Methods: A retrospective chart review of all patients who presented to a single surgeon over a seven-year period with diagnosis of a nerve tumor was undertaken. Patients were included only if a pre-operative MRI was obtained and the patient subsequently underwent operative exploration. The pre-operative MRIs were compared to the final pathology report for the accuracy of both the diagnosis and for the determination of benign versus malignant disease. Pre-operative symptoms were also analyzed to determine possible predictors of malignancy.

Results: 32 patients (18 male, 15 female) underwent 33 operations (one bilateral) on 34 nerves (one with involvement of 2 nerves on the same extremity). The mean age was 43.5 (range 14- 80), with a median age of 39. All patients with malignancy presented with symptoms of pain, while only 1 of the 4 had symptoms of severe pain and weakness. 72% (21/ 29) of patients with benign disease presented with pain and 24% (7/ 29) presented with severe pain. Only 7% (2/29) of patients in the benign nerve tumor group presented with weakness. Statistical analysis of all patients diagnosed with nerve tumors either on MRI or on final pathology revealed that MRI has a sensitivity of 83% and a positive predictive value of 89% when used to identify nerve tumors. When used to identify benign nerve tumors, MRI had a sensitivity of 93%, a specificity of 25%, a positive (benign) predictive value of 89% and a negative (malignant) predictive value of 33%. MRI correctly predicted malignancy in only 1 of 4 cases.

Conclusion: While MRI is useful as an aid in the evaluation and operative planning in patients with nerve tumors to define the nerve of origin and involvement of adjacent structures, diagnostic errors are common.