TECHNIQUE ASSESSMENT

OBSERVATIONS ON SPINE DEFORMITY

AND SYRINGOMYELIA

Ulrich Batzdorf, M.D.
Department of Neurosurgery,
David Geffen School of Medicine,
University of California, Los Angeles,
Los Angeles, California
Larry T. Khoo, M.D.
Department of Neurosurgery,
University of California, Los Angeles,
Medical Center,
Los Angeles, California
David L. McArthur, Ph.D.,
M.P.H.
Department of Neurosurgery,
University of California, Los Angeles
Medical Center,
Los Angeles, California
Reprint requests:
Ulrich Batzdorf, M.D.,
Department of Neurosurgery,
University of California, Los Angeles
Medical Center
Box 956901,
Los Angeles, CA 90095-6901.
Email: ubatzdorf@mednet.ucla.edu
Received, May 15, 2006.
Accepted, February 23, 2007.
OBJECTIVE: Spine deformities, notably scoliosis, are known to occur in conjunction
with syringomyelia. This study aims to analyze the effect of laminectomies performed
in the course of treatment of syringomyelia. It examines the incidence, severity, and type
of spine deformity as it relates to the extent and location of laminectomies performed.
METHODS: Records of 169 patients were analyzed for evidence of spinal deformity
on imaging studies for the extent of the syringomyelic cavities and for previous surgi-
cal procedures on the spine. This analysis included patients with syringomyelia related
to Chiari malformation, as well as patients with primary spinal pathology.
RESULTS: Spinal deformities were encountered in 41% of Chiari-syringomyelia patients
who had not undergone previous surgery and in 57% of such patients who underwent
reoperation. Scoliosis, the most common type of deformity encountered, was likely to
be mild in patients who had not undergone previous surgery and severe in reoperated
patients. Spine deformity was significantly more common in those patients who had
more extensive bone removal.
CONCLUSION: Complete laminectomy should be avoided whenever possible in patients
with syringomyelia because local denervation of the axial musculature, added to loss of
medial anterior horn cells from syringomyelia, favors the development of spine deformi-
ties. This is particularly true of laminectomies performed at the junctional areas of the
spine, i.e., cervical-thoracic and thoracolumbar. Hemilaminectomy usually suffices for shunt
placement; instrumented stabilization should be considered in patients undergoing full
laminectomy, especially those considered to be at high risk of developing deformity.
KEY WORDS: Kyphosis, Laminectomy, Scoliosis, Shunting, Syringomyelia
Neurosurgery 61:370–378, 2007 DOI: 10.1227/01.NEU.0000279971.87437.1F www.neurosurgery-online.com

D
eformities of the spine related to syringomyelia, notably
neurogenic scoliosis, have been recognized for many
years (8, 9, 14, 22). This association is known to be par-
ticularly high in children younger than 10 years old (15). When
compared with patients with idiopathic scoliosis, children with
associated syringomyelia and Chiari malformation have an
increased preponderance of atypical curve patterns, often with
higher apices (9, 21). A variety of explanations have been
offered for the observation that neurogenic scoliosis is distinct
from idiopathic scoliosis; that of Huebert and MacKinnon (8),
which proposes loss of neurons to the axial musculature in
syringomyelia, seems to be one of the most plausible.
New or additional spinal deformities that develop after sur-
gical procedures have been performed for syringomyelia are a
less well-recognized problem. Our experience over the course
of more than 20 years has brought to light a group of patients
with syringomyelia whose spine deformity seemed to develop
as a consequence of a posterior surgical approach to the spine
by laminectomy in the course of treating some aspect of the
underlying syringomyelia. It is proposed that the underlying
pathophysiology may have some common features with that
postulated for patients with neurogenic scoliosis.
In the past 20 years, 169 surgical procedures were performed
on 169 patients by the senior author (UB); 105 of these were for
syringomyelia related to Chiari malformations and 64 were
related to primary spinal pathology. Of the Chiari-related cases,
63 procedures were the patient’s first procedure (“Chiari initial
surgery”); 42 procedures were performed on patients who had
previously undergone a surgical procedure for Chiari-related
syringomyelia at another institution (“Chiari reoperation”). Of
the 64 patients who underwent surgical procedures for
syringomyelia related to primary spinal pathology, the leading
underlying causes were trauma (n ⫽ 51), previous tumor or
related surgery (n ⫽ 5), and meningitis (n ⫽ 2).
The records of these patients, all of whom were under the
care of the senior author (UB), were analyzed for evidence of
spinal deformity on imaging studies, for the extent of the
syringomyelic cavity, and for previous surgical procedures on

370 | VOLUME 61 | NUMBER 2 | AUGUST 2007 www.neurosurgery-online.com

 OBSERVATIONS ON SPINE DEFORMITY AND SYRINGOMYELIA

the spine. Identification of spine deformities was based on

magnetic resonance imaging scans and, in many patients, TABLE 1. Number and percent of patients with spine deformity
x-rays of the spine. These studies were obtained as part of the by group
usual evaluation of patients with syringomyelia in planning Chiari Chiari
treatment. Because these patients were referred for their neuro- Deformity initial surgery reoperation
logical problems rather than the associated spine deformity, (n ⴝ 63) (n ⴝ 42)a
standing spine x-rays were not obtained. For the purposes of
None or nonsignificant 37 (58%) 18 (43%)
this review, spine deformities were defined as mild when they
were not associated with symptoms attributable to the defor- Significant deformity 26 (41%) 24 (57%)
mity and were evident only on imaging studies, moderate Thoracic scoliosis 24 (38%) 14 (33%)
when they were not symptomatic but were evident on clinical Mild 13 (20%) 1 (2%)2
evaluation, and severe when they required treatment in the Moderate 7 (11%) 6 (14%)3
past or were presently symptomatic and the deformity, per se, Severe 4 (6%) 7 (17%)5
related instability or sequelae required treatment. All patients
Kyphosis 2 (4%) 8 (19%)
were initially evaluated by the senior author (UB).
Statistical analyses included the Fisher exact test for categor- Cervical, mild 1 (2%) 1 (2%)
ical data, analyses of variance for grouped continuous data, Cervical, severe 0 1 (2%)1
and event history analysis for follow-up data using package Thoracic, mild 1 (2%) 2 (5%)
eha within R version 2.3.1 (19). The study was undertaken with Thoracic, severe 0 2 (5%)4
approval of the University of California, Los Angeles (UCLA)
Hyperlordosis 0 6 (14%)
Institutional Review Board, Office for Protection of Research
Subjects, protocol 04-05-100-03. Cervical straightening 0 3 (7%)

RESULTS
Table 1 shows the type and severity of spine deformities
encountered in patients with Chiari-related syringomyelia, both
those who underwent their initial Chiari surgery at UCLA and
those who underwent Chiari reoperation. Spinal deformities
were encountered in a minority (41%) of patients with Chiari-
syringomyelia who had not undergone previous surgery, but in
a majority (57%) of patients who underwent reoperation (P ⫽
0.002). Among all deformities, scoliosis was the most common
deformity in these two groups. Nearly all (92%) of the Chiari
initial surgery group, but fewer (58%) of the Chiari reoperation
group, demonstrated scoliosis (P ⫽ 0.008). If scoliosis was pres-
ent, its severity varied significantly between initial surgery and
reoperation groups (P ⫽ 0.007); the former was predominantly
rated mild, whereas the latter was predominantly rated severe.
Kyphotic deformities were not significantly more frequent in the
reoperation group (P ⫽ 0.132), and both hyperlordosis and
straightening were encountered only in the reoperation group.
Figure 1 shows the extent of bone removal in patients with
Chiari-related syringomyelia for the Chiari initial surgery group
as a cluster, and then separately for those within the Chiari reop-
eration group who demonstrated no deformity (n ⫽ 18) and
those who demonstrated deformity (n ⫽ 23). Analysis of only the
latter two groups showed a highly significant difference in total
extent of bone removal (P ⫽ 0.001). Those with no deformity
averaged bone removal spanning 2.6 vertebral segments,
whereas those with deformity averaged 4.7. Bone removal at
multiple locations was seen only in those with deformity.
Figure 2, A and B, reflects the extent of the syringomyelic
cavities as related to the deformities in those patients with pri-
mary spinal pathology only and in the Chiari initial surgery
group and Chiari reoperation groups, respectively. Both fig-
a
Superscript numerals identify patients shown in Table 3.
ures demonstrate that the length of the syringomyelia cavities
did not relate to the severity of the spine deformity; no statis-
tically significant difference was identified either in the initial
surgery group (F ⫽ 1.41; df ⫽ 2, 21; P ⫽ 0.514) or in the reop-
eration group (F ⫽ 0.62; df ⫽ 4, 18; P ⫽ 0.65).
Table 2 indicates the principal underlying cause in patients
with syringomyelia of primary spinal origin, i.e., not related to
craniovertebral junction deformities (n ⫽ 60), categorized by
presence (n ⫽ 16) or absence (n ⫽ 44) of significant deformity
(the latter including those with stable fixation). Trauma with or
without early fusion was the most frequent presumed cause
(47%) followed by tumors of all types (22%). Presence or
absence of deformity was not statistically related to the nature
of the underlying pathology (P ⫽ 0.363).
For those persons with primary spinal syringomyelia of trau-
matic etiology, Figure 3A shows the extent of bone removal
during surgery with patients grouped by presence or absence
of deformity and also by the presence or absence of fusion.
The total extent of bone removal was not statistically different
between these subgroups (F ⫽ 0.32; df ⫽ 3, 24; P ⫽ 0.991). For
both those with and without spinal deformity, a third had
undergone previous spinal fusion. Laminectomies were evenly
distributed throughout the spine in patients with posttraumatic
syringomyelia. In those patients who went on to develop defor-
mities, laminectomies in the vicinity of the thoracolumbar junc-
tion (⫾2 vertebral levels) were significantly more frequent than
in those without deformity (P ⫽ 0.035); all were found to be
kyphotic deformities.
Figure 3B shows levels of bone removal for those persons
with primary spinal syringomyelia stemming from arachnoid
cysts (n ⫽ 7), tumors (n ⫽ 10), or meningeal inflammation

NEUROSURGERY VOLUME 61 | NUMBER 2 | AUGUST 2007 | 371

BATZDORF ET AL.

median, 4 yr). Follow-up rates

classified according to Chiari
initial surgery (n ⫽ 62), Chiari
reoperation (n ⫽ 41), and pri-
mary spinal pathology (n ⫽ 60)
were 98, 98, and 94%, respec-
tively. The sole question for the
purposes of this article was
whether the individual had
improved, stayed the same, or
worsened relative to their status
immediately after recuperation
from surgery or had died.
A statistically significant dif-
ference was found in status on
follow up. Improved status was
noted for 18 of the patients with
Chiari malformation who had
initial surgery (30% of those
interviewed), for five of the
patients with Chiari malforma-
tion who underwent reopera-
tion (13%), and for 14 of the
patients with primary spinal
pathology (26%). Worsening
was found for three of the
patients with Chiari malforma-
tion who had initial surgery
(5%), for nine of the patients
with Chiari malformation who
underwent reoperation (24%),
and for 18 of the patients with
primary pathology (34%) (P ⬍
FIGURE 1. Bar graph showing the extent of bone removal by laminectomy in 105 patients who underwent surgi- 0.001). This evidence of differ-
cal treatment for Chiari-related syringomyelia. Patients with significant spine deformities who are also included in
ential outcome was intertwined
Table 3 are identified by an asterisk. Note that all 63 patients who had Chiari initial surgery underwent only occip-
ital craniectomies (designated as “0 cervical”) and C1 laminectomies.
with differing event histories
for each of the three groups. In
the initial surgery group, no
(n ⫽ 8). With the latter two categories, further subdivided by deaths were identified after 2 years and no worsening was iden-
presence or absence of deformity, no statistically significant tified after 6 years postsurgery. In the reoperation group, no deaths
difference between subgroups was found (F ⫽ 1.66; df ⫽ 4, 22;
P ⫽ 0.196). Bone removal for this group was generally limited
to two or three vertebral levels and, with one exception, in- TABLE 2. Number and percent of patients with selected princi-
pal precipitating event by spine deformity
volved only the thoracic spine. Laminectomies for spinal cord
tumors or for syringomyelic cavities developing after tumor No deformity Significant
removal (n ⫽ 4) similarly were mostly located at thoracic lev- Event or stable fix- deformity
els. Of these four patients, two had a laminectomy that crossed ation (n ⴝ 44) (n ⴝ 16)a
the cervical-thoracic junction and one had a laminectomy that Trauma 18 (41%) 9 (56%)
crossed the thoracolumbar junction. Spinal surgery for menin- With early fusion 6 (14%) 3 (19%)6
geal inflammatory processes was also located primarily at tho- Tumor-related 9 (20%) 4 (25%)
racic spinal levels.
Arachnoid cyst 8 (18%) 0
Follow-up information was available for 163 of the study
patients either by brief personal interview with the senior author Meningeal inflammation 7 (16%) 1 (6%)
(UB) (n ⫽ 152) or by receipt of information that a patient was Developmental abnormality 2 (5%) 1 (6%)
deceased (n ⫽ 11). A total of six patients could not be located. a
Superscript numeral identifies a patient shown in Table 3.
Follow-up periods varied from 1 to 22 years (mean, 5.25 yr;

372 | VOLUME 61 | NUMBER 2 | AUGUST 2007 www.neurosurgery-online.com

 OBSERVATIONS ON SPINE DEFORMITY AND SYRINGOMYELIA

A special subgroup of six

A patients demonstrated severe
spine deformities that coin-
cided both in time and location
with surgical procedures that
involved dorsal muscle strip-
ping performed in conjunction
with laminectomies directed at
the underlying syringomyelia
pathology. Table 3 summarizes
clinical data on these six pa-
tients, five of whom had Chiari-
related syringomyelia and one
of whom had posttraumatic
syringomyelia. This cluster of
patients, all of whom had oper-
ations performed at UCLA
Medical Center, illustrates the
spectrum of abnormalities that
may be encountered.
The spine deformities were
regarded as stable except in two
patients (Patients 1 and 6) who
underwent fusion procedures
aimed at providing stability and
B partial correction of the deformi-
ties. Patient age did not distin-
guish between patients with
Chiari-syringomyelia who were
reoperated and who showed
deformities (average age, 41.8
yr) and those who did not (aver-
age age, 38.4 yr) (P ⫽ 0.331).
Progression of spine deformities
was not encountered during the
period of observation.

Illustrative Case
Patient 3
A 20-year-old man presented to
the UCLA Medical Center with
neck pain and progressive balance
and gait impairment. At age 17,
he had developed impaired right
upper extremity function and gait
difficulty, and a Chiari malfor-
mation with syringomyelia was
diagnosed. At age 18, he had
FIGURE 2. A and B, bar graphs showing the longitudinal extent of syrinx cavity formation in patients with Chiari-
undergone a suboccipital decom-
syringomyelia with spine deformities. A, patients who underwent their initial Chiari surgery at UCLA Medical
pression, which included a C1
Center (UCLA MC) are shown. B, patients who underwent Chiari reoperation are shown; patients included in Table
and partial C2 laminectomy as
3 are identified by an asterisk.
well as a C7 laminectomy extend-
ing to C6 and T1 for placement of
were identified after 2 years and no worsening after 6 years. In the a syrinx shunt. Surgery was performed at another hospital. On exam-
primary spinal group, no improved or unchanged patients were ination, he had weakness and atrophy of the right upper extremity
identified after 8 years of follow up. and lower extremity spasticity with extensor plantar responses.

NEUROSURGERY VOLUME 61 | NUMBER 2 | AUGUST 2007 | 373

BATZDORF ET AL.

mately 6 years. Review of his

A diagnostic studies demonstrates
progressive upper thoracic ky-
phosis with compensatory cervi-
cal hyperlordosis developing over
a 2-year period after his initial
operation.

DISCUSSION
Patients with syringomyelia
who underwent a laminec-
tomy, other than the first cer-
vical lamina, seem to be at
risk for spine deformity at the
level of their laminectomy. We
postulate that this is the result
of a combination of impair-
ment of the nerve supply to
the supporting spinal mus-
cles, resulting from nerve cell
destruction in the cord attrib-
utable to syringomyelia and
mechanical denervation of the
axial musculature attributable
B to laminectomy. Cord trauma
from the original injury may
a d d t o t h e n e u ro n a l l o s s .
Gravity affecting the thoracic
viscera may also play a role.
The effects of gravity on a
poorly supported spine seem
to be a common feature of pos-
tural deformities seen with
syringomyelia. This applies
equally to syringomyelia asso-
ciated with Chiari malforma-
tion, trauma of the spine, or
other causes. The association
of neurogenic scoliosis with
syringomyelia is extensively
documented in the literature
(4, 6, 9, 15, 17, 18, 21, 22), espe-
cially in children. Numerous
authors have pointed out the
very high incidence of neuro-
genic scoliosis in children
FIGURE 3. A, bar graph showing the extent of bone removal by laminectomy for patients with posttraumatic younger than age 10 (15) and
syringomyelia. The patient included in Table 3 is identified by an asterisk. B, bar graph showing the extent of bone younger than age 16 (8, 9).
removal in patients with syringomyelia of differing underlying pathology, including arachnoid cysts, tumor or There does not seem to be gen-
tumor-related surgery, and meningeal inflammation. eral agreement on the predom-
inance of left versus right tho-
Suboccipital reexploration at the UCLA Medical Center approxi-
racic curvatures (4, 6, 18). Although levorotatory curves have
mately 3 years after his initial operation demonstrated extensive scar traditionally been associated with the presence of intra-
tissue and a constrictive dural band at the level of the foramen mag- medullary spinal cord lesions, more recent studies have sug-
num (Fig. 4). The patient’s gait and right upper extremity function gested that the incidence of right and left curvatures is approx-
improved after reoperation and he has been followed for approxi- imately equal (9, 17). However, there have been no rigorous

374 | VOLUME 61 | NUMBER 2 | AUGUST 2007 www.neurosurgery-online.com

 OBSERVATIONS ON SPINE DEFORMITY AND SYRINGOMYELIA

TABLE 3 . Clinical data for patients who demonstrated severe spine deformitiesa

Diagnosis Age at Indication for Type of

Patient Laminectomy Indication Deformity Deformity Time
(age of Surgery additional additional
no. levels for surgery levels type interval
diagnosis, yr) (yr) surgery surgery
1 CM, Suboccipital ⫹ 44 Decompression C4-C5 Kyphosis Pain, tethered 6 mo Untethering
syringomyelia C1–C6 cord C7–T3 duraplasty
(44) syrinx Occipital-
cervical fusion
2 Syringomyelia Suboccipital ⫹ 39 Decompression C3–4 Kyphosis Tethered at T5 4 yr Declined
(38) C1, C2 44 Thoracic
CM (39) T5–T6 scoliosis
Scoliosis (?)
3 CM Suboccipital, 19 Decompression C3–T2 Cervical Myelopathy 2.5 yr Reexplore
Syringomyelia C1⫹ partial C2 Syrinx shunt hyperlordosis suboccipital
(18) C5–T1 C7–T2 craniectomy,
Scoliosis kyphosis intradural lysis
Scoliosis (left)
4 CM Suboccipital ⫹ 27 Decompression C3–T4 Cervical Myelopathy 11 yr Decompression
Syringomyelia C1, C2 39 Syrinx shunt hyperlordosis
(27) Scoliosis T4–5, T10–11 45 Syrinx shunt Thoracic
C7–T4 kyphosis
5 Scoliosis (10) T2–T3 28 Syrinx shunt T2–T4 Kyphosis Pain 14 yr Untethering
Syringomyelia T3–T4 Tethered at Duraplasty
(25) Tethered C7–T3, syrinx
CM (41) C7–T3
Syrinx
6 Posttraumatic T1–T2 29 Fusion C7–T2 Cervical Myelopathy 17 yr Posterior
syringomyelia T4–T5 41 Syrinx shunt hyperlordosis cervical fusion
(22) T1–T3 46 Lysis of
intradural scar

a
CM, Chiari malformation.
b
Time interval to consultation or surgery for problems resulting from deformity.

studies correlating the rostrocaudal extent of the syrinx cavity or
its lateral expansion, when present, with the scoliotic deformity.
In the absence of such information, one may speculate that
when there is a symmetrical, centrally located syringomyelic
cavity that extends into the thoracic spine and asymmetry of
the thoracic viscera, the heart lying on the left side would
encourage gravity to deform the spine. The relatively large
number of children younger than 10 years of age who still
have a residual central canal of the spinal cord (23) also corre-
lates with the higher incidence of syringomyelia and scoliosis
in children younger than 16 compared with those older than
16 (8). Huebert and MacKinnon (8) assumed that the syrinx
cavity expands asymmetrically and offered asymmetrical loss
of medial anterior horn cells as a result of the syrinx cavity as
a possible explanation for neurogenic scoliosis. Isu et al. (9)
also postulated asymmetry of the syrinx cavity. The medial
anterior horn cells are believed to be the cells of origin for the
motor innervation of the paraspinal axial musculature.
Experience with spinal curvature problems developing after
poliomyelitis, which also destroys anterior horn cells, offers
many parallels with regard to pathogenesis (10, 20).
Reversibility of scoliosis with posterior fossa decompression
(4, 9, 15) in younger patients suggests that some of the spinal
cord changes in this more plastic population may be partially
reversible (20); collapse of the syrinx cavity often follows pos-
terior fossa decompression in younger patients. Gravity can
also produce spinal deformity in the sagittal plane after
laminectomy over an area of syringomyelia (13). Depending
on the level of syringomyelia and of the laminectomy, one
may also see the effects of weight of the head, of the thoracic
viscera, or of the upper torso in producing a combination of
sagittal and rotational imbalance.
Laminectomy as well as preparation for spinal fusion neces-
sitates dissection and retraction of the axial musculature from
the laminae, even beyond the level of bone removal, and results
in some degree of local denervation. Permanent changes in
imaging and muscle strength, particularly of the spinal exten-
sor muscles, have been demonstrated after lumbar laminec-

NEUROSURGERY VOLUME 61 | NUMBER 2 | AUGUST 2007 | 375

BATZDORF ET AL.

tomies (7, 12). Segmental add to the neurological problem. Occipital craniectomy with
A B denervation and atrophy of removal of only the C1 lamina in patients with Chiari malfor-
the multifidi has also been mation does not produce spine deformity, but more extensive
confirmed by electrodiagnos- bone removal may be associated with greater likelihood of
tic studies in a significant spine deformity (Figure 1). All but one of the patients with
proportion of patients who Chiari malformation with deformities shown in Table 3 had
underwent dorsal lumbar undergone at least a C1 and C2 laminectomy; one patient
muscular exposure and underwent a high thoracic laminectomy only. Three other
retraction (11, 13). Similar patients had laminectomies at or very close to the cervical–
changes must occur after thoracic junction. Bone removal limited to two thoracic lami-
laminectomy at any level of nae rarely results in spinal deformity (Figure 3B; Table 2). As
the spine. Postlaminectomy shown in Table 1, 41% of patients with Chiari malformation
FIGURE 4. T1-weighted mag- kyphotic deformities seen in who had their initial Chiari surgery at UCLA had an existing
netic resonance imaging scans of a the absence of syringo- scoliotic deformity of the thoracic spine, and two others had
20-year-old man (Patient 3) with myelia (1) are primarily the mild kyphosis. Preexisting spine deformities add to the risk of
Chiari malformation and result of compromise of the further deformity with surgery and may account for the
syringomyelia. A, preoperative T1- facet joints. higher incidence of deformity, 57%, in patients with Chiari
weighted magnetic resonance We propose that the com- malformation who underwent reoperation. Patients with
imaging scan of the cervical spine bination of surgical trauma Chiari malformation who underwent reoperation also had
demonstrating severe tonsillar
from dorsal muscle retrac- more extensive syrinx cavities (Fig. 2, A and B), possibly the
descent to the lower margin of the
tion, resulting in muscle result of longer duration of their disease.
C1 lamina and a syrinx cavity
beginning at the C7 level. B, 2- injury and in local denerva- Procedures such as syrinx shunting through laminectomy
year postoperative T1-weighted tion, added to partial ante- may not be avoidable in some patients. The considerations pre-
magnetic resonance imaging scan rior horn cell loss from the sented here suggest that such procedures be performed with
of the cervical spine demonstrat- syrinx cavity creates a partic- minimal disruption of soft tissue and bone. Thus, hemilaminec-
ing evidence of C1, C2, partial C6, ularly weak axial muscula- tomy, leaving the interspinous ligament intact, would appear
C7, and T1 laminectomies. ture. This allows the forces preferable to complete laminectomy. When laminectomy can-
Markedly increased cervical lordo- of gravity to act in such a not be avoided, consideration should be given to immediate
sis is evident on this study. The manner as to result in spine instrumented stabilization, particularly if the laminectomy is at
patient also has an upper thoracic
deformity over a period of or near a junctional area of the spine.
kyphosis intrawedge-shaped
months or years. Although The present study is sufficiently well powered to adequately
T1–T2 intervertebral disc.
there is likely overlap in the enumerate the principal deformities of, and precipitating
innervation pattern of sple- events leading to, syringomyelia and to describe the varying
nius capitis, semispinalis, and multifidi, the longitudinal patterns of bone removal during surgery. However, the issue of
extent of the syrinx cavity favors segmental or multilevel par- outcomes after surgical intervention for Chiari malformation is
tial denervation. The unique combination of both underlying not completely resolved in the present study. Although signif-
and iatrogenic neuromuscular instability may be partially icantly differing outcomes were observed, such differences can-
responsible for the clinical syndrome encountered in our six not be disentangled from varying lengths of follow up and
patients. Junctional areas of the spine, where the normal cur- other factors in the present sample.
vatures reverse, such as the cervical-thoracic junction or the
thoracolumbar junction, seem particularly prone to develop REFERENCES
such deformities as suggested in Figures 1 and 3A. The excep-
tional importance of muscles inserting into the C2 spinous 1. Albert TJ, Vacarro A: Postlaminectomy kyphosis. Spine 23:2738–2745, 1998.
process and perhaps also into C7 (16) point out the signifi- 2. Batzdorf U: Syringomyelia, Chiari malformation and hydromyelia, in
Youmans JR (ed): Neurological Surgery. Philadelphia, W.B. Saunders Company,
cance of posterior structures in maintaining normal spinal 1995, ed 4, pp 1090–1109.
alignment. 3. Batzdorf U: Treatment of syringomyelia associated with Chiari I
Depending on the extent and anatomic level of the laminec- Malformation, in Batzdorf U, Tamani N, Nagashima T (eds): Syringomyelia,
tomy, and the coexistence of cord trauma, one may see Current Concepts in Pathogenesis and Management. Tokyo, Springer-Verlag,
2001, pp 121–135.
kyphotic deformities (Patients 1, 2, and 5) or hyperlordosis
4. Brockmeyer D, Gollogly S, Smith JT: Scoliosis associated with Chiari I malfor-
(Patients 3, 4, and 6). The explanation for the development of mations: The effect of suboccipital decompression on scoliosis curve progres-
hyperlordosis in some patients and kyphosis in others must sion. A preliminary study. Spine 28:2505–2509, 2003.
involve a number of different factors. Head support and sagit- 5. Dyste GN, Menezes AH, VanGilder JC: Symptomatic Chiari malformations.
tal balance are key elements. High thoracic loss of axial sup- An analysis of presentation, management, and long-term outcome.
J Neurosurg 71:159–168, 1989.
port seems to favor the development of thoracic kyphosis 6. Eule JM, Erickson MA, O’Brien MF, Handler M: Chiari I Malformation asso-
with compensatory cervical hyperlordosis. Kyphosis of the ciated with syringomyelia and scoliosis. A twenty-year review of surgical and
spine favors anterior tethering of the spinal cord, which may nonsurgical treatment in a pediatric population. Spine 27:1451–1455, 2002.

376 | VOLUME 61 | NUMBER 2 | AUGUST 2007 www.neurosurgery-online.com

 OBSERVATIONS ON SPINE DEFORMITY AND SYRINGOMYELIA
7. Gejo R, Matsui H, Kawaguchi Y, Ishihara H, Tsuji H: Serial changes in trunk
muscle performance after posterior lumbar surgery. Spine 24:1023–1028, 1999.
8. Huebert HT, MacKinnon WB: Syringomyelia and scoliosis. J Bone Joint Surg
Br 51B:338–343, 1969.
9. Isu T, Chono Y, Iwasaki Y, Koyanagi I, Akino M, Abe H, Abumi K, Kaneda K:
Scoliosis associated with syringomyelia presenting in children. Childs Nerv
Syst 8:97–100, 1992.
10. Levine DB: Poliomyelitis, in Hardy JH (ed): Spinal Deformity in Neurological
and Muscular Disorders. St. Louis, CV Mosby, 1974, pp 111–139.
11. MacNab I, Cuthbert H, Godfrey CM: The incidence of denervation of the
sacrospinales muscles following spinal surgery. Spine 2:294–298, 1977.
12. Mayer TG, Vanharanta H, Gatchel RJ, Mooney V, Barnes D, Judge L, Smith S,
Terry A: Comparison of CT scan muscle measurements and isokinetic trunk
strength in postoperative patients. Spine 14:33–36, 1989.
13. McLaughlin MR, Wahlig JB, Pollack IF: Incidence of postlaminectomy kypho-
sis after Chiari decompression. Spine 22:613–617, 1997.
14. McRae DL, Standen J: Roentgenologic findings in syringomyelia and
hydromyelia. Am J Roentgenol Radium Ther Nucl Med 98:695–708, 1966.
15. Muhonen MG, Menezes AH, Sawin PD, Weinstein SL: Scoliosis in pediatric
Chiari malformations without myelodysplasia. J Neurosurg 77:69–77, 1992.
16. Nolan JP Jr, Sherk HH: Biomechanical evaluation of the extensor musculature
of the cervical spine. Spine 13:9–11, 1988.
17. Özerdemoglu RA, Denis F, Transfeldt EE: Scoliosis associated with
syringomyelia: Clinical and radiologic correlation. Spine 28:1410–1417, 2003.
18. Piper JG, Menezes AH: Chiari malformation in the adult, in Menezes AH,
Sonntag VKH (eds): Principles of Spinal Surgery. New York, McGraw-Hill,
1996, pp 379–394.
19. R Language Definition: http://cran.r-project.org/doc/manuals/R-lang.html.
Accessed February 27, 2007.
20. Robin GC: Neurological Disease and Scoliosis: Scoliosis. New York, Academic
Press, 1973, pp 37–42.
21. Spiegel DA, Flynn JM, Stasikelis PJ, Dormans JP, Drummond DS, Gabriel
KR, Loder RT: Scoliotic curve patterns in patients with Chiari I malformation
and/or syringomyelia. Spine 28:2139–2146, 2003.
22. Williams B: Orthopaedic features in the presentation of syringomyelia. J Bone
Joint Surg Br 61B:314–323, 1979.
23. Yasui K, Hashizume Y, Yoshida M, Kameyama T, Sobue G: Age-related mor-
phologic changes of the central canal of the human spinal cord. Acta
Neuropathol 97:253–259, 1999.
Acknowledgments
Support for David L. McArthur, Ph.D., M.P.H., was provided by RO1NS40777,
P50NS044378, and RO1NS049471.
COMMENTS
T he authors performed a retrospective analysis of an extensive series
of 169 patients treated for syringomyelia. Although a standardized
radiographic analysis of coronal and sagittal spinal alignment was not
performed, sufficient information obtained from magnetic resonance
imaging and limited plain x-rays was used to identify patients with
spinal deformity. This analysis found that 41% of Chiari-related
syringomyelia patients who had not undergone previous surgery had
spinal deformity. Seventeen percent were found to have moderate or
severe thoracic scoliosis. The risk of spinal deformity was even higher
in patients undergoing reoperation for Chiari-related syringomyelia,
with 31% having moderate or severe thoracic scoliosis and 14% having
cervical or thoracic kyphosis.
The authors found worsened status relative to their status immedi-
ately after recuperation in 5% of the Chiari initial surgery patients,
24% of the Chiari reoperation patients, and 34% of the primary pathol-
ogy patients. Although the information was not provided, it would
have been helpful to know if there was a relationship with deformity
progression.
NEUROSURGERY
Supine imaging tends to underestimate the true incidence of spinal
deformity. This study indicates that spinal deformity may be more
common than has been generally appreciated. Standing x-rays of the
spine should be considered in this patient population before surgical
intervention to potentially reduce the likelihood of a progressive defor-
mity by adjusting the surgical procedure. In cases of significant preex-
isting deformity, the additional spinal instrumentation may reduce
deformity progression. A prospective analysis of the true incidence,
magnitude, factors associated with risk for progression and impact on
functional outcome of spinal deformity in these patients would further
build upon this important preliminary data.
Christopher I. Shaffrey
Charlottesville, Virginia
T he authors studied 169 patients with syringomyelia, including
patients with Chiari-related syringomyelia and primary spinal
syringomyelia. They evaluated whether or not deformity had devel-
oped after surgical procedures treating these entities. They divided
the 169 patients into the Chiari initial surgery group, composed of
63 patients with first time Chiari-related syringomyelia treated by
the authors. Another 42 patients who had previously undergone
surgical procedures for Chiari-related syringomyelia at another
institution and then treated by the author were assigned to the
Chiari reoperation group. Another 64 patients underwent surgical
procedures for syringomyelia related to primary spinal pathology,
including trauma (n = 51), previous tumor or related surgery (n = 5),
and meningitis (n = 2). The authors analyzed these patients with
spinal deformity on images for the extent of the syringomyelic cav-
ity and previous surgical procedures. They found 41% deformity in
the Chiari initial surgery group and 52% deformity in patients who
had Chiari reoperation.
The authors suggest that to avoid these deformities, surgery with
minimal disruption of soft tissue and bone (i.e., hemilaminectomy)
leaving the interspinous ligament intact, would be preferable to com-
plete laminectomy. Also, when laminectomy cannot be avoided to treat
a Chiari-related syringomyelia, consideration should be given to imme-
diate instrumented stabilization, particularly if the laminectomy is at or
near a junctional area of the spine.
The authors’ observance and subsequent recommendations are cer-
tainly reasonable. In the future, experience from the authors or pos-
sibly other groups who followed these suggestions would then hope-
fully show a decreased percentage of deformity in patients with
syringomyelia.
Volker K.H. Sonntag
Phoenix, Arizona
B atzdorf et al. examined the occurrence of spine deformities in rela-
tion to syringomyelia in 169 patients. The patients were divided
into two groups: those operated on primarily by the authors and those
who had surgery at another institution. In general, both of these groups
had approximately a 50% incidence of deformity. The deformity often
occurred after the operative procedure. Although the exact cause for
the deformities is unknown, they are probably related to muscle dissec-
tion, bony resection, and anterior horn cell dysfunction, as discussed by
the authors. The recommendations provided by Batzdorf et al. in the
final paragraph are reasonable and will most likely result in a signifi-
cant decrease in the incidence of deformity after Chiari surgery.
Vincent C. Traynelis
Iowa City, Iowa
VOLUME 61 | NUMBER 2 | AUGUST 2007 | 377
BATZDORF ET AL.
B atzdorf et al. have provided information from their experience that
is invaluable for surgeons caring for patients with Chiari malforma-
tion and syringomyelia. Other than at C1, laminectomies in patients
with syringomyelia place the patient at significant risk for postopera-
tive deformity. Neural injury from syringomyelia (i.e., anterior horn cell
injury), combined with structural integrity disruption as a result of
laminectomy, augments the chance of deformity and deformity pro-
gression. When such operations are indicated, limited laminectomy
(e.g., hemilaminectomy) and/or structural augmentation via instru-
mentation and fusion are prudent. The authors are to be congratulated
for solidifying our thought processes regarding the clinical decision
making process.
Edward C. Benzel
Cleveland, Ohio
378 | VOLUME 61 | NUMBER 2 | AUGUST 2007
T he authors share with us their extensive experience in treating
syringomyelia and associated spinal column deformities. Their ret-
rospective analysis of these 169 cases is a monumental work, but the
lack of detailed preoperative deformity measurements makes it difficult
to draw conclusive treatment recommendations from the data.
Ultimately, the prediction of a progressive spinal column deformity
after laminectomy is predicated on basic biomechanical principles,
including the spinal level of treatment, extent of bone and joint removal,
global spinal alignment, age, and the presence of any neuromuscular
disorders. The authors, with their significant experience, have com-
mented on all of these factors, highlighting their importance in consid-
ering the long-term consequences of surgery for intramedullary cysts.
Michael Y. Wang
Los Angeles, California
www.neurosurgery-online.com