Diagnostic yield of brain biopsy in neurodegenerative disorders

The diagnostic yield and therapeutic implications of brain biopsy were determined in a series of 50 consecutive brain biopsies that were performed in 48 patients between 1990 and 1995 at The Johns Hopkins Hospital to assess progressive neurodegenerative disorders of unclear origin. Severely immunocompromised patients and patients undergoing biopsies for suspected neoplastic lesions were excluded from this analysis. Before surgery, the patients had undergone extensive laboratory and radiographic tests, including lumbar puncture (all 48 patients), electroencephalography (26 of 48 patients), magnetic resonance imaging (all 48 patients), and angiography (17 of 48 patients). Despite the results of these studies, diagnoses could not be established, and thus, brain biopsies were undertaken. Only 10 of the 50 biopsies (44 open procedures and 6 stereotactic procedures) led to diagnoses, resulting in a diagnostic yield of 20%. An additional three biopsies (6%) were only suggestive of diagnoses. The results of 33 biopsies (66%) were abnormal but nonspecific, and the results of 4 (8%) were normal. Minor complications associated with biopsy occurred in five cases (10%), and there were no deaths. Of the 10 patients whose biopsies were diagnostic, only 4 underwent meaningful therapeutic intervention as a result of the procedure, resulting in an overall therapeutic benefit in only 8% of all the cases. An analysis of patient subgroups to elucidate a correlation with diagnostic biopsy revealed that patients with focal magnetic resonance imaging findings had the highest likelihood of a diagnostic biopsy (odds ratio, 4.00). Electroencephalography and laboratory abnormalities were not predictive of a diagnostic biopsy. We conclude that the current diagnostic yield (20%) of brain biopsy for progressive neurodegenerative disorders is lower than that of earlier reports and that the therapeutic benefits of the procedure are limited.