Y. Morio et al., Correlation between operative outcomes of cervical compression myelopathy and MRI of the spinal cord, SPINE, 26(11), 2001, pp. 1238-1245
Study Design. Magnetic resonance images of cervical compression myelopathy
were retrospectively analyzed in comparison with surgical outcomes.
Objectives. To investigate which magnetic resonance findings in patients wi
th cervical compression myelopathy reflect the clinical symptoms and progno
sis, and to determine the radiographic and clinical factors that correlate
with the prognosis.
Summary of Background Data. Signal intensity changes of the spinal cord on
magnetic resonance imaging in chronic cervical myelopathy are thought to be
indicative of the prognosis. However, the prognostic significance of signa
l intensity change remains controversial.
Methods. The participants in this study were 73 patients who underwent cerv
ical expansive laminoplasty for cervical compression myelopathy. Their mean
age was 64 years, and the mean postoperative follow-up period was 3.4 year
s. The pathologic conditions were cervical spondylotic myelopathy in 42 pat
ients and ossification of the posterior longitudinal ligament in 31 patient
s. Magnetic resonance imaging (spin-echo sequence) was performed in all the
patients. The transverse area of the spinal cord at the site of maximal co
mpression was computed, and spinal cord signal intensity changes were evalu
ated before and after surgery. Three patterns of spinal cord signal intensi
ty changes on T1-weighted sequences/T2-weighted sequences were detected as
follows: normal/ normal, normal/high-signal intensity changes, and low-sign
al/high-signal intensity changes. Surgical outcomes were compared among the
se three groups, The most useful combination of parameters for predicting p
rognosis was determined using a stepwise regression analysis.
Results. The findings showed 2 patients with normal/ normal, 67 patients wi
th normal/high-signal, and 4 patients with low-signal/high-signal change pa
tterns before surgery. Regarding postoperative recovery, the preoperative l
ow-signal/high-signal group was significantly inferior to the preoperative
normal/high-signal group. There was no significant difference between the t
ransverse area of the spinal cord at the site of maximal compression in the
normal/high-signal group and the low-signal/high-signal group. A stepwise
regression analysis showed that the best combination of surgical outcome pr
edictors included age (correlation coefficient R = -0.348), preoperative si
gnal pattern, and duration of symptoms (correlation coefficient R = -0.231)
.
Conclusions. The low-signal intensity changes on T1-weighted sequences indi
cated a poor prognosis. The authors speculate that high-signal intensity ch
anges on T2-weighted images include a broad spectrum of compressive myeloma
lacic pathologies and reflect a broad spectrum of spinal cord recuperative
potentials. Predictors of surgical outcomes are preoperative signal intensi
ty change pattern of the spinal cord on radiologic evaluations, age at the
time of surgery, and chronicity of the disease.