• Aaron A Cohen-Gadol, Ofer M Zikel, Gary M Miller, Allen J Aksamit, Bernd W Scheithauer, and William E Krauss.
• Department of Neurologic Surgery, Mayo Clinic and Mayo Foundation, Rochester, Minnesota, USA. cohengadol.aaron@mayo.edu
• Neurosurgery. 2003 Apr 1;52(4):806-15; discussion 815-6.

ObjectiveNeurosurgeons are frequently asked to evaluate patients for spinal cord biopsies when preoperative magnetic resonance imaging studies demonstrate nonspecific features. These lesions often appear unresectable, but surgeons must decide whether a biopsy is warranted. To determine the best approach to these cases, we evaluated the clinicopathological findings for patients with unknown spinal cord lesions who underwent spinal cord biopsies.MethodsThirty-eight consecutive patients who underwent spinal cord biopsies at the Mayo Clinic (Rochester, MN) between August 1988 and July 1998 were studied. A detailed review of the case histories, radiological results, surgical notes, histological findings, and outcomes was performed.ResultsSpinal cord biopsies were performed for 21 male and 17 female patients (mean age, 42.1 yr) with progressive neurological deficits related to spinal cord lesions. All patients underwent preoperative magnetic resonance imaging evaluations. High T2-weighted signal intensity and spinal cord expansion were identified in 92 and 87% of cases, respectively. After gadolinium infusion, the majority (94%) of the inflammatory lesions demonstrated patchy and often peripherally situated enhancement. This neuroradiological pattern was less common for neoplasms (50%) and benign lesions (40%). The most common pathological findings were inflammatory changes of demyelination or sarcoidosis, which together accounted for 13 cases (34%). Nonspecific changes or benign lesions were observed in 10 cases (26%). Neoplasms were identified in eight cases (21%). One case of tuberculosis and one of schistosomiasis were found. Overall, 47% of the preoperative diagnoses made by the attending surgeon were correct. For 26% of the patients, specific treatment was based on the biopsy results. The average follow-up period was 12 months (standard deviation, 14 mo; range, 0-50 mo).ConclusionPreoperative laboratory and imaging studies are often diagnostically inconclusive in cases of spinal cord lesions with nonspecific features. Biopsies should be considered for patients with progressive symptomatic lesions.

34 keywords...

hide…