536
Thoracicand Thoracolumbar Spine: Posterior Approach
.
PostoperativeCare
Posterior column and motor function is monitored in the
recovery room. Attention is also directed toward ensuring
good sphincteric function in the days following surgery. A
lumbar x-ray with a lateral view is obtained in the days
followingsurgery.
.
ChoiceofOperation
Once the necessity of a shunt
insertion
to treat
sy-
ringomyelia is determined. one has to decide which opera-
tion is most indicated for a particular patient. This will de-
pend on key factors listed in Table 72- t. These are listed in
orderof importance. The type of syrinx one is dealing with is
most important. Aholocord syrinx. even in the presence of a
patent subarachnoid space. is not likely to respond to a pas-
sivesyringosubarachnoid shunt. Bydefinition. for this syrinx
typethe thoracic area is available for shunting and the cervi-
calsegments are not put at risk by the treatment. Finally.the
pleural site is close. A syringopleural shunt is therefore
Table 72-1
KeyFactors Influencing the Choice of the Type of
Shunting Needed in Syringomyelia
1. Typeof syrinx(holocord vs segmental)
2. Site of proposed myelotomy along spinal axis
3. Patency of subarachnoid space
4. Patency of pleura or peritoneum
5. Previous shunting procedure
Figure 72-1
Abdominal site. Note the horizon-
tal skin incision to respect skin folds followedby
vertical fascial incision to respect muscle fiber
orientation.
indicated. Aperitoneal recipient site is less optimal because
the course of the catheter is longer and more vertical.favor-
ing dislodgment. If one chooses a lower myelotomysite.
there may be more risk for sphincters as one approachesthe
conus region. Finally.the peritoneum as a recipient usually
forces repositioning during surgery. making the surgery
longer and the risk of contamination higher. The riskofin-
fection in our hands is higher with this type of shunt.
For a segmental syrinx. the syringosubarachnoid shunt
efficacyis higher than with a holocordsyrinx.andtherisk
of dislodgment of the catheter less than with other shunt
types. The operative morbidity is also theoretically less.
However. one must be satisfied with the patency of the
subarachnoid space and with its ability to absorb the sy-
ringomyelic fluid. Ifsuch is not the case. then one shouldre-
ally resort back to either a syringopleural or a syringoperi-
toneal shunt depending on the position of the syrinxalong
the spinal axis. A corollary to this would be a previous
attempt at syringosubarachnoid shunting that has failed.
Conversely,the only exception to this would be a veryhigh,
circumferentially large but longitudinally short cervical
syrinx where not only spine movements but head move-
ments make anchoring the tube going to the pleura very
difficult. There. an attempt at reconstituting the subarach-
noid space. perhaps with drainage toward the cisterna
magna and enlargement of the latter with a duraplasty is
indicated.
To best treat syringomyelia with shunting procedures
when they are indicated one should be familiar with the
indications
and the operative
procedures
involved in
syringopleural. syringosubarachnoid. and syringoperitoneal
shunting.
References
72
SyringoperitonealShunting
537
1. VernetO.Farmeri-PoMontes]L. Comparison of syringopleural and sy-
ringosubarachnoid
shunting
in the treatment
of syringomyelia
in
children.] Neurosurg 1996;84:624-628
2. Phillips TW. Kindt GW.Syringoperitoneal
shunt for syringomyelia:
a
preliminary report. Surg Neurol1981 ;16:462-466
3. Lesoin F.Petit H. Thomas CE III. et al. Use of the syringoperitoneal
shunt in the treatment
of syringomyelia.
Surg Neurol
1986;25:
131-136
4.
Barbaro NM. Wilson CB. Gutin PH. et al. Surgical treatment
of sy-
ringomyelia:
favorable results with
syringoperitoneal
shunting.
]
Neurosurg 1984:61:531-538
5.
Batzdorf U. Klekamp J.Johnson JP.A critical appraisal of syrinx cavity
shunting procedures.J Neurosurg 1998;89:382-388
6.
Lund-Johansen M. Wester K. Syringomyelia treated with a nonvalved
syringoperitoneal
shunt: a follow-up
study. Neurosurgery
1997:41:
858-864
73
Posterior Approach and In Situ Fusion
of the Thoracic Spine
Trent L.Tredway and Richard G. Fessler
.
Indications
.
Contraindications
.
Advantages
.
Disadvantages
.
Objective
.
Surgical Techniques
Anesthesia and Positioning
Incision and Musculocutaneous
Flap
Decompressive thoracic laminectomy
(DTL) was once the uni-
versal approach
performed
for extradural
spinal cord com-
pression. Indications for this approach included compression
secondary to neoplastic disease, hematoma, or abscess as well
as canal or cord compromise
secondary
to traumatic
frac-
tures.! This approach has been abandoned
over the years with
the increasing
popularity
of the posterolateral
approach.
Moreover, the circumferential
approaches
are directed at the
site of pathology
and have been demonstrated
to be safer
when compared with DTL.However, DTLremains the stan-
dard approach for intradural
pathology as well as dorsally lo-
cated hematoma, abscess, tumor, and laminar fractures.
.
Indications
. Dorsal epidural hemorrhage with incomplete cord in-
jury or cauda equina syndrome
. Spinal cord tumors of the thoracic spine
.
Dorsal laminar fractures with neural compression (may
be associated with higher rate of progressive neurologi-
calloss and deformity when compared with posterolat-
eral approach)
.
Contraindications
.
Ventral neural compression
.
Lifeexpectancy less than 3 to 6 months
.
Medical illness precluding major surgery
538
Decompression
Decortication
and Prepartion for In Situ Fusion
In Situ Fusion
Closure
.
Postoperative Care
. Advantages
. Familiar surgical approach
.
Avoidsthe risks of placing instrumentation
.
Disadvantages
. Does not allow access to ventral compression
.
Potentially lower fusion rates compared with instru-
mented cases
.
Objective
. Toobtain dorsal neural decompression and providesub-
sequent stabilization of the thoracic spine
. Toobtain exposure to the dorsal thoracic spinal cordfor
resection of tumors or arteriovenous malformations
. To obtain tissue biopsy of dorsally located pathology
when
previous less invasive attempts
have failed;
namely,computed tomographically(CT)guidedbiopsy
.
SurgicalTechnique
Anesthesia
and Positioning
General endotracheal anesthesia and placement of a Foley
catheter are performed in the supine position. The patientis
then placed in the prone position on chest rolls with the
T
73
Posterior Approach and InSitu Fusion
539
armstuckedat the side. Caremust be emphasized when
positioningthe arms so that all bony prominences and un-
derlyingneurovascular structures are protected. This will
reducethe occurrence of postoperative pressure sores and
nervepalsies. It is also imperative to allow the abdomen to
hangfreelyto decreasethe venousdistention.especiallyin
thelowerthoracic/upper lumbar region. The anterior chest
wallshould also have adequate room for expansion during
theprocedure.
Incisionand Musculocutaneous Flap
Afteradequate cleansing of the skin with antiseptic solu-
tionand sterile draping of the patient, an incision is made
overthe spinous processes of the levels involved. The lo-
calizationof the incision may be facilitated with the use of
fluoroscopyor with the use of radiopaque markers and
Figure73-1
Anatomy of posterior thoracic spine.
plain radiographs. The incision is carried down to the fas-
cia of the superficial muscular layer. This layer consists of
the trapezius muscle (C2-T12), latissimus
dorsi muscle
(T6-lumbosacral fascia), and rhomboid muscles (C7 and
T1 insertion for rhomboid minor and T2-TS insertion for
rhomboid major) (Fig. 73-1). With the use of a Bovie
electrocautery
and a Cobb periosteal elevator, a subpe-
riosteal dissection is performed through the superficial
muscular layer. The intermediate layer. consisting of the
serratus
posterior
muscle. is then dissected. The deep
muscles
are then
encountered
and dissected
subpe-
riosteally. The deep
layer includes the sacrospinalis.
semispinalis. multifidi. and rotator muscles (Fig. 73-2).2
The use of self-retaining retractors will facilitate the dis-
section and will allow for adequate visualization of the
levels exposed. Furthermore. fluoroscopy or intraoperative
radiographs will help identify and verify the correct levels.
The dissection needs to be performed laterally so that the
transverse
processes are generously exposed. This will
facilitate the decortication
and placement of bone graft
later in the procedure.
Decompression
Once the posterior elements
have been adequately ex-
posed, a dorsal decompression is performed with removal
of the spinous processes and laminae of the involved levels
Figure 73-2
Subperiostealdissectionofthe thoracicspine muscula-
ture.
540
Thoracicand Thoracolumbar Spine: Posterior Approach
Figure 73-3
Dorsal decompression with removal of spinous
processesand laminectomies.
(Fig. 73-3). This is performed with a Kerrison rongeur or
a high-speed drill with a cutting bur. An advantage of
utilizing the Kerrison rongeur is that the bone may be used
for subsequent grafting. One advantage of using the high-
speed drill with a cutting bur is that it may decrease the
operative time and thus the patient's exposure to anesthe-
sia and wound contamination. However, risks associated
with these instruments, including dural tears and injury to
neural structures, should be taken into consideration. At
this time, attention
is turned toward removing the ex-
tradural pathology, which may include tumor, abscess,
hematoma,or bone fragments.
Decortication and Prepartion
for InSitu Fusion
After adequate exposure and decompression
have been
performed, prepartion of the transverse
processes and
facet joints for bone grafting is addressed. The transverse
processes are decorticated with the use of sharp cupped
curettes. A meticulous decortication
will enhance the
fusion process. Next, facet joints should be resected and
the cartilage removed. Thisprocess may also be performed
with the use of a high-speed drill with a cutting bur (Fig.
73-4 ).
Figure 73-4
Decorticationofthe transverse processes and removal
ofthe facetjoints.
In Situ Fusion
Once the transverse processes and the facet joints have
been prepared, autologous bone graft is packed around the
areas of decortication (Fig. 73-5). The graft may be com-
posed of the local bone obtained during the dorsal de-
compression but most often consists of autologous iliac
bone graft.3
Closure
Once hemostasis has been achieved and the bone graft
has been carefully placed, the paras pinal muscles are re-
turned to midline and the fascial layer is reapproximated
with an absorbable suture (0 Vicryl). The subcutaneous
layer is then closed with a 2-0 Vicryl suture and the skin
is closed with a 3-0 nylon suture. A 7 mm Jackson-Pratt
drain may be placed overlying the fascial layer, which
may reduce the formation of postoperative
hematomas
and seromas (Fig. 73-6).
If a cerebrospinal
fluid leak
occurs, the dura should be repaired primarily if possible.
In the event that the dura cannot be repaired
or the
leak is not observable, placement
of a lumbar drain is
recommended.
...
73
PosteriorApproachand In Situ Fusion
541
Figure 73-5
Placement
of bone graft
on transverse
processes and
facetjoints.
.
PostoperativeCare
Althoughthe articulation with the rib cage provides some
stability to the thoracic spine, an external orthosis
is
recommendedwhen performing a fusion without internal
References
1. PopovicEA. Decompressive
thoracic
laminectomy.
In: Kaye and McL
Black, eds. Operative
Neurosurgery.
London:
Churchill
Livingstone;
2000:1817-1826
2.
Papadopolous
SM, Fessler RG.The thoracic
spine. In Benzel. ed. Spine
Surgery:
Techniques.
Complication
Avoidance,
and
Management.
Pennsylvania:
Churchill Livingstone:
1999:157-168
Figure
73-6
Closure of the fascia with placement
of drain.
instrumentation.
Postoperativeradiographsat 1-,3-, and
6-month time intervals will assess any new deformity as
well as a pseudarthroses.
3.
Simpson JM. An H. Posterior exposures of the thoracic spine. In An
and Riley,eds. AnAtlas of Surgery of the Spine. NewYork:Lippincott-
Raven; 1998:31-43
74
18
Pedicle Screw Instrumentation
of the Thoracic Spine
Trent L.Tredway and Richard G. Fessler
.
Indications
. Contraindications
.
Advantages
.
Disadvantages
.
Objectives
There are numerous fixation systems that have been used in
spinal surgery; however.the pedicle screw fixation system is
the only one that engages all three columns. This unique as-
pect relies mainly on the pedicle as the bridge connecting the
anterior. middle.and posterior columns. The pedicle also rep-
resents the strongest point ofattachment to the spine and al-
lowssignificantapplication of forcesin multiple planes.!
The pedicle screw fixation system offers many advan-
tages over other fixation devices. First. the system does not
require that the dorsal elements be intact. Thus these sys-
tems are quite adept for postlaminectomy procedures as
well as for traumatic injuries. Second. the system avoids
placement of instrumentation
into the spinal canal. Hook
systems and sublaminar wires require hardware within the
spinal canal and pose more of a potential risk. Finally,the
fusionrates are reportedly higher.2
Thepediclescrew fixation system isnot without disadvan-
tageswhencomparedwith other systems.Thesystem requires
a sound understanding of the anatomy of the pedicle. espe-
ciallyinthe thoracicspine. Caudaland medial penetration of
the corticalbonecan result in unintended durotomy or neural
injury.Thissytemalsorequiresan increased exposure to allow
forimplantationofscrewsalongthe optimal trajectory.
.
Indications
.
Stabilization of the thoracic spine after neural decom-
pression for tumor or infection
.
Thoracic spinal column fractures
.
Kyphoticdeformities without significant osteoporosis
542
.
Radiological Evaluation
.
Surgical Techniques
Anesthesia and Positioning
Incision and Musculocutaneous
Flap
Placement of Thoracic
Pedicle Screws
Closure
.
Contraindications
.
Significant osteoporosis
(relative contraindication)
.
Medical illness precluding
major surgery
. Lifeexpectancy less than 3 to 6 months
.
Advantages
. Does not require the dorsal elements to be intact
.
Avoids placing instrumentation
within the spinalcanal
.
Higher fusion rates
.
Disadvantages
. Smaller pedicles may result in higher incidenceof duro-
tomy and/or nerve injury.
. Requires increased
exposure
to obtain necessary trajec-
tory
.
Objectives
. To relieve pain, preserve neurological function, and
maintain spinal alignment
.
To obtain stabilization of the thoracic spine utilizinga
three-column construct
. Radiological Evaluation
74
Pedicle Screw Instrumentation
543
Priortosurgicalintervention,anteroposterior(AP)/Iateral
radiographsof the regions to be stabilized and fused are
required.Furthermore, computed tomographic (CT)scan
examinationof the region should be obtained to evaluate
the sizeof the pedicles. Ifsignificant osteoporosis is sug-
gestedby the radiographs and CTscan, then bone densito-
metrymay also be performed because osteoporotic bone
willlead to poor fusion rates and increased instrumenta-
tion pullout.
. SurgicalTechnique
Anesthesia and Positioning
Generalendotracheal intubation is performed and the pa-
tientis then placed in the prone position. Care is taken to
avoidpressure on the bony prominences and neurovascular
structures.The anterior chest wall should have adequate
roomforexpansion during the surgery.
Incisionand Musculocutaneous
Flap
Theskinincision is performed over the spinous processes of
thelevelsinvolved.The construct should include three nor-
mal levels above the pathology and two normal levels
belowthe pathology, and the incision should allow for ade-
quate exposure. The incision is carried
down subpe-
riosteallyalong the spinous processes and laminae and
shouldbe performed to expose the transverse processes.
Placement ofThoracic
Pedicle Screws
Afteradequate exposure of the dorsal thoracic spine has
beenachieved, the pathological process is addressed.
Decompression if indicated. removal of traumatic bone
fragments,and biopsy or resection of tumor or abscess are
performedprior to the stabilization and fusion.
Thethoracic pedicle is oriented in a posterolateral to an-
teromedialdirection by -10 degrees along most of the tho-
racicspine (Fig. 74-1). There is a slight anterior and lateral
angulationof the pedicle at 112.3 It is recommended that
fluoroscopybe available to assist in the proper trajectory. If
the dorsal elements are intact. then a small laminotomy
maybe performed to enable direct palpation of the medial
aspectof the pedicle. It is also important to note that the
thoracicpedicle height is greater than the width and there-
foresmaller-diameter
pedicle screws should be utilized
inthe thoracic region as compared with the lumbar spine
(Fig.74-2).
Theentry site for the pedicle screw is identified by the in-
tersection of a horizontal line connecting the transverse
processeswith a vertical line connecting the middle of the
Figure 74-1
Anatomyofthoracicpedicle.
Figure 74-2
The relationshipofthe pedicle height and widthinthe
thoracicspine.
facet joints (Fig. 74-3).4 A high-speed drill with a cutting
bur is used to remove the cortical surface overlying the en-
try point. Next, a pedicle probe is directed along the path of
the pedicle screw. Asounding probe is then utilized to en-
sure the path has cancellous bone in all directions. AK-wire
is then placed along the path and the trajectory is assessed
with fluoroscopy. After confirmation of the trajectory, a tap
is used to prepare the bone for the pedicle screw. Newer
systems use self-tapping screws, which reduce weakening
of the cancellous bone when compared with the non-
self-tapping screws. The length of the pedicle screw is
544
Thoracicand Thoracolumbar Spine: Posterior Approach
Figure74-3
Entry site for pedicle screw.
determined by the K-wires and should provide penetration
of 50 to 80%of the vertebral body (Fig. 74-4). The screw
strength is determined
by the core diameter
(minor)
whereas the outside diameter (major) is an important factor
in screwpullout.!
Oncethe pedicle screw has been placed. the screws are
connected to a longitudinal member (Fig.74-4). The lateral
aspect of the facet joint and the transverse processes are
decorticated with a curette or bur, and bone graft is placed
over the decorticated areas (Fig.74-5).
Closure
The wound is irrigated with antibiotic irrigation and hemo-
stasis is obtained utilizing electrocautery. The deep fascia is
closed with an absorbable suture (0 Vicryl)and the subcu-
taneous tissue is closed with a 3-0 absorbable suture. The
skin is reapproximated and closed with a 3-0 nylon suture
or with skin staples. Adrain (7 mm Jackson-Pratt) may be
placed overlying the deep fascial layer.The drain is brought
through a percutaneous incision.
Figure 74-4
Pediclescrews penetrating 50 to 80%of the vertebral
body and connecting to a longitudinal member.
Figure 74-5
Placement of bone graft along the transverse processes
andfacet joints after pediclescrewplacement.
References
1. Halliday AL. Zileili M.Stillerman
CB, Benzel E. Dorsal thoracic
and
lumbar screw fixation and pedicle fixation techniques.
In: Benzel. ed.
Spine Surgery: Techniques,
Complication
Avoidance.
and Manage-
ment. Philadelphia: Churchill Livingstone;
1999: 1053-1064
2, Stillerman CB.Gruen JP. Roy R.Thoracic and lumbar fusion: techniques
for posterior stabilizatio~. In: Menezes A. Sonntag VKH.eds. Principles
ofSpinalSurgery.NewYork:McGraw-Hili;1996:1199-1224
74
PedicleScrewInstrumentation
545
3. Papadopolous
SM. Fessler RG.The thoracic spine. In: Benzel, ed. Spine
Surgery:
Techniques.
Complication
Avoidance.
and
Management.
Philadelphia:
Churchill Livingstone;
1999:157-168
4.
Simpson
JM, An H. Posterior
exposures
of the thoracic
spine.
In An
and Riley, eds. An Atlas of Surgery of the Spine. New York: Lippincott-
Raven; 1998:31-43
Section V
Lumbarand LumbosacralSpine
A. Pathology of the Lumbosacral Spine
. 75. Spondylolisthesis in Children
. 76. LumbarDegenerative DiskDisease
. 77. Tumorsof the lumbar and
Lumbosacral Spine
. 78. Trauma of the Lumbar Spine
B.AnteriorApproach
. 79. Computer-Assisted Lumbosacral
Fusion
. 80. Anterior Transperitoneal Approach
to the Lumbosacral Junction
c. AnterolateralApproach
. 81. Anterolateral Retroperitoneal
Approach to the Lumbosacral Spine
D.PosteriorApproach
. 82. Approach to the Lumbar Spine
. 83. Lumbar Laminectomy
. 84. Hemilaminectomy
. 85. Foraminotomy
. 86. Microdiskectomy
. 87. Far-Lateral Microdiskectomy
. 88. Repairof a Cerebrospinal Fluidleak
. 89. Lumboperitoneal Shunt
. 90. Rhizotomy of the Lumbosacral
Nerve Roots
.
91. Resection of Cauda Equina
Ependymomas
. 92. Release of the Tethered SpinalCord
.
93. Resection of Lumbosacral Lipomas
.
94. Transverse Process Fusion
. 95. Posterior Lumbar Interbody Fusion
and Transpedicular Screw Fixation
.
96. Posterior Fusion with Hook-Rod
Instrumentation
.
97. Posterior Fusion with Luque
Instrumentation
.
98. Placement of Morphine or
Baclofen Pumps
. 99. Repair of Myelomeningoceles
. 100. Excisionof a Spinal Congenital
Dermal Sinus/Dermoid
. 101. Resection of Sacrococcygeal
Teratoma
.
102. Surgical Management of Spinal
Dysraphism
.
103. Repair of Diastematomyelia
.
104. Sacral Agenesis
.
105. Selective Dorsal Rhizotomy of
the Lumbosacral Spinal Cord for
Spasticity in Cerebral Palsy
.
106. IliacCrest Bone Grafting
75
Spondylolisthesis in Children
PaulA.Grabb,
MichaelP.B.Kilburn,
BrendaC.Grabb,and
MarkN.Hadley
. Definitions
. ClinicalPresentation
. Diagnosis
. Management
ParsInjurywithoutSpondylolisthesis
Spondylolisthesis
Spondyloptosis
. SurgicalIntervention
. Conclusions
Spondylolisthesis
is derived from the Greek terms spondylo
("spine")and olisthesis ("to slip"). The phenomenon
of
spondylolisthesis
is unique to mammals that load the spine
whenassumingan erectbiped posture. The clinical disorder
hasnotbeendetected in humans who have not walked at
sometime.Ahistory of repetitive loading of the erect spine
hasbeendescribedin populations with a relatively high oc-
currenceofspondylolisthesis.such asathletes and Eskimos.
Theremaybeageneticpredisposition for spondylolisthesis,
withupto 15%of caseshaving associated familial cohorts
withspondylolisthesis.
Childrenwith symptomatic spondylolisthesis most com-
monlypresentwith musculoskeletal back pain. Occasion-
allytheywillhaveassociated radicular symptoms as well.
Plainradiographsare often enough to make the diagnosis
andleadtoreferralto the appropriate specialist. This chap-
teraddressesthe definition. presentation. diagnosis, man-
agement,and outcome of children
with
symptomatic
spondylolisthesis.
We will review the difficulties in the care
ofthesepatients,which are usually related to determining
if,when,andhowthey should be treated.
. Definitions
Wiltsetalclassifiedspondylolisthesisinto five different
typesaccordingto thepresumedetiology: congenital.isth-
mic,degenerative.
traumatic.and pathological.! Thefirst
two
types account
for
the vast majority
of pediatric
spondylolisthesis.
with
the isthmic type being the most
common.
Congenital or dysplastic spondylolisthesis
involves the
lumbosacral junction. There is often a bony abnormality of
the upper surface of the sacrum or posterior
arch of the
most caudal mobile vertebra. The bony abnormalities allow
the spondylolisthesis to occur because of facet subluxation
and pars lengthening. without
the necessity of a pars de-
fect. If the dorsal vertebral arch remains intact. then neural
compression is likely asthe slip advances.
Isthmic spondylolisthesis
is the most common type in
children.
The most frequent
site of involvement
is L5
followed by L4. By definition there is a defect in the pars in-
terarticularis.
which is the bony bridge connecting the su-
perior and inferior zygapophyseal joints. Isthmic spondy-
lolisthesis
is divided
into three subtypes; A. B, and C.
Subtype A is thought to be the result of repetitive bilateral
fatigue fractures of the pars secondary to repetitive exces-
sive axial loading. The bony defect is replaced with fibrous
tissue. Subtype A represents the most common subtype of
isthmic spondylolisthesis. Subtype B is similar to A except
that there is partial bony healing of the pars from repeated
callous formation. This process allows for pars elongation
and anterior slipping of the rostral vertebra over the caudal
vertebra. Subtype Cis very rare and results from acute trau-
matic loading of the pars interarticularis.
which fractures
and is acutely unstable.
549
550
Lumbar and Lumbosacral Spine: Pathology of the Lumbosacral Spine
Isthmic spondylolisthesis is twice as common in males as
in females. There is a 4%incidence of isthmic spondylolis-
thesis in children under 6 years of age, most of whom are
asymptomatic.2.3This is in contradistinction
to children
with congenital spondylolisthesis, of which -30%are symp-
tomatic with cauda equina symptoms.2 The entity of
spondylolysis represents an injury to the pars interarticu-
laris without vertebral slip or spondylolisthesis. Spondylol-
ysis is very likelythe precursor to isthmic spondylolisthesis
in children and adolescents. The presence of spondylolysis
without spondylolisthesis is important to diagnose or ex-
clude in a child or adolescent presenting with back pain, es-
pecially with an acute onset.4.5The term spondyloptosis
refers to complete subluxation of the superior vertebral
body anterior to the inferior vertebral body. Other types of
spondylolisthesis are exceedingly rare in children and will
not be discussed.
. ClinicalPresentation
The uniform complaint of children with spondylolisthesis is
back pain. Unlike adults, in whom back pain is a very com-
mon complaint with an oftentimes unexplained etiology,
back pain in children almost always has an underlying
structural abnormality as its cause. The back pain in chil-
dren with spondylolisthesis carries the hallmarks of me-
chanical pain: activity related and relieved by rest. It is typi-
cally chronic in nature but in certain cases has a rather
abrupt onset. True radiculopathy, either unilateral or bilat-
eral, occurs infrequently. Afrank cauda equina syndrome is
exceedingly unusual as a mode of presentation but has been
reported.
Children who acquire spondylolisthesis
are often ath-
letic or are considered by their parents to be very physi-
cally active relative to other children. There may be a re-
mote
history
of trauma
that
correlates
with
the
beginning of the pain syndrome. The musculoskeletal ex-
amination may reveal tenderness over the lower lumbar
flank region, paravertebral
muscle spasm, and, rarely, a
palpable step between spinous processes at the level of
the slip. In cases discovered
late, an abnormal
lum-
bosacral kyphosis may be present, associated with a com-
pensatory thoracolumbar
lordosis. In this circumstance,
the child may have a shortened gait with a tendency to
flex at the knee and hip joints to maintain an erect pos-
ture. In advanced cases, a prolonged stooped posture is
an effort to relocate the center of gravity shifted by the
slip. Hamstring tightness may develop and its etiology re-
mains unclear. The neurological examination in children
with spondylolisthesis is usually normal, even if they re-
port intermittent
neurological symptoms. In the rare cir-
cumstance when a neurological deficit is present, it most
often involves the L5 nerve root because of the more
common spondylolisthesis of L5over 51.The neural com-
promise resulting in neurological dysfunction is typically
due to both stretch and compression.
In cases of ad-
vanced spondylolisthesis,
multiple nerve roots can be
stretched
and
compressed
creating
the
neurological
cauda equina syndrome.
. Diagnosis
Children who present with low back pain should have plain
radiographs
performed
of the lumbar spine as a first diag-
nostic step. Anteroposterior
and lateral radiographs should
be obtained
with the child standing because the supine po-
sition may reduce the spondylolisthesis.
The lateral viewis
the
most
sensitive
for detecting
spondylolisthesis
(Fig.
75-1),
though oblique views may be helpful by depictinga
defect
in the
pars interarticularis.
Quantification
of the
spondylolisthesis
has been described
by several different
investigators
utilizing different
scales. The presence of> 2
mm of anterolisthesis
of one vertebra
over another on a
standing
lateral radiograph
is diagnostic of spondylolisthe-
sis. Meyerding divides the rostral aspect ofthe inferiorver-
tebral body into quarters.6 The posterior inferior edgeofthe
superior vertebral body is noted to rest overthe mostposte-
rior (grade I) to the most anterior
(grade IV)quarter ofthe
inferior vertebral body. Taillard more preciselyquantifies
the spondylolisthesis
by measuring
the percentage ofslip.
By either
means,
a slip of greater
than 50% or grade IIis
considered
significant.? Sometimes
standing dynamic
lateral radiographs in flexion and extension are helpfulin
depicting the amount of abnormal
motion.
Various determinants
have been cited to attempt to de-
fine the prognosis
of spondylolisthesis,
such as the slipan-
gIe, the lumbar index, the sacral inclination, and the sagittal
pelvic tilt index. Efforts to predict progression by imaging
criteria in isthmic spondylolisthesis have failedtoidentifya
clear predictive
factor, with the possible
exception ofthe
sagittal
pelvic tilt index, which
measures
the relationship
between
52, the center of the hip, and L5.8Higher gradesof
slip at presentation have correlated with the femalesex.
dysplastic 51 segment, and disk height reduction.
Ifa child presents with mechanical back pain, and multiple-
view plain radiographs
are unrevealing,
then radionucleotide
bone scanning and single photon emission computed tomog-
raphy (SPECT)may reveal a pars abnormality (Fig.75-2).In
addition, the radioisotope studies may provide some informa-
tion about the acuteness or chronicity of a knownparsde-
fect,4.9Ifthere is not an increase in uptake the defect canbeas-
sumed
to be chronic. When there is increased uptake, the
defect either may be acute or may be the result of traumatoa
previously existing lesion. In cases of normal plain radiographs
and an abnormal bone scan, computed
tomography (a) may
help define the abnormality
discovered by the bone scan.cr
may be helpful in defining the bony anatomy of the vertebral
bodies involved if one is contemplating
surgical intervention
(Fig. 75-3). Well-defined
sclerotic margins either on plainra-
diographsor on the CTstudies suggesta chronicparsdefect
Children who present with neurological complaintsor
signsshouldbe evaluatedwith magneticresonanceimag-
ing (MRI).MRIis the imaging modality of choiceforchil-
dren in this circumstance and is invaluable in decpictingthe
relationshipsofonevertebralbodyto another, diskdisease.
nerve root compression, and posterior element anomalies.
as well as any previously unrecognized intraspinal pathol-
ogy. MRIwill also reveal signal change within an injured
pars before bony changes are visible by plain radiographsor
CT.In children it is important to diagnose or excludetheex-
istence of an acutely symptomatic pars defect or injuryat
75
Spondylolisthesisin Children
551
Figure 75-1
Lateral radiograph of the lumbar spine showing a
Meyerdinggrade Ispondylolisthesisat L4.
--
--
. I
Figure75-2
Single photon emission computed tomography study
showingincreased uptake at the region of the pars interarticularis
ofLS.
thetime of initial presentation. Appropriate early manage-
ment of children with acute spondylolysis is likely to pre-
ventrecurrent stresseson an acutely injured pars that oth-
erwise might lead to progression of the spondylolisthesis. s
.
Management
Oncethe diagnosis of spondylolisthesis
has been established
a contemporary
management
scheme must
be outlined.
In
the rare circumstance,
spondylolisthesis
in a child will be an
incidental finding,
the child being asymptomatic.
These chil-
dren do not require acute management.
Young athletes with
asymptomatic
spondylolisthesis
have been prospectively fol-
Figure 75-3
Computed tomographic scan through LSshowingthe
bilateralbonydefects inthe parsinterarticularis.
lowed through years of rigorous athletic training. No child
studied had evidence of significant worsening.10In children
with mild spondylolisthesis discovered incidentally there is
no role for activity restriction or, it appears,any other form of
intervention. The natural history of asymptomatic spondy-
lolisthesis is seemingly quite benign. The more clinically rele-
vant circumstance, however, is not the incidental, asympto-
matic patient, but the patient who presents neurologically
normal with back pain of mild to moderate severity.
Pars Injury without
Spondylolisthesis
In the child or adolescent who presents with the acute on-
set of back pain and in whom imaging reveals evidence of
spondylolysis, a pars injury
without
a spondylolisthesis,
there will usually be an activity or event that can be traced
to the onset of pain. These children are most often effec-
tively treated
with nonsteroidal
medication,
strict activity
restriction
for 6 to 12 weeks, and stretching
exercises. They
can then return to activity as long as they remain asympto-
matic. It is unusual
for such a child to require
bracing for
pain relief. Recurrent symptoms or persistence of symp-
toms warrants
reinvestigation
for evidence
of progression
to a spondylolisthesis.
Bracing for 3 to 6 months is recom-
mended
should symptomatic
progression
be noted. Symp-
tomatic spondylolysis presenting
late with evidence of
chronicity has been treated with 3 to 6 months of bracing
with a TLSOwith -80% successratesy,12
Spondylolisthesis
The most common clinical scenario we see in our clinics is
an adolescent child with mechanicalbackpain of weeks
to monthsduration, who is neurologicallynormal and has
a Meyerding grade I slip. The severity of the patient's
pain, the neurologicalstatus, the degree of slip, the type
of spondylolisthesis, and the activity level of the child
must all be considered
to arrive at an individualized
552
Lumbarand Lumbosacral Spine: Pathology of the Lumbosacral Spine
management scheme. In children with a Meyerding grade
I slip or evidence of a pars stress fracture only, moderate
mechanical back pain. and a physically active or athletic
lifestyle, the treatment regimen of nonsteroidal medica-
tions, strict activity restriction
for 6 to 12 weeks, and
stretching exercises often results in long-term symptom
resolution. The more severe the mechanical pain the child
has, coupled with a higher-grade
slip on radiographic
studies, warrants bracing with an external orthosis. Exter-
nal orthoses do not provide predictable
reduction
in
translational motion of the spine but may improve pos-
ture, prevent hyperextension with an antilordotic brace,3
and certainly serve as an activity deterrent
to a child. If
the child presents with a congenital or dysplastic cause of
the spondylolisthesis, concern for failure of conservative
therapy as already outlined is increased.
The vast majority of children with symptomatic spondy-
lolisthesis will respond to nonoperative management. A
small proportion
will remain symptomatic
or worsen
despite these efforts and require operative intervention
(Fig.75-4),13Indications for operative treatment of sympto-
matic spondylolisthesis in children are as follows:
.
Inadequate response to conservative therapy
.
Progression of a slip to greater than Meyerding grade II
.
Imaging evidence of instability in a patient with neuro-
logical deficit (Fig.75-5)
.
Symptomatic Meyerding grade III
. .Symptomatic or asymptomatic Meyerding grade IV
Spondyloptosis
The presence of a neurological deficit should greatly lower
the threshold for surgical intervention. Although a mild
L5 radiculopathy may very well respond to conservative
Figure 75-4
Lateralradiograph of the lumbosacral spine of a 15-
year-oldathletic girlwitha symptomaticgrade Ispondylolisthesisthat
did not respondto nonoperativetherapy.
Figure 75-5
Sagittal, T2-weighted magnetic resonanceimagingof
the childinFig.75-4.
measuresandcanbefollowedclinicallyfora time,thepres-
ence of bowel or bladder dysfunction is an indicationfor
immediate intervention.
. Surgical Intervention
Because all children with spondylolithesis who require
surgery will have intersegmental
instability, somesortof
fusion procedure must be part of the surgical plan.Itisim-
portant for the surgeon to remember that at the levelofthe
spondylolisthesis, the neural arch is free-floating,andthat
the fusion must be performed posterolaterally betweenthe
transverse processes, with or without an interbody fusion.
Ifthe attempted fusion is performed over a singlemotion
segment dorsally between the spinous processes andlami-
nae, stability will not be achieved because, by definition,
the lamina is disconnected from the vertebral bodybythe
pars defects. Issues of reduction, decompression, andin-
strumentation
must be individualized. Dubousset makes
the distinction between the child with a vertical andhori-
zontal sacrum when prescribing surgical therapy.14For
children with spondylolisthesis
and a horizontal sacrum
without neurological involvement he advocates a simple
posterolateral fusion without decompression, reduction,or
instrumentation.
In the unusual circumstance of a child
with a neurological deficit, Dubousset advocates decom-
pression and instrumentation across a single level.Hedoes
not advocate any attempt at reduction of the spondylolis-
thesis beyond postural reduction during the positioningof
the child on the operating table. In a child with a vertical
sacrum and no neurological deficit Dubousset advocatesin
situ fusions, following gradual reduction over 2 to 5days.
He does not attempt to reduce the spondylolisthesis intra-
operatively due to concerns for iatrogenic neurological
deficits. These children remain casted in hyperextension
for 4 months.
75
Spondylolisthesis inChildren
553
Incontrast, Fabriset al treated 12 children with advanced
spondylolithesis by removal of the loosened arch, diskec-
tomy,placement of a temporary reduction device between
the sacrum and U, followed by posterior interbody fusion
and pedicle screw fixation. IS They had no neurological
com-
plications and required no postoperative
casting. They
achievedsolidfusions in all 12 children and mean reductions
of80%of the spondylolisthesis. Muschik et al compared an
anterior procedure alone to a combined anterior-posterior
procedurein children with severe spondylolisthesis.16The ra-
diographic results were better in the anterior-posterior
group,but they could not demonstrate a difference in clinical
outcomebetween the two groups. Molinari et al described
complicationsassociated with three different approaches for
thesurgicaltreatment of high-grade isthmic spondylolisthe-
sisinchiidren.17They concluded that in situ fusion (Group 1)
resultedin a 45%incidence of pseudarthrosis but no surgi-
callyinduced neurological complications. Twoof seven chil-
drenwhounderwent posterior decompression, instrumenta-
tion,and fusion (Group 2) developed pseudarthroses, yet no
childof the 19 they treated with anterior-posterior fusions
with reduction
and
fixation
(Group
3)
developed
pseudarthrosis. They reported no neurological deficits in
Group1 but three transient and one permanent deficit in
Groups2 and 3.They concluded that the increased neurolog-
ical risks of circumferential fusion-reduction procedures
wereacceptable given the fusion success rate, the degree of
reductionachieved, and the improved patient outcome with
regardto pain and function.
It isnot clear from a review of the literature that one ap-
proach or surgical management strategy is clearly better
thananother when considering the surgical management
ofchildren with symptomatic spondylolisthesis. Some ba-
sicconcepts, however, can be forged from the conflicting
literature: (1) fusion alone carries a higher pseudarthrosis
ratecompared with instrumentation with fusion, (2) a sin-
gle motion segment fusion is usually all that is required
(e.g., L5-S1), (3) neural decompression should be accom-
plished when there is a preoperative neurological deficit,
(4) intraoperative attempts at reduction carry an increased
risk of neurological deficits, and (5) reduction
of the
spondylolisthesis
may improve fusion rates and patient
outcome.
Our own approach to these patients, including our surgical
management strategy,embraces manyof these concepts. We
are aggressivewith these patients fromthe outset.Weattempt
to identify the pathology early (pars defect or injury with or
without spondylolisthesis). We insist on absolute activity
restriction and aggressive, consistent medical management.
We will not hesitate to use bracing as an adjunct to our
medicaltherapy and we providelongitudinalfollow-up.Inthe
rare circumstance in which surgery is required,we treat only
the levels of segmental instability in question. We provide
careful neural decompression, do not attempt forcedreduc-
tion, utilizeinterbody fusiontechniques forgradeIandIIslips,
and apply internal fixation instrumentation in addition to
autologousdorsolateralintertransverse fusion(Fig.75-6).
.
Conclusions
In summary, pediatric spondylolisthesis is a relatively com-
mon condition that is uncommonly symptomatic. Conserva-
tive measures of activity restriction, physical therapy, and at
times bracing are usually adequate to allow for healing and
the resolution of symptoms. Infrequently a child with a
spondylolisthesis will fulfill the criteria to undergo opera-
tive therapy. When operative intervention is planned, the
specific procedure must be individualized based on the
presence or absence of neurological deficits, the degree of
the deformity, and risk:benefit ratio of the use of internal
fixation as an adjunct to bony fusion.
A
B
Figure 75-6
(A) Anteroposterior
and (8) lateral radiographs 6 months following posterior decompression,
posterior interbody fusion, and pedicle
screw fixation.
Note the partial
reduction
of the spondylolisthesis
compared
with preoperatively.
References
554
Lumbarand Lumbosacral Spine: Pathology of the Lumbosacral Spine
1.
Wiltse
LL.Newman
PH. McNab I. Classification
of spondylosis
and
spondylolisthesis.
Clin Orthop 1976; 117:23-29
2.
Baldwin NG. Ehni BI..Spondylolisthesis:
sagittal
plane lumbar
spine
deformity
correction.
In: Benzel
EC. ed. Spine
Surgery.
New York:
Churchill-Livingstone;
1999:435-452
3.
LonsteinJE.
Spondylolisthesis
in Children. Spine 1999;24:2640-2648
4.
Anderson
K.SawarkJF.
Conway JJ.Logue ES. Schafer ME Quantitative
assessment
with SPECT imaging
of stress injuries
of the pars interar-
ticularis and response
to bracing.J
Pediatr Orthop 2000;20:28-33
5.
Ralston S. Weir M. Suspecting
lumbar
spondylolysis
in adolescent
low
back pain. Clin Pediatr (Phila) 1998;37:287-293
6.
Meyerding
HW. Spondylolisthesis.
Surg Gynecol
Obstet
1932;54:
371-377
7.
Taillard WE Etiology of spondylolisthesis.
Clin Orthop
1976;117:30-39
8.
Schwab FJ. Farcy JPC. Roye DP.The sagittal
pelvic tilt index as a crite-
rion in the evaluation
of spondylolisthesis.
Spine 1997;22:1661-1667
9.
Mandell GA. Nuclear medicine
in pediatric
musculoskeletal
imaging.
Semin Musculoskelet
Radiol 1999;3:289-316
10.
Muschik
M. Hahnel
H. Robinson
PN. Perka C. Muschik C. Competitive
sports
and
the
progression
of spondylolisthesis.
J Pediatr
Orthop
1996;16:364-369
:..
11.
BlandaJ.
Bethem
D. Moats W.Lew
M. Defects of pars interarticularis
in athletes:
a protocol
for non operative
treatment.
J Spinal Disord
1993;6:406-411
12.
Morita T. Ikata T. Katoh S. Miyake R. Lumbar spondylolysis in children
and adolescents.
J Bone Joint Surg Br 1995;77:620-625
13.
Smith JA. Hu SS. Management
of spondylolysis
and spondylolisthesis
in the pediatric
and adolescent
population.
Orthop Clin North Am
1999;30:487-499
14.
Dubousset J. Treatment
of spondylolysis
and spondylolisthesis in chil-
dren and adolescents.
Clin Orthop 1997;337:77-85
15.
Fabris DA. Costantini
S. Nena U. Surgical treatment
of severe 1.5-51
spondylolisthesis
in children
and adolescents.
Spine 1996;21 :728-733
16.
Muschik M. Zippel H. Perka C. Surgical management
of severe spondy-
lolisthesis
in children and adolescents.
Spine 1997;22:2036-2043
17.
Molinari
RW. Bridwell KH. Lenke LG.Ungacta FF.Riew KD.Complica-
tions in the surgical
treatment
of pediatric
high-grade.
isthmic dys-
plastic spondylolisthesis.
Spine 1999;24:1701 - 1711
..
I
I
76
lumbar Degenerative Disk Disease
MickJ. Perez-Cruet
and Dino Samartzis
.
History of Surgical
Approaches
.
Indications
for Surgery
Patient Selection
Radiological
Evaluation
.
Preoperative
Planning
.
Surgical Technique
Anesthesia
Positioning
Dissection
Disk Removal
Closure
Chroniclowbackpain disablesmillionsofAmericansand
createsa tremendous financial burden on society exceeding
$50billionannually.1 Aretrospectiveanalysisofworkers'
compensation claims at a large industrial manufacturer
withinthe United States found 19%of claims were related
tobackinjuries and these claims were responsible for a dis-
proportional41%oftotal injury costs.2Lowback pain causes
nearly80%of injured workers to miss at least 8 weeks of
work following a back injury.3 In persons younger than
45years old, low back pain is the most frequent cause of
activitylimitation.4 Although only 2%of patients with low
backpainhavean acutediskherniation,a disproportionate
30%of U.S.annual costs for the treatment of low back pain
arespent on this relatively small percentage of patients.s
Clearlythe demographics of lumbar disk disease can allow
forminor improvements in patient care and surgical tech-
niquesto have a major impact.6
There are nearly 15 million physician office visits each
year for low back pain of which 90%of patients improve
within 1 month regardless of treatment.6-8 Failure of non-
operative treatment leads to almost 300,000 patients un-
dergoingvarious surgical procedures to relieve back pain
and sciatica secondary to herniated
lumbar disks, with
mixed results.9 The incidence of lumbar disk herniation
peaksbetween 24 and 45 years of age, with the cases lead-
ingto surgery occurring most often in the 30- to 39-year-
old range.1OHistorically, there has been a male predomi-
nanceinthe incidenceoflumbardisksurgeryrangingfrom
.
Complications
Wound Infection
Injuriesto GreatVesselsor Bowel
CerebrospinalFluidLeakage
Diskitis
.
Outcomes
Difficulty in Rating Outcomes
Outcome of Lumbar Microdiskectomy
Herniated Disk Recurrence
.
Conclusions
1.3:1to 2:1.9-12Kelseyand Ostfeld argue there is no true dif-
ference in incidence of lumbar disk herniation (LDH)be-
tween sexes but that men are chosen more often for surgery
due to their reluctance to miss more work and impatience
awaiting improvement with conservative measures.13Oth-
ers contend that males are more susceptible to disk hernia-
tion because they have more significant disk degeneration
than females, at most age groups compared, due to higher
mechanical stress and worse disk nutrition due to longer
nutrient diffusion pathways.14,ISOther risk factors for LDH
include smoking, presence of narrower lumbar vertebral
canal, sedentary occupations, prolonged motor vehicledri-
ving, and operating vibrating machinery.II.16-21
.
History of Surgical Approaches
The operative treatment of lumbar disk disease has chal-
lenged spine surgeons since the first reported case in 1929
by Walter Dandy.22In 1934, Mixter and Barr published a
description of the pain and neurological symptoms associ-
ated with LDHin a review of 19 patients.23 Mixter and
Barr's lumbar diskectomy technique involved an aggressive
diskectomy. Their approach was used during much of the
cold war era and underwent only subtle alteration until a
less invasive microsurgical dissection was introduced
in
the late 1970s, resulting in shortened hospital stays and
555
556
Lumbar and Lumbosacral Spine: Pathology of the Lumbosacral Spine
faster return to work.24-31Initial published series on this
less invasive lumbar microdiskectomy technique showed
operating times could be cut in half and hospitalizations
reduced by threefold.28 Adebate then ensued over which
microsurgical technique produced the best outcome while
preventing recurrences
and the need for reoperation.
Williams suggested a more conservative approach of re-
moving only a portion of the disk.28Whereas Caspar and
others suggested a more extensive removal of the disk was
needed to prevent recurrent disk herniation and provide
better relief of nerve root compression.24.32Though signifi-
cant differences in technique emerged with the introduc-
tion of lumbar microdiskectomy, the superior outcomes of
the microdiskectomy versus the traditional lumbar diskec-
tomy technique became clear.32-35In part, the success of
these newer, less invasive techniques can be attributed to
the excellent operative optics made possible by the surgical
microscope.33.36.37
There are several forms of therapy available to patients
suffering from an LDH,including chemonucleolysis38-42
percutaneous diskectomy,43-47arthroscopic diskectomy,48-50
and thermal coagulation of the disk.51.52However, a review
of publications
comparing microsurgical diskectomy to
alternative
treatments
shows microdiskectomy
to be a
superior option. Chemonucleolysis is less cost effective
due to fewer successful outcomes and higher recurrence
rates leading to a need for subsequent surgical manage-
ment.53-56 A successful
outcome
with
microsurgical
diskectomy is four times more likely than with chemonu-
cleolysis and 6.8 times more likely than with percutaneous
diskectomy while decreasing the risk of recurrence. Com-
pared with microdiskectomy, reoperation is 16 times more
likely following chemonucleolysis
and 49 times more
likely after percutaneous diskectomyY Automated percu-
taneous diskectomy has proven to be less efficacious and
less cost effective than microsurgical diskectomy.52.58.59
Arthroscopic diskectomy requires considerable training to
be proficient, although similar outcomes have been re-
ported compared with open diskectomy.6o Currently, the
open lumbar diskectomy is the gold standard treatment of
LDHrefractory to nonsurgical
therapies
and to which
other diskectomy techniques should be measured in terms
of cost savings, technical ease, and patient and surgeon
satisfaction.57.61
.
Indications for Surgery
Patient Selection
A careful preoperative
history and physical exam are neces-
sary in selecting the best candidates
for lumbar diskectomy.
The duration
of radicular
symptoms
is important because
patients with radiculopathy and back pain longerthan
6 months have more severe disk pathology and poorerout-
comes.62-65Physical findings such as lumbar range ofmotion
and crossed Lasegue's sign may be useful in predicting the
degree of herniation and thus which patients will havebetter
outcomes.66 The difficulty
in any technique
used to treat
lumbar disk herniations
is determining
which patient char-
acteristics
predict a good clinical outcome. Because lowback
pain has a multitude
of causes with only a small minority
due to disk herniation,
patients must undergo careful clinical
and radiographic evaluation to determine if microdiskectomy
is likely to result in a symptom-free
lifestyle worth the cost
and risk of surgical intervention.
Most surgeons will follow
patients with low back pain and radicular symptoms initially
with a trial of bed rest, analgesics,
and physical therapy for
several weeks before pursuing expensive radiographic stud-
ies and entertaining
surgical options.67 For those with more
severe symptoms and physical findings, surgical treatment is
associated with greater improvement
than continued conser-
vative treatment.68.69 Several objective
rating systems that
often include neurological signs, root tension signs,radi-
ographic findings, and psychosocial factors have shown some
success in predicting
surgical outcomes.7°-72 Cerebrospinal
fluid (CSF)markers such as neurofilament and S-100proteins
may provide additional evidence of axonal and Schwanncell
damage due to lumbar disk herniation compressingthe
nerve root,73although these marker studies are rarelyor-
dered. Nevertheless, radiographic studies cannot replacea
careful preoperative history and physical examinationin
helping to determine the etiologyof clinicalsymptoms.
Radiological Evaluation
If low back pain and radicular symptoms persist despite
conservative measures, diagnostic studies are warranted
(Fig.76-1A,B),though a positiveradiographicresultmust
A
Figure 76-1
(A) Sagittal and (B) axial T2-weighted magnetic resonance image showing LS,51 herniated disk.
B
76
LumbarDegenerative DiskDisease
557
besupported by clinical evidence of nerve root compression
because false-positive radiographs are not uncommon.74
Weiselet al studied neuroradiologists' readings of lumbar
computed tomographic
(Cf) scans on 52 volunteers
with no
back or leg pain and found radiographic abnormalities in
37%of these "normal" individuals, particularly
in those over
40years 01d.75Jensen et al performed
magnetic
resonance
imaging (MRI) examinations
on 98 people
without
back
painand found 52% had a bulge at one level, 27%had a pro-
trusion, and one patient had an extrusion.76 Progressive de-
generation and weakening
of the spine and associated
disk
dessication and collapse
are normal
processes
of aging77
and current or future symptoms
cannot be predicted
on ra-
diographic evaluation
alone. Weber prospectively
studied
283 patients
with
clinical and myelographic
evidence
of
lumbar disk herniation
and followed 126 patients
with un-
certain indications
for surgery by randomizing
them into a
continued conservative
therapy group and a surgically man-
aged group.12 He found that surgery
offered better
short-
term outcomes: 90%patient satisfaction in the operative
group versus 66% satisfaction
in the nonoperative
group at
1year.However,this difference lost statistical significance
at 4 years with 80%versus 68% patient satisfaction,
respec-
tively.Shvartzman
et al reviewed
55 male truck drivers fol-
lowing 3 months
of conservative
therapy
trial due to an
acute WH. They found no significant
difference
in outcome
or costs between
continuing
conservative
therapy
versus
surgery, but the
nonsurgical
group
did miss
18.5 more
weeksof work than the surgical group over 5 years.5There
isa clear advantage
to surgical treatment
when patients are
seeking immediate
relief of pain and a faster return to work.
For a patient with persistent
lumbar radiculopathy,
radi-
ographic evaluation
is performed
to determine
if nerve root
compression by an LDHcorresponds
to the clinical picture.
MRIhas replaced CTand myelography as the initial test of
choicefor several reasons. MRIis less invasive,obviating the
needfor lumbar puncture and contrast injection and shows
soft tissue disk pathology
very well. MRI appears
to corre-
latemore accurately with surgical findings at a rate of 90%
for MRI versus
77% for Cf,77 and MRI is more
sensitive
(100%)than diskography
(86%) in detecting
LDH.78If MRI
failsto provide clear evidence of LDH,myelography
is a use-
fuladjuvant study, particularly
when no neurological
deficit
exists, multiple
nerve roots are involved, or a centrally
her-
niateddisk affects only a single root.79Regardless of which
radiographic study is chosen, clear evidence
must be seen
onthe images that a protruding disk is compressing a nerve
rootand causing the clinical symptoms to prevent clinical
failure and produce a good outcome
from surgery.80
The type of disk herniation (Le., protrusion versus se-
questration)
seen on imaging studies
has also been shown
to be of some predictive value. The etiology of a lumbar disk
herniation is felt to be the protrusion of the soft inner nu-
cleuspulposus through the outer anulus fibrosus, which is
anatomicallyweaker in the posterior portion of the disk due
tofewerlamellar bundles and the greater percentage of dis-
continuous lamellar bundles.81 In a 2-year outcome study,
anatomical continuity of the anulus fibrosis following disk
herniation was important in predicting surgical success be-
causepatients with disk fragments extruded through a rup-
tured anulus had better clinical outcomes, at 2 years, com-
pared with patients with contained herniation or negative
disk pathology.82Other predictors of good outcomes from
surgery include: no preoperative comorbidity, duration of
sciatica less than 6 months, lack of previous nonspinal
surgery, absence of back pain or work-related injury, radic-
ular-distribution
pain extending
to the foot, positive
straightlegraise(SLR)examwithoutbackpainon SLRtest,
and reflex asymmetry; each independently predicted a
good outcome from lumbar disk surgery.62.82-85
Afactor that has been found to correlate with a poor sur-
gical outcome is work-related injury or accident-related
compensation.86Compensationpatientshaveahigherinci-
denceof pain,disability,unemployment,andpsychological
problems when these groups are compared with noncom-
pensation back injury patients, and claim settlement did
not reduce morbidity up to 5 years laterP Davisconcluded
compensationpatients were the leadingdisincentivesfor
good functional outcome because over half of the 30 pa-
tients with a poor result in his reviewof 984patientshad
compensationclaimspendingat the time of surgery.88The
presenceof psychosocialissuessuchas lowpainthreshold
and obvious secondary gain certainly complicate patient
selection
for successful
lumbar
microdiskectomy
and
emphasize the need to have indisputable radiographic evi-
denceof lumbardiskherniationcompressingthe nerveroot
correspondingwith aneurologicaldeficitbeforedecidingto
perform surgery.
.
Preoperative Planning
Aftercarefullyselectinga patient
to
undergo
lumbar
microdiskectomy,the patient should have routine preopera-
tive screening
tests
including
complete
blood count,
urinalysis, and coagulation profile (PT/PIT).An electrocar-
diogram and chest x-ray are performed if the patient isover
40yearsold.Patientsareconsentedand advisedof possible
operative and postoperative complications.
Patient education plays an important roll in successful
spinal surgery.89 Patients are educated preoperatively on
their diagnosis, treatment plan, possible complication, ex-
pected postoperative course, wound care, and follow-up.
Postoperative education and discharge instructions help to
reinforcemanyof these conceptsand reducepatientanxi-
ety and unnecessary returns. Complications include the fol-
lowing: persistent
incisional pain, infection, significant
bleeding, dural tear, nerve root injury resulting in motor
and or sensory deficits, bowel and bladder incontinence,
sexual impotence, visceral or major vascular injury, no im-
provement in or worsening symptoms, and recurrent disk
herniation.31.61.89-95
It is important to explain to the patient
the significanceof a dural tear because this has been a
sourceof litigationdespiteitsrelativelybenignclinicalsig-
nificance if treated properly. Additionally, risks associated
with general anesthesia are discussed.89
Patients are asked to report to the hospital on the day of
surgery and to follow routine preoperative
anesthesia
protocol including nothing by mouth after midnight the
night before surgery. In addition, patients are asked not to
take anynarcoticsthe morningof surgeryinthe eventthat
558
Lumbar and Lumbosacral Spine: Pathology of the Lumbosacral Spine
questions arise concerning surgical consent. Any medica-
tion that could contribute to increased bleeding should be
stopped 10-14 days before surgery to allow the patient's
coagulation status to normalize. This is particularly impor-
tant in patients taking aspirin daily.Microdiskectomy is not
advised in patients with drug- or metabolic-induced coagu-
lopathies before correction because even mild oozing will
significantly compromise surgical exposure and risk inad-
vertent nerve root injury. Usually, a simple bleeding time
and coagulation profile is all that is needed to evaluate pa-
tients in question.
. SurgicalTechnique
Anesthesia
Oncein the operating room the patient is placed under gen-
eral endotracheal anesthesia with an inducing agent and a
short-acting muscle paralytic agent. A Foley catheter is
placed to prevent bladder distention and is removed before
the patient is awakened after the procedure. This is particu-
larly important in older males, who may have a hyper-
trophic prostate, and helps reduce postoperative voiding
difficulties.
Positioning
The patient is then placed prone onto a Wilson frame or
suitable substitute cushioned with an egg crate mattress,
with arms placed to the side, or bent at < 90 degrees on
arm supports in larger individuals. All pressure points are
properly padded. The back is shaved and cleansed thor-
oughly during preparation for surgery and an intravenous
cephalosporin is administered prior to skin incision to mini-
mize the risks of postoperative infection. Palpation is used
to estimate the level of the incision, the iliac crest correlat-
ing roughly to the L4-L5level, a no. 18-gauge spinal needle
is placed, and then lateral flouroscopy is taken to confirm
the levelof incision.
Dissection
A midline skin incision is made averaging 2 cm in length.
The subcutaneous
fat is separated
with
Bovie electro-
cautery,
using an insulated
tip to prevent
skin burns, until
the lumbodorsal
fascial plain is reached. The fascial fibers
leading to the midline should be visualized and the spinous
processes easily palpated. Alinear incision is then madein
the lumbodorsal
fascia connecting
spinous process to spin-
ous process.
Using a straight,
sharp 1 cm wide osteotome
the perispinous
muscles are carefully dissected in a subpe-
riosteal fashion from the spinous processes and the laminar
portion closest to the operative inter-space. It is not neces-
sary to strip muscle off the entire spinous process and lam-
ina for adequate
exposure because this will only cause more
bleeding
and postoperative
patient
discomfort.
Usingthis
method
the surgeon always has a clear midline anatomical
reference
point. A subperiosteal
muscle dissection usinga
sharp osteotome
allows the muscle to fall back toward the
spinous process and lamina after closure and reduces intra-
operative
bleeding.
It is important
that the osteotome is
kept on bone to prevent
inadvertently
entering
the spinal
canal. The dissection
of muscle
should
not be forced but
rather a teasing away of the muscle off the bone.
To reduce the amount
of tissue trauma while optimizing
surgical field visualization, a retractor (P-C Microsurgical
Spinal Retractor,
Codman,
Randolph,
MA) was developed
that allows excellent soft tissue retraction
(Fig. 76-2). Other
retractor systems that have been used effectively include
the METRxMD and METRxMED (Fig. 76-3, Sofamor-Danek,
Memphis, TN).
The P-C Microsurgical Spinal Retractor was designed with
several features
to improve the surgeon's
operative fieldof
view. The retractor
has a hinged handle, which allows it to
conform to the lordotic curvature
of the spine and reduces
its profile. This also helps to prevent
rotation
of the retrac-
tor once placed and allows the retractor blades to beangled
in a caudal or rostral direction in the wound. The medial
blade is flat and toothed to improve retraction
of the medial
structures,
limit rotation,
and prevent
tearing of the inter-
spinous ligament. The fanned tooth blade at the depthof
the retractor
assures adequate
exposure
of the medial facet
A
Figure 76-2
(A) The P-C microsurgical
spinal retractor,
(B) in place for lumbar microdiskectomy
procedure.
B
rrrrrrrr
76
Lumbar Degenerative DiskDisease
559
....-.
,..
..
A
Figure76-3 (A)METRxMDand(8) METRxMEDretractorsystems.
complexfordrillingbyproperlyretractingsofttissueat the
depthofthe wound.Cautionneeds to be exercisedwhen
makinga small skin incision because it is relatively easy to
anglethe dissection toward the wrong level; therefore, a fi-
nallateralflouroscopicview or plain x-ray is taken to assure
theproperlevelwith the retractor in place.
TheMEDTRxMDretractor approaches the spine using a
seriesof progressively larger muscle dilators (Fig. 76-3A).
Usingthis technique the muscular attachments
to the
midlinebone structures are preserved. Initially a K-wire is
advancedunder fluoroscopic guidance a fingerbreadth lat-
eralto the midline at the disk space of interest and docked
on the facet-laminar junction
(Fig. 76-4A,B). An initial
muscledilator is then passed over the K-wire (Fig. 76-4C)
and the K-wire is removed (Fig. 76-4D). Subsequent
muscledilators are then passed, over which a tubular
retractorworking channel is placed (Fig. 76-4E). The re-
mainderof the procedure is then performed through the
tubular
retractor
under
microscopic
visualization
(Fig.76-4F). The METRxMEDsystem is similar except for
the operation
being
performed
under
endoscopic
visualization (Fig. 76-5A,B).96 Each of these techniques
hasdistinct advantages and disadvantages (Table 76-1).
TheMETRxMDgives better optical resolution and three-
dimensional visualization. The METRxMEDsystem also
usesa 30 degree angle endoscope, which allows visualiza-
tionofanatomical structures which increases the field of
vision.Additionally, the METRxMEDsystem allows heads
upviewingof the operative field on the video monitor re-
gardless of the position
of the tubular
retractor
(Fig.
76-5A,B).Disk removal is performed similarly using the
threementioned retractor systems, and all result in cos-
meticallypleasing skin incisions (Fig. 76-5A,B) and can be
performedon an outpatient basis.
Lumbarmicrodiskectomy remains the gold standard to
whichother techniques for treatment of LDHshould be com-
pared.Thereare advantages and disadvantages to the METRx
MDandMETRxMEDtechnique compared with the more tra-
ditionallumbar microdiskectomy (Table 76-1). Individual
surgeonsmust determine, based on clinicaltraining and com-
fortlevel,which technique is most appropriate for their clini-
calpracticeto assure patient safetyand improve outcomes.
....-
B
Disk Removal
Asmall amount of muscle and soft tissue is removed from
the medial facet complex and interlaminar space to allow
adequate bone visualization. Using a high-speed drill with
a 4 to 6 mm diameter circular cutting bit, bone is removed
from the caudal aspect of the rostral lamina and medial as-
pect of the inferior facet of the superior vertebral body.The
drill bit is then changed to a diamond bur 4 to 6 mm in di-
ameter, which is less likely to cause dural perforation or
nerve root injury. Drilling of the medial portion of the
superior facet of the inferior vertebral body and the rostral
aspect of the more caudal lamina is performed. Care is
taken in drilling the caudal lamina (S1lamina in the figure)
because the nerve root is typically pushed up against its
underside. Cool irrigation is continuous
during drilling
with the diamond bur to prevent heat injury to the nerve
root.
Intraoperative
irrigation
with
cool saline
may
also have additional benefits in reducing postoperative
pain.97
The thinned-out
bone over the ligamentum flavum is
removed with an upgoing curette
or Kerrison punch.
Adequate bone removal ensures exposure of the nerve root,
limits the need to retract an already compromised nerve
root, and allows adequate visualization of the herniated
disk. An upgoing curette is then used to sweep under the
lamina and detach the ligamentum flavum, which is re-
moved with a Kerrison punch. Throughout the removal of
the ligamentum flavum a ball-ended probe lifts the liga-
ment and protects the underlying nerve root from injury.
The epidural fat can be reduced using a bipolar to allow
adequate visualization of the nerve root and underlying
disk herniation. With adequate medial facetectomy,very lit-
tle if any nerve root retraction is needed. A long, tapered
bipolar; long, thin bite plate Kerrisonpunches; and long, ta-
pered curettes ensure optimal visualization and surgical
working space. Drilling can also be performed more easily
with a long, thin, streamlined drill, which does not impede
the surgeon's view.
The disk anulus is coagulated and then cut with a no. 11
blade, and the disk is removed, initially with a small micro-
pituitary
that
minimizes nerve root compression
and
.
560
Lumbarand Lumbosacral Spine: Pathology of the Lumbosacral Spine
c
D
A
,:,
~
I
1:
.~... B
E
---
F
Figure 76-4
(A) placement
of K-wire through
skin. (B) under fluoroscopic
guidance.
(C) with first dilator
placed over K-wire. (D) under fluoro-
scopic guidance. (E) sequential muscle dilators.
and (F) tubular
retractor
in place and affixed to flexible arm assembly.
trauma. Oncethe nerverootisdecompressed.a largerpitu-
itary rongeur is used to remove additional disk material.
Downgoing curettes can then be used to pass under the
nerve root below the plane of the disk anulus to push
additional herniated disk fragments into view. Aggressive
scraping of the vertebral end plates is avoided to reduce
postoperative
back pain. aseptic diskitis. and the risk of
damage to the great vessels.98Inspection under the nerve
root is performed using a ball-ended probe to assure com-
plete decompression.
Free disk fragments
located medially
can be teased
into view using
the
ball-ended
probe or
downgoing
curette and removed with the pituitary rongeur.
The nerve root is then followed
out through
the foramen
and additional
bone removed
over the foramen
to prevent
any compression
ofthe root that may occur with postopera-
tive disk subsidence.
r
76
Lumbar Degenerative DiskDisease
561
A
Figure76-5
(A) Endoscopic attachment
to tubular retractor.
(B) Fluoroscopic images showing working through tubular retractor with Kerrison
punch.
B
Closure
Thewoundis irrigated with cool saline solution and all
bleedingcontrolled with electrocautery or a small piece of
Gelfoam.NoGelfoamis left in place after closure. One 2-0
Vicrylstitch is placed deep to reapproximate the muscles
loosely.Thelumbodorsal fascia is approximated with one or
two2-0 Vicrylstitches, the subcutaneous fat is approxi-
matedwith2-0 Vicrylstitches, and the skin is closed with a
runningsubcuticular 4-0 chromic suture. Toprovide addi-
tionalpostoperativepain comfort, 10 mLof 0.5%bupivacaine
(AbbottLaboratories, Chicago, IL) is injected around the
incision.99Mastisol (Ferndale Laboratories, Ferndale, MI),
Steri-Strips(3M Health Care, St. Paul, MN), or Dermabond
(Ethicon,Somerville, NJ) followed by sterile dressing is
applied.The patient is reminded to remove the wound
dressing covering in the morning but to leave the Steri-
Strips in place for 4 to 7 days to allow adequate skin approx-
imation and healing. The patient is advised not to submerge
the wound for 10 days but that showers are allowed.
Postoperative pain is controlled with oral analgesics. In-
structions are given to avoid strenuous activity for the first
1 to 2 weeks and no heavy lifting (> 30 pounds) for at least
30 days to minimize chances of reherniation
of disk
material. Using minimally invasive approach techniques
the majority of patients can be discharged on the day
of surgery. Patients can usually return to work within 1 to
2 weeks. The patient
is seen in clinic 2 weeks after
discharge and is advised to return sooner if there is any sign
of wound infection, fever, persistent
wound drainage,
unremitting pain, or progressive neurological deficit. Exer-
cise for the first months should be restricted to walking,
Table76-1
Advantages and Disadvantages of Traditional vs METRxMDand METRxMED
Lumbar Microdiskectomy
Techniques
Advantages
Disadvantages
TraditionalMicrodiskectomy
Excellent visualization
Familiar approach
Mayrequire 2- to 3-day hospital stay
Requires muscle retraction
METRxMDMicrodiskectomy
Outpatient procedure
in most cases
Excellentvisualization
with operative
microscope
Reduced pain and rapid return
to work in most cases
Learning curve to master in most cases
METRxMEDMicrodiskectomy
Outpatient procedure in most cases
Reducedpain and rapid return to work in most cases
Endoscopecan look beyond confines of tubular retractor
Fairto good visualization
Steep learning curve to master
Two-dimensional visualization
562
Lumbarand Lumbosacral Spine: Pathology of the Lumbosacral Spine
stationary bike, or swimming and should not include run-
ning,heavyweight lifting,or contact sports.89
.
Complications
Wound Infection
Woundinfections appear to be the most common complica-
tion, accounting for up to 75%of complications in some se-
ries.88.92The use of perioperative intravenous antibiotics can
help to reduce wound infection rates to < 2%,10with the an-
tibiotic given just before surgery being the most effective
dose.lOO.1Ot
Injuries to Great Vessels or Bowel
Potentially the most lethal complication is injury to great
vessels or the bowel due to penetration of the anterior disk
anulus and longitudinalligament.95.102.103
Fortunately, these
insults occur rarely, with a frequency < 0.2%,but do require
immediate emergency laparotomy and repair of injury with
appropriate fluid and blood replacement.104.105Careful at-
tention when inserting instrument tips into the disk space
and use of three-dimensional
microscopic visualization
helps to prevent this complication.
Cerebrospinal Fluid Leakage
CSFleaks following intraoperative dural tears are also infre-
quent and can lead to pseudo meningocele if not repaired.
Williams found CSF leakage occurred in 5% of his mi-
crodiskectomy cases whereas Davis'sreview showed a < 1%
frequency of this complication, perhaps reflecting improve-
ments in microscope optics and technique.28.32.88Persistent
CSFleaks leading to formation of pseudomeningocele can
be prevented by immediate recognition and repair of a
dural tear. Because the procedure described here uses a
smallincision, a primary dural repair can be difficult. Dural
tears can be avoided by using a diamond bur for deep bone
removal,careful dural retraction with a probe or cottonoid,
and adequate visualization at all times, especially when us-
ing the Kerrisonpunch to remove bone tissue. However, if a
leak does occurs and cannot be repaired primarily, a small
piece ofDuraGen(Integra LifeSciencesCorporation, Plains-
boro, NJ)can be applied and covered with fibrin glue. Pa-
tients should be well advised of this complication prior to
surgery.1Ot
Diskitis
One of the least common yet extremely painful complica-
tions of lumbar disk surgery is diskitis, which can be di-
vided into two groups; septic or aseptic.90Septic diskitis can
be distinguished by the presence of fever,leukocytosis, pos-
itive bacteriologic cultures or histologicalfeatures on biopsy
as well as by an increased erythrocyte sedimentation rate
and C-reactive protein.90.106-109
Sterile or "aseptic" diskitis,
due to chemical irritation and inflammation ofthe cartilagi-
nous end plates of the vertebrae, results in no woundinfec-
tion or fever, and negative cultures. Wilson and Harbaugh
reported this occurred in 2.3%of cases and blamed it on
pieces of cartilage flaked off during radical disk removal.32
The two entities need be distinguished becausesepticdiski-
tis requires antibiotic therapy whereas aseptic diskitiscan
be treated with rest, analgesics, and physicaltherapy.
In the limited number of published series involvingout-
patient lumbar microdiskectomy patients, complications
occurred infrequently. Ofthe 116patients reported inthree
different series, CSFleakage occurred in just two patients
whereas there were no injuries to major vessels or bowel,
no episodes of diskitis, and no wound infections.21.94.11o.1
The decrease in wound infection rate and diskitis withan
outpatient technique may be in part due to less tissuedis-
section and reduced hospital exposure to nosocomialor-
ganisms.
.
Outcomes
Difficulty in Rating Outcomes
There are numerous
articles published
regarding outcomes
of lumbar
disk surgery;24.28.30.85.112-115
however, comparing
outcomes
of different
series can be difficult due to different
outcome
rating scales, surgical approaches,
selection crite-
ria, and outcome
interpretation.
Success rates vary signifi-
cantly (60-97%) based on methodology
of the study as well
as criteria
used to select a surgical
candidate,116.117Results
are often based
on patient
response
to mailed question-
naires or retrospective
reviews of clinical charts, offeringa
subjective
or biased impression
of surgical outcomes. Vari-
ous outcome
scales have been developed
to better define
outcomes.84.101.116.118-121Of these
the
Prolo
scale, which mea-
sures economic and functional status of the patient before
and after treatment, is relatively easy to apply but hasnot
been universally accepted to rate results of lumbar mi-
crodiskectomy.93However, the simplicity of this scalemay
not tell the full story because there is no pain scaleoranal-
gesic use rating. Technical variations between surgeonsin
performing lumbar microdiskectomy also make compara-
tive studies difficult. Some authors describe microdiskec-
tomy as any diskectomy using a microscope regardlessof
the lengths of the incision, whereas others use the termto
describe a "small"surgical incisionoperating with loupe
magnification.
Outcome of Lumbar Microdiskectomy
Results of lumbar microdiskectomy
have been excellent and
this procedure remains the gold standard to which other
diskectomy
techniques
should be compared.
Williams de-
fined surgical "cures" as those patients
being economically
productive
and comfortable
without
the need for analgesic
medications. He reported a 91%cure rate in 530 patientsfol-
lowed up to 5.5 years after surgery.28 His extended follow-up
series of 1051 patients over 20 years showed that 98.4%
were surgical "cures."122More recent outcome reports using
theProloscaleshow the majority of patients are able to re-
turntotheir previous occupation on a limited basiswithout
backor leg pain but do experience occasional symptoms of
sciatica.88.92
Long-term retrospective analysis of lumbar
microdiskectomyhas shown successful outcome of 91%at
6 months, which
declined
to
83% success rating
at
10years.123
However,closely controlled prospective studies
havereportedsignificantly lower successrates ranging from
73to77%.86This may be a more realistic outcome expecta-
tionbecausedegenerativedisk diseaseis an ongoing process
thatismultifactorial in nature, with patients having persis-
tent back pain despite relief
of
radicular
symptoms.
Compensationalso plays a role in predicting outcome be-
causehalfof patients with a Prolo score of 5 out of 10or less
hadcompensationcasespending at the time of surgery.88
HerniatedDiskRecurrence
Adebateoverthe correlation between the amount of disk
materialremovedand recurrence rates has been ongoing
sincethefirst reports of microsurgical lumbar diskectomy.32
Theconservativesurgical approach by Williams to remove
onlytheprotruding disk fragment rather than more exten-
sivediskectomyinitially encountered a recurrence rate of
9%,whereashis later series reported a recurrence rate of
only5.5%.28,122Heconcludedthe spontaneousreherniation
ratewasonly 1.9%becausetwo thirds of the total rehernia-
tionsweredueto new lumbar trauma.122The work ofWiI-
sonandHarbaughusing a more aggressive diskectomy to
removethe entire disk material in hopes of preventing re-
herniationshowedan only slightly lower recurrence rate of
4%.32
Adrawbackto this technique was an increased risk of
postoperativediskitis and back pain, possibly due to greater
irritationof the vertebral end plates.30,98One study of out-
patientmicrodiskectomy casesrevealed a more staggering
differencein recurrencesbased on amount of disk removal
as21%had recurrence when
only
the ruptured
disk
fragmentwas removed compared with
no recurrences
whenthe technique was modified
to remove more disk
material.94
Resultspooled from nine different series show a
3.7%recurrencerate in casesinvolving more extensive disk
excision,compared with
a 7% rate after
limited
disk
removaI.57Striffeler et al provided
conflicting
evidence
References
1. Frymoyer
JW.Cats-Baril WL An overview
ofthe
incidences
and costs
oflowback pain. Orthop Clin North Am 1991 ;22:263-271
2. Spengler DM,Bigos SJ.Martin NA. ZehJ. Fisher L. Nachemson
A. Back
injuries in industry: a retrospective
study, I: Overview and cost analy-
sis.Spine 1986;11 :241-245
3. MayerT. Gatchel J. Mayer H. A prospective
two-year
study of func-
tional restoration
in industrial
low back
injury.
JAMA 1987:258:
1763-1767
4. Kelsey
J. PastidesH.Bisbee G.Musculoskeletal Disorders: Their Fre-
quency of Occurrence
and Their
Impact
on the
Population
of the
UnitedStates. Edited. New York: Prodist: 1978
5. Shvartzman L.Weingarten
E. Sherry H, Levin S, Persaud A. Cost-effec-
tiveness analysis of extended
conservative
therapy
versus surgical in-
tervention in the management
of herniated
lumbar
intervertebral
disk.Spine 1992:17:176-182
76
LumbarDegenerativeDiskDisease
563
because their
comparison
of Williams's
"conservative"
microsurgical diskectomy with the technique of Casparand
Lowe revealed no significant difference between the two
groups in reoperation rates due to recurrence.124Their find-
ing is substantiated by another retrospective review report-
ing no increase in reherniation or reoperation rates when
disk fragment excision alone was compared with diskec-
tomy and end plate curettage.98
Dealing with reherniation or return of sciatica symptoms
can be one of the most difficult
issues of lumbar spine
surgery. The incidence of repeat surgery for recurrence
ranges from 1.7%to 8%.95,113,114.125-128
Reportsvary greatly as
to how successful repeat microdiskectomy can be because
44 to 100%are able to return to work and 56 to 88%havere-
lief of sciatica.129-132The time interval between the initial
procedure and recurrence of symptoms is an important pre-
dictor of successwith reoperation becausepatients with re-
current sciatica at least 1 year after the first operation have
better outcomes, with successrates comparable to those of
primary microsugicallumbar
diskectomy.13o.131
Thesurgical
findings during repeat surgery for recurrent sciaticaare also
important
in predicting outcome because the presence of
scar formation and lack of definitive reherniation results in
less satisfactory
outcomes.133 We
recommend
repeat
surgery for recurrence of lumbar disk diseasebe performed
as an inpatient
procedure because these patients may re-
quire more extensive dissection and exploration to relieve
their symptoms, thus requiring
more rigid postoperative
pain management and observation.
. Conclusions
Lumbar microdiskectomy can be performed safely and
successfullyon carefullyselected patients. Toachievethe best
results, patients must have radiographic evidence of a herni-
ated lumbar disk compressing the nerve root to a degree
significant enough to cause a clinical radiculopathy at the
corresponding level.Patients with secondary gain fromtheir
symptoms, such as pending litigation or compensation sur-
rounding an injury,have delayed recoveryperiods and poorer
overallsurgical results.134-137
Realisticpatient expectationsare
achievedwith adequate preoperative patient education.
6.
Spengler
DM. Lumbar diskectomy:
results with limited disk excision
and selective foraminotomy.
Spine 1982;7:604-607
7.
Dixon A. Progress and problems
in back pain research. Rheumatol
Re-
habil1973;12:165
8.
Hart LG. Deyo RA. Cherkin
DC. Physician
office visits for low back
pain:
frequency.
clinical
evaluation.
and treatment
patterns
from a
U.S. national
survey. Spine 1995;20:11-19
9.
Bruske-Hohlfeld
I. Merritt JL. Onofrio BM. et al. Incidence
of lumbar
disk
surgery:
a population-based
study
in Olmsted
County.
Min-
nesota.
1950-1979.
Spine 1990:15:31-35
10.
Spangfort
EV.The lumbar disk herniation:
a computer-aided
analysis
of2,504
operations.
Acta Orthop Scand SuppI1972:142:1-95
11.
Kelsey JL.Githens PB. O'Conner T. et al. Acute prolapsed
lumbar
inter-
vertebral
disk: an epidemiologic
study with special
reference
to dri-
ving automobiles
and cigarette
smoking. Spine 1984:9:608-613
43.
Davis GW. Onik G. Helms
C. Automated
percutaneous
diskectomy.
Spine 1991 ;16:359-363
44.
Hijikata
S. Percutaneous
nucleotomy:
a new concept technique and
12 years' experience.
Clin Orthop
1989;238:9-23
45.
Hijikata
S. Nakayama
T. Yamagashi
M. Ichihara
M. Percutaneous
nucleotomy:
a new treatment
method
for lumbar
disk herniation.]
Toden Hosp 1975;5:39-44
46.
Mayer HM. Brock M. Percutaneous
endoscopic
diskectomy:
surgical
technique
and preliminary
results compared
to microsurgical diskec-
tomy. j Neurosurg
1993;78:216-225
47.
Onik G. Mooney V.Maroon jC. et al. Automated
percutaneous diskec-
tomy:
a
prospective
multi-institutional
study.
Neurosurgery
1990;26:228-233
48.
Kambin
p. O'Brien
E. Zhou L. Schaffer jL. Arthroscopic
microdiskec-
tomy and selective
fragmentectomy.
Clin Orthop
1998;347:150-167
49.
Kambin p. Zhou L.Arthroscopic
diskectomy
of the lumbar spine. (Iin
Orthop
1997;337:49-57
50.
Onik GM. Kambin P. Chang MK. Minimally
invasive disk surgery: nu-
cIeotomy versus fragmentectomy.
Spine 1997;22:827-830
51.
Choy DS. Ascher PW. Ranu HS. et al. Percutaneous
laser disk decom-
pression:
a new therapeutic
modality.
Spine 1992;17:949-956
52.
Stevenson
RC. McCabe CJ. Findlay AM. An economic
evaluation ofa
clinical trial to compare
automated
percutaneous
lumbar diskectomy
with microdiskectomy
in the treatment
of contained
lumbar disc her-
niation. Spine 1995;20:739-742
53.
Crawshaw
C. Frazer
AM. Merriam
WF. Mulholland
RC.Webb jK.A
comparison
of surgery and chemonucleolysis
in the treatment ofsci-
atica: a prospective
randomized
trial. Spine 1984;9: 195-198
54.
Maroon jC. Abla A. Microdiskectomy
versus chemonucleolysis.
Neuro-
surgery
1985;16:644-649
55.
Norton WL. Chemonucleolysis
versus
surgical
diskectomy. Compari-
son of costs and results in workers' compensation claimants.Spine
1986;11 :440-443
56. Watters WC III.Mirkovic S. Bossj. Treatment of the isolatedlumbar
intervertebral
disk herniation:
microdiskectomy
versus chemonucle-
olysis. Spine 1988;13:360-362
57.
Apostolides
Pj. Jacobowitz
R. Sonntag
VK. Lumbar
diskectomy mi-
crodiskectomy:
"the gold standard:'
Clin Neurosurg
1996;43:228-238
58. Chatterjee S. Foy PM.Findlay GF.Report of a controlled clinicaltrial
comparing
automated
percutaneous
lumbar
diskectomy
and mi-
crodiskectomy
in the treatment
of contained
lumbar disk herniation.
Spine 1995;20:734-738
59.
Kahanovitz
N, Viola K. Goldstein
T. Dawson
E. A multicenter
analysis
of percutaneous
diskectomy.
Spine 1990; 15:713-715
60.
Hermantin
FU. Peters T. Quartararo
L. Kambin P. A prospective. ran-
domized
study comparing
the results
of open diskectomy
with those
of video-assisted
arthroscopic
microdiskectomy.
j Bone joint SurgAm
1999;81 :958-965
61.
Gibson jN. Grant IC. Waddell
G. The Cochrane
review of surgery for
lumbar disk prolapse and degenerative lumbar spondylosis.Spine
1999;24: 1820-1832
62.
Hurme
M. Alaranta
H. Factors
predicting
the result
of surgery for
lumbar
intervertebral
disk herniation.
Spine 1987; 12:933-938
63.
Kotilainen
E. Valtonen S. Carlson CA. Microsurgical
treatment
oflum-
bar disk herniation:
follow-up
of 237 patients.
Acta Neurochir (Wien)
1993;120:143-149
64.
Nygaard OP. Romner B.Trumpy jH. Duration
of symptoms
as a predic-
tor of outcome after lumbar disk surgery. Acta Neurochir (Wien)
1994;128:53-56
65.
Vucetic N, de Bri E.Svensson O.Clinical history in lumbar disk herniation:
a prospective
study in 160 patients. Acta Orthop Scand 1997;68:116-120
66. Vucetic N. Svensson O. Physical signs in lumbar disk hernia. (lin
Orthop
1996;333: 192-201
67.
Fast A. Low back disorders: conservative management. ArchPhys
Med Rehabil1988;69:880-891
68.
Alaranta
H. Hurme
M. Einola S. et al. A prospective
study of patients
with sciatica:
a comparison
between
conservatively
treated patients
and patients
who have undergone
operation.
II: Results after 1-year
follow-up.
Spine 1990:15:1345-1349
69.
Atlas Sj. Deyo RA. Keller RB. et al. The Main Lumbar Spine Study. II:
1-year outcomes
of surgical and nonsurgical
management
of sciatica.
Spine 1996;21 :1777-1786
70.
Finneson
BE. A lumbar
surgery
predictive
score
card.
Spine
1978;3:186-188
71.
Herron
LD. Turner
j. Patient
selection
for lumbar
laminectomy
and
diskectomy
with
a revised
objective
rating
system.
Clin Orthop
1985;199:145-152
564
Lumbar and Lumbosacral Spine: Pathology of the Lumbosacral Spine
12.
Weber
H. Lumbar disk herniation:
a controlled.
prospective
study
with ten years of observation.
Spine 1983;8:131-140
13.
Kelsey jL. Ostfeld AM. Demographic
characteristics
of persons
with
acute
herniated
lumbar
intervertebral
disk.
j
Chronic
Dis
1975;28:37-50
14.
Miller jA. Schmatz C. Schultz AB. Lumbar disk degeneration:
correla-
tion with age, sex. and spine level in 600 autopsy
specimens.
Spine
1988:13:173-178
15.
Nachemson
A, Lewin T, Maroudas
A. Freeman
MA. In vitro diffusion of
dye through
the end-plates
and the anulus
fibrosus of human
lumbar
inter-vertebral
disks. Acta Orthop Scand 1970;41 :589-607
16.
Damkot DK. Pope MH. Lordj. Frymoyer jW. The relationship
between
work history.
work environment
and low-back
pain in men. Spine
1984:9:395-399
17.
Heliovaara
M. Vanharanta
H. Korpi j. Troup jD. Herniated
lumbar
disk
syndrome
and vertebral
canals. Spine 1986:11 :433-435
18.
Kelsey jL. An epidemiological
study
of the
relationship
between
occupations
and acute
herniated
lumbar
intervertebral
disks.
Int j
EpidemioI1975;4:197-205
19.
Kelsey jL. Golden AL. Mundt
Dj. Low back pain/prolapsed
interverte-
bral disk. Rheum Dis Clin North Am 1990;16:699-716
20.
Ramani
PS. Variations
in size of the bony lumbar
canal
in patients
with
prolapse
of
lumbar
intervertebral
disks.
Clin
Radiol
1976;27:301-307
21.
Winston
K. Rumbaugh
C. Colucci V. The vertebral
canals
in lumbar
disk disease. Spine 1984;9:414-417
22.
Dandy WE. Loose cartilage
from intervertebral
disk simulating
tumor
ofthe
spinal cord. Arch Surg 1929;19:660-672
23.
Mixter
Wj. Barr jS. Rupture
of the intervertebral
disk with involve-
ment of the spinal canal. N Engl j Med 1934:211 :210-215
24.
Caspar W. A new surgical procedure
for lumbar
disk herniation
caus-
ing less tissue
damage
through
microsurgical
approach.
Adv Neuro-
surg 1977:4
25.
Goald HJ. Microlumbar
diskectomy:
followup
of 147 patients.
Spine
1978;3: 183-185
26.
Hudgins RW. The role of microdiskectomy:
follow-up
of 477 patients.
Orthrop
Clin 1980;2:95-100
27.
Kahanovitz
N, Viola K. Muculloch
j. Limited surgical
diskectomy
and
microdiskectomy:
a clinical comparison.
Spine 1989; 14:79-81
28.
Williams
RW. Microlumbar
diskectomy:
a conservative
surgical
ap-
proach to the virgin herniated
lumbar disk. Spine 1978:3:175-182
29.
Wilson DH. Kenning j. Microsurgical
lumbar
diskectomy:
preliminary
report of 83 consecutive
cases. Neurosurgery
1979;4:137-140
30.
Yasargil MG. Microsurgical
operations
for herniated
lumbar
disk. Adv
Neurosurg
1977;4:81-82
31.
Zahrawi F. Microlumbar
diskectomy:
is it safe as an outpatient
proce-
dure? Spine 1994;19:1070-1074
32.
Wilson
DH. Harbaugh
R. Microsurgical
and standard
removal
of the
protruded
lumbar
disk:
a
comparative
study.
Neurosurgery
1981 :8:422-427
33.
Caspar W, Campbell
B, Barbier
DD. Kretschmmer
R. Gotfried
Y.The
Caspar microsurgical
diskectomy
and comparison
with
a conven-
tional standard
lumbar
disk procedure.
Neurosurgery
1991 :28:78-87
34.
Rogers LA.Experience
with limited
versus
extensive
disk removal
in
patients
undergoing
microsurgical
operations
for ruptured
lumbar
disks. Neurosurgery
1988;22:82-85
35.
Silvers HR. Microsurgical
versus standard
lumbar
diskectomy.
Neuro-
surgery 1988;22:837-841
36.
Cooper D. Feuer H. Lumbar
microdiskectomy.
j Indiana
State
Med
Assoc 1981;74:674-675
37.
Williams
RW. Microsurgical
lumbar
diskectomy.
Report
to
the
American Association
of Neurology
and Surgery.
1975. Neurosurgery
1979:4:140
38.
Benoist M. Deburge A. Heripret
G. Busson j. Rigotj, Cauchoix J. Treat-
ment of lumbar disk herniation
by chymopapain
chemonucleolysis.
A
report on 120 patients. Spine 1982;7:613-617
39.
Dabezies
Ej. Brunet M. Chemonucleolysis
vs. laminectomy.
Orthope-
dics 1978; 1:26-29
40.
Fraser RD. Chymopapain
for the treatment
of intervertebral
disk her-
niation:
a
preliminary
report
of
a
double-blind
study.
Spine
1982;7:608-612
41.
javid Mj. Nordby Ej. Ford LT.et al. Safety and efficacy of chymopapain
(Chymodiactin)
in herniated
nucleus
pulposus
with sciatica:
results
of a randomized.
double-blind
study.jAMA
1983;249:2489-2494
42.
Nordby EJ.Chymopapain
in intradiskal
therapy. j Bone joint Surg Am
1983:65:1350-1353
76
LumbarDegenerativeDiskDisease
565
72. Spengler
DM, Ouellette
EA, Battie
M, Zeh J. Elective
diskectomy
for
herniation
of a lumbar
disk:
additional
experience
with
an objective
method.
J Bone Joint
Surg Am 1990;72:230-237
73. BrisbyH,OImarker K, Rosengren L,Cederlund
CG. Rydevik B. Markers
or nerve
tissue
injury
in the cerebrospinal
fluid
in patients
with
lum-
bar disk herniation
and sciatica. Spine 1999;24:742-746
74. Boden SD.Davis DO. Dina TS, Patronas
NJ,Wiesel SW. Abnormal
mag-
netic-resonance
scans or the lumbar
spine in asymptomatic
subjects:
a prospectiveinvestigation. J BoneJoint Surg Am 1990;72:403-408
75. WeiselSW.Tsoumas N, Fefrer HL,Citrin CM,Patronas NJ.Astudy of
computer-assisted
tomography:
the incidence
or positive
CATscans
in an asymptomatic
group of patients.
Spine 1984;9:549-551
76. Jensen MC,Brant-Zawadzki
MN. Obuchowski
N. Modic MT, Malkasian
D,RossJS. Magnetic resonance
imaging of the lumbar
spine in people
withoutbackpain. NEnglj Med 1994;331:69-73
77. YamadaH.Strength of Biological Materials. Evans FG,ed. Baltimore:
Williams and Wilkins; 1970:255-282
78. BirneyTJ.White m, Berens D. Kuhn G. Comparison
of MRI and discog-
raphy in the diagnosis
or lumbar
degenerative
disk disease.
J Spinal
Disord 1992;5:417-423
79. Simeone FA.Lumbar disk disease.
In: The spine (2 voL). Philadelphia:
Williams and Wilkins; 1975
SO.
FagerCA,Freidberg SR. Analysis of failures and poor results or lumbar
spine surgery. Spine 1980;5:87-94
81. TsujiH,HiranoN.Ohshima H,Ishihara H,Terahata N.Motoe T.Struc-
tural variation or the anterior
and posterior
anulus
fibrosus in the de-
velopment or human
lumbar
intervertebral
disk: a risk factor for in-
tervertebral disk rupture.
Spine 1993;18:204-210
82. Vucetic N,Astrand P, Guntner
p. Svensson O.Diagnosis and prognosis
in lumbar disk herniation.
Clin Orthop
1999;361: 116-122
83. AbramovitzIN.NeffSR.Lumbar disk surgery: results of the Prospec-
tive Lumbar Diskectomy Study ofthe Joint Section on Disorders ofthe
Spine and Peripheral
Nerves
of the American
Association
of Neuro-
logical Surgeons and the Congress
or Neurological
Surgeons.
Neuro-
surgery1991;29:301-308
84. DvorakJ,Gauchat MH,Valach L.The outcome of surgery for lumbar
disk herniation,
I: A 4-17 years' follow-up
with emphasis
on somatic
aspects. Spine 1988;13:1418-1422
85. LewisPJ,Weir BK.Broad RW, Grace MG. Long-term
prospective
study
oflumbosacral
diskectomy. J Neurosurg
1987;67:49-53
86.
Quigley
MR. Bost J. Maroon
JC, Elrifai
A. Panahandeh
M. Outcome
af-
ter microdiskectomy:
results
of a prospective
single
institutional
study. Surg NeuroI1998;49:263-268
87. GreenoughCG.Fraser RD.The effects of compensation on recovery
fromlow-backinjury. Spine 1989;14:947-955
88. DavisRA.A long-term analysis of 984 surgically treated herniated
lumbar disks.J Neurosurg
1994;80:415-421
89.
Perez-Cruet MJ, Rice-Wyllie
L,Pieper DR. Patient education.
In: Perez-
Cruet MJ, Fessler RG.eds. Outpatient
Spinal Surgery. St. Louis: Quality
Medical Publishing;
2002: 35-47
90. Fouquet B. Goupille P, Jattiot F, et aL Diskitis after lumbar disk
surgery: features
of "aseptic"
and
"septic"
forms.
Spine
1992;17:
356-358
91. Kelly A,Griffith H,Jamjoom A.Results of day-case surgery for lumbar
disk prolapse. Br J Neurosurg
1994;8:47-49
92. Pappas CT. Harrington
T. Sonntag
VK. Outcome
analysis
in 654 surgi-
cally
treated
lumbar
disk
herniations.
Neurosurgery
1992;30:
862-866
93. Prolo DJ,Oklund SA, Butcher
M. Toward uniformity
in evaluating
re-
sults of lumbar spine operations: a paradigm applied to posterior
lumbarinterbody fusions. Spine 1986; 11:601-606
94. RogersLA.Outpatient microsurgical management of ruptured lumbar
disks. N C MedJ 1987;48:117-120
95. Schwartz AM, Brodkey JS. Bowel perforation
following
microsurgical
lumbardiskectomy: a case report. Spine 1989;14:104-106
96. Perez-Cruet
MJ, Smith
M. Foley K. Endoscopic
lumbar
microdiskec-
tomy. In: Perez-Cruet MJ.Fessler RG.eds. Outpatient Spinal Surgery.
St. Louis: Quality Medical Publishing;
2002: 171-183
97. Fountas KN,Kapsalaki EZ. Johnston KW.Smisson HF III. Vogel RL,
Robinson JS Jr. Postoperative
lumbar
microdiskectomy
pain: minimal-
izationby irrigation and cooling. Spine 1999;24:1958-1960
98. BalderstonRA,Gilyard GG.Jones AA.et aLThe treatment of lumbar
disk herniation:
simple
fragment
excision
versus
disk space curet-
tage.J Spinal Disord 1991 ;4:22-25
99. Mullen JB. Cook WA Jr. Reduction
of postoperative
lumbar
hemil-
aminectomy
pain
with
Marcaine:
technical
note.
J Neurosurg
1979;51:
126-127
100. Horwitz NH. Curtin JA.Prophylactic
antibiotics
and wound
infections
following
laminectomy
for
lumber
disk
herniation.
J Neurosurg
1975;43:727-731
101. Perez-Cruet
MJ, Fessler RG, Perin NL Review:
complications
of mini-
mally
invasive
spinal
surgery.
Neurosurgery
2002;51(5
Suppl):
S26-S36
102. Salander JM. Youkey JR, Rich NM. Olson DW, Clagett GP. Vascular
in-
jury related to lumbar disk surgery.J Trauma 1984;24:628-631
103. Shaw ED.ScarboroughJT.BealsRK.Bowelinjury as a complication of
lumbar
diskectomy:
a case report and review ofthe
literature.
J Bone
Joint Surg Am 1981 ;63:478-480
104. Anda S, Aakhus S, Skaanes KO,Sande E, Schrader
H. Anterior
perfora-
tions in lumbar
diskectomies:
a report of four cases of vascular
com-
plications
and a CT study ofthe
prevertebrallumbar
anatomy.
Spine
1991 ;16:54-60
105. Smith
EB, DeBord JR. Hanigan
We. Intestinal
injury
after
lumbar
diskectomy.
Surg Gynecol Obstet 1991 ;173:22-24
106. Bircher MD. Tasker T,Crawshaw C.Mulholland Re.Diskitisfollowing
lumbar
surgery. Spine 1988;13:98-102
107. Dall BE.Rowe DE,Odette WG,Batts DH. Postoperative discitis: diag-
nosis and management.
Clin Orthop
1987;224:138-146
108. El-Gindi
S, Aref S. Salama
M, Andrew J.Infection
of intervertebral
disks after operation.J
Bone Joint Surg Br 1976;58:114-116
109. Lindholm TS. Pylkkanen
P.Diskitis following
removal of intervertebral
disk. Spine 1982;7:618-622
110. Bookwalter JW III.Busch MD,Nicely D.Ambulatory surgery is sare
and effective in radicular
disk disease. Spine 1994; 19:526-530
111. Cares HL.Steinberg RS.Robertson ET, Caldini P.Ambulatorymicro-
surgery
for ruptured
lumbar
disks: report
of ten cases. Neurosurgery
1988;22:523-526
112. Ebeling U, Reichenberg
W, Reulen HJ. Results of microsurgical
lumbar
diskectomy:
review
on
485
patients.
Acta
Neurochir
(Wien)
1986;81:45-52
113. SachdevVP.Microsurgicallumbar diskectomy: a personal series of 300
patients with at least 1 year of follow-up. Microsurgery 1986;7:55-62
114. Williams RW. Microlumbar diskectomy: a 12-year statistical review.
Spine 1986; 11:851-852
115. Young PH. Results
of microsurgery.
In: Williams
RW, McCulloch JA.
Young PH, eds. Microsurgery
of the Lumbar Spine. Rockville: Aspen;
1990:215-222
116. Howe J.Frymoyer JW. The effects
of questionnaire
design
on the de-
termination
of
end
results
in
lumbar
spinal
surgery.
Spine
1985;10:804-805
117. McCulloch JA. Focus issue on lumbar
disk herniation:
macro-
and mi-
crodiskectomy.
Spine 1996;21(24
Suppl):45S-56S
118. Fairbank JCT, Mboat JC, Davies JB, O'Brien JP. The Oswestry
low back
pain disability
questionaire.
Physiotherapy
1980;66:271-273
119. Millikan CH.The problem of evaluating treatment of protruded lum-
bar intervertebral disk: observations of results of conservative and
surgical treatment
in 429 cases.JAMA
1954;155:1141-1143
120. Pace N, Pace P. Morici D. Clinical evaluation of late results of surgical
therapy
for protruded
lumbar intervertebral
disk. Arch Orthop
Trauma Surg 1980;96:91-94
121. Rish BL.A critique of the surgical management of lumbar diskdisease
in a private neurosurgical
practice.
Spine 1984;9:500-504
122. Williams
RW. Lumbar disk disease:
microdiskectomy.
Neurosurg Clin
N Am 1993;4:101-108
123. FindlayGF.Hall BI. Musa BS,OliveiraMD,FearSe.A10-yearrollow-up
of the outcome of lumbar microdiskectomy. Spine 1998;23:1168-1171
124. Striffeler
H, Groger
U. Reulen HJ."Standard"
microsurgical
lumbar
diskectomy
vs. "conservative"
microsurgical
diskectomy:
a prelimi-
nary study. Acta Neurochir
(Wien)
1991 ;112:62-64
125. Goffin J. Microdiskectomy for lumbar disk herniation. Clin Neurol
Neurosurg
1994;96:130-134
126. Hirabayashi S, Kumano K.Ogawa Y. Aota Y,Maehiro S.Microdiskec-
tomy and second operation
for lumbar disk herniation.
Spine
1993;18:2206-2211
127. Nystrom B.Experience
of microsurgical
compared
with conventional
technique
in lumbar
disk operations.
Acta Neurol
Scand
1987;76:
129-141
128. Thomas AM. Afshar F.The microsurgical treatment
of lumbar disk
protrusion:
follow-up of 60 cases. J Bone Joint Surg Br 1987;69:
696-698
129. Ebeling U,Kalbarcyk H. Reulen HJ. Microsurgical reoperation follow-
ing lumbar
disk surgery:
timing. surgical
findings. and outcome
in 92
patients.J
Neurosurg
1989;70:397-404
566
Lumbar and Lumbosacral Spine: Pathology of the Lumbosacral Spine
130. Haglund MM. Moore AJ. Marsh H, Uttley
D. Outcome after repeat
lumbar microdiskectomy. BrJ Neurosurg 1995;9:487-495
131. Patel N. Pople IK. Revisionallumbar
microdiskectomy:
an analysis of
operative
findings
and
clinical
outcome.
Br
J
Neurosurg
1995;9:733-737
132. Silvers HR.Lewis PJ,Asch HL. Clabeaux DE.Lumbar diskectomy for re-
current disk herniation. JSpinal Disord 1994;7:408-419
133. Fandino J. Botana C. Viladrich
A. Gomez-Bueno J. Reoperation after
lumbar disk surgery:
results in 130 cases. Acta Neurochir
(Wien)
1993;122:102-104
134. Beals RK. Hickman NW. Industrial
injuries
of the back extremities:
comprehensive
evaluation-an
aid in prognosis and management:a
study of 180 patients. JBoneJoint Surg Am 1972;54:1593-1611
135. Derebery VJ.Tullis WHo Delayed recovery in the patient with awork
compensible injury. JOccup Med 1983;25:829-835
136. Haddad GH. Analysis of 2932 workers'
compensation back injury
cases: the impact on the cost to the system. Spine 1987;12:765-769
137. Young IN. Shaffrey CI. Laws ERJr, Lovell LR.Lumbar disk surgeryina
fixed compensation
population:
a model for influence of secondary
gain on surgical outcome. Surg NeuroI1997;48:552-559
77
Tumorsof the Lumbarand LumbosacralSpine
DarylR.FourneyandZiya L.Gokaslan
. ClinicalPresentation
. RadiologicalEvaluation
. Diagnosis
. Indicationsfor Surgery
MetastaticTumors
PrimaryTumors
. PreoperativeEmbolization
. SurgicalTechnique
Posterior/PosterolateralApproach
Dissection
Awiderangeofneoplasticprocessesmayaffectthe lumbar
andlumbosacralspine.Amethodofclassificationforthese
tumorsis presented in Table 77-1. Tumors of the conus
medullarisinclude glial neoplasms, of which ependymomas
are the most common,! followed by hemangioblastoma,
metastases,and lymphoma. Myxopapillary ependymoma
arisesfrom the conus, cauda equina, or filum terminale;
other primary intradural-extramedullary
neoplasms in-
cludeschwannoma, neurofibroma, meningioma, and para-
ganglioma.
Thischaptermainlyfocusesonthe extradurallumbarand
lumbosacraltumors; these are best divided into primary
andmetastaticlesions.Primarytumorsare rareandinclude
vertebralaxis lesions and paravertebral tumors with spinal
extension(e.g., rhabdomyosarcoma). Primary tumors may
beclassified by the cell of origin or the biological behavior
(Table77-2).
Thelumbar spine is second only to the thoracic spine for
theoccurrence of metastatic disease.2Approximately 50%of
casesof metastatic epidural compressionin adults arise
frombreast,lung,or prostate cancer.3Other commonpri-
marycancers include lymphoma, renal cell carcinoma, sar-
coma,andmultiplemyeloma.Themajorityofspinalmetas-
tases(80%)involve the well-vascularized vertebral bodies
rather than the dorsal vertebral elements.4 Thus epidural
compression of the neural structures from metastatic le-
sionsismost often ventral.
Themanagementoflumbarand lumbosacraltumorsre-
quires careful patient selection, appropriate preoperative
planning, a keen appreciation
of the regional anatomy,
Anterior/Anterolateral Approach
PosteriorStabilization
SimultaneousCombinedAnterior-Posterior
ApproachandTotal EnBlocSpondylectomy
.
Complications
Postoperative
Local
Systemic
familiarity with the advantages and limitations of the dif-
ferent exposures, and a clear sense of the surgical objective.
Adjuvant radiation therapy, chemotherapy, and tumor em-
bolization are also decided on a case-by-case basis.
.
ClinicalPresentation
Lowback pain is the most common presenting symptom for
most lumbar and lumbosacral tumors.s Benigntumors typi-
cally present with an insidious onset of nonspecific back
complaints. As a consequence, there is often a significant
and possibly avoidable delay from the onset of symptoms to
the establishment of a correct diagnosis. Slow-growing lo-
cally aggressive tumors may achieve a significant size be-
fore clinical detection. Patients with malignant neoplasms
tend to have more rapidly progressive symptoms.
Three basic types of pain are encountered in patients
with spinal tumors: (1) local pain. (2) radicular pain, and (3)
axial pain. Localpain is described as constant, and generally
does not worsen with movement or improve with recum-
bency. This type of pain has been attributed to stretching of
the vertebral body periosteum by the growing tumor mass
or local tumor-induced
inflammatory process. Radicular
pain is related to nerve root compression by tumor. This
pain characteristically radiates in the dermatomal distribu-
tion of the compressed nerve root and is usually associated
with dysesthesia. The third category of pain is axial pain,
which is a mechanical pain secondary to instability of the
567
Neuological deficits are a less common presentingcom-
plaint in patients with lumbar and lumbosacral tumors.s
Obviously, lower motor neuron findings predominate inthis
region of the spine. The loose arrangement
of nerve rootsin
the lumbosacral
spinal
canal accounts
for asymmetrical
motor and sensory findings in many patients with cauda
equina compression
compared
with the more characteristi-
cally symmetric
deficits in patients with conus lesions.Sad-
dle anesthesia
and bladder or bowel dysfunction are typical
findings in patients
with either cauda equina or conus com-
pression. Ofcourse, patients with new or progressiveneu-
rological deficits such as these require emergent evaluation.
568
lumbar and lumbosacral Spine: Pathology of the lumbosacral Spine
Table 77-1
AClassification of Lumbar and Lumbosacral
Tumors
EXTRADURAL
Metastasis
Primary vertebral tumor (see Table 77-2)
Primary paravertebral tumor with spinal extension
(e.g., rhabdomyosarcoma)
INTRADURALIEXTRAMEDULARY
(TUMORSOFTHECAUDA
EQUINAORFILUMTERMINALE)
Myxopapillaryependymoma/ependymoma
Schwannoma
Metastasis
Neurofibroma
Meningioma
Paraganglioma
INTRAMEDULLARY(TUMORSOFTHECONUS)
Glialtumor: myxopapillaryependymoma, ependymoma,
astrocytoma,
glioblastoma. oligodendroglioma
Hemangioblastoma
Metastasis
lymphoma
Table 77-2
Classificationof PrimaryVertebral Axis Tumors With
Referenceto Tissueof Origin and Biological Behavior
True Bone Tumors
Associated
Bone Tumors
Osseous
Bone island/osteoma
(a)
Osteoid osteoma (a)
Osteoblastoma
(a)
Osteosarcoma
(b)
Cartilagenous
Enchondroma
(a)
Osteochondroma
(a)
Chondrosarcoma
(b)
Fibrous
Fibroma (a)
Fibrosarcoma (b)
Fibrous Histiosarcoma (a to b)
Osteoclastic
Giant cell tumor (b)
Vascular
Hemangioma
(a)
Aneurysmal
bone cyst (a)
Hemangiopericytoma
(b)
Angiolipoma
(a)
Angiosarcoma
(b)
Marrow
Plasmacytoma
(a)
Myeloma (b)
Ewing's sarcoma (b)
Adipose
lipoma
(a)
Angiolipoma
(a)
Notochord
Chordoma
(b)
a, benign tumor; b,locally aggressive or malignant tumor.
spinal column. Axial pain worsens with movement and is
characteristicaIly relieved with recumbency. Imaging stud-
ies in these patients commonly reveal vertebral body col-
lapse from pathological fractures. Spinalreconstruction and
stabilization are usually very effective in relieving this type
of pain.
.
Radiological Evaluation
Inadditionto a completehistoryand physicalexamination,
the judicious and informed use of imaging techniquesises-
sential in the assessment of vertebral neoplasia. Plainradi-
ographs are usuaIly obtained first. It must be remembered
that the lossof50to 70%ofthe trabecularbonemustoccur
before radiolucent defects are visible on plain radiographs.6
Thesize,location,and pattern ofbone destructionhelpto
differentiatebenign frommalignantprocesses.Forbenign
lesions, there is frequently a modest periosteal reactionor
sclerosis surrounding
any bony destruction.
Common plain
film findings in metastatic spinal disease include pedicle
erosion (the "winking owl" sign), paraspinal soft tissue
shadow, vertebral collapse, and pathological fracture-dislo-
cation. Although the diagnostic yield of plain radiographsis
low, especiaIly for early lesions, this imaging modalityre-
mains an important tool for assessing the alignment ofthe
spinal column. When necessary,dynamic studiesareob-
tained to assess spinal stability.
Radioisotope bone scans are very sensitive in detecting
vertebrallesionsand areveryusefulinthe identificationof
occult tumors at remote vertebral levels.However,bone
scansarealsononspecificand maybe abnormalincasesof
trauma, infection, and degenerative disease.
Computed tomography (CT)is the most sensitive testto
evaluate subtle changes in bone mineralization and pro-
vides osseous detail unmatched by plain radiographs or
magnetic resonance imaging(MRI).Myelographycanbe
combinedwitha toprovidean assessmentofnerverootor
cord compression; however, MRI provides a noninvasive
methodto evaluatethe spinalcordand nerverootsandal-
lows detailed visualization of the soft tissues in multiple
planes. MRImay be the most sensitive test for detectingle-
sions of the vertebral body.7The administration of intra-
venousgadoliniumis usefulfordelineatingthe vascularity
of spinal tumors and often provides superior visualization
of intrameduIlary tumors.s
Knowledgeofthe tumor type is essentialin planningap-
propriate treatment. Atissue diagnosis is almost invariably
requiredforpatients with primaryvertebral lesions;how-
ever, it is not necessary for most patients with metastatic
spinaldiseasewhen the primary tumor is known.Fluoro-
scopic or a-guided
biopsy of the spinal lesion by open
or percutaneous
needle techniques
can be performed.
a-guided
percutaneous fine-needle aspiration biopsy of
..
77 TumorsoftheLumbarandLumbosacralSpine
569
spinalesionsyielded a definitive cytological diagnosis in
88.5%of patients in a recent report. 9
Patientsreferred for the resection of lumbar and lum-
bosacraltumors should be carefully staged for evidence of
localand systemic spread of disease. In addition to a thor-
oughhistoryandphysicalexamination,stagingmayrequire
imagingstudies such as CTof the chest and abdomen as
wellas a bone scan. Repeated radiographic and clinical
stagingstudies may be indicated for patients with primary
lumbaror lumbosacral tumors initially considered inopera-
bleasa result of advanced local spread of disease because
thesepatients may become candidates for surgery follow-
ingneoadjuvantchemotherapy or other therapy.
. Diagnosis
Oncea definitive diagnosis has been made, and the extent
ofdiseaseboth locally and systemically has been evalu-
ated,further treatment strategy can be planned. The role
of adjuvant therapies
must also be considered
at this
stage,includingchemotherapy, radiation therapy, and pre-
operativeembolization. Forexample, current management
protocolsfor osteogenic sarcoma include neoadjuvant
chemotherapy.
10
. Indications for Surgery
Metastatic Tumors
Themanagement of patients with metastatic spine disease
shouldbe highly individualized and depends upon several
factors,including the clinical presentation,
duration of
symptoms,tumor type, anticipated radiosensitivity, tumor
location,extent of extras pinal disease, integrity of the
spinalcolumn, and the medical fitness and life expectancy
ofthepatient. The surgical management of metastatic spine
diseaseisgenerally palliative and includes (1)decompres-
sion for neurological compromise or radiculopathy sec-
ondaryto compression of the neural elements by epidural
disease,and (2) spinal stabilization for painful pathological
compressionfractures or instability.
Therelative roles of radiotherapy and surgery for patients
withmetastatic spinal tumors continue to be debated. Pre-
dictorsof a favorable response to radiotherapy include fa-
vorablehistology, earlier diagnosis, and less neurological
deficitat presentation.11 Highly radiosensitive tumors in-
cludelymphoma, neuroblastoma, seminoma, and myeloma.
Breastand prostate carcinoma are considered moderately
radiosensitive. Even for patients harboring radiosensitive
neoplasms,improvement from radiation therapy begins a
fewdaysafter the initiation of therapy. The rate of ambula-
toryrecovery in patients who are paraparetic at presenta-
tionis < 5%.12Therefore, in patients with progressive cauda
equinaor conus medullaris syndrome who are medically fit,
urgent surgical decompression
should
be considered
because the therapeutic
effects of radiation therapy are
delayed.
Afactor that deserves special emphasis is the importance
of considering the bony integrity of the spine and the spinal
alignment. Once deformity of the vertebral body has oc-
curred, radiotherapy does not prevent further collapse.13,14
Only soft tissue epidural compression responds to radiation
treatment-and
radiotherapy is not effective in cases of
spinal cord impingement involving bone or disk frag-
ments.IS In addition, radiotherapy is not a treatment
for
painful instability due to pathological compression frac-
tures. Therefore, surgery should be considered the first-line
therapy for patients with spinal deformity associated with
significant pain or neurological compromise.s Table 77-3
summarizes the indications for surgicalmanagement in pa-
tients with metastatic spine tumors.
Primary Tumors
Most primary lumbar and lumbosacral tumors are resistant
to radiation therapy and chemotherapy. Enbloc resection,
even if only marginal, therefore remains the most effective
treatment for long-term disease control and potential cure
in these patients. Primary tumors can be classified on the
basis of their biological behavior as benign or malignant;
the latter includes locally invasive lesions as well as high-
grade malignancies (Table 77-2). Benign encapsulated tu-
mors, such as schwannoma, may be treated with simple le-
sional resection. En bloc resection that includes a margin of
uninvolved tissue is necessary to effect cure for patients
with low-grade malignancies, such as giant cell tumor or
chordoma (Fig. 77-1 ).16.17
The surgical treatment goals are
similar for some high-grade
malignancies,
such as os-
teogenic sarcoma, when the lesion is localized to the spine
and amenable to en bloc resection. However, some very
high-grade malignancies may be best managed with radia-
tion therapy and chemotherapy. Simple intralesional curet-
tage or cryosurgery for low-grade tumors is generally
discouraged because of unacceptably high rates of local re-
currence.IS
The advantages of total en bloc resections for establish-
ing a cure are well recognized by surgical oncologists.
Unfortunately, because of anatomical considerations, it is
often not possible to attain this goal for spinal column
tumors. Enbloc tumor resections in the spine may involve
Table 77-3
Indications for Surgery versus RadiationTherapy
for Metastatic Tumors
Indications for
Surgery
Indications for Radiation
Therapy
1 Radioresistant tumor
2 Pathological fracture-
dislocation
3 Progressive neurological
deterioration
4 Failure of radiation therapy
4 Poor medical status/limited
life expectancy
1 Radiosensitive tumor
2 No evidence of instability or neural
compression by bone/disk
3 Stable neurological findings
5 Uncertain diagnosis
570
Lumbar and Lumbosacral Spine: Pathology of the Lumbosacral Spine
Figure 77-1
Axialcomputed tomographic (CT)and magnetic resonance imaging (MRI)scans of a patient with destructive l2 lesionwith
paraspinaland epidural extension. Sagittal MRIshows complete collapseof l2 vertebral bodywithepidural tumor. Bonescan ofthe samepatient
demonstrating
the extent of neoplastic process. CT-guided needle biopsy revealed giant cell tumor.
a complete vertebral level ("en bloc spondylectomy"), or
only a portion of a vertebral level, as long as the tumor
margins are not compromised.19The biological rationale for
en bloc spondylectomy is supported by a study examining
the local spread of metastatic
vertebral tumors, which
demonstrated that each vertebra acts as a separate com-
partment surrounded by anatomical barriers to neoplastic
cells.20Fortotal en bloc spondylectomy to be feasible, the
tumor must not invade adjacent visceral organs and must
not be adherent to the vena cava or the aorta.21Tomita et
aI21have established a new surgical classification ofverte-
bral
tumors,
modified
from
the
staging
system
of
Enneking22in which the anatomical location of tumor in-
volvement (vertebral body, pedicle, lamina and spinous
process, epidural space, paravertebral area) is used to de-
termine whether a patient will benefit from a total en bloc
spondylectomy. Reports of en bloc spondylectomy are
currently limited to small retrospective series with limited
follow-up
periods.21.23-27
.
PreoperativeEmbolization
Preoperative spinal angiography and embolization arein-
valuable in certain cases, especially with tumors knownto
be highly vascularized such as renal cell carcinoma.giant
cell tumor,
thyroid
tumors,
sarcoma,
and melanoma
(Fig. 77_2).28.29As a result of reductions in intraoperative
blood loss, preoperative arterial embolization helpsto facil-
itate the safe resection of a variety of spinal tumors.
..
77
Tumors ofthe Lumbar and Lumbosacral Spine
571
Figure77-2
Pre- and postembolization
radiographs showing tumor blush and complete obliteration of vascular supply via L2segmental vessels,
respectively,in the patient in Figure 77-1.
. SurgicalTechnique
Posterior/Posterolateral Approach
Thegeneralfamiliarity and relative technical ease of poste-
riorapproachesmakesthem appealingto most neurosur-
geons.These are the approaches of choice for the resection
ofprimaryor metastatic tumors involving the dorsal spinal
elements.Likewise, they provide the best exposure for tu-
morsof the intraspinal compartment, including lesions of
theconusmedullaris, cauda equina, and filum terminale. In
thelumbararea, in view of the greater mobility of the cauda
equina(below the level of L2),it is sometimes possible to
achieve satisfactory
intralesional
decompression
of
metastaticlesions by a posterior approach, even when the
massis located ventral to the thecal sac. Ventral access is
improvedin many cases by removal of the facets and pedi-
cIe.s.30.3!
Factorsdeterminingthe approachto spinaltumors
aresummarized in Table 77-4.
Withprone positioning, it is important that the abdomen
befreeof pressure so that bleeding is minimized. Aswith
allsurgeries, pressure points should be routinely checked
andpadded if necessary. Plain radiography or fluoroscopy
maybe used to plan the appropriate incision length.
Dissection
Amidline incision is made and the Bovie electrocautery is
usedto proceed through the subcutaneous fatty tissue. The
thoracodorsal fascia is incised, and a subperiosteal dissec-
tion is performed down the spinous processes and across
the laminae. Depending on the exposure needed, the dis-
sectionmay be carried further laterally over the facet joint
to fully expose the pedicles and transverse
processes.
Laminectomies may then be completed with the high-
speed power drill using a cutting and diamond bur attach-
ment. The ligamentum flavum is carefully removed to ex-
pose the thecal sac.
Depending upon the goals of surgery, the lateral ele-
ments, including the facets and pedicles, may be removed
to access the vertebral body. Excavation of tumor from the
vertebral body is performed with suction (in the case ofsoft
tumor) or with the drill. A ventral rim of cortical bone
should be left intact to prevent injury to the great vessels.
The dural sac is decompressed by careful displacement of
the anterolateral epidural mass into the excavated vertebral
Table77-4
FactorsDeterminingthe Approach
for SpinalTumors
1. Location of tumor (Le., anterior, posterior, both anterior
and posterior, intradural?)
2. Spinal level
3. Number of levels involved
4. Extent of involvement of adjacent structures (Le., large
paravertebral component?)
5. Primary versus metastatic lesion (Le., en bloc versus
intralesional resection?)
6. Estimated vascularity (Le., need for circumferential
devascularization
prior to resection?)
7. Bony integrity/spinal alignment (Le.. need for vertebral
body reconstruction,
anterior or posterior stabilization?)
8. Patientcondition
572
Lumbar and Lumbosacral Spine: Pathology of the Lumbosacral Spine
body cavity before removal. When this technique of pos-
terolateral decompression is applied bilaterally, effective
circumferential decompression of the dural sac may be ob-
tained.
Polymethyl methacrylate (PMMA) secured with Stein-
mann pins30.31is placed into the vertebrectomy defect. Sta-
bilization with anterior plating techniques is not possible
with posterolateral approaches, and therefore segmental
fixation is always applied to the posterior spine (see Poste-
rior Stabilizqtionsection).
Posterolateral decompression is particularly useful for
metastatic lesions at LS-S1,where a transperitoneal expo-
sure, with its accompanying potential for morbidity, is the
only other feasible approach (Fig. 77-3). As a side-note,
PMMAreconstruction of the Sl vertebral body is somewhat
problematic because the S2vertebral body has only a small
surface area on which to fix the bone cement. Even if the
polymer is successfully secured to the rostral and caudal
vertebral bodies, the reliability of the lateral fixation to the
alae may be questionable. These constructs should there-
fore always be supplemented with posterior segmental fix-
ation.32
The posterolateral approach avoids some of the pitfalls of
ventral procedures while providingadequate decompression,
even for patients with tumor circumferential to the neural el-
ements. Unfortunately, Akeyson and McCutcheon30found a
high incidence of wound infection (12%),persistent cere-
brospinal fluid (CSF)leakage (16%),and graft dislodgment
(16%)with this technique. Wound dehiscence and infection
are especially common among patients with metastatic
spinal tumors who undergo radiation therapy.s Even with
preoperative tumor embolization, intraoperative blood loss
may be profuse because intralesional decompression pro-
ceeds without ligation of the segmental feeding arteries.
In summary,
the authors
reserve single-stage
posterolat-
eral decompression
for patients
with metastatic
tumors at
LS-S1, or for patients
with
comorbid
illness or previous
surgery that renders an anterior approach nonfeasible.28
Forpatients with localized primary epidural tumors,tra-
ditional
posterior
approaches
will facilitate an en bloc re-
section only when the disease is limited to the dorsal spinal
elements.
Ventral resections
utilizing
a posterolateral ap-
proach are always intralesional
and therefore
likely to be
associated with tumor recurrence.
Total En Bloc Spondylectomy
Stener33 pioneered a method of performing a complete
spondylectomy for thoracic and lumbar tumors viaa poste-
rior approach. Tomita et aJ21.26later described a modification
of the technique called "total en bloc spondylectomy,"
which involves removal of the entire vertebra in twopieces
(anterior and posterior to the spinal cord) through a single
posterior midline incision. After transecting the pediclesto
remove the posterior neural arch, blunt dissection isusedto
free the anterior part of the vertebral body from the sur-
rounding soft tissue. Acorpectomy is then completed using
a threadwire saw while carefully protecting the spinalcord
and other soft tissues. The limitations of this approach are
greatest in the lumbar spine, where the psoas and iliac
muscle insertions on the vertebral body make a single-stage
posterior
dissection extremely
difficult. There are also
safety issues: injury to the major vessels or visceral organs
may occur during blind dissection of the anterior aspectof
Figure 77-3
Lateraland anteroposterior postoperative radiographs of a patient with metastatic renal cellcarcinoma involvingthe L1vertebral
body. The patient underwent
combined resection of the tumor with left thoracoabdominal
and posterior midline approach. Also noted are methyl
methacrylate
reconstruction
of L1, anterior plate fixation, and posterior thoracolumbar
instrumentation
with pedicle screws.
thevertebral body. In the lumbar region, the nerve roots are
important for motor function of the legs and cannot be sac-
rificedwithout resulting in neurological
morbidity.
Manip-
ulation of the neural structures
while the vertebral
body is
"rolledout"isalso a concern. Finally,a strictly posterior ap-
proachdoes not allow for augmentation
of the anterior con-
structwith a plate.28In the authors' opinion, complete en
blocspondylectomy,when indicated, is best performed us-
inga combination of anterior and posterior approaches
(see
Simultaneous Combined Anterior-Posterior
Approach
and
TotalEnBlocSpondylectomy section).
/
Anterior/Anterolateral Approach
Ventralapproaches facilitate the greatest access to the ante-
riorandmiddlecolumnsofthe spine.Inthe tumor setting,
theseapproachesare mostvaluableformetastaticlesions,
becausemetastases frequently involve the vertebral body.
Thetypeofventral exposure is largely determined by the
levelofthe lesion (Table 77-4). Formost ventral pathology
intheL2-LSregion, a retroperitoneal exposure through the
flankis utilized. Acombined thoracotomy/retroperitoneal
approach(the "thoracoabdominal approach") is used for le-
sionsinthe region of the thoracolumbar junction (T11-L1).
Ventralexposure of LS-S1 is best achieved via a midline
transperitonealroute. After a discussion of the anterior and
anterolateral approaches used in the lumbar spine, the
techniquesof lumbar vertebrectomy and reconstruction of
thevertebralbody defect are described.
RetroperitonealApproach
Thepatient is placed in the lateral decubitus position. The
sideofsurgeryis first dictated byany laterality to the tu-
mor;however,a left-sidedapproachis preferableto avoid
retractionofthe liver and the fragile vena cava. It is helpful
toflexthe upper hip to relax the psoas muscle. The table
shouldbe flexed slightly to open the space between the ribs
andthe iliaccrest.
Theincision begins in interval between the iliaccrest and
thecostalmargin. The ventral direction of the incision is de-
terminedby the level of the tumor. This may range from an
essentiallytransverse incision that follows the twelfth rib
andends above the umbilicus for the exposure of L2,to a
caudallyoblique incision that reaches the interval between
thesymphysis pubis and the umbilicus for the exposure of
L5.Theincision should extend no further ventrally than the
lateralborder of the rectus abdominis muscle.
Themuscle layers are opened in line with the skin inci-
sion.Beginningsuperficially, these are: (1) latissimus dorsi,
(2)external oblique, (3) internal oblique, and (4) transver-
sus abdominis. The retroperitoneal
space is entered by
openingthe transversalis fascia. The retroperitoneal fat is
identified.Great care is taken to leave the peritoneum intact
while it is bluntly dissected from the fascia. Occasional
carefulsharp dissection may be necessary in ventral regions
towardthe midline, where the peritoneum is typically thin
andsometimes adherent. Acommon mistake is to enter the
perirenal fat; however, the correct plane isjust in front of
the quadratus lumborum and psoas muscles. Likewise, it is
77
Tumors of the Lumbarand LumbosacralSpine
573
possible to mistake the transverse processes for vertebral
bodies and enter the retropsoas space, a blind pouch be-
tween the quadratus lumborum and psoas muscles. The
ureter
is identified and retracted
medially
with
the
retroperitoneal
fat. Gently stroking the ureter produces
characteristic peristaltic movements that aid in its identifi-
cation. The genitofemoral nerve is identified on the surface
of the psoas muscle and should be protected.
The lumbar spine is medial to and partially obscured by
the psoas muscle. The psoas muscle can be mobilized with
subperiosteal dissection to facilitateexposure of the ventro-
lateral borders of the spine. The sympathetic trunk is lo-
cated ventrally, medial to the psoas muscle, and should be
preserved if possible. The segmental vessels run over the
midportion of each vertebral body; these should be identi-
fied, ligated, and transected close to the neuroforamina
with minimal retraction or manipulation. Afterligation of
the segmental vessels, the aorta can be mobilized slightly
medially to facilitate the ventral exposure of the spine. Ra-
diographs may be obtained at this stage to confirm the ap-
propriate levels if the tumor is not immediately obvious.
Vertebrectomy and reconstruction proceed in the standard
fashion, as described in later sections. The muscle closure is
performed in multiple layers.
If exposure of L1or L2is required, the twelfth rib is re-
sected. The left diaphragmatic crus, which extends to the
second vertebral body, needs to be taken down for exposure
ofL1. Forlesions at L2,the crus may need to be taken down
as well.
Transperitoneal
Approach
Although the retroperitoneal
approach may be used for
L4-LStumors, exposure of LSand the upper sacrum is best
achieved through a direct anterior approach. Bilateral ven-
tral exposure of L2-L4 may also be achieved with this ap-
proach; however,access is limited by the aorta and the infe-
rior vena cava, and we have seldom found an indication for
such an exposure in patients with lumbar tumors.
Preoperatively, the patient should be prepared with a
bowel cathartic. Positioning is supine with a bolster to ele-
vate the sacrum. In addition, placing the patient in a mild
degree ofTrendelenburg allows the peritoneal contents to
fall away with gravity. Alongitudinal incision is made from
a point several centimeters above the pubis to just below
the umbilicus. The incision is deepened in the midline be-
tween the rectus abdominis muscles, and the preperitoneal
fat is identified. The peritoneum is picked up carefully be-
tween two pairs of forceps and, after making sure that there
is no viscera trapped beneath it, incised with a scalpel.
Upon exposure of the peritoneal cavity, the greater omen-
tum and the bowel are packed cephalad.
The sacral promontory is the landmark for LS-S1. Infil-
trating the tissue over the sacral promontory with a few
milliliters of saline facilitates dissection by elevating the
peritoneal tissue. The superior hypogastric plexus is directly
ventral to the fifth lumbar vertebra and the LS-S1 disk
space. Sympathetic fibers within this plexus are necessary
for coordinating reflexes responsible for normal ejaculatory
function in males. There is less potential for damage to the
used to makea centrallyplaced"anchoringhole"through
the end plates above and below the defect. The depth
should be approximately to the center of each vertebral
body. The chest tube is cut to the appropriate lengthtofill
the two holesand spanthe defect.Smallnotchesaremade
at both ends of the chest tube and a center holeisfashioned
with a Leksell rongeur-the
center hole is necessarytoac-
commodatethe 60 mLsyringethroughwhichthePMMAis
injected. The PMMAmust extrude through both endsofthe
chest tube to provide a cement-bone
interface withinthe
anchoring hole in each vertebral body.The spacedirectlyin
front of the thecal sac is kept free of PMMAwith two
wooden tongue blades, while a Penfield no. 2 or 3 instru-
ment is used to fashion the cement filling the vertebrec-
tomy defect to the appropriate shape. Copious irrigation
should be used while the PMMApolymerizesto prevent
thermal injury to the neural structures.
PMMAreconstruction is ideal in the cancer settingbe-
cause it accomplishes immediate stabilization afterradical
tumor resection, allowing patients to ambulate withoutthe
use of external orthoses and permitting the administration
of radiation therapy without delay.36Although PMMAisvery
stable in compression, we advocate the addition ofanterior
fixation(lockingplate and screwsinserted intotherostral
and caudal vertebral bodies) to prevent distraction failure.
574
Lumbar and Lumbosacral Spine: Pathology of the Lumbosacral Spine
superior hypogastric plexus by retracting it laterally rather
than proceeding through it; therefore, it is no longer recom-
mended to incise the dorsal peritoneum in the midline. In-
stead, this tissue is incised along the entire extent of the
right common iliac artery. Great care must be taken not to
injure the right ureter, which crosses the common iliacves-
sels near the lateral margin of the L5-S1 disk space. The
ureter is easily identified by its characteristic peristaltic
movements. It should be mobilized laterally. The plane be-
tween the prevertebral tissue and the~anterior longitudinal
ligament is identified, and this tissue, including the hy-
pogastric plexus, is mobilized laterally. The middle sacral
artery and vein may cause troublesome bleeding-these
vessels may be ligated and transected.
Vertebrectomy and reconstruction proceed in the fashion
described following here. The peritoneum, dorsally and
ventrally, is closed with a running absorbable suture. The
rectus sheath and skin are closed in multiple layers.
Vertebrectomy
After ligation of the segmental vessels and radiological con-
firmation of the pathological level, the lumbar vertebrec-
tomy is performed. Leksell rongeurs are used to begin re-
moval of soft tumor from the vertebral body; we then
switch to a high-speed power drill with a diamond bur at-
tachment for the bone resection. The diamond bur is ideal
because there is much less chance of entraining adjacent
soft tissue. The rostral and caudal end plates are drilled
away until the soft anulus is encountered. Likewise, the
posterior longitudinal ligament should be exposed directly
in front of the thecal sac. Acommon mistake during verte-
brectomy with a retroperitoneal exposure is to incompletely
removeportions of the vertebral body at the base of the
contralateral pedicle; tilting the bed slightly facilitates visu-
alization of this region. Following bone removal, diskec-
tomies are completed with a no. 15 blade scalpel, pituitary
forceps,and curettes. Finally,the posterior longitudinalliga-
ment is carefully incised with the scalpel and dissected
away from the dural sac. It is important to open the poste-
rior longitudinal ligament and visualize the nerve roots ex-
iting at the level of the resection to ensure complete de-
compressionof the thecal sac.
VertebralBody Reconstruction
Reconstruction of the spinal column after vertebrectomy is
required to obtain a stable construct. The vertebral body
may be replaced with various materials, including auto-
graft/allograft bone, PMMA,or spacers. Abone graft is often
recommended for patients with a life expectancy beyond 6
months;
however, vertebral
body reconstruction
with
PMMA is probably the best option for all patients with
metastatic tumors.s
Several methods to fixPMMAto the vertebral body have
been described. Of these, the chest tube technique34 has
been used extensively by the authors with excellent results,
even in patients who have survived beyond 3 years.3sln the
lumbar spine, a 32 to 36 French chest tube is used, depend-
ingon the sizeofthe vertebralbodies.Aright-angledrillis
Posterior Stabilization
The definition, and therefore the management, ofinstability
in the tumor setting are controversial. Denis's three-column
concept of the spine is widely accepted as the mechanical
model for thoracolumbar fractures and has applicationto
neoplastic involvement of the spinal columnY Generally
accepted criteria for spinal instability secondary to trauma
include (1) at least two-column injury, (2) > 50%collapseof
vertebralbodyheight,(3)> 20to 30 degreekyphoticangu-
lation, or (4) involvement of the same column in twoor
more adjacent levels.37.38Additionally, the quality ofthe sur-
rounding bony and ligamentous structures should be con-
sidered in patients with neoplastic disease.
Spinal instability may be caused or worsened by the
approach to, or resection of, spinal tumors. A posterioror
posterolateral decompression in the presence of neoplastic
involvement of the anterior or middle columns may result
in instability. Bydefinition, a two-column injury (andthere-
fore an unstable injury) is created when performing averte-
brectomy, regardless of the state of the dorsal columns.
Both of these examples may require posterior spinal stabi-
lization, although in selected cases, reconstruction of the
vertebrectomy defect alone may provide satisfactory sup-
port. In a recent biomechanical analysis of various stagesof
tumor lesions using human cadaver specimens, anterior re-
construction alone provided stiffness equivalent to circum-
ferential reconstruction, as long as only vertebrectomy or
subtotal
spondylectomy
had
been
performed.
Total
spondylectomy, on the other hand, significantly reducedthe
anterior reconstruction stiffness, suggesting the need for
combined anterior and posterior procedures.39 Posterior
stabilization is indicated for most patients who undergo
two-level vertebrectomies
with anterior
reconstruction
77
Tumorsof the Lumbarand LumbosacralSpine
575
Figure77-4
Sagittal magnetic resonance imaging (MRI)of a patient with metastatic
renal cell carcinoma with 51 destruction
and partial collapse.
Thepatient presented
with intractable
low back pain and bilateral
sciatica.
because
lengthyanterior constructs(with orwithout ante-
riorplating)mayproveunstable.Inour opinion,posterior
stabilizationin addition to anterior reconstruction is also
indicatedafter the resection of thoracolumbar junction tu-
mors-this region represents a "transition zone" between
therelatively rigid thoracic levels and the more mobile
lumbarsegments, resulting in unique biomechanical stress
(Fig.77-4).Finally,posteriorstabilizationisnecessaryafter
resectionof most tumors caudal to L4, where anterior
fixationisdifficult to apply. Indications for posterior stabi-
lizationafter the resection of lumbar and lumbosacral
tumorsand vertebral body reconstruction are summarized
inTable77-5.
Posteriorstabilization of the lumbar region is best per-
formedwith pedicle screws. Perhaps the greatest advantage
ofpediclescrew fixation for patients with spinal tumors is
theabilityto instrument at levels where wide laminectomies
or even partial pedicle resections have been performed.
Becausepedicle screws do not rely on the laminae or facets
Table77-5
Indicationsfor PosteriorStabilizationafter
Resection of Lumbar and Lumbosacral Tumors
and Anterior Reconstruction
1. Extensiveposterior decompression
(includes all patients who
undergo single-stage posterolateral decompression)
2. Vertebrectomy of two or more levels
3. Thoracolumbar junction "transition zone"
4. Tumorscaudal to L4(anterior stabilization difficult)
for attachment, this form offixation obviates the need forthe
placement of hooks or sublaminar wires within the spinal
canal. Biomechanical studies have found pedicle screw con-
structs to be more rigid than traditional forms of internal fix-
ation, thus improving the potential for successful fusion.40
Segmental distracting, compressing, lordosing, rotating, and
antero- or retrolisthesing forces can be applied, depending
on the clinical situation. Finally, fixation can be limited to
only adjacent spinal segments, allowing motion segments
above and below the levelof spinal disease to be spared.
In the authors' recent review of 100 consecutive cases in-
volving pedicle screw fixation for malignant spinal disease,
the rate of screw-related late instrumentation
failure was
only 2%.5The low rate of pedicle screw failure, even in long-
term survivors, was attributed to an emphasis on the con-
current
use of anterior
approaches
to reconstruct the
weight-bearing capacity of the spine.
Instrumented fusion across the lumbosacral junction de-
serves special mention. The Sl pedicle is larger than the
lumbar pedicles and may often be fitted with 7 to 8 mm
screws. Additional bony purchase can be obtained if the
screw penetrates the anterior Sl cortex or the superior end
plate of S1. Medially directed Sl pedicle screws that are
cross-linked and attached to rods create a triangulation ef-
fect, which greatly increases torsional stability and resists
pullout.41-44Triangulation with an oblique orientation also
interferes less with the superjacent facet joint and allows
greater bony purchase with a longer screw.45
A method to enhance sacral fixation with screws is
to place an additional
pair of laterally directed
bone
screws into the sacral alae below the Sl level. This has a
576
Lumbar and Lumbosacral Spine: Pathology of the Lumbosacral Spine
biomechanical advantage over a single pair of S1 pedicle
screws;43however,the bone of the ala is usually of low den-
sity, and purchase may be tenuous.46 Moreover, the risk of
neurovascular injury from laterally directed screws in this
regionmust be kept in mind.
S2(or lower level) pedicle screws are often of little use
because the pedicles are very short. Biomechanical testing
has shown that pedicle screws placed below the S1level do
not significantly enhance stability.43 In addition, the thin
sagittal dimension of the sacrum at lower levels increases
the risk of penetration of its ventral surface, with a potential
for injury to the adjacent vascular and visceral structures.
Screws at lower sacral levels are often prominent dorsally
and may even tent the overlying skin.
Sacralscrew fixation may be sufficient for cases in which
the fixation length is short (Le.,one or two levels) and there
is minimal instability. If a longer construct is placed, the
sacral attachment is subjected to large cantilevered forces
that may lead to screw pullout. Finally, the use of sacral
screws may be precluded in certain cases, such as when the
pedicles, body,or ala of the sacrum are involvedwith tumor.
A simple method of sacropelvic fixation involves the
placement of long, variable-angle bone screws obliquely
across the sacroiliacjoint into the iliacbones. Atripod effect
may be gained by combining the sacroiliac fixation with ad-
ditional sacral fixation points.47However, the authors' pre-
ferred method of fixation to the pelvis is the Galveston tech-
nique (Fig. 77-5).48 It involves the insertion of an angled
distal limb of a spinal fixation rod into the posterior iliac
bones, just above the sciatic notch. Custom bending and in-
sertion of the rod require some technical skill to achieve the
correct position with the rod remaining
intracortical.
Preformed rods are also available. Iliac screws (ISOLAiliac
screws;
Acromed, Cleveland,
OH) can be placed indepen-
dently of the spinal rod, with the two subsequently linked
together. In the biomechanical testing of 10 differentlum-
bosacral instrumentation
techniques
in a bovine model,Mc-
Cord et al43found that the most effective construct entailed
medially directed S1 pedicle screws and an iliac purchasein
the Galveston-type
fashion. The key structural
element of
this construct is related to the resistance to flexion conferred
by the anterior purchase obtained with iliac rods wedged
between the hard cortical bone above the sciatic notch.
In the senior authors' series of 13 patients who underwent
Galveston fixation following
the resection
of metastatic or
locally aggressive
neoplasms
of the lumbosacral
region,J2
solid bone fusion was achieved
in four (31 %)and partial or
unilateral fusion occurred in three (23%).Fivepatients(38%)
had no convincing evidence
of fusion on radiographic stud-
ies. Because three of these five patients improved clinically,a
satisfactory
fibrous
union
may have occurred.
Radiation
therapy,
chemotherapy,
and neoplastic
disease
processes
may have contributed
to the lack of bony fusion in some pa-
tients. Only one hardware-related
complication
occurred:
the rods fractured bilaterally at the transition
point between
the lumbar and sacral segments
in one patient. This was cor-
rected with double iliolumbar rod fixation.
Simultaneous
Combined Anterior-Posterior
Approach and Total En Bloc Spondylectomy
Primary and metastatic lumbar tumors that invade boththe
anterior and the posterior columns are a challenging surgical
problem if one contemplates a complete resection. Tradi-
tionally, a posterior approach followed by a separately
staged anterior approach has been used in the surgical
Figure 77-5
Postoperative anteroposterior
and lateral radiographs of the patient in Figure 77-4, who underwent transpedicular
S1 vertebrectomy
and posterior lumbar-iliac instrumentation
using Galveston L-rodfixation. The patient had complete resolution of his back pain and sciatica.
managementof these complex lesions. However, there are
manydisadvantages to traditional staged procedures. When
tumorinvolves both the anterior and posterior columns, it
mustby definition involveat least one pedicle. Therefore, a
two-stageprocedure will alwaysresult in violation ofthe tu-
mormass.Significantblood loss may occur during the pos-
teriorstage, when ventral access to the blood supply of the
tumorand major vessels is limited. Recently, the authors
havedescribedindetaila simultaneouscombinedanterior-
posteriorapproach that can be used for lumbar and thoracic
tumor~.28Although the approach is very useful for patients
withmetastatic tumors, it is perhaps most indicated when a
totalen blocspondylectomyjs contemplated. The patient is
placedin the lateral decubitus position for the procedure.
Thetechnique essentially involves a retroperitoneal expo-
surecombinedwith a dorsal midline approach.
Themain advantages of the simultaneous combined ap-
proachare: (1) direct visualization of the entire extent of
thetumoris possible,which enables radical resection while
safelyavoidingmajor vessels; (2) large paraspinal masses
maybeencompassed en blocwith the specimen; (3) hemo-
stasisisoptimized because the dissection proceeds around
oroutsideofthe tumor margin; (4) the patient is in the lat-
eralrather than a prone position, and thus venous return is
notimpeded(another hemostatic advantage); and (5) com-
pletereconstruction and stabilization of the spine both an-
teriorlyand posteriorly can be performed during the same
procedure.The major disadvantage of the simultaneous
combinedapproach is that the lateral positioning does not
readilypermitfluoroscopicguidanceduring placementof
theposteriorfixation. However, direct visualization of sev-
eralanatomical landmarks by this wide exposure does aid
correctplacement of the instrumentation.
Althoughthe simultaneous combined approach works
wellintheupperlumbarspine,thereareanatomicallimita-
tionstothis type of exposurefor the lower lumbarverte-
brae.especially when a total en bloc spondylectomy is the
goal:(1)the nerve roots in this region are important for mo-
torfunctionof the legs and cannot be sacrificed without re-
sultingin neurological morbidity; (2) the iliopsoas muscles
areattached to a significant portion of the lateral aspect of
thevertebralbodies bilaterally and must be dissected offor
includedwith the specimen if they are infiltrated by tumor;
(3)thevertebral bodies are larger than elsewhere; (4) the iI-
iacvesselsare in close proximity and may be adherent to the
spine;and (5) the ureters, located just laterally, are at risk
forinjury in the lower lumbar levels. As a result of these
considerations,we prefer to perform the total en bloc resec-
tionoflower lumbar tumors via separately staged anterior
andposteriorapproaches.25Bydefinition, this is only possi-
blewhenthe tumor does not involveeither pedicle.
. Complications
Postoperative
Postoperativecomplications, at least for patients with
metastatictumors (for which large series are available).
havebeen extensively studied.5.35,49-53Some of the factors
citedto significantly influence the risk of complications
77
Tumorsof the LumbarandLumbosacralSpine
577
include: (1) emergent surgery51; (2) previous radiother-
apy5.50-52;(3) advanced age50;(4) significant neurological
deficit50;(5) nutritional depletion53;(6) high intraoperative
blood IOSS53;and (7) the presence of at least one "serious"
medical condition (diabetes, heart disease, congestive ob-
structive pulmonary disease. liver failure, or rheumatoid
arthritis).5
Local
Intraoperative blood loss may be reduced by preoperative
embolization, especially for characteristically hemorrhagic
tumors such as renal cell carcinoma. One should try to
maintain meticulous hemostasis throughout the surgery.
Although we do not frequently employ the cell-saverdevice
during complex spine tumor surgery. we do not believethat
the presence of a neoplastic process necessarily contraindi-
cates its use.
Neurological complications during surgery may be re-
duced with careful surgical technique and the routine use of
somatosensory-evoked potential monitoring.
Wound dehiscence and infection are among the most fre-
quently cited complications in metastatic tumor series.
with rates reported as high as 25 to 34%.50The risk is high-
est among patients with dorsal wounds. compared with
those who undergo ventral approaches.5.3o Several au-
thors5.50-52have found an association between radiation
therapy (especially preoperative radiotherapy51)and subse-
quent wound breakdown and infection. High intraoperative
blood IOSS,53nutritional depletion,53 corticosteroid use,53
advanced age,50and the presence of paraparesis50 have also
been associated with wound complications. Our findings
and those of others emphasize the importance of avoiding
preoperative radiation therapy in patients who are candi-
dates for surgery as first-line therapy. Unfortunately, sur-
geons are often asked to evaluate patients after they had al-
ready received radiation treatment.
At our institution, we have introduced several measures
in an attempt to avoid wound complications. At the end of
the surgical procedure, extensive wound irrigation is per-
formed prior to obtaining a hemostatic closure. We also ob-
tain better soft tissue coverage by mobilizing the paraspinal
muscles with simple relaxing incisions. Wound dehiscence
is avoided by meticulous closure of the fascia. Finally,we
place drains in all wounds to help prevent hematoma or
seroma formation.
With regard to instrumentation
failure. we had only one
case of pedicle screw loosening and one case of pedicle
screw fracture in our series of 100 consecutive surgeries in-
volving pedicle screw fixation for malignant spinal tumors.5
This low rate of failure, even among long-term survivors. is
likely related to our emphasis on reconstruction of the ante-
rior spinal column to restore the weight-bearing capacity of
the spine. In this setting, the pedicle screws are used as a
supplemental method of fixation rather than as the primary
mechanical support devices. Although we always attempt to
limit the extent of posterior instrumentation, we often ex-
tend the rostral and caudal ends of the posterior construct
several levels above and below the lesion. especially when
the tumor involves the thoracolumbar junction. Likewise.
long lever-arm constructs (Le.,four or more levels) involv-
578
Lumbarand Lumbosacral Spine: Pathology of the Lumbosacral Spine
ing the lumbosacral junction are secured with Galveston
rod fixation.This is an attempt to increase the load sharing
of the screws.
Systemic
Closepostoperative monitoring and constant vigilance are
required for the early detection of cardiac or respiratory
problems. including congestive heart failure, pulmonary
edema, and pneumonia. Older patients with comborbid
medical illness are at high risk for these complications.5
The frequency of postoperative confusion in patients un-
dergoing surgery for metastatic spinal tumors is reported as
high as 10%,51although in most cases these episodes resolve
within 72 hours.
Thromboembolic complications may be reduced with the
routine use of external sequential pneumatic compression
stockings. Lowmolecular weight heparin is recommended
for paraplegic patients, although the initiation of anticoagu-
References
1.
Fourney DR. Fuller GN. Gokaslan ZL Intraspinal
extradural
myxopap-
iIIary ependymoma
of the sacrum
arising
from the filum terminale
externa:
case report. J Neurosurg
Spine 2000;93:322-326
2.
Brihaye J.Ectors
P, Lemort
M. Van Houtte
P. The management
of
spinal
epidural
metastases.
Adv
Tech
Stand
Neurosurg
1988;
16:121-176
3.
Byrne TN. Spinal cord compression
from epidural
metastases.
N Engl J
Med 1992;327:614-619
4.
Perrin RG. McBroom RJ.Anterior
versus
posterior
decompression
for
symptomatic
spinal metastases.
CanJ Neurol Sci 1987;14:75-80
5.
Fourney DR. Abi-Said D. Lang FF.McCutcheon
IE, Gokaslan
ZL. Use of
pedicle
screw
fixation
in the management
of malignant
spinal
dis-
ease:
experience
in 100 consecutive
procedures.
J Neurosurg
Spine
2001 ;94:25-37
6.
Algra PRoHeimans
JJ. Valk J, Nauta 11.Lachniet
M. Van Kooten B. Do
metastases
in vertebrae
begin in the body or in the pedicles?
Imaging
study in 45 patients.
AJRAmJ RoentgenoI1992;158:1275-1279
7.
Traill Z, Richards
MA. Moore
NR. Magnetic
resonance
imaging
of
metastatic
bone disease. Clin Orthop
1995;312:76-88
8.
Sze G. Magnetic
resonance
imaging
in the evaluation
of spinal
tu-
mors.
Cancer
1991 ;67:1229-1241
9.
Kang M. Gupta
S. Khandelwal
N. Shankar
S. Gulati
M. Suri S. CT-
guided
fine-needle
aspiration
biopsy
of spinal
lesions.
Acta Radiol
1999;40:474-478
10.
Sundaresan
N. Rosen G. Huvos A. Krol G. Combined
treatment
of os-
teosarcoma
of the spine. Neurosurgery
1988;714-719
11.
Maranzano
E. Latini P. Effectiveness
of radiation
therapy
without
surgery in metastatic
spinal
cord compression:
final results
from a
prospective
trial.lntJ
Radiat Oncol Bioi Phys 1995;32:959-967
12.
Gilbert RW. KimJH. Posner JB. Epidural spinal cord compression
from
metastatic
tumor: diagnosis
and treatment.
Ann NeuroI1978;3:40-51
13.
Asdourian
PL. Mardjetko
S. Rauschning
W. Jonsson
H. Hammerberg
KW. Dewald RL. An evaluation
of spinal
deformity
in metastatic
breast cancer. J Spinal Disord 1990;3: 119-134
14.
Rao S. Badani K.Schildhauer T, Borges M. Metastatic
malignancy
of the
cervical spine: a nonoperative
history. Spine 1992;17:S407-$412
15.
Tomita
T. Galicich JH. Sundaresan
N. Radiation
therapy
for spinal
epidural
metastases
with
complete
block.
Acta
Radiol
Oncol
1983;22: 135-143
16.
Turcotte RE.Sim FH. Unni KR.Giant cell tumor
of the sacrum. Clin Or-
thop 1993;291 :215-221
17.
York JE. Kaczaraj A, Abi-Said D. et al. Sacral chordoma:
40-year
expe-
rience at a major cancer center. Neurosurgery
1999;44:74-80
18.
Camins
M. OppenheimJ.
Perrin R.Tumors
of the vertebral
axis: be-
nign. primary
malignant.
and metastatic
tumors.
In: Youmans J.ed.
Neurological
Surgery.
Vol 2. 4th ed.
Philadelphia:
WB Saunders;
1996:3134-3167
lant prophylaxis should probably be delayed forat least48
to 72 hours after surgery.54
In an effort to control pain and maintain neurological
function. most clinicians use corticosteroid therapyasa
supplement
to surgery or radiation therapy for patients
with metastatic spinal tumors. High-dose corticosteroid
therapy carries a significant risk of toxicity.Theincidenceof
serious toxic side effects was 14%in one study55andin-
cluded one fatal ulcer hemorrhage. rectal bleeding.andgas-
trointestinal perforation. These risks must be weighed
against the anticipated benefits. In addition to the dose.the
duration of steroid therapy also determines its toxicity.A
significant increase in severe complications. includingfatal
sepsis, has been noted when dexamethasone was usedfor
more than 40 days to treat malignant epidural spinalcord
compression.56 Because major complications may occur
with prolongedtherapy,the use of steroidsshouldbedis-
continued as rapidly as possible or tapered down to the
lowest possible level compatible with clinical benefit.
19.
Hasegawa
K, Ogose A. Kobayashi
H. Morita T. Hirata Y.Simultaneous
anterior-posterior
approach
for excision
of malignant
paraspinal tu-
mor
and
subsequent
reconstruction:
technical
note. J Neurosurg
Spine 1999;91 :236-240
20.
Fujita
T, Ueda
Y. Kawahara
N. Baba H, Tomita
K. Local spread of
metastatic
vertebral
tumors:
a histologic
study.
Spine
1997;22:
1905-1912
21.
Tomita K. Kawahara
N. Baba H. Tsuchiya H. Fujita T. Toribatake Y.Total
en bloc spondylectomy:
a new surgical
technique
for primary malig-
nant vertebral
tumors.
Spine 1997;22:324-333
22.
Enneking
WF. Spanier
SS. Goodman
MA. A system
for the surgical
staging
of musculoskeletal
sarcoma.
Clin Orthop
1980;153: 106-120
23. Abe E. Kobayashi T, Murai H. Suzuki T. Chiba M. Okuyama K.Total
spondylectomy
for primary
malignant,
aggressive
benign, and soli-
tary metastatic
bone tumors
ofthe
thoracolumbar
spine. J Spinal Ois-
ord 2001 ;14:237-246
24.
Bauernhofer
T. Stoger H. Kasparek
AK. et al. Combined
treatment of
metastatic
osteosarcoma
of the spine. Oncology 1999;57:265-268
25.
Marmor
E. Rhines LD.Weinberg
JS. Gokaslan
ZL Total en bloc lumbar
spondylectomy:
case report. J Neurosurg
Spine 2001 ;95:264-269
26.
Tomita K. Kawahara
N, Baba H, Tsuchiya H. Nagata S. Toribatake Y.To-
tal en bloc spondylectomy
for solitary
spinal
metastases.
Int Orthop
1994;18:291-298
27.
van Dijk M. Cuesta MA. Wuisman
PI. Thoracoscopically
assisted total
en bloc
spondylectomy:
two
case
reports.
Surg Endosc
2000;14:
849-852
28.
Fourney DR.Abi-Said D. Rhines LD. et al. Simultaneous
anterior-pos-
terior approach
to the thoracic
and lumbar
spine for the radical resec-
tion of tumors
followed
by reconstruction
and stabilization.
J Neuro-
surg Spine 2001 ;94:232-244
29.
Jackson
R. Gokaslan
Z. Loh S. Metastatic
renal cell carcinoma
of the
spine:
surgical
treatment
and
results.
J Neurosurg
Spine 2001 ;94:
18-24
30.
Akeyson E. McCutcheon
I. Single-stage
posterior
vertrebrectomy
and
replacement
combined
with
posterior
instrumentation
for spinal
metastasis.
J Neurosurg
1996;85:211-220
31.
Bilsky MH. Boland P. Lis E. Raizer 11.Healey JH. Single-stage
postero-
lateral
trans pedicle
approach
for spondylectomy.
epidural
decom-
pression,
and
circumferential
fusion
of spinal
metastases.
Spine
2000;25:2240-2249
32.
Jackson
RJ.Gokaslan
ZL. Spinal-pelvic
fixation
in patients
with lum-
bosacral neoplasms.J
Neurosurg
Spine 2000;92:61-70
33.
Stener B. Complete
removal
of vertebrae
for extirpation
oftumors.
A
20-year experience.
Clin Orthop
1989;245:72-82
34.
Errico TJ. Cooper PRoA new method
of thoracic
and lumbar
body re-
placement
for
spinal
tumors:
technical
note.
Neurosurgery
1993;32:678-680
77
Tumorsof the Lumbarand LumbosacralSpine
579
35.Gokaslan ZL.YorkJE. Walsh G. et al. Transthoracic
vertebrectomy
for
metastatic spinal tumors.J
Neurosurg
1998;89:599-609
36. YorkJE.Gokaslan ZL.Instrumentation
of the spine in metastatic
dis-
ease. In: Errico TJ. ed. Spine: State of the Art Reviews. Vol 13. no. 2.
Philadelphia:
Hanley
& Belfus;
1999:335-350
37. Denis F.The three column spine and its significance
in the classifica-
tion ofacute thoracooumbar
spinal injuries. Spine 1983;8:817-831
38. Kern MB. Malone
DG. Benzel
EC. Evaulation
and surgical
manage-
ment of thoracic
and
lumbar
instability.
Contemp
Neurosurg
1996;8:1-8
39. Kanayama M. NgJT.Cunningham BW.Abumi K. Kaneda K.McAfeePc.
Biomechanical analysis of anterior
versus
circumferential
spinal
re-
'onstruction
for various
anatomic
stages
of tumor
lesions.
Spine
1999;24:445-450
40. VaccaroAR.Garfin SR.lnternal
fixation (pedicle
screw fixation) for fu-
sionsofthe lumbar spine. Spine 1995;20:157S-165S
41. Carson W. Duffield R. Arendt M. Ridgely B. Gaines JR.Internal forces
andmoments in traspedicular spine instrumentation:
the effect of
pediclescrew angle and transfixation-the
4R-4bar linage concept.
Spine1990;15:893-901
42. CarlsonGD.Abitbol JJ.Anderson DR.et al. Screw fixation in the hu-
man sacrum: an in vitro study of the biomechanics
of fixation.
Spine
1992;17(6Suppl):S196-S203
43. McCordDH.Cunningham BW. Shono Y. MyersJJ.McAfeePc. Biome-
chanical analysis
of
lumbosacral
fixation.
Spine
1992;17(8
Suppl):S23S-S243
44. SmithS.AbitbolJ.Carlson G. Anderson D. Taggart K.Garfin S.The ef-
fectsof depth of penetration.
screw orientation.
and bone density
on
sacral screw
fixation.
Spine
1993;18:1006-1010
45. Krag M. Beynnon
B. Pope
M. DeCoster T. Depth of insertion of
transpedicular
vertebral
screws
into human
vertebrae:
effect upon
screw-vertebra interface
strength.J
Spinal Disord 1988;1 :287-294
46.
Ogilvie J. Transfedt E. Wood K.Overview of fixation
to the sacrum
and
pelvis in spinal surgery. In: Margulies J.Floman
Y. FarcyJ-P. Neuwirth
M. eds. Lumbosacral
and Spinopelvic
Fixation.
Philadelphia:
lippin-
cott-Raven;
1996:191-198
47.
Baldwin N. Benzel E.Sacral fixation using iliac instrumentation
and a
variable angle screw device.J Neurosurg
1994;81 :313-316
48.
Allen BJ.Ferguson
R. The Galveston technique
for L rod instrumenta-
tion of the scoliotic spine. Spine 1982;7:276-284
49.
McAfee PC. Zdeblick TA. Tumors of the thoracic
and lumbar
spine:
surgical
treatment
via
the
anterior
approach.
J Spinal
Disord
1989;2:145-154
50.
Sundaresan
N. Steinberger
M. Moore F. et al. Indications
and results
of combined
anterior-posterior
approaches
for spine tumor surgery. J
Neurosurg 1996;85:438-446
51.
Pascal-Moussellard H.BrocG. Pointillart V. Simeon F.VitalJM. Sene-
gas J. Complications of vertebral metastasis surgery. Eur Spine J
1998;7:438-444
52.
Wise JJ. Fischgrund JS.Herkowitz HN.Montgomery D. Kurz LT.Com-
plication.
survival rates. and risk factors of surgery for metastatic
dis-
ease of the spine. Spine 1999;24:1943-1951
53.
McPhee lB. Williams RP. Swanson CEoFactors influencing wound
healing
after
surgery
for metastatic
disease
of the
spine.
Spine
1998;23:726-732
54.
Geerts WHoHeit JA.Clagett GP.et al. Prevention of venous throm-
boembolism.
Chest 2001 ;119: 132S-175S
55.
Heimdal K. Hirschberg H. Slettebo H. Watne K.Nome O.High inci-
dence
of serious
side effects of high-dose
dexamethasone
treatment
in patients
with
epidural
spinal
cord
compression.
J Neurooncol
1992;12:141-144
56.
Martenson
JA Jr. Evans RG. Lie MR. et al. Treatment
outcome
and
complications
in patients
treated
for malignant
epidural
spinal cord
compression.J
NeurooncoI1985;3:77-84
78
Trauma of the Lumbar Spine
Patricia B. Raksin
. Anatomy
.
Classification
.
Mechanismsof Injury
.
Diagnosis
.
Treatment
.
Conclusion
Lumbar and lumbosacral spine injuries reflect the transi-
tional nature of the bony anatomy in the region. and in turn.
the differing biomechanical forces to which individual
levels may be subjected. Denis's three-column model pro-
vides the basis for modern considerations of stability in the
face of injury. Diagnosis of lumbar and lumbosacral spine
injuries requires a complete neurological examination.
coupledwith multiple imaging modalities. to determine the
nature of the injury and degree of instability. The appro-
priate treatment
regimen takes into account traumatic
and medical comorbidities with the dual goals of decom-
pressionofneural elements and stabilization.
. Anatomy
The particular anatomy of the lumbar spine predicts struc-
tural integrityinthe faceofapplied. supraphysiologicalforces.
Vertebral bodies are composed of a cancellous. trabeculated
core. surrounded by cortical bone. Diminution ofthese hori-
zontal cancellous trabeculations predisposes the vertebral
bodies to an increasedriskoffailureunder axial loadwith ag-
ing.1Intervertebral disks are composed ofan outer anulus fi-
brosis and inner nucleus pulposus. an acellular meshwork of
collagen fibers in a glycosaminoglycanmatrix. The superior
and inferior cartilaginous end plates help distribute vertebral
forces to the disk surface.The posterior elements. the pedi-
cles. facets. laminae, and spinous processes. permit move-
ment and provide strength.Theseelements contribute to flex-
ion stability and resist rotation. Facetjoints alone have been
shown to bear up to 18%ofa compressiveload.2
The ligaments that bind adjacent vertebral bodies do not
resist movements within a physiological range; rather, they
580
serve to limit supraphysiological
motions. The anterior lon-
gitudinalligament
limits extension
whereas
the posterior
longitudinal
ligament
limits
flexion.
The
ligamentum
flavum that inserts on the lamina at each level assists inre-
sistance
to flexion. The capsular
ligaments.
oriented per-
pendicular
to the facet joint plane, both assist flexion load
bearing and resist axial rotation. In general. ligaments resist
sudden loading better than bone; significant
force is neces-
sary to rupture
ligamentous
structures.
However, ligaments
do tend to increase in laxity and hysteresis with age.3
Consideration of lumbar spine trauma is necessarilycom-
plicated by the anatomical architectural transition that occurs
between the thoracolumbar and lumbosacral junctions.The
thoracic spine is fairly rigid due to the presence of ribs.The
thoracolumbar junction, therefore. is subjected to greatstress
because it links the stiff thoracic spine with the relatively
more mobile lumbar spine. Segmental mobility of the lumbar
spine increases in a caudal direction. The L4and L5segments
are not only the most mobile but also bear the greatest bur-
den under compressive loads. It is for this reason that degen-
erative changes occur most frequently at these levels.Simi
larly,more flexionand extension occur at the lower segment
than at those closer to the thoracic spine. The anterior longi
tudinalligament
and anterior anulus resist hyperextension
whereas the posterior longitudinal ligament and posterioran-
ulus limit hyperflexion.The intervertebral disk and face
joints each contribute -45%of resistance to axial rotation;in-
terspinous and supraspinous ligaments contribute the re-
maining 10%.4The articular processes and capsular ligament
provide support for lateral bending. Aroughly equal amoun
of lateral bending occurs at all levels of the lumbar spineex
cept at L5-Sl, where bending is more restricted.s
In addition. although the sagittal diameter of the cana
remains relatively constant from T12 to L5,the transverse
diameterincreasesin a caudal direction: from 20 mm at
T12to22-23 mm at L3,to 26 mm at L5.There is effectively
a25%increasein transverse area (350-450 mm2) from up-
pertolowerlumbar spineP Asthe canal size increases, the
sizeof the neural elements occupying it decreases. The
samespinal cord that fills the canal at T12 tapers to the
conusbyL2and to the cauda equina more distally. It follows
thatthesamefracturethat causesparaplegiaat T12mayre-
sultinno demonstrable deficit if sustained at L5.
Thethoracolumbar junction also constitutes a region of
transItionfromrelative kyphosisto lordosis.The transition to
lordosisbegins at Ll-L2 and increases in a caudal direction.
Normalordosis is -20 degrees at L4-L5and 28 degrees at
the lumbosacraljunction.s Overall, the kyphotic thoracic
spineandsacrum are relatively immobile as compared with
thelordoticlumbar segment. This relative mobility has im-
portantimplications for susceptibility to certain fracture
typesand, in turn, influences the construct considerations
necessaryto providestability in the faceof injury.9-n
. Classification
Existingclassification systems employ a mechanistic ap-
proachto lumbar fractures. Holdsworth's model for spinal
stabilitygrouped structures into two columns (Fig. 78-1 ).12
Theanterior column was composed of the vertebral body,
intervertebraldisk, and associated longitudinal ligaments.
The posterior column
included
the
posterior
neural
elementsas well as the interspinous ligamentous struc-
tures.Theposterior column functioned as a tension band,
balancingthe load carried by the anterior column. Disrup-
2-Column Model
3-Column Model
Posterior Column
Posterior Column
Middle
Column
Anterior
Column
Anterior
Column
Figure78-1
Two- and three-column
models
proposed
for evalua-
tion of spinal stability in traumatic
injuries.
78
Trauma of the Lumbar Spine
581
Table78-1
Whiteand PanjabiGradingof ClinicalInstability
for the Lumbar Spine
Element
Point Value
Caudaequinadamage
3
Relativeflexion sagittal planetranslation> 8%
2
or extension sagittal planetranslation> 9%
2
Relativeflexion sagittal planerotation < -9 degrees
2
Anterior elementsdestroyed
2
Posteriorelementsdestroyed
2
Dangerousloading anticipated
1
tion of more than one column constituted instability.
Denis later built upon Holdsworth's system, devising a
three-column model for the assessment of spinal stability.13
Denis's posterior column, like Holdsworth's, consists of the
posterior bony elements and ligamentous complex.Denisdi-
vides Holdsworth's anterior column into middle and anterior
segments. The middle column is composed of the posterior
longitudinal ligament, the posterior anulus, and the posterior
one third of the vertebral body.The anterior column consists
of the anterior longitudinal ligament, the anterior anulus, and
the anterior two thirds of the vertebral body. The columns
form the basis for his assessment of spinal stability in the face
of disruption of various elements. In general, disruption of
two or more columns constitutes an unstable spinal injury.
White and Panjabi have developed a checklist for the di-
agnosis of clinical instability of the lumbar spine.14Their
criteria are the result of cadaveric biomechanical studies of
each segment of the spinal axis, as well as an extensive lit-
erature review. A state of clinical instability exists when a
patient earns fiveor more points (Table 78-1).
The White and panjabi system is applicable not only in
the setting of trauma but also in the setting of instability
secondary to neoplasm, infection, or prior surgery.
.
Mechanisms of Injury
Patterns of lumbar spine fractures are influenced by the in-
herent lordosis as well as the capacity for flexionand exten-
sion. The fact that the lumbar spine is normally in lordosis
means that a significantly great flexion force is necessary to
impart a compression type injury. Most injuries, therefore,
occur by distraction or direct axial loading. Injuries involv-
ing only a portion of the posterior column (fractures of the
transverse or spinous processes or fractures of the pars in-
terarticularis) do not cause acute instability. Clinically sig-
nificant lumbar spine fractures generally fall into one of
four categories: compression, burst, seat belt. or fracture-
dislocation, corresponding to the type of force applied.
Compression fractures result from failure of the anterior
column under axial loading (Fig.78-2). The middle column
remains intact (and may, in fact, act as a hinge). More
severe compression
fractures
may be associated
with
582
Lumbar and Lumbosacral Spine: Pathology of the Lumbosacral Spine
posterior ligamentous injury. Although common in the tho-
racolumbar spine, the natural lordotic curvature of the
lower lumbar spine renders these fractures the least preva-
lent. Roentgenographic studies demonstrate anterior wedg-
ing, with normal height of the posterior vertebral body.
There is no disruption of the pedicles. There is no subluxa-
tion relative to the vertebral body above or below.Likewise,
computed tomography (CT)demonstrates a loss of height
with minimal or no retropulsion of bony fragments intothe
spinal canal. In general, these fractures are considered sta-
bleand requirenosurgicalintervention.IS-I?
Burst fractures occur in response to direct axial loading
with the spine in a neutral or flexed position (Fig.78-3).
The burst fracture signifies disruption of both the anterior
and the middle columns. Ifin a neutral position at the time
of impact, comminution of the vertebral body, fracture of
the pedicles, and retropulsion of fragments into the spinal
canal result. Ifin a flexed position, the result is central com-
pression of the body as well as retropulsion of fragments
into the canal. Roentgenographic studies demonstrate a
> 50%loss of height (fracture of the posterior body cortex),
a widened interpedicular distance, splaying of the posterior
joints, and evidence of canal compromise by retropulsionof
bony fragments from either end plate. CT demonstrates
Figure78-2
Compressionfracture.
A-C
D
~
E
Figure 78-3
(A)Burstfracture type A.(8) Burstfracturetype B.(C)Burstfracturetype C.(D)Burstfracturetype D.(E)Burstfracturetype E.
Table78-2 BurstFractureTypes
Type Pattern
Applied Force
A
B
(
o
E
Fractureofbothendplates
Fractureofthesuperiorendplate
Fractureofthe inferiorendplate
Burstrotation
Burstlateralflexion
Pureaxialloading
Axial loadingwith flexion
Axialloadingwith flexion
Axialloadingwith rotation
Axial loadingwith flexion
78
Trauma of the Lumbar Spine
583
Type A fractures result from pure axial loading. Type B
fractures. the most common, result from axial loading in the
presence of flexion. These fractures typically occur at the
thoracolumbar junction. TypeCfractures are rare but are be-
lieved to occur by the same mechanism as type B.Type D
fractures occur by axial loading with rotation and may be
mistaken for fracture-dislocation injury.Type Efractures dif-
fer from pure lateral compression fractures insofar as widen-
ing
of the
interpedicular
distance
does
occur. The
retropulsed bony fragment may be more lateral than that
seen in the other subtypes. Burst fractures.by definition, are
unstable and require decompression as well as stabilization.
Seat belt-type
fractures result from application of a
flexion-distraction
force (Fig. 78-4). Smith and Kaufer
described an injury to the posterior elements-bony or liga-
mentous or both-accompanied
by a longitudinal splaying
of those elements.19 This injury represents failure of both
anteriorandposteriorfractures of the vertebral ring as well
asretropulsionofa bonyfragment into the spinal canal. Most
burstfracturesinvolveonly one end plate. Denis classified
thesefracturesinto five subtypes on the basis of applied
forceandresultantend plate involvement18(Table78-2).
Cf
II
I
D
Figure78-4
(A)Seatbelt-type injury. One-level type of injury with posterior failure through the ligament. (8) Seatbelt-type injury. One-level
injurywithposteriorfailurethrough the posterior arch (Chancefracture). (e) Seatbelt-type injury. Two-leveltype of injury with failure of the middle
column
atthelevelofthe disk.(D) Seatbelt-type injury. Two-leveltype of injury with failure through the posteriorvertebral body.
A
\lll
I
B
,..
584
Lumbar and Lumbosacral Spine: Pathology of the Lumbosacral Spine
A-C
D,E
F,G
Figure 78-5
(A) Fracture-dislocation. shear
subtype (lateral
view).
(B) Fracture-disloca-
tion,
flexion-rotational
subtype
(posterior
view).
(C) Fracture-dislocation,
slicefracture
subtype
(lateral
view).
(D) Fracture-disloca-
tion,
slice fracture
subtype
(posterior view).
(E) Fracture-dislocation,
shear subtypewith
failure
in posteroanterior
direction. (F)Frac.
ture-dislocation,
shear subtype with failurein
anteroposterior
direction.
(G) Fracture-dislo-
cation,
flexion-distraction
subtype with pre-
dominantly
ligamentous/disk
failure in all
three columns
leading
to subluxation.
theposteriorand the middle columns. There is no loss of
vertebralheight,nor is there displacement of the body.The
intactanteriorportion of the vertebral body acts as a hinge
suchthatthesefractures are unstable in flexion. D_enisdi-
videsseatbelt-type injuries into two subtypes.2o A one-
levelinjury-the so-called Chance fracture-may
be ori-
entedeitherthrough the ligaments or bone.21Two-level
injuriesinvolverupture of the middle column through ei-
therboneordiskand may be associated with intra-abdomi-
nalinjuries(Fig.78-4).
Roentgenographicallyvisualized horizontal splitting of
thetransverseprocesses and pedicles is pathognomonic for
thistypeof injury.This may occur in concert with a hori-
zontalsplittingof either or both the spinous process and
theparsinterarticularis. There is also an increase in the
posteriorheight of the disk space or of the vertebral body.
Insofarasaxialslices are configured in the plane of injury,
standarda
imaging (without
sagittal
and coronal recon-
struction)isoflimited use in this setting.
Fracture-dislocation injuries, involving failure of all three
columnsat a given level, are the least common
and most
unstable22-24(Fig. 78-5).
Denis
devised
a three-group
systemtocharacterize these fractures on the basis of mech-
anism.Theflexion-distraction subtype is a seat belt-type
fracturecoupledwith tearing of the anterior anulus and in-
volvementofthe anterior longitudinal ligament. This injury
allowsforsubluxationand dislocation.
Theshearsubtype occurs in extension. Disruption of
theanterior longitudinal ligament leads to tearing of the
disksuchthat translation
of one vertebral body on an-
otheroccurs.Translation may occur in an anteroposterior
orposteroanterior
direction,
leaving
a floating
posterior
arch.Suchfractures are often associated with a dural tear
and complete paraplegia.
The flexion-rotation
subtype
couplesa rotational component to subluxation or distrac-
tion.Rupture of the posterior
and middle
columns
may
beaccompanied by failure
of the anterior
column
under
rotation.This subtype is often accompanied by multiple
transverseprocess fractures. Unilateral or bilateral facet
fracturedislocation of the lumbosacral spine constitutes
afourth, exceedingly rare subtype.25.26The pathogno-
monicroentgenographic sign of these injuries is subluxa-
tion or dislocation seen on anteroposterior
or lateral
views.
. Diagnosis
Traumaticlumbar and lumbosacral
fractures
often occur in
concertwith injuries to other organ systems. Therefore, it is
importantto attend to the "ABCs":airway, breathing, and
circulation,prior to addressing
the potential
spine injury. A
systematic survey of neurological
function
is crucial. Level
ofconsciousnessand cranial nerve exam should be docu-
78
Trauma of the Lumbar Spine
585
mented. The motor exam should include testing of individ-
ual muscle groups as well as assessment of tone and deep
tendon reflexes.
The ASIA(American Spine Injury Association) motor
score is commonly employed as a relative measure of func-
tion.27The ASIAscore grades five groups of muscles in each
extremity on a scale of 0 to 5 for a total possible score of
100 (in a motor intact individual)(Table 78-3).
The presence of long tract signs may indicate a more
cephalad level of injury. Likewise,it is possible that more
than one spinal level may be involved. The sensory exam
should include modalities of light touch, pinprick, and pro-
prioception. Arectal exam (for tone and perianal sensation)
and bladder function should be noted. The Frankel scale,
grades I through VII,is also used to classify neurological
function.28
Most trauma centers will obtain anteroposterior
and
lateral x-rays of the entire spinal column for a newly ad-
mitted trauma patient. If a lumbar or lumbosacral frac-
ture is suspected,
these films should be supplemented
with oblique views. Initial review of plain x-rays should
be geared toward identification
of fractures and obvious
disruption
of bony alignment.
Further inspection may
yield evidence of loss of vertebral height, widening of the
interpedicular distance, loss of lordosis, subluxation, ro-
tation, foraminal narrowing, and/or spinal canal compro-
mise.
Evidence of any abnormality
on plain x-ray should
prompt further evaluation with a CTscan. Our center rou-
tinely requests 3 mm axial sections through the region of
interest, with sagittal and coronal reconstruction.
CTal-
lows for more detailed assessment of the bony anatomy,
including the amount of canal compromise, extent of ver-
tebral body comminution, and interpedicular distance. CT
also provides better visualization of the architecture
of
fractures through the pedicles, facets, laminae, and spin-
ous processes. Reconstructed
images add an additional
Table78-3
ASIAClassificationSystem
Grade
Motor Exam
o
1
2
Total paralysis
Palpable or visible contraction
Active movement, full range of motion with gravity
eliminated
Active movement,
full range of motion versus
gravity
Active movement, full range of motion versus
moderate resistance
Active movement, full range of motion versus
normally surmountable
resistance (normal)
3
4
5
586
Lumbar and Lumbosacral Spine: Pathology of the Lumbosacral Spine
dimension to investigation of fracture architecture
and
stability.
Although cr is an excellent adjunct for the evaluation of
bony deformity, it fails to define adequately
injury to
neural, ligamentous, and soft tissue structures. An MRIor
cr myelogram is necessary to characterize further the na-
ture of the injury. CTmyelogram will illustrate thecal sac
and nerve root compression in the axial and sagittal planes
if present. However, the utility of this diagnostic modality
is limited by its invasive nature and poor soft tissue resolu-
tion. Therefore, cr myelogram is used preferentially only if
MRIis not immediately available or is not feasible given the
presence of metallic foreign bodies. MRIwill demonstrate
disk disruption, epidural hematoma,
root compression,
spinal cord contusion (to the level ofthe conus), soft tissue
injury, and splaying of the spinous processes (consistent
with ligamentous injury). It is reasonable to perform a pre-
operative angiogram to define the course of the artery of
Adamkiewicz if a lateral extra cavitary approach is contem-
plated.
.
Treatment
Indicationsfor surgery include neurologicaldeficit secondary
to spinal canal compromise, gross instability from a complex
fracture,severe misalignment in the sagittal plane, and dural
laceration.Neurologicaldeficit resulting from lumbar or lum-
bosacral fractures may include various degrees of lower ex-
tremity weakness (level-dependent) plus or minus bowel
and bladder dysfunction (cauda equina). In general, neuro-
logicalfunction will improve to a certain extent followingde-
compression.Durallacerations, if not repaired primarily, may
lead to herniation of cauda equina, scarring, and subsequent
developmentofa chronic pain syndrome.
There does not appear to be a correlation between earlier
surgicalintervention and better neurological outcome.29-32
Instability, in this context, may be defined as a complex
fracture involving both anterior and posterior elements or
as ligamentous injury with fracture dislocation and facet
compromise. Correction of deformity in the sagittal plane
approximates correction of kyphosis. Indications for surgi-
cal intervention in the absence of a neurological deficit re-
main less clear.Outcomes with surgical intervention versus
conservative management are split.
Conservative management entails 6 to 12 weeks of strict
bedrest.33-35Guttman introduced the concept of postural re-
duction as a nonsurgical means of treating spinal frac-
tures.36Guttman and, later, Frankel reported that approxi-
mately 60% of patients treated by conservative means
improved at least one (Frankel) neurological gradeY Pa-
tients undergoing surgical intervention showed no greater
improvement. Still, postural reduction has limitations. The
potential
for systemic
complications
is great. Aggressive
deep venous thrombosis
prophylaxis
and pulmonary toilet.
as well as attention
to skin integrity, are necessary adjuncts
to conservative
treatment.
These therapies
require a dedi-
cated and vigilant team to implement.
And although correc-
tion of displacement
of a fracture-dislocation
frequently oc-
curs, abnormal
angulation
often follows.
The timing of surgical intervention
also remains contro-
versial. Early operation
(within 3 or 4 hours) advocates be-
lieve early decompression
maximizes neurological recovery.
However,
published
outcome
studies
do not necessarily
support
this option
over delayed intervention
(within 3-4
days).38Delayed intervention allows for appropriate diag-
nostic evaluation and stabilization of the patient priorto
surgery and minimizes
complications
without a demonstra-
ble effect on neurological
recovery.
The primary
goals of surgical
intervention
are decom-
pression
of neural elements
and stabilization.
The latteris
undertaken
with an eye toward restoration
of lumbar lordo-
sis and fixation of as few motion segments as possible while
preserving maximal neurological function. Secondarygoals
include
prevention
of systemic
(pulmonary,
gastrointesti-
nal, genitourinary,
and cutaneous)
complications as wellas
early mobilization
and rehabilitation.
The particular surgi-
cal approach
depends
on the anatomy
of the individual in-
jury. Anterior
and posterior
routes
for decompression are
available. The anterior
procedure
involves a retroperitoneal
approach
for vertebral
body resection
and placement ofa
strut graft. Posterior
procedures
include laminectomy, lat-
eral extracavitary,
and posterolateral
approaches.
Posterior laminectomy
is most appropriate
in the settings
of posterior
element
injury,
dural
laceration,
epidural
hematoma,
or isolated radiculopathy
secondary to neuralel-
ement compression.
With retraction
of the cauda, laminec-
tomy may also permit retrieval of ventral bone fragmentsim-
pinging on the canal. Although laminectomy
is preferredfor
its simplicity,
it offers potential
for further
destabilization
and may not provide adequate
exposure
for retrieval ofven-
tral or lateral fragments.39 Furthermore,
laminectomy cou-
pled with removal of interspinous
ligament and ligamentum
flavum increases the possibility of progressive flexiondefor-
mity and decreases stability of the injured spinal column.
A lateral extracavitary
(LECD)approach
is most appropri-
ate for lesions lateral or ventral to the spinal cord. Capener
first
described
this
approach
in
the
setting
of spine
trauma.40 Larson et al later
refined
the procedure.41 The
LECDapproach
involves resection
of the transverse process
to expose the lateral aspect of the vertebral column. Posteri-
orly extruded
fragments
may then be removed via a pos-
terolateral
decompression.
A transpedicular
posterolateral
approach
allows for decompression
of fragments ventralto
the spinal cord as well as for instrumentation
without a
separate
incision. The advantage
of these approaches isdi-
rect visualization
of the spinal canal without
the potential
78
Traumaof the LumbarSpine
587
A
Figure78-6
(A) Anterolateral approach with verte-
brectomy and decompressionofspinalcord. (B)Antero-
lateralapproach withstrut graft and Z-plateinstrumen-
tation.
B
Figure78-7
Posterior stabilization with laminar hooks and rod
construct.
Figure 78-8
Posterior stabilization with pedicle screw and longitudi-
nal rod construct.