Diagnostic yield of next-generation sequencing applied to neurological disorders.

The exponential knowledge on the genetic etiology and the trend towards genetically-specific therapies for previously untreatable disorders, requires neurologists to be familiar with the strengths and weaknesses of Next-Generation Sequencing (NGS). Our aim was to assess the diagnostic yield of NGS studies in clinical practice in our setting. We performed a retrospective, cross-sectional, 18 months long study, from a single Portuguese center, of consecutive neurological patients for whom a NGS study was requested. A diagnosis rate (DR) of 33.2% was achieved for a total of 190 patients (89 children). It was higher for muscle diseases (DR 61.1%). In 20%, an inconclusive molecular diagnosis was obtained. The rate of incidental findings (IF) was 5.3%. We found better DR for clinical exome (52.6%, p < 0.05) although only 14% of patients were characterized using this approach. The performance of gene panels for muscle diseases was better but not statistically significant (DR 56.3% vs. 31.7% overall, p > 0.05). The reduced number of patients in several phenotypic groups limits the interpretation of specific diagnostic yields. The better yield of gene panels for muscle diseases suggests that gene panels may be a more cost-effective first-line test in well-defined phenotypes. For heterogeneous phenotypes and overall, WES-based virtual panels or clinical exome should be favored. We present daily practice evidence that, with the constraints of our health system, for one third of the patients with neurological disorders of undetermined etiology a definitive diagnosis can be reached with NGS.