We present the case of a 15-year-old boy with symptoms due to Klippel–Feil syndrome. Radiographs and CT scans demonstrated basilar impression, occipitalisation of C1 and fusion of C2/C3. MRI showed ventral compression of the medullocervical junction. Skull traction was undertaken pre-operatively to determine whether the basilar impression could be safely reduced. During traction, the C3/C4 junction migrated 12 mm caudally and spasticity resolved. Peri-operative skull-femoral traction enabled posterior occipitocervical fixation without decompression. Following surgery, cervical alignment was restored and spasticity remained absent. One year after surgery he was not limited in his activities. The surgical strategy for patients with basilar impression and congenital anomalies remains controversial. The anterior approach with decompression is often recommended for patients with ventral compression of the medullocervical region, but such procedures are technically demanding and carry a significant risk of complications. Our surgical strategy was an alternative solution. Prior to a posterior cervical fixation, without decompression, skull traction was used to confirm that the deformity was reducible and effective in resolving associated myelopathy.

Signal intensity changes of the spinal cord on MRI in chronic cervical myelopathy are thought to be indicative of the prognosis. However, the prognostic significance of signal intensity change remains controversial.

The purpose of this study was to investigate the characteristics of MR findings in cervical compression myelopathy that reflect the clinical symptoms and the prognosis and to determine the radiographical and clinical factors that correlate to the prognosis.

The subjects were 73 patients who underwent cervical expansive laminoplasty. Their mean age was 64 years, and the mean postoperative follow-up was 3.4 years. The pathological conditions were cervical spondylotic myelopathy in 42 and ossification of the posterior longitudinal ligament in 31.MRI (spin-echo sequence) was performed in all patients. Three patterns of spinal cord signal intensity changes on T1-weighted sequences/T2-weighted sequences were detected as follows: normal/ normal.

(N/N); normal/ high signal intensity changes (N/Hi); and low signal intensity changes/high signal intensity changes (Lo/Hi). Surgical outcomes were compared among these three groups. The most useful combination of parameters for predicting prognosis was determined.

There were 2 patients with N/N, 67 with N/Hi and 4 with Lo/Hi signal change patterns before surgery. Regarding postoperative recovery, the preoperative Lo/Hi group was significantly inferior to the preoperative N/Hi group. The best combination of predictors for surgical outcomes included age, preoperative signal pattern and duration of symptoms.

The low signal intensity changes on T1-weighted sequences indicated a poor prognosis. We speculate that high signal intensity changes on T2-weighted images include a broad spectrum of compressive myelomalacid pathologies and reflect a broad spectrum of recuperative potentials of the spinal cord. Predictors for surgical outcomes are preoperative signal intensity change pattern of the spinal cord on radiological evaluations, age at the time of surgery and chronicity of the disease.