Neurological diseases of unknown etiology: Brain-biopsy diagnostic yields and safety

Background: For nonneoplastic neurological diseases, no recommendation exists regarding the place or appropriate timing of brain biopsy. The aim of this study was to evaluate the diagnostic yield and safety of brain biopsies from patients with neurological diseases of unknown etiology. Methods: We performed a retrospective cohort study from January 1, 2008 to December 31, 2018. We analyzed 1847 brain-biopsied patients, including 178 biopsies indicated for neurological diseases of unknown etiology. Specific histological and final diagnosis rates, positive diagnosis-associated factors, complication rate and complication-associated factors were assessed. Results: Specific histological diagnosis and final diagnosis rates were 71.3% and 83.1%, respectively, leading to therapeutic management change(s) for 75.3% of patients. Brain- biopsy-related mortality and permanent neurological morbidity occurred in 1.1% and 0.6% of the patients, respectively. The multivariable logistic-regression model retained (odds ratio [95% CI] only immunodepression (2.2 [1.1-4.7]; P = .04) as being independently associated with specific histological diagnosis, while supratentorial biopsy-targeted lesions (4.1 [1.1-15.2]; P = .04) were independently associated with a final diagnosis. Biopsies obtained from comatose patients were less contributive to the diagnosis (0.2 [0.05-0.7]; P = .01). Prebiopsy platelet count <100 G/L (28.5 [1.8-447]; P = .02), hydrocephalus (6.3 [1.2-15.3]; P = .02) and targeted lesions <1 cm (4.3 [1.2-15.3]; P = .03) were independently associated with brain biopsy-related complications. Conclusion: For highly selected patients with neurological diseases of unknown etiology, brain biopsy has a high diagnostic yield and low frequency of severe complications. We advocate that this procedure be considered early in the diagnosis algorithm of these patients.