quire more complex operations such as osteot-

om ies and long fusions.

! Diagnosis o f Leg Pain

The classic presen tat ion of nerve root comp res-

sion is buttock pain that radiates to the lower

extremities and neurogenic claudication. Ra-dicular or leg pain points t o spinal or foram inal

stenosis caused by facet joint an d ligamen tum

avum hypertrophy, foraminal narrowing due

Fig. 2 .6 Late ral subluxat ion is measured by the

horizontal distance bet ween t he supe rolateral

corner of the caudal vertebra and the inferolateral

corner of the cephalad vertebra.

to vertebral rotatory subluxation, or reduct ion

of the inte rped icular distance on th e concavity

of the curve. These patients usually develop

bu rning or ach ing pain st ar t ing in the but tock

that radiates down the appropriately involved

derm atom e to the lower leg. Clinical identi ca-tion of the involved de rm atom e provides a good

clue to t he likely nerve root a ected . This can

then be con rmed if there is corresponding

weakn ess in th e sam e m yotome. Typically, im -

pingem ent of th e L4 n er ve root lead s to an te rior

shin numbness with ankle dorsi exion weak-

ness (t ibialis anter ior), an L5 ne rve root involves

the posterolateral calf and foot dorsum, with

extensor hallucis longus weakness, and an S1

ner ve root involves the posterior calf and sole,

w ith weakn ess of the exor hallucis longus.

Neu rogenic claudication com m on ly presents

w ith insidious onset of but tock, thigh, and calf

pain t rigge re d by walking. The usu al disa bilit y

is thus diminished walking tolerance. Vascular

claudication is an imp orta nt di erent ial diag-

nosis. Patients w ith vascular claudication m ay

also present with diminished walking toler-

ance due to calf cramping on exertion or a

sensation of tightn ess that p roceeds from dis-

tal to proximal. This contrasts to neurogenicclaudication whe re d iscom fort with num bness

pro ceeds fro m proxim al to dist al. Object ive

sensory examination should pinpoint the spe-

ci c der m atom e or suggest wh ich ner ve root is

compressed. Motor we akness u sually suggests

a m ore long-standing n erve compression. Vas-

cular examinat ion should be performe d, includ-

ing observation for troph ic changes in th e skin

and nails of the lower limbs and diminished

distal pulses, which wou ld suggest a vasculo-

genic cause for the pain.

Radiographic assessment was described

above (see Diagnosis of Back Pain).

! Factors that May Lead to

Curve Progression, Hence

the Need for Surgical

Treatment

In general, the issue of whe the r curves progress

is debated in th e literatu re, and t he rate of pro-




gression is highly variable. Curves may pro-

gress 1 to 6 degrees pe r year (average 3 degrees

per year ).6,7 Risk factor s for progre ssion include

a pr ior history of progression an d rad iograph ic

risk factors such as a symm etrical disk degen-

eration, lateral disk wedging, and osteophyteformation.8–10 Comparing the two sides of a

spinal segm ent , less than 80% of lateral disk

wedge and more than 5 mm of lateral osteo-

phyte di ere nce m ay indicate an unst able seg-

ment .9 Progression has been suggested t o occur

w ith Cobb angles greater t han 30 de grees, loss

of lum bar lordosis, apical rotation larger t han

Nash-Moe grad e 2 (convex pedicle m igrates 25%

of the vertebral body width and the concave

pedicle gradually disappea rs), lat era l list hesis

of 6 mm or more, or a prominent or deeply

seated L5 disk (in relationship to the inter-

crestal line).11–14 The presence of rotatory

subluxation, lateral spondylolisthesis, and disk

degeneration in the upper lumbar levels also

suggests a risk of progressive deformity.12 Due to

spine coupling, rotatory deformity of th e spine

is related to t he d evelopme nt of lateral spondy-

lolisthesis. Thus, apical vertebral rotation may

also predict scoliosis progression. Spinal seg-

m ent s proxim al to the scoliosis share t he loadto comp ensate for spine im balance. With d isk

degeneration, this compensatory mechanism

fails, an d p rogressive deform ity occurs. Fusion

surgery is required to prevent curve progres-

sion, and the length of fusion is dependent on

the presen ce of coronal or sagitt al imbalance.

! ManagementManagement of adult deformities should be

tailored to each patient because the symptom-

atology is di eren t in ever y case. Treat m en t is

dependent on the experience of the surgeon,

the patient ’s preference, the p atient’s age and

functional status, magnitu de of deform ity, the

rate of progression, and the presen ce of comor-

bidit ies. Som et im es the cause of pain is di -

cult to di eren tiate based solely on clinical and

radiological exam ination. In these cases, trans-foraminal epidural injections, selective nerve

root blocks, and facet joint blocks are comm only

utilized to iden tify the pain gen erator.

Non op era t ive m anagem ent is usu ally re -

served for patients with mild symptoms aris-

ing from stenosis, radicular or back pain, curve

m agnitude of less than 30 degrees, lateral sub-

luxation of less than 2 mm, and reasonable

coronal and sagitt al balance.14 Com m on indi-cations for surgery include axial back pain,

symptom atic deform ity, neurologic symptom s,

and dissatisfaction w ith appea rance. The nal

decision shou ld be a balance of the m agnitude

of surgery, th e qu ality of life gain, and t he risk

of surgical complications. Patients with severe

deform ity may require m ajor surgery to achieve

full correction, and the risk of complications

w ill dram atically increase. Complication rates of

up to 80%have been reported in some series.15,16

Conversely, decompression only or limited fu-

sion m ay be su cient to provide reasonable

and lasting relief for patients. The following

sections discuss the au thors’ experience in su r-

gical decision making, choosing between de-

compression only and sh ort or long fusion, and

the pitfalls of man aging adu lt deform ity.

! Critical Factors in Decision

Making for Surgery

The goal of surgery for these p atients sh ould be

to perform th e smallest operation possible that

would help relieve the symptoms and prevent

a recurre nce. We nd it helps to break down

the componen ts of patient complaints in order

to make an app ropriate decision.

1. Leg pain

a. Nerve root comp ression/spinal stenosis—

local decompression

b. Degenerative spon dylolisthes is—local de-

comp ression ± fusion

2. Back pain

a. For local degeneration or instability—

short fusion

b. For ex ible or corre ct able sagit tal or cor -

ona l imbalan ce—long fusion

c. For sti or un correct able sagittal or coro-

na l im balan ce—long fusion + osteot om ies

3. Progressive deformitya. Long fusion to prevent progression, sel-

dom would be performed alone in the

absence of symptom s above



18 Chapte r 2

! Surgical Decision Making

Decompression Only

Decompression alone is indicated for p atients

with neurogenic claudication without back pain or a sym ptom at ic or progress ive defor-

mity. Decompression is usually in the form

of posterior fenestration or laminectomy, al-

though other approaches, including anterior

indirect decompression and endoscopic trans-

foraminal approaches, have been described.

Each approach has m erit and w ould be depen-

dent on the surgeon’s experience. A detailed

discussion of the ir relative m erits is beyond th e

scope of this chapter. As a general principle,

in those patients undergoing decompression

alone, as much facet joint and as many poste-

rior ligame ntous str uctu res as possible should

be pre se rved to re duce the risk of future pro -

gression and iatrogenic instability.

One shou ld always be prepared t o carry out

a local fusion if more extensive bon e resect ion

is necessary. This is not uncommon as such

individuals often have tight spinal canals, and

incidental durotomies are not infrequent. If a

w ide lam inectomy is perform ed for the repair,a local fusion with pedicle screw xation m ay

be advisable.

Usual indications for decompression-only

surgery include leg pain with minimal or no

back pain , Cob b angles o f less than 30 degre es,

less than 2 m m of lateral subluxation, and n or-

mal coronal and sagittal balance indicated by

the center sacral vertical line and C7 plumb-

line.14 Despite combined back and leg symp-

toms that m ay warrant fusion, decompression

alone may be indicated for patients with sig-

ni cant ly high surgical risk. This may be the

be st op t ion for an eld erly pat ient w ith neuro-

genic claudication, a m ild deform ity, and poor

bo ne qu alit y. Pat ients undergoing decompre s-

sion alone should always be war ned of a risk of

progression of the deform it y that m ay re qu ire

a fut ure fusion procedu re.

FusionFusion surgery is indicated for th e tre atm ent

of back pain du e to de generative changes or in-

stability, as well as to prevent progression of

th e deform ity. It m ay be performe d alone or in

combination with decompression in patients

w ith radicular symp tom s. Shor t fusion m ay be

useful to stabilize curves with signi cant ap i-

cal rotation or when translation or lateral lis-thesis is greater than 3 m m .6,12,17,18 However, if

the re is symptom atic coronal or sagitt al im bal-

ance, realignm ent and long fusion is advisable.

Determination of fusion levels for adult

scoliosis is based o n th e severit y of spinal de-

form ity and the global appea rance and degen-

erative changes of the entire spinopelvic axis.

There is no universal agreement on the length

of fusion an d th e selection meth od of the en d

vertebrae for instrum entation.

Short fusion within the deformity not ex-

ceeding the end vertebrae aim s to stabilize th e

spinal segm ents w ithout correcting the wh ole

deform ity. Its m ajor advantage is th e lower risk

of complications from the anesthesia or the

surgery; th us it is indicated in th ose with back

pain but w ithou t coron al o r sagit tal im ba lance.

Long fusions or fusions exten ding beyond t he

end vertebrae is useful for correction of large

curvature s with coronal or sagittal imbalance,

but th is procedure needs to be balanced againstincreased comp lication rates.

Typical patient s wh o may be m anaged w ith

short fusions (Fig. 2.7) are those w ith smaller

Cobb angles (less than 30 degrees) and minor

rotatory subluxation (lateral subluxations of

more than 2 mm ).14 Back and leg pain an d seg-

mental instability caused by wide decompres-

sions can all be treated w ith short fusions.

Long fusions (Fig. 2.8), which generally

m eans fusion to L5 or th e sacrum , and to T10

or above, yield bette r sur gical correct ion of the

scoliosis and restoration of lumbar lordosis.

They are typically indicated for p atients wh ose

curves are likely to p rogress, such as patient s

w ith Cobb angles greater th an 45 degrees, m ore

than 2 mm of lateral subluxation, and coronal

and sagittal imbalance.14 The aim of the long

fusion is to achieve balance in both the coronal

and sagitt al planes, not absolute Cobb an gle cor-

rection.19 Glassman et al 20 demonstrated that

pos it ive sagit tal balan ce is the sin gle biggest pre dict or of clin ica l symptom s in ad ult defor-

m ity and takes priority over other para m eters.




Thu s, long fusions sho uld always be considered

in patients with global coronal or sagittal im-

balance to ach ieve bet ter funct ional ou tcom es .

It should be borne in mind th at instrum en-

tation sh ould not e nd at the level of junctional

kyph osis or spon dylolisthesis. Any level of se-vere rotatory subluxation should be included

within the fusion block. To balance the spine,

the most horizontal vertebra should be the

upper instrumented vertebra (UIV).21 Instru-

m ent ation should not end at a level with p os-

ter ior column d e ciency, w ith listhesis in any

direction, at a level of a rotated segmen t, with

junct ional kyp hosis, at the ap ex of the defor-mity in the coronal or sagittal plane, or at a

degene rated level.

Fig. 2.7a–e A 71-year-old man with com plaint s of

axial back pain and bilate ral lower limb claudication.

(a) The patient has degenerat ive scoliosis from L3 to

L5, with a Cobb angle of 25 degrees. (b) L3–4 and

L4–5 spondylolisthe sis and sp inal stenosis were

noted. (continued on page 20)

a

b

(text cont inues on page 23)



20 Chapte r 2

Fig. 2.7 a–e (continued) (c) The L5-S1 was well

hydrated, and there was no oblique take-o .

(d,e) Shor t fusion from L3 to L5 was performe d

with good correction of the segmental instability.

c

d

e




Fig. 2.8 a–d A 58-year-old man with axial back pain

and lumbar hypolordosis. There is (a) an oblique

L5-S1 with de generat ive changes and (b) a positive

sagitt al imbalance. (continued on page 22)

a

b



22 Chapte r 2

Fig. 2.8 a–d (continued ) (c) The pat ient was treated

with an L3 pedicle subt ract ion osteotomy and

posterior spina l fusion from T10 to the sacrum with

S1 and iliac instrument at ion. (d) Good restoration

of sagittal balance is observed postoperatively.

c

d




Upper Instrumente d Vertebra for

Long Fusions

In gen era l, for long fusions, the aut hor s’ prefer-

ence is to en d at T10 or above, because instru -

m ent ation an d fusions ending at T12 to L1 have been show n t o have a h igher revision rate, likely

related to the hypermobile t horacolumbar re-

gion as it transitions from an immobile tho-

racic spine to a m obile lum bar spine. There are

also chan ges in facet orient ation from coronal

to sagittal and changes in sagittal alignment

from kyph osis to lordosis. Exten ding th e UIV to

T10 (level with tru e r ibs) or fur the r p roxim ally

can provide relative protection to t he a djacent

segment with the increased stability provided

by the rib cage. The rib cage lengt hens the

transverse dimensions of the spine and gives

the thoracic spine greater resistance to bend-

ing stresses in m ultiple p lanes. T11 or T12 do es

not have costostern al articulations; hence, these

levels lack the biomechan ical advantage of th e

upp er levels.

In add ition, other factors that could a ect

long-term survival of that segment need to be

taken into consideration. These factors include

healthy adjacent spinal segments with no de-generation or instability in any plane, and a

UIV adjacent to spinal segments with normal

sagitta l, coronal, and axial alignm ent and near-

neu tral rotat ion. The UIV should lie w ithin t he

“stable zone” de ned by the cente r sacral ver-

tical line, before surgery, or could be placed

into that zone a fter surgery.

Curre ntly, there is no consen sus stu dy avail-

able to recommend T10 instrumentation in all

pat ient s to im prove long-term resu lt s. Dis-

advantages of the UIV above T10 include in-

creased risk of perioperative complications,

and with longer instrumentation across the

thora colum bar spine th ere is also a greater risk

of pseudar throsis. Thu s, the rat ionale for stop -

ping at T10 m ay not be ap plica ble in all cases.

Important decisions on the extent of instru-

mentation and fusion should depend on the

pos it ion of t he UIV in rela t ionsh ip to t he globa l

spine. An exte nsion to T5 or even higher w ould

depend on th e ability of the lumbar surgery tocorrect th e sagittal imbalance. With control of

more spinal segments, better sagittal balance

m ay be achieved m ore easily.

Ultimately, th e surgical procedure shou ld be

tailored to each patient’s needs and based on

the goal of achieving a well-balanced, stable,

pain less, an d dura ble sp ine w ith the few est

num ber of fused segment s wh ile reducing the

risk of complications associated with large-scale op erations.

Low er Instrumente d Vertebra for

Long Fusions

For th e lower instru m ent ed vert ebra (LIV), m ost

long fusions will extend to the sacropelvis or

stop at L5. In adolescen t idiop ath ic scoliosis, it

m ay be possible to stop at L3 or L4 in a lum bar

curve, but be cause of str uctu ral chan ges and a

xed tilt found at the caudal spinal segments

like L4-L5 in adult deformity, stopping at a

m ore cranial segment is gene rally not ad vised.

Stopping at L5 en ables retent ion of the lum bo-

sacral m otion, avoidance of sacroiliac (SI) joint

stress, decreased operative time and instru men-

tation complications, and a lower pseud art hro-

sis rate. Pelvic xation can also be avoided . On

the other hand, this procedure places a lot of

stress at th e L5-S1 disk, being the only residua lmobile segment, and the patient needs to be

warned of future breakdown and th e need for

surgery to fuse this segment. In general, for

m any pat ient s, th e L5-S1 disk is already degen -

erated, and in su ch cases it is probably better to

fuse to the sacrum. Preservation of the L5-S1

disk enables some pelvic motion, which may

be im por tant for som e funct ional dem ands of

pat ients, such as r id ing a bicycle.

Fusion to th e sacrum is required for disk de-

gener ation at L5-S1, spon dylolisth esis or spina l

stenosis at the sam e segm ent , as well as oblique

take-o at L5-S1 or in fractiona l cur ves greate r

than 15 degrees.22 Balancing is di cult w ith-

out fusion down to the sacrum in cases of

oblique take-o at L5-S1. In addition , th e fora-

men is smaller on one side, leading to unilat-

eral L5 radiculopathy. It is not uncommon to

see pat ients w ith foram inal, central, or lateral

recess ste nosis at L5-S1. If stenosis is p resen t

at L5-S1 an d m ore exten sive decomp ression isrequired, fusing down to th e sacrum is inevi-

table. The obvious disadvantages of fusion to

the sacropelvis include increased operat ive tim e



24 Chapte r 2

and m ore exten sive surgical dissection to reach

the sacru m . Anter ior colum n supp ort m ay also

be re qu ired to reduce the rat e of pseudarthro -

sis. Lost m otion at L5-S1 may also alter t he pa-

tien t’s gait.

Osteotomies may be required if less than30%deformity correction can be obtained on

bendin g radiographs. This is not uncom m on

because adult deform it ies ar e usu ally st i . To

avoid overloading the instrumentation at the

m etal–bone interface, releases and rebalancing

would be required. There are t wo t ypes of sag-

ittal imbalance in adult deformity. First, the

spine is globally balanced but a segm ent al por-

tion of spine is at or kyph otic. Second , th ere

is global and segmental imbalance. Coronal

imbalance can also be classi ed into two types

w ith the sh oulders and p elvis tilted in opposite

directions or w ith tilting in the sam e direction.

Posterior column osteotomies are the best

choice for segmental imbalance of the spine.

A prere quisite w ould be m obile d isk spaces to

allow extension correction. If the disks are al-

ready degenerated and sti , ante rior release is

also required. If the bone stock is inadequate,

anterior structural grafts can be used to im-

prove fusio n rat es. For globa l im ba lance, bot hSmith-Petersen and pedicle subtraction osteot-

omies can be used. Typically, a Smith-Petersen

osteotomy is indicated if the weight-bearing

line falls w ithin 3 cm of the sacrum , and a p ed-

icle subtra ction osteotom y is reserved for cases

w ith poor bone stock, and it can provide 30 de-

grees of lordotic correction. In patients with

combined coronal and sagitt al im balance, ped-

icle subtraction osteotomies are a viable op-

tion if the shoulders an d pelvis are tilted in the

same direction, but a vertebral column resec-

tion is the better option if the shoulders and

pelvis a re t ilted in op posite direct ions.

Anterior procedures are required only in

rigid d eform ities that are not p assively correct-

able with posterior instrumentation. They are

usually used only in combination with poste-

rior instr um ent ation, as inte rbody fusions alone

may not be able to correct the overall sagittal

alignment.23 Anter ior spinal fusion can furth er

correct lumbar hypokyphosis and imbalance, p rovid e in direct decompress ion by foram inal

distraction, prevent posterior instrumentation

failure by load sha ring, and decrease t he r ate of

pseudar thro sis , which is e sp ecially co m m on in

smokers, diabetics, and osteoporotic patient s.14

! Complications

Adult deformity surgery is challenging, and

there are many associated complications. Re-

por ted complica t ion rat es reach 80 % for adult

deform ity, with up to 58%of patient s requiring

reoperation.15,16 These degene rative conditions

usually occur in the elderly with multiple co-

m orbidities such as pulmon ary and card iac dis-

ease, osteoporosis, and nutritional de ciency.

These conditions shou ld be prope rly optim ized

prior to su rgery to d ecrease periop erat ive risks .

Any of the ab ove comorbidities m ay a ect the

tim ing of surgery a s well as the scale of surgi-

cal correction.

Deform ity correction can indirectly decom -

pre ss the neural st ruct ure s by rod derot at ion,

cantilever reduction maneuvers, and particu-

larly by increasing vertebral disk height with

anter ior interbody fusion. Overdistraction onthe concave side may lead to loss of lumbar

lordosis. To redu ce rigid cur ves, poste rior col-

umn osteotomies at multiple levels are likely

required to mobilize the spinal segments. Fu-

sions should avoid stopping at a level of ro-

tatory subluxation to prevent aggravating the

subluxation.

With limited instrumentation and fusion,

degeneration may be accelerated in the re-

m aining cur ve as a result of adjacent segm ent

disease. Stopping the fusion within the de-

formity may provoke these adjacent segment

pro blem s. Stopping the fusio n at the thor aco-

lumb ar junction also leads to adjacent segmen t

disease cranial to th e segm ent of fusion. Fusion

to T10 or above m ay avoid t his. However, some

consider adjacent segment degeneration un-

preventable in fusio n surgery as it could be d ue

to th e nat ural age-related p rogression of a de-

generative process coupled with the postsurgi-

cal e ect of spinal sti en ing created by fusionor instrum entation procedures.24,25




Proximal adjacent segment degeneration

is detected by progressive narrowing of disk

height, progressive decrease in lordosis or in-

crease in kyphosis, osteophyte form ation, scle-

rosis of an adjacent end plate, or translation

in the coronal or sagittal planes. Proximal junct ional prob lem s su ch as ad jacent segm ent

degeneration, compression fracture, or screw

failure in th e UIV occurs m ore frequen tly with

fusions en ding at T11 to L2 as comp ared w ith

those at T10 or above.26

For the LIV, depending on fusion to L5 or to

the sacrum , di eren t comp lications m ay occur,

including L5-S1 disk degenerat ion, loss of cur ve

and balance correction, iliac screw implant

prob lem s, and pse udarthro sis . If the fusio n is

stopped at L5 where there is xed sagitt al im-

balance and disk degenera t ion at L5-S1, the

rate of disk degeneration w ill furt her increase,

leading to loss of sagitta l pro le correct ion and

L5-S1 spon dylolisth esis.22 In osteoporot ic bone,

fusion to L5 has a high risk of xation failure, as

the L5 pe dicles are m ostly cancellous an d t here

are trajectory problems in obtaining a medial

angle for placing th e p edicle screws. Failure of

L5 screws w ith loosening leads to k yphosis or

hypolordosis of the L4-L5 segment. L4-L5 kyphos is m ay be tolerat ed in a sh or t fusio n, b ut

with longer fusions the degree of sagittal im-

balance becom es an issue.

Fusions to the sacru m sh ould be reserved for

L5-S1 spondylolisthesis, stenosis, oblique take-

o , m oderate or severe L5-S1 degenerat ion, and

pr ior lam inect om y. Problem s w ith lon g fusions

to the sacru m include higher complication rates

du e to a large-scale op erat ion, risk of sacroil-

iac joint degeneration, altered gait mechanics

and increased pseudar throsis. Instrum entation

complications for these long fusions include

bre akage and back-ou t or loosening of screws.

To avoid this, S1 screws should be bicortical

through the promontory anteriorly. S1 screws

should also be d irected m edially to avoid pen -

etrating the L5 nerve root. Bone grafting an-

terior to L5-S1 and iliac screws may further

protect the S1 xat ion. To im prove the L5-S1

xation, distal hooks, iliac screws, and inter-

bod y cages for anterior colum n su ppor t ar ealso options. Hooks are an alternat ive xation

especially in osteoporotic bone but m ay cause

sten otic problem s at L5-S1.

Iliac screws entail the risk of pullout 27 and

are usua lly more p rom inent . Screws should be

bu ried if possib le, but , in th in p at ients, re m oval

may be required and should be done around2 years afte r xat ion. Technica lly, iliac screw s

are m ore di cult to insert with previous pos-

terior iliac bone harvesting. There is also a

higher pseudarthrosis rate at L5-S1, but this

m ay be salvaged by revision surgery w ith an te-

rior reconstru ction and iliac xation as well as

using bone morphogenetic protein to improve

fusion rates. The lowest pseudarthrosis rate of

L5-S1 fusions is associated w ith complete sacro-

pelvic xation an d su rger y in pat ients younger

than 55 years of age.28

! Chapter Summary

In adult deform ity, the re is di culty in m atch-

ing a patient’s symptoms and concerns with the

surgical plan. Clinicians must weigh potential

gains and risks, and all surgical decisions should

be individ ually tailore d to the pat ient . Com or - bidit ies sh ou ld be addre ssed p rior to su rgery to

avoid p er ioperat ive complications. Usually, the

surgical opt ions include decomp ression alone,

decompression with limited arthrodesis, and

deformity correction with long fusion (Table

2.1). Decompre ssion surgery is reser ved for pa-

tients with leg pain but minimal or no back

pa in, scoliosis Cobb an gles less t han 30 degrees ,

less than 2 m m of subluxation, no th oracic hy-

perk yp hos is, and acceptable coron al and sagit -

tal balance, or if they have a poor premorbid

state. For short fusions, patients should have

scoliosis Cobb angles less th an 30 d egrees, seg-

mental instability (more than 2 mm of lateral

sublu xation ), back and leg pain , no signi cant

imbalance issues, and, if destabilizing, decom-

pressio n is re qu ired for adequ at e re lief of sp i-

nal stenosis an d n erve r oot comp ression. Long

fusions are reserved for scoliosis Cobb angles

greater than 45 degrees, more than 2 m m of sub-

luxation, and coronal and sagitt al imbalance. Toavoid complications related to instrum en tat ion,



26 Chapte r 2

fusion shou ld not end at a level with junctionalkyphosis or spondylolisthesis, posterior col-

um n de ciency, a rotated segm ent , a level at

the apex of the deformity, or a degenerated

level. Levels of rotatory subluxation must be

included within the fusion block. For balance,

the m ost hor izont al verteb ra shou ld be th e UIV.

Extension of the fusion to T10 p rovides th e in-

creased stability o ered by th e rib cage. The

LIV at L5 is only feasible with a normal L5-S1

disk, and no spondylolisthesis or spinal steno-

sis or oblique take- o at L5-S1. Fusions to the

sacrum should b e avoided if possible to avoid

iliac screw implant problems, pseudarthrosis,

sacroiliac joint problem s, and gait d istu rban ces,

bu t is u sually m an datory for lon g-stan ding sag-

ittal and or coronal imbalances.

Pearls

All surgical decisions for degenerative scoliosis

should be individually tailored to the patient. Decompression surgery is reserved for patients

with leg pain, minimal or no back pain, scoliosis

Table 2 .1 Deco mpressio n Surgery Only Versus Short Fusion Versus Long Fusio n

Sympto m or

Condition Decompression Only Short Fusion Long Fusion

Pain Radicular pain, minimal

or no back pain

Back and leg pain Back and leg pain

Scoliosis Cobb angle < 30

degrees

Cobb angle < 30 degrees Cobb angle > 30 degrees

Subluxation < 2 mm < 2 mm > 2 mm

Overall balance Acceptable coronal andsagitt al balance

Acceptable coronal andsagitt al balance

Global coronal andsagittal imbalance

St abilit y St able m ot ion segm ent Segm ent al inst abilit y, > 50%

pars/ facet excision fordecompression

Segm ental and regional

kyphosis

Ope rat ed le vels St enot ic le vels only Rotatory subluxat ionsegme nts within fusion

block, segm en tal

instability caused by widedecompression

UIV: T10LIV: L5 if no degeneration,

spondylolisthesis,

stenosis or obliquetake-o at L5-S1

Lim it at ions Cannot a ddress global ba lance, progressive

deformity, segme ntalinstability with wide

decom pre ssion

Higher surgical risk, cannotaddress global balance,

adjacent level disease

Highe st surgical risk,compromised xation

with osteoporosis, highrisk of pseudarthrosis,

iliac screw prominence

Abbreviations: UIV, upp er inst rum ented vert ebra; LIV, lower instrumented vert ebra.

Cobb angles less than 30 degree s, less than 2 mm

of subluxation, no thoracic hyperkyphosis, accept-

able coronal and sagittal balance, or those with

po or p remorb id state.

Short fusions are for scoliosis Cobb angles less

than 30 de grees, segm ent al instabilit y, back and

leg pain, and no signi cant imbalance issues.

Long fusions are for scoliosis Cobb angles greater

than 45 degrees, more than 2 mm of sublux-

ation, and coronal and sagitta l imb alance.

Extension of th e fusion to T10 provides increased

stab ility o ered by the rib cage.

The m ost horizont al vert ebra should be the UIV.

Pitfalls

Fusion should not end at a level with junctional

kyphosis or spondylolisthesis, posterior column

de ciency, a rotated segm ent , at the ape x of the

deformity, or a dege nerated level.

Avoid the LIV ending at L5 with an abn orm al

L5-S1 d isk, spo ndylolisthesis, sp inal stenos is, or

oblique take-o at L5-S1.

Fusions to t he sacrum should be avoided if possi-

ble due t o the increase d risk of iliac screw implan t

pro blems, pseu dart hro sis, sacroiliac joint pro b-

lem s, and gait disturbances.




References

Five Must -Read Referen ces

1. Schwa b F, Dub ey A, Gam ez L, et al. Adu lt sco liosis:

pre valence, SF-36, an d nut rit ion al pa ram et er s in an


1085 PubMed

2 . Kirka ldy-Willis WH, Wedge JH, Yon g-Hin g K, Reilly J.

Pathology and pathogenesis of lumbar spondylosis

and stenosis. Spine 1978;3:319–328 PubMed

3. Dubousse t J. Thre e-d ime nsion al analysis of th e sco-

liotic deformity. In: Weinsteid S, ed. The Pediatric

Spine: Princip les and Pra ctice. New York: Raven Press;

1994

4. Luk KD, Che un g KMC. Lum bar spin al in st ab ility.

Hong Kong Journ al of Orthopa edic Surgery 1998 ;2

5. Luk KD, Chow DH, Holmes A. Vertical instability in

spondylolisthesis: a traction r adiographic assessme nt

technique and the principle of managem ent. Spine

2003;28:819–827 PubMed

6. Bradford DS, Tay BK, Hu SS. Adult scoliosis: surgical

indications, operative management, complications,

and outcomes. Spine 1999;24:2617–2629 PubMed

7. Grubb SA, Lipscomb HJ, Coonrad RW. Degenerative

adu lt onset scoliosis. Spine 1988 ;13:24 1–245 PubMed

8. Benner B, Ehni G. Degenerative lumbar scoliosis.

Spine 1979;4:548–552 PubMed

9. Jim bo S, Kobayas hi T, Aon o K, Ats ut a Y, Mats uno T.

Epidemiology of degenerative lumbar scoliosis: a

comm unity-based cohort study. Spine 2012;37:1763–

1770 PubMed

10. Kobaya sh i T, Ats ut a Y, Takem itsu M, Mat su no T,

Takeda N. A prospective study of de novo scoliosis

in a community based cohort. Spine 2006;31:178–

182 PubMed

11. Grubb SA, Lipscom b HJ. Diagnost ic nd ings in pain-

ful adult scoliosis. Spine 1992;17 :518–527 PubMed

12. Pritchet t JW, Bort el DT. Degen erat ive sym ptom atic

lum bar scoliosis. Spine 19 93;18:70 0–703 PubMed

13. Robin GC, Spa n Y, Ste inb er g R, Makin M, Men czel J.

Scoliosis in the elderly: a follow-up study. Spine

1982;7:355–359 PubMed

14. Silva FE, Len ke LG. Adu lt de gen erat ive scoliosis: eva -

luation and management. Neurosurg Focus 2010;

28:E1 PubMed

15. Car reon LY, Puno RM, Dima r JR II, Glassm an SD, John -

son JR. Perioperat ive com plications of poster ior lum -

ba r decom pre ssion an d ar th ro des is in old er ad ult s. J

Bone Joint Surg Am 2003;85 -A:2089–20 92 PubMed

16. Edwards CC II, Bridwell KH, Patel A, Rinella AS, Berra

A, Len ke LG. Long a du lt de form ity fus ions t o L5 an d

the sacrum. A matched cohort analysis. Spine 2004;

29:1996–2005 PubMed

17. Sapkas G, Efstat hiou P, Bad ekas AT, Ant on iadis A,

Kyratzoulis J, Meleteas E. Radiological parameters as-

sociated w ith t he evolution of degener ative scoliosis.

Bull Hosp Jt Dis 1996;5 5:40 –45 PubMed

18. Tribu s CB. Degen erat ive lum bar scoliosis: evaluation

and management. J Am Acad Orthop Surg 2003;11:

174–183 PubMed

19. Simm ons ED. Surgical treat m ent of patient s w ith

lumb ar sp inal stenosis w ith associated scoliosis. Clin

Orthop Relat Res 2001 ;384:45– 53 PubMed

20. Glassm an SD, Ber ven S, Brid well K, Hor ton W,

Dim ar JR. Correlation of radiograp hic param eter s and

clinical symptom s in ad ult scoliosis. Spine 2 005;30:

682–688 PubMed

21. Simm ons ED Jr, Simm ons EH. Spin al ste nos is w ith

scoliosis. Spine 1992 ;17(6, Supp l):S117– S120 PubMed

22. Bridwell KH. Selection of instr um ent ation an d fusion

levels for scoliosis: wher e to star t an d w here t o stop.

Invited submission from the Joint Section Meeting

on Disorders of the Spine and Peripheral Nerves,

March 2004. J Neurosu rg Spine 2004;1:1 –8 PubMed

23. Cho KJ, Suk SI, Park SR, et al. Shor t fusion versu s lon g

fusion for degenerative lumbar scoliosis. Eur Spine J

2008;17:650–656 PubMed

24 . Ghiselli G, Wang JC, Bha tia NN, Hsu W K, Dawson EG.

Adjacent segment degeneration in the lumbar spine.

J Bone Joint Sur g Am 2 004 ;86- A:14 97–1 503 PubMed

25. Kum ar MN, Baklan ov A, Chop in D. Corr elation be-

twe en sagittal plane changes and adjacent segment

degeneration following lumbar spine fusion. Eur

Spine J 2001;10:314–3 19 PubMed

26. Shu ebar ger H, Suk SI, Mardjet ko S. Debate: deter -

mining the upper instrumented vertebra in the ma-

nagement of adult degenerative scoliosis: stopping

at T10 versus L1. Spine 2006;31(19, Suppl):S185–

S194 PubMed

27. Weist ro er JK, Perr a JH, Lonste in JE, et a l. Com plica -

tions in long fusions to th e sacrum for adult scoliosis:

m inimum ve-year analysis of fty patien ts. Spine

2008;33:1478–1483 PubMed

28 . Kim YJ, Bridw ell KH, Len ke LG, Rhim S, Cheh G. Pseu-

darthrosis in long adult spinal deform ity instrum en-

tation and fusion to the sacrum : prevalence and r isk

factor analysis of 144 cases. Spine 2006;31:2329–

2336 PubMed



! Introduction

The use of spinal osteotomies in severe spinal

deformities has enabled corrections that were

not considered possible in the past. With ad-

vanced posterior-based techniques, excellent

corrections are achieved through a single ap-

proach , shor tening the d urat ion of surgery and

redu cing the n eed for multiple position chan ges

dur ing surgery. Although t he m ajority of cor-rections can be perform ed from th e posterior

direction, selective deformities may require

combined an terior procedures.

With the improvement in surgical tech-

niques and neuromonitoring modalities, ob-

taining correct ions of severe spinal deform ities

is now both possible an d reason ably safe.1 Thor -

ough kn owledge of advanced an atomy, careful

pre op erat ive planning, an d sp ecialized inst ru-

me ntation and implants provide the necessary

tools for successful surgery. This chapter re-

views t he various osteotom ies, the indications

for their use, as well as the methods of maxi-

m izing corrections and m inimizing both short-

and long-term complications.

! Planning the Deformity

Correction

Although deciding whet her or n ot to operate is

the rst m ain decision to be made, planning the

ner d etails of the procedure will help ensure a

smooth er ow of the surgery. The main plan-

ning should be d one p reoperat ively, and an a l-

gorithm for key decisions shou ld be esta blished

pre op era t ively and discussed w ith the pat ient

and fam ily. For exam ple, if th e pat ient d oes not

wish to assume the increased risk associated

with achieving a more complete deformity

correction, it is important to discuss what can

be achieved w ith lesser re leases. Con versely, if

correction is a key comp onen t of the patient ’sexpectat ions and t he su rgical team can reliably

achieve t hese goals safely, a th ree-colum n oste-

otomy can be performed if lesser osteotomies

are unsuccessful.

Determining the Flexibility

of the Deformity

Using the least risky procedure to obtain a cor-

rection is key to the safe outcome of deform ity

surgery. If a similar correct ion can be obtained

through multiple posterior column releases, a

three-colum n osteotomy m ay not be necessary.

Determ ining the exibility of the curve can

often be di cult, and intraoperat ive adjust-

m ents m ay be required in cases wh ere the curve

is m ore sti or less sti tha n expected.

Helpful clues to cur ve exibility include th e

pre se nce of w ide disk sp aces, disk sp aces that

open and close on bending lms, and curve

magnitudes that decrease when the p atient isin the prone position or w ith tra ction views. If

computed tomography (CT) imaging demon-

strates ant erior fusions, either congenital or from

3

The Use of Osteotomies forRigid Spinal Deformities

Ste phe n J. Lew is and Simo n A. Harris



The Use of Osteotom ies for Rigid Spinal Deformities 29

p revious surgery, th ese fu sions w ill not corre ct

with posterior releases, and three-column os-

teotomies will be required. In contrast, good

corrections can be a chieved w ith poster ior col-

um n releases through previous posterior fusion

masses that h ave not undergone previous an-terior fusions. Proper preope rative worku p w ith

long-cassette anteroposterior (AP) and lateral

side bend ers, CT scan, and m agnetic resonance

imaging (MRI) should be done preoperatively,

so that the best possible preoperative plan

can be made. Newer technologies with three-

dime nsional (3D) printers can provide surgeons

w ith preop erative m odels of the spine, to even

bet ter pre pare for the ult im at e pro cedure .

Exposure

Excellent exposure is an essential component

of the procedu re. Severe deform ities can m ake

this m ore challenging; however, taking the t ime

to obtain the necessary exposure will greatly

facilitate implant insertion, and generally im-

prove t he ow of t he p rocedure . It is im por tant

to ident ify the spine levels, areas w ith previous

decompre ssions, fusion m asses, and previous

implants. In cases of revisions, knowledge of p rev ious sp in al in st rum entat ion w ill ensu re

that t he required instrum ents are available to

facilitate implant removal.

Spinal Cord Blood Flow

The blood ow to the spinal cord ente rs thedura t hrough vessels that t ravel with th e exit-

ing nerve root. Although n erve roots a re com -

m only sacri ced in th oracic-level osteoto m ies,

taking a ner ve root at the level of the arte ry of

Adamkiewicz could lead to signi cant detri-

m ent t o the spinal cord circulation.2 This arter y

has variable anatomy, but is present between

T8 and L1 on the left side in the majority of

peop le. When consid ering ost eotom ies a rou nd

the thoracolum bar junction, protecting and sav-

ing the nerve roots may preserve key sources

of blood ow.

For thoracolumbar three-column osteoto-

mies, preoperative angiography can be per-

formed to determ ine the exact location of the

arte ry of Adam kiewicz. The ar ter y run s a char-

acteristic intradural “hairpin” loop on imag-

ing3 ( Fig. 3.1). The location of the artery may

in uen ce the choice of level of th e osteotom y,

and th e surgeon m ay choose a level other t han

the ap ex if the ar ter y is present at t he ap ex. In- ju ring this vessel, especially in the pre sence of

Fig. 3.1a,b Spot image (a) and inverse (b) shots

of angiography of the left T11 segme ntal arte ry

showing the characteristic intradural hairpin loop

(white arrow), represent ing the artery of Adam -

kiewicz. In th is pat ient , the vessel enters the dura

through t he left T11 forame n and forms the loop

that extends up to T10. With the vessel arising two

levels proximal to the apex, a vert ebral column

resect ion was performed at L1 without incident .

a b



30 Chapte r 3

hypotension, can lead to a loss of intraopera-

tive motor evoked potential (MEP) monitoring

that is often delayed from the time of injury.

With spinal cord infarction as one of the m ain

risks of spinal cord level osteotomies, knowl-

edge of and at tention to th is artery m ay help preven t th is devast at in g com plicat ion , esp e-

cially in patient s with previous anter ior proce-

dures, whe re segmen tal vessels may have been

ligated.

Fixation

Achieving adequ ate and stable xation is es-

sent ial to obtaining and m aintaining deform ity

correction. Although the pedicle screw is themain anchor in the majority of constructs, al-

ternatives such as hooks, laminar screws, fu-

sion mass screws or hooks, wires, and bands

should be considered whe n ped icle screw xa-

tion is not possible.4 Obtaining adequat e prox-

imal anchors is generally the key determinant

of successful constructs in thoracic osteoto-

mies. Osteotomies should not be attempted

unless solid proxim al and distal xation is es-

tablished. Careful planning from the preop era-

tive im ages will help to ident ify and select t he

approp riate anchor for each level.

During osteotom y closure, various m ethod s

can be utilized to protect the main implants.

Tem porar y devices or implant s can be u sed to

close th e osteotom ies, such as cent ral rod con-

structs, sparing the main screws.5 The use of

per iapical reduct ion scre w s, tubes, or ot her

extenders on the screws, linking of multiple

anchors to th e rod before cantilevering the re-

duction, and the use of a three- or four-rodtechnique with connectors can facilitate reduc-

tion of the osteotom y and correction of the d e-

form ities wh ile protecting the m ain anchors.

Determining the Desired

Correction

The imaging should be carefully studied to

identify the deform ity and determ ine the t ype

and m agnitude of the desired correction. Care-ful underst anding of the nor m al sagittal align-

m ent, the pelvic param eters, and th e m agnitude

of the deformity will help to identify which

osteotomies would be required to gain the de-

sired correction.6,7

For xed kyph otic deform ities, correct ion

w ill be achieved t hrough a nter ior length ening,

pos terior shor tening, or a com binat ion of both.For coronal deformities, correction will be

achieved through concave lengthening, convex

shorte ning, or a combination of both. For xed

lordosis, correction can be achieved through

anterior shortening, posterior lengthening, or

a com bination of both. For m ultiplanar d efor-

mities, it is important to identify the primary

deform ity or deform ities, and tailor an osteot-

omy or combination of maneuvers to achieve

the desired correct ion. For example, for a xed

kyphotic scoliosis, a combination of posterior

shorte ning and convex shorten ing could be the

pr im ary m od e of correct ion. If a verteb ra l col-

um n re section (VCR) were to be per form ed, a

larger anterior cage placed on the concavity

could maximize correction. For xed hyper-

lordosis, a form al anter ior release or resection

could be combined with posterior colum n re-

leases to achieve the d esired correction.8

The magnitude of the deformity must be

considered. Rough estimates of potential cor-rection th rough a single osteotomy include 10

degrees of sagitt al or coronal plane correction

through a single posterior column release, 30

to 35 d egrees of sagitt al and 10 to 15 d egrees of

coronal plane throu gh a single pe dicle subtrac-

tion osteotomy (PSO), and 30 to 50 degrees of

correct ion th rough a VCR in th e coron al or sag-

ittal plane.9,10 For a VCR, more correction will

be ach ieved t hrough a d efor mity w ithou t a pre -

vious fusion compared with one that is previ-

ously fused , as correction w ill be achieved only

through the osteotomy site and not th rough the

adjacent segments in cases of previous fusion

masses. Properly estimating the desired cor-

rection relative to t he deform ity w ill help plan

the num ber and types of osteotomies required

to achieve th e desired correction.11

Deciding the Leve l of the

OsteotomyFor p osterior column releases (Sm ith Petersen ,

Ponte), multiple periapical osteotomies will




help achieve a gradual, multilevel correction

for deformities with m obile ant erior colum ns.

For t hr ee- column osteoto m ies (PSO, PSO vari-

ants, VCR), the preferred vertebra would be at

the apex of the deformity, not t ilted in th e cor-

onal or sagittal planes, and would be appropr iat e for proxim al an d dist al xat ion. Other

considerations include the location of the ar-

tery of Adam kiewicz and the presence of pseu-

darthrosis in cases of revisions, in which cases

it would preferable be include the nonfused

levels in th e osteot omy.

Planning an Osteoto my

Putting all the information together will help

to determ ine the best opt ion for deform ity cor-

rection. A represen tative case is a 66-year-old

woman with ankylosing spondylitis (Fig. 3.2).

Preoperative im aging w ith long-cassette radio-

graphs and CT demonstrated the autofusion

of her spine. Her chief complaints are sagitt al

imbalance and di culty w ith forw ard gaze.

Her pelvic incidence m easures 55 degrees, thelumba r lordosis 10 degrees, w ith a sacral slope

of 5 degrees and a pelvic tilt of 50 degrees.

With th e desired lum bar lordosis being 10 de-

grees less than the pelvic incidence, and the

desired pelvic tilt being less than 25 degrees,

she wou ld require ~ 35 degrees of lum bar lor-

dosis. This can best be achieved throu gh a single

lum bar PSO.

As for her th oracic spine, she has signi cant

complaints related to her gaze. Her thoracic

kyphosis from T5 to T12 m easures 25 degrees,

wh ich is within the norm al range. Her T2-T5,

Fig. 3.2 a–d Represent ative case of a 66-year-old

woman with ankylosing spondylitis as demonstrated

on the p reoperative st anding posteroant erior radio-

graph (a) and the sagitt al CT reconstruction (b).

Abnorm al sagit ta l alignm ent is characterized by

a low sacral slope (SS), high pelvic t ilt (PT) and

insu cient lum bar lordosis (LL, T12-S1) for the

given pelvic incidence (PI). To maintain a balanced

relationship of the PI and LL, an L2 ped icle subt rac-

tion osteotomy (PSO) was performed. Forward gaze

was improved with a T3 PSO to correct the p roximal

tho racic kyphot ic deformity. Stabilizat ion of t his

correction was achieved with a C2 to pelvis construct

as demonstrated on the standing postoperative

long-casset te posteroant erior (c) and lateral (d)

radiographs.

a b c d



32 Chapte r 3

however, m easures 45 d egrees, which is greater

tha n th e 10 to 15 degrees expected for this re-

gion. A single PSO in t his region w ould p rovide

the n ecessary correction to improve her gaze.

This patient und erw ent a T3 an d an L2 PSOs

with a C2-to-pelvis stabilization through asingle-stage procedur e, addressing both of her

deformities and providing her with the neces-

sary sagittal balance.

Single Proce dure or Staged

Although it m ay be preferable to complete t he

surgery in one stage, certain factors may ne-

cessitate performing the procedure in two or

more stages. These factors include excessive

bleeding, long durat ion of t he su rgery, medica l

comorbidities, and di culties w ith neu rom on-

itoring. Recognizing th ese di culties pre ope r-

atively may help to electively plan performing

these surgeries over two separate days. The

bene ts of st aging include m inim izing op era -

tive team fatigue, postpon ing the bleeding por-

tion of the procedure to the second day, and

the possibility of obtaining prop er im aging to

check the position of the instrum entation pr ior

to th e second stage. The t iming betw een stagesis controversial. Some advocate a short time

of 1 to 2 days, wh ereas others recomm end 1 to

2 weeks to allow p atients to achieve their nor-

mal nutritional status before proceeding. Lo-

gistical issues of operative time and surgical

team availability, as well as patient and fam-

ily issues, also need to be considered in the

decision.

The Surgical Team

Having a strong, cohesive surgical team with

open communication is essential to the suc-

cess of these complex reconstructions. Ideally,

the team should include an experienced spine

surgeon and anesthesiologist, skilled surgical

assistant s, a nursing team fam iliar w ith the in-

strum entation and procedure, an experienced

neuromonitoring and radiology technologist,

and a blood conservation team . Open com m u-

nication is impor tant , and such issues as blood pressu re par am et ers, b lood conse rvat ion st rat -

egies, neurom onitoring changes, and inform a-

tion about the surgical eld and the stage of

the procedure should be reviewed frequently

throughout the case.12

Obtaining Fusion Across the

Osteotomy

Obtaining a solid fusion across the osteotomy

is imp orta nt in preventing early imp lant failure

at t he level of the osteotom y. Although m ultiple

rods can increase t he rigidity of the construct s,

having stable anterior and posterior columns

w ith bridged struct ural bone across all defects

is key to obtaining fusion. Anterior grafts are

not su cient to overcom e large posterior col-

umn defects. Resected ribs can be preservedin the procedure an d used to bridge posterior

column defects following osteotomy closure.13

Techniques of fashioning the rib and the host

bed, w ir ing r ibs in place, or usin g m in i-screw s

from the craniofacial intern al xation sets to

secure the ribs will help re-create the struc-

tu ral continu ity of the p osterior column .

! Osteotomy OptionsSpinal osteotom ies can be divided into six main

types14 ( Fig. 3 .3):

Posterior colum n:

1. Partial facet

2. Complete facet

Partial body:

3. Pedicle subtr action osteot omy (PSO)

4. Transd iskal ped icle subt raction osteotomy

Complete body:5. Ver teb ra l colum n resect ion (VCR)

Multiple vertebra e:

6. Multiple vertebr al colum n resection

In this classi cation, the approa ch m odi er

was added. If the procedure was performed

from posteriorly, the osteotom y would h ave a

“P” after th e n um ber. If a combined anter ior

and posterior sur gery w as pe rformed , an “A/P”

would be added after the number. For exam-

ple, if a PSO was p erform ed fro m poster iorly, itwould be considered a type 3P osteotomy. A

VCR performed through a combined anterior




and posterior approach wou ld be considered atype 5 A/P.

Types 1 and 2: Posterior Column

Osteotomies

Release of the facet joints an d th e post erior lig-

am entous stru ctures, including the ligament um

avum , provides signi cant m obility to the pos-

terior column . For any correction to occur, the

anterior column has to be mobile. With thecombination of a m obile anter ior column and a

released posterior column, signi cant correc-tion can be achieved in both the coronal and

sagitt al planes (Fig. 3.4). For kyphosis correc-

tion, this osteotomy provides a combination of

po ster ior sh or te ning and ante rior lengthening.

Type 1 o steotom ies involve resection of th e

infer ior facets. This can pr ovide some m obilit y

in th e poster ior colum n. Type 2 osteotomies

involve removal of the superior facet and liga-

mentous structures. Resection of the superior

facet is the key to the release. This can beachieved w ithout resecting the inferior facets,

Fig. 3 .3a–f Schemat ic of the compre-

hensive anatom ic spinal osteotom y

classi cation proposed by Schwab et al.

In th is classi cat ion, Type 1 (a) is a part ial

facet resect ion, t ype 2 (b) is a complete

facet resect ion, t ype 3(c)

is a ped iclesubtraction osteotomy, type 4 (d) is a

transdiskal pedicle subtract ion osteotomy,

type 5 (e ) is a vert ebral column resect ion,

and type 6 (f) is a multi-level vert ebrec-

tomy. (From Schwab F, Blondel B, Chay E,

et al. The comprehensive anatomical

spinal oste otomy classi cation. Neurosur-

gery 2014;74:112–120, discussion 120.

Reprinte d with perm ission.)

a b c

d e f

Fig. 3 .4a–f Long cassett e standing (a) posteroante-

rior, (b) lateral, and (c) sagitt al computed tomogra-

phy (CT) reconst ruction of a 17-year-o ld boy with an

L2 congenital kyphosis. Note the global compe nsa-tion of the deformity through thoracic and lumbar

hyperlordosis. Posterior column osteotomies were

pe rform ed at L1-L2 and L2-L3, with correct ion of

the deformity and stabilization from (d) L1 to L3,

allowing for (e ) th e spont aneous no rmalization of

the thoracic kyphosis and a decrease in (f) thecompensatory lumbar hyperlordosis.

a b c d e f



34 Chapte r 3

especially when signi cant distraction of thefacets occurs, as is the case w ith large kyphot ic

deformities. Preservation of the inferior facet

during the osteotomy can help to maintain

poste rior -colu m n bone stock , aid ing in the p os-

terior fusion. Use of a hook-based temporary

central rod to facilitate osteotomy closure fol-

lowing the posterior column release produced

~ 10 degrees of correction per osteotomy level

in the t horacic spine (Fig. 3 .5).

Type 3: Pedicle Subtraction

Osteotomy

The PSO is a posterior-based closing-wedge

osteotomy. It is ideally suited for kyphosis

correction and can reliably produce 25 to 35

degrees of lordosis even in the presence of a

solidly fused anter ior colum n15 ( Fig. 3.6). Per -

forming the PSO asymmetrically can enable

concomitant coronal plane correction. A PSO

can be performed in both the thoracic andlum bar spines. In cases of pelvic inciden ce (PI)

and lumb ar lordosis (LL) m ismatch w ith an as-

sociated abnormally high PI, a sacral PSO can be perform ed to decrease the PI an d nor m alize

th e PI–LL relat ionsh ip.16

The main complications associated with

PSOs are bleeding, potential nerve root injury

or ent rapm ent, and pseudarthrosis. The tech-

nique is discussed below. Careful attention to

detail can help m inimize the poten tial m orbid-

ity that can be seen with th ese cases.

Technique

Multiple variations of the technique h ave bee n

described, but the principles of the procedure

are com m on to all of them .

Decompression

Following exposure an d implant insertion, the

pedicle is iso lat ed fro m all of it s bon y at tach-

ments: laterally, the transverse process; dis-

tally, the pars; and proximally, the superiorfacet. A complete laminectomy of the involved

level is performed as well as of some or all of

Fig. 3.5a,b (a) Schem atic of a poste rior column

osteotomy with resection of the superior facets.

(b) Osteotomy closure is achieved with a temporary

central hook-rod construct reducing the inferior

facet to the proximal surface of the pedicle. Note

the exible anterior column allowing ant erior

lengt hening with oste otomy closure. (From Lewis SJ,

Goldstein S, Bodrogi A, et al. Comparison of pedicle

subtraction and Smith-Pete rsen osteotom ies in

correct ing thoracic kyphosis when closed with a

central hook-rod construct. Spine 2014;39:1217–

1224. Reprinted with perm ission.)

a b




the adjacent levels to ensure adequate space

for th e d ural sac cent rally up on closure. Com -

plete resect ion o f th e p edicle is requ ired t o cre-

ate a single foram en for t wo n erve roots—the

ner ve root of the osteotomy level as well as thener ve root of the level proxim al. The complete

pos te rior ele m ents of the verteb ra of the os te-

otomy level should be resected. A triangular

poste rior -b ased w edge o f bone is then rem oved

from the body, leaving a small amoun t of ante-

rior bone of the vertebral body. The anterior

colum n acts as a h inge du ring closure.

For thoracic-level osteotomies, the trans-

verse processes are rem oved to reveal the m e-

dial rib. The r ib is dissecte d free from all its soft

tissue attachm ent s, taking care to avoid ente r-

ing the pleura l space. The r ib is the n cut 5 to

6 cm lateral from th e vertebra. Subpe riosteal

dissection is done to free up the medial rib,

wh ich is then d etached from the lateral aspect

of the vertebral body. Dissection along the

lateral pedicle and body is then performed to

free th e m ediastinum from the ventral verte-

bra l bod y. Spoon re t rac tor s can then be placed

around th e anterior vertebra to furth er protect

the m ediastinal str uctu res. Retract ing the spi-nal cord should be avoided dur ing the decom -

pre ssion to m in im ize iat rogenic in ju ry.

Minimizing Bleeding During the Osteotomy

The epidural veins run a predictable course;

ident ifying, coagulating, and cut ting the m can

m inimize blood loss dur ing the procedu re. The

veins run t hrough the epidural fat and sh ould

be coagulat ed w hile separa t ing the fat from the

dura. A second series of veins run along the

medial aspect of the pedicle, distally along

the cour se of th e exiting ner ve root and proxi-

mally over the pedicle and deep to the su-

perior facet . W hen reach ing arou nd ventrally,

care should be made to avoid the segmental

vessels running along the midportion of the

lateral vertebral body. As well, failure to sepa-

rate the plane of the mediastinum from theventra l body can lead to signi cant m ediasti-

nal venous bleeding during dissection lateral

to the vertebral body. It is imperative to stay

along the lateral aspect of the vertebra w hen

dissecting an teriorly.

Osteotomy Closure

Closure of the osteotomy is perform ed after en -

suring adequate rese ction of the posterior wall

of the vertebral body and after complete re-section of the p edicles has bee n pe rform ed. If

di culty is encoun tered closing the osteotomy,

Fig. 3.6 a–e Long casset te stand ing (a) lateral and(b) sagitt al CT reconst ruction of a 67-year-old man

who underwent a previous anterior and posterior

L2-L4 fusion for an L3 bu rst fracture. Intraoperat ive

views with (c) th e t emporary cent ral rod in place

and postoperative long casset te (d) late ral and(e ) sag it ta l T2 m agne tic resonance imaging (MRI)

dem onstrating restored sagitt al alignme nt following

an L3 PSO and T10 to pe lvis construct .

a b c d e



36 Chapte r 3

the surgeon should consider resecting more

bon e an ter iorly for adequ at e decom press ion.

Inadequate bone resection is the m ain reason

osteotom ies do not close.

To judge th e red uction of the p osterior col-

um n, the inferior facet of th e level proximal tothe osteotomy can be preserved and reduced

to the superior facet of the level distal to the

osteotomy. This w ill ensure a stable posterior

column with structural bone continuity and

preve nt oversh or tening during closu re. Closu re

can be achieved through the use of a hook-

based centra l ro d (Fig. 3.7), through com pres-

sion of the periapical anchors w ith tem porary

short rods, or with th ree- or four-rod constructs

using side-to-side rod connectors. If posterior

colum n continuity cannot be achieved th rough

osteotomy closure, struct ural bone graft (from

adjacent ribs or large spinous processes) can be

used to ll the posterior defects.

Type 4: Transdiskal Variant

Modifying the proximal resection of the PSO

to exten d across th e d isk space provides for a

greater resection and enables bone-on-bone

contact th rough th e ante rior colum n. This vari-

ation is particularly useful in cases of diskitis

w ith kyph otic collapse (Fig. 3 .8) and posttrau-

m atic kyphosis.17

PSO w ith Previous Ante rior Implants

Anterior implants at the level of the planned

osteotomy present a challenge wh en perform -

ing posterior-based procedures. The implants

can be rem oved either th rough a form al ante-

rior approach or through an anterior reach-

around procedure from a p osterior approach18

(Fig. 3.9). Posteriorly th e tran sverse processes

are rem oved, and dissection is performe d along

the lateral aspect of the pedicle. The anterior

implant s are ident i ed. Taking care to preserve

the exiting nerve root, a m etal cutting bur can

be used to cu t the anterior ro d proxim al an d

distal to the ante rior screw. Som e of the lateral

bod y is then re m oved to ident ify the neck of

the screw, which is then cut with the bur. The

segment of the anterior screw with the at-

tached rod is rem oved. The osteotom y is then

Fig. 3.7a,b Schematic of (a) a p edicle subt raction

osteotomy closed with (b) a cent ral rod. Note the

reduction of the inferior facet of the proximal level

to the superior facet of the distal level, re-creatinga new facet joint and continuity of the posterior

column. (From Lewis SJ, Goldstein S, Bodrogi A,

et al. Comparison of pedicle subtraction and

Smith-Pete rsen oste otom ies in correct ing t horacic

kyphosis when closed with a cent ral hook-rod

construct. Spine 2014;39:1217–1224. Reprintedwith perm ission.)

a b




perfor m ed in the usu al fash ion and the re-

m aining shaft of the screw is rem oved w ith the

vertebral body resection (Fig. 3 .10 ).

Type 5: Vertebral Column

Resection

For large m ultiplanar deform ities, resection of

a complete vertebral body can provide the

mobility in the spine to achieve the neededcorrection.19 Severe kyp hot ic deform ities, like

those seen following tu berculosis, often requ ire

more extensive resections involving multiple

vertebrae to achieve the needed correction.

Common indications for VCR include severe

kyphoscoliosis, congenital deformities (Fig.

3.11 ), and rigid deform ities secondar y to pre -

vious surgery.20

The procedure is performed in a similar

fashion to a PSO. Although the PSO is often

perfor m ed for p r im arily sagit tal plane defor -

m ities, the VCR can a ccom m odate m ultiplanardeform ities. These deformities often have m ajor

rotational an d translational components, causing

Fig. 3.8a–h Long cassette (a) lateral and (b) sagit talT2-weighted MRI of a 73-year-old woman with

known tuberculosis unresponsive to medical

treatm ent . Note t he destruct ion and kyphotic

collapse of the T10-T11 d isk space and (c) adjacent

vert ebral bodies with an associated e pidural abscess

noted on (d) gadolinium-enhanced T1-weighted

MRI. Long casset te (e ) posteroant erior and

(f) lateral views demonst rate a T4 to L2 post erior

reconstruction. (g ) A transdiskal pedicle subt raction

osteotomy was performed by resecting the posterior elemen ts and ped icles of T11, the

proximal vert ebral body of T11, the T10-T11 disk,

and t he d istal vertebral body and end p late of T10.

(h) A new vertebral body was creat ing by reducing

the proximal body of T10 to the dista l verteb ral

body of T11. Note t he inferior face t o f T10 was

reduced to t he supe rior facet of T12 to maintain

the integrity of the poste rior column.

a b

c

d e f

g h



38 Chapte r 3

Fig. 3.9 a–h Long casset te (a) posteroant erior and

(b) lateral view, sagittal T2 MRI (c) , and CT coronal

(d), sagitta l (e ), and rep resent at ive axials (f) of a

69-year-old woman with coronal and sagittal

malalignm ent following a previous ante rior T12to L5 fusion and circumferent ial extension to t he

sacrum and pe lvis. Thoracic kyphosis (T5-T12)

measures 45 deg rees, lumbar lordosis (T12-S1)

measures 7 degrees, the pelvic incidence measures

51 degrees, the sacral slope m easures 5 degrees,

and t he sagit tal vertical axis measures 12 cm.

(g,h) The pat ient underwent an o set L2 pedicle

subtraction osteotomy through a poste rior

approach and proximal extension to T4, as de mon-strated in the standing postoperative long-casett e

posteroanterior (g ) and lateral (h) ragiographs.

The L2 anterior screw was removed from the sam e

posterior approach .

a b c d e f g h

Fig. 3.10 a–e (a) Int raoperative view of the lateral

and ant erior dissection performed to ident ify the

previously placed anterior instrum entat ion t hrough

a posterior exposure. Note the preservation of the

exiting nerve root. A me ta l-cut ting high-speed drill

is used t o cut the ante rior rod proximal and d istal to

the screw. After rem oval of some o f the lateral

vertebral body, a further cut is made along the neck

of the screw. (b) The screw head with t he at tached

rod is removed. (c) The shaft of the screw is

extracted when completing the oste otomy. (d,e) A

schemat ic demonstrates the removal of the ante rior

implant. (From Lewis SJ, David K, Singer S, et al.

A technique of ante rior screw removal through a

posterior costo transversectomy approach for

posterior-based osteo tomies. Spine 2010;35:

E471–E474. Reprinted with perm ission.)

a

b

c d e

signi cant challenges to the exposure, the dis-

section, and the decomp ression, especially on

the concave side (Fig. 3.12a,b). Care must be

taken when dissecting around the vertebral

bod y on the concavit y, t o ensu re that the dis-section does not ente r the m ediastinum . Simi-

larly, with the severe rotation, the spinal cord

will be shifted against the concavity (Fig.

3.12c,d), making it vulnerable to injury with

removal of the concave pedicle. These chal-

lenges are not as di cult w hen per form ing a

PSO for sagitt al plane deform ities.

The step s for a VCR are similar t o t hose fora PSO: exposure, followed by insertion of im-

plan ts, removal of the transverse processe s, re-

m oval of the m edial ribs and rib head s, exposu re




Fig. 3.11 a–g (a) Anteroposterior and (b) lateral

long cassette radiographs of a 19-year-old man with

congenital kyphoscoliosis. (c,d) Three -dimensional

reconstructions demonst rates a T11 hem ivertebra

at t he ape x of the de formity. The pat ient under-

went (e ) posterior resect ion of T11 and T12 and(f,g) post erior reconstruct ion from T5 to L4. A

port ion of the re sected vertebra was use d as an

ante rior strut be tween T10 and L1 to maintain the

integrity of the anterior column. Closure of the

osteotomy was performed with proximal to distal

convex rod placeme nt with the t emporary concave

rod, with loosened set screws in place t o preventtranslation.

a b c d e f g

Fig. 3.12a–d Axial (a) CT and (b) MRI of a non-

rotate d t horacic spine with kyphosis. Note the

position of the rib heads and the cent ral position of

the spinal cord. (c) Comparat ive CT of a pat ient with

a severe scoliosis. Note the marked rotation of the

vertebra, the convex late ral vertebral body abut ting

the posterior aspect of the convex rib, the very

ventral position of the concave rib, and the posterior

position of the convex rib head. (d) Axial MRI shows

the spinal cord shifted aga inst the concave rib.

a b

c d



40 Chapte r 3

of the ventral vertebral body, posterior de-

compression, removal of the concave pedicle,

tem porar y concave rod, convex rem oval of the

pedicle , and re lease of the proxim al a nd dist al

disks. The vert ebral body can be rem oved p iece-

m eal or en b loc. For piecem eal rem oval, a shellof vent ral vertebr al body cortex can be left be-

hind to protect the med iastinal structures and

serve as a barrier for the anterior strut graft

or cage. For en-bloc resection, circumferential

release of the disks needs to be performed to

perm it adequ at e release and re m oval. Rele ase

of the concave side of the disk is the m ost di -

cult. Often th e adjacent m edial ribs on th e con-

cavity need to be rem oved to provide su cient

access for the release. The tap for the pedicle

screws can be u sed as a joystick from the con-

vexity t o facilitate th e com plete rem oval of the

vertebral body. The use of proper retractors to

protect the m edias t inal st ruct ures during an-

terior body resection is paramoun t.

When performing VCR with marked rota-

tion, it is easiest to enter t he canal through t he

convex foramen to start the decompression.

The complete posterior elem ent s of the level

to be resected should be removed, along with

the laminae of the adjacent levels. This will provide good visualizat ion of the sp in al cord

up on oste otom y closure. The con cave pedicle is

carefully rem oved an d a tem porar y rod is then

placed on the con cavit y. The m ajorit y of the

remaining dissection and vertebrectomy can

be perform ed fro m the convex it y w ithou t the

temp orary rod being in t he w ay.

A structural anterior support graft, either a

par t of the resect ed verteb ra or a cage, is in -

serted anteriorly to guide the reduction and

preven t over sh or ten ing. A lam inar sp read er can

be use d to dist ra ct ventrally from the concave

side to facilitate the graft/cage insertion. Clo-

sure of the VCR should be done with a convex

rod. The t em pora ry concave rod is left in place,

with the set screws loosened, preventing trans-

lation without hindering osteotomy closure.

Redu ction is often easiest from proximal to dis-

tal. Single- and dual-rod reduction techniques

have been described. Redu cing a p roxim al and

distal convex rod to a central connector hasalso been described. Being familiar with mul-

tiple techniques and the equipment available

will enable the surgeon to tailor the m ethod to

the given situation.

Type 6: Multilevel Vertebral

Column ResectionSevere kyph otic angular d eform ities, often sec-

ondary to remote infections, are amenable to

pos te rior-based vertebr al re se ct ions. A sin gle

level is often insu cient . Multiple levels of th e

remnants of the deformed vertebrae are re-

sected (Fig. 3 .13 ). Follow ing resect ion, ven tr al

distraction aids in length ening the an terior col-

umn for placement of an anterior cage/strut.

Posterior shorten ing through t he rod w ill com -

plete the corre ct ion.

! Osteotomies for Fixed

Lordosis

The correction of xed hyperlordosis requires

a combination of anterior shortening and pos-

terior lengthening. This is most reliably ac-

complished t hrough a form al anterior release

followed by a p osterior correction (Fig. 3.14).Similar to severe kyphosis, w here the vertebral

column is displaced poste riorly, in xed hype r-

lordosis the spine is displaced ventrally. This

ventral displacement favors an anterior ap-

pro ach , w ith the sp ine be ing su per cial t o the

anterior abdominal wall. In cases of thoracic

hyperlordosis, severe narr owing of the m edias-

tinum occurs, w ith bronchial comp ression oc-

curring in th e m ore severe cases. Even in cases

w ith respiratory issues, these p atient s paradox-

ically bene t from formal anterior approaches

to decrease the lordosis and help increase the

kyphosis, thereby increasing the anteroposte-

rior diam eter of the m ediastinum , relieving the

br on ch ial com pre ssion.

Anterior shortening can be accomplished

with multiple-level diskectomies for global

hyperlordosis or through resection of disk and

bo ne for m ore foca l deform it ies (Figs . 3.15 an d

3.16). A posterior-column release and instru-

me ntation is then perform ed. Contouring the pos te rior ro d in the approp riate sagit tal plane

w ill th en reduce th e lordotic deform ity. Form al




Fig. 3.13 a–e (a) Anteroposterior and (b) lateral

views of a 59-year-old m an with severe kyphosis

secondary to remote infect ion. (c) CT sagitt al view

shows four vert ebrae autofused ventrally with a

severe focal kyphosis, and hyperlordosis of the d istal

lumbar and thoracic spines. (d,e) Long casset te

radiographs demonstrating correction following

multilevel vertebrectomy, placement of an anterior

cage, and posterior T6 to pelvis instrumentation.

a b c d e

Fig. 3.14 a–d Supine (a) anteroposterior and

(b) late ral long casset te radiographs of a 17-year-old

boy with severe neuromuscu lar lordoscoliosis with

previous Baclofen pum p inse rt ion. (c,d) Following

L1 to S1 anterior diskectom ies, intraoperative

traction and a posterior T2 to pe lvis instrumentation

and fusion was performed.

a b c d



42 Chapte r 3

Fig. 3.15a–f (a) Anteroposterior and (b) lateral

long cassette radiographs of a 47-year-old woman

with a remote Harrington rod instrumentation

and fusion for adolescent idiopat hic scoliosis. She

presented with distal degenerat ion and sag ittal

malalignment. (c,d) An L3 pedicle subt ract ion

osteotomy and ante rior lumbar interbody fusions at

L4–5 and L5-S1 were performe d, resulting in xed

lumbar hyperlordosis. Because of the patient ’s

severe unhappiness with her sagitt al alignme nt,

an ant erior L2–3 diskect omy and resection of the

proximal port ion of the L3 vertebral body followed

by a L2–3 post erior colum n release were perform ed.

The poste rior fusion m ass release was gent ly

distracte d and held open with mesh cages, while

an ap propriately contoured rod was inserted from

distal to proximal to reduce the o steotomy.

(e,f) This resulted in a more balanced sagit ta l plane.

(From Lewis SJ, Gray R, David K, Kopka M, Magana S.

Technique of Reverse Smith Petersen osteotomy

(RSPO) in a patient with xed lumbar hyperlordosis

and negative sagittal imbalance. Spine 2010;35:

E721–E725. Reprinted with perm ission.)

a b c d e f

Fig. 3.16 a–c Late ral radiographs of the pat ient in

Fig. 3.15 demonstrating (a) the L3 pedicle subtrac-

tion osteotomy and (b) the planned resection for

the reverse Smith–Petersen osteotomy. (c) Close-up

lateral view of the lumbar spine following closure of

the combined anterior/posterior osteotom y. (From

Lewis SJ, Gray R, David K, Kopka M, Magana S.

Technique of Reverse Smith Petersen osteotomy

(RSPO) in a patient with xed lumbar hyperlordosis

and negative sagittal imbalance. Spine

2010;35:E721–E725. Reprinte d with permission.)

a b c




pos te rior dist ra ct ion after the circu m ferential

release m ay cause unwan ted distraction of th e

entire spine instead of just the posterior column,

making the reduction to an under-contoured

rod the preferred m ethod.

! Chapter Summary

The approach to severe spinal deform ities has

signi cantly changed with the improved tech-

niques, imaging, and instrumentation that

are available. An improved outcome will be

achieved w ith careful preoperative planning, a

deep understanding of the deformities, and

th e knowledge and ability to perform t he vari-

ous correct ion techn iques. Creating an environ-

m ent with experienced and skilled surgical and

per ioperat ive team s w ill help to predict and

manage the complexities associated with the

successful treat ment of these challenging cases.

Pearls

Obtaining appropriate preoperative imaging

studies can help to bett er understand the complexities o f the de form ity and the patient’s anat-

omy, and to plan for potent ial di culties in the

pro cedure.

Careful preoperative clinical and radiographic

evaluat ion will help to assess the exibility of the

deformity, to determ ine the xation options, and

to decide on the location, number, and type

of osteotom ies required to achieve the desired

correction. Accurate intraoperative spinal cord monitoring,

including mot or evoked pot ent ials, is essen tial to

the safe comp letion of these procedures. Under-

standing the timing and magnitude of neuro-

mo nitoring changes will direct key intraope rative

decisions.

Although ped icle screw instrume ntat ion is the p ri-

mary method of curve control, alternative xation

methods such as fusion mass screws or laminar

hooks are importan t b ackup strategies, especially

in revision surgery or dysplastic anato my.

Pitfalls

The artery of Adamkiewicz has a variable anat-

omy from T8 to L1, mo st com mon ly on the left

side. When considering ost eoto mies around t he

thoracolumbar junction, protecting and saving

the nerve roots may preserve key sources of blood

ow.

Careful and controlled reduct ion of three -column

osteotomies is essential to prevent cord trans-

lation and subsequent injury. Complete visuali-

zation o f the cord and harmonious collaboration

with the surgical team , electrophysiological moni-toring team , and nursing sta is essen tial for spi-

nal cord safet y.

References


1 . Dorw ard IG, Len ke LG. Osteotom ies in the pos ter ior-

only treatment of complex adult spinal deformity:

a com parat ive review. Neurosurg Focus 201 0;28:E4

PubMed

2. Dom m isse GF. The blood su pp ly of th e spina l cord. A

critical vascular zone in spinal surgery. J Bone Joint

Surg Br 1974;5 6:225–2 35 PubMed

3. Boll DT, Bulow H, Black ha m KA, Ascho AJ, Schm itz

BL. MDCT angiogra phy o f the spina l vasculatu re a nd

the artery of Adamkiewicz. AJR Am J Roentgenol

2006;187:1054–1060 PubMed

4. Lewis SJ, Arun R, Bodrogi A, et al. The use of fusion

mass screws in revision spinal deformity surgery.

Eur Spine J 2014;23(Suppl 2):18 1–186 PubMed

5. Lew is SJ, Golds te in S, Bod rogi A, et a l. Com pa rison o f

pedicle su bt ra ct ion an d Smith-Pe ter se n oste otom ies

in correcting thoracic kyphosis when closed with a

central hook-rod construct. Spine 2014;39:1217–

1224 PubMed

6. Schw ab F, Patel A, Ungar B, Farcy J-P, Lafage V. Adu lt

spinal deformity-postoperative standing imbalance:

how m uch can you tolerate? An overview of key pa-

rameters in assessing alignment and planning cor-

rective surgery. Spine 2010; 35:2224– 2231 PubMed

7. Rose PS, Bridwell KH, Lenke LG, et al. Role of pelvic

incidence, thoracic kyphosis, and patient factors on

sagittal plane correction following pedicle subtrac-

tion osteotom y. Spine 2009;34 :785–79 1 PubMed

8 . Lewis SJ, Gray R, David K, Kopka M, Magan a S. Tech -

nique of Reverse Smith Petersen osteotom y (RSPO) in a

pat ien t w ith xed lum bar hyp er lordosis an d negative

sagittal imbalance. Spine 2010;35:E721–E725 PubMed

9. Cho K-J, Bridwell KH, Lenke LG, Berra A, Baldus C.

Comparison of Smith-Petersen versus pedicle sub-



44 Chapte r 3

traction osteotomy for the correction of xed sagit-

tal imb alance. Spine 2005 ;30:2030 –2037, discussion

2038 PubMed

10. Dorward IG, Lenke LG, Stoker GE, Cho W, Koester LA,

Sides BA. Radiogr aph ic and clinical out com es of pos-

terior column osteotomies in spinal deformity cor-

rection. Spine 2014;39:870–880 PubMed

11. Bridw ell KH. Decision m aking regard ing Sm ith-

Petersen vs. pedicle su btraction osteotomy vs. ver-

tebral column resection for spinal deformity. Spine

2006;31(19 , Supp l):S171–S178 PubMed

12. Jar vis JG, Str ant zas S, Lipkus M, et al. Respo nd ing to

neurom onitoring changes in 3-colum n posterior spi-

nal osteotom ies for rigid pediatric spinal deform ities.

Spine 2013;38 :E493–E503 PubMed

13. Lewis SJ, Kulkarni AG, Rampersaud YR, et al. Poste-

rior colum n re construction with autologous rib graft

after en bloc tumor excision. Spine 2012;37:346–

35 0 PubMed

14. Schwa b F, Blonde l B, Chay E, et al. The com preh en sive

anatomical spinal osteotomy classi cation. Neuro-

surgery 2014;74:112–120, discussion 120 PubMed

15. Lafage V, Schwab F, Vira S, et al. Does vertebral level

of pedicle subtraction osteotomy correlate with de-

gree of spinopelvic parameter correction? J Neuro-

surg Spine 2011;14:184–191 PubMed

16. Lafage V, Bha ru cha NJ, Schw ab F, et al. Mult icen te r

validation of a formula predicting postoperative

spinopelvic alignment. J Neurosurg Spine 2012;16:

15–21 PubMed

17 . Halper n EM, Bacon SA, Kitagaw a T, Lew is SJ. Poster ior

transdiscal three-colum n shortening in th e surgical

treatment of vertebral discitis/osteomyelitis with

collapse. Spine 201 0;35:13 16–1322 PubMed

18 . Lew is SJ, David K, Singer S, et al. A techn ique of ant e-

rior screw rem oval through a posterior costotransver-

sectomy approach for posterior-based osteotomies.

Spine 2010;35: E471–E474 PubMed

19. Ham zaoglu A, Alanay A, Ozturk C, Sarie r M, Karad er-

eler S, Ganiyusufoglu K. Posterior vertebral column

resection in severe spinal deform ities: a total of 102

cases. Spine 2011;36 :E340–E344 PubMed

20 . Len ke LG, O’Lea ry PT, Bridw ell KH, Side s BA, Koest er

LA, Blanke KM. Poster ior vert ebra l column rese ction

for severe pediatric deformity: minimum two-year

follow-up of thirt y- ve consecut ive patien ts. Spine

2009;34:2213–2221 PubMed



! Introduction

Fusion attempts across L5/S1 in adult spinal

deformity are plagued by a high rate of pseu-

dart hrosis and implant breakage/failure du e to

the unique anatomic and biomechanical char-

acteristics of the lumbosacral junction.1,2 In

a single-institution review of adult deformity

pat ients w ith const ruct s great er than fou r lev-

els, construct s ending at S1 had a signi cantlyhigher rate of pseudarthrosis compared with

constru cts end ing at L5 or m ore cepha lad lev-

els.1 Othe r pseu dart hrosis risk factors were age

older than 55 years, more than 12 levels in-

cluded in the construct, and T10-L2 kyphosis

of greater than 20 degrees. The addition of

sacral-pelvic xation increases the strength

and stability of constructs spanning the lum-

bosacral junct ion and is a form idable too l in

the spinal deform ity surgeon’s arm am entar ium

for correcting spinal imb alance.

! Anatomic and

Biomechanical

Considerations

As a transition zone from the mobile lumbar

spine to the sti pelvis, the lumb osacral junc-tion exper iences signi cant forces challenging

arthrodesis attempts across the segment. De-

spite bearing axial loads of more than double

bod y weigh t , the osseou s anatom y of the sa-

crum p rovides relatively litt le strength for x-

ation.3 The sacrum consists of a thin rim of

cortical bone surrounding a cancellous core,

w ith large pedicle diamete r precluding the e n-

gagement of both medial and lateral cortical

walls via ped icle screw instr um ent ation.

The lumbosacral junction is a biomechani-

cally distinct location that is subjected to thehighest level of translational shear force and

the most limited range of motion within the

spine, with the L5/S1 disk bearing the largest

summation of load vectors.2,4–6 These u nique

stresses, combined with the relatively small

am oun t of sacral cort ical bon e available for x-

ation, result in increased pseudarthrosis and

implant breakage/failure in long instrumenta-

tion constru cts end ing at S1.2,7

McCord et al6 introduced the concept of the

lumbosacral pivot point at the junction of the

L5-S1 disk and the middle osteoligamentous

colum n to describe the considerable exion

moments and cantilever forces acting at the

lumbosacral junction. Extending xation an-

terior to this pivot point increases construct

strength. Screw insertion into the ilium pro-

vides th e longest xation length ante rior to

this pivot point, and w as found to be t he on ly

instrum entation type at t he lum bosacral junc-

tion that signi cantly increased the maximumexion moment at failure. Compared with the

weak cancellous composition of the sacrum , the

4

Indications and Techniquesfor Sacral-Pelvic Fixationin Adult Spinal Deformity

Kriste n E. Jones, Robe rt A. Morgan, and David W. Polly, Jr.



46 Chapte r 4

pos ter ior iliu m o ers abu ndant cort ica l bon e

for anchoring instrumentation and enables

increased screw length and diameter, making

sacral-p elvic xation a useful tech nique for in-

creasing constr uct strength .

! Indications for Sacral-Pelvic

Fixation

The rigid xation provided by sacral-pelvic

instrum ent ation is a useful adjunct to treating

a wide array of pathological entities. Sacral-

pelvic xa t ion is in dicat ed for lu m bos acral

arth rodesis extending ceph alad to th e L2 ver-

tebra, augmentation for poor quality or os-

teoporotic bone, sacrectomy for tumor or

infection, un stab le sacral fractur es, correct ion

of at-back syndrom e via lumb ar osteotomy,

correction of pelvic obliquity, and high-grade

spondylolisthesis.8 The addit ion of iliac xa-

tion in these conditions signi cantly reduces

the stress placed on S1 instrumentation and

increases construct strength.

As w ith all spinal surgery, selection of the ap-

propriate ap proach for each individual pat ientand m eticulous atten tion to surgical technique

is the key to successful treatment. Although

sacral-pelvic xation is not required for many

pat ients undergoing lum bosacral ar thro des is,

th e force vectors required for th e creation an d

m aintena nce of proper sagittal alignm ent rela-

tive to the patient’s individual bone quality

m ust be considered.

! Sacral-Pelvic

Instrumentation Selection

and Techniques

Sacral Fixation

Screws at S1 can be placed th rough th e pedicles

in a medially convergent m ann er w ith bicort ical

end-p late or tricortical purchase, or through th e

ala in a divergent m anner (Fig. 4 .1a). Sublami-nar h ooks and w ires and S2 pedicle screws can

be used to su pplem ent S1 pedicle /a lar screw s

bu t shou ld n ot be relied on for anchor ing a long

construct.

S1 Pedicle Screw s

The sizable medially convergent sacral pedicles

accomm odate large screw length an d diameter

while simultaneously preventing “ lling” the

pedicle to ach ieve bicortica l purchase of the

m edial and lateral pedicle w all with a single

screw. The largely cancellous sacral pedicles

provid e relat ively lit t le pullo ut st rengt h in

un icortical xation, an d un icortical ped icle

screws should be avoided.3 Bicortical xation

anchored into the anterior sacral cortex pro-

vides increased p ullout strength compared w ith

unicortical S1 pedicle screws; however, addi-

tional trajectories can be employed to further

enh ance pu llout str ength . Luk et al9 compared

bicor tical S1 ped icle screw insert ion torque and

pullout st re ngt h to that of S1 pedicle screws

advanced through the S1 superior end plate.

S1 pedicle screws traversing the end plate

had signi cantly higher inser tional torque and

pullout st re ngt h compar ed w ith bicortica l S1

screws.Tricort ical xation, de ned as a screw tra-

jector y tow ard the m edial sacral pro m on tor y,

captures purchase in the dorsal, anterior, and

superior end-plate cortex (Fig. 4.1b). Lehman

et al10 found that this tricortical trajectory

doubles the insertional torque compared w ith

bicort ica l S1 pedicle screws par allel to the S1

end plate. Tricortical S1 screw strengt h has n ot

been direct ly compared w ith t rans– end-plate

screw strength; both provide enh anced strength

compared with typical bicortical purchase

paralle l to the S1 end plate. Triangu lat ion of

the pedicle screw t rajectory increases pullout

strength compared with straight-ahead trajec-

tory and should be u niversally emp loyed.

Sacral Alar Screw s

Alar screw insertion ut ilizes a lateral trajectory

into low-d ens ity cancellous sacral bone. Bicor-

tical alar xation is techn ically possible butfraught w ith risk of injuring the L5 n erve root s



Indications and Techniques for Sacral-Pelvic Fixat ion in Adult Spinal Deformity 47

Fig. 4.1a,b Sacral screw trajectories. (a) Trajecto-

ries for S1 pedicle (A) and alar (B) screw placement.

(b) Intraoperative view de monst rating probe

insertion in tricortical S1 screw trajectory. Tricortical

purchase ut ilizing dense sacral promontory cort ical

bone should be employed to m aximize screw

pullout st rength.

a

b



48 Chapte r 4

or comm on iliac vessels draped ante riorly across

the ala.3 Thus, alar screws are ut ilized m ainly as

unicortical supp lement s to bicortical/tricort ical

S1 p edicle screw const ru cts. A Chop in or Colo-

rado p late or a Tacoma block can be u tilized to

connect S1 and alar screws. Disadvantages ofthis technique include constrained screw st art -

ing point and p otent ial imp airm ent of the ideal

trajectory.

Sacral Sublaminar Wires and Hooks

Although sublaminar wires and hooks lack

su cient biom echanical strength to serve as

anchors to long constructs, they can be used

as supp lements for short-segment fusions.3,11

Hooks are optim ally placed in th e do rsal sacral

neuroforamina where improved cortical pur-

chase can b e achieved. The Harrington instru-

mentation system initially employed sacral

hooks as anchors to long constructs, but the

rate of pseudarthrosis and hook dislodgment

at L5/S1 was u naccept ably high. Sacral sub lam -

inar wires and hooks should not be used as

anchors to long constr ucts.

S2 Screws

S2 pedicle screw strength is typically limited

by a sh or t pedicle lengt h and a locat ion dor sa l

to the lumbosacral pivot point described by

McCord et a l6 an d Kebaish. 11 S2 pedicle screw s

and S2 screws directed laterally into th e ala can

be used as adjunct su ppor t to sh or t-segm ent

fusions but lack the biom echanical stren gth to

anchor long construct s.12

Jackson Intrasacral Rod Technique

Jackson intra sacral rods are inserted vertically

throu gh th e ala from S1 to th e level of S2 an d

the n can be connecte d to a constr uct including

S1 p edicle screws. Insert ion is techn ically dif-

cult and can be preclude d by alar anatomic

variations. This technique h as been shown to

be biom ech anically in fer ior to a lu m bosa cral

pedicle screw s– iliac screw const ruct and ism ent ioned for historical context only.5

Transsacral Fixation

Kellogg Spee d rst describe d tra nsver tebra l

strut grafting at L5/S1 from an anterior ap-

pro ach for pat ients w ith high -grade spon dylo-

listh esis. The Speed techn ique involves dr iving

a bular stru t graft through the L5 vertebral

bo dy and in to the sacrum via an te rior expo-

sure, and is a u seful technique in lieu of inter-

bo dy cage placem ent , w hich has an increas ed

risk of anterior subsidence for patients with

high-grad e spondylolisthesis.13

Due to the risk of the anterior exposure to

the lumbosacral junction, including injury to

the great vessels during mobilization or sym-

pat het ic plexu s dysfu nct ion cau sin g re trograd e

ejaculation in m ales, H.H. Bohlman popu larizedthe posterior approach for tran svertebra l b-

ular strut grafting at L5-S1 for patients with

high-gra de spon dylolisthesis. Ante rior xation

through L5/S1 can also be performed via a

paracoccyge al approach in a m in im ally in -

vasive fashion u tilizing synth etic imp lants . As

with all constructs, the addition of posterior

colum n sup port increases stability.

Iliac FixationAnchor ing a constr uct with iliac xation cre-

ates a longer lever arm to resist cantilever

forces across the lumb osacral junct ion via ex-

tension ant erior to the lum bosacral pivot point,

increasing biomechanical strength of sacral-

pelv ic const ruct s.6 Incorporation of the ilium

into a construct o oads stress from sacral

screws an d decreases the r ate of sacral instru-

m entation failure and pseudart hrosis across the

lumb osacral junction.4,14

Transiliac Fixation

Harrington Threaded Sacral Rod

Developed for pelvic xation in adjun ct w ith

Harr ington distraction rod s, this device is men-

tione d for historical pur poses. Two sep arate

pos te rior iliac incis ions are used to inse rt the

threa ded rod th rough the p osterior iliac wings

with compression applied. Pseudarthrosis rateshave been rep orted above 40%m and t he d is-




traction forces of Harrington rods can result in

sagitta l plane imba lance.2,8

Kostuik Transiliac Bar

Inser ted from the m idline, the Kostuik tran sil-iac bar is placed 1 to 2 cm anter ior to the pos-

terior superior iliac spine and then attached

via custom connectors to S1 pedicle or alar

screws. The bar is sm ooth and has a contoure d

shape that accommodates the midline sacral

dorsal prominence, and has been reported to

have a h igh fusion rate of up to 97%.8

Iliac Fixation

Luque L-f xation

The rst to develop segment al instru m entation,

Luque exten ded lum bosacral construct s to the

pelvis usin g L-shaped ro ds w hose ends were

inserted into the posterior ilium at the poste-

rior sup erior iliac spine. This circum vented the

distraction problem associated with the Har-

rington technique and improved fusion rates,

but pist on ing occu rring be t ween the rod s de-

creased t he stability of th e Luqu e L- xation intorsion and exion.2,8

Galveston Technique

Ben L. Allen and Ron L. Ferguso n d evelope d t he

Galveston te chnique in t he 1980s. Sm ooth rod s

inserted from th e poster ior superior iliac spine

into the ilium were connected to segmental

lumbosacral instrumentation. Rod contouring

required signi cant expertise. To obviate the

need for this, rods with pre-bent sagittal con-

tour and bilateral iliac xation were created in

a single piece.3 This constru ct still required sub-

stantial rod management skills. Because the

smooth rods had inferior pullout strength com-

pared w ith thread ed screws, iliac screw s quickly

be cam e a m ore p op ular m et hod of xation .

Iliac Screws

Iliac xation with single or stacked unilateralor bilateral screws is performed with fully or

par t ially threa ded screw s. The pullout st rengt h

and rotational stability is superior to non-

threa ded rod techniques. The star ting point for

screw insertion is at the level of the posterior

super ior iliac spine w ith a t rajectory targeting

either the supra-acetabular notch or th e ante-rior superior iliac spine (Fig. 4.2). The screw

trajectory is planned via visualization of the

“iliac teard rop” on eith er uoroscopy or com-

pu ted tom ography (CT) im age gu idance. The

iliac teardrop is a region above th e acetabulum

bo rd ering t he m edial iliac w all, the lat era l iliac

wall, and the zenith of the sciatic notch. The

screw achieves greatest pu rchase in th e lateral

margin of the teardrop, directed through the

cortical bone just above the sciatic notch. San-

tos et al15 analyzed various screw lengths and

diamete rs and found a signi cant increase in

insertional torque for iliac screws of length

" 80 mm and diameter " 9.5 m m . No di eren ce

in insertional torque existed bet ween the supra-

acetabular notch and the anteroinferior iliac

spine- directe d tr ajectory. Given t he r isk of ace-

tabular joint violation w ith th e supra-acetabular

trajectory, the authors concluded that optimal

iliac screws are inserted in the anteroinferior

iliac spine trajectory w ith length " 80 mm anddiameter " 9 .5 mm .

Iliac screws can be used in combination w ith

sacral screws to provide increased construct

strength and are biomechanically superior to

oth er pelvic xation opt ions. Com par ing th e

Galveston techn ique to iliac screws in a series

of 20 neuromuscular scoliosis patients, iliac

screws enabled better correction of pelvic

obliquity and d ecreased implant breakage.16

In a biom echanical compar ison bet ween the

m odi ed Galveston technique with iliac xa-

tion but no S1 xation versus S1 pedicle screws

plus iliac screw s, or S1 and S2 screws w ithou t

iliac xat ion, Tis et al17 found that constructs

w ith iliac screws conferred signi cant strength

via decreased range of motion in multidirec-

tional exibility testing and increased load to

failure.

Bridwell’s group 4 performed a laboratory

investigation compar ing m ultidirectional exi-

bi lit y and exu ra l load to failure am on g thefollowing constru ct t ypes: lum bosacral pedicle



50 Chapte r 4

screws, interbody cage, and iliac screws. Iliac

screw or interbody cage placem ent signi cantly

redu ced m ultidirectional exibility at the lum -

bosacral junct ion com pared w ith pedicle screws

alone, but iliac screw xation was superior in

protect ing sacral screws from pullout or p low -

through.4

Lebwohl et al5 performed a laboratory bio-m echanical analysis compar ing construct sti -

ness, S1 screw strain, and u ltimate failure load

among several techniques of supplementary

sacral xation, using S1 ped icle screws with

and w ithout S2 screws, as well as an int rasacral

rod and iliac screws. All techniques decreased

the S1 screw strain in exion-extension, but

only iliac screws decreased S1 screw strain in

axial loading. In destru ctive testing w ith exionloading, only iliac screw s signi can tly increased

Fig. 4.2 a–d Iliac screw placement u sing intraopera-

tive computed tomography (CT)-based framelessnavigation. (a) Iliac screw insert ion point is at the

posterior superior iliac sp ine, d irected toward the

anterior inferior iliac spine. (b) The teardrop view

is utilized t o opt imize screw purchase in the lateral

cort ical wall, just above t he sciatic notch . (c) The

sagittal view is used to dem onstrate screw angula-

tion toward the anterior inferior iliac spine, avoiding

violation of the acetabulum. (d) Stacked iliac screw

construct in patient with progressive deformity from

high-grade spondylolisthe sis p reviously fused in

situ, requiring sacral oste otomy for deformity

correction.

a

b

c

d




the load to failure. The auth ors concluded that

th e add ition of iliac xation signi cant ly in-

creases the biomechanical strength of sacral

constructs.

A signi cant d isadvant age to iliac screws is

the potential prominence of the screw headat th e p oster ior sup er ior iliac spine (PSIS). Al-

though the starting point can be m odi ed or

the PSIS notched to allow bur ying of the screw

head, implant prominence cannot be com-

pletely e lim inat ed. Anot her d isadvantage is the

requirement of an o set connector for seg-

me ntal lumbosacral instrum entation.

Transsacral Iliac Fixation

S2-Alar-Iliac Screws

Developed to addre ss the problem atic imp lant

pro m inence and o se t locat ion of iliac screw s,

S2-alar -iliac (“S2-iliac”) screws are associated

with a lower complication rate related to im-

p lan t prom in ence-relat ed pain , necess it at ing

screw rem oval, compare d w ith t raditional iliac

xation.11 Implant prominence is minimized,

with th e m ean distance of the insertion point

to the skin 15 m m deep er for the S2-iliac tech-nique compared with traditional iliac screw

inser tion at th e PSIS.18 The S2-iliac screw st ar t-

ing point is 2 to 4 m m lateral and 4 to 8 m m

caudal to the dorsal S1 foramen, with a tra-

je ct or y tow ard the anterior infer ior iliac sp ine

(Fig. 4.3a) at an angle of ~ 40 degrees lateral

and 20 to 30 d egrees caudal.11,18 The iliac tear-

drop is again employed for trajectory align-

ment. Unlike traditional iliac screws, S2-iliac

screws do not typically require an o set con-

nector for joining rods to lum bosacral pedicle

screw constructs (Fig. 4.3b). Biomechanical

test ing of S2-iliac screws has sh own e quivalent

stability to conventional iliac screws.19 The

S2-iliac screw trajectory results in crossing the

sacroiliac joint, which has not been shown to

be prob lem at ic in the rst 5 year s o f follow-up

of this technique, but which requires ongoing

surveillance (Fig. 4.3c,d). Although haloing

around iliac screws m ay be observed in over

25%of patien ts, iliac screw pu llout o r breakageis very r are.1,11,14

Adjunctive Anterior

Interbody Support

Polly and colleagues 20 found that load-b earing

interbody structural grafts increase construct

sti ness and therefore can decrease the strainon posterior instrumentation, in addition to

increasing th e su rface a rea available for arth ro-

desis. They also found that the location of the

interbody graft in the sagittal plane has bio-

m echanical signi cance, with ant eriorly placed

grafts having increased sti ness compare d with

central or posteriorly placed inter body grafts.

Comp ared w ith stand- alone pedicle screw and

combination pedicle-iliac screw constructs,

pedicle -iliac screw const ruct s com bined w ith

interbody cages signi cantly redu ce segmen tal

movement across the lumbosacral junction in

laboratory analysis.4

! Patient Positioning

Patients undergoing sacral-pelvic instrumenta-

tion via open or minimally invasive technique

should be positioned prone on an operating

table that enables the creation or m aintenance

of anatomic lumbar lordosis. Fixation of the

lumbosacral junction in a at or kyphot ic an-

gulation m ust be absolutely avoided due to the

resultant sagitt al imbalance. Allowing th e abd o-

men to hang freely w ithout ventral compression

helps minimize intra-abdominal pressure and

venous bleeding.

! Operative Technique s

Sacral Fixation

Pedicle screw pullout stre ngth is increased by

m edialization of insert ion trajectory compared

with “straight-ahead” insertion without me-

dial angulation of the screw tip.3,12 An obese

bo dy hab itus, an iliac crest ove rh an g, o r t rian -

gulated vertebral bodies may presen t obstacles

to adequ ate m edialization of pedicle screw t rajectory (Fig. 4.4a). If this problem is encountered



52 Chapte r 4

Fig. 4.3 a–d S2-alar-iliac screws. (a) Int raoperative

CT-guided planning dem onstrating st arting point

and trajectory of S2-alar-iliac screw. (b) Postoperative

CT scan showing alignment of S2-alar-iliac screw

head with lumbosacral pedicle screws, eliminat ing

the need for an o set connector typically required

by trad itional iliac scre ws. (c,d) X-rays dem onst rat-

ing usage of S2-alar-iliac screws to anchor long

construct for deformity correction.

a b

c d




while using a midline open skin incision, the

bailout technique re qu ires sp lit t ing the fascia

in a paramedian fashion for a transmuscular

approach to enable a more lateral starting

point and m edializat ion of the t ra jector y. Vigi-

lance is required to prevent an unintended

straight-ahead trajectory from violating the

ante rior sacral or lum bar cortex with resultant

injury to neurovascular structures. The aortic

bifu rcat ion occurs at ap proxim ately L4/ L5, w ith

the common iliac vessels traveling laterally

from t he b ifurcat ion. The L4 and L5 ne rve root s

traverse the anterolateral sacral cortex pr ior to

join ing the lum bosacral p lexu s located at the

level of the sacral ala, and t he colon is in close

opposition to the ventral surface of S2. A rela-

tive “safe zone” exists in the ventral midline

of the sacral promon tory; h owever, individual pat ient vascular anat om y sh ould be in cid en-

tally visualized an d ap preciated on preope ra-

tive spine imaging prior to proceeding with

instrumentation.

Iliac Fixation

Misdirection o f iliac or S2-iliac screw s th rough

the sciatic notch can cau se injur y to the su pe-

rior gluteal artery or sciatic nerve, which is a

rare but serious complication. Violation of the

acetabulum m ust be avoided. Fam iliarization

with the iliac teardrop view enables correct

screw t rajectory selection.

Placeme nt of screws into th e ilium requires

signi cant insertional torque that can cause

breakage o f th e screw driver unless care is t ake n

to sequentially tap the screw trajectory com-

plete ly to the des ired depth. Once iliac screw

insertion is initiated, one should not pauseduring insertion, as the mechanical thermal

energy generated by screw insertion in the

Fig. 4.4 a–c Complicat ions of sacral-pelvic xat ion.

(a) Failure t o adequately direct pedicle screws in

me dial trajectory, combined with signi cant screw

pe rforation of ant erior cortex, result ing in screws

abutting internal iliac veins b ilate rally. (b,c) Asymp-

tom atic halo formation (arrows; dotted line) around

iliac screws doe s not necessita te revision unless

resulting in pain or lumbosacral pseudarthrosis.

a

b

c



54 Chapte r 4

bone assis ts in temporar ily lessening inser t ion al

torque, a bene t that is lost whe n screw inser-

tion pauses. Iliac screw head prominence and

o set distance from lum bosacral pedicle screw

heads is minimized utilizing the S2-alar-iliac

trajectory.

Revision of Screws

As mentioned, iliac screws may undergo as-

ymptomatic haloing (Fig. 4.4b,c). Unless pain,

implant breakage, or instability of the constr uct

occurs, screw revision is not required. If iliac

screw rep lacem ent is required, using larger di-

am eter screws confers more stability than using

longer screws.

! Chapter Summary

Sacral-p elvic xation is a power ful tool for in-

creasing the st rength and sta bility of lumb osa-

cral construct s. Extrem e biomechan ical forces

at th e lum bosacral junct ion and relatively poor

bon e qu alit y of the sacrum re su lt in a h igh ra te

of lumbosacral pseudarthrosis and implantfailure in adult spinal deformity correction.

Sacral-p elvic xation is indicated for lum bo-

sacral arthrod esis extend ing ceph alad to th e L2

vertebra, augm ent ation of construct s for poor-

quality or osteoporotic bone, sacrectomy for

tumor or infection, unstable sacral fractures,

correction of at-back syndrom e via lumb ar

osteoto my, correct ion of pelvic obliquit y, high-

grad e spon dylolisth esis, or as a salvage me cha-

nism d uring revision for pseu dart hrosis.

Sacral xation is optim ally per form ed using

bi- or t ricort ica l S1 m edially direct ed pedicle

screws. Alar and S2 screws an d h ooks through

the dorsal sacral foramina may be added for

supplementation, but lack the biomechanical

strengt h to anchor long constr ucts in adult spi-

nal deform ity. Care m ust be t aken to opt imize

screw size for cort ical pu rchase w hile avoiding

injury to neurovascular structures anterior to

the lumbosacral junction.

Iliac xation is opt ima lly per form ed using

iliac or S2-alar-iliac threaded screws, in a tra-

jector y toward the antero in fer ior iliac sp ine,

with length " 80 mm and diameter " 9.5 mm.Know ledge of sacral-pelvic anatom y and u se of

the iliac teardrop view on intra operat ive imag-

ing is key for iliac screw p lacemen t.

The addition of interbody structural grafts

increases the surface area for arthrodesis and

sti ness of the construct and should be per-

formed for long constructs at the lumbosacral

junct ion.

The spinal deformity surgeon must have

excellent knowledge of indications, instru-

mentation options, and techniques of sacral-

pelvic xat ion . Fam iliar ization w ith sacra l-p elvic

anatomy is necessary to optimize the size

and trajectory of instrumentation and avoid

complications.

Pearls

Sacral-pelvic xation increases the strengt h and

rigidity of constructs spanning the lumbosacral

junc tion. Iliac xation decrease s the rate of sacral instru-

mentation failure and reduces the incidence of

lumb osacral pseud arthrosis.

S2-alar-iliac screws provide stron g biome chanical

xation, m inima l implant prominen ce, and favor-

able implant alignment with lumbosacral pedicle

screws for ease o f rod con touring.

Pitfalls

Sacral pedicle and alar screws have inadequate

strength to anchor constructs extending cepha-lad t o L2, predisposing t he p atient t o lumbo -

sacral pseu dart hrosis unless sacral-pelvic xation

strategies are em ployed.

Aceta bular joint impingem ent or sciatic not ch vi-

olation with resultant neurovascular injury can

occur during p lacement of iliac screws un less in-

traoperative imaging with teardrop view is used.




References


1. Kim YJ, Bridw ell KH, Len ke LG, Cho KJ, Edw ards CC II,

Rinella AS. Pseuda rth rosis in ad ult sp inal deform ity

following multisegmental instrumentation and ar-

throdesis. J Bone Joint Surg Am 2006;88:721–728 PubMed

2. Kostuik JP. Treatment of scoliosis in the adult thora-

columbar spine with special reference to fusion to

the sacr um . Orthop Clin North Am 198 8;19:371 –381

PubMed

3. Santos ER, Rosn er MK, Per ra JH, Polly DW Jr. Spinopel-

vic xation in deform ity: a review. Neurosurg Clin N

Am 2007;18:373–384 PubMed

4. Cun nin gha m BW, Lew is SJ, Long J, Dm itr iev AE, Lin-

ville DA, Bridwe ll KH. Biom ech an ical evaluat ion of

lumbosacral reconstruction techniques for spondy-

lolisthesis: an in vitro porcine model. Spine 2002;27:2321–2327 PubMed

5. Lebwohl NH, Cunningham BW, Dmitriev A, et al.

Biome chanical compar ison of lumb osacral xation

techniques in a calf spine model. Spine 2002;27:

2312–2320 PubMed

6. McCord DH, Cun nin gha m BW, Sho no Y, Myers JJ,

McAfee PC. Biomechanical analysis of lumbosacral

xation. Spine 1992;17(8, Suppl):S235–S243 PubMed

7. Kim YJ, Brid well KH, Len ke LG, Rine lla AS, Edw ards C

II. Pseudarthrosis in primary fusions for adult idio-

pat h ic scolios is: incid en ce, ri sk fac to rs , an d ou tcom e

analysis. Spine 2 005;30:4 68–474 PubMed 8 . Mosh irfar A, Rand FF, Spon seller PD, et al. Pelvic xa-

tion in spine surger y. Historical overview, indications,

biom echan ica l re levan ce, an d cu rre nt te chniqu es . J

Bone Joint Surg Am 2005;87(Suppl 2):89–106 PubMed

9. Luk KD, Chen L, Lu WW. A stronger bicortical sacral

pedicle s crew xation t hro ugh t he S1 e ndplat e: an in

vitro cyclic loading and pull-out force evaluation.

Spine 2005;30:525–529 PubMed

10 . Leh m an RA Jr, Kuklo TR, Belmon t PJ Jr, Ande rsen RC,

Polly DW Jr. Advan tage of ped icle screw xation di-

rected into the apex of the sacral promontory over

bicor t ica l xa tion : a biom echan ica l an alysis . Spin e2002;27:806–811 PubMed

11. Kebaish KM. Sacropelvic xation: techniques and

complications. Spine 2010;3 5:2245– 2251 PubMed

12. Koller H, Zenn er J, Hem p ng A, Fer rar is L, Meier O.

Reinforcement of lum bosacral instru me ntation u singS1-pedicle screws combined with S2-alar screws.

Oper Orthop Traumatol 2013;25:294–314 PubMed

13. Cunningham BW, Polly DW Jr. The use of interbody

cage devices for spinal deformity: a biomechanical

per sp ec t ive. Clin Orthop Relat Res 20 02 ;3 94 :7 3– 83

PubMed

14. Tsuchiya K, Bridwell KH, Kuklo TR, Lenke LG, Baldus

C. Minimum 5-year analysis of L5-S1 fusion using

sacrop elvic xation (bilateral S1 an d iliac screw s) for

spinal deformity. Spine 2006; 31:303–3 08 PubMed

15. Santos ER, Sembrano JN, Mueller B, Polly DW. Opti-

m izing iliac screw xation: a biomecha nical stud y onscrew length, trajectory, and diameter. J Neurosurg

Spine 2011;14:219–225 PubMed

16 . Peelle MW, Len ke LG, Brid we ll KH, Sides B. Com pari-

son of pelvic xation techniques in neu rom uscular

spinal deformity correction: Galveston rod versus

iliac and lumbosacral screws. Spine 2006;31:2392–

2398, discussion 2399 PubMed

17 . Tis JE, Helgeson M, Lehm an RA, Dmit riev AE. A bio-

m echanical comparison of di erent types of lumbo-

pelvic xat ion. Spin e 2 00 9; 34 :E866 –E872 PubMed

18. Cha ng TL, Spons eller PD, Keba ish KM, Fishm an EK.

Low pro le pelvic xation: anatom ic param eters forsacral alar-iliac xation versus trad itional iliac xa-

tion. Spine 2009;3 4:436–4 40 PubMed

19. O’Brien JR, Yu W, Kaufman BE, et al. Biomechanical

evaluation of S2 alar-iliac screws: e ect of length

and quad-cortical purchase as comp ared w ith iliac

xation. Spine 2013;38: E1250–E1255 PubMed

20. Polly DW Jr, Klem m e W R, Cun ningh am BW, Bur net te

JB, Haggerty CJ, Oda I. The biomechanical signi -

cance of anterior colum n sup port in a simulated sin-

gle-level spin al fusion. J Spina l Disord 20 00;1 3:58 –62

PubMed



! Introduction

Osteoporosis is an imbalance between bone

form ation and resorption that primarily a ects

trabe cular bone . Progressive bone m ineral loss

and concomitan t bony architectu re changes re-

sult in pain, deformity, increased risk of frac-

tu re, and possible ne ural compression.

The spine is the m ost comm on site of osteo-

porot ic fr act ure s. Althou gh m ost pat ients w ithacute vertebr al comp ression fractures imp rove

regardless of the t reatm ent ap plied, no patient

experiences spontaneous restoration of the

vertebral he ight and achieves a realigned spine.

Therefore, spinal instru m ent ation is event ually

required for some pat ients.

With aging comes a higher incidence of

comorbidities that further complicates the

management of osteoporotic spine. The el-

derly today have more active lifestyles than

did the elderly of previous generations, and

they refuse to accept disability an d d eform ity

as a part of the aging process. Modi able

conditions su ch as p ulmon ary, coronar y, and

cerebrovascular disease an d diabetes m ellitus

should be addressed in collaboration with the

consulting medical and anesthesiology spe-

cialists to m inimize the surgical risk and op ti-

mize the outcome. Patients who smoke, have

a nut ritional de ciency, are dep ressed, or are

subject to oth er life stressors shou ld be coun-seled preoperatively to reduce the impact of

these factors.

Performing adult spinal reconstruction in

pat ients w ith os teop enia re qu ires careful pre -

operative planning, as osteopenia has impact

on both idiopathic and degenerat ive disorders.

Similarly, careful preoperative planning is re-

quired when performing a reconstruction on

younger patients with secondary osteoporosis

due to factors such as hypercort isolism, hyper-

thyroidism, hyperparathyroidism, alcohol abuse,

and immobilization.In p atients with low bone m ineral density

(BMD), spinal imp lant s cann ot be p laced as se-

curely as in patients with normal BMD, and

thus application of corrective forces through

the weak bone–implant interface is di cult. To

avoid failure in such situations, it is imp orta nt

to understand the biomechanics of the osteo-

porot ic sp ine an d to re cognize that os te op oro-

sis is a system ic disease. The m ain su rgical goal

should be set to tre at the symptom s. This chap -

ter discusses the pre- and postoperative mea-

sures that can be taken in treating patients with

osteoporosis, and the surgical strategies that

can be u sed to re duce t he r isk of failure.

! Understanding the

Modes o f Failure in

Osteoporotic Spine

In the osteoporotic spine, the two most com-

m on surgical problem s are failure of the xation

5

Instrumentation Strategiesin Osteoporotic Spine:How to Prevent Failure?

Ahmet Alanay and Caglar Yilgor



Instrument ation Strateg ies in Osteoporot ic Spine 57

or of the bone –implant inter face, and a djacent

segmen t failure, either of wh ich m ay result in

pseudar thro sis .

In the early postoperative period, pedicle

and adjacent vertebral fractures are the most

common failures, whereas in the late phase, pse udar throsis w it h in st rum entat ion failu re ,

adjacent disk degene ration, and late compres-

sion fractu res w ith progressive kyph osis occur

m ore frequently.

Because t he osteoporotic spine is less able to

withstand force, even the stresses and strains

th at are e nt ailed in th e activities of daily living

can cause postoperative implant failure, which

may present as a sudden pain, a neurologic

problem , or im plant pro m inence.

Fixation Failure

Because the elastic modulus of the bone is

smaller than th at of the imp lant, and because

the force tran smissions follow t he path of least

resistance, the bone surrounding the screws

fails before t he implant does. This phe nom e-

non is called screw toggling, and , under repet i-

tive cycling loading, pedicle screws t ypically fail

by cephalocaudal toggling. The n loosening andeventu ally pullout occur, stripping or fracturing

the pedicle. The t hinne r lateral wall of the pe d-

icle is more often fractured than the medial wall.

The p acking of a stripp ed screw h ole with cort i-

cocancellous graft does not usu ally augment t he

pullou t strength o f osteop orot ic pe dicles, which

is a possible salvage m ethod in healthy bone.1

In a cement -augmented pedicle, the screw

can be pu lled out alone, causing no dam age to

the bone or the cement, or the screw and the

cement can be pulled out together, creating

either an enlarged h ole in th e pedicle or a ped-

icle fracture.

The dorsal lamina has a thicker cort ical shell

than does the ventral aspect, which contributes

to its success in the osteoporotic spine. The

main failure mechanism of the laminar hooks

is lam ina breakout, breaking the “ring” form ed

by the lam ina, posterior verteb ra l bo dy, and

m edial pedicle walls.

Fracture of the upp er-instrum ented vertebrais anoth er comm only seen failure in the osteo-

porot ic spine.

Adjacent Segment Problems

After xation of the osteoporotic spine, almost

80%of the proxim al junctional kyph osis occurs

due to adjacent vertebra fractures.2 Instability

and adjacent disk degeneration are other pos-

sible m echanisms of adjacent segm ent failure.

The p reoperat ive statu s of the adjacent seg-

ment and disk is the greatest predictor of the

developm ent of postoperative adjacent segm ent

failure. One m ust avoid end ing a fusion adjacent

to a severely degenerated disk or to a segme nt

w ith xed obliquity or subluxation.

Nonunion and Pseudarthrosis

Similar to a h ealthy bone, an osteop orotic boneis also subject to pseudarthrosis, especially in

fusions extending to the sacrum . Known r isk fac-

tors include thora colum bar kyphosis, positive

sagitt al balance greater than 5 cm , presen ce of

hip osteoar thr itis, and incomplete sacropelvic

xation.

! Preoperative Measures

Quantifying Bone Quality

Grad ing scales from X-rays, du al-en ergy X-ray

absorpt iomet ry (DEXA), quant itative compu ted

tom ography (QCT), and m icroden sitomet ry can

be used to diagnose and qu ant ify os te op oros is

in an adult surgical candidate. QCT provides

separate BMD estimates of trabecular an d cor-

tical bone, and has a higher sensitivity due to

its imaging in a cross-section al plane. Althou gh

QCT is useful in predicting the fracture risk,

there is no clear consensus on a correlation

between the qu ant it y of os teop oros is and the

type of strategies that shou ld be applied.

The DEXA values acquired from th e fem oral

neck should be interpreted with caution be-

cause the bone density in t he spine decreases

earlier than in other skeletal sites in the early

pos tm enop ausa l years due to turnover in th is

highly trabecu lar bone. Bone den sity at various

skeletal sites begins to coincide at a bou t age 70.Also, DEXA acquired from th e verteb rae m ay

be fa lse ly elevat ed due to de generat ive ch an ges.



58 Chapte r 5

Therefore, the surgeon must be ready to deal

with a weak bone regardless of the preopera-

t ive DEXA values .

Medical TreatmentIt is well documented in the literature that

BMD correlates w ith imp lant p ullout strengt h.

Therefore, preop erative m edical treat m ent w ith

bisp hosp honates, recombin an t parathyroid hor-

mone (rPTH), calcitonin, selective estrogen

receptor modulators, calcium, or vitamin D

should be considered. It is also important to

determ ine wh ether the bene t of m edical treat-

m ent is su cient enough to o set the delay in

surgical treatmen t.The choice, timing, and duration of post-

operative pharmacological treatment for osteo-

porosis also remain con trover sial because these

drugs m ay inter fere w ith bone healing.

! Intraoperative Measures

The loss of the quan tity as w ell as the architec-

tu re of the osteop orotic bone m ay increase the

risk of spinal surgery or make the surgical

goals di cult to achieve. In these situ ations ,

speci c pedicle screw characteristics and in-

sert ion techniques can be adopted , and su rgical

strategies such as addressing the pathom orphol-

ogy of the osteoporotic vertebrae, han dling soft

tissue m eticulously, enhan cing anchor points,

applying prophylactic vertebroplasty, using

interbody support, and protecting the bone–

implant interface are utilized to improve therate of successful xation. These techniqu es

and strategies are discussed in the following

subsections.

Pathomorphology o f the

Oste oporot ic Vertebrae

It is well established th at th e bone quality varies

in di eren t part s of the vertebrae . The verte-

bra l bod y it self is the m ost a ected par t of theosteoporot ic vertebrae. The lam ina, on th e other

han d, wh ich is pred om inantly cortical, is rela-

tively spared and is potentially a stronger an-

chor. The morphometry of the pedicles are

variable. This pattern of bone loss causes the

pedicle screw xat ion t o be le ss e ect ive in the

osteoporotic bone. The xation of the ped icle

screws is achieved either by taking advantageof the relatively stronger cortical bone w ithin

the pedicle by increasing the screw diameter

and avoiding tap ping the screw path , or by aug-

menting the pedicle screw in various ways.

Sublam inar xation w ith w ires, cables, hooks,

and bands is also a good alternative because

the lamina is less a ected by osteoporosis.

The BMD also var ies in di erent r egions of

the sacrum . Medial side h as a h igher BMD than

the lateral side, and the superior sacral end

plat e h as t he h igh est . The screw s shou ld there -

fore be directed m edially in a t riangular fash-

ion an d toward t he sacral promontory.

The T2 pedicle is generally stronger than

T3–T6 p edicles, making T2 a good opt ion for

screw xation or pedicle hooks as a strong

upper an chor point.3

Pedicle Screw Factors

No conse nsu s h as yet been reach ed on the op -timal screw diameter, length, and shape for

xation in the osteoporotic bone. However,

several p edicle screw characteristics, together

with th e hole preparation and screw insertion

tactics, are shown to achieve a bette r xation

and p revent implant failure.

Pedicle Screw Characte ristics

Double-th readed ped icle screws have a cancel-

lous thread ed t ip followed by a cortical thread.

The w ider pitch of cancellous threa d p rovides

additional grip in the cancellous bone, and

the screw advances faster with higher inser-

tion torque. The cortical thread in the pedicle

area provides higher grip and less toggle du e to

denser threads.4

Conical (tapered) screws also increase in-

sertional torque, but they cannot be reversed

or backed ou t, because doing so eradicates th e

screw’s cont act w ith the b one.The expand able ped icle screw u ses a n ovel

screw design th at enables the d istal part of the




screw to enlarge within the vertebral body as

a posteriorly directed force is applied to the

screw to resist pullout failure. The tip of the

screw becomes anchored against the inne r cor-

tex of the dorsal vertebral body, resulting in a

76%increase in holding strength in compari-son to convent ional pedicle screws,5 by taking

advantage of the relatively uncompromised

cortical bone rather than depending solely on

weakene d osteop orotic cancellous bone. How-

ever, in patients with severely low BMD, ex-

pandable screw s m ay be unable to ove rcom e

the extreme biomechanical disadvantage, re-

sulting in failure. Moreover, screw revision

rem ains an issue in th e clinical application of

these screws.

Pedicle Screw Tract Augme ntat ion

It is possible to augmen t th e ped icle screw t ract

by pre paring the hole, inject ing polym et hyl-

m etha crylate (PMMA) bone cem ent into the

hole, and inserting th e screw afterw ards. Aug-

mentation may also be done with bioactive

ceme nts, calcium p hosph ate, or calcium sulfate

using trad itional or fenestr ated ped icle screws.

Coating the pedicle screw w ith hydroxyapatiteis a time-dependent augmentation technique

that increases osteointegration.

Cement Augmentation and Fenestrated Screws

A cadaveric b iome chan ical ana lysis of PMMA-

augm ente d pedicle screw xation using a novel

fenestrated bone tap increased the pullout

strength by 199%and 162%in primary and re-

vision p rocedu res, respect ively.6 Clinical ser ies

also demonstrated good outcome with no

screw loosening, m igration, or pullout detected

in th e follow- up X-rays, no fractu re at the aug-

m ente d levels, and no imp lant failure requiring

reintervention.7,8

Although PMMA augme ntat ion of the pedi-

cle screws provides good xation in patient s

with low BMD, it is not free of complications.

Extravasat ion, intr acan al leakage, hypoten sion,

increase in pulmonar y artery p ressure, pulmo -

nar y ceme nt em boli, super cial infections, andthermal nerve injuries were reported. There-

fore, strategies were developed to reduce the

likelihood of cement leakage. Higher viscosity

cement can be used, and uoroscopy can pro-

vide additional assistance. It is generally recom -

m end ed to inject 1 to 3 m L of ceme nt becau se

using a larger am ount fails to dem onstrate an y

signi cant bene t in pullout strength .6

Attention was also paid to the method of

PMMA augm ent ation. Injecting cemen t into a

cavity prepared by an in atable balloon fol-

lowed by insert ion of the p edicle screw dem on-

strated almost twice the pullout strength of

screws augmente d w ith standard cemen t injec-

tion.9 Fenestrated screws have been used m ore

recently, with promising results.10 Although

clinical long-term results are yet to be seen,

there is a potential theoretical advantage of

using fenestr ated screws over injecting ceme nt

followed by screw insertion. Injecting the ce-

me nt into the prepared hole lls the tract, and

wh en inserting the p edicle screw, the cem ent

coats the screw threads and thereby reduces

e ect ive screw purcha se. Altern atively, cem en t

injection through a fenestrated screw enables

the cem ent to in ltrate in the vertebral body

without altering the bone–implant interface.11

Although it is widely used with promising

results, PMMA is toxic, is unable to undergoremodeling after microfracture within the ce-

m ent , and is di cult to rem ove in revision

surger y. Hence, osteob iologic cem en t is an area

of interest and development for screw aug-

m ent ation. Calcium ph osphate an d calcium su l-

fate avoid the exother m ic reaction and reduce

the risk of leakage. Moreover, they are bio-

resorbable and poten tially osteoconduct ive, and

integrate in the nat ural process of bony rem od-

eling. A cadaveric study comparing osteobio-

logic cement and PMMA for the use of screw

augme nt ation found no signi cant di eren ces

in axial pullout stre ngth.12

Hydroxyapatite Coating

The increased osteointegration of the hydroxy-

apatite-coated pedicle screws is time depen-

dent ; w ith tim e, optim um stability is achieved.

It has been shown in an osteoporotic animal

model that hydroxyapat ite-coated pedicle screwsare 1.6 times m ore resistant to pu llout an d th at

they have superior biological bonding to the



60 Chapte r 5

surrou nd ing bone, occurring as early as 10 days

after surger y.13 However, th ey do not allow for

the application of add itional forces dur ing in-

traoperative correction maneuvers.

Inse rtion Technique and

Inse rtional Torque

The star ting point, hole preparat ion, tap ping, the

insertion a ngle, and the trajectory of a pedicle

screw, as well as its length, depth of penetra-

tion, and diamet er, a ect its insert ional torque

and the reby its resistance to failure.

Length of Screw and Depth of

Screw Penetration

The length of a pedicle screw is linea rly related

to its pullout stre ngth . As the screw p enet rates

furth er into the vertebr al body, the cu tout load

to failure increases. By engaging the ventral

cortex of the vertebral body, screws can be

p laced in a bicor t ical fash ion to provid e up

to an ad ditional 30%of pullout strength .14 The

risk–bene t ratio should be considered, and

care m ust be t aken to avoid injur y to adjacent

structures w hen using this technique.

Diameter

The d iam eter of the p edicle screw shou ld be as

w ide as possible to enable better cortical bone

purch ase. Increasing the diam eter increases t he

pullout st re ngt h in m ilder cases; how ever, in

severe osteoporosis the p ullout force is low re -

gardless of th e screw diamete r. Instead, using

larger diam eters m ay cause d ilation or fracture

of the p edicle that d ecreases its strengt h.15

Starting Point, Insert ion Angle, and Trajectory

When the screws are placed parallel, the vol-

um e of cancellous bone betw een t he t hreads

of the screw deter m ines the resistance to pull-

out for each screw. Triangulated screws provide

bet ter p ullout st rengt h w ith a lar ger volum e o f

cancellous bone available for resistance to p ull-

out because the construct is contributed by thevolume of bone within the trapezoid area in

the vertebr al body form ed by longer and tr ian-

gulated screws.16

Thoracic Spine

In t he anatom ic trajectory, the screw is in line

w ith the ped icle axis and the refore is directed

to the inferior corner of the vertebral body in

the sagitt al plane. In th e straightforw ard te ch-

nique, the screw is parallel to th e vertebral end

plat e an d t rian gu lat ion in the t ransve rse plane

can be achieved. This technique provides at

least 39%h igher m axim um insertional torque

and 27%greater pullout strength.17 Pedicle-r ib

screws increase the e ective transverse diam-

eter wh en compared w ith the pedicle alone and

can be used for safer insertion of the screws,although it m ay decrease the pullout strength

by 25 %.18 The sta rting point sh ould be selected

in accordance w ith the t rajectory used.

Lumbar Spine

Placing the pedicle screws in convergence also

increases the pullout strength in the lumbar

spine.

Sacrum and PelvisSacral xation is a big challenge in the osteo-

porot ic sp ine. Restorat ion of the sagit tal bal-

ance is more imp ortant than the xation itself.

When the fusion is extended to the sacrum

and a long fusion is performed, multiple and

bicort ica l screw xat ion sh ou ld be use d in ad-

dition to considerat ion of ante rior colum n sup -

por t or iliac xation. The t ricort ica l t echniqu e,

which entails directing the screws into the

sacral promontory, increases the insertional

torque.19

Hole Preparation and Tapping

Appropriate preparation of the hole improves

screw purchase. High insertional torque im-

proves the screw pullout st re ngt h. In hea lthy

vertebral bodies, the screws are placed after

tapping to avoid microfracturing within the

dense bony matrix of the bone during screw

insert ion. In osteoporotic cancellous bon e, how-ever, tapping results in rem oval of bone w ithin




the pedicle track and prevents bone compres-

sion around t he screw t hread s. Even screw re-

m oval and imm ediate reinsertion decreases the

m echanical insert ion torque. Therefore, und er-

tapp ing or not tap ping at all is advised in oste-

oporotic bone. When compared with same-sizetapping, under-tapping by 0.5 and 1.0 mm

increases the insert ional torque by 47% and

93%, resp ectively.20

Under-t apping is m ore bene cial in the lum -

bar sp ine than in the thoracic sp ine. This m ay

be due to the fact that the thor acic pedicle

screws are probably more dependent on cor-

tical purchase w ithin the ped icle walls.

Enhancing Anchor PointsThe weak link in the osteoporotic spine in-

strumentation is the implant–bone interface.

Fixation strategies for osteoporotic bone are

targeted toward taking advantage of the rela-

tively stronger cortical bone. Anchor options,

in addition to screws, include hooks, wires,

cables, and b ands.

Load Sharing by Multileve l Fixation

Because th e implant–bon e interface in th e os-

teoporotic bone is prone t o failure, the num ber

of points of xation m ust be increased to dis-

tribute the contact forces more evenly. Longer

construct s w ith at least thre e sets of xation

points at e ach end can be bene cial, keeping in

mind t he added morbidity entailed w ith using

additional screws.

As previously stated, using hooks, wires, ca-

bles, an d ban ds as w ell as cross- links h elp in im -

proving the p er form an ce of the p edicle screw s.

Select ion of Fusion Leve ls

End-instrumented vertebrae should be care-

fully selected. Ending the construct in a ky-

phot ic region or at the apex of kyphosis shou ld

be avo ided.

Anoth er frequen t decision-m aking dilem m a

in the osteoporotic spine is wh ether or not to

fuse to sacrum. Certain scenarios that requirelum bosacral xation are symp tom atic L5-S1

spond ylolisth esis, over 15 degre es of scoliosis

at the L5-S1 segmen t, and th e nee d to achieve

prop er sagit tal ba lance.21 Stop ping at L5 ent ails

the risk of increased adjacent segm ent disease,

wh ereas fusing to the sacru m is found to have

more complications.21 L5 pedicles are usuallyshort and cont ain more cancellous bone. There -

fore, it may be r isky to end a long fusion at L5

in osteoporotic patients because L5 pedicle

screws m ay fail.

Cross-Link

The use of a rigid or semirigid cross-link, es-

pecially w hen the screws are t rian gu lat ed, in-

creases the torsional sti ness by making the

construct per form e ectively as a quadr ilateral

frame. The use of a cross-link is especially ad-

vantageous in longer constr ucts, as it p revents

rods from telescoping.

Hoo ks, Wires , Cables, and Bands

The use of sublaminar and pediculolaminar

hooks, wires, cables, and bands takes advan-

tage of the cortical bone composition of the

spinal lam ina.A polyester band may be used to increase

the surface of bony contact and to t any anat-

omy. It m ay be used in a sub laminar, subpa rs,

tran sversal, or lam inotran sversal fashion to en -

able translation, distraction and compression,

in situ be nding, and rod derotat ion.

Prophylactic Vertebroplasty

In the setting of osteoporosis, junctional fail-

ure, especially in th e cran ial levels, is not a rar e

occurrence. Prevention is the best way to

overcome adjacent segment failure. Prophy-

lactic vertebroplasty ent ails cemen t augm ent a-

tion of the adjacent noninstrum ented segmen t/

segments. Although there is a paucity of clini-

cal and biomechanical studies, prophylactic

vertebroplasty seem s to be helpful in de creas-

ing the revision arthrodesis rates because of

adjacent vertebr ae fractures.22

Furt her stu dies are needed to clarify the op -timum amount of cement required and how



62 Chapte r 5

many levels should be prophylactically ce-

m en ted . E cacy at the distal adjacent level

also need s to be furt her an alyzed.

Interbody Support in

Osteoporotic Spine

Ante rior column supp ort is bene cial in load

sharing because the graft or cage lessens the

stress directed toward t he screw–rod constru ct.

Anterior interbody support may further im-

prove sagit tal sp in al balance and rat es of

arthrodesis.

Inter body grafts serve a m ore critical role at

the cau dal end of the constru ct, particularly atthe lumb osacral jun ction. Grafts can be placed

w ith a bias toward the concavity of the defor-

m ity to assist correction. Figs . 5.1, 5 .2, and 5.3

Fig. 5.1a,b (a) Preope rative post eroante rior X-ray. (b) Preoperat ive lateral X-ray.

a b




dem onstrate the u se of the intraoperative m ea-

sures m entioned above in a 64-year-old pat ient

with a BMD of –3.2 complaining of back and

leg pain.

In severe oste oporosis, th ere is a risk of sub -

sidence of the graft or cage into the end plates

that may lead to anterior column collapse

and subsequent kyphotic deform ity. The cage

should be placed to contact the peripheral

apophyseal ring where the cortical bone is

stronger. A graft with an elastic modu lus that is

similar to th e nat ive bone also reduces the r isk

of subsidence. Choices of interbody graft m ate-

rials include bon e (au tograft or allograft), m etal,

carbon ber, polyaryleth eret he rketon e (PEEK),

and other synthet ic mater ials. Iliac crest auto-

graft is typically the be st m atch, but it is asso-ciated with well-established harvest-related

morbidity.

Role of Ante rior Fixation in the

Oste oporotic Spine

Cont inuous loading of ant erior screw constructs

on a low-BMD spine can lead to screw cutout.

Although newer implant designs demonstrate

improved anchorage,23 anterior xation has alim ited role in th e osteoporot ic spine because it

is the m ost a ected part of the vertebrae.

Role of Semirigid Fixation in the

Oste oporot ic Spine

The loading of the spine in various axes of

motion creates increased stress at the bone–

implant inter face. The di eren ce of the r igidity

within th e instrum ented and n oninstrumentedsegments of the spine can accelerate adjacent

segment degeneration and potentially cause

pse udarthrosis. Sem irigid xation m ay p rovide

su cient stab ilization to facilitat e bony fusion

wh ile per m itting some degree of exibility to

o oad stress at the adjacent segm ent and the

bo ne–im plant in terface .

Protection o f the Bone–Implant

Interface

Hand ling int raoperative soft tissue m eticulously,

providing extensive release , perform ing oste-

otomies to increase exibility and thu s m ini-

m ize the corrective forces, maint aining sagittal

alignm ent , and obt aining a solid fusion ar e es-

sential for the protection of the bone–implant

interface.

Meticulous Soft Tissue Handling

Care should be taken to preserve the supra-

spinous ligamen t, int raspinous ligamen t, and

Fig. 5.2a,b (a) Preoperative sagitt al magnetic resonance imaging (MRI) view. (b) Preoperative transverse

MRI view.

a b



64 Chapte r 5

Fig. 5.3a,b (a) Last follow-up posteroanterior X-ray

of the same patient demonstrating the use of

cement-augmented pedicle screws, ante rior

interbody support, prophylactic vertebroplasty,

and cross-link. (b) Last follow-up lateral X-ray of

the same patient dem onstrating the use of ceme nt-

augmented pedicle screws, ante rior interbody

support, p rophylact ic verteb roplasty, and cross-link.

a b




ligament um avum between the cranial fused

level and th e adjacent segm ent , as well as be-

tween the cranial levels of fusion. Doing so

provid es a segm ent of higher p oste rior tensio n,

wh ich m ay prevent the developmen t of junc-

tional deformity and instability. Supra- andinfra-adjacent facets should be preser ved and

the cranial disk space should not b e violated

w ith pedicle screws.

Exte nsive Release

The release of the d iskoligamen tous an d bony

constraints such as diskectomy, facetectomy,

or various osteotomies should be optimized

before r educt ion to d ecrease the st ress ap plied

to the bone–implant interface w hen perform -

ing spinal corrective m ane uvers.

Maintaining Sagittal Alignment

Aligning th e oste oporot ic spine to physiologi-

cal coronal and sagittal contours neutralizes

the deforming forces, reduces the junctional

forces, and d ecreases the ener gy required for

ambulation.

Obtaining Fusion

Obtaining a rapid and solid fusion ensures

long-ter m spinal stability, redu cing th e load on

instruments and on the relatively poor bone–

implant interface. A thorough fusion p rocedure

with appropriate bone-bed preparation and

approp riate use of bone grafts or substitu tes is

the refore of par ticular imp orta nce in the oste-

oporotic spine. The u se of bone m orph ogenetic

pro te in m ay facilit at e an earlier an d m ore vig-orous fusion, and d ecrease th e risk of implant-

related failure. The use of bone morphogenetic

pro te in m ay also be associat ed w ith complica -

tions related to soft tissue swelling, inappro-

priat e bon e form at ion ar ou nd n eura l elem ents,

and subsequ ent ra diculitis.

! Postoperative Measures

Enh ancing the purchase of intern al xation and

pro te ct ing t he bo ne–im plant in terface by han-

dling soft tissue meticulously, providing ex-

ten sive releases, maint aining sagittal alignm ent ,

and using anter ior interbody suppor t decreases

the demand on xation in the postoperative

period. Ext ernal br ace im m ob ilizat ion and re -

striction of the spine r ange of motion a s wellas physical therapy and reh abilitation are m ea-

sures that can be taken in the postoperative pe-

riod to serve the sam e pu rpose. As stated ea rlier,

the timing of postoperative pharmacological

osteoporosis treat m ent re m ains cont roversial.

If a brace is to be used, it must be custom-

molded p ostoperat ively, after surgical deform ity

correct ion is established . Rehab ilitation sh ould

focus on gait training, balance, and general

conditioning, together with range-of-motion

and exibility exercises of th e hip an d kne e.

There is no consensus yet on the duration of

br ace applicat ion.

! Chapter Summary

Prim arily a ecting the tr abecular bone, osteo-

poros is causes progre ssive bon e m ineral loss

and concomitant bony architecture changesthat result in pain, deformity, increased frac-

ture risk, and possible neural compression.

Although most patients with acute, painful

vertebral compression fractures improve re-

gardless of the treatment applied, no patient

spontaneously restores the vertebral height

and achieves a realigned spine. Spinal instru-

m ent ation is eventu ally required in som e oste-

oporotic patient s. In the sett ing of osteoporosis,

however, the xation of the spinal implant s is

insecure, and application of corrective forces

through a weak bone–imp lant interface is di -

cult, comp licating the surgical treatm ent . The

rst step for an adult spinal surgical candidate

is the diagnosis and quant i cation of the oste-

oporosis. Unde rstan ding the biom echanics and

the modes of failure of the osteoporotic spine

is important. The vertebral body itself is the

most a ected part of the osteoporotic verte-

br ae. The lam ina, o n the ot her han d, w hich is

predom inantly cort ica l, is relat ively spar ed andis potentially a stronger an chor. The m orph om -

etry of the pedicles is variable. Failure of the



66 Chapte r 5

xation or of the bone–imp lant inter face, adja-

cent segment failures, and pseudarthrosis are

the t hree m ain problems in osteoporot ic spine.

Several pre- and postoperative measures may

be taken, as well as app lying several su rgical

strategies intraoperatively to prevent failure.The pedicle screw characteristics together

with hole preparation and screw insert ion

tact ics are show n to achieve a bette r xation.

Ante rior column supp ort is bene cial in load

sharing, improving sagittal balance and re-

ducing the rates of arthrodesis. The results of

cement augment ation of the pe dicle screws and

the adjacent noninstrum ented vertebrae seem

prom ising, bu t it is not a com plicat ion-free

procedure.

Pearls

In elective cases, increasing t he BMD preoperat ively

with parathyroid hormone might be considered.

Under-tapping is advised in the osteoporotic

bone to increase the inse rt iona l torque.

Triangulated screws provide better pullout

strength, with a larger volume of cancellous bon e

available for resistance to pullout because the

construct is contributed by the volume of bone

within the trapezoid area in the vertebral bodyformed by the longer a nd t riangulated screws.

The use of a cross-link is especially advant ageous

in longer constructs, as it prevents rods from

telescoping.

The use of hooks, wires, cables, and bands take

advantage of the relatively stronger cortical bone

for xation of osteop orot ic bon e.

Cem ent -augmen ted p edicle screws are advanta-

geo us for bet te r xation and for allowing add i-

tional correct ive forces.

Prophylact ic vert ebrop lasty is helpful in decreas-

ing t he revision art hrodesis rates because of adja-

cent vertebrae fractures.

The release of the diskoligamentous and bony

constraints such as d iskectom y, facetect omy, or

various oste oto mies should be opt imized before

reduction when performing spinal corrective

pro cedures to de crease the st ress app lied to the

bo ne –implant inter face.

A custom-m olded posto perative brace h elps pro-

tect the bone–implant interface.

Pitfalls

Avoid ending a fusion adjacent to a severely

de gene rated disk or to a segm ent with xedobliquity or subluxation.

DEXA scans in elderly patients must be inter-

pret ed with caution because degenerative changes

may falsely elevate th e BMD values.

In th e oste oporotic bone, tap ping results in re-

moval of bone within the pedicle track and

prevent s bo ne compression aro un d the screw

threads.

Avoid ending a construct in a kyphotic region or

at t he a pex of kyphosis.

Avoid dam aging the supra- and intraspinous liga-

ment s and ligamentum avum between the cra-

nial fused level and the adjacent segm ent , as well

as bet ween the cranial levels of fusion.

Avoid damaging the supra- and infra-adjacent

facets and violating the cranial disk space with

pe dicle screws.

References


1. Halvorson TL, Kelley LA, Thom as KA, Wh itecloud TS

III, Cook SD. E ect s of bon e min era l densit y on ped icle

screw xation. Spine 1994;19:2415–2420 PubMed

2. DeWald CJ, Stan ley T. Inst ru m ent ation- related

complications of multilevel fusions for adult spinal

deformity pat ients over age 65: surgical considera-

t ions and treatmen t options in patients w ith poor

bo ne qu alit y. Sp in e 2006 ;3 1(1 9, Sup pl): S14 4–S1 51

PubMed

3. But ler TE Jr, Ash er MA, Jayar am an G, Nun ley PD,

Robinson RG. The str ength and st i ness of thoracic

implant anchors in osteoporotic spines. Spine

1994;19:1956–1962 PubMed

4 . Mu m m an en i PV, Hadd ock SM, Liebsch ne r MA, Keav-

eny TM, Rosen ber g WS. Biom echa nical evaluat ion of

a doub le-threaded pedicle screw in elderly vertebrae.

J Spina l Disord Tech 20 02;1 5:64 –68 PubMed

5. McKoy BE, An YH. An exp an da ble an chor for xa-

tion in osteoporotic bone. J Orth op Res 2001;19:545–

547 PubMed

6. Fra nkel BM, D’Agost ino S, Wan g C. A biom ech an i-

cal cadaveric analysis of polymet hylmethacr ylate-

augm ente d pedicle screw xation. J Neurosurg Spine

2007;7:47–53 PubMed

7 . Chan g MC, Liu CL, Che n TH. Polymethylm etha crylate

augmentation of pedicle screw for osteoporotic spi-




nal surger y: a novel techn ique. Spine 200 8;33:E317–

E324 PubMed

8. Aydogan M, Oztu rk C, Karat opr ak O, Tezer M, Aksu N,

Hamzaoglu A. The pedicle screw xation with verte -

br op last y a ugm en tat ion in the su rg ica l t re at m ent of

the severe osteoporotic spines. J Spinal Disord Tech

2009;22:444–447 PubMed

9. Burval DJ, McLain RF, Milks R, Inceoglu S. Primary

pedicle screw au gm en tat ion in oste op orot ic lum ba r

vertebrae: biomechan ical analysis of ped icle xation

strength. Spine 2007;32:1077–1083 PubMed

10 . Cha ng MC, Kao HC, Ying SH, Liu CL. Polymethylm e-

thacrylate augmen tation of cannulated ped icle screws

for xation in osteoporot ic spines and comparison of

its clinical results an d biom echanical characteristics

with the needle injection method. J Spinal Disord

Tech 2013;26 :305–31 5 PubMed

11. Wang MY, Hoh DJ. Bone m etab olism and osteop o-

rosis and its e ects on spinal disease and surgical

tr eat m en ts. In: W inn HR, ed . Youm an s’ Neurological

Surgery. Philadelphia: Elsevier; 2011

12 . Roh m iller MT, Schwa lm D, Glat te s RC, Elalayli TG,

Spengler DM. Evaluation of calcium sulfate paste for

augmentation of lumbar pedicle screw pullout

stren gth. Spine J 2002;2:2 55–260 PubMed

13 . Hase gaw a T, Inu fus a A, Imai Y, Mikaw a Y, Lim TH,

An HS. Hydroxyapatite-coating of pedicle screws im-

proves re sis tan ce against pull-ou t force in the oste o-

por ot ic canine lum ba r sp ine m od el: a pilot st udy.

Spine J 2005;5:239–2 43 PubMed

14 . Zind rick MR, Wilt se LL, Wide ll EH, et al. A biom e-

chanical stud y of intrap edu ncular screw xation in

the lumbosacral spine. Clin Orthop Relat Res 1986;

203:99–112 PubMed

15 . Yazic i M, Pekm ezci M, Cil A, Alan ay A, Acar oglu E,

Oner FC. The e ect of pedicle expan sion on pe dicle

m orph ology and biom echanical stability in the im -

mature porcine spine. Spine 2006;31:E826–E829

PubMed

16 . Hadjipavlou AG, Nicodem us CL, al-Ham da n FA, Sim -

mons JW, Pope MH. Correlation of bone equivalent

m ineral density to pull-out resistan ce of triangulated

ped icle screw const ruct . J Spin al Diso rd 19 97 ;1 0:

12–19 PubMed

17. Leh m an RA Jr, Polly DW Jr, Kuk lo TR, Cun nin gha m B,

Kirk KL, Belmo nt PJ Jr. Str aight- forwa rd versu s an ato-

mic trajectory technique of thoracic pedicle screw

xation: a biomechanical analysis. Spine 2003;28:

2058–2065 PubMed

18 . White KK, Oka R, Mah ar AT, Low ry A, Gar n SR. Pullout

strength of thoracic pedicle screw instrumentation:

comparison of the transpedicular and extrapedicular

techniques. Spine 200 6;31:E355–E358 PubMed

19 . Leh m an RA Jr, Kuk lo TR, Belmon t PJ Jr, And er se n RC,

Polly DW Jr. Advant age of ped icle screw xation d i-

rected into the apex of the sacral promontory over

bicor t ica l xat ion : a biom echan ica l an alysis . Spin e

2002;27:806–811 PubMed

20. Kuk lo TR, Lehm an RA Jr. E ect of variou s t ap pin g

diameters on insertion of thoracic pedicle screws: a

biom echan ica l an alysis. Spin e 20 03 ;2 8: 20 66 –2 07 1

PubMed

21. Bridwell KH, Edwards CC II, Lenke LG. The pros and

cons to saving the L5-S1 motion segment in a long

scoliosis fusion constru ct. Spine 2003;28 :S234–S242

PubMed

22 . Chian g CK, Wan g YH, Yan g CY, Yan g BD, Wa ng JL.

Prophylactic vertebroplasty may reduce the risk of

adjacent intact vertebra from fatigue injury: an ex

vivo biomechanical study. Spine 2009;34:356–364

PubMed

23. Goldhah n J, Reinh old M, Stau ber M, et al. Impr oved

anchorage in osteoporotic vertebrae with new im-

pla nt designs. J Orthop Res 2006;24:917–925 PubMed



! Introduction

Signi cant advances have been m ade in adult

spinal deformity (ASD) surgery over the past

several years. Improvements in pedicle screw

technology and the increasing use of three-

column osteotomies have m ade m ore powerfuldeformity corrections possible. Nonetheless,

loss of neurologic function after ASD surgery

remains a serious and potentially devastating

complication, with profound conseque nces for

health -related qua lity of life. Int raoperative neu-

rophysiology mon itoring has be come a reliable

and e ective moda lity to optimize neu rologic

safety d ur ing ASD surger y.1,2 Furth erm ore, the

increasing use of im age-guided navigation sys-

tems has led to signi cant improvements in

th e accuracy of ped icle screw placem ent .3,4 This

chapter discusses the management of acute

ne urologic com plications in com plex ASD sur-

gery and proposes a treatment algorithm t o deal

w ith these complications in safe and e ective

manner.

! Prevalence

The incidence of neurologic complications in

ASD patients un dergoing deform ity correction

surgery has been di cult to deter m ine. Mul-

tiple aws exist in the published data, w ith a

lack of high-qu ality prosp ect ive st ud ies, signif-

icant d ata var iability, an d a lack of rigorous an d

validated measurements of neurologic func-

tion. Nonetheless, the incidence of neurologic

de cits following ASD surger y has been p re-viously report ed as ran ging from 0% to over

10%.5– 9

Two previous studies, one from a single in-

stitution, an alyzed prospectively collected d ata

to identify comp lications, but n either on e used

a validated scoring system to quantify neuro-

logic funct ion.10,11 All studies could more ac-

curately be described as retrospective studies

of prospect ively collected dat a. But an accurate

rate of neu rologic comp lications after complex

ASD surgery is critical for informed decision

making for both patients and surgeons. Fur-

thermore, it is crucial to be able to measure

changes in neurologic complication rates in a

standa rdized fashion so as to accurately evalu-

ate new techniques, technologies, and thera-

pies in ASD su rgery.

The Scoli-Risk-1 t rial is a recen t p rospective,

multicenter observational study attempting to

accurately assess the neurologic complication

rate following complex ASD surgery using theAm er ican Spinal Injur y Associat ion (ASIA) scor-

ing system (Fig. 6.1).12 A total of 276 pat ients

6

The Incidence and Managementof Acute Neurologic ComplicationsFollow ing Complex Adult SpinalDeformity Surgery

Joseph S. Butler and Law rence G. Lenke



Neurologic Complicat ions Following Adult Spinal Deformity Surgery 69

Fig.6.1

TheAmericanSpinalInjuryAssociation(ASIA)scoringsystem.



70 Chapte r 6

were enrolled from 15 international centers.

At hospital discharge, 23.1% of patients had a

m easurable lower extre m ity motor de cit (i.e.,

less tha n 5/5 m otor strength in all ve m ajor

leg muscles) decreasing to 17.8%at 6 weeks

and 10.7% at 6 mon ths postoperat ive. Thisstudy gives a clearer indication of the expected

neurologic complication rate in ASD patients,

pot ent ially se t t ing a st andard for future clin ica l

trials aim ed at lowering the rate of postopera-

tive neuro logic de cits.

! Mechanisms of Neurologic

ComplicationsThere are several proposed causes of intraop -

erative neurologic de cits during ASD sur-

gery. Int raop erat ive spinal cord, caud a equin a,

or nerve root de cits may result from direct

neurologic trauma during instrumentation

such as t he p lacem ent of pedicle screws, hooks,

or sublaminar w ires. Furt her m ore, intraope ra-

tive corrective maneuvers may lead to neuro-

logic de cit secondary to either d istract ion of

the neural elements or excessive tension onlocal vasculature, leading to decreased blood

ow and cord ischemia. Spinal cord ischemia

may also result from prolonged extreme hy-

pot ension (m ean ar ter ial p ressure [MAP] < 55

mm Hg), hypoxia secondary to decreased he-

moglobin level, or vascular compromise after

ligation of the segm ent al vessels in an anter ior

procedure.13

! Patient Evaluation and

Preoperative Planning

Neurologic complicat ions are m os t st rongly

associated with prolonged complex surgery, a

large am ount of blood loss, combined anter ior/

pos ter ior procedures , m ult ist age su rgery, con-

genital kyphosis or scoliosis, large or rigid spinal

cur ves (Cobb a ngle > 90 degre es), preexisting

myelopathy or neurologic de cit, and intra-medullary spinal cord tumors. Further risk

factors include tethered cord, Arnold-Chiari

malformation, syringomyelia, and split cord

malformations.

A complete patient history and thorough

physical exam inat ion st ill rem ain e ssent ial ele-

ments to an adequate preoperative workup.

Patients should be assessed for a history ofcongenital deformities such as kyphosis and

scoliosis, neu ro brom atosis, and skeletal dyspla-

sia, which would infer a considerably increased

risk of iatrogenic neurologic complications. The

physical exam inat ion sh ou ld inclu de a thre e-

dimensional assessment of the spine to evalu-

ate pat ient postu re, neurologic statu s, hip exion

contract ures, leg length inequality, pelvic obliq-

uity, body habitu s, and nut ritional status. Me-

ticulous examination of the motor, sensory,

and re ex fun ction as well as gait assessmen t

is critical in screening patients for potential

intraspinal and brainstem anomalies such as

teth ered cord, Arn old-Chiari m alform ation, sy-

ringomyelia, and split cord malformations.

Adeq uate rad iological im aging is crucial for

optim al surgical and n eurologic outcome. How-

ever, it is technique-dep ende nt, requ iring vi-

sualization of the entire spine in the coronal

and sagittal planes, including the hip joints,

with all im aging taken w ith the patient stand-ing with th e knees fully extende d for accurate

measurement of sagittal balance (sagittal ver-

tical axis [SVA]), th oracic kyph osis, lum bar lor-

dosis, and spinopelvic parameters including

pelvic in cid ence (PI), sa cral slop e (SS), an d pel-

vic tilt (PT). Lateral dynamic standing lumbar

X-rays m ay ident ify focal instab ility or sp ond y-

lolisthesis. Ben ding lms an d sup ine X-rays

w ithout th e e ects of gravity help assess the

exibility of a deformity. Once the appropri-

ate radiographic studies have been obt ained,

the sagittal and coronal balance can then be

assessed.

Magnetic resonance imaging (MRI) is used

as a routine p reoperat ive radiographic study to

assess central can al stenosis, facet hyper troph y,

pedicu lar anom aly, foram inal encroachm ent ,

and degenerat ive d isk disease. It also helps de-

term ine the p resence of intraspinal anom alies.

Patients with suspected low bone m ass or with

established osteoporosis should have a dual-en ergy X-ray absor pt iome tr y (DEXA) scan p er-

form ed to optim ize surgical planning.




! Intraoperative Preparation

Meticulous intraoperative preparation is re-

quired for the safe and e ective man agement of

intraoperative complications. Before induction

of anesthesia, the surgical team should discusswith the anesthesia and neuromonitoring

teams and operating room sta the patient’s

m edical situation, the inten ded procedu re, and

surgical time frame. An arterial line may be

used to monitor MAP. Somatosensory evoked

potent ial (SSEP) and m otor evoked pot ent ial

(MEP) leads a re p laced and checked before th e

pat ient is turned to the prone pos it ion. The

upper extremities are padded and posit ioned

to avoid stret ch or compr ession of th e brachial

p lexu s, and ca re is t aken to ensu re that the

pressure areas are well padded. Forced-air

warming blankets prevent hypothermia, par-

ticularly in pr ocedures of long durat ion.

Maintaining adequate blood pressure is es-

sential. However, a balance should be main-

tained to minimize intraoperative blood loss

and t ransfusions, yet en suring adequate spinal

cord p erfusion . A MAP of < 55 m m Hg has been

associated w ith an increased risk of spinal cord

ischemia.14 However, mild hypotensive anes-thesia is often used to minimize blood loss,

par t icu larly during the surgical a ppro ach , with

the MAP maintained at 65 to 70 mm Hg. Ap-

proxim ately 30 m inutes before per form ing cor-

rective man euvers, the an esthesia team should

gradually elevate the MAP to > 70–80 mm Hg

to maintain adequate cord perfusion during

spinal column manipulation and deformity

correction.

! Intraoperative

Neuromonitoring

Stagnara Wake-Up Test

The Stagnara wake-up test has been widely

used in the intraop erative assessme nt of neu -

rologic function. It assesses primary motor

cortex, anterior motor pathways of the spinalcord, ner ve roots, and periph eral ner ves. How-

ever, it gives only a gross approxim ation of th e

function of these elements and does not di-

rectly m easure any components of the sensory

system.15 This test involves a tem porar y redu c-

tion in anesthesia, after which the patient is

asked to move the upper and lower extremi-

ties. The test is limited, as it is entirely relianton patient compliance and cannot be u sed in

pat ients unab le to follow com m ands because

of intellectual and developmental disability,

young age, or p reoperat ive w eakness. The test

itself carries risk, including self-extubation,

loss of intravenous access or of safe patient-

pos it ion ing on the t able, air em bolism , and

posto perat ive re collect ion of t he event .

The w ake-up test w as historically the bench-

m ark for intra operative neu rologic assessmen t,

and is still used at som e center s in conjunct ion

with advanced neuromonitoring techniques

as a means of con rming neurologic status.

Properly administered, the wake-up test should

be 100% accurate in detect ing gross m ot or

changes.15 Although th e limitations of the te st

prevent assessm ent of ne m otor ch an ges, it

w ill alert t he surgeon to the m ost clinically sig-

ni cant neurologic de cits. It is used when

there is any problem obtaining spinal cord

monitoring (SCM) signals (such as in a patientw ith th oracic myelopathy) and also in p atient’s

who have had SCM changes meeting evoked

potent ial w ar ning cr iteria w hen the re spon ses

cannot be improved. It also should always be

perform ed at the e nd of t he surgical p rocedure

pr ior to the pat ien t’s leaving th e ope rating room.

Som atosensory Evoked Potentials

Somatosensory evoked potentials assess the

po ster ior colu m ns of the sp inal cord , in ad dit ion

to the cerebral cortex and mixed peripheral

ner ves. The posterior column s are respon sible

for proprioception as opposed to pain and te m -

pera ture . Althou gh pro priocept ive loss is n ot as

debilitating as a mot or de cit, it can have a sig-

ni cant imp act on activities of daily living. As

SSEPs are sensitive to focal posterior column

and global spinal cord issues, they act as a good

surrogate for othe r n eural p athw ays. However,

there are situations when a motor de cit mightnot be d em onstrated on SSEP m onitoring. For

example, anterior vascular territory comprom ise



72 Chapte r 6

w ithout concomitan t posterior vascular changes

m ight not be addressed by SSEP mon itoring.

Somatosensory evoked potentials continue

to be the most frequently used intraoperative

monitoring method to assess the integrity of

the dorsal column but cannot be relied on tomonitor motor function directly. Reports of

post operat ive paraparesis in the abs ence of

intraoperative SSEP signal change underscores

this importan t lim itation.16 SSEP de pend ability

falls o wh en applied to patient s w ith preex-

isting neurologic conditions. Individual nerve

root injury is not e ect ively monitored by SSEPs.

Missed nerve root or isolated motor pathway

complications are not failures of the m odality

bu t rat her are issues outsid e of SSEP mon itor ing

capability, highlighting t he need for a lternat ive

or adjunct m onitoring approaches. Warn ing cri-

ter ia for SSEPs at out inst itut ion include greater

than a 60% decrease in amplitude or a 10%

increase in latency of the signal as compared

w ith baseline values.

Motor Evoked Pote ntials

Motor evoked poten tials m onitor corticospinal

tract activity via stimulation at the level of themotor cortex or spinal cord and are selective

for motor pathways. MEP monitoring relies

on intervening thalamic synapses to prevent

ant idrom ic ring of spinal sensory tracts. The

stim ulation site for t ran scranial MEP (tcMEP) is

the cerebral cortex. MEP end-point data are as-

certained from the spinal cord (D-wave) or

from t he end m uscle compound m otor action

pot en t ial (CMAP). Stim uli are prese n ted as

single high voltage or multiple small stimuli.

Sources of stimulation include magnetic and

electrical. For magnetic tcMEP, a coil over the

cortex provides the stimulation. Electrical

stimulus of the motor cortex is provided by

subdermal electrodes. Although occasionally

associated w ith scalp edem a and u nreliable re-

cordings, corkscrew electrodes are preferable

given their low impedance and secure posi-

tioning in the scalp. Peripheral data are com-

m only elect rom yograp hic via CMAP. The CMAP

is best m onitored at sites rich in corticospinaltract innervation such as the distal limb mus-

cles. Com m on r ecording sites are abductor pol-

licis brevis or adductor hallucis brevis with

viable alternat ives of long forearm exors and

extensors in the upper extremity and tibialis

ante rior in the lower extrem ity. Although the re

does not app ear to be any monitoring advan-

tage to increasing the number of monitoredm uscles, increased m uscle group test ing might

p rovid e a be ne t in ident ifying pos it ioning-

related injury.

Electromyography

The clinical applications of electromyography

(EMG) an d its speci city for the m otor system

led to the introduction of spontaneous EMG

(sEMG) recordings. sEMG myotomes are pre-

selected to coordinate with operative levels,

and m uscle relaxants m ust not be utilized du e

to dam pen ed or even absen t activity. Continu -

ous electrical activity to a myotom e is recorded

and observed and may be indicative of root

irritation. When a n er ve root is noted to be ex-

cessively ma nipu lated or imp inged, tr igger ing

a burst of activity, with more severe nerve

manipulation and stretch of a nerve root train

activity is also noted. One would gen erally note

silen ce if th e n erve root is cleanly severed. Dis-tal recording sites are typically paired w ith an

intramuscular needle or wire electrodes in-

serted after induction but before surgery.

Triggered EMG (tEMG) has also been used

as it is postulated that a high stimulus inten-

sity tEMG w ill dem onstrate an int act cortex of

a ped icle hole through wh ich a screw is passed.

In application, bone has high impedance re-

quiring high thresh old to stimulate the adjacent

ner ve. Wh en tEMG requires high stim ulation,

it demonstrates the integrity of the pedicle

cortex and lack of perforation. Direct stimula-

tion of a m isplaced pedicle h ole with a breach,

can activate th e adjacent ner ve root and evoke

a CMAP in t he appr opriate m yotom es at lower

stimulus intensities than would be expected

with an imperforate pedicle cortex. Clinical

correlation is of course re quired , but t EMG at-

tempts to p rovide data on a pathway from p ed-

icle screw or tract to the distal site an d can be

used in thoracic spine operations if the rectusabdominis or intercostals musculature are

m onitored as th e distal recording site.




! Multimodality

Intraoperative Monitoring

Somatosensory evoked potentials are the

m ost comm on neurom onitoring modality em-

ploye d, bu t a re not always a su cient p roxy for

all cord function. Failing to recognize the lim-

itation s of SSEPs can lead t o devast ating conse -

quen ces. It m ust be em pha sized th at no single

m odality su cient ly m onitors all spinal cord

pat hways . If the goal of in t raop era tive neuro -

m onitoring is to detect the onset of de cits for

bot h se nso ry and m ot or pat hways, then no

single modality meets the goal; however, a

combination of testing met hods m ight. Multi-

modality intraoperative monitoring uses allelectrophysiological techniques and can pro-

vide intraop erative inform ation about the n eu-

ral structu res at risk. It per m its assessm ent of

bot h asce nding and descending p at hways con-

curren tly, providing a cert ain degree of redu n-

dancy because many types of intraoperative

injuries will compromise both motor and sen-

sory pathways.

! Intraoperative Neurologic

Complications

Signi cant ASD surgery requires continuous

neu rom onitoring, especially dur ing placeme nt

of instrumentation and deformity correction.

Imm ediate action is required whe n dam age to

the spinal cord or a peripheral nerve is sus-

pect ed at any t im e during the pro cedure in

response to changes of > 50% am plitu de and

> 10% laten cy in t he SSEP/MEP signa ls. An

algorithm m ay aid the p rimar y surgeon in de-

termining the causative factor and initiating

appropriate treatment. Reassessment of neu-

romonitoring signal strength is performed after

each step. The speci c timing of each of the

steps listed below is not un iversal; rathe r, tim -

ing should be determined on a case-by-case

basis. Each subs equ ent st ep is in it iat ed if the

pat ient fails to dem onst rat e im prove m ent inneu rologic function after the previous sequen-

tial corrective maneu vers have been p erformed.

Here is a general checklist of factors to con-

sider w hen SCM changes occur:

Ten- Item Che cklist in Respo nse to Losing

SCM Data or Meeting Warning Criteria

(SSEPs or Neurogenic MEPs)

1. Check with personne l to ma ke cert ain SCM data

issue is real (experience matters).

2. Be aware that an increase in blood pressure

(MAP " 80–90 mm Hg/systolic blood pressure

> 120 mm Hg) may require a quick dose of epi-

nephrine/norepinephrine or a dopamine drip;

provide blood pro duct s if ne ed ed (he mog lobin

# 9).

3. Release any tract ion on pat ient’s spinal column

(halo, halo-fem oral, etc.).

4. Palpate the dura, checking for impingem ent (ifspinal canal is open), such as prior oste otomies/

laminectomies.

5. Reverse any correct ive maneuver; also consider

shortening of spinal column.

6. Con rm the absence of spinal subluxation.

Consider using temporary bilateral rods during

closure.

7. Consider implant malposition (screw/hook/wire)

if temporally related, which might indicate dural

impingement.

8. Order a wake-up test if the monitoring data

have not improved or reached baseline. Also,

this is a good time to take a deep breath and

re ect on possible add itiona l issues.

9. Con rm that elevated MAP is being maintained.

10. Consider apical spinal cord decompression to

relieve tight neu ral tissue.

Increase Spinal Cord Perfusion

Im m ediately following iden ti cation of neu ro-

monitoring signal changes, the hemodynamic

and oxygenation status of the patient should be op t im ize d to im prove perfusio n to the sp i-

nal cord. The MAP is elevated t o > 80 m m Hg or

20% above baseline.17 Hemoglobin and blood

glucose levels are evaluated and corrected if

required. Body temperature should be main-

tained at > 36.5°C to opt imize ne urom onitoring.

These measures have been shown to increase

spinal cord per fusion.18

Stagnara Wake-Up Test

Changes in neuromonitoring signal suggestive

of persistent n eurologic de cit m ay be cause



74 Chapte r 6

for considering a con rmatory test. Prior to

the induction of anesthesia, patients should be

counseled t hat t hey w ill be asked to per form

several commands upon awaking from anes-

thesia. Frequent assessment of the patient’s

neuromonitoring status is recommended asinstrum entat ion is placed and corrective ma-

neuvers per form ed. This enables the surgeon

to most reliably pinpoint the possible incit-

ing factor, such as a malpositioned implant

or tension on the cord caused by corrective

maneuvers. Thus, potentially problematic in-

strum entation can be rem oved or correction

relaxed while waiting for the patient to

awaken from an esthesia for the wake-up test.

However, one must also understand that a

time lag can occur bet ween the application of

a corrective force (e.g., distraction) and the

dim inut ion/loss of SCM signals. Thu s, pru den t

responses to any SCM change that meets the

warning criteria must be made based on the

many factors involved, including the MAP,

any recent correction m aneuvers, and th e type

of data .

Release of Correction

The surgeon shou ld consider releasing the ten-

sion on the spine when all parameters (e.g.,

MAP, temperat ure, he m oglobin level) have been

reasonably addressed and there is still an ab-

sence of expected motor function with the

wake-up test. After release of correction, a

second wake-up test can be considered w hen

SSEP/MEP signals ind icate pe rsisten t abnor m al-

ity. If an improvemen t in th e wake-u p test or

in ne urom onitoring is discovered after t he re-

lease of surgical correction, the surgeon has

the option of fusing the spine in situ or at-

tempt ing a m ore modest correction. When no

improveme nt in neu rologic funct ion is elicited

after the release of tension, all screws and

hooks shou ld be rea ssessed. The stability of the

spine also shou ld be evaluated . Wh en re m oval

of instrum entation would compromise the sta-

bilit y of the sp inal colum n, su ch as a fter verte-

bra l bod y resect ion, th e surgeon m ay be forced

to maintain the existing instrumentation andfuse the spine un der the least amoun t of ten-

sion. If osteotom ies have bee n p erforme d, the

canal should be exam ined for fragments of bone,

Gelfoam , or bone wax, wh ich m ay be contrib-

uting to cord comp ression.

Pedicle screw position should be critically

exam ined in light of a mon itoring chan ge. The

pos it ion of each screw can be re assessed usingone m easure or a com bination of several. High

stimu lation th resholds of each xation point,

as indicated by triggered electromyography,

th eoret ically indicate intracort ical screw p osi-

tion secondary to increased resistance to cur-

rent ow through cort ical bone. Any ped icle

screw w ith a m arkedly lower electromyogra-

phy th resh old (< 60%) in relat ion to the rest

of the construct should be reassessed, as this

m ay ind icate a p ossible ped icle wall breach.19

Screw position can also be evaluated with the

use of intraop erat ive uoroscopy. A ped icle

screw tip past the midline of the vertebral

bod y not ed on posteroa nterior radiographs is

suggestive of a medial pedicle breach. In the

pre se nce of any o r all of thes e signs, t he screw

may be removed to reassess the tract with

direct palpation. A small laminotomy may

also be performed to evaluate the integrity

of the medial pedicle cortex with or without

screw removal. Early removal of instrumen-tation m ay increase th e possibility of neu ro-

logic improvement, provided the spine will

not be signi cantly destabilized with rem oval

of instrum entation.

The ability to obtain adequ ate postope rative

imaging studies is one potential advantage

of removal of instrumentation. The quality of

computed tomography (CT) and MRI scans is

superior when no instrumentation is present

to create ar tifact. Even titan ium construct s can

pro duce ar t ifac t on CT or MRI scans. An MRI

scan may be done in the p resence of titanium

instrumentation; otherwise, a CT scan can be

ordered . If the instru m ent ation is retained and

stan dard CT or MRI scanning is inconclusive, a

CT m yelogram can be p erform ed. If an ab nor -

mality (e.g., screw malposition, hematoma)

is identi ed, urgent return to the operating

room is indicated for decompression or re-

m oval of instrum ent ation. If su cient im aging

can be performe d and there is no identi ablesite of compression, close patient observation

is adequ ate.




! Steroid Protocol

Although its u se is debated, m ethylpredniso-

lone is currently the only recognized pharma-

cologic inter vention for the treat m ent of acute

spinal cord injury (SCI).13 The use of steroidshas not been extensively studied in the intra-

operative setting; however, we currently ad-

minister steroids to patients who have a

continued negative w ake-up test (i.e., absence

of motor fun ction) after release of tension from

corrective an d distractive man euvers. The cu r-

rent re comm end ed protocol is a loading int ra-

venous bolus dose of 30 mg/kg administered

over 15 minutes, followed by 5.4 mg/kg/h as

a 23-hour infusion (if started within 3 hoursfrom the time of injury).13 The u se of met hyl-

predniso lone for the m anage m ent of in t ra op -

erative SCI is not well documented. Thus, the

surgeon mu st weigh the potential bene ts of

improved ne urologic recover y against t he pos-

sible increased risk of infection. The Am erican

Association of Neu rological Surgeon s/Congress

of Neurological Surgeons (AANS/CNS) Joint Sec-

tion on Disorders of the Spine and Peripheral

Nerves Guidelines Com m it tee has indicated

that methylprednisolone for either 24 or 48hours is an option in th e treatm ent of patients

w ith acute SCIs that shou ld be und ert aken only

w ith the kn owledge that t he evidence suggest-

ing harm ful side e ects is more consistent th an

any suggestion of clinical ben e t.

Intravenous lidocaine (2 mg/kg) for vasodi-

latation has been described for treatment of a

pos tulat ed ischem ic sp inal cord after segm en-

tal vessel ligation.20 In experimental animal

models, intrathecal and intravenous vasodila-tors have been shown to enhance spinal cord

perfusio n and neuronal protect ion. How eve r,

we h ave no clinical experience w ith this m edi-

cation and thus cannot comment speci cally

on its u sefulness.

! Postoperative Manageme nt

A patient w ith an intraope rative ne urologic in-sult should be admitted to the intensive care

unit postoperatively for close monitoring of

hem odynamic parameters as w ell as for neuro-

logic examinations. MAP must be maintained

at > 80 m m Hg w ith the judicious use of intra-

venous uid replacem ent , blood tran sfusion (if

indicated), or vasopressors wh en necessary to

m aintain cord per fusion. A $-agonist (e.g., dopam ine) can be u sed to m ain tain m ean ar ter ial

blood pressu re if uid re placem ent alon e is in-

su cient . A neu rologic examinat ion should be

perform ed an d docum ented every h ou r for t he

rst 12 to 24 hours. This may pose a problem if

the p atient rem ains intubated and sedated. In

this case, it is paramount that the patient be

lighten ed from sedation on an h ourly basis for

e ective assessmen t of neu rologic function.

! Delayed Postoperative

Neurologic Complications

Neu rologic complica t ions in the postop erat ive

per iod sh ou ld be m anaged w it h the sa m e dil-

igence and meticulous care as described for

an intraoperative SCI. Although relatively un-

common, delayed postoperative SCI may be

attr ibuted to progressive spinal cord ischem iasecondary to traction or to the developm ent of

an epidural hem atoma.

As w ith an y acute SCI, ade quat e pe rfusion of

the spinal cord is paramount. Blood pressure

should be meticulously monitored, and MAP

should be m aintained at > 80 mm Hg in an ef-

fort to sust ain spinal cord pe rfusion. Vasopres-

sors (e.g., dopam ine) m ay be required to attain

adequate blood pressure and cord perfusion.

Hemoglobin levels should also be checked to

avoid excessive postoperative anemia. Patient

temp erature should be maintained above 36.5°C.

A steroid protocol m ay be initiated as indicated

above for the patient with continued neuro-

logic loss.

Obtaining imaging studies before retu rning

the patient to the operating room may aid in

delineat ing the cause of the de cit. This will

enable the surgeon to plan th e proper course of

action, whet her that involves reexploration for

localized decomp ression of an evolving epiduralhem atom a, release of correction, or rem oval of

instrumentation to correct spinal cord isch-



76 Chapte r 6

emia secondary to excessive tensioning. Reim-

aging with CT or MRI scans can be omitted if

obtaining these studies would result in a sub-

stantial time delay. Early decompression may

improve neurologic outcome for the patient

with new -onset neurologic de cit in the acute pos top erat ive period. Con versely, if no abnor -

m ality is iden ti ed on CT or MRI scan, the p a-

tient m ay be observed closely with sup port ive

treatment.

! Chapter Summary

The surgical treatm ent of comp lex adult spi-

nal deform ity has advanced signi cantly in re-cent times with the use of pedicle screw-based

instrumentation and the increasing role of

complex three-colum n osteotomies to optimize

deformity correction. The correction of large

magnitude coronal and sagittal plane defor-

m ity is becoming more com m on. However, th e

technical demands involved in restoration of

spinopelvic alignm ent and sagitt al and coronal

balance in large-m agnitude deform it ies has a

signi cant risk of neu rologic complications, w ith

pot ent ially devast at ing clin ica l an d funct ion al

sequelae. It is important that spinal surgeons

use an algorithm for the safe and e ective m an-

agement of neu rologic sequelae associated w ith

ASD surgery so as to opt imize patient m anage-

m ent an d functional outcome.

Pearls

Neurologic safety during spinal de form ity sur-

gery requires preoperative p reparation, intraop -

erative multimodality spinal cord m onitoring, and

po stop era tive d iligence .

A com bina tion of SSEP, MEP, and EMG mon itor-

ing is necessary for comprehensive assessment

and evaluation of neurologic function during ASD

surgery.

Appropriate responses to any loss of degradation

of SCM data and to warning criteria should in-

clude a spectrum of responses aime d at optimiz-

ing spinal cord b lood supply and m inimizing a ny

direct or indirect tension or pressure on the neu-

ral element s.

To con rm the patient ’s neurologic integrity, one

must always perform a detailed motor exam of

the lower extremities at the end of the surgical pro cedure be fore th e pa tient is extuba te d and

leaves the operating room .

Pitfalls

One must ident ify tho se patients who are at high

risk of neurologic complications during ASD sur-

ger y, including t hose with p reexisting neuro logic

abnormalities, an abnormal neural axis on MRI

exam, or large and sti kyphot ic deformities.

Not resp onding to or trusting t he SCM pe rsonnel

and da ta in a timely fashion can have devastat ing

neurologic sequelae. One must be careful with those patients kept

intubated/sedated following extensive ASD sur-

gery in order not to miss a delayed neurologic

complication due to the inability to obtain an

adequ ate ne urologic exam o n a frequent b asis.

References


1 . Dorman s JP. Establishing a standard of care for neu-

rom onitoring dur ing spinal deformity sur gery. Spine

2010;35:2180–2185 PubMed

2 . Malhot ra NR, Sha rey CI. Intra ope rat ive electr oph ys-

iological monitoring in spine surgery. Spine 2010;

35:2167–2179 PubMed

3. Gelalis ID, Pasch os NK, Pakos EE, et al. Accur acy o f

ped icle s crew placem en t: a sys te m at ic re view o f pro-

spective in vivo stud ies compar ing free han d, uoro-

scopy guidance and navigation te chniques. Eur Spine

J 2012;21 :247–25 5 PubMed

4. Tian NF, Hua ng QS, Zho u P, et al. Ped icle screw in ser -

tion accuracy with di erent assisted met hods: a sys-

tematic review and meta-analysis of comparative

stud ies. Eur Spine J 2011;20:846 –859 PubMed

5. Daubs MD, Len ke LG, Cheh G, Stob bs G, Brid we ll KH.

Adult spinal deformity surgery: complications and

outcomes in patients over age 60. Spine 2007;32:

2238–2244 PubMed

6. Kim YB, Len ke LG, Kim YJ, et a l. The m orbid it y of

an anterior thoracolumbar approach: adult spinal

deform ity patients with greater than ve-year fol-

low-up. Spine 2009;34:822–826 PubMed

7. Kim YJ, Brid well KH, Len ke LG, Cheh G, Baldu s C. Re-

sults of lumbar pedicle subtraction osteotomies for

xed sagittal im balance: a m inimum 5-year follow-up

study. Spine 2007;32:2189–2197 PubMed

8. Lap p MA, Brid we ll KH, Len ke LG, et al. Long-

term complications in adult spinal deformity pa-

tients having combined surgery a comparison of




p r im ar y to rev ision pat ien ts. Spine 200 1; 26: 97 3–

983 PubMed

9. Rhee JM, Bridwell KH, Lenke LG, et al. Staged poste-

rior sur gery for severe adult spinal deform ity. Spine

2003;28:2116–2121 PubMed

10 . Bridw ell KH, Lew is SJ, Len ke LG, Bald us C, Blanke K.

Pedicle subtraction osteotomy for the treatment of

xed sagittal imbalance. J Bone Joint Surg Am 2003;

85-A:454–463 PubMed

11 . Buch owski JM, Brid we ll KH, Len ke LG, et al. Neuro-

logic complications of lumbar pedicle subtraction

osteotomy: a 10-year assessment. Spine 2007;32:

2245–2252 PubMed

12 . Len ke LG, Feh lings MG, Sha rey CI, et al. Prosp ect ive,

multicenter assessment of acute neurologic compli-

cations following complex adult spinal deform ity sur-

gery: t he Scoli-Risk-1 trial. Spine 2014 subm itted

13. Winter RB. Neurologic safety in spinal deform ity sur-

gery. Spine 1997; 22:1527 –1533 PubMed

14. Owen JH. The application of intr aoperat ive m onitor-

ing during su rgery for spinal deform ity. Spine 1 999;

24:2649–2662 PubMed

15. Vauzelle C, Stagn ara P, Jouvin rou x P. Fun ction al m on-

itoring of spinal cord activity dur ing spinal surger y.

Clin Orthop Relat Res 19 73;93:17 3–178 PubMed

16. Lesser RP, Raud zen s P, Lüd ers H, et al. Postope rat ive

neurological de cits may occur despite unchanged

intraoperative som atosensory evoked poten tials. Ann

Neuro l 19 86 ;1 9: 22 –2 5 PubMed

17 . Naslund TC, Hollier LH, Mon ey SR, Facu nd us EC,

Skenderis BS II. Protecting the ischemic spinal cord

dur ing aort ic clamp ing. The in uen ce of anest hetics

and hypot her m ia. Ann Surg 1992;215:40 9–415, dis-

cussion 415–416 PubMed

18. Raynor BL, Len ke LG, Kim Y, et al. Can tr iggere d elec-

trom yograph thresh olds predict safe thoracic pedicle

screw placement? Spine 2002;27:2030–2035 PubMed

19. Bracken MB, She par d MJ, Holford TR, et a l. Adm ini-

stration of methylprednisolone for 24 or 48 hour s or

tirilazad m esylate for 48 h ours in th e treatm ent of

acute spinal cord injury. Results of t he Third National

Acute Spinal Cord Injury Randomized Controlled

Trial. National Acute Spinal Cord Injury Study. JAMA

1997;277:1597–1604 PubMed

20 . Klem m e WR, Bur khalte r W, Polly DW Jr, Dahl LF,

Davis DA. Reversible ischemic myelopathy during

scoliosis surgery: a possible role for intravenous li-

docaine. J Pediatr Orth op 1999; 19:763– 765 PubMed



! General Introduction of

Adult Scoliosis

Adult scoliosis is de ned a s an abnorm al defor-

m ity in an a dult, with Cobb angle greater than

10 degrees in the coronal plane, with or with-

out sagittal imbalance or ab norm al pelvic ori-

entation.1 In the elderly population, a variety

of prevalence rates have been repor ted as a re-sult of di erences in de nitions of scoliosis,

sample size, ethnicity, and screening tools. In

a study of volunteers who were over 60 years

of age, Schwab et al2 found that 68% of the

subject s met t he de nition of scoliosis. As re-

por ted by Xu et al ,3 the prevalence of scoliosis

was 13.3%in a cohort of 2395 adu lts older t han

40 years o f age.

Adult scoliosis may stem from progression

of scoliosis during childhood or adolescence

(Fig. 7 .1), or m ay be new ly developed in adu lt-

hood through degenerative changes (Fig. 7.2).

The former condition is referred to adult idio-

pat h ic scoliosis, w hereas the lat ter is term ed

degen erat ive scoliosis or de novo scoliosis. In

contrast to adolescents, adults with scoliosis

characteristically present w ith back pain, ra-

diculopat hy, an d ne urogen ic claud ication. Cos-

m esis is a concern of some young adult scoliosis

pat ients. They oft en complain of waist asym -

metry and ribs abutting the pelvis, as a resultof im balance in the coronal plane or th e sagit-

tal plane.

For ad ult scoliosis, non ope rat ive care is usu -

ally the rst-line treat ment option. Nevert heless,

surgery may be inevitable when nonoperative

m easures fail. The prim ary ind ications for sur-

gery of adu lt scoliosis are (1) progressive defor-

m ity, (2) poor spinal balance causing funct ional

di culties, (3) large deform ity th reate ning

cardiopulmonary compromise, (4) neurologic

m anifestations, (5) persistent pain t hat fails to

respond to nonoperative treatm ent, and (6) un-acceptable cosmetic appearance.1,4–6 Bess et al, 7

in a multicenter review of 290 patients with

adult scoliosis, reported that operative treat-

ment for older patients was primarily driven

by pain and disabilit y, independent of rad io-

graphic measurements, and, for younger pa-

tients, by increased coronal plane deformity.

Although ope rative m anagem ent of adult scoli-

osis is a growing ch allenge, a variety of sur gical

options h as been em ployed, include p osterior,

anterior, or combined approaches. Silva and

Lenke6 proposed six distinct levels of surgical

options for adult degenerative scoliosis: I, de-

compression alone; II, decompression and

limited instrumented posterior spinal fusion;

III, decompression and lumbar curve instru-

m ente d fusion; IV, decomp ression w ith an te-

rior and posterior spinal instrumented fusion;

V, thoracic instrum ent ation and fusion exten -

sion; and VI, inclusion of osteotomies for spe-

ci c deform ities.Fusion levels should start proximally at a

stable vertebra, typically above T6 or below

7

Postoperative CoronalDecompensation in Adult Deformity

Yong Qiu



Postoperative Coronal Decompensation in Adult Deformity 79

T10, and end distally at a neutral and stable

vertebra. The proxim al level never start s at t hethora cic kyph osis apex, avoiding proxima l junc-

tional kyphosis, wh ereas t he distal level never

ends at a level with rotatory subluxation. The

decision of whether to include L5 or the sa-

crum in the fusion is controversial.8,9 Fusion

distally to L5 o ers the th eoretical bene ts of

preserved lum bosacral m ot ion, sh or ter su rgi-

cal time, and a decreased likelihood of pseu-

darthrosis; on the other hand, it carries the

potential for accele rate d sym pt om at ic advan ced

degeneration at the L5/S1 disk, which in turn

ultimately results in axial discomfort, radicu-

lopathy, and loss of lumbosacral lordosis. In

contrast to L5, fusions extende d to th e sacrum

achieve a higher stability of xation as well as a

be t te r cor rect ion of sagit tal im ba lance, but this

procedure also runs the risk of an increasin g

chance of pseudarthrosis, a greater frequency

of major complications, and a higher rate of

instrum entation failure.8,9 A recent study a lso

found fusion to th e sacrum to be one of the riskfactors of proximal junctional kyph osis.10 Non-

controversial indications for fusion to the sa-

crum include the following6: (1) an oblique

take-o of L5 on the sacrum , (2) a lum bosacralfractional cur ve > 15 degrees, (3) advanced de -

gene ration of the L5/S1 disk or th e L5/S1 facet

join ts, (4 ) L5/S1 sp on dylolisthesis, and (5) p rior

history of decompression at this segment.

Wh en fusion to the sacrum cannot be avoided,

it is important to perform an interbody fusion

be tween L5 and S1 to decrease the risk of a

nonunion.

! Di erentiating Between

Degenerative and

Idiopathic Scoliosis

An essent ial prem ise of the treat m ent of spinal

deformity in particular is understanding its

etiology. Aebi1 developed in 2005 a classi ca-

tion for adult scoliosis based on the etiology:

type 1, de novo scoliosis; type 2, progressive

idiopath ic scoliosis; t ype 3a, secondary degen-erative scoliosis, due to a preexisting condition,

either intrinsic (adjacent curve) or extrinsic

Fig. 7.1 a–e (a,b) A 57-year-old wom an with a

history of ado lescent idiopat hic scoliosis. (c–e) Both

coronal and global sagitt al balance was well main-

tained, and t he L2/3 d isk height was much bet ter

preserved on the convex side t han on the concave

side.

a b c e

d



80 Chapte r 7

Fig. 7.2 a–d Adult de generat ive scoliosis in a

65-year-old woman. (a) Stand ing X-ray lms showed

a right-sided lumbar curve, rotatory subluxation at

L3/4, and reg ional kyphosis from L3 to L5 as well as

sagittal imbalance. (b) Magnetic resonance imaging

of advanced disk degene ration showed dark disks

with narrowed disk height . (c,d) Computed

tomography scans demonstrated the vacuum

phenomenon as well as canal stenosis in the lowe r

lumbar region.

a b

c d




(lower limb length discrepancy) to the spine;

and type 3b, scoliosis secondary to metabolic

bon e disease . How ever, it is di cu lt to di er-

ent iate t he clinical etiology in elderly patient s

who are unable to provide either a medical

history or information about their spinal de-formity. As there is a paucity of information

in th e literature, this chapter d escribes several

experience-based radiographic criteria t hat may

help to di erentiate between d e novo and idio-

pat hic scoliosis .

Apical Disk Height

In de novo scoliosis, asymmetrical disk de-

generation has been regarded as the initiatingfactor that contributes to the occurrence of

degenerat ive scoliosis. Asymm etric degenera -

tion in the apical disks leads to asymmetric

disk collapse, ind ucing a cur vature by pivoting

on the apical facet joint at the concave side,

wh ich in turn exacerbates more degeneration

of the disks on the concave side than on the

convex side and then start the vicious circle.

Bao et al11 also found that regional lum bar disk

degenerat ion correlated w ith th e coronal Cobb

angle, con rm ing that asymm etric disk degen-eration contributed to th e developm ent of de

novo scoliosis. We observed that the convex

disk height w as signi cant ly less in de novo

scoliosis (Fig. 7.2) than in idiopathic scoliosis

(Fig. 7 .1).

Curve Pattern

The m ajority of degen erat ive scoliosis a ect s

the lumbar or t horacolumbar spine, and t heircurve pattern s may be di erent from th at of

idiopathic scoliosis. Because the original patho-

genesis of de n ovo scoliosis is th e degen erat ion

of disk and facet joints, the ap ex of the lum bar

curve is often located at the intervertebral

space. The m ost comm on ap ex of de novo sco-

liosis is the inter verteb ral space of L2/3 or L3/4,

often with a shorter curve span. In addition,

the levels involved in the degenerative curve

are gene rally thre e to four levels, wherea s four

to six levels are more common in idiopathic

curves.

Regularity of Apical Vertebra

In add ition to cur ve patter ns, the apical verte - br a in de novo sco liosis is ofte n irre gu larly

wedged. Osteophytes, end-plate abruption, and

osteoporotic minor fracture are comm only seen

in degenerat ive vertebrae, so the sh ape of ver-

tebra m ay not be regularly trap ezoid. In con-

trast, wedging of apical vertebra, if any, is

usu ally regular in idiopath ic lum bar scoliosis.

Compensatory Curve Above the

Main Curve

Regular compensatory curves proximal to the

m ain cur ve in idiopath ic scoliosis m ay develop

during adolescence, and serve as a way to re-

ba lan ce the dist al thoraco lum bar/ lum bar curve

in the coronal plane (Fig. 7.1). That explains

why global coronal imbalance is less frequen t

in adolescent idiopathic scoliosis cases with

double curve patte rns, owing to the compensa-

tory curve and its compensatory ability from

the diskal, pelvic, and certainly spinal muscu-

lar structure as well. This balancing pattern

m ay continu e into adulthood and last for a long

time. This featu re of compensatory curves could

serve as an imp ortan t rad iograph ic sign to dif-

feren tiate these tw o entities.

Patient s with de n ovo scoliosis often presen t

early w ith coronal or sagittal imbalance due to

the less e ective compensative curves above

the imbalance. In contr ast w ith sagitt al im bal-

ance, there is paucity of informat ion in the lit-eratu re on th e inciden ce of coronal im balance

in de novo scoliosis. Based on our study,11

about one third of patients with d e novo sco-

liosis may presen t w ith coronal im balance due

to a lack of compensatory curves. Similar to

the functional scoliosis seen in young individ-

uals with disk herniation or other lower back

diseases, tr unk shifting is not un comm on in de

novo scoliosis with sten osis because of the p ain-

alleviating mechanism. This trunk shifting or



82 Chapte r 7

coronal imbalance with spinal curve may be-

com e structural with time.

Correlation Betw een the Cobb

Angle and Imbalance

The discrepancy between the Cobb angle and

imbalance in de novo scoliosis is an interest-

ing nd ing. Degen erat ive cur ves always are

shorter th an t he idiopathic curves and have a

smaller Cobb angle. As mentioned earlier, cor-

onal imbalance is an important feature of de

novo scoliosis, whatever the Cobb an gle, whereas

signi cant imba lance is mostly witnessed in

adult idiopathic scoliosis with severe curves.

The refore, a sm all Cobb an gle with obvious cor-

onal imbalance indicates that the curve might

be degenerat ive (Fig. 7 .2).

Correction Ability o f Rotatory

Subluxation

Ver tebral rotat ory su bluxat ion (VRS) is a tr iax-

ial deform ity, pre dom inan tly at t he L3/L4 level,

w ith fem ale predom inance. Although it is m ore

likely to occur in patients with de novo scolio-

sis at the early stage of th e d eform ity, it couldalso occur du ring the late course of adult idio-

pat hic sco liosis . At ou r center, we fou nd that

the correct ability of VRS under tract ion or on

side-bend ing lms could be used to di erent i-

ate de novo scoliosis and idiopath ic scoliosis. In

the form er, redu ction of VRS und er t raction or

on side bending may not be achieved because

of the rigidity from vertebral degeneration at

the subluxation level, including disk collapse,

osteophytes, and spont aneou s vert ebral or facet

fusion. In contrast, in idiopathic scoliosis, VRS

could be partially reduced un der tr action or on

side bending.

Discrepancy Between the Cobb

Angle and Rotatory Subluxation

At our center w e n oted t hat VRS was m ore like

to occur wh en t he Cobb angle increased in cases

of idiopathic scoliosis. However, in de novo

scoliosis, VRS may occur at its ea rly stage, and

its onset and severity may not necessarily be

correlated w ith th e Cobb angle. In ot her words,

VRS in de novo scoliosis may develop even in

cases with a sm all curve.

The Origin o f Stenosis

According to th e de nition of de novo scoliosis,

its primar y cause is the degenerat ion of spine,

including disks, the m uscle–ligament s comp lex,

and the facet joint. Lumbar stenosis is more

comm only seen in p rima ry degene rative scoli-

osis than in adult idiopathic curves. Therefore,

radicular leg pain and claudication should be

m ore comm on in de novo scoliosis, even w ith

small curves. This is in accordance with our

clinical observation th at m echanical back pain

is the m ost comm on complaint in m any adult

idiopath ic scoliosis patients d ue to deformit y-

induced paraspinal muscle fatigue, whereas

neu rogenic back pain in comb ination with leg

pain is t he m ost com m on complain t in de n ovo

scoliosis patien ts (Fig. 7.2 ).

Lum bar Lordosis

In addition to di eren t curve presen tations in

the coronal plane, sagitt al alignm ent m ay also

be di ere nt be tween de novo and idiopat hic

scoliosis, especially in th e early stages. Lum bar

lordosis m ay rema in norm al in idiopathic sco-

liosis because disk height may be maintained

for a long time. In de novo scoliosis, however,

lumb ar lordosis m ay not be preserved because

of early disk collapse. Bao et al11 also dem on-

strated t hat d e novo scoliosis patient s with se-

vere disk degeneration have lumbar hypolordosis

or kyphosis (Fig. 7.2). Moreover, osteoporotic

fracture is more frequen tly observed in de novo

scoliosis, particularly in female patients, greatly

contributing to lumbar kyphosis, whereas in

idiopathic scoliosis, the degenerative patholo-

gies are not the primary cause, and osteopo-

rotic fractu re m ay be less com m on.




! Contribution of Disk

Degeneration to Spinal

Imbalance and Curve

Severity

Spine imbalance in the sagittal or coronal

plane has an im por tant im pact on the health

status and treatment options in patients with

de novo scoliosis. Sagittal im balan ce is closely

correlated with poor health-related quality of

life (HRQOL). Import an tly, it h as been w ell doc-

um ente d that coronal imbalance is also one of

the m ain causes of unsat isfactory app earan ce,

impaired function, and back pain.12 Because

the established consensus in terms of the ori-

gin of de novo scoliosis is that it is triggered

by asym m et rica l disk degenerat ion, ou r team

conducted a stu dy speci cally focused on th e

correlation between disk degeneration and

spinal imbalance.11

We quanti ed disk degener ation using the

P rrm ann classi cation, wh ich describes ve

grades of disk degene ration on m agnetic reso-

nan ce imaging.13 Each grad e of disk was scored

w ith a speci c num ber to enable doing calcu-

lations; for exam ple, grad e I was given a scoreof 5, whereas grade V was given a score of 1.

Thus, higher scores represent ed h ealthier disk

conditions.

The results of our study revealed that disk

degeneration at the lower end vertebra (EV)

was strongly correlated with sagittal imbal-

ance (Fig. 7.2). We found that the grad e of the

lower EV disk reached a mean degeneration

score of 2.32, being the second m ost severely

degenerated disk after the apical disk. There

m ay be th ree stages of disk degenerat ion, cor-

related w ith stability an d m otion: dysfunct ion,

instability, and stabilization. With moderate

disk degenerat ion, the disk might be come u n-

stable. Also, there m ight b e a tend ency for in-

stability to lie in m oderately degenerated disks

with well-preserved disk height, whereas

mobility may decrease and restabilize in the

collapsed disks. This nding supp orted our as-

sum ption t hat lower EV disk degeneration w as

more responsible for the sagittal imbalance

because it s st abilit y was jeopardized. How-

ever, we failed to nd signi can t correlation

be tween coro nal im ba lance an d disk degener-ation. Cert ainly, degenerat ion of th e posterior

elements, including the facet joints and the

paraspinal m uscle, is anot her accepte d factor

accounting for de novo scoliosis; therefore, it

is assumed that unstable posterior elements

instead of disk degeneration may be the im-

por tan t cause of coro nal im balance in lum ba r

degenerative scoliosis. The degenerative facet

join ts w ith os teoa rthrit is m ay be the prim ary

cause, or may be se condary to t he loss of disk

height, leading to vertebral instability an d in-

creased segmental axial mobility, which may

contribute to coronal imbalance. Asymmetric

atrophy of paraspinal muscles is another pos-

sible factor in uencing coronal balance. The

degree of instability varies in each individual,

ba sed on the slip in the sagit tal plane, t ra ns-

lational dislocations in th e coronal plane, and

three-dimensional rotational subluxation.

Correlation between the Cobb angle and

apical disk degeneration was also noted. Them ore degenerat ion the ap ical disk presented,

th e larger is the Cobb an gle. Such a close rela-

tionship between degeneration of the apical

disk and the Cobb angle can be explained by

the path ology of degene rative scoliosis: asym -

m etric degenerat ion in the apical disk will lead

to asym m etr ic disk collapse, inducing the spine

to bend the apical facet joints, which in turn

exacerbates the degeneration of the concave

side. In addition, we also found that regional

lumbar disk degeneration grade is correlated

with sagitt al malalignm ent , including an an te-

verted C7PL and lumbar kyphosis (Fig. 7.2).

Decreases in lumbar lordosis in patients with

disk degeneration, as evidenced in our study,

explain w hy de n ovo scoliosis patients w ith se-

verely degenerate d disks had lum bar hypolor-

dosis or kyph osis.



84 Chapte r 7

! Coronal Balance of

Adult Deformity

Coronal Balance Assessment

The concept of balance of the spine has beenextensively described by the Scoliosis Research

Society (SRS). The concept imp lies th at, in bot h

the coronal and sagitt al planes, the hea d is po-

sitioned correctly over th e sacrum and pelvis, in

bot h a t ran sla t ional and an angu lar sense . Fro m

the frontal view of the trunk, balance implies

horizontal shoulders and th e tru nk evenly dis-

tribute d about t he vertical line pa ssing through

th e cent er sacral vert ical line (CSVL). Spina l bal-

ance in the coronal plane can be determ ined as

the displacem ent of the m ost cephalad verte-

bra fro m the CSVL in bo th a dist an ce (frontal

plane o set) an d an angle (o set angle). In p rac-

tice, the de ned cephalad vertebra usually is

C7 or T1 (Fig. 7.3). Compen sation in th e coronal

plane is usu ally referred t o as the t ra nsla t ion o f

the midpoint of C7 in relation to CSVL (mea-

sured in the same man ner as coronal balance

[CB]). It prim arily describes t he position of the

head over the pelvis. Decompensation occurs

when this alignment strays from the midline

by m ore t han a thre sh old value speci ed by theinvestigators, usually repor ted as 2 cm.

The w ord balance implies a static alignm ent

in the standing (or unsupported seated) posi-

tion, whereas compensation and decompensa-

tion refer to the result of dynamic alignm ent .

In detail, compensation signi es the active

process of becom in g balanced, w hereas de-

compensation indicates a failure to achieve

balance , e sp ecia lly aft er an in terven t ion su ch

as surgery.

Relationship Betw een Coronal

Balance and Quality o f Life

In patients with adult scoliosis, the impact of

sagittal balance on clinical health status has

Fig. 7.3 a–c Examples of the classi cation of the

coronal balance pat te rn in adult scoliosis. (a) Type A

in a 64-year-old woman without obvious t runcal

asymmetry. (b) Type B in a 63-year-old woman with

the trunk shifting toward the concave side. (c) Type C

in a 61-year-old woman with the t runk shifting

toward the convex side.

a b c




been extensively discussed, w here as the im -

pac t of coron al balance on fu nct ion al ou tcom es

is less clear. In contrast to patients with idio-

pat h ic scoliosis, pat ien ts w it h degenerat ive

lumbar scoliosis have an increased likelihood

of imbalance in the coronal plane. Because ofasymm etrical degenerat ive changes and verte-

bra l wedging in the apica l region, coron al im -

balance is fre qu ently ob se rved. According to

the study by Daubs et al,14 13 of 85 (1 5%) adu lt

scoliosis patients with preoperative coronal

imbalance had worsening coronal balance of

m ore than 1 cm after surgery. This nding sug-

gests that th e inciden ce of postoperative imbal-

ance in the coronal plane is underest imated .

It ha s been found that coronal imbalance cor-

relates w ith signi cant clinical m anifestation s

such as p elvic obliquit y, sitting or sta nd ing im -

balance, as w ell as severe cosm et ic t runca l de-

form ity. Moreover, coronal im balan ce is one of

the m ain und erlying causes of the progression

of deformity, back pain, and functional com-

prom ise . Axia l pain usu ally derives fro m the

convexity of the curve, and leads to furth er de -

terioration of coronal imbalance. Radicular pa in

and neurogenic claudication mainly originate

from the compression on the concavity of thecurve, or from dynamic overstretching on the

convex side.1 Deterioration of these sym ptom s

runs in parallel with the increase in coronal

imbalance to som e exten t. To address th ese

prob lem s, it is im por tant to corre ct the pre op -

erative coronal imbalance.

To help elucidate the factors th at are m ost

crucial for im proved ou tcomes, several studies

have attem pted to correlate radiographic nd-

ings with clinical symptoms in adult scoliosis.

Glassm an et al15 repor ted that signi cant coro-

nal imbalance was associated with pain and

dysfunction in unoperated patients, and coro-

nal imbalance was not as critical a parameter

as sagittal imbalance in prediction of symp-

tom s. However, Daubs et al14 showed that sagit-

tal balance is the stron gest predictor of imp roved

functional outcom es in ad ult scoliosis patients.

They found that restoring sagittal balance in

pat ients w ith com bined coro nal an d sagit tal

imbalance seems to be the key to improvingthe functional outcomes. In terms of patients

w ith coronal imbalance alone, improveme nt in

coronal balance was a signi cant pre dictor of

improved surgical outcom es.14

In som e stu dies, coronal im balance has also

be en re por te d to lead to d ecreas ed HRQOL and

increased risk of implant failure in adult scoli-

osis patient s.15–17 In the study by Glassman etal,15 signi cant coronal imbalance of greater

than 4 cm w as associated w ith more p ain and

dysfun ction for unop erated pat ients but n ot for

operated patient s. Ploumis et al16 reported that

pat ients w ith coro nal im balance of great er

than 50 m m showed worse physical function

scores. Cho et al17 demonstrated that preoper-

ative coronal imbalance led to more implant

failures, requiring removal of the implant.

Therefore, improved postoperative coronal

balance sh ou ld be the goal in order to im prove

th e HRQOL as w ell as to r edu ce th e n eed for re -

vision surgery. At our center, postoperative

coronal imbalance was one of the factors that

contributed to implant failure.

Decom pensation in the

Coronal Plane

As mentioned above, decompensation in the

coronal plane implies dynamic malalignmentof the spine, com m only measu red a s CB (trans-

lation of th e cen ter of C7 in relat ion to CSVL)

beyon d a sp eci ed threshold value. In adoles-

cent idiopathic scoliosis patients, decompen-

sation was usually de ned as coronal imbalance

of more than 2 cm (measured in the same

manner as CB). In adult scoliosis patients, the

thresh old value of decomp ensation in th e cor-

onal plane varied am ong stud ies. Glassm an et

al15 reported the association between coronal

imbalance of greater than 4 cm and deteriora-

tion in clinical symptoms in nonoperated pa-

tients. Daubs et al14 and Ploumis et al 16 de ned

coronal imbalance as C7PL > 4 cm and > 5 cm

lateral to CSVL, respectively. In a recent study,

with emphasis on coronal imbalance in adult

spinal deform ity patients t reated w ith long fu-

sions, Ploum is et al18 also emp loyed a criterion

of 4 cm.

In a stu dy by th e SRS th at classi ed ad ult

scoliosis according to t he King/Moe and Len keclassi cations, coronal im balance was consid-

ered to be one of the global balance m odi ers.19



86 Chapte r 7

Im balance was considered to be presen t if the

C7PL was located > 3 cm t o th e r ight or left of

th e CSVL. We recen tly sur veyed a consecu tive

series of degen erat ive lum bar scoliosis pat ients

to evaluate coronal balance, with the balance

thre shold set at 3 cm . We found a high preva-lence of preoperative coronal imbalance in

adults with degenerative lumbar scoliosis. In

these patients, imbalance occurred either on

the concave side, namely C7 deviating toward

the concave side of the m ain curve w ith refer-

ence to CSVL (Fig. 7 .3b), or on th e convex side,

w ith C7 deviating toward th e convex side (Fig.

7.3c).

Postoperative CoronalDecompensation

Postoperative coronal decompensation is a major

complication in adult scoliosis.14,18 In ad dition

to the Daubs et al14 nding repor ted above (see

section Relationship Between Coronal Balance

an d Quality of Life), Ploum is et a l,18 in a cohort

of 54 adult patient s treated w ith long fusions,

found p ostoperat ive coronal decompen sation in

seven patients with preoperative coronal im-

balance and in fou r w ithou t , at six weeks p ost-operatively. At a m inimum 2-year follow-up , four

m ore patients w ithout initial im balance were

observed with coronal decompensation. The

authors reported that postoperative coronal

decompen sation was found in an increased num-

ber of adult spine deformity patien ts.18 But so far,

the underlying factors that predict postopera-

tive coronal decomp ensation rem ain unclear.

In theory, decreased compensation of the

segments above and below the fusion pre-disposes the patient to postoperative coronal

decompensation. Multiple deformity- and

surger y-related factors are probably associated

with the occurrence of postoperative coronal

decompensation.

Deformity-Relate d Facto rs

We found that the preoperative coronal im-

balance pat te rn plays an im por tant ro le in the

occurrence of postoperative coronal decomp en-sation. A curve with imbalance to the convex

side predisposes to further decompensation,

par t icu larly w hen osteotom ies of the posterior

elements, such as Smith-Petersen osteotomy

(SPO), or th rough t hre e colum ns, such as ped-

icle subtraction osteotomy (PSO), are under-

taken.6,20 For cases with imbalance to t he convex

side, compression maneuvers on the convexside at t he level(s) of the osteotom y, wh ich a re

perform ed to close the os teot om y gap, m ay

lead to further inclination of the trunk toward

the convex side. As in congenital thoracolum-

bar kyp hoscoliosis, we also not ice d that pa-

tient s with preop erative convex im balance had

a higher rate of postoperative coronal decom-

pensat ion after thre e-colum n osteotom ies.

At the same time, decreased compensation

above and below the instr um ent ation also play

an import ant role in the d evelopm ent of post-

operative coronal decompensation, because

the compensation potential of the unfused

segments comes mainly from the disks and

parave r teb ral m usculature. Hence , the m or e

degenerative changes the adjacent vertebrae

cephalad or caudal to the fusion levels mani-

fest, the worse the potential ability for these

unfused segments to compensate, resulting in

an increasing likelihood of postoperative coro-

nal d ecompen sation.

Surgery-Related Factors

Among the surgery-related factors that have

impact on the occurre nce of postoperat ive cor-

onal decompen sation, the lower instrume nted

vertebra (LIV) selection is of upmost impor-

tance. Ending LIV at a vertebra that cannot

becom e hor izontal during su rgery carr ies the

poten tial r isk of postop erat ive decom pen sat ion .

If th ere is a resid ual ob liquit y of LIV in t he cor-

onal plane, an inclination of the tr un k is bound

to occur, because the disk below LIV provides

lim ited range of m otion. As m ent ioned pr evi-

ously, we found that the disks of the lower

lumb ar region showed signi cant degenerative

changes. The physiological funct ion of these disks

is correspond ingly comp rom ised. Apparen tly, fu-

sion distally stopping at a vertebra th at cann ot

be com e hor izontal places the coro nal ba lance

pat te rn at r isk of decom pen sat ion aft er surger y.In addition, proper determination of the

upp er instru m ente d vertebra (UIV) can dim in-




ish the inciden ce of coronal decomp ensation.

A location of UIV below t he en d ver tebra of the

m ain curve can result in inclination of the fu-

sion with reference to the CSVL. If fusion ex-

tend s into the thoracic region, however, a m uch

m ore cepha lad location of the UIV tha n th e en dvertebra also carries the risk of decompensa-

tion, because it lowers the n um ber of thoracic

vertebrae with compensation potential. More-

over, inclination of UIV to the convex side of

the main curve probably impedes balance in

the coronal plane through the adverse impact

on the auto-compensation mechanism.

An inapp ropriate osteotomy algorithm m ay

contribute to postoperative coronal decompen-

sation as well. Three -colum n osteotom ies usu-

ally begin at the apex and from the convex

side.20 This is e ective in the correction of

cases with p reoperative coronal imbalance to

the concave side. For a case with preexistent

imbalance to the convex side, however, such a

maneuver might aggravate the imbalance be-

cause of the comp ression forces at the osteoto-

m ized site from the convex side.

In practice, surgeons perform sagittal bal-

ance restoration more than in the coronal

p lane, and the balance pat tern and the com - pensa tion pot ent ial in the coron al p lane are

sometimes ignored. Such an attitude toward

coronal plane balance is evidently an under-

lying risk factor for postoperative coronal

decompensation. Multiple studies have dem on-

strated that coronal imbalance accompanied

by sagit tal im balance is a m or e com m on clin i-

cal scenario.14,16,19 Therefore, adequate atten-

tion need s to be paid n ot only to the sagittal

p lane but also t o t he coron al p lane. In p at ien ts

complaining of sagittal imbalance, Bridwell20

classi ed the coexistent coronal imbalance

with into type A and type B. In type A, the

pat ient ’s sh ou lders and pelvis are t ilt ed in op-

pos ite direct ions; the sh ou lder is elevat ed at

the side where the pelvis is lower. Conversely,

w ith type B, both the shoulders and the pelvis

tilt in the same direction. An asymmetrical

PSO is often useful in correcting type A bi-

p lanar deform it ies.21 The more radical tech-

niques such as vertebral column resection(VCR) are som et ime s useful for the rar e t ype B

deformities.20

! Prevention o f

Postoperative Coronal

Decompensation

A Nove l Classi cation of CoronalBalance Pattern

A discrepancy exists between sagittal imbal-

ance, which is well account ed for in th e t radi-

tional treatment algorithm, and imbalance in

the coronal plane, wh ich th e algorithm ignores.

Furthermore, postoperative coronal decom-

pensat ion is an im por tan t complicat ion that

a ects surgical outcom e and increases the re-

vision rate. To address this problem, we have

established a novel classi cation regard ing cor-onal balance pat tern s for adult scoliosis.

This classi cat ion is based on CB, w hich is

m easured as th e distance of the m idpoint of C7

relative to the CSVL on standing posteroante-

rior X-ray lms (Fig. 7 .3). Ver tebral alignm en t in

the coronal plane is considered to be balanced

if CB is less th an 3 cm at eith er side; oth er w ise,

it is considered to be im balanced. Patients w ith

a balanced coronal pattern are categorized as

typ e A (Fig. 7 .3a). Patients w ith an im balanced

coronal pattern (CB m ore th an 3 cm ) are cate-

gorized as type B if the imbalance is on the

concave side of the m ain curve (Fig. 7 .3b) and

typ e C if th e imb alance is on th e convex side of

the m ain curve (Fig. 7 .3c).

Osteotomy Options Based on

This Classi cation

For a coronal pattern of type A or type B, the

three-column osteotomy, if necessary, should

be perform ed r igh t at t he apex from the convex

side, so as to restore lumba r lordosis and to re-

establish coronal balance when the compres-

sion forces are app lied to close th e osteotom y

gap. This osteotom y opt ion is ver y e ect ive in

correcting patients with a type B coronal pat-

tern. But in type C patients with preoperative

coronal im balance on the convex side, this os-

teotomy option m ight be inappropriate due to

the compression forces at th e ap ex. Although itis rare, intraoperative dislocation after three-

colum n osteotom y can occur as a severe com-



88 Chapte r 7

plicat ion, due to compulsive ly re ct ifying the

imbalanced trunk, which is accompanied by a

relatively rigid lumbosacral hemicurve.

To restore a balanced spine with t he tr unk

centrally over the pelvis, a novel osteotomy

strategy has been suggested for cases withtyp e C (Fig. 7 .4). First, a th ree- column osteot-

omy needs to be performed at a more distal

level, usually at th e L4 ver tebra or th e L4/5 disk

from the concave side of the m ain curve to re-

store the balance of the trunk over the pelvis.

Second, the apical region is then corrected.

In cases with kyphosis, an additional three-

column osteotomy can be done at the apex.

The optimal strategy of osteotomy for type C

begins from the concave side, and convert s

the previous imbalance pattern into a type A

pat tern.

In a ddition, an asym m etrical PSO might be

an alternate option for patients with a type C

pat tern. Toyone et al21 described t he technique

of asym m etrical PSO through w hich a coronal

correction was well achieved upon closure of

the osteotom y wedge on the convex side.

! Revision Surgery for

Instrumentation Failure

Due to Postoperative

Coronal Imbalance

Long spinal instr um ent ation is often indicated

in adult spinal deformity, which immobilizes

a long span of spinal segmen ts, leading to in-

creased m otion of the adjacent segm ents an d

the potential for degenerative pathology. Be-

cause the lumbosacral junction presents high

mechanical demand, a high rate of complica-

tions has been well docum ented for long fusions

to the sacrum . One of the implant-related com-

plicat ions is ro d fra ct ure , w hich is associated

w ith th e use of iliac screws or sm all-diamet er

rods, operating at inappropriate fusion levels,

resulting in postoperative coronal imbalance,

and failing to add ress sagitt al imbalance.

Postoperat ive coronal imbalance often re-quires add itional revision surgery. In our prac-

tice, the inciden ce of rod breakage is 15% in

adult spinal deformity patients (9/59) with a

minimum of 2-year follow-up, particularly in

pat ients w ith postop era tive re sid ual k yp hosis.

We speculate th at rod fracture m ay partly result

from overloaded mechanical forces imposed

on instrumentation in cases with postopera-tive coronal im balance (Fig. 7 .5).

The use of iliac screws might increase the

risk of implant failure because of the increased

sti ness of the lumbosacral constructs. The

excessive stress of rod contouring is necessary

to connect iliac screws and S1 pedicle screws,

bu t it can lead to ro d fra ct ure .22 In particular,

we found that, in patients with postoperative

coronal decompen sation, the location of the rod

fracture is often close to the level of the iliac

crest or th e osteotom y level (Fig. 7 .6).

Postoperative coronal imbalance th at is com-

plica te d by a symptom at ic rod fra ct ure is a de-

nitive indication for revision surger y. Several

m odalities have been emp loyed to x the frac-

tured rod. Traditionally, the entire incision is

reopened an d the fractured rod is replaced with

a new one. Alternatively, revision with a com-

binat ion of in -lin e rod connector s an d cross-

links can restore th e sti ness of the original

construct w ithout the need to replace the en-tire construct. To reinforce the local construc-

tion of the fractured rod and decrease the risk

of comp lications, we u se satellite rod s (Fig. 7 .6).

This local direct-repa ir strategy requ ires the

reopening of only the area surrounding the

fractured rod rather than the entire opening

along the instrumentation. More importantly,

satellite rod s can en able the restoration of cor-

onal balance through local comp ression at t he

convex side or distraction at the concave side.

Recently, we st arte d u sing the satellite r ods in

the index surgery at the osteotomy level or

wh en the instrum entation bridges the lumbo-

sacral junct ion.

! Chapter Summary

Adult scoliosis may stem from the progression of

scoliosis in children or ad olescents (idiopath ictype), or may newly develop in adulthood

through degenerative changes (degenerative

(text cont inues on page 93)




Fig. 7.4a,b (a) A 65-year-old woman withdegene rative scoliosis and a t ype C coronal

ba lance pat tern. Lum bar kyphosis and severe

sagitt al imba lance was note d. As per t he

surgical algorithm, a long fusion from T6 to

S1 was starte d with osteotomy at L4/5 to

ba lance the spine in the coronal plane

followed by a pedicle subtraction osteotomy

(PSO) at L1. (b) At 2-year follow-up, the spinal

ba lance was well maint ained.

a

b



90 Chapte r 7

Fig. 7.5a–h (a,b) A 56-year-old woman with

degene rative lumbar kyphoscoliosis complicate d by

lumbar st enosis. Posterior spinal fusion from T5 to

pelvis was done t oget her with L4-L5 decompre ssion.

(c,d) Postoperative coronal imbalance was noted.

(e ) Both rods were fractured at 8 months’ follow-up.

(f) Revision surgery with a domino connector was

pe rform ed to restore coronal balance, which was

well maintained at (g,h) 2 years, follow-up.

a b c d

e f g h




Fig. 7.6 a–e (a,b) A 64-year-old woman with

degenerative lumbar scoliosis was treated with

Luque instrumentat ion 9 years ago in another

hospital. Poste rior instrume ntat ion from T5 to S1

with an L1 PSO was performed in the revision

surgery.(c)

However, immediate postoperativecoronal imbalance toward the convex side was

noted. (continued on page 92)

a

b c



92 Chapte r 7

Fig. 7.6 a–e (continued ) (d) Two years later, the

rod fractured at t he right side of L2. (e ) A second

revision surgery was per forme d with sat ellite rods.

Both coronal and sagitt al balance was restored at

6 m onths’ follow-up.

d

e




type). Surgical interventions are mainly in-

dicated for patients complaining of pain and

disability. In our experience, several factors

di erentiate between degenerative and idio-

pat hic sco liosis . In degenerat ive sco liosis , d isk

degeneration contributes to the occurrenceof spinal imbalance. Disk degeneration at the

lower end vertebra strongly correlates with

sagittal imbalance, whereas that at the apex

correlates with cu rve m agnitude. In degen era-

tive lumba r scoliosis, imbalance in the coronal

plane is fre qu ently obse rved an d usu ally asso-

ciated w ith deter ioration of symptom s such as

back pain and rad icu lop at hy. A r ecent su rvey

found a high rate of preoperative coronal im-

balance in degenerat ive lum bar sco liosis , on

either the concave or the convex side. Correc-

tive surgery m ight result in coronal decompen -

sation in adult scoliosis patients, leading to

trunk shifting and possibly implant failure.

Deformity- and surgery-related factors might

lead to t his complication.

A novel classi cation system has bee n de-

vised for the coronal balance pattern in adult

scoliosis: typ e A, balanced; t ype B, imb alanced

on th e concave side; and type C, im balanced

on the convex side. For a preoperative coronal pat tern of t yp e A or B, t he thre e-colum n oste -

otomy should be p erform ed right at the apex

from the convex side. For type C, the optimal

strategy of osteotomy begins from the concave

side of the main curve, at a more distal level,

usu ally at the L4 vert ebra or L4/5 disk, followed

by corre ct ion of the apica l re gion. For inst ru-

m ent ation failure due to postoperat ive coronaldecompensation, revision surgery focuses on

reinforcing th e local constr uct ion, using in-line

rod conn ector s, cross-links, or satellite rod s.

Pearls

Consult the classi cation system that has been

devised for preo perative coronal balance p att ern

in adult scoliosis.

Ident ify the factors that di erent iate de gene ra-tive and idiopat hic adult scoliosis.

Keep in m ind that disk degen eration may con-

tribute to sp inal imbalance and curve severity.

Evaluate t he risk factors of postop erat ive coron al

decompensation.

Be aware of revision options for instrumentation

failure due to postoperative coronal imbalance.

Pitfalls

Postope rative coronal decompen sation can o ccur

after osteotomy at the apex.

Postoperative coronal decompensat ion may result

in instrume nta tion failure and loss of correction.

References


1. Aebi M. The adult scoliosis. Eur Spine J 2005;14:

925–948 PubMed

2. Schwa b F, Dub ey A, Gam ez L, et al. Adu lt sco liosis:

pre valen ce, SF-36 , a nd nut rit ion al par am et er s in an


1085 PubMed

3. Xu L, Sun X, Huang S, et al. Degenerative lumbar

scoliosis in Chinese Han popu lation: prevalence and

relationship to age, gender, bone mineral density,

and body mass index. Eur Spine J 2013;22:1326–

1331 PubMed

4. Glassm an SD, Schw ab FJ, Brid we ll KH, Ond ra SL, Ber -

ven S, Lenke LG. The selection of operative versus

nonoperative treatm ent in patients w ith adult scoli-

osis. Spine 20 07;32:93 –97 PubMed

5. Smith JS, Sha rey CI, Ber ven S, et al; Spinal Deform ity

Study Group. Operative versus nonoperative treat-

m ent of leg pain in adu lts with scoliosis: a ret rospec-

tive review of a prospective multicenter database

with two-year follow-up. Spine 2009;34:1693–1698

PubMed

6. Silva FE, Len ke LG. Adu lt d egen er ative s coliosis:

evaluation and ma nagem ent . Neurosurg Focus 2010;

28:E1 PubMed

7. Bess S, Boachie- Adjei O, Bur ton D, et a l; Inter nat ional

Spine Study Group. Pain and disability determine

treatm ent m odality for older patients w ith adult sco-

liosis, wh ile deformit y guides treatm ent for younger

pat ien ts. Spine 2 00 9; 34 :2 18 6– 21 90 PubMed

8. Bridwell KH, Edwards CC II, Lenke LG. The pros and

cons to saving the L5-S1 motion segment in a long

scoliosis fusion construct. Spine 2003;28(20, Sup-

pl): S234– S242 PubMed

9 . Polly DW Jr, Hamill CL, Brid well KH. Debat e: t o fu se

or not to fuse to the sacrum, the fate of the L5-S1

disc. Spine 2006;31(19, Suppl):S179–S184 PubMed



94 Chapte r 7

10 . Yagi M, King AB, Boach ie-Adjei O. Incid en ce, r isk fac-

tors, and natural course of proximal junctional ky-

phos is: su rg ica l outcom es re view of ad ult id iop at h ic

scoliosis. Minim um 5 years of follow-u p. Spine 2012;

37:1479–1489 PubMed

11. Bao H, Liu Z, Zhu F, et a l. Is t he sacro -fem oral-p ubic

angle predictive for pelvic tilt in adolescent idio-

pat h ic sco liosis pa t ien ts? J Spin al Diso rd Tech 20 14 ;

27:E176–E180 PubMed

12 . Mac-Thion g JM, Transfeld t EE, Mehb od AA, et a l. Can

c7 plumbline and gravity line predict health related

quality of life in adult scoliosis? Spine 2009;34:

E519–E527 PubMed

13. P rr m ann CW, Metz dor f A, Zanet ti M, Hodler J, Boos

N. Magn et ic res on an ce classi cat ion o f lum ba r in ter -

vertebral disc degeneration. Spine 2001;26:1873–

1878 PubMed

14 . Daubs MD, Len ke LG, Bridwe ll KH, et a l. Does cor rec-

tion of preop erative coronal im balance make a di e-

rence in outcomes of adult patients with deformity?

Spine 2013;38:476–483 PubMed

15. Glassm an SD, Ber ven S, Bridw ell K, Horton W, Dima r

JR. Correlation of radiograph ic param eters and clini-

cal symptom s in adult scoliosis. Spine 2005 ;30:682–

68 8 PubMed

16 . Ploum is A, Liu H, Mehb od AA, Transfeld t EE, Win -

ter RB. A correlation of rad iographic and functional

m easurem ent s in adult degenerat ive scoliosis. Spine

2009;34:1581–1584 PubMed

17. Cho W, Mason JR, Sm ith JS, et a l. Failur e of lum bop el-

vic xation after long constru ct fusions in patients

with adult spinal deformity: clinical and radio-

graphic risk factors: clinical article. J Neurosurg

Spine 2013;19:445–453 PubMed

18. Ploum is A, Simpson AK, Cha TD, Herzog JP, Wood KB.

Coronal spinal balance in adu lt spine deform ity pati-

ents w ith long spinal fusions: a minimum 2–5 year

follow-up study. J Spinal Disord Tech 2013 Sep 27.

[Epub ah ead of print] PubMed

19. Lowe T, Berven SH, Schwab FJ, Bridwell KH. The SRS

classi cation for adult spinal deform ity: building on

the King/Moe and Lenke classi cation systems. Spine

2006;31 (19, Suppl):S119–S125 PubMed

20. Bridwell KH. Decision m aking regarding Sm ith-

Petersen vs. ped icle subtra ction osteotom y vs. verte-

br al colum n re se ct ion for sp inal deform it y. Spin e

2006;31 (19, Supp l):S171–S178 PubMed

21. Toyone T, Shib oi R, Ozawa T, et a l. Asym m et rica l ped-

icle subtraction osteotomy for rigid degenerative

lumbar kyphoscoliosis. Spine 2012;37:1847–1852

PubMed

22. Sche er JK, Tan g JA, Deviren V, et al. Biom ech an ical

analysis of revision str ategies for rod fract ure in p ed-

icle subtraction osteotomy. Neurosurgery 2011;69:

164–172 , discussion 172 PubMed



! Introduction

The de nition of a health-related outcome var-

ies in the literature an d m ay encomp ass a spec-

tru m of m easures. Clinical outcom e is the en d

result of health care delivered to patients or

pop ulat ion s, and entails su ch considerat ions

as quality, patient -based assessment , and value

of care. Measuring outcomes is complex, and

there is no single m easure that sum m arizes the pat ient’s exp erience, the hospital persp ect ive,

the payer perspective, and the treating physi-

cian’s p erspect ive. Therefore, outcomes must be

considered broadly and encompass a spectrum

of perspectives and m easures. This chapter d is-

cusses measures of health-related quality of

life (HRQOL), an d the a pp lication o f these m ea-

surem ent s in assessing outcom es of spinal de-

form ity surgery.

Outcome m easuremen t is an important as-

pect of su rgeon accountabilit y an d is vit al in

determining the quality and value of health

care. Quality may be evaluated based on pro-

cess measures, objective health outcomes,

pat ient-re por te d ou tcom e m ea su re m ents, an d

cost of care. Value is a broad er m easure that in-

corporates an analysis of both quality and

cost. This chapter provides an overview of

various outcome measures used for spinal de-

form ity, and ndings from the literature on

outcom es in adult spinal deform ity surgery.

! Process Measures

Process m easures are a re ection of how care

is delivered. Examples include compliance

with antibiotic or thromboembolic prophy-

laxis guidelines, the u se of surgical “tim e-ou ts”

before su rgery, preop erat ive risk assessm ents

of patients, and implementing postoperative

care protocols for the prevention of common

po stop erative com plication s. The utilit y of pro -cess measures depends on how reliably they

are linked to clinical outcomes. For example,

m easuring comp liance with pre operat ive anti-

biot ic gu idelines is u sefu l in that it m ay p re dict

a reduction in the incidence of surgical site

infections. Although th ey are an indirect m ea-

sure of quality, the implementation and mea-

surement of such guidelines are important in

standard izing care an d imp roving quality.

! Physiological Outcome

Measures

Physiological ou tcomes repre sent clinical h ealth

metrics that may be measured objectively. In

adult spinal deformity, these may include ra-

diographic outcomes, implant survival, and

fusion rat es. Cobb angle, sagitt al vert ical axis

(SVA), and spinopelvic parameters including

8

Measuring Outcome and Valuein Adult Deformity Surgery

Robert Waldrop and Sigurd Berven



96 Chapte r 8

pelvic in cid ence (PI), pelv ic t ilt (PT), sa cral

slope (SS), an d lum bar lordosis (LL) are a ll im -

por tan t radiogra phic m ea su rem ents in sp inal

deformity. Although these outcomes are easy

to measure and interpret , i t is important to

evaluate clinical measurem ent s in relation t otheir correlation with quality measurements

(e.g., do radiographic outcom es pre dict reope r-

ation rates?).

! Quality Measures

Traditional measurements of quality include

outcom es such as operative tim e and length of

hosp ital stay, as w ell as rates of comp lications ,

reoperations, and readmissions. Such measures

provide im por tant in form at ion that m ay be

used to comp are the p erforma nce of individual

providers and hospitals and to es tablish m et -

rics for pe rforman ce and goals for qua lity im -

prove m ent . However, ove rall qu alit y of ca re

encompasses mu ch more than these tradition-

ally reporte d quality met rics.

Quality metrics are valuable in identifying

outliers, and in improving care processes and pat hways. How eve r, overall qu alit y m easure s

are distinct from pat ient-cente red clinical out-

come measures. One important concern re-

garding reliance on quality measures is the

poss ib ilit y that m easu rin g qu alit y alone m ay

lead to a focus on outcomes that are not pa-

tient-centered. If the target for outcome were

only length of stay or avoidance of readmis-

sion, the n th at goal m ay incentivize signi cant

undertreatment of complex spinal disorders.

Fig. 8 .1 provides an examp le of a case in wh ich

a patient un derw ent a limited decompression

and posterior-based tethering procedure for

a comp lex spinal deform ity. Measured by only

length of stay or com plications of care, the lim -

ited decompression surgery would be rated as

a high-quality outcome. However, the patient

had n o improvem ent in he r health status or in

her deformity measures. The patient under-

wen t a revision surger y 3 years after th e index

procedure an d w as t reated w ith a three-colu mnosteotomy for multiplanar realignment of the

spine. The revision sur gery resu lted in a longer

stay, higher cost, and more risk and potential

for comp lication th an the index surger y. How-

ever, the p atient rep orted a dr am atic imp rove-

men t in health status that is not captured by

the quality metrics alone. It is important to

avoid myopic focus on qu ality m etr ics w ithoutgiving priority to patient-centered measures

of clinical outcom es in spinal deform ity.

! Patient-Reported

Outcomes

Although process measures, physiological out-

comes, and trad itional quality m etr ics are im-

por tant too ls for assessin g health care qu alit y,

they do not re ect the patient’s health care

experience or the impact of care on HRQOL.

There has been an increasing emphasis on pa-

tient- based health assessme nts in recent years.

Patient-reported outcome measures (PROMs)

may include a spectrum of domains to assess

HRQOL. Frequently used domains include

disability/functional status, pain and other

symptom s, em otional/psychological w ell-being,

general health status, and satisfaction withhealth care experience. The Visual Analogue

Scale (VAS) for pain assessm ent is anoth er com -

mon ly used outcome m easure.

Measurement tools for patient-reported out-

comes include both disease-speci c and gen-

eral health status measures. Disease-speci c

measures focus on domains associated with a

par t icu lar con dit ion or pat ient pop ulat ion, an d

have the advantage of increased responsive-

ness to change (ther e is a m ore reliable change

in outcome score as the underlying condition

changes) compared with general health status

m easures. Examp les of speci c outcome tools

includ e th e Scoliosis Research Societ y (SRS-22)

questionnaire, the Oswestry Disability Index

(ODI), and th e Neck Disability Ind ex (NDI).

General health status outcomes tools are

advantageous in that they may be used in any

pat ient pop ulat ion and allow for bro ad com -

par isons across a spect rum of m edica l an d sur-

gical conditions. However, they are often lessresponsive to changes in par ticular conditions

or disease states. Examples of gene ric pro les



Measuring Outcom e and Value in Adult Deformity Surgery 97

Fig. 8.1a,b (a) A 73-year-old woman

presented with sag ittal and coronal

plane deformity, back pain, and

neurogen ic claudicat ion. She was

unable to live independent ly. The

patient was t reate d with a limiteddecompression and a posterior-based

tet hering device. Although the length

of stay was 3 days, and the re was no

complication or readm ission within 90

days, the re was also no improvement

in radiographic or patient-cente red

clinical outcomes, and the patient

remained disabled. (b) Postoperative

X-rays 2 years after a revision surgery

in which the patient was treated with a

three -column osteotomy for realign-

ment of the spine. The pat ient stayedin the hospital for 6 days and her

pe rioperative course was com plicated

by a supravent ricular t achycardia that

required cardioversion. At 2-year

follow-up, she was living independently

and walking without limits.

a

b



98 Chapte r 8

include t he Shor t Form 36 (SF-36), Shor t Form

6 Doma ins (SF-6D), th e EuroQOL ve dim en -

sions questionnaire (EQ-5D), and the Health

Utilities Ind ex (HUI).

Both general health statu s m easures and cer-

tain disease-speci c outcom e tools m ay be usedas indirect m easures t o calculate u tility scores.

A ut ility score re ects societal preferen ces for a

health state. Di eren t health states are rated

on a continuous scale from 0 to 1, with the

value re ect ing a m easu re of well years of life.

Utility scores derived from patient-reported

outcome questionnaires using validated in-

struments provide information on a patient’s

health status and the value that society places

on t hat health state. Consideration of a ut ility

score over time yields a quality-adjusted life

year (QALY), calculated as t he ut ility score m ul-

tiplied by the nu m ber of years that h ealth state

is maintained. Thus, the durability of an out-

com e re sult s in increased QALYs over t ime. A

QALY is an ou tcome m easure that represen ts a

standa rdized unit for compar ison across elds

and can be assigned value by society.

! Commonly Used Outcome

Measurements

The follow ing instrum ent s are comm only used

pat ient -repor ted outcom e m easu rem ents in

adult spinal deform ity that h ave validated con-

version s t o u tility scores /QALYs.

Short Form 36 (SF-36) and ShortForm 6 Domains (SF-6D)

The SF-36 is a w idely used generic health sur-

vey consisting of 36 quest ions w ith four physi-

cal health scales (physical funct ioning, physical

role lim itation, bodily pain, and general h ealth)

and four mental health scales (vitality, social

functioning, emotional role limitation, and

mental health). The SF-6D is an abbreviated

version of the SF-36 t hat h as been established

as a preference-based health state classi ca-tion th at m ay be converted to a ut ility score.1

EuroQOL Five Dimensions

Questionnaire (EQ-5D)

The EQ-5D is anoth er validated and w idely used

general health questionn aire that is used to es-

tab lish a ut ility score. It includes ve healthdimensions: mobility, self-care, usual activities,

pain/discom for t , an d anxie ty/depressio n.

Scoliosis Research Society

Questionnaire

The Scoliosis Research Society (SRS) question-

naire measures how spinal deform ity a ects a

pat ient’s HRQOL based on ve dom ains: pain ,

function, self-image, men tal he alth, and satis-

faction. The 22-item questionnaire (SRS-22) is

the most widely used and validated version,

although several other versions exist (SRS-24,

SRS-29, SRS-30). SRS-22 has been validated as

a reliable instrument with high internal con-

sistency, responsiveness, reproducibility, and

discriminator y capacity for patients w ith adult

deformity.2,3 A m odel has been established for

translating SRS-22 scores to SF-6D scores to

deter m ine utility scores.4,5

Oswestry Disability Index (ODI)

The ODI mea sure s HRQOL in p atien ts w ith low

back pain . It rat es a pat ien t ’s d isabilit y score

base d on 10 m easu res: pain , perso nal ca re,

sitt ing, stan ding, walking, lifting, sleep ing, sex

life, social life, an d t raveling. Highe r scores cor -

respond to a greater degree of disability. The

ODI is a validated and w idely used m easure that

can be reliably translated to a ut ility score.6

! Cost and Value

In our cur rent health care e conomy, cost has be-

come an increasingly important consideration

in the assessmen t of health care inter ventions.

Econom ic analyses of health care intervent ions

include cost-m inimization stud ies, cost-e ec-

tiveness an alyses, an d cost-u tility ana lyses. Anassessment of costs may include direct costs,




charges, and reimbursements. Indirect costs

such as loss of productivity due to time o

from work, transpor tation to health care facili-

ties, and th e cost of caregivers m ay also be in-

clude d in cost analyses and incorporate a w ider

view of total costs from a societal perspect ive.Although cost in itself is an impor tan t con-

sideration, the value of care provides th e m ost

meaningful assessment of a health care inter-

vention. Value of care en compasses both out-

com e and cost and is de ned as the net bene t

of care relative to t he net cost of care, or wh at

we get for wh at we spend. The m easurement

of bene ts and costs in spine surgery is not

uniform and m ay vary depen ding on the per-

spective of the stakeholder in the health care

economy. Hospitals and other health care facil-

ities may emphasize outcomes and costs that

a ect a single admission such as length of

hospit al stay, imp lant u tilization, and com pli-

cations. Third-party payers often focus on

outcomes and costs in a medium-term time-

frame including readmissions within 90 days

or the cost of outpatient care. The value of a

health care intervention to the physician and

pat ient is es tablished ove r a longer t im efram e

than a single admission; its imp act is m easured based on HRQOL over a lifet im e.

Cost-ut ility stud ies provide t he m ost useful

information about the value of a health care

intervention because a utility score is able to

captu re a patien t’s preference for di eren t health

states over tim e. An outcom e m easure th at di-

rect ly re ects HRQOL and is tran slatable across

disease states, such as QALYs, is an imp ort ant

prerequ isite for est imat ing th e value of or thope-

dic care. The length of follow- up is also an im-

por tan t considerat ion when measuring value, as

the cost of a single episode of care will be signi -

cantly discounted by the duration of the ben e t.

! Outcomes of Adult Spinal

Deformity Surgery

Several stu dies have reported various outcomes

for the operative and nonoperative manage-ment of adult spinal deformity. Estimates of

the prevalence of adult spinal deform ity in the

United States range from 2.5 to 25%7 However,

many of these patients do not seek medical

care for their condition, and of those who

do, many m ay have successful ma nageme nt of

their symptoms without surgery. Nonopera-tive care may include physical therapy, core

strengt hen ing, weight loss/aerobic activity, pain

m edications, steroid injections, and altern ative

m odalities such as acupunct ure and chiroprac-

tic care. For m ost patien ts, a trial of nonop era-

tive care should be initiated before sur gery is

considered. Exceptions include patients with

neurologic de cits or signi cant instability.

Surgery may also be indicated in patients

w ith p rogressive curves, substantial deform ity-

related pain, and those who have failed ap-

prop riat e nonop erat ive t reat m ent . Stu dies of

operative and nonoperative management of

adult spinal deformity h ave d emon strated im-

proved pat ient-re por ted ou tcom es w ith su rgi-

cal man ageme nt.8–12

In a review art icle on ad ult spinal deformity,

Youssef et al13 sum mar ize the ndings of 49

studies rep ort ing outcomes for various sur gical

strategies including decomp ression alone ver-

sus decompression with fusion; anterior, pos-terior, or combined surgical approaches; the

use of vertebr al osteotomies; an d levels of in-

strumented vertebrae. A variety of outcome

m easuremen ts are reported for each technique.

Radiographic Outcomes

A systematic review of adult scoliosis outcomes

by Yadla et al14 found a range in Cobb angle

correction from 9.1 to 53.9 degrees (m ean 26 .6

degrees, representing an average 40.7%curve

correction) in a ser ies of 49 ar ticles pu blished

be tween 195 0 and 2 00 9 w ith m in im um 2-yea r

follow-up.

Radiographic outcomes h ave also been com-

pared be twee n di erent su rgical ap proaches .

Crand all and Revella15 foun d no signi cant dif-

ference in coronal curve correction between

pat ient s undergoing pos ter ior in st rum ented

fusion in addition to either anterior lumbar

interbody fusions (ALIF), with an average cor-rection 69.5%, or t ransforam inal lumb ar inter-



100 Chapte r 8

bod y fusion (TLIF), w ith an ave rage corre ct ion

68.7%. A liter atu re r eview by Mun dis et al16 re -

por te d im proved coronal and sagit tal correct ion

with a lateral transpsoas approach compared

w ith open anter ior procedures. A retrospective

report by Pateder et al17 comparing patientswh o under went posterior only surgery (n = 45)

versus combined anterior-posterior surgery

(n = 35) found no signi cant di eren ce in coro-

nal or sagittal curve correction between the

two groups.

Youssef et al13 also reviewed radiographic

outcom es for di eren t types of posterior oste-

otomies including Smith-Petersen osteotomy

(SPO), pedicle subt ract ion osteotom y (PSO), and

vertebral column resection (VCR). These tech-

niques are used to achieve varying degrees of

lordosis correction and restoration of sagittal

balance. SPO provid es the sm alle st degr ee of

curve correction, achieving up to 10 degrees

of lordosis per vertebral level; however, multi-

level osteotomies may achieve a large overall

correction. Repor ts of PSO have dem onstrated

an average 30 degrees of lordotic correction

per level. In a comparison of SPO an d PSO, Cho

et al18 found an average total correction of 33

degrees for patients undergoing three or moreSPOs and 31.7 degrees for patient s un dergoing

PSO, but a signi cant ly lower imp rovemen t in

sagitta l balance for t he SPO group tha n for th e

PSO group. VCR achieves the highest degree of

curve correction. Suk et al19 reported a mean

deformity correction of 59%(109.0 degrees to

45.6 degrees) in 16 patients who u nderw ent

pos ter ior VCR. Papadop oulos et al20 reported

that of 45 patients w ho under went p osterior

VCR, the average correction of kyphosis was

from 108 degrees to 60 degrees with one pa-

tient sustaining a complete spinal cord injury.

Complications

The inciden ce of comp lications is an imp orta nt

quality m easure in adu lt spinal deform ity. Re-

por ted com plica t ion rates for sp inal deform it y

surgery are high, but a stan dardized de nition

or classi cation for repor ting complications in

the literature has not been established. Com- plicat ion rat es have be en classi ed in various

ways, including major versus m inor comp lica-

tions, early versus late complications, and sur-

gical versus medical complications. Reported

complications in deformity surgery include

pseudar throsis, adjacent segm en t d isease, dura l

tears, super cial or deep wound infections,implan t complications, neu rologic de cits, epi-

dural hematoma, wound hem atoma, pulmonary

embolism, deep vein thrombosis, systemic com-

pl icat ions, and deat h . The in cid ence of com -

plica t ions m ay b e in uenced by pat ient factor s

(e.g., age, comorbidities, severity of deformity)

or sur gical factor s (e.g., app roach t ype, need for

osteoto my, nu m ber of levels fused).

The 49 ar ticles reviewed by Yadla et al14 re-

por t complicat ion ra tes rangin g from 0 to 53 %,

with a combined total of 897 complications

am ong 2 ,175 pat ient s (41.2%). Char osky et

al31 repor ted an overall 39%comp lication rate

among 306 patients over age 50 undergoing

adult deformity surgery with either an ante-

rior on ly, posterior on ly, or comb ined app roach.

Sansur et al22 reviewed a total of 4,980 cases

of adult scoliosis from the SRS morbidity and

mortality database and found an overall com-

plica t ion rat e of 13.4% and a m or talit y ra te of

0.3%. Signi can tly higher complicat ion rat esresulted from revision sur geries, osteotom ies,

and combined anterior-posterior surgery.

Youssef et al13 summarized several studies

repor ting complication rates of various p roce-

du res. Tran sfeldt et al23 repor t a 10%complica-

tion rate among adu lt deform ity patients w ho

unde rwent decompression alone compared with

56% in patients wh o underwent decompres-

sion and fusion. Burn eikiene et al24 reported a

31% inciden ce of system ic comp lications and

49%ha rdware or surgical technique comp lica-

tions in 29 patients undergoing TLIF. Compli-

cations of ALIF may include vascular injuries,

ilioinguinal and iliohypogastric nerve injuries,

damage to the bladder or ureters, pseudar-

th rosis an d su bsidence, ileus, lymph ocele, and

retrograde ejaculation.13,25 Most of these com -

plicat ions are uncom m on , althou gh ra tes of

m ajor and m inor complications vary in th e lit-

eratu re. In a stu dy of 447 p atients, McDonnell

et al26 found a complication rate of 11% form ajor comp lications and 24%for m inor compli-




cations. Comp lications of th e lateral t ranspsoas

approach are often related to manipulation of

the lumber plexus. In a prospective multi-

center evaluation of 107 adult degenerative

scoliosis patients undergoing extreme lateral

inter body fusion, Isaacs et al27 reported a 12.1%m ajor comp lication rate.

Reoperations

Scheer et al28 an alyzed dat a from a prospe ctive,

multicenter adult spinal deformity database,

and exam ined th e rates, indications, timing, and

risk factors for reoperat ion as well as the e ect

of reope ration on HRQOL m easu res. In a coh ort

of 352 patients (268 with at least 1-year fol-

low-up), they found a total reoperation rate

of 17%, the m ajority of w hich occurred w ithin

1 year of the index operation. The most com-

mon indications for reoperation included in-

strumentation complications and radiographic

failure.

There w as a 19% reoperat ion rate for pa-

tients undergoing a three-column osteotomy

and a 16% reoperation rate for patients not

requiring three-column osteotomy; however,

three-colum n osteotomy was not signi cantly predict ive of reop era t ion at 1 year. The upper-

m ost instrum ented vertebra was also not pre-

dictive of reoperat ion. There w ere no signi cant

di erences in the Am erican Societ y of Anes-

thesiologists (ASA) grade, Charlson comorbid-

ity index rating, preop erative body m ass index

(BMI), or smoking history between patients

wh o did not undergo reoperation and those wh o

did. Patients wh o needed reoperation w ithin 1

year h ad w orse ODI and SRS-22 scores at 1-year

follow-up than did patients not needing reop-

erat ion. However, th ere wa s no signi cant dif-

ference in HRQOL scores at 2 years between

pat ients w ho re qu ired re op era t ion at 1 year

and those wh o did not.

Other studies have dem onstrated similar re-

ope rat ion rates , ran ging from 10 to 21%.21,29–31

Reasons for revision surgery in adult spinal

deformity include pseudarthrosis, curve pro-

gression, infection, painful/prom inent implant s,

adjacent segment disease implant failure, andneuro logic de cits.29,30

HRQOL Outcomes in Adult

Spinal Deformity

Despite high comp lication and reope ration rates

in adult spinal deform ity surgery, patient sat-

isfaction with these procedures is high. Bothcondition-speci c and genera l HRQOL outcomes

that can be converted to utility scores and

compared across the literature are important

prere qu isi tes for det erm in ing the value of sp i-

nal deform ity surgery.

Several prospective m ulticenter st udies have

dem onstrated the bene ts of operative treat-

men t of adult spinal deform ity compared with

nonoperative care in regard to patient-rep orted

health m easures includ ing ODI, SRS-22, EQ-5D,

and numeric rating scale scores for leg and

ba ck p ain .8–12

Yad la et al14 reported th at in the 49 studies

included in their systematic review, ODI and

SRS were the most commonly used patient-

ba sed ou tcom e inst rum ents, w ith 11 st udies

reporting pre- and postoperative ODI scores

and 10 studies reporting pre- and postopera-

tive SRS scores. The re w as an average decre ase

of 15.7 points (range 3.1–32.3) in ODI score

am ong 911 patient s. This imp rovem ent in dis-ability outcome correlates with previous re-

por ts of signi can t clin ical im prove m ent of

ODI scores ran ging from 4 to 15 p oints.32 Of th e

999 patients with pre- and postoperative SRS

scores in Yadla et al’s review, the re w as a m ean

increase in SRS scores of 23.1 points, well

above the minimal important di erence for

SRS scores of 13 point s rep ort ed b y Bagó et al.33

You ssef et al13 summarized the results of

studies reporting HRQOL outcomes for pa-

tien ts und ergoing various surgical approa ches.

Crandall and Revella15 found nonsigni cant

di ere nces in VAS and ODI out come s betw een

pat ients undergoing posterior fusion w ith ei-

ther ALIF or TLIF. Mundis et al16 foun d signi -

cantly improved VAS and ODI scores in a

literature review of the lateral approach for

adult spinal deformity. Various studies have

reported improved patient-reported outcomes

following PLIF, including improved ODI, SF-36,

and VAS scores.34–36 Good et a l 37 reported sim-ilar SRS and ODI scores for bot h p oster ior on ly



102 Chapte r 8

and combined fusions, with both having im-

provem ents a t 2-ye ar follow-up.

Cost and Value in Adult

Spinal DeformitySeveral recent articles have reported on the

costs of adult spinal deformity surgery. Mc-

Carth y et al38 studied t he tot al costs of 484 pa-

tient s undergoing operat ive treatm ent of adult

spinal deformity with an average follow-up

of 4.8 years, and found an average total hospi-

ta l cost of $120,394. Total cost for pr ima ry su r-

gery averaged $103,143, which increased to

$111,807 at 1-year follow-up and $126,323 at

4-year follow-u p. Hospital read m issions wererequ ired in 130 p atien ts (27%), w ith an aver-

age readmission cost of $67,262. Another cost

analysis by McCarthy et al39 found higher di-

rect costs w ith increasing age, length of hospi-

tal stay, length of fusion, and fusions to the

pelvis.

Cost-utility studies of adult spinal defor-

mity are lacking in the literature. Although

several recent ly published system atic reviews

repor t on cost-ut ility analyses in spine care,40,41

none of the reviewed articles include valueassessments in adult deformity. One study by

Glassm an et al42 examined the costs and bene-

ts of non ope rative care for adult scoliosis, an d

questioned the value of nonoperative treat-

m ent given the ir ndings of a $10,815 m ean

treatment cost over a 2-year period with no

signi can t chan ge in HRQOL.

! Improving Outcomes inDeformity Surgery

Measurem ent of clinical outcom es and value is

an imp orta nt goal in spine surger y and is criti-

cal in establishing accountability for the end

resu lt of care. Ern est A. Codm an w as a sur geon

in the early 20th centu ry and a pioneer in ad-

vocating outcom e m easureme nt an d report ing.

He proposed an “end results system ” in wh ich

pat ients’ symptom s, diagnos is, t reat m ent , andoutcom es would be tracked over tim e in an ef-

fort to redu ce complications and improve qual-

ity of care. At the time, Codman’s ideas were

seen as radical and m et w ith strong resistance,

leading to his dismissal from his faculty posi-

tion at Massachusetts General Hospital. Al-

though great strides have been made since

Codman’s time in recognizing the importanceof outcome measurement, there is still much

room for improvement in the e ort to establish

regular and reliable system s for outcom e m ea-

suremen t and reporting.

There is a high variability in spine surgery

w ith regard to surgical rates, surgical strategies,

and costs.43–45 High variability indicates a lack

of consensus on the optimal treatment strat-

egy and a need for fur the r comp arative e ec-

tiveness re search. Redu cing variability in spine

surgery requires an evidence-based approach

to care. The establishm ent of large m ulticente r

pro cedural an d diagnos is- based regist ries for

spine surgery has been an important step to

improving outcom e measurem ent and report -

ing. These registries provide a reliable system

for th e rep ort ing of comp lications, clinical out -

comes, and HRQOL out comes, and facilitat e t he

evaluation of alternat ive intervent ions in com-

parat ive e ect iveness rese arch . With the ac-

curate measurement of complications, qualitym ay be improved t hrough the establishm ent of

clinical protocols based on stand ards of care in

an e ort to reduce complications. The w ide-

spread use of patient-reported outcome tools

that m ay be translated to a u tility score is nec-

essary to address t he lack of cost-u tility a naly-

ses and value-b ased assessmen ts in adult spinal

deformity. An increased emphasis on measur-

ing and im proving value in spine care w ill re-

sult in improved outcomes and reduced costs

over time. Although we support an e ort to

redu ce variability in spine surgery th rough an

eviden ce-based app roach to care, we also rec-

ognize th at care is not m onolithic, and patient

and physician preference must be considered

to obtain optimal outcomes.

! Chapter Summary

Surgical treatm ent of adult spinal deform ity is

a high-cost intervention that consistently draws




the attention of the lay press and the me dical

pro fessio n due to a perceived lack of e ect . In

the evolving health care economy, dem onstra-

tion of value, through cost data coupled with

pat ient-re por ted ou tcom es , w ill be cr it ica l in

m aintaining patient access to care. To best p ro-tect our patients’ ability to receive care that

can e ect change in the ir health stat us, it is

imperative for spinal surgeons to understand

and collect patient reported outcomes. Surgi-

cal treatment of adult spinal deformity has

been sh ow n to have a signi cant e ect on pa-

tient-reported outcomes. Although the initial

cost of spinal deformity surgery is high, the

cost pe r QALY decreases w ith increasing d ura -

bilit y of the intervent ion. Thu s, it is imperat ive

that we as a profession continue to track and

report secondary interventions and compli-

cations of care, so that optimal intervention

strategies can be created. Determination of

appropriate quality metrics and process mea-

sures for the delivery of spine care can help

achieve improved patient outcom es and poten-

tially lower cost, thus maximizing societal re-

tu rn on investm ent for the care of adult spinal

deformity.

Pearls

In a value-based health care econom y, m easures

of cost an d clinical outcome are importan t t o de -

ne cost-e ective inter vent ions.

Patient-centered measures of outcomes provide

the most useful assessme nt of value of inte rven-

tions in deformity surgery.

Utility scores are a useful measure of general

health status preference that has a de nable unit

of well-years of life/year.

Cost pe r QALY is a measu re o f value th at is sensi-

tive to the m agnitude of the health status change

and t he durability of change.

Selection of disease-speci c, pat ient -reported out-

come should consider validated met rics that can

poten tially be co nver ted to a utility score.

Pitfalls

Sole focus on quality metrics and process mea-

sures creates a dissociation bet ween intervent ions

and pat ient-centered outcomes.

Optimizing qualit y and p rocess metrics in the ab -

sence of patient-centered information may in-

correctly guide evidence-based care, and provide

incen tives for inappropriate care.

Cost-minimization strategies or focus on cost

without regard to e ect of treatment on patient

reporte d ou tcome s will not b e a value-optimizing

strategy.

References


1. Brazier J, Roberts J, Deverill M. The estimation of a

pre fer en ce-ba se d m ea su re of hea lth fro m the SF-36 .

J Health Econ 2 002;21:2 71–292 PubMed

2. Berven S, Deviren V, Demir-Deviren S, Hu SS, Brad-

ford DS. Stu dies in th e m odi ed Scoliosis Resear ch

Society Outcomes Instrument in adults: validation,

reliability, and discriminator y capacity. Spine 2003;

28:2164–2169, discussion 2169 PubMed

3. Bridwell KH, Berven S, Glassman S, et al. Is the SRS-

22 instru m ent resp onsive to change in adult scoliosis

pat ients having prim ar y sp inal deform it y su rgery?

Spine 2007;32:2220–2225 PubMed

4. Bridwell KH, Cats-Baril W, Harrast J, et al. The valid-

ity of the SRS-22 instr um ent in an adult spinal defor-

mity population compared with the Oswestry and

SF-12: a study of response distribution, concurrent

validity, internal consistency, and reliability. Spine

2005;30:455–461 PubMed

5. Brazier JE, Rober ts J. The estim ation of a prefere nce-

ba se d m ea su re of hea lth from the SF-12 . Med Care

2004;42:851–859 PubMed

6. Carr eon LY, Glassm an SD, McDono ugh CM, Ram per -

sau d R, Ber ven S, Sha inline M. Pred icting SF-6D ut i-

lity scores from the Oswestry disability index and

numeric rating scales for back and leg pain. Spine

2009;34:2085–2089 PubMed

7. Unite d Stat es Bone a nd Joint Init iative. The Burd en of

Musculoskeletal Diseases in the United States, 2nd

ed. Rosem ont , IL: Amer ican Academ y of Orth opaed ic

Surgeons; 20 11

8. Everet t CR, Patel RK. A system atic literat ure review of

nonsurgical treatment in adult scoliosis. Spine

2007;32(19 , Supp l):S130–S134 PubMed

9. Sm ith JS, Sha rey CI, Ber ven S, et al; Spinal Defor-

mity Study Group. Improvement of back pain with

operative and n onoperative treatm ent in adults with

scoliosis. Neurosurgery 2009;65:86–93, discussion

93–94 PubMed

10. Li G, Passias P, Kozanek M, et a l. Adu lt sco liosis in pa -

tient s over sixty- ve years of age: outcomes of oper-

ative versus nonoperative treatment at a minimum

two-year follow-up. Spine 2009;34 :2165–2170 PubMed



104 Chapte r 8

11 . Smith JS, Sha rey CI, Ber ven S, et al; Spinal Defor-

mity Study Group. Operative versus nonoperative

treatm ent of leg pain in adu lts with scoliosis: a ret-

rospective review of a pr ospective m ulticenter da-

tabase with two-year follow-up. Spine 2009;34:

1693–1698 PubMed

12. Bridw ell KH, Glassm an S, Horton W, et al. Does

treatment (nonoperative and operative) improve

the two-year quality of life in patients with adult

symptomatic lumbar scoliosis: a prospective multi-

center evidence-based medicine study. Spine 2009;

34:2171–2178 PubMed

13 . You ssef JA, Orndo r DO, Patt y CA, et al. Cur rent st a-

tus of adult spinal deformity. Global Spine J 2013;

3:51–62 PubMed

14 . Yad la S, Malt en fort MG, Ratli JK, Har rop JS. Adu lt

scoliosis surgery outcomes: a systematic review.

Neu ro su rg Focu s 2 01 0; 28 :E3 PubMed

15. Cran dall DG, Revella J. Tran sforam inal lum bar inte r-

body fu sio n ver su s a nte r ior lum ba r in te rb od y fu sio n

as an adjunct to posterior instrum ented correction

of degenerative lumbar scoliosis: three year clinical

and radiographic outcomes. Spine 2009;34:2126–

2133 PubMed

16. Mundis GM, Akbarnia BA, Phillips FM. Adult defor-

mit y correction th rough m inimally invasive lateral

approach techniques. Spine 2010;35(26, Suppl):

S312–S321 PubMed

17 . Pate de r DB, Keba ish KM, Cascio BM, Neubaeur P, Ma-

tusz DM, Kostuik JP. Posterior only versus combined

anterior and posterior approaches to lumbar scolio-

sis in adults: a r adiographic analysis. Spine 200 7;32:

1551–1554 PubMed

18 . Cho K-J, Bridw ell KH, Len ke LG, Ber ra A, Bald us C.

Comparison of Smith-Petersen versus pedicle sub-

tract ion osteotom y for the correction of xed sagittal

imbalance. Spine 2005;30:2030–2037, discussion

2038 PubMed

19 . Suk S- I, Chun g E-R, Kim J-H, Kim S-S, Lee J-S, Choi

W-K. Posterior ver tebral column resection for severe

rigid scoliosis. Spine 2005;30: 1682–16 87 PubMed

20 . Papadopoulos EC, Boach ie-Adjei O, Hess WF, et al;

Found ation of Orthop edics and Comp lex Spine, New

York, NY. Early out com es an d com plication s of post e-

rior vertebr al colum n rese ction. Spine J 2013 Apr 2 5.

[Epub ah ead of print] PubMed

21 . Cha rosky S, Guigu i P, Blam ou tie r A, Roussou ly P, Cho -

pin D; Stu dy Group on Scolio sis . Com plica t ion s an d

risk factors of prima ry adu lt scoliosis surgery: a m ul-

ticenter study of 306 patients. Spine 2012;37:693–

70 0 PubMed

22 . Sansu r CA, Sm ith JS, Coe JD, et a l. Scoliosis r ese arch

society morbidity and mortality of adult scoliosis

surger y. Spine 2011;36 :E593–E597 PubMed

23 . Transfeld t EE, Topp R, Mehb od AA, Win te r RB. Surgi-

cal outcom es of decom pression, decomp ression with

limited fusion, and decompression with full curve

fusion for degene rative scoliosis with radiculopathy.

Spine 2010;35:1872–1875 PubMed

24. Bur ne ikien e S, Nelson EL, Mason A, Rajpal S, Serx-

ne r B, Villavicencio AT. Com plication s in p atien ts u n-

dergoing combined transforaminal lum bar interbody

fusion and posterior instrumentation with defor-

m ity correction for degen erative scoliosis and spinal

stenosis. Surg Neurol Int 201 2;3:25 PubMed

25. Than KD, Wang AC, Rahman SU, et al. Complication

avoidance and m anagemen t in anterior lum bar inter-

body fusion . Neurosu rg Focu s 2011; 31 :E6 PubMed

26 . McDonn ell MF, Glassm an SD, Dima r JR II, Pun o RM,

Johnson JR. Perioperative complications of a nter ior

pro cedu re s on the sp ine. J Bone Join t Sur g Am 19 96 ;

78:839–847 PubMed

27 . Isaacs RE, Hyde J, Goodr ich JA, Rodge rs WB, Phil-

lips FM. A prospect ive, nonran dom ized, multicenter

evaluation of extrem e lateral inte rbody fusion for the

treat m ent of adult degene rative scoliosis: periopera-

tive outcomes an d comp lications. Spine 2010 ;35(26,

Suppl):S322–S330 PubMed

28 . Sche er JK, Tang JA, Sm ith JS, et a l; Inte rn at iona l Spin e

Study Group. Reoperation rates and impact on out-

come in a large, prospective, multicenter, adult spi-

nal d eform ity dat abase: clinical article. J Neurosurg

Spine 2013;19:464–470 PubMed

29 . Pichelm an n MA, Len ke LG, Bridwe ll KH, Good CR,

O’Lear y PT, Sides BA. Revision rat es following p rim ar y

adult spinal deformity surgery: six hundred forty-

three consecutive patients followed-up to twenty-

two years postoperative. Spine 2010;35:219–226

PubMed

30 . Kelly MP, Len ke LG, Bridw ell KH, Agar wal R, Godzik J,

Koester L. Fate of the adult revision spinal deformity

pat ient: a sin gle in st it u t ion experience. Spin e 20 13 ;

38:E1196–E1200 PubMed

31. Acosta FL Jr, McClendon J Jr, O’Shaughnessy BA, et al.

Morbidity and mor tality after spinal deform ity sur-

gery in patients 75 years and older: complications

and predictive factors. J Neurosurg Spine 2011;15:

667–674 PubMed

32. Fairb an k JC, Pynse nt PB. The Oswes tr y Disabilit y

Index. Spine 2000;25:2940 –2952, discussion 2952

PubMed

33. Bagó J, Pér ez- Gru eso FJS, Les E, Her nán de z P, Pel-

lisé F. Minimal impor tan t di eren ces of the SRS-22

Patient Questionnaire following surgical treatment

of idiopathic scoliosis. Eur Spine J 2009;18:1898–

1904 PubMed

34 . Wu C-H, Wong C-B, Chen L-H, Niu C-C, Tsai T-T, Chen

W-J. Instrumented posterior lumbar interbody fu-

sion for pat ients w ith degen erative lumba r scoliosis.

J Spinal Disord Tech 2008;21:310–315 PubMed

35. Zimm er m an RM, Moham ed AS, Skolasky RL, Robin-

son MD, Kebaish KM. Fun ction al outcom es an d com -

p licat ions after pr im ary sp inal s urge r y for scoliosis

in adults aged forty years or older: a prospective




study with minimum two-year follow-up. Spine 2010;

35:1861–1866 PubMed

36 . Tsai T-H, Huang T-Y, Lieu A-S, et al. Funct ional o ut -

come analysis: instrum ented posterior lum bar inter-

bo dy fu sio n for degen er at ive lum ba r sco liosis . Act a

Neu ro ch ir (Wien) 201 1; 15 3: 54 7– 55 5 PubMed

37. Good CR, Lenke LG, Bridwell KH, et al. Can posteri-

or-only surgery provide similar radiographic and

clinical results as combined anterior (thoracotomy/

thoracoabdominal)/posterior approaches for adult

scoliosis? Spine 2010;35 :210–21 8 PubMed

38. McCar th y IM, Hostin RA, Am es CP, et a l; Int er na -

tional Spine Study Group. Total hospital costs of

surgical treatment for adult spinal deformity: an

extended follow-up study. Spine J 2014;14:2326–

2333 PubMed

39. McCarthy IM, Hostin RA, O’Brien MF, et al; Interna-

tional Spine Study Group. Analysis of the direct cost

of surgery for four diagnostic categories of adult spi-

nal deform ity. Spine J 2013;13:1843– 1848 PubMed

40. Indr aka nt i SS, Web er MH, Takem oto SK, Hu SS,

Polly D, Berven SH. Value-based care in the man-

agemen t of spinal disorders: a systemat ic review of

cost-u tility an alysis. Clin Ort hop Relat Res 2012 ;47 0:

1106–1123 PubMed

41. Kep ler CK, Wilk ins on SM, Radcli KE, et a l. Cost -

utility analysis in sp ine care: a systematic review.

Spine J 2012;12:676–6 90 PubMed

42 . Glassm an SD, Car reon LY, Sha rey CI, et a l. The cost s

and bene ts of nonoperative managem ent for adult

scoliosis. Spine 2010;35 :578–582 PubMed

43. Irw in ZN, Hilibra nd A, Gust avel M, et a l. Var iation in

surgical decision m aking for de generat ive spinal dis-

orders. Part I: lumbar spine. Spine 2005;30:2208–

2213 PubMed

44 . Sand er s JO, Haynes R, Light er D, et al. Var iation in

care among spinal deformity surgeons: results of a

survey of the Shriners hospitals for children. Spine

2007;32:1444–1449 PubMed

45. Deyo RA, Mirza SK. Tren ds a nd variation s in th e u se

of spine surger y. Clin Orthop Relat Res 200 6;443:1 39–

14 6 PubMed



! Introduction

Proxim al junct ional kyph osis (PJK) was rst

de ned and characterized in the literature by

Glattes and colleagues.1 These authors pre-

sented a retr ospective series of 81 adult defor-

m ity patients and de ned abn orm al PJK using

two criteria: (1) proximal junctional sagittal

Cobb angle " 10 degrees ( Fig. 9 .1) and (2) post-

operative proximal junctional sagittal Cobbangle at least 10 degrees greater than the pre-

operative m easuremen t. They reported an in-

cidence o f PJK of 26%,1 and have been supported

by su bsequ ent st udies re por ting ra tes of PJK

ran ging from 17 t o 61.7%.2,3 Although PJK is

generally asymptomatic, there is a subset of

pat ients (1.4–4%) w ho present w ith symptom s

requiring furt her surger y.4,5

Risk factors for the development of PJK in-

clude advanced age, surgical app roach, greater

rigidity of construct, greater magnitude of

sagitt al correction, the p resence of preexisting

proxim al kyp hosis, dam age to the pos ter ior

ligamentous complex, damage to the adjacent

facet when instrumenting the upper instru-

m en ted ver tebra (UIV), xation to th e ilium ,

type of instrumentation (hooks versus pedicle

screws), and the presence of osteoporosis.1,2,6

Given t he large num ber of iden ti ed risk fac-

tor s, the et iology of PJK is most likely mu ltifac-

tor ial in nat ure . Noneth eless, advanced age is afactor that seem s to be u niform across the m a-

jorit y of st udies. In a ddit ion, th e curre nt litera-

tur e suggests that PJK m ay be m ore prevalent

than the rates initially reporte d 20 years ago.

This chapter synthesizes our current un-

derstanding of PJK in adults by reviewing the

literatu re u nde rlying the various etiologies of

PJK, discussing the impact of PJK on clinical

outcome s, exam ining the risk factors th at lead

to revision surgery due to PJK, providing con-

sensus exper t opinion on possible m ethod s for

minimizing PJK development, and describingindications for surgical treat m ent .

! Etiology and Risk Factors

for Proximal Junctional

Kyphosis

The etiology of PJK is m ultifactorial an d can b e

divided into surgical, radiograph ic, and p atient -related factors. These are sum m arized in Table

9.1. We w ill closely exam ine the literat ure ab out

these various causes.

Surgical Factors

Disruption of the Posterior Soft Tissues

In t he ir classic pap er, Panjabi and W hite 7 high-

lighted the role of the posterior spinal liga-

m ent s in prevent ing excessive motion betw eenthe vertebrae. Given these ligaments’ role as a

stabilizer in th e spine, the disrup tion of poste-

9

Junctional Issues Follow ing AdultDeformity Surgery

Han Jo Kim, Sravisht Iyer, and Christopher I. Sha rey, Sr.



Junc tional Issues Following Adult Deformity Surgery 107

rior soft tissues has always been viewed as a

potential contribu tor to the develop m en t of PJK.

The relative contribution of the posterior

soft tissues has been examined in a cadaveric

model.8 The authors of this study performed

one of several procedures on m otion segmen ts

obtained from six hum an cadavers. These pro -

cedures included bilateral transverse h ook site

preparat ion, sublam inar h ook site p re par at ion,

pedicle screw placem ent , supra- an d in tersp i-

nous ligamen t t ransection, and tran section of all

poster ior st ructures. Follow ing these inte rven -

tions, the auth ors measured the torque need ed

to produce 2.8 degrees of angular displace-

men t, and, based on th is measuremen t, calcu-

lated the total exion sti ness of the motion

Fig. 9.1a,b The proximal junct ional sagitt al Cobb

measurem ent (proximal junct ional angle). This is

de ned as the Cobb angle between the inferior end

plat e of the upper instrum en ted vert ebra and the

superior end plate of the vertebra two levels above.

(a) Postoperative radiograph. (b) Radiograph at

6 months t hat m eet s both criteria of proximal

junctional kyphosis (PJK): proximal junct ional angle

> 10 degrees and a progression of the proximal

junctional angle > 10 deg rees.

a b

Table 9.1 A Summ ary of Various Risk Factors for PJK Proposed in the Literature

Surgical Radiographic Patient-Specif c

• Disruption of poste rior soft tissues• Rigidity of instrume ntat ion

• Combined ante rior-posterior approach/fusion

• Upper instrumented vertebrae in the upperthoracic spine

• Fusion to the sacrum• Degree of correction

% Increased lumbar lordosis% High SVA corre ct ion

% Failure to respect global sagit ta lalignment

• Increased preoperative thoracickyphosis

• Increased preope rative proximal

junctional angle

• Advanced age

• High BMI• Osteoporosis

Abbreviations: BMI, body mass index; SVA, sagit tal vertical axis.



108 Chapte r 9

segment . Their data showed that sectioning the

supra- and int erspinous ligamen ts led to a sig-

ni cant ( p = 0.02) loss of exion sti ne ss. They

found that the posterior ligamentous complex

contributed 6.59% to the sti ness of the m o-

tion segmen t and that the sti ness loss couldroughly double (12.62%) w ith t he exposure re-

quired for placement of instrum entation at th e

same motion segment .

Similarly, a biom echanical mo del developed

by Cam m ar at a et a l6 showed th at complete fac-

etectomy and posterior ligament resection both

independently increased the proximal junc-

tional kyphotic angle; com bining th e tw o pro-

cedures resulted in an even greater increase in

th e kyphot ic angle.

Denis et al9 present ed a series of 67 patient s

with Scheuermann’s kyphosis treated with an

instru m ente d fusion. They highlighted t he im -

por tance of the posterior ligam entou s st ruc-

tu res in their series. They separate d th eir cohor t

of patients w ith PJK into a group w here the fu-

sion stopped short of the proximal end verte-

bra as well as a grou p w here the proxim al end

vertebra w as include d in t he fusion. The latte r

group had three patients who developed PJK.

All thre e wer e noted t o have disrup tion of the ju nct ional ligam entum avum w ith su blam i-

nar hooks or sublaminar wires. In all, the au-

thors noted that disruption of the posterior

ligaments was implicated in 25%(5/20) cases

of PJK obser ved in th eir se ries.

Although the above ndings all hint at the

importan ce of the p osterior structures, the tru e

impact of posterior dissection has been di cult

to isolate in clinical studies. Although all sur-

geons would agree that disruption of the m us-

cular, ligamen tous, and bony t issue likely occurs

cephalad to th e UIV, the degree of disruption is

di cult to quant ify, let alone standa rdize.2 De-

spite this limitation, the spinal reconstructive

community generally agrees that damage to

the posterior soft tissues likely contributes to

th e d evelopm en t of PJK.2

Rigidity of Instrumentation

In addition to the posterior soft tissues, nu-merous investigators have commented on the

rigidity of instrum ent ation as a risk factor for

development of PJK. This has been a focus in

the eld, par ticularly in light of the m ore rigid,

all-pedicle screw constructs that have gained

in popularity over the p ast tw o decades.

The biomechanical study by Cammarata et

al6 described in the prior section nicely high-lights t he impact of increasing construct rigid-

ity on t he proxim al junct ional angle. Similarly,

Thaw rani et al10 used a porcine cadaver model

to show that transverse process hooks pro-

vided decreased sti ness compare d with an all

pedicle screw con st ruct . In t he a ll pe dicle screw

group, the majority of the motion occurred

at t he m otion segmen t im med iately proxima l

to th e UIV, wherea s this tran sition w as m ore

gradual in the t ransverse process group.

However, clinical stu dies h ave not clearly af-

rm ed the nd ings of the above biomechan ical

studies.5,11,12 Y.J. Kim et al 12 found th at the use

of all pedicle screw construct s wa s associated

with an increased rate of PJK compared with

hybrid or hook constructs ( p = 0.04), but this

di erence did not rem ain signi cant when ad-

ju st ing for age ( p = 0.33). Similarly, other pub-

lished series of adult scoliosis patients have

not demonstrated that all pedicle screw con-

stru cts were m ore likely to be associated w ithPJK tha n w ere hybrid screw–hook constr ucts.5

Hassanzad eh an d colleagues11 published a series

of 47 consecutive adult patients with 2-year

follow-up who underwent long spinal fusion

w ith h ooks or screw s at t he UIV. They found

no instances of PJK in the 20 patients treated

w ith a hook at t he UIV compare d w ith a 29.6%

(8/27 patients) rate of PJK in the screws group

( p = 0.01).

Surgical Approach

Som e t ypes of surgical approach h ave also been

associated w ith PJK. Y.J. Kim an d colleagu es12

found that a com bined anterior and posterior

approa ch was a risk factor for developmen t of

PJK, even w hen adjusted for age ( p = 0.04). This

nding has been consistently shown to be th e

case in othe r ser ies as well.13,14 In a ret rospec-

tive series of 249 patients (adults and adoles-

cents) who underwent surgery for idiopathicscolios is, H.J. Kim et al14 performed a m ultivar-

iate analysis to identify risk factors for the de-




velopm ent of PJK. They found that patient s w ho

underwent an anterior-posterior approach were

th ree tim es as likely (odds rat io [OR], 3.04; 95%

con den ce int er val [CI], 1.56–5.93) to develop

PJK compared with individuals who under-

wen t posterior-on ly fusion. Additionally, a com - bined anter ior and p osterior approach was one

of the few factors that could be consistently

identi ed as a risk in a system atic review of the

eld.2

Upper Instrume nte d Vertebrae

The cont ribut ion of the UIV to th e development

of PJK was suggested in the initial description

of the phenomenon. Glattes and colleagues1

foun d a signi cant ly higher level of PJK w he n

the instr um ent ation stopped at T3 (53%) when

comp ared w ith T4 (12.5%) ( p = 0.02). A subse-

quent larger series from the same institution

did show that an upper thoracic UIV (T2–6)

demonstrated a higher prevalence of PJK

(33.67%, 33/9 8) compa red w ith lower t horacic

and upp er lum bar UIV ( p = 0.036). 12 However,

this di erence did not rem ain signi cant wh en

adjusting for age ( p = 0.65). 12 The UIV, along

with a combined anterior-posterior approach,was only one of two independent risk factors

associated with the development of PJK in a

series of 249 pat ients.14

Bridw ell et al15 found that p atients w ith more

advanced PJK (PJK " 20 degrees) were more

likely to have a lower n um ber of levels fused (8

versus 11) and were more likely to have a UIV

in the lower thoracic spine ( p < 0.001). Ha et

al16 examined the di eren ce betw een a UIV in

the lower and upper thoracic spine an d found

that the mechanism of failure was di erent

bet ween the two scenarios . Failu re occu rre d

sooner ( p < 0.01) and was more likely to occur

due to fracture in the lower thoracic spine,

whereas subluxation was more prevalent in

the upper thoracic spine.

Other series, however, failed to identify the

UIV as a risk facto r for d evelopm en t o f PJK.5,17 A

systematic review found low-level evidence

that the UIV was am ong th e r isk factors associ-

ated w ith t he developm ent of PJK.2

The m echanism of how th e UIV m ight con-

tribute to the development of PJK is incom-

plete ly understood. Pro posed theor ies inclu de

both dam age to the adjacent facet join t that can

occur more easily in th e up per t horacic spine18

as well as the interface between the mobile

cervical and relat ively stat ic th oracic spine.1

Instrumentation to the Sacrum/Ilium

In cases of adult scoliosis, extension of the

fusion to the sacropelvis and the subsequent

increase in sti ness of the constru ct has been

thought to contribute to the development of

PJK. In th eir ser ies o f adult pat ien ts , Y.J. Kim

and colleagues12 found a higher rate of PJK in

pat ients w hose low er inst rum ented verteb ra

(LIV) was S1 compared with patients with an

LIV of L5 or above (51 %vs 3 0%, p = 0.009). This

remained a strong trend ( p = 0.059) even afte r

adju stin g for age. Yagi et al5 observed a sim ilar

tren d; in their series, fusion to the sacrum was

associated w ith a signi cantly higher incidence

of PJK (an increase of 27.6%, p = 0.02). A more

recent clinical series also found that patients

w ith PJK requ iring revision wer e m ore likely to

have fusions exten ding to the pelvis (74% vs

91%, p = 0.02). 19 Fusion to the sacru m was also

associated with an increased risk of progres-sion of PJK to greate r th an 2 0 degrees.15

Magnitude of Correction

More recently, as we have begun to un dersta nd

the import ance of global sagittal alignm ent , in-

vestigators have sought to dete rm ine if param -

eters of sagittal alignment correlate with the

incidence of PJK. In gene ral, stud ies have foun d

that an increase in sagittal balance correction

and an increase in lumbar lordosis correlate

w ith th e developm en t of PJK.5,17,19,20 The m ech-

anism underlying this increased rate of PJK is

un know n. In t he ir retrospect ive ser ies, H.J. Kim

et al19 found that patients requiring revision

sur gery for PJK ha d a lum bar lordo sis (LL) th at

was closer to t he p elvic inciden ce (PI), w he reas

th ose w ithou t PJK had a LL m uch lower t han PI.

Their ndings are sim ilar to those of Maruo

and colleagues,17 who showed that increasing

LL more than 30 degrees was associated witha signi cant ly higher incidence of PJK (58% vs

28%, p = 0.003).



110 Chapte r 9

Similarly, patients requiring revision surgery

with PJK had a lower postoperative sagittal

vertical axis (SVA) (0.8 vs 4.1 cm) and a higher

m agnitu de of SVA correct ion (9 vs 4 cm ) com-

pared w ith those w ithou t PJK.19 These nd ings

are gen erally in keep ing with th ose of Yagi andcolleagues,5 w ho saw th at a SVA correct ion

greate r th an 5 cm led to a 5 0%inciden ce of PJK

( p = 0.01). Anoth er ser ies of 54 pat ients d em on-

strat ed th at t he risk of PJK decreases by 30 %for

every centimeter increase in the C7-plumb-

line.20 In the same series, the C7-plumbline

had ret urn ed closer to its preoperat ive position

in all pat ient s by na l follow -up (average 2.23

years).

The interplay betw een th oracic kyphosis and

lumb ar lordosis, that is, the global sagitt al align-

ment (GSA), is also becoming an increasingly

import ant concept in u nde rstand ing PJK. In t he

same study where they looked at sagittal cor-

rec tion , Yagi and colleagues5 showed that a

non ideal postop erat ive GSA (th oracic kyph osis

[TK] + LL + PI > 45°) led to a 7 0%r ate of PJK ( p <

0.001). Maruo et al17 also showed that ideal

global sagittal alignm ent protected against t he

development of PJK. The importance of spinal

balance is also high ligh te d by Mendoza- Lat tesand colleagues,20 who found that the di erence

be tween TK and LL was inversely pro por t ional

to t he risk of developing PJK.

Taken together, these studies add to our

growing understanding of global sagittal bal-

ance. They suggest that the goal of restoring

th e SVA to 0 cm m ay not be opt ima l for all pa-

tient s. Inde ed, studies of asymptom atic volun-

teers h ave consistent ly shown increased age to

be corre lat ed w ith great er SVA.21,22 They also

highlight the need for further studies to es-

tablish the optimal spinopelvic parameters

for surgeons to target (i.e., to understand the

di eren ce bet we en LL + 9° and LL – 9°).19

Radiographic Factors

Preoperative Thoracic Kyphosis

High preoperative TK may predispose to PJK

in both adult and pediatric populations. Inthe adult population, Maruo and colleagues17

showed that preoperative TK greater than 30

degrees was a risk factor for developing PJK

(62% vs 29%, p = 0.002). Similarly, Mendoza-

Latt es and colleagues20 found th at patients with

PJK ha d a larger d i ere nce be twe en TK an d LL

at baseline ( p = 0.012). These patients also

prese nted w ith low er sacral slope and signs of pelvic ret roversio n.

Proximal Junctional Angle

Lee and colleagues ,23 one of the rst groups to

descr ibe PJK in pat ient s w ith idiopat hic scolio-

sis, found that a preoperative PJ angle greater

than 5 degrees was a risk factor for develop-

ment of subsequent junctional kyphosis. This

work h as been supp orted by Denis et al’s9 se -

ries of Scheuer m ann patient s. The large m ajor-

ity of cases of PJK obser ved in t hat ser ies were

noted w hen the p roximal extent of the fusion

did not include t he kyph otic proxim al end ver-

tebra. Maruo and colleagues17 were able to

demonstrate that a proximal junctional angle

(PJA) greate r t han 10 d egrees ( p = 0.016 ), in ad-

dit ion t o a PI > 55° ( p = 0.037), was a r isk factor

for t he developm en t o f PJK.17

Patient Factors

Patient -speci c factor s such as advanced age,

high body mass index (BMI), the presence of

osteoporosis, smoking, and the presence of

other comorbidities are always important to

consider prior to spine surgery. Not surpris-

ingly, m any of th ese factors have been linked to

th e developm en t of PJK. In t he adu lt liter atu re,

increasing age has been associated with the

incidence of PJK and PJK requiring revision in

num erous case series.1,2,15,19,24

Interestingly, the link between high BMI

and the development of PJK is less clear in the

literature. Bridwell et al15 were able to show

th at h igher BMI ( p = 0.015) and the presen ce of

a comorbidity ( p = 0.001) were associated w ith

th e developm en t of PJ angle > 20 degrees . How-

ever, other series from the same institution

have failed to show th e sam e link betw een high

BMI and PJK (de ned in th e trad itional m ann er

using 10 degrees as a cuto ).19,24

Given that a large proportion of PJK occurs

du e to fract ure at t he UIV,17 i t is not sur prising




that osteoporosis plays a critical role in the

developm en t of PJK. In o ne series, osteop orotic

pat ients ove r the age of 65 w ho underwent a

m inimum ve-level fusion were found to have

pedicle and compression frac ture s at a rat e of

13%, w ith PJK occurr ing in 26%of pat ien ts .25 Ina small series of 10 adult patients, Watanabe

and colleagues26 found th at osteopenia and p re-

operative comorbidities were common among

pat ients w ith proxim al verteb ra l fra ct ure an d

subluxation. Case series in ad ult pat ients h ave

found that osteoporosis is m uch m ore preva-

lent in individuals with PJK than in those

without.24

! Timing of Proximal

Junctional Kyphosis

The m ajority of th e cases of PJK are ident i ed

w ithin the rst year postoperat ively.5,14 The

pat ients w ho do go on to develop PJK progress

to abou t on e ha lf (53%) of th eir tot al degree of

PJK by 3 months.5 Sim ilar ly, Y.J. Kim an d col-

leagues12

repor ted th at 59%of progression ofthe PJA occurs within the rst 8 weeks. Maruo

et al17 rep orted that 62%of cases with PJK were

ident i ed with in 8 weeks, w ith fractu re being

the m ost comm on cause.

! Clinical Outcomes Afte r

Proximal Junctional

Kyphosis

Clinical outcomes after PJK are summarized in

Table 9.2. In gener al, m ost stud ies have show n

that most cases of PJK are asymptomatic, and

that this condition does not substa nt ially alter

clinical outcome. There are, however, two re-

cent series th at do show that patient s with PJK

have increased p ain levels compared w ith those

pat ients w ithou t PJK.19,24 The di erence in pain

levels between the t wo groups met the m ini-m al clinically import ant di erence.24 The in-

cidence of symptom atic pain (i.e., upp er back

pain re por ted by the pat ient at follow-up) w as

also markedly higher in the group with PJK

(29 .4% vs 0.9%, p < 0.001). These more recent

results highlight the importance of furthering

our understanding of PJK; as we shall see in

the following sections, the early descriptionsof PJK as a rad iograph ic nd ing th at wa rra nt s

follow-up m ight be understating the t rue im-

pact of t he condit ion.

Investigators are now turning a closer eye

to t he concept of symp tom atic PJK to see h ow

these cases might impact clinical outcomes.

Yagi an d colleagu es5 found t hat th eir patients

w ith symptom atic PJK had a signi cantly higher

Oswestry Disability Index (ODI) score com-

pared w it h pat ien ts w it hou t PJK.5 Similarly,

H.J. Kim et al19 foun d lower ODI and pain score s

in patients with PJK and lower pain scores in

pat ients undergoing revis ion for PJK. Impor -

tantly, they found lower outcomes across all

domains of the Scoliosis Research Society (SRS)

questionn aire in patient s with PJK, though th ese

outcome s did not reach statistical signi cance.

! Revision Surgery forProximal Junctional

Kyphosis

The m ajority of cases of PJK are a symp tom atic

and d o not require inter vention. Repor ted rates

of revision d ue to PJK have ra nged from 1.4 to

11.2%w ith pain being the most comm on rea-

son for revision.4,19,27,28

Severe cases of PJK can lea d to signi can t

sagittal imb alance an d disability. In a sm all se-

ries of 10 patients, Watanabe et al26 reported

vertebral subluxation and severe neurologic

de cit in two of the ir 10 patient s as a result of

progressio n of PJK.

Hart et al28 reported on a case series from

the Invasive Species Specialist Group (ISSG) da-

tabase. Their de nition of proximal junct ional

failure (PJF) was “severe PJK,” which was furt he r

de ned as a change of more than 10 degrees of

kyphosis between the UIV and the vertebratwo levels above the UIV (UIV +2), along with

one or more of the following: fracture of the



114 Chapte r 9

vert ebral bod y of UIV or UIV +1, post er ior os-

seoligamentous disruption, or pullout of in-

stru m ent ation at the UIV. In t heir ser ies, they

identi ed 57 patients from a series of 1,218

consecutive adult patients wh o met this de -

nition of PJF (4.68 %). Of th e 5 7 ca ses of PJF, 27(47.4%) un der wen t revision su rgery w ithin 6

m onth s of the ir index ope ration. Of the causes

of PJF, fract ure was t he m ost com m on (56%),

followe d by soft tissue failure (3 5%) an d screw

pullout (9%). Of the r isk factor s ident i ed for

revision surgery, a combined an ter ior/posterior

approach ( p = 0.001) an d h ighe r PJK angu lation

( p = 0.034) were found to be signi cant . Of th e

various modes of failure, only the presence

of traum atic mechan ism of failure, wh ich oc-

curred in six patients, was deemed to predis-

pose p at ients to a revis ion ( p = 0.019). A higher

SVA also tren ded towa rd be ing a signi cant

p redictor of re vision ( p = 0.090) along with

fem ale sex ( p = 0.066). The overall rate of revi-

sion su rgery w as 2.21%.

A similar st udy w as cond ucte d by Yagi and

colleagues4 as p art of the Comp lex Spine Study

Group. They utilized a consecutive series of

1,668 p atients t reated for adult spinal defor-

m ity and greater than ve levels of fusion.The ir series had longer follow-up (2 to 12 year,

m ean 4 .3 years), and they also focused on pa-

tient s older th an 50 at th e time of surgery. They

de ned PJF as PJK requ iring revision an d iden -

ti ed 23 p atien ts (1.4%). The large m ajority of

th ese 23 patien ts (17 patient s, 74%) had be en

revision surgery cases at the tim e of their long

segment fusion and all had received posterior

ped icle scr ew con st ruct s. Osteop enia w as prev-

alent (10/23, 43 %), but, inte rest ingly, non e of

these patients h ad osteoporosis. Sixteen patients

were revised for intolerable pain, another six

for a neurologic de cit, an d one for head pt osis.

The a uth ors found that a m ajority of these fail-

ures occurred early, with a mean time to PJF

(revision) of 10.5 ± 9.3 m onth s; 87%h ad be en

revised within 2 years of surgery. H.J. Kim et

al19 reported that a higher lum ber lordosis and

an increase d SVA correct ion ar e r isk factors for

revision due to PJK. They reported a revision

rate of 10 .7%.Moving forward, important work remains

to be d one regard ing our u nd erst an ding of PJF.

A classi cat ion system for PJK an d a clear de -

nition of PJF are needed before we progress

tow ard d e nin g the r isk factors for PJF. Add i-

tionally, the optimal magnitude of correction

also rem ains to be determ ined.

Preve nting Proximal Junctional

Kyphosis

To date, no de nitive m ethods have been de-

scribed to prevent PJK, although several ap-

proaches h ave bee n su ggested . Two studies h ave

repor ted on t echnical tricks to reduce the inci-

dence of PJK.11,29,30 Hassanzadeh et al 11 found

no insta nces of PJK in 20 patient s treate d w ith

a hook at the UIV compare d w ith a 29.6%(8/27

pat ien ts) rat e of PJK in pat ien ts w ho were

treated with a pedicle screw at the UIV ( p =

0.01). Additionally, they found that patients

w ith hooks had signi cantly higher funct ional

scores compared with those w ith screws ( p <

0.01). These data have not been replicated to

date.

Given that vertebral fractures represent a

com m on et iology for PJK an d PJF, invest igators

have also studied the impact of prophylactic

one- and two-level vertebroplasty above longfusions. Results rep orted include both biome-

chanical29 and clinical 30 data. In a biom echani-

cal mo del using 18 cad averic spines, Kebaish e t

al29 were able to show a signi cant redu ction

in vertebral compression fractures when two-

level vertebroplasty (UIV and UIV +1) was

comp ared w ith on e-level (UIV only) or no ver-

tebroplasty. A clinical series from the same

group followed 38 patien ts w ith tw o-level ver-

teb rop last y (UIV an d UIV +1) for 2 years .30 They

re por te d a low er r ate o f PJK or PJF (PJK 8%, PJF

5%, com bined 13%) th an previou sly published

rates. Their stu dy did n ot include a control co-

hor t (i.e., patients w ho h ad n ot received verte-

bro plast y), and did not sh ow any signi cant

di erences in clinical outcomes between the

group s w ith an d w ithou t PJK or PJF.

Fina lly, Yan ik et al31 reported on a series of

60 patient s treated for Scheuer m ann k yphosis.

To reduce th e sti ness of the proximal con-

struct, they studied the impact of leaving twoscrew t hreads out of the posterior cortex whe n

placin g pedicle screws at the UIV. They theo-




rized that this would reduce the sti ness at the

proxim al aspect of t he con st ruct . At an ave rage

of 2-year follow-u p, they found that the screws

w ith thread s left out of the cortex ha d a lower

PJ angle (4.44 ± 1.55 d egrees vs 8.08 ± 2.96 de -

grees) when compared with standard pediclescrew insertion ( p = 0.001). This grou p also h ad

no cases of PJK, compared with a 17.2%(5/29,

p = 0.02) rate with the standard screw tech-

nique. Finally, they were also able to show an

improveme nt in the p hysical compon ent of the

Short Form 36 score in the group treated with

the modi ed screw insertion.

Revision Strategies

The approa ch to t he patient w ith PJK is sim ilar

to the app roach to th e patient w ith other sagit-

tal plane deform ities. Our indications for a re -

vision op erat ion are as follows:

1. Progressive deform ity

2. Pain th at has failed non operative measures

for man agem ent

3. Implant prominence with imminent skin

bre akdow n

4. Neurologic de cit or cord comp ression

The decision ab out th e UIV level selection is

based on several facto rs, bu t generally sp ea k-

ing, PJK cases w ith a UIV in t he lower th oracic

spine should be extended u p to the upper th o-

racic spine, and cases w ith a UIV in th e u ppe r

thoracic spine should be extended up to T1–2

or into th e cervical spine.

Osteotomies may also be necessary in the

treatm en t of PJK. Osteoto my selection is based

on th e r igidit y of PJK. Flexib le PJK (Fig. 9 .2) can

usually be treated without an osteotomy, or

with a posterior column only osteotomy (Fig.

9.2d), w herea s rigid d eform ities (Fig. 9 .3) may

necessitate a three-column osteotomy (i.e.,

pedicle su btract ion ost eot om y or vertebral

colum n resection) (Fig. 9.3c). Flexibility can be

assessed with hyperexten sion or supine radio-

graphs as well as the scout images on some

comput ed t om ography (CT) scans (as long as a

head support was not used du ring the scan).

For cases where a neurologic de cit orsymptomatic cord compression is present, a

vertebral column resection m ay be necessary

to decompress the kyphotic area where the

cord compression is likely to be present. Ana-

tomic realignment is essential to relieve the

cord compression.

The goal for the revision operation should

avoid the temptation for overcorrection; in-stead , the goal sho uld b e an SVA close to 4 to

5 cm. Overcorrection can lead to a recurrence

of PJK and necessitate ad ditional operat ions and

unn ecessary risk for pat ients.

! Chapter Summary

Proximal junct ional kyphosis is de ned u sing

two criteria: (1) proximal junctional sagittal

Cobb angle " 10 de grees, and (2) postoperative

proxim al junct ion al sagit tal Cobb angle at least

10 degrees greater than the p reoperative mea-

surement .1 PJK is gene rally an ea rly postop er -

ative phenomenon, and most cases typically

are recognized in the rst year after surgery.5,14

Rates of PJK reported in the literature range

from 17 t o 61 .7%.2,3

Proxim al junct ional kyph osis is likely multi-

factorial in origin. Risk factors for the devel-opment of PJK can be categorized as surgical,

radiographic, and patient-related factors. Ad-

vanced age appears to be the most important

pat ient-re late d r isk factor. Surgical r isk factor s

to consider include greater m agnitude of sag-

ittal correction, dam age to the adjacent facet

wh en instr um ent ing the UIV, dam age to the

posterior ligam entou s complex, a com bined

anter ior and posterior appr oach, xation to the

ilium, and certain t ypes of instr um ent ation. Of

the ab ove, the m agnitude of correction is par-

ticularly im por tan t, as an SVA of 0 cm m ay not

be op t im al for all pat ients.21,22 A higher post-

operative LL and an SVA correction of greater

than 5 cm have both been associated with th e

development of PJK.17,19 To date, no de nit ive

methods have been described to prevent PJK.

Some investigators have described technical

tr icks to red uce t he inciden ce of PJK.11,31

Patients with PJK are generally asymptom-

atic. However, recent stud ies have show n t hatthese patients may have increased pain levels

and worse functional outcome measures.19,24



116 Chapte r 9

Fig. 9.2 a–d A pat ient present ing with PJK who was

treate d with extension of the posterior fusion and

posterior ost eotom ies only. (a) Late ral radiograph.

(b) Hyperextension view, clearly showing a exible

deformity. (c) Computed t omography (CT) scan.

(d) Preope rative (left) and postoperative (right)

standing radiographs.

a b

c d




In the subset of patients who are symptom-

atic, pain is the m ost com mon complaint and

th e m ost com mon reason for revision. Repor ted

rates for revision for PJK range from 1.2 to

11.4%.4,19,27,28 Patien ts requ iring revision for PJK

are gene rally approache d similarly to patient s

with other sagittal plane deformities. Osteoto-

m ies may be necessary and are chosen based on

the rigidity of the PJK. When revising PJK, the

tem ptat ion for overcorrect ion should be avoided.

Fig. 9.3 a–c A pat ient with a rigid deformity.

(a) Preoperat ive image s of the proximal

kyphotic de formity. (b) CT scan and recum bent

lm show a solid fusion extending up to the

uppe r instrume nted vert ebra (UIV) with a rigid

deformity. This patient required a three-columnoste otomy to correct the PJK. (c) Pre- and

postrevision radiog raphs.

a

b c



118 Chapte r 9

Pearls

The appropriate magnitude of correction and

global sagitt al alignme nt are critical to a chieving

successful outcomes and avoiding the develop-

me nt of PJK. An SVA of 4 cm is a reaso nab le go al,especially for those patient s over 60 years of age.

Proximal junct ional kyphosis is an early postoper-

ative ph enom enon, a nd m ost cases can t ypically

be observed with in 2 to 6 months post operatively.

Similarly, cases of junctional failure typically are

evident within the rst year.

When revising cases of PJK, it is crucial to con-

sider the rigidity of th e deform ity, as exible PJK

may be treated with instrumentation and fusion

only, witho ut an ost eoto my or poste rior column

osteot omies only.

Pitfalls

Careful att ent ion m ust be paid to global sagitt al

alignment; attempting to aggressively correct all

pa tients to an SVA of 0 cm has be en sho wn to

pre dispose patients to PJK. Althoug h no consen sus exists on prevent ing PJK,

surgeons must pay close attention to the integ-

rity of the posterior soft tissues and the rigidity

of the construct, and m ust select an appropriate

UIV. Failure to consider t hese factors may lead to

the developm ent of PJK.

Although most cases of PJK are asymptomatic,

these pat ient s may have increased pain and worse

functional scores. They must be followed regu-

larly to ensure stable kyphosis and acceptable

outcomes.

References


1 . Glatt es RC, Bridw ell KH, Len ke LG, Kim YJ, Rinella A,

Edwards C II. Proximal junctional kyphosis in adult

spinal deform ity following long instru m ente d poste-

rior spinal fusion: inciden ce, outcomes, and risk fac-

tor ana lysis. Spine 2005 ;30:1643 –1649 PubMed

2. Kim HJ, Len ke LG, Sha rey CI, Van Alst yn e EM, Skelly

AC. Proximal junct ional kyphosis as a distinct form

of adjacent segment pathology after spinal defor-

m ity surgery: a system atic review. Spine 201 2;37(22,

Suppl):S144–S164 PubMed

3 . Lee JH, Kim JU, Jan g JS, Lee SH. Analysis of t he inc i-

dence and risk factors for the progression of proxi-

m al junct ional kyphosis following surgical treat m ent

for lumbar degenerative kyphosis: minimu m 2-year

follow-u p. Br J Neurosu rg 2014 ;28:252 –258 PubMed

4. Yagi M, Rahm M, Gaines R, et a l; Comp lex Spin e

Study Group. Characterization and surgical outcomes

of proximal junctional failure in surgically treated

pat ien ts w ith ad ult sp inal de for m it y. Spin e 2 01 4; 39 :

E607–E614 PubMed

5. Yagi M, King AB, Boach ie-Adjei O. Incid en ce, risk

factors, and natural course of proximal junctional

kyphosis: surgical outcomes review of adult idiopa-

thic scoliosis. Minimum 5 years of follow-up. Spine

2012;37:1479–1489 PubMed

6. Cam m arat a M, Aub in CE, Wang X, Mac-Thion g JM.

Biomechanical risk factors for proximal junctional

kyphosis: a detailed numerical analysis of surgi-

cal instrum entat ion variables. Spine 20 14;39:E500–

E507 PubMed

7. Panjabi MM, White AA III. Basic biom echa nics of th e

spine. Neurosurgery 1980;7:76–93 PubMed

8 . Ande rso n AL, McI TE, Ash er MA, Bur ton DC, Glatte s

RC. The e ect of posterior th oracic spine an atom ical

structu res on mot ion segment exion sti ness. Spine

2009;34:441–446 PubMed

9. Denis F, Sun EC, Winter RB. Incide nce a nd risk factors

for proximal and distal junctional kyph osis following

surgical treatmen t for Scheuerm ann kyphosis: m ini-

m um ve-year follow-up. Spine 2009;34: E729–E734

PubMed

10 . Thaw rani DP, Glos DL, Coom bs MT, Bylski-Aust row

DI, Sturm PF. Transverse process hooks at upper in-

strumented vertebra provide more gradual motion

tran sition than p edicle screws. Spine 2014; 39:E826–

E832 PubMed

11. Hassanzadeh H, Gupta S, Jain A, El Dafrawy M, Sko-

lasky RL, Kebaish KM. Type of an chor at t he proxim al

fusion level has a signi cant e ect on the inciden ce

of proximal junctional kyphosis and outcome in

adults after long po sterior sp inal fusion. Spine Defor-

mity 2013;1:299–305

12 . Kim YJ, Bridw ell KH, Len ke LG, Glat te s CR, Rhim S,

Cheh G. Proximal junct ional kyphosis in adult spinal

deformity after segmental posterior spinal instru-

m entation and fusion: m inim um ve-year follow-up.

Spine 2008;33:2179–2184 PubMed

13 . Wan g J, Zhao Y, Shen B, Wang C, Li M. Risk fact or

analysis of proximal junctiona l kyphosis after poste-

rior fusion in patients with idiopathic scoliosis. In-

ju ry 201 0; 41 :4 15 –4 20 PubMed

14 . Kim HJ, Yagi M, Nyugen J, Cun nin gha m ME, Boach ie-

Adjei O. Com bined a nter ior-poster ior surger y is the

most important risk factor for developing proximal

junct ional kyphosis in id iop at h ic sco liosis . Clin Or-

thop Relat Res 2012;470 :1633–1 639 PubMed

15 . Bridwe ll KH, Len ke LG, Cho SK, et al. Proxim al ju nc-

tional kyphosis in pr imary adult deformity surgery:




evaluation of 20 degrees as a critical angle. Neuro-

surgery 2013;72:899–906 PubMed

16 . Ha Y, Maru o K, Racin e L, et al. Proxim al jun ct ional

kyphosis and clinical outcom es in adult spinal defor-

mity surgery with fusion from the thoracic spine

to the sacrum : a comp arison of proximal and distal

upper instrumented vertebrae. J Neurosurg Spine

2013;19:360–369 PubMed

17. Maru o K, Ha Y, Inou e S, et al. Predictive facto rs for

proxim al junct ion al ky ph os is in lon g fu sio ns to the

sacrum in adult spinal deformity. Spine 2013;38:

E1469–E1476 PubMed

18. Helgeson MD, Shah SA, Newt on PO, et al; Harm s

Stud y Group. Evaluation of pr oxim al junct ional ky-

phos is in ad ole sce nt id iop at h ic sco liosis follow ing

pe dicle screw , hook, or hybrid in st rum en tat ion . Spine

2010;35:177–181 PubMed

19 . Kim HJ, Brid we ll KH, Len ke LG, et a l. Patien ts w ith

proxim al junct ional kyp hos is r equ ir ing revision su r-

gery have higher postoperative lumbar lordosis and

larger sagittal balance corrections. Spine 2014;39:

E576–E580 PubMed

20 . Mend oza- Lat tes S, Ries Z, Gao Y, Weinste in SL. Proxi-

mal junctional kyphosis in adult reconstructive

spine surgery results from incomplete restoration

of the lumbar lordosis relative to the magnitude of

the thoracic kyphosis. Iowa Orthop J 2011;31:199–

20 6 PubMed

21 . Vedan ta m R, Len ke LG, Kee ne y JA, Brid we ll KH. Com -

par iso n of st an ding sagit tal sp inal al ign m en t in as -

ymptom atic adolescents and adults. Spine 1998;23:

211–215 PubMed

22. Gelb DE, Lenke LG, Bridwell KH, Blanke K, McEnery

KW. An analysis of sagittal spinal alignment in 100

asymptomatic middle and older aged volunteers.

Spine 1995;20:1351–1358 PubMed

23 . Lee GA, Bet z RR, Clem en ts DH III, Huss GK. Proxim al

kyphosis after posterior spinal fusion in patients w ith

idiopathic scoliosis. Spine 1999;24 :795–79 9 PubMed

24. Kim HJ, Bridwe ll KH, Len ke LG, et al. Proxim al ju nc-

tional kyph osis results in inferior SRS pain subscores

in adult deformity patients. Spine 2013;38:896–901

PubMed

25. DeWald CJ, Stanley T. Instrumentation-related com-

plica t ion s of m ult ilevel fu sio ns for ad ult spinal d efor-

mity patients over age 65: surgical considerations

and treatment options in patients with poor bone

quality. Spine 2006;31(19 , Supp l):S144– S151 PubMed

26. Wata na be K, Len ke LG, Brid we ll KH, Kim YJ, Koest er

L, Hensley M. Proximal junctional vertebral fracture

in adults after spinal deform ity surgery using ped icle

screw const ruct s: analysis of m orph ological featu res.

Spine 2010;35:138–145 PubMed

27. Ream es DL, Kasliw al MK, Smit h JS, Ham ilton DK,

Arlet V, Sha rey CI. Time to developm en t, clinical and

radiographic characteristics, and management of

proxim al junct ion al kyphos is follow ing a dult thor a-

colum bar instr um ente d fusion for spinal deformity. J

Spina l Disord Tech 2 014 PubMed

28. Hart R, McCar thy I, O’Brien M, et al. Iden ti cation o f

decision criteria for revision surgery am ong patient s

with proximal junctiona l failure after surgical treat-

ment of spinal deformity. Spine 2013;38:E1223–

E1227 PubMed

29. Kebaish KM, Martin CT, O’Brien JR, LaMotta IE, Voros

GD, Belko SM. Use of vert ebrop lasty to preven t

p roxim al junct ion al fr ac ture s in ad ult d eform it y su r-

gery: a biome chanical cadaveric study. Spine J 2013;

13:1897–1903 PubMed

30. Martin CT, Skolasky RL, Mohamed AS, Kebaish KM.

Preliminar y results of the e ect of proph ylactic ver-

tebroplasty on the incidence of proximal junctional

complications after poster ior spinal fusion to th e low

thor acic spine. Spine Deform ity 2013;1:1 32–138

31 . Yan ik HS, Kete nci IE, Polat A, et al. Prevent ion of

p roxim al junct ion al kyph os is a fter p oste rior s urger y

of Scheuermann kyphosis: An operative technique.

J Spinal Disord Tech 2014 Jul 29. [Epub ahead of

p rin t] PubMed



! Introduction

The rst application of a metallic implant to

the human spine was reported by Hadra1 in

1891. Silver w ires were placed in th e t horacic

spine for treatment of a spinal fracture. The

most important historical event in spinal re-

construction surgery was the invention of the

Harrington instrumentation in the middle of

the 20th century.2 Paul Harrington developedthese spinal im plants, consisting of hooks and

rods m ade of stainless steel, for treat ment of se-

vere spinal deformity and fracture-dislocations

of the spine. Since the introduction of his de-

vice, surgeries to correct and stabilize the

spine with m etallic imp lants have undergone

dramatic development, and surgeries using

metallic implants to correct and stabilize the

dam aged spine became kn own as spinal instru-

m ent ation surgery. The Harrington instr um en-

tation surgery was m odi ed by his successors,

who added sublaminar wires, tapes, and pedi-

cle screws.3 Since 2000, pedicle screws and

rods have been widely used for spinal defor-

mity surgery due to their biomechanical su-

periorit y. Of the biom at erials used for sp inal

deform ity surgery, titanium alloys are th e m ost

pop ular m at erial at the prese n t t im e due to

their imp roved biocompatibility and because

they en tail fewer m etal-related art ifacts in mag-

netic resonance imaging (MRI).4 Historically,stainless steel and cobalt-chromium (Co-Cr)

were discovered much earlier than titanium

alloys (Table 10 .1). Spinal implants made of

stainless steel or Co- Cr are cur ren tly used for

pat ient s w it h r igid sp in al cu rves due to their

superior mechanical properties to titanium

alloys.

There is no consensu s on how t o select m e-

tallic m aterials for spinal deformity correction

today, and spine surgeons de pen d on th eir per-

sonal experience through t heir m edical career.This chapter discusses met allic spinal imp lants

and the biomechanics of the deformity cor-

rection of the spine. It is imperat ive th at pra c-

titioners be fam iliar w ith th ese topics, as they

are indispensable in providing patients with

safe and e ective spinal deform ity correction.

! Mechanical Properties

of Metals

Metals have a com m on patt ern of stress–strain

curve consisting of th e elastic deform ation zone

and the plastic deform ation zone. In t he elastic

deformation zone, the stress–strain relation-

ship is linear and th e m etal deform s in propor-

tion to the applied force (Fig. 10.1). After the

yield point, the stress–strain curve of metals

becom es non linea r. If increas ing force is applied

to a me tallic imp lant, the implant w ill reach th eultim ate strength and nally ruptu re or break.

10

Biomechanics and Material Sciencefor Deformity Correction

Manabu Ito , Yuichiro Abe, and Remel Alingalan Salming o



Biom echanics and Mate rial Science for Deformity Correction 121

Although all meta ls follow th e sam e pat tern of

stress–strain relationship, the re are di eren ces

in their yield strength and ultimate strength

and t he slope of the stress–strain curve. Within

the elastic deform ation zone, met als are able

to retur n to th e original shape a fter the force

is removed. Once metals were over-bent to

th eir plastic deform ation zone, the m etal is notable to return to the original shape and p erm a-

nent deformation of metallic implants occurs.

If plastic deform ation of th e rod s occurs after

scoliosis corre ction , signi cant loss of correc-

tion may result, and the original purpose of

correcting the spinal deform ity would not be

ful lled. For this reason, it is important for

spine surgeons to know how much force is

put on the sp inal im plan ts during corre ct ion

pro cedure s and m echanica l re spon ses of the

m etallic implants t o the forces created by cor-

rection procedures.

The m etals curren tly used for spinal defor-

m ity surgery include stainless steel, pure tita-

nium, titanium alloys, and Co-Cr alloys. The

m echanical prope rt ies of each met al are shown

in Fig. 10.2 . Commercial pure titanium (cpTi)

has four grades based on its mechanical prop-

ert ies. Grade 1 h as th e h ighest value of elon-

gation at break, but it has the lowest tensile

strength. As the grade increases, the tensilestrengt h increases, and t he cap ability of elonga-

Table 10.1 History of Metallic Implants Used for

Spine Surge ry

Year Event

1890s Application of sliver wires for spinal

fractures1910s Development of stainless steel

1920s Development of cobalt-chromium

alloys (Vitallium® in 1932)

1940s Development of SUS316, 317 stainlesssteel (Harrington instrumentat ion)

1960s Development of titanium alloys(Grade 1–4: com me rcially pure

titanium, other titanium alloys:Ti-6Al-4V, Ti-6Al-7Nb,

Ti-6Al-2.5Fe, Ti-13Zr-13Ta, et c.)

Abbreviation: SUS, steel use stainless.

tion decreases. The tensile strength of Ti alloys

is m uch higher th an th at of pure titanium , but

the elongation capacity of Ti alloys is the low-

est am ong all the t ypes of titanium . Although

titanium rods are bent by surgeons during sur-

gery to the desired contour, the mechan ical sti -

ness of titan ium decreases signi cantly around

the bending points. Co-Cr shows the highest

tensile strength and relatively high break p oint

for elongation. This sti er m echanical proper ty

of Co-Cr is favored by sp ine su rgeon s for t reat -

m en t of rigid spina l deform ities. Sta inless steel

(grade, SUS316 L) is a litt le we aker in its t en sile

strength compared with Co-Cr, but it shows

m uch bet ter elongation dura bility. New stain-

less steel m aterials w ith higher ten sile strengt h

have been invente d re cently and w ill be avail-able for surger y very soon.

Fig. 10.1 The st ress–strain curve of a typical

struct ural me tal. Line A shows the apparent st ress

and line B shows the true stress. Point 1 shows the

ultimate strength, and point 2 shows the yield

strength . A material demonstrate s rupture at point

3. The area of linear relationship between the stress

and strain indicates the elastic deformation region

before point 2. After point 2, t he st ress increases

up to t he ultimat e te nsile strengt h (point 1) in

region 4. Beyond point 1 , a neck forms where thelocal cross-sectional area signi cant ly decreases

and t he m aterial becomes weaker in region 5.



122 Chapte r 10

! Changes in Mechanical

Properties of Rods

After Manual Bending

(Table 10 .2)

Because t he m etallic rods supp lied by m edical

device companies for use during surgery are

straight, the surgeon needs to bend them by

hand to the desired contour just before per-

form ing the correction procedure. With regard

to the rods’ mechanical properties, the yield

strength of titanium alloy rods with a 5.5-mm

diameter decreases from 803.9 N to 324.0 N

(40.3%) after t hree -point ben ding by 20 d e-

grees.5 The yield strength of titanium alloy rods

w ith a 6.0-m m d iameter is reduced 54.1%after20-degree three-point bending. Co-Cr alloy

rods with a 6.0-m m diameter showed the sam e

tendency as the 6.0-mm titanium alloy rods,

w ith their yield signi cantly decreasing to 56.4%

after 20-degree three-point bending. If multi-

ple rod bending op erat ions were perform ed,

even the m echan ically sti est Co- Cr rods can

exhibit a signi cant reduction in their mechan-

ical prope rt ies. Spine sur geons should be aw are

of the reduction in mechanical properties of

each m etal to avoid m echanical failures of rods

Fig. 10.2 The relationship bet ween tensile strength

and elongation at t he break of each met al. There are

four grades for comm ercial pure titanium (cpTi),

with small di erences in mechanical prope rties.

Though titanium alloys show greater te nsile

strength than cpTi does, the break points of

titanium alloys under elongation are much lower

than those of cpTi. Sta inless steel shows the highest

capab ility for elongat ion and Co-Cr shows the

highest tensile st rength among all.

Table 1 0.2 Changes in Mechanical Properties After Rod Bending5

Yield Strength (N) No Bend

Bend Back

One Time

20-Degree

Bend

40-Degree

Bend

6.0-mm Ti rod 1,004 748 (74.5%) 544 (54.1%) 509 (50.6%)

6.0-mm Co-Cr rod 865 689 (79.6%) 488 (56.4%) 476 (55.0%)

Sti ness (N/m m)

6.0-mm Ti rod 160 151 (94.6%) 143 (89.2%) 120 (74.9%) 6.0-mm Co-Cr rod 317 278 (87.8%) 261 (82.3%) 208 (65.8%)

Abbreviations: Ti, titanium; Co-Cr, cob alt-chromium.




dur ing surgery and postoperative follow-u p pe-

riods before solid bony fusion is obtain ed.

The mechanical forces on spinal implants

decrease over time after surgery, as biological

bon y fusio n m at ure s and becom es solid . The

maximum force on the spinal implants mayoccur d uring the correction procedure . Precise

in-vivo forces on spinal implants during cor-

rection p rocedures are still unkn own. Previous

biom echanica l st udies have t r ied to m ea su re

in-vivo forces on rods by various engineering

methods.6,7 It is di cult to obt ain in-vivo dat a

during operative procedures due to ethical re-

strictions, and r eliable in-vivo data in spinal de-

formity correction are lacking. Medical devices

such as m etallic rods and screws for spinal de-

form ity should be designed and man ufactured

based on the biom echanica l and biologica l en-

vironment in which they will be used in the

human body. Reliable in-vivo biomechanical

information regarding the force on spinal im-

plants during spinal d eform it y corre ct ion m ay

advance the safety and e ectiveness of defor-

m ity surgery in the future.

! Viscoelasticity of the Spine

The spinal column, consisting of bone, liga-

m en ts, and inter vertebral disks, is a comp osite

material with signi cant viscoelasticity. Vis-

coelastic materials show two biomechanical

characteristics: creep ph enom enon and stress

relaxation. The creep phenomenon states that

when stress is held constant, the strain on the

m aterial increases w ith t ime. Stress relaxation

states that wh en t he strain is held constant, the

stress decreases with t ime.

Considering th e viscoelastic prope rt y of the

spine, rapid correct ion procedures such as quick

rod rotation m aneuvers may m ake th e spine

m uch sti er and may hinder e cient spinal

correction, resulting in a lower correction rate

than the surgeon anticipated preoperatively.

Thu s, dest abilization p rocedu res, such as bilat-

eral facetect om ies, diskectom ies, an d release of

costotransverse ligaments, are very importantin obtaining better correction. Besides these

technical issues, the biomechan ical pr inciples

indicate that slow rod rotation an d tran slation

procedures can ob tain be t ter nal correct ion

rates. After the correction procedure is com-

plete d, the ro d w ithin an elas t ic deform at ion

range will tend to spring back to the original

shape due to the stress relaxation e ect aftercorrection procedures (Fig. 10.3). Fast rod ro-

tation m ay cause signi cant increase of m e-

chanical loads on the implants and result in

dramatic changes in the shape of the rods. If

the forces on the rods were within the elastic

deform ation zone of met al, the rods wou ld still

have a potent ial to retu rn to the original shape.

Spine surgeons shou ld be fam iliar with the m e-

chanical character istics of m etals and t he sp inal

column to obtain better correction rates and

provide pat ients w ith sa fe surgery.

! Correction Procedures for

Adolescent Idiopathic

Scoliosis

There are m any surgical procedures to correct

adolescent idiopathic scoliosis (AIS) reportedafter the introdu ction of the Harrington instru-

mentation. The Cobb angle correction on an-

teroposter ior (AP) radiographs was 40% w ith

Harrington rods, 55%with the dual-rod multi-

hook system (CD, Cotrl-Dubousset instru m ent ),

and 65% with th e dual-rod multiple pedicle

screw constru cts in th e coronal plane. Recent

studies reported t hat p edicle screw constructs

in the sagittal plane increased the lordosis of

the thoracic spine.8 While the surgeon is per-

form ing the direct vertebr al rotation techn ique

to decrease rotational deformity around the

apex of the thoracic curve, the m ajor force on

the spine pushes the th oracic rib hump down

to lessen the rotational deform ity of the spine,

wh ich eventu ally causes dekyphosis in the th o-

racic spine .9

There have been several attempts to create

thoracic kyphosis by posterior spinal instru-

mentation surgery. Because a titanium rod is

mechanically weaker than a stainless steel orCo-Cr rod, some surgeons ut ilized stainless steel

or Co-Cr rods rathe r th an t itanium rods so as



124 Chapte r 10

not t o yield to th e force on the rods. Anothe r

m ethod was to use an in-situ rod bending tech-

nique after a single rod-rotation to create tho -

racic kyph osis. One of th e prob lems of an in-situ

rod-bending procedure is that a m uch greater

load would be applied to the pedicle screws

around the area of rod bending, which may

increase the possibility of vertebral fractures

or screw loosening due to a higher concentra-

tion of mechanical force on the screws. Also,

multiple rod-bending procedures will reduce

the original mechanical strength of the metal-

lic implant .

There have been new surgical techniques

for correcting spinal deformity to maintain or

create th oracic kyphosis. One technique is th e

vertebral coplanar alignment (VCA) reported

by Vallesp ir et al.10 This technique u ses slotted

tubes attached to each pedicle screw on the

convex side of the th oracic cur ve. Two longitu-

dinal rods are inserted an d separated along theslots, driving the t ubes into on e plane, ma king

th e axis of th e vertebrae coplanar, th us correct-

ing transverse rotation and coronal translation.

For creating th oracic kyphosis, the e nds of the

tubes are spread in the thoracic spine. After

locking a de nitive rod on the concave side

and retrieving tubes on the concave side, the

convex-side rod is inserted and tightened . The

curve correction rate in the m ain thoracic curve

was 73%on average, and the average preope ra-

tive th oracic kyphosis of 18 degrees re m ained

un changed after surger y. Anoth er techn ique is

the simultaneous translation technique using

two rods as reported by Clement et al.11 This

technique uses polyaxial pedicle screws and

polyaxial claw s consist ing of a ped icle h ook and

an opposing transverse counter-hook placed

at th e m ost cephalad end of the rod. The t wo

6.0-mm t itanium rods are bent rst and are in-

serte d pre-or iented. Redu ction of the d eform ity

is obtained by gradu al and alternate t ightening

of all nuts on both rod s, allowing the vertebrae

to gradually approach the rods. Another tech-nique uses a Universal clamp consisting of a

woven polyester band, a t itanium alloy clamp,

Fig. 10.3 The shape changes of the t wo rods on

both side s of the curve before (red) and just aft er

(blue) rod rotation and 1 week after surge ry (white).

The original contours of the two rods showed

signi cant reduct ion just after the rod rotation

procedure. The rods, however, tended to spring

back to the ir original shapes as long as the forces

were within the elastic deformat ion zone of the

meta l (titanium a lloy). The m echanical stress on

rods tends to decrease with t ime due to the

stress-relaxation e ect of the spine.




and a locking screw as w ell as pe dicle screws.12

Pedicle screws were placed in two or m ore ver-

tebrae at th e distal extrem ity of the curve with

m onaxial screws on t he convex side an d poly-

axial screws on th e concave side. Thoracic levels

were instrumented with three to seven sub-lam inar Universa l clamp s (UCs) on th e concave

side an d one sublam inar UC at the apex on th e

concave side. Correction of the thoracic curve

was p erform ed using posterom edial translation

by t igh tening the UCs for the spine t o appro ach

the pre-ben t double rods.

The kinem atic concept of how to correct th e

deform ity with p edicle screws is show n in Fig.

10.4 . The ap ex verteb ra should be m oved from

ante rior to posterior and from lateral to me dial

with two anchor points corresponding to the

tips of pedicle screws. The screw tip on the

concave side should be moved more posteri-

orly than that of the convex side. By providing

a bigger bend to the concave side rod than to

the convex side rod and rotating the two rodssimultaneously, the screw tip of the concave

side at the apex of the curve m oves m ore pos-

teriorly than that of the convex side does. This

technique w as reported by Ito et al13 and named

the simultaneous double rod rotation tech-

nique, which allows simultaneous correction

of the coronal plane d eform ity and restoration

of the thoracic kyphosis. The biomechanically

strongest correction procedu re, which utilized

a solid frame between the pedicle screws on

Fig. 10.4 The locations of the apex vertebra in the

axial plane. The apex vertebra is located antero-

laterally before surgery and the vertebra is to be

relocated post eromedially during surgery. The

pedicle screw at t he apex vert ebra on the concave

side of the curve should be m oved m ore toward

the back (red line) than that on t he convex side (blue

line) to relocate t he apex vertebra to the normal

position. In the simultaneous double rod rotat ion

technique,13 the concave side rod should be bent

more than the convex side rod, which allows the

head of the pe dicle screw on t he concave side t o

move m ore toward the back than that on the

convex side.



126 Chapte r 10

bot h the convex and concave sid e of t he cu rve,

was reported by Chang and Lenke.14 By con-

necting the t wo rods w ith a solid meta l frame,

the spinal implant is able to produce the most

pow erful force on the spin e, but the s t re ss con -

centration m ay not occur on speci c pediclescrews.

! Intraoperative Mechanical

Forces on Rods During Rod

Rotation Maneuvers

During rod rotation procedures to correct the

deformed spine three dimensionally, the in-

serted rods frequently show dram atic contour

changes due to signi cant mechanical loads

on th e rods and screws. A recent biom echani-

cal study has used nite elemen t analysis of

the mechanical properties of metals and the

geometrical changes of metallic rods before

and after surgery.15–17 These auth ors measured

three-dime nsional (3D) contour chan ges of the

rods before and after surgery using postopera-

tive 3D computed tomography (CT) images.

Speci c m athem atical assumptions and bound-

ary conditions were put into the computer sim -ulation . In the ir nite eleme nt ana lysis (FEA)

m odel, the distal end of the rod was xed com-

plete ly, an d the upperm ost end was ab le to

m ove p arallel to the axis of the t run k. By calcu-

lating the forces on rods in AIS patients with a

single thoracic curve, pullout forces of about

150 N were exerte d on t he concave side screws

aroun d th e apex of the cu rve if pedicle screws

were inserted at all the fusion levels (implant

dens ity 100%) (Fig. 10.5). At bot h e nds of the

rods, push-in forces of about 200 N were on the

pedicle screw s on the concave side of the cu rve.

Push-in forces on ped icle screws ra rely result in

clinical problem s, but pullout forces on screws

can lead to screw loosening or bony fractures,

wh ich can create serious complications for th e

Fig. 10.5 The force on each pedicle screw during

rod rotat ion p rocedure in an ado lescent idiopathic

scoliosis (AIS) patient. The contour of the concave

side rod shows signi cant reduct ion, and pullout

forces around the apex of the curve have reached

about 200 N according to the calculation. At both

ends of the concave side rod, maximum pushing-in

force was exert ed on the pe dicle screws. The m inus

sign indicates pull-out force on each screw. The tot al

amoun t of the force acted on t he concave side rod

has topped 1,400 N.




spinal cord or great vessels. According to the

FEA m odel, the pu llout forces on ped icle screw s

around t he ap ex of the curve m ay exceed 500 N

if the n um ber of pedicle screws were redu ced.

Maximum pullout forces of thoracic pedicle

screws in cadaveric spines are about 600 N, sothat there may be an increasing risk of screw

pullouts a ro und the apex in pat ients w ith rigid

curves and fewer pedicle screws.

One of the e ective solut ions to reduce me-

chanical forces on pedicle screws is to place

cross-links or a fram e over the t wo rod s. Many

rod rotation procedures utilize only the con-

cave-side rod for correction of scoliosis, and

the convex-side rod is placed after comp letion

of rod rotation of the concave side. In these

correction procedures, the force on the con-

cave side rod is m uch higher th an th at on th e

convex side rod. The re sults show ed th at 50%of

the rod contour on the concave side was lost

after rod rotation m aneuvers and the opposite

side rod showed a lmost no chan ge in its shape .

If a cross-link was placed between the two

rods, 15%of the total load was shared by the

convex side rod, which showed some contour

changes. The biomechanically strongest con-

struct with pedicle screws is bilateral screw p lacem ent w it h a cross- lin k, w hich for m s a

triangular shape in each vertebra. From this

biom echanica l poin t of v iew, verteb ra l colum n

manipulation with a rigid frame cross-linking

the two rods and pedicle screws is the most

pow erful corre ct ion pro cedure in sp inal defor-

m ity surgery.14

! Implant Density andCorrection Rate

According to the present consensus among

spine experts worldwide, AIS curves of less

than 70 degrees w ith exibility need less tha n

80%of screw density.18 From a biome chanical

standpoint, a small number of anchors with

less rigid m etallic rods are not a ble to correct

the spinal deform ity su cient ly because they

can sustain only small amoun ts of mechan icalload. It seem s reasonable to assum e th at m ore

implant density and more rigid rods will pro-

vide bette r correction rates and nal outcom es.

Several recent stud ies, however, found that the

nal outcome of the surgical correction did

not show any signi cant improveme nt even if

surgeons used m ore rigid and thicker rods w ith

a higher implant density.19,20 Implant densityand mechanical sti ness of the rods are of

some importance to obtain better correction

rates. The more impor tan t steps to a ect the

nal outcome of deformity correction may be

perform ing pre op erat ive cu rve exib ilit y and

destabilization pr ocedures, such as Ponte oste-

otomies, before performing correction proce-

dures including rod rotation and t ranslation.

! Chapter Summary

Harrington started to use his spinal implants,

consisting of hooks and rod s m ade of stainless

steel, for treat m ent of spinal deform ity almost

50 years ago. Since the introdu ction of his de-

vice, surgeries to correct and stabilize the spine

with m etallic implants have shown dram atic im-

provem ent in thre e-dim ensio nal corre ct ion of

the cu rves; an excellent correction rate has beenobtained in recent years by using pedicle screws

and r igid rods. Titanium alloys are th e most pop -

ular material for recent spinal surgery due to

their biocompatibility and fewer m etal-related

artifacts on MRI. However, bending a titanium

rod multiple times may lead the material to

plast ic deform at ion, w hich m akes it signi -

cantly weaker m echanically. Surgeons should

be fam iliar w ith the m ech an ica l chara cter ist ics

of each m aterial used for deform ity surgery and

refrain from excessive manual bending of the

rods to maint ain the original me chanical prop-

erty of each metal. Spinal implants made of

stainless steel or Co-Cr are commonly used for

correction of rigid spinal deformities, such as

severe scoliosis and rigid kyphosis, because of

their superior mechan ical sti ness and ability

to obtain b ette r correction rates. Surgical treat-

m ent of spinal deformity requires fam iliarity

with spinal biomechanics and the mechanical

characteristics of each biomaterial. This chap-ter discussed the fundamen tals of biom echan-

ics of spinal deformity correction, mechanical



128 Chapte r 10

behaviors of m et allic implan ts, in-vivo forces o n

rods and pedicle screws, biom echanical bene-

ts of destab ilization procedures for rigid curves,

and viscoelastic proper ties of the spine. Reade rs

can ap ply the concepts of spinal biom echanics

and material science to their own correction procedures to p rovide their pat ien ts w ith a safe

and e ective surgical procedure an d excellent

clinical outcomes.

Pearls

Spinal deformity correction relies on metallic

imp lants such as screws, hooks, and rod s.

Popular metals used for spinal implants are tita-

nium alloys, stainless steel, and cobalt-chromium.

Each me ta llic mate rial has di eren t mechan ical

behaviors for th e st ress–strain relat ionship and

repetitive loading.

The viscoelastic property of the spine should be

considered when operating on cases with rigid

large spinal deform ity.

Pitfalls

Mechan ical load s on spinal imp lants d uring cor-

rection procedures often exceed the limit of

bo ne st i ness, which may lead to surge ry-relate d

complications.

Di erent deformity correction procedures have

be ne ts and limitat ions.

Rapid rod rotation for rigid deformity correction

will signi cant ly increase mechanical loads on the

spinal imp lant.

References.


1. Hadra BE. Wiring of the vertebrae as a means of

imm obilization in fracture and Pott’s disease. The

Times and Register, Medical Press, Philadelphia,

1891:1–8

2. Harrin gton PR. Treat m en t of scoliosis. Corr ect ion and

intern al xation by spine instru m ent ation. J Bone

Joint Surg Am 1962;44- A:591–6 10 PubMed

3. Suk SI, Lee CK, Kim WJ, Chu ng YJ, Par k YB. Seg-

men tal pedicle screw xation in the treatment of

thoracic idiopathic scoliosis. Spine 1995;20:1399–

1405 PubMe d

4. Uhtho HK, Bard os DI, Liskova-Kiar M. The advan-

tages of titan ium alloy over stainless steel plates for

the inter nal xation of fractures. An experim ent al

study in dogs. J Bone Joint Surg Br 1981;63-B:427–

48 4 PubMed

5. Dem ura S, Murakam i H, Hayashi H, et al. The in u-

ence of rod contouring of di eren t spinal constru cts

on strength an d sti ness. Orthope dics, in press

6. Wang X, Aub in CE, Crand all D, Labelle H. Biom ech an -

ical modeling and analysis of a direct incremental

segmental translation system for the instrumenta-

tion of scoliotic deformities. Clin Biomech (Bristol,

Avon) 2011;2 6:548–55 5 PubMed

7 . Aubin CE, Labelle H, Chevr e ls C, Desr och es G, Clin J,

Eng AB. Preoperative planning simulator for spi-

nal deformity sur geries. Spine 2008;33 :2143–2152

PubMed

8. Sucato DJ, Agrawal S, O’Brien MF, Lowe TG, Richards

SB, Lenke L. Restoration of thoracic kyphosis after

operative treatment of adolescent idiopathic scolio-

sis: a mu lticenter comparison of three surgical ap-

p ro aches . Spin e 2 00 8; 33 :2 63 0– 26 36 PubMed

9 . Lee SM, Suk SI, Chu ng ER. Direct vert ebra l rot ation :

a new technique of three-dimensional deformity

correction with segmen tal pedicle screw xation in

adolescent idiopathic scoliosis. Spine 2004;29 :343–

34 9 PubMed

10 . Vallespir GP, Flores JB, Trigu eros IS, et a l. Ver teb ra l co-

plan ar align m en t: a st an da rd ized te chniqu e for th ree

dimensional correction in scoliosis surgery: techni-

cal description and preliminar y results in Lenke typ e

1 curves. Spine 2008;3 3:1588– 1597 PubMed

11. Clem en t JL, Cha u E, Geo ray A, Vallad e MJ. Sim ult a-

neous translation on two rods to treat adolescent

idiopathic scoliosis: radiograph ic results in coronal,

sagittal, and transverse plane of a series of 62 pati-

ents w ith a m inimu m follow-up of two years. Spine

2012;37:184–192 PubMed

12. Mazd a K, Ilha rr ebo rd e B, Even J, Lefevre Y, Fitou ssi F,

Penn eçot GF. E cacy and safety of post erom edial

translation for correction of thoracic curves in ado-

lescent idiopathic scoliosis using a new connection

to th e spine: the Universal Clamp. Eur Spine J 2009;

18:158–169 PubMed

13. Ito M, Abum i K, Kotan i Y, et al. Sim ulta ne ous dou ble-

rod rotation technique in posterior instrum entation

surgery for correction of adolescent idiopathic sco-

liosis. J Neurosurg Spine 2010;12 :293–30 0 PubMed

14. Chan g MS, Len ke LG. Ver tebr al de rot ation in a doles-

cent idiopathic scoliosis. Oper Tech Orthop 2009;

19:19–23

15. Salmin go R, Tad an o S, Fujisak i K, Abe Y, Ito M. Cor rec-

tive force analysis for scoliosis from implant rod

deformation. Clin Biomech (Bristol, Avon) 2012;27:

545–550 PubMed




16 . Salm ingo RA, Tad an o S, Fujis aki K, Abe Y, Ito M. Rela-

tionship of forces acting on implant rods and d egree

of scoliosis correction. Clin Biomech (Bristol, Avon)

2013;28:122–128 PubMed

17. Salm ingo RA, Tad an o S, Abe Y, Ito M. In ue nce o f im-

plan t r od cu rvat ure on sagit tal co rre ct ion o f sco liosis

deformit y. Spine J 2014;14:143 2–1439 PubMed

18. de Kleuver M, Lew is SJ, Ger m scheid NM, et al. Opti-

m al surgical care for adolescent idiopathic scoliosis:

an international consensus. Eur Spine J 2014;23:

2603–2618 PubMed

19. Prin ce DE, Matsu m oto H, Cha n CM, et al. The e ect of

rod diameter on correction of adolescent idiopathic

scoliosis at two years follow-up. J Pediatr Orthop

2014;34:22–28 PubMed

20. Che n J, Yang C, Ran B, et al. Cor re ction of Len ke 5

adolescent idiopathic scoliosis using pedicle screw

instrum entation: does implant density in uence the

correction? Spine 201 3;38:E946–E951 PubMed



! Introduction

Adult spinal deformity (ASD) is among the

m ost challenging pathologies treate d by spine

surgeons. Preoperative preparation to correct

neu ral elem ent compression as well as coronal

and sagittal malalignment requires strict at-

tent ion to detail and n ear-pe rfect technical ex-

ecut ion. Surger y to treat ASD can be fru strat ing,

however, as a seemingly “perfect” surgery maystill be complicated in th e both n earan d long-

term per iods by comp lications requiring unan -

ticipated revision surger y. Two such com m on

complications are pseudar throsis and infection.

Our group has reported a 9%rate of unantici-

pat ed se cond pro cedure s in adult s undergoing

prim ar y surgery for ASD, where pseuda rthrosis

(4%) and infect ion (1%) were com m on causes

for revision.1 Even more trou blesome w as a 21%

rate of repeat revision surgery, where pseu-

dar th rosis (5%) an d in fect ion (2%) we re again

tw o comm on causes.2 These nu m bers are in line

with reports from other cohorts; thus, pseu-

dart hrosis remains a comm on cause of revision

surgery in prim ary and revision ASD.3,4

! Pseudarthrosis

Successful fusion requires optimal biologicaland biomechanical environm ent s, wh ich rely on

bot h pat ient select ion and su rgical technique.

Although advances in imp lant te chnology and

osteobiologics have improved union rates, pseu-

darthroses persist in as many as 16%of three-

column osteotomy patients.5,6 This is a r esult

of the risk factors for pseu dart hrosis that e xist

by the nat ure of ASD surgeries. Lon g-segm ent

fusions are at a higher risk for pseudarthrosis

because of the large su rface area of bon e that

m ust hea l. In add ition, the com monly employed

m idline approach is associated w ith a den er-vation of the paraspinal musculature, with an

associated d ecrease in vascularity to th is mus-

culature. This is an insult to the local healing

environment. ASD surgeries are not uncom-

m only associated w ith spinal stenosis, requir-

ing decomp ression, or r igid coronal and sagitt al

plane deform it ies, re qu ir ing posterior colum n

osteotomies. The greater the amount of bone

resected, the greater the theoretical risk of a

pseuda rthrosis d evelop ing. Desp ite m od er n im-

plants, a long segm en t fusion w ill rem ain som e-

what mobile, through elastic deformation of

the constru ct. This microm otion, in som e cases,

m ay be excessive, w ithout the rigidity required

to promote fusion. In the absence of a fusion

mass, implants will almost certainly fracture.

This m ay occur a s early as 6 m onth s to 1 year,

though w e h ave en countered pseudarthrosis

that present ed a late as 7 years postope ratively

(Fig. 11.1). Finally, ASD often involves fusion

of junctional segments (e.g., lumbosacral junc-tion, thoracolum bar junction), wh ich are at a

higher risk of nonu nion.6 In som e cases, ante-

11

Pseudarthrosis and InfectionMichae l P. Kelly and Sigurd Berven



Pseudarthrosis and Infect ion 131

rior column support, through either a trans-

foraminal lumbar interbody fusion (TLIF) or

an anter ior lum bar in terbody fusion (ALIF), mayassist w ith imp roving un ion rates. Approp riate

use of recombinant human bone morphoge-

netic protein-2 (rh-BMP2) at t he lumbosacral

ju nct ion m ay obv iat e the use of TLIF/ALIF at

L5-S1.

Nico t in e exp os ure is associated w it h de-

creased fusion rates in spine surgery. In the

case of ASD, the risk of nonunion with multi-

level surgery is great a nd, in ou r p ractice, sur-

gery is not o ered to patients wh o are unable

to cease nicotine use. To test for nicotine use,

we routinely check urine cotinine levels. This

m etabolite of nicotine is excreted in t he u rine

and o ers reliable values for active and pa ssive

exposure to cigarette sm oke, m aking it a suit-

able screening test for this patient popu lation.

Nicot ine has proven an t ian giogenic e ect s on

fusion m asses, increasing th e likelihood of non-

union. Furth erm ore, nicotine has been shown to

have an adverse in uence on patient-repor ted

outcomes in several areas of spine surgery,independent of other risk factors, further de-

creasing th e p otent ial for success with surgery.

Given the costs an d r isks associated w ith spinal

deformity surgery, nicotine cessation is neces-

sary. In most cases, we require 3 months of preop erat ive abst inence, to prove that the ces-

sation is lasting.

Osteop orosis is increasingly com m on in ASD

pat ients, as t he age of p at ients seeking surgery

rises. Perhap s even m ore comm on is hypovita-

minosis D, which has been shown to be com-

mon in general orthopedic surgery practices

and in spine surger y–speci c practices.7 Vita-

m in D plays an essent ial role in bone m etabo-

lism and hom eostasis, and low vitam in D levels

are associated w ith osteom alacia (hypominer-

alized bone). We routinely check serum 25-

hydroxyvitam in D levels at pre ope rative visits,

and w e prescribe supplemen tation with oral

vitamin D as needed. In most cases, 50,000

Int ern ation Units (IU) w eekly for 6 we eks, fol-

lowed by 1,000 IU daily. In addition, we rec-

ommend supplementation with 1,000 mg of

calcium daily. In some cases, hypovitaminosis

D exists du e to som e oth er system ic pathology,

rather than malnutrition. For these patients,we consult en docrinologists w ith a par ticular

interest in bone m etabolism .

Fig. 11.1 A 50-year-old woman prese nte d 8 years

afte r L3 pedicle subtract ion osteotomy with broken

implants and pseudarthrosis causing xed sagit tal

plane m alalignm ent . She was treated with revision

posterior spinal fusion; not e t he four rods spanning

the level of the pedicle subtraction.



132 Chapte r 11

As previously mentioned, however, osteo-

poros is is a com m on com orbidit y encountere d

by sp in e su rgeon s. This is of conce rn to su r-

geons, as bone mineral density (BMD) is di-

rectly correlated with insertional torque and

pullo ut st rengt h. St rengt h of screw purchase ,in tu rn, a ects fusion rates by deter m ining the

dura bility and r igidity of the constru ct and its

stability while the fusion mass matures. We

routinely obtain bone densitometry tests such

as dual-energy X-ray absorptiometry (DEXA)

scans preoperatively. These tests provide the

tr ue value of th e BMD and th e T-score, which

is the number of standard deviations above

or below the mean for young-adult reference

stan dard s. Osteop orosis is de ned as a T-score

of –2.5 or lower. Osteopen ia exists at betw een

–1 and –2.5 st an dard deviat ions. When order-

ing a DEXA scan, one must remember that

in many cases the lumbar spine BMD may be

falsely elevated due to spondylosis, with end-

plate sclerosis and os teop hyte form at ion (Fig.

11.2). The DEXA scan shou ld include den sities

at the hip and distal radius. We have found

th e d istal radius p art icularly useful for m aking

the diagnosis of osteoporosis, which facilitates

pharm acologica l m an agem ent of t he diso rd er.

Although a multitude of options exist for

the management of osteoporosis, only one

anabolic therapy exists, teriparatide (Forteo,

Eli Lilly, Ind ianap olis, IN).8 Ter iparat ide is a re-combinan t form of a port ion of the p arathyroid

hormone (PTH). Although PTH increases os-

teoclastic act ivit y, via osteoblast signaling, pul-

satile administration of teriparatide increases

osteoblastic activity to a greater degree, in-

creasing bone format ion. An adverse e ect of

teriparatide observed in animal models was os-

teosarcoma formation, however, and patients

with risk factors for osteosarcoma, including

Paget’s disease and prior radiation therapy,

should avoid teriparatide exposure. In cases

of a contraindication to teriparatide, we often

emp loy denosum ab, a m onoclonal antibody that

binds re ceptor act ivat or of nuclear-' B (RANK)

ligand (RANKL). The b ind ing o f RANKL prevents

RANK activations, thereby suppressing osteo-

clast activation. This is an “anti-catabolic” mech-

anism of osteoporosis management, however,

similar to bisphosphonates. Both denosumab

and bisphosphonates delay callus maturation

in fracture models, which likely behave simi-larly to a fusion m odel, but t here is no hu m an

evidence to support a correlation between ex-

posure t o t hese ant i-catab olic m edica tions and

pse udarth ros is.8 Nonetheless, we attempt to

avoid bisphosphonate exposure early in the

spine fusion process, as there are animal data

to support a negative e ect of bisphospho-

nates. There is some eviden ce, however, that a

combination of teriparatide and bisphospho-

nate therapy may be ideal to maximize callus

form ation and rigidity, w ith early teriparatide

administration followed by conversion to bis-

phosphon at e therap y.

Surgical techn iques may play a role in instr u-

menting the osteoporotic spine as well. Spe-

cially designed screw t hread s have been show n

to increase insertional torque, which may ben-

e t the durability of a construct in osteopo-

rotic bone. Hydroxyapatite coating has also

been sh ow n to in crease screw purchase . As the

pedicle is often patu lous in the os teoporot icspine, one may choose to avoid tapping the

channel of the pedicle screw. If tapping is per-

Fig. 11.2 Upright and supine radiographs of a

65-year-old woman with degenerative lumbar

scoliosis. Note the hypertrophic osteoarthritis

through the apex of the deformity, which would

cause a falsely elevated bone mineral density.




formed, it should undersize the anticipated

screw diameter by 1 mm or more. Screw di-

ameters should be maximized, at least 70%of

pedicle diam et er, t o ensu re an ap pro priate in -

terference t w ithin the pedicle. Screw length

should be maximized and, in extreme cases,one m ay choose to tap th e anterior cortex and

achieve b icort ical purchase of the screw, as this

w ill signi cantly improve pullout stren gth. One

should attem pt to leave the dor sal cortex int act

as well, as this may minimize screw toggling

and loosening.

Implant materials are a matter of prefer-

ence and are d ebated am ong spinal deform ity

surgeons. We prefer to use 5.5-mm cobalt-

chromium (Co-Cr) rods in cases of degenera-

tive scoliosis with poor bon e qu ality. In larger,

sti er reconstructions, we commonly use a

6.0-m m Co-Cr rod on t he “working side,” and

place a 5.5-m m Co-Cr on t he contralat era l sid e.

Som e surgeons pr efer com m ercially pure t ita-

nium rods, w hich are a bit less sti th an Co-Cr,

believing that th is w ill st re ss t he bon e–implant

interface less, thereby de creasing th e incidence

of adjacent segment problem s. Conversely, some

surgeons prefer 6.35-mm stainless steel (SS)

rods, believing this is the most durable metalfor spinal deformity surgery. In a series of spi-

nal deformity patients, implant fracture was

less comm on am ong those xed w ith Co-Cr

rods.9 In cases of thr ee-colum n osteotom y (i.e.,

pedicle su btra ct ion osteotom y, vertebra l col-

um n resection), dominoes are used to span t he

osteotomy level with more than two rods, in-

creasing the r igidity of the constru ct at th e os-

teotomy site and decreasing micromotion. In

any case, secure xation, with accurate screw

placem ent , is re qu ired alon g the lengt h of the

deformity. This increases the rigidity of the

construct , decreasing nonun ion rates. We per-

form high-den sity instru me ntation (1.8 screws/

level) in the vast m ajorit y of ASD surger ies. Al-

though this issue is debated regarding adoles-

cent idiopathic scoliosis, we feel strongly that

high-den sity instr um ent ation is needed in ASD.

In all cases instrumented to S1, we place

distal supporting screws, often S2-alar-iliac

(S2AI) or iliac screws.10 Iliac screws d istal to S1decrease strain on S1 xation in exion, m ini-

mizing micromotion at the lumbosacral junc-

tion. S2AI screws have been proposed as an

alternative to iliac screws, with a m ore vent ral

starting point and “in-line” tulip heads, facili-

tating rod placemen t. Midter m results of S2AI

screws are encouraging, and we employ this

technique frequently. A third option for distalxation is S2-alar screw s. These screw s o er

suppor t to S1 pullout, with out crossing or im-

mobilizing the sacroiliac joint. If one chooses

to use this method, attention must be paid to

the course of the L5 root, as it courses over the

front of the sacral ala. Bone graft materials

should consist of locally obtained autograft,

allograft, and iliac crest bone graft (ICBG) or

rbBMP-2. We routinely use fresh-frozen can-

cellous allograft, as its osteoinductive proper-

ties are likely better than those of cortical

chips. In long fusions, inadequate volumes of

ICBG may indicate the use of rh-BMP2 in an

o -label fashion. We have shown this to be an

e ective met hod of improving fusion rates. We

neithe r use nor advocate other so-called osteo-

biologic prod uct s, as the evidence to su ppor t

their use is weak. The dorsal elements should

be decort icated, u sing gauges or burs, to fre sh

bleeding bo ne. Bleeding bo ne is a re qu isite for

compet itive fusion rates. If using m ore th an t worods or cross-links, they are xed after bone

graft placement, to minimize disruption and

interference w ith contiguous grafting.

In the vast m ajority of cases, the diagnosis

of pseudarth rosis is made with t he p resence of

loose or fractured implants. We intervene w hen

ther e is a p rogression of deformit y or p ain. In

the absence of progression, we often observe

unilateral rod fractures. When both rods have

fractu red, we usually recomm end revision sur-

gery. In a small number of cases, a pseudar-

throsis may present as symptoms consistent

with neu ral elem ent compression (e.g., radicu-

lopathy, claud ication) due to scar tissue/ brous

callus accumu lation (Fig. 11.3). Comp uted to-

m ography (CT), w ith ne cuts and m etal sub-

traction, is the p referre d im aging m odality for

diagnosing pseudarthrosis.11 Flexion/exten sion

radiographs have been used, though they are

not as sen sitive as CT scann ing.

Management of pseudarthrosis consists ofrevision surgery. The use of rh-BMP2 in pos-

terolateral fusion revisions has received Food



134 Chapte r 11

and Drug Adm inistrat ion (FDA) app roval. Wh enrevising a pseudarthrosis case, inspection of

the entire fusion m ass is recom men ded, as one

pse udarth ros is m ay beget anot her.12 In our

experience, most pseudarthrosis presents at

L5-S1, L4-L5, and at the thoracolumbar junc-

tion, in that order. Revision surgery often con-

sists of anter ior colum n sup port , if not already

done, in the form of a TLIF, ALIF, or transpsoas

interbody fusion. Options for graft material

include t itan ium , polyethe ret he rketone (PEEK),

and allograft. Our preferen ce is to use t itanium ,

as it bond s to bon e, increasing the stability of

the construct and likely improving fusion rates.

The p seudar throsis tissue sh ould be debrided,

as these tissues are unable to undergo mine ral-

ization and m ust be re m oved to achieve union.

In th e lumbar spine, we often p erform a posterior

column osteotomy through the pseudarthrosis

tissue an d compress through th e pseuda rthrosis.

In addition to encouraging fusion by exposure

of bleeding cancellous bon e, the osteotomy aidsin restoration of lordosis, taking tension o of

the fusion mass and prom oting compression at

Fig. 11.3 A 61-year-old man presente d with a

sagitt al plane malalignment and L4 radiculopat hy

due to pseudarthrosis. Note the vacuum disk within

the instrument ed levels, indicative of a

pseudarthrosis.

th e pseudar th rosis level. If rh- BMP2 is not used ,

we u se autogenou s iliac crest graft, as th e bio-

logic activity of allograft is not su cient to

create a comp etitive environm ent for fusion in

the setting of a previous pseudarthrosis. For

cases of lumbosacral junction pseudarthrosis,we ensure that we have adequate distal xa-

tion, consisting of ped icle screws at th e rst

sacral vert ebra an d iliac screws below t hat. In

the case of patulous, eroded S1 pedicles, we

place m ult ip le iliac screw s on ea ch sid e.

The postoperative routine is unchanged in

the m anagem ent of pseudart hrosis. No bracing

is used and th e patients are mobilized on post-

operative day 1. Strict activity precautions

are established, however, and patients are in-

structed on how to safely rise from bed and

are given a front-wheeled walker to use for 6

weeks to encourage an upright posture, dis-

couraging exion throu gh the revision fusion

m ass. Teriparatide the rapy is continued for a

minimum of 3 m onths after surgery. We do not

perform CT scanning for evaluat ion of fusio n

masses, as the exposure to radiation is exces-

sive, and ndings are un likely to a ect our man-

agem ent of the patient.

Careful attention to detail in preoperative planning, intraoperat ive p er form an ce, and post -

operative rehabilitation helps surgeons mini-

m ize pseudar throsis in their pract ice. Sm oking

cessation is absolutely mandatory to ensure

competitive results. Appropriate choices of bi-

ologics and instrumentation help increase fu-

sion rates. However, a nonunion rate of 0%is

not realistic, and t he informed decision-m aking

pro cess shou ld include d iscussion of t he r isk o f

reop erat ion for non un ion in ASD surger y.

! Infection

Perioperative surgical-site infection (SSI) is a

signi cant cau se of m orbidit y in ASD surger y,

with reported rates ranging from 0.3% to

20%.1,2,13 These rates include both super cial

and deep wou nd infections; the treatm ents and

imp lication s for each di er. In many inst an ces,a super cial infection can be managed on an

outpatient basis, with oral antibiotics alone.




Conversely, a deep wound infection is a ca-

tastrophe, nearly universally managed with

rehospitalization, revision surgery, and pro-

longed intr avenou s ant ibiotics followe d by oral

antibiotics.

A review of the Scoliosis Research Society(SRS) Morbidity and Mortality database re-

vealed an overall infection rate of 2.1%in cases

perfor m ed by part icip at ing m em bers.13 This

cohort include d a heter ogeneous m ix of proce-

dures, including n oninstrum ented degenerative

lumbar surgeries, in addition to instrum ented

ASD procedures. Not surprisingly, less exten-

sive surgeries, including lumbar diskectomies

and minimally invasive TLIF procedures were

associated with lower rates of infection. Those

cases that w ere associated w ith instrumen ted

spinal fusions had higher rates of infection.

Neurom uscular scoliosis (1 4%) and pos tlam i-

nectom y kyph osis (5.1%) had the highest rates

of postoperat ive wou nd infection. Revision sur -

geries were m ore com m only a ected by infec-

tion s (3.3%vs 2.0%), an d d eep infect ions we re

m ore comm on in this situation as well.

Although the SRS database provides good

epidemiological data regarding rates of infec-

tions, it does not provide the granu lar data thatallow for ner conclusions regarding postoper-

ative infections in th e ASD popu lation. Several

smaller series have reviewed the rates of re-

oper ation for pat ients und ergoing ASD recon-

structions. Pichelmann et al1 published our

group’s exper ience w ith p rima ry ASD surger-

ies, not ing a rate of reoper ation for infection o f

1.4%and account ing for 15.5%of revision sur-

geries. One must note that this rate is likely

lower th an the t rue value, as super cial infec-

tions are unlikely to have undergone reopera-

tion. As a follow-up study, we reviewed the

rates of unan ticipated reoper ation for revision

ASD proced ures, noting infection in 14 %of re-

vision surgeries performed.2 These rates are

similar to those presen ted by others, with in-

fect ion accoun ting for 15%of revision surger ies

in ASD.3

Adult spinal deformit y surgeries are at h igher

risk for perioperative infection than other or-

thop edic or neurologic surgeries because of thelong durat ion, high est imated blood losses, large

surface ar eas, and exte nsive u se of implant s. It

stands to reason that t he longer a w ound is ex-

posed to air, t he m ore likely som e contam ina-

tion may occur. High estimated blood losses

are often associated with the need for periop-

erative allogeneic transfusions. Although allo-

geneic transfusions are associated with majorcomplications, such as t ransfusion-related acute

lung injury (TRALI), m ore com m on a re p er iop-

erative infections, including SSI, urinary tract

infections, and respiratory tract infections.14

The exposure t o allogeneic blood an d prote ins

is associated with an immunomodulatory ef-

fect that may depress the immune response

to pathogens, making the patient susceptible

to SSI. This relationsh ip has bee n sh own in less

extensive lumbar degenerative fusion proce-

dures and the sam e is likely tr ue in ASD.

Imp lants render p atients susceptible to deep

wound infections, as there is a race between

native cells and bacteria to t he implant surface.

Bacteria form a glycocalyx on implants, which

helps the m adhe re to the surface. The glycoc-

alyx “protects” the bacteria from antibiotics,

due t o poor pe net rance by antibiotics, and a lso

decreases the value of woun d culture, as bacte-

ria becom e adh eren t to the glycocalyx and are

not easily shed into the wound bed. There isevidence that titanium and cobalt chromium

imp lant s are less suscep tible to glycocalyx for-

m ation t han stainless steel. Our preference is

to u se Co-Cr rods for their m aterial propert ies

in ASD.

Patient demographics certainly help iden-

tify those at risk for developing SSI. Thus, as

w ith pseudar throsis, it is imp erative that t hese

pat ient s are id en t i ed and that approp riate

steps are taken to m inimize t he risk of SSI. A

comprehensive review of patients treated at

our institu tion, with an overall deep infection

rate of 2.0%, foun d th at a concom itan t diagno -

sis of diabetes had the strongest association

w ith a pe riope rat ive infection (odd s ratio [OR],

3.5).15 The impor tan ce of controlled blood glu-

cose levels was emphasized, as even patients

without a diagnosis of diabetes but with epi-

sodes of hyperglycem ia showe d a higher rate of

infection. These ndings were later supp orted

by Richards et al,16 w ho found an increased rateof infection in orthopedic traum a patients w ith

poorly controlled periopera tive bloo d glu cose



136 Chapte r 11

levels. Obesit y (body ma ss index 30– 35 kg/m 2)

was also associated with an increased risk of

infect ion (OR, 2/2).

Preincision antibiotic prophylaxis is man-

datory, and antibiotics must be given at the

appropriate time. We have shown a 3.4-foldincreased risk of infection in those p atient s wh o

did not receive intravenous cefazolin within

1 hou r of incision. In a st ud y of ped iatric spinal

deformity surgeries, Milstone et al17 found a

similar relationship between infections and

inappropriate antibiotic administration tim-

ing. The choice of proph ylactic antibiotics m ay

vary by institu tion, but it shou ld consist of an

antibiotic with broad-spectrum gram-positive

coverage of comm on skin ora. At our insti-

tu tion, prophylaxis is pr ovided w ith cefazolin

and vancomycin. In the case of penicillin or

cephalosporin allergy, we use a ztreona m . Both

ant ibiotics are given w ithin 1 hou r of the skin

incision, with vancomycin started earlier to

provid e an ap pro priat e rat e of adm in ist rat ion

and to m inim ize the risk of red m an syndrome.

During surger y, we re-dose ant ibiotics at half

the time of normal administration. For exam-

ple, cefazolin is g iven every 4 hou rs , as it is nor -

m ally given every 8 h ours.Int ra-site ant ibiotics, comm only in t he form

of lyophilized vancomycin powder, adminis-

tered at th e time of wound closure are becom-

ing increasingly comm on. Several retrospect ive

analyses have shown decreased rates of SSI

w ith the institution of this technique. Sweet et

al18 published the rst report on this m ethod,

noting a decreased prevalence from 2.6% to

0.2%with the use of intra-site vancomycin in

adult spine patient s. This bene t has been sup -

por ted by several ot her st udies in adult and

pediat r ic sp ine su rgeries. The re have be en no

comp lications de nitively linked to the u se of

intra-site vancomycin powder, though there

are concerns about pseudarthrosis, anaphylaxis,

and “super-in fect ion.” Vancomycin is acidic and

changes the environment within the wound

bed, t hou gh no e ect on union ra tes has been

observed. Vancomycin powder is e ective in

eliminating gram-positive contaminants, but

some concern over an increase in gram-nega-tive an d p olymicrobial infection s exists. A sin-

gle random ized controlled trial did not suppor t

the e ectiveness of lyophilized vancomycin

pow der.19 Nonetheless, our expe rience m irrors

those of others, and we continue to employ

this pra ctice.

We place both super cial and dee p drains atthe time of wound closure. There is limited bu t

not strong evidence suppor ting their use in spi-

nal d eform ity surger y. Blank et al20 perform ed a

rand om ized t rial of surgical drains in patient s

undergoing surgery for adolescent idiopathic

scoliosis. They found increased rate s of wou nd

drainage in those patients treated without a

drain. The study w as under powered and thus

could not detect a di eren ce in infection rates,

however. It stan ds to reason that an ade quately

pow ere d s tudy wou ld su ppor t a decreased r at e

of postoperative infection w ith decreased wound

drainage. Postoperative drains h ave be en asso-

ciated with increased rates of perioperative

blood t ra nsfu sions, and this m ust be balanced

with the potential bene t of wound drainage.

Diagnosis of spine infections can often be

m ade w ith a history and physical exam ination.

Fevers and chills as well as lethargy/malaise

are comm on—the form er are more frequent

with acute infections, and the latter are m orecommon with chronic, deep infections. In the

case of acute infections, some m ay show w ound

eryth em a, uctu ance, and woun d drainage.

These signs are not ubiquitous, however, and

th e clinician m ust h ave some level of suspicion.

In chronic infections, a small draining sinus

is com mon , or a new and enlarging uctuant

mass may be present. Upon presentation, one

should draw a complete blood count an d serum

C-re act ive p rotein (CRP). We h ave found CRP to

be m ore usefu l t han the eryt hro cyte sedim en-

tation rate (ESR) in diagnosing postoperative

infections. In t he im m ediate postoperat ive pe-

riod, the half-life of CRP is ~ 2.5 days, whereas

th e k inet ics of ESR are of little u tility. Im aging

of the spine should begin with plain radio-

graphs, which may show evidence of screw

loosening, with halos, in cases of chronic in-

fection (Fig. 11.4). Follow ing plain rad iograph s,

CT or magnetic resonance imaging are of lim-

ited utility, as a seroma forms in all cases, re-gardless of bacterial contamination. Aspiration




of a serom a and uid analysis m ay be per-

form ed w hen there is uncertainty regarding the

pre se nce of an in fect ion. Physician su sp icionand concern sh ould drive the d ecision to inter -

vene for a suspected infection.

Management of a super cial infection is

often su ccessful w ith ant ibiotics alone, as this is

usually a cellulitis related to th e surgical woun d.

Acute, deep wound infections are treated ag-

gressively in our p ract ice. The st an dard of care

is irrigation and debr idem ent . In th e woun d is

grossly contaminated, we may remove loose

graft. In the absence of gross contamination,

graft and implant s are retained . E ort should

be m ade to ach ieve good decontam inat ion of

the wound, so that implants can be retained.

Although t heir p resence as a foreign bod y might

concern surgeons, the instability caused by

implant removal will make eradication of the

infection m ore di cult. If th ere is any ques-

tion regarding the level of contamination of

the w ound, we place a wound vacuum -assisted

closure (WVAC) dressing and ret urn to t he o p-

erating room in 72 hours for repeat irrigationand debr idem en t, with p ossible WVAC change

Fig. 11.4 An 11-year-old girl with neuromuscular

scoliosis presente d with pain 11 mont hs after

posterior spina l fusion . Explorat ion revealed a deep

wound infect ion. Note the lucencies (arrows)

surrounding the midthoracic pedicle screws.

versus delayed primary closure. A chronic, de-

layed deep woun d infection is treated di erent ly,

however. These infections are m ore com m only

associated with less virulent bacteria, such as

Propion ibacterium acn es or Staphylococcus epi-

dermidis. As these bacteria are less virulent,and grow m ore slowly, intr aoperat ive cultures

should be taken and incubated for a longer

per iod than nor m al, 1 4 to 28 days . In cases of

chronic, delayed infection, we rem ove im plants

and con rm t he arthrodesis and the absence

of pseud ar th rosis. Patient s are followe d for ev-

idence of curve progression or pseudarthrosis

in follow-up and are re-instrumented only

wh en needed for deform ity progression.

When successfully treated, patients with

deep wound infection following ASD surgery

can expect good outcomes, equivalent to

those of patients who did not encounter this

complication.21

! Chapter Summary

Pseudarthrosis and infection are two comm on

reasons for revision surgery in adult spinaldeformity. Both of these pathologies have risk

factors that are m odi able by both th e patient

and the surgeon. Nicotine avoidance is m an-

datory in ASD, given the already high rate of

per ioperat ive com plicat ions assoc iat ed w it h

these surgeries. Evaluat ion of bone h ealth, with

bo ne m inera l densit y te st ing, and t reat m ent

of osteoporosis shou ld be routine. Approp riate

management of diabetes mellitus, including

tight control of perioperative blood glucose

levels, will help minimize risks of periopera-

tive infection. Ant ibiotic pro phylaxis shou ld be

given w ithin an h our of incision and be tailored

to prophylaxis against comm on ora in the

community.

As techn iques evolve, w ith concom itant im-

provem ents in im plants and b iologic therap ies,

the rates of these two complications will fall.

Ultimately, met iculous at tent ion to detail be-

fore, during, and after surgery w ill ma ximize

the likelihood of success and m inimize complicat ions in these pat ients.



138 Chapte r 11

Pearls

Smo king cessat ion is necessary in ad ult spinal de-

formity surgery.

Evaluation for osteoporosis, with bone mineral

density testing, and appropriate pharmacologi-cal intervention should be performed in ASD

patients.

Teriparatide is an anabolic agent available for

managem ent of osteoporosis and may have some

be ne t in achieving art hro de sis.

Facetect omies must b e performed , and decort i-

cation of dorsal elements is necessary.

Recombinant hum an bone m orphogene tic pro-

tein-2 decreases rates of reoperation for pseu-

darthrosis in ASD.

Informe d consent spe ci cally ta ilored to t he use

of rh-BMP2 in bot h on-labe l and o -label app lica-

tions (addressing risks of pain, seroma, ectopic bone, and po te nt ial for malignancy) should be

sought from the patient.

Iliac or S2-alar-iliac screws should be placed rou-

tinely, when fusing to S1.

High-density (" 1.8 screws/level) constructs are

recommend ed in ASD.

Patient facto rs associate d with surg ical-site infec-

tions in ASD include high body mass index and

poorly con trolled diab et es m ellitus.

Prophylactic, intravenous antibiotics must be ad-

ministered with in 1 hour of incision.

Prophylactic, intra-site, lyophilized vancomycin

po wder may de crease rat es of surgical-sit e

infection.

Aggressive management of acute, deep wound

infections o ers patients a chance for equivalent

outcom es once the complication has resolved.

Pitfalls

Nicot ine exposure increases rat es of pseud ar-

throsis. One must resist t he de sire to o perate o n

pa tien t s who are using nicot ine prod uc ts an d

insist that they be compliant with smoking ces-

sation prior to surgery.

Implants and osteobiologics do not compensate

for poo r planning and e xecution of te chnique in

ASD. Poor planning and performance increase

rates o f pseudar throsis and surgical-site infection.

The use of rh-BMP2 without informed consent

from the p atient is ill-advised.

Inadequate debridement of pseudarthrosis

tissue increases the likelihood of recurrent

pseu dart hrosis.

Prophylactic antibiotics must be administered

within 1 hour of incision. Cefazolin should be

given e very 4 hours during surgery.

Poorly controlled postoperative blood glucose

levels increase t he risk of infect ion.

References


1. Pichelm an n M A, Len ke LG, Brid we ll KH, Good CR,

O’Leary PT, Sides BA. Revision rates following pri-

mary adult spinal deformity surgery: six hundred

forty-three consecutive patients followed-up to

twenty-two years postoperative. Spine 2010;35:

219–226 PubMed

2. Kelly MP, Len ke LG, Bridwe ll KH, Agar wa l R, Godzik J,

Koester L. Fate of the adult revision spinal deformity

pat ien t: a sin gle inst it ut ion experience. Spin e

2013;38:E1196–E1200 PubMed

3 . Richard s M. Unan ticipated revision surger y in adult

spinal deformity: an experience with 815 cases at

one institution. Spine 2014;39(26, Suppl 1):S174–

182 PubMed

4. Mok JM, Cloyd JM, Bradford DS, et al. Reoperation

after prim ary fusion for adult spinal deform ity: rate,

reason, and tim ing. Spine 2009; 34:832– 839 PubMed

5. Kim HJ, Buchow ski JM, Zebala LP, Dickson DD,

Koester L, Bridwell KH. RhBMP-2 is superior to iliac

crest bone graft for long fusions to the sacrum in

adu lt spinal deform ity: 4- to 14-year follow-u p. Spine

2013;38:1209–1215 PubMed

6. Kim YJ, Brid well KH, Len ke LG, Rhim S, Cheh G. Pseu-

darthrosis in long adult spinal deform ity instrumen -

tation and fusion to the sacrum : prevalence and r isk

factor analysis of 144 cases. Spine 2006;31:2329–

2336 PubMed

7. Stoker GE, Buchowski JM, Bridwell KH, Lenke LG,

Riew KD, Zebala LP. Preoper ative vita m in D st atu s of

adults u nder going surgical spinal fusion. Spine 20 13;

38:507–515 PubMed

8. Hirsch BP, Unnanuntana A, Cunningham ME, Lane

JM. The e ect of therap ies for osteoporosis on spine

fusion: a systematic review. Spine J 2013;13:190–

19 9 PubMed

9. Smit h JS, Sha rey CI, Am es CP, et al; Inte rn at ional

Spine Study Group. Assessme nt of symp tom atic rod

fracture after posterior instrumented fusion for

adult spinal deformity. Neurosurgery 2012;71:862–

86 7 PubMed

10. Kebaish KM. Sacropelvic xation: techniques and

complications. Spine 2010; 35:2245 –2251 PubMed

11. Car re on LY, Djurasovic M, Glassm an SD, Sailer P. Di-

agnostic accuracy and reliability of ne- cut CT scans




with reconstructions to determine the status of an

instrum ented posterolateral fusion w ith surgical ex-

plor at ion as re fer en ce st an da rd . Spin e 2 00 7; 32 :8 92 –

89 5 PubMed

12. Pateder DB, Park YS, Kebaish KM, et al. Spinal fusion

after revision surgery for pseudarthrosis in adult

scoliosis. Spine 2006;31:E314–E319 PubMed

13 . Smit h JS, Sha rey CI, Sansu r CA, et al; Scoliosis Re-

search Society Morbidity and Mortality Committee.

Rates of infection after spine surgery based on

108,419 procedu res: a re port from t he Scoliosis Re-

search Society Morbidity and Mortality Committee.

Spine 2011;36:556–563 PubMed

14. Woods BI, Rosario BL, Chen A, et al. The association

be twee n pe riop er at ive a lloge neic transfu sion volu m e

and postoperative infection in patients following

lum bar sp ine su rgery. J Bone Joint Surg Am 2013;95:

2105–2110 PubMed

15. Olsen MA, Nepple JJ, Riew KD, et al. Risk factors for

surgical site infection following orthopaedic spinal

operations. J Bone Joint Surg Am 2008;90:62–69

PubMed

16. Richard s JE, Kau m ann RM, Zucker m an SL, Obre m s-

key W T, May AK. Relation ship of hype rglycemia a nd

surgical-site infection in or thop aedic sur gery. J Bone

Joint Surg Am 2012;94: 1181–11 86 PubMed

17. Milsto ne AM, Maraga kis LL, Tow ns en d T, et a l. Tim ing

of preop erative antibiotic prop hylaxis: a m odi able

risk factor for deep surgical site infections after p edi-

atric spinal fusion. Pediatr Infect Dis J 2008;27: 704–

70 8 PubMed

18. Swe et FA, Roh M, Sliva C. Intra wou nd app lication of

vancomycin for prophylaxis in instrumented thora-

colum bar fusions: e cacy, dru g levels, and patien t

outcomes. Spine 2011;36:2084–2088 PubMed

19. Tub aki VR, Rajasekar an S, She tt y AP. E ect s of usin g

intravenous antibiotic only versus local intrawound

vancomycin antibiotic powder application in addi-

tion to intravenous antibiotics on postoperative in-

fection in spine sur gery in 907 patient s. Spine 2013;

38:2149–2155 PubMed

20. Blank J, Flynn JM, Bron son W. The use of post ope ra-

tive subcut aneou s closed suction dra inage after pos-

terior spinal fusion in adolescents with idiopathic

scoliosis. J Spina l Disord Tech 2 003 ;16(6 ):508 –51 2

21. Mok JM, Guillaum e TJ, Talu U, et a l. Clinical ou tcom e

of deep wound infection after instrumented poste-

rior spinal fusion: a matched cohort analysis. Spine

2009;34:578–583 PubMed



Note: Page references followed by f or t indicate p ages or tab les, respect ively.

A

Acetabu lum , iliac screw xation- related injur y to,

53, 54

Adjacent segm ent disease/failure, 24, 56–5 7,

65–66

proxim al, 25

Adolescent idiopath ic scoliosis

corrective procedu res for, 123– 127, 125 f, 126 f

fusion levels in, 23

pro gressio n to ad ult sco liosis , 78 , 79 f

unt reated, 1

Adu lt degen erat ive spinal deform ities, 1. See also

Adu lt scoliosis, degene rative (de novo)

clinical evaluation of, 2–4

epide m iology of, 1–2

pre op era t ive eva luat ion of, 2–4

pro gressio n ra te of, 2

Adu lt scoliosis, 1–11

classi cation of, 79, 81

de nition of, 78

degen erat ive (de n ovo), 80 f

apical disk height in , 80 f, 81coronal imbalance in, 81–82

de nition of, 78

di eren tiated from adu lt idiopath ic scoliosis,

79, 81–82, 93

lumbar kyphosis associated w ith, 80 f, 83

pat hology of, 83

spinal curve pattern s in, 81

stenosis-related, 82

surgical option s for, 78–79

tru nk shifting in, 81–82

fusion levels for, 18

idiopathic

as back pain cause, 82de nition of, 78

di eren tiated from adu lt degen erat ive scoliosis,

79, 81–82, 93

proxim al ju nct ional kyp hos is in , 11 0

nonop erative m anagemen t of, 78

surgical treat m ent of, indications for, 78

treatm ent d ecision making about, 12–27

Adu lt spina l deformit ies. See also Adu lt scoliosis

de nition of, 1

nonop erative m anagemen t of, 4, 17, 99

pat hogenes is o f, 1

pre se nt ing age of, 2 , 5

prevalence of, 99

progr ession of, 2 , 4, 17

typ es of, 1

Adult spinal deformity surgery

indications for, 4–5, 78, 99

levels of, 8–9

outcomes measures for. See Health-related quality

of life (HRQOL) m easu res

Album in levels, preope rative assessm ent of, 5

Am er ican Spin al Injur y Association (ASIA) scor ing

system , for spin al cord injur y, 68, 69 f, 70Ank ylosing sp ond ylitis, osteotom y plan ning for,

31–32, 31 f

Anterior approach. See also Comb ined ante rior/

pos te r ior a ppro ach

in ad ult scoliosis, 78

in rigid spinal deform ity surger y, 24

Anter ior column su pport

e ect on union rates, 130–131

in osteopo rotic spine, 62

in pseu doarth rosis revision surgery, 134

Ante rior xation, in osteop orotic spine, 63

Anterior lumbar interbody fusion (ALIF), 130–131,

13 4Arte ry of Adam kiewicz, in ost eotom y, 29–3 0, 29 f, 31

Index



142 Index

B

Back pain, adu lt spinal d eform ity-related , 2, 9, 78

axial, 12, 85

clinical evaluat ion of, 2

diagnosis of, 12–16 , 13 f, 14 f, 15 f, 16 f

as indication for surger y, 17

mechanical, 82neurogenic, 82

radiologic assessm ent of, 13–1 6, 13 f, 14 f, 15 f

Biome chanics, of spinal deform ity correct ion,

120–129

implant mater ials, 133

intraoperative mechanical forces, 126–127,

12 6 f

me chanical proper ties of imp lants, 120–121

after man ual bending, 122–123, 122t

spinal viscoelasticity, 123, 1 24 f, 128

stress-strain curves, 120–121, 121 f

Bisphosphonates, 132

Blood m anagemen t, perioperative, 5–6

Blood t ran sfusions, comp lications and risks of,

6, 135

Bone grafts, 133, 134

Bone m ineral d ensit y (BMD)

correlation w ith

implant pu llout stren gth, 58, 132

insertional torque, 132

falsely elevated, 132, 132 f

low, as spinal imp lant failure cause, 56

in osteop orotic spine, 56, 58, 132, 138

pre op era t ive m ea su re m ent of, 7 , 57

Bracingcontraindication to, 4

pos to pera t ive, of ost eop or ot ic spin e, 65, 66

But tock pain, 13, 16

C

Cages, for osteop orotic spin e, 63

Cen tr al sa cral ver tical lin e (CSVL), 14–1 5, 15 f, 84

Claudication

neu rogenic, 2, 78, 82

di eren tiated from vascular claud ication, 16

level of surgical treatm ent for, 8

ner ve root comp ression-related, 16

pharm acot her ap y for , 4

relationship to coronal imbalance, 85

vascular, di eren tiated from neur ogenic

claud ication, 16

Cobb an gle, in ad ult spin al deform ities, 2, 82

correlation with apical disk degeneration, 83

discrepancy w ith rotatory su bluxation, 82

less than 30°, 18, 19 f, 25

pre op era t ive m ea su re m ent of, 3

in proximal junct ional kypho sis, 106, 107 t , 11 5

in spinal curve pr ogression, 17

as surgery outcome m easure, 95–96

Codm an , Ern est A., 102Combined anterior/posterior app roach


as proximal junctional kyph osis cause, 107 t ,

108–109, 114

as respiratory system injury cause, 6

Comorbidities, in adult spinal deformity patients,

5, 7, 56

Compensatory spinal curves, 81–82

Com plications, of adu lt spinal deform ity surger y,3, 24–25, 100–101. See also Neurologic

complications, of adu lt spinal deform ity

surgery; Pseudoarthrosis; Surgical-site

infections

causes of, 24–25

pre op er at ive assessm en t for, 5

pre vent ion of, 7–8

rate of, 17, 24

Com put ed to m ography (CT)

pos top era t ive, 74 , 76

pre op er at ive, 3, 57

of pseudoart hrosis, 133

Compu ted t omography m yelography

for back pain assessm ent , 13

intraoperative, 74

“Con e o f econo my,” 13

Cont ractu res, of hip or kn ee, 3

Corona l balance

in ad ult scoliosis, 84, 84 f

assessm ent o f, 14–1 5, 15 f, 84

Corona l comp ensat ion, de nition of, 84

Coronal decompensation. See also Coron al

imbalance

de nition of, 84, 85

pos top era t ive, pr eve nt ion of, 87– 88Corona l im balance

in ad olescent idiop athic scoliosis, 79 f, 81

in ad ult scoliosis

in adu lt degen erat ive scoliosis, 81–8 2

correlation w ith qu ality of life, 84–85

de nition of, 85–86

with associated sagittal imbalance, 87

e ects of, 13

greater than 4 cm , 8

pos top era t ive, 85

as instrum entat ion failure cause, 88, 90 f –9 2 f, 93

revision sur gery for, 88, 90 f –92 f, 93

pre op er at ive, 85

typ es of, 24

typ e A, 84 f, 87, 93

t ype B, 84 f, 87, 93

t ype C, 84 f, 87, 88, 89 f, 93

Cort icosteroid epidu ral injections, diagnostic, 4

Cost, of adu lt spinal deform ity sur gery, 95, 98–99

Cross-links

in osteop orotic spine, 61, 66

in ped icle screw xation, 127

C7 plum b line, 14–15, 15 f, 11 0

DDecompression, 8

e cacy of, 17



Index 143

with fusion

anterior an d p osterior fusion, 8, 78

complication rat e of, 100

limited p oster ior fusion, 8, 78

lumbar cur ve instru me nted fusion, 8, 78

with pedicle subt raction osteotomy, 34–35

Decompression-only surgeryfor ad ult d egener ative scoliosis, 78

complication rat e of, 100

indications for, 18, 25

versus shor t or long fusion, 26 t

Degene rative cascade, of adu lt spinal deform ities,

12

Degree of slip, radiologic measu rem ent of, 13–1 4,

13 f, 14 f

Denosum ab, 132

Diabetes m ellitus, as su rgical-site infection risk

factor, 135–13 6, 138

Disk degener ation

correlation with spinal imbalance, 83

ima ging of, 16

lum bar, 17

Disk height, measu rem ent o f, 14

Dual-energy X-ray absorptiometry (DEXA),

pre op era t ive, 7, 57–5 8, 66 , 13 2

E

Elderly patient s

adu lt spinal deform ity onset in, 2, 5

spinal surger y-related comp lications in, 24

Electromyography

spontane ous, 72triggered, 72

Euro QOL Five Dim en sion s (EQ-5D) qu est ionn aire,

96, 98

F

Facet nerve blocks, 4

Fenest ration, poste rior, 18

Fibular str ut grafting, 48

Flat-back syndrome, with lumbar lordosis, 8

Fractures

of imp lants, 130, 133

osteop orotic, 7, 56, 65, 81, 82

as proximal junctional kyphosis cause,

110–111 114

Fusion. See also Interbody fusion/supp ort

in adu lt degene rative scoliosis, 78–7 9

ant erior, 24

with decompression, 8, 78, 100

indications for, 18–1 9

lim ited, 17

local, w ith pedicle screw xation, 18

long-segment, 18–19, 20 f –2 2 f, 23

complications of, 25, 130

versus decompression-only surgery, 26 t

indications for, 26, 26t w ith lower inst rum ent ed verteb ra (LIV), 23–24 ,

25–26

as pseu doar th rosis risk factor, 130

with uppe r instru me nted vertebra (UIV),

19, 23

in osteop orotic spine, 65, 132

to th e pelvis, as proximal junct ional kyphos is risk

factor, 109

to th e sacrum , 23–24, 25, 26, 79as proxim al junct ional kyph osis risk factor,

10 7t , 109

short-segme nt, 18, 19 f –20 f

indications for, 26, 26 t

G

Gait assessm ent , 3, 13

Galveston tech nique, 49

Groin p ain, 13

H

Harrington , Paul, 120, 127

Harrington instrum ent ation system

developm en t of, 120, 127

with sacral hooks, 48

Harrington t hreaded sacral rods, 48–49

Health -relat ed q ua lity of life (HRQOL) m easu res,

95–105

comm only used measures, 98

correlation w ith

coronal imb alance, 83, 85

sagittal imbalan ce, 83

cost an d value-related m easures, 95, 98–99

de nition of, 95

need for improvement in, 102 pat ient-re por ted m ea su re s (PROMs), 96, 98 , 10 3

physio logical m ea su re s, 9 5– 96

pro cess m easure s, 95

quality m easures, 96

radiographic outcomes, 99–100

ut ility scores, 96, 98, 103

Health Utilities Inde x (HUI), 96, 98

Hemoglobin levels, preo perat ive, 5–6

Hemostasis, intraoperative, 6

High-risk patients

decom pression- only surgery for, 18

pro toco l for, 5

Hydroxyapatite, as ped icle screw coating, 59–60,

13 2

Hyperglycemia, as surgical-site infection risk factor,

135–136, 138

Hypovitam inosis D, 6–7, 9, 131

I

Iliac screw xation, 48–5 1, 50 f, 53–54

asymp tom atic haloing of, 53 f, 54

complications of, 53, 54

implantation problem s with, 25, 26

in postoperative coronal imbalance revision

surger y, 88S2-alar- iliac, 51, 52 f, 53, 54

in fusions to S1, 133, 138



144 Index

Ilium, fusion to, as proxim al junction al kyphosis

risk factor, 106, 109

Infections. See Surgical-site infections

“Instability catch,” 12–13

Instru m entat ion, spinal. See also Osteoporot ic spine,

instrum entat ion strategies for

implication for magnet ic resona nce imaging, 3levels of, 18

rigidity of, as proxim al junction al kyph osis risk

factor, 108

th oracic, w ith fusion exten sion, 8

Interbody fusion/suppor t

anterior lumbar (ALIF), 130–131, 134

complication rate o f, 100– 101

in osteopo rotic spine, 62–6 3, 64 f

in sacral-pelvic xation, 51

tran sforam inal lum bar (TLIF), 130–1 31

Intraop erative m onitoring. See Neuromonitoring,

intraoperative

Ischem ia, spinal, 75

J

Jackson int rasacral rod te chnique , 48

K

Kostuik transiliac bar, 49

Kyphosis

lumbar, adult degenerative scoliosis-related, 80 f, 83

osteotom y for, 30, 34–36, 35 f

pos tlam in ectom y, 13 5

proxim al ju nct ional , 79 , 10 6– 119

clinical outcom es after, 111, 112 t –11 3 t etiology of, 106–1 11, 115

pre vent ion of, 114– 11 5

revision surger y for, 111, 114–1 15, 116 f –11 7 f,

117, 118

risk factors for, 106– 111, 107 t , 11 5

tim ing of, 111

Scheu erm ann ’s, 108, 114–115

thoracic

osteotom y for, 31–32 , 31 f

pos ter ior spin al in st rum entat ion-r ela ted,

123–124

pos top er at ive, 11 0, 1 23– 124

pre op era t ive, 10 7 t , 110

techn iques for creation of, 123– 125

L

Laminectomy

in decompression-only surgery, 18

as kyph osis cause, 135

Leg-length discrepan cy, 2–3, 79, 81

Leg p ain, 16, 18 , 82

Lidocaine, as spinal cord ischem ia treatm ent , 75

Listhesis. See also Spond ylolisthesis

anteroposterior, 3

later al, 3, 16Lordosis

xed, osteotom y for, 40, 42 f, 43

lumbar

adu lt degene rative scoliosis-related , 80 f, 82

adu lt idiopath ic scoliosis-related, 82

anterior fusion app roach to, 8

with at-back syndrom e, 8

as proxim al junction al kyph osis risk factor,

107 t , 109, 110, 115radiological measurement of, 15

as spinal surgery outcome measu re, 95–96

Lower in stru m ent ed ver tebrae (LIV)

in long-segmen t fusion, 23–24, 25–26

in proximal junctional kyphosis, 109

Lum bar spin e. See also Lordosis, lum bar

degen erative adu lt scoliosis of, 81

pedicle scre w xat ion in , 60

Lum bosacral xation, in osteop orotic spine, 61

Lumbosacral junction

biom echan ics of, 45

sacral-pelvic xation of, 45–5 5

Lum bosacral pivot point , 45

Luque instr um entat ion, 91 f

Luq ue L- xat ion, 49

M

Magnetic reson ance imaging (MRI)

of disk degene ration, 16

lum bar, for back pain assessm ent , 13

pos top er at ive, 74 , 76

pre op er at ive, 3, 70

of spinal steno sis, 16

Mean ar ter ial pressu re (MAP), intra oper ative

m onitor ing of, 71, 73, 74, 75Metab olic bon e disease, as adu lt scoliosis cause,

1, 79

Metallic implan ts, for spinal deform ity correct ion,

120–129

for ad olescent idiopat hic scoliosis, 123– 127, 125 f,

126 f

nite elem ent ana lysis of, 126–1 27

fractures of, 130, 133

histor y of, 120, 121 t

implant d ensity, 133

relationship w ith correction rate, 127

infection of, 135

intraoperative mechanical forces on, 126–127,

126 f

m aterials for, 133

m echanical properties of, 120–121

after man ual bending, 122–123, 122 t

spinal viscoelasticity an d, 123, 124 f, 12 8

stress-st rain cur ves of, 120– 121, 121t

Methylprednisolone, 75

Microdensitometry, preoperative, 57

Motor evoked pote nt ial (MEP) intra oper ative

m onitor ing, 29–3 0, 43, 71, 72, 73, 74

N Nar cot ics pain m edica t ions, 4 , 10

Neck Disa bi lit y In dex (NDI), 96



Index 145

Nerve ro ot com pre ssion , sym ptom s of, 16

Nerve ro ot /t ra nsfora m in al inject ions, d iagnos t ic,

3– 4

Neuro log ic co m plica t ions, o f ad ult sp in al deform it y

surgery, 68–77

delayed postoperative, 75–76

intraoperative managem ent of, 71intraoperative neuromonitoring for, 29–30, 43,

71–73 , 74, 75

mechan isms of, 70

pos top er at ive m an age m ent of, 75

pre op era t ive risk as sessm en t for, 7 0

pre valence of, 68–7 0, 68 f

steroid protocol for, 75

Neuro log ic d e cit s, p re op er at ive eva luat ion of, 2

Neuro m on itor ing, in tra op er at ive, 74 , 75

m otor evoked pote nt ial (MEP) m onitor ing, 29–3 0,

43, 71, 72, 73, 74

som atosen sory evoked pote nt ial (SSEP) m onitor-

ing, 71–72, 73, 74

Neurovascular e xam in at ion, pr eop era t ive, 3

Nicot ine. See also Smoking cessation

e ect on patient outcomes, 131

Non stero idal ant i- in am m at or y d rugs (NSAIDs), 4

Non union . See also Pseudoarth rosis

nicotine-related, 131

rates of, 133, 134

reope ration r isk of, 134

Nut rit ional st at us, p re op era t ive asse ssm ent of, 5 , 10

Nut rit ional su ppor t , in sp inal su rger y p at ien ts, 5

OObesity

as proxima l jun ctiona l kyphosis risk factor, 110

as sur gical-site infection r isk factor, 136, 138

Open anter ior surgery, pulm onary function e ects

of, 6

Osteop enia, 56, 111, 114, 132

Osteop hytosis, preope rative evaluation of, 3

Osteoporosis, in adult spinal deformity patients,

138. See also Osteoporotic spine, instru m en-

tat ion strategies for

bo ne m inera l densit y in , 56 , 58 , 13 2, 1 38

diagnosis of, 132

man agement of, 132, 138

pre op era t ive m edica l t re at m ent for, 58

as proxim al junction al kypho sis risk factor, 106,

110–111

in spinal surgery pat ient s, 7, 10

Osteoporotic spine, instrumentation strategies for,

56–67

adjacent segment failure in, 56–57, 65–66

anchor points en hancement, 61

anterior xation, 63

bo ne- im plan t in te rface p ro tect ion , 63 , 65

xation failure in, 56–5 7, 65–6 6

fusion, 132for instru me ntat ion failure p revention

intraoperative measures, 58–65

pos top era t ive m ea su re s, 6 5

pre op era t ive m ea su re s, 5 7– 58 , 66

interbody suppor t, 62–63, 64 f

nonu nion and, 57

ped icle scre w xat ion , 58 –61

insertion technique and insertion torque in,

60–61tract augmen tation in, 59–60

pro phylact ic ve r teb ro plast y, 61 –62 , 66

pse udoa rthro sis an d, 56 –5 7, 6 5– 66

sem i-rigid xation, 63

Osteotomy, 8. See also Vertebral column resection

(VCR)

closure of, 30, 35–3 6, 36 f

for xed lordosis, 40, 42 f, 43

with long-segment fusion, 24

ped icle su btra ct ion

closure of, 35–36 , 36 f

comparison w ith Smith-Petersen osteotomy, 36 f

complications of, 34

for coronal im balance corre ction, 88, 91 f

outcomes, 100

with previous anterior implants, 36–37, 38 f

revision surger y of, 131 f

for rigid spinal deform ities, 30, 31–3 2, 31 f, 33 f,

34–36, 35 f, 36 f

techn ique of, 34–3 6, 35 f

for typ e A coronal imba lance correction , 87

pos te r ior

outcom es of, 100

in pseu doarth rosis revision surgery, 134

Smith-Petersen, 36 f, 10 0as spinal cord infarction cause, 29–30

for proxim al junction al kyphos is, 115, 116 f, 11 7

for rigid spinal deform ities, 28–4 4

anterior app roach in, 28

closure of, 30

xation in, 30

fusion across, 32

level of osteotom y, 30–3 1

m ulti-level vertebre ctomy, 32, 33 f, 40, 41 f

num ber of osteotomies required, 30

pedicle su btra ct ion os teo to my, 30 , 31 –32 , 31 f,

33 f, 34–36 , 35 f, 36 f

p lann ing of, 28–3 2

pos terior a ppro ach in , 28

pos terior colum n os teot om y, 33 –3 4, 3 3 f, 34 f

as single procedur e, 32

as staged procedure, 32

transdiskal pedicle subtraction osteotomy, 32,

33 f, 36–37 , 37 f

typ es of, 32–4 0

three-column , 8–9, 10

for coronal imb alance correct ion, 87–88, 89 f

deform ity exibility determination in, 28–29

indications for, 28–2 9

level of, 31versus m ultiple posterior column releases, 28,

29



146 Index

Osteotomy (continued )

for proxim al jun ctional kyph osis, 117 f

reope ration rate for, 101

spinal cord blood ow in, 29–30

spinal cord protect ion in, 43

surgical exposure in, 29

thoracolumbar, 29, 29 f Oswestry Disability Index (ODI), 8, 96, 98, 101–102,

11 1

Outcome m easures, in spinal deform ity surgery.

See Healt h- relate d q ua lity o f life (HRQOL)

measures

P

“Painfu l catch,” 12– 13

Pain man agement , in adult spinal deformities, 4, 10

Parathyroid horm one, 132

Patient-reported outcome measures (PROMs), 96,

98, 103

Pedicle rods, biomecha nical proper ties of, 120

Pedicle screw xation

altern atives to, 30, 43

intraoperative assessme nt of, 74

w ith local fusion, 18

in osteoporotic spine, 58–61, 132–133

as proxima l jun ctional kyph osis risk factor, 108

sacral, 46–48, 47 f

m edial insert ion tr ajectory for, 51, 53, 53 f

Pedicle screw s

biom echan ica l p ro per t ies o f, 120

hydroxyapatite-coated, 59–60, 132

m echan ical forces on, 126–127 , 126 f Pediculolaminar instrum entat ion, for osteoporot ic

spine, 61

Pelvic incidence (PI), as spinal d eform ity su rgery

outcome m easure, 95–96

Pelvic obliquit y, 2–3

Pelvis

fusion to, as proxima l jun ctiona l kyphosis risk

factor, 109

ped icle scre w xat ion in , 60

Periapical redu ction screw s, 30

Physical examin ation, preop erat ive, 70

Physiological outcome m easur es, 95–9 6

Polyme thylm eth acrylate (PMMA) bone cem ent , 59

Posterior app roach, 9. See also Combined anterior/

pos te r ior ap pro ach


in severe adu lt spinal deformit ies, 6

Posterior column release, multiple periapical

osteotom ies for, 30–3 1

Posterior spinal ligaments, surgery-related dis-

rupt ion of, 106–108, 107 t , 11 5

Postm enopau sal wom en, bone mineral density

screen ing in, 7

Postural imb alance, as back pain cau se, 13

Prealbum in levels, preop erat ive assessm ent of, 5Preoperative evaluation, of adult degenerative

spinal deformit ies, 1–4

Preoperative planning, of adult spinal deformity

surger y, 8–9

Process m easure s, 95

Provocative testing, preo per ative, 3–4

Proxim al jun ctional angle, in proxima l jun ctiona l

kyph osis, 107 t , 110

Pseudoar throsis, 130–134, 131 f, 13 3 f, 13 4 f diagnosis of, 133

Harrington threade d sacral rod-related,

48–49

L5-S1, fus ion- relate d, 25

long-segment fusion-related, 24

m anagemen t of, 133–134

rate of, 130

as revision surgery cause, 130, 133– 134

risk factors for, 130–13 3

sacral hook-related , 48

Psychological disorders, in adu lt spinal deform ity

pat ien ts, 7

Pulmonary disorders, postoperative, 8

Pulmonary function testing, preoperative, 6, 9

Q

Quality- adju ste d life yea rs (QALYs), 98, 99 , 103

R

Radicular pain, 4, 16, 82, 85

Radiculopathy, 2, 78

Radiographic measurements, of spinal deformity,

95–96

Radiograp hic outcom es, of adu lt spinal deform ity

surgery, 99–100Radiological imaging, preoperative, 3, 9, 70

Radiological instability, 13–14, 13 f

Recombinant human bone morphogenetic

pro te in -2 (rh -BMP2), 13 3– 134, 1 38

Recombinant human erythropoietin (rEPO), 6

Reop eration /revision surgery, 101

for oste otom y, 101

for ped icle xation osteoto my, 131 f

for post operat ive coronal imb alance, 88, 90 f –92 f,

93

for p seudoart hrosis, 130, 133–134

for sur gical complications corre ction, 24

surgical-site infection -related, 135

Rigid spinal d eform ities

combined an terior/posterior approach to, 8–9

osteotom y for. See Osteotom y, for r igid spinal

deformities

proxim al ju nct ional kyp hos is- re lat ed , 11 7 f,

11 8

Rod rotation m aneuvers, intraoperative m echanical

forces in, 126–12 7, 126 f

S

Sacral-pelvic xation, 45–5 5

adjunctive ante rior interbody support in, 51anatom ic and biomechanical considerations in,

45–46



Index 147

indications for, 46

instrum entat ion selection and techniques for,

46–51

iliac xation, 48–51 , 50 f, 52 f

sacral xation, 46–4 8, 47 f

oper ative technique s in, 51, 53, 53 f

pat ien t pos it ioning for , 51Sacral slope, 95–9 6

Sacrum

bo ne m inera l densit y o f, 58

fusion to, 23–24 , 25, 26, 79

as proxim al junction al kyph osis risk factor,

10 7 t , 10 9

pedicle scre w xat ion in , 60

Sagittal alignm ent , global, 107 t , 109–110, 118

Sagittal b alance

pre op era t ive assessm ent of, 8

radiological assessme nt of, 14–15 , 15 f

as surgical outcom e pred ictor, 85

Sagittal imb alance

adu lt degen erative scoliosis-related, 83

with associated coronal imbalance, 87

correlation w ith d isk degenerat ion, 80 f, 83

in osteopor otic spine, 65

pre op era t ive eva luat ion of, 10

proxim al ju nct ional kyp hos is- re lat ed , 11 1

surgical correction of, 87

typ es of, 24

Sagittal plane d eform ities, 13

Sagittal p lum b line, C7, 14–15, 15 f, 11 0

Sagitt al vert ical axis (SVA), in p roxim al jun ction al

kyphosis, 107t , 110, 114, 115Sciatic ner ve, iliac screw xation-r elated injur y to,

53, 54

Scoliosis, 1. See also Adolescent idiopath ic scoliosis;

Adu lt scoliosis

neu rom uscular, as surgical-site infection risk

factor, 135, 137 f

pr im ar y d egenera t ive, 1

secondar y, 1

surgery-related, 1

Scoliosis Research Societ y, 84

Morbidity and Mortality d atabase, 135

SRS-22 quest ionna ire, 8, 11, 96, 98, 101–1 02

Sem i-rigid xation, in osteop orotic spine, 63

Severe adu lt spinal deform ities, surgical decision

making regarding, 17–26

Short Form 6 Domains (SF-6D), 96, 98

Shor t Form -36 (SF-36), 96, 98

Slip an gle, radiological m easure m ent of, 13–1 4,

13 f, 14 f

Sm oking cessation, preo perat ive, 6, 8, 131, 138

Som atosen sory evoked p oten tial (SSEP)

intraoperative monitoring, 71–72, 73, 74

Spee d techn ique, of bular str ut grafting, 48

Spinal cord injur y

acute, medical treatm ent for, 75delayed intraoperative, 75–76

Spinal cord per fusion, int raop erat ive, 73

Spinal cu rves

compensatory, 81–82

progr ession of, 16– 17

Spinal instab ility, diagnosis of, 12–13

Spinal surger y

decision m aking for, 17–2 6

high-risk protocol for, 5indications for, 17

as scoliosis cause, 1

Spin e, viscoelasticity of, 123, 124 f, 12 8

Spinopelvic imbalance, radiographic determination

of, 3

Spinopelvic parameters, as spinal deformity surgery

outcome m easures, 95–96

Spondylolisthesis

degenerative, 12

L5-S1

fusion sur gery for, 25

slip angle m easurem ent in, 14

lateral, 17

Stagnara Wake-Up Test, 71, 73–74

Sten osis, spinal

im aging of, 16

lumbar

as adu lt degen erative scoliosis cause, 82

at L5-S1, 23

as pseu doar th rosis risk factor, 130

Sublaminar instr um entat ion, 48, 61

Sublu xation, rotator y, 15–1 6, 16 f, 82

de nition of, 82

discrepan cy with Cobb an gle, 82

w ithin fusion block, 26 pre op era t ive eva luat ion of, 3

as spinal cu rve pr ogression risk factor, 17

Sup erior gluteal arte ry, iliac screw xation- related

injury to, 53

Surgical-site infections, 134–13 7

antibiotic prophylaxis/treatm ent for, 134–135,

136, 137, 138

diagnosis of, 136– 137, 137 f

man agement of, 137

as revision surger y cause, 130

risk factors for, 135–13 6, 138

T

Teripar atide, 132 , 134, 138

Thoracic instru me ntat ion, with fusion extension, 8

Thora cic spine, ped icle screw xation in, 60

Thora columb ar spine, adult degen erat ive scoliosis

of, 81

Thoracolum bar/th oracolum bosacral orthoses

(TLO/TLSO), 4

Transforaminal lumbar interbody fusion (TLIF),

130–131

Tran siliac xat ion, 48– 49

Tran ssacra l iliac xation , 51, 52 f

Transver tebral bular str ut grafting, 48Tricor tical xat ion, of sacral pe dicle screw s, 46,

47 f



148 Index

U

Universal clamp t echniqu e, 124– 125

Upper instrum ented vertebrae (UIV)

in long-segmen t fusion, 19, 23

in proximal jun ctional kyphosis, 107t , 108 , 109,

110–111, 114, 118

Urinar y tract infections, per ioperat ive, 7–8, 135

V

Value, of health care, 95, 98–99

of adult spinal deform ity surger y, 102, 103

Vertebrae

apex/apical

in adu lt degen erat ive scoliosis, 81

in ped icle screw xation, 125, 125 f

osteoporotic, pathomorphology of, 58

proxim al neutra l, 15

Ver teb ral colum n re sect ion (VCR)

out comes of, 100

of proxim al jun ctional kyph osis, 115

of rigid sp inal deform ities, 32, 33 f, 37–40 , 39 f

of severe ad ult spina l deform ities, 6

for t ype B coronal imba lance correction , 87

Ver teb ral copla na r alignm en t (VCA), 124Vert ebrectom y, mult i-level, 32, 33 f, 40, 41 f

Vert ebroplast y, prop hylactic, in osteop orotic spine,

61–62, 66

Visual Analogue Scale (VAS), 96, 10 1

Vitam in D de ciency, 6–7, 9, 131

W

Wound infections, postoperative. See Surgical-site

infections

ao spine masters series volume 4 adult spinal (bookzz.org)

Documents