Purpose: We assessed the performance of metagenomic next-generation sequencing (mNGS) in the diagnosis of infectious encephalitis and meningitis. Methods: This was a prospective multicenter study. Cerebrospinal fluid samples from patients with viral encephalitis and/or meningitis, tuberculous meningitis, bacterial meningitis, fungal meningitis, and non-central nervous system (CNS) infections were subjected to mNGS. Results: In total, 213 patients with infectious and non-infectious CNS diseases were finally enrolled from November 2016 to May 2019; the mNGS-positive detection rate of definite CNS infections was 57.0%. At a species-specific read number (SSRN) ≥2, mNGS performance in the diagnosis of definite viral encephalitis and/or meningitis was optimal (area under the curve [AUC] = 0.659, 95% confidence interval [CI] = 0.566–0.751); the positivity rate was 42.6%. At a genus-specific read number ≥1, mNGS performance in the diagnosis of tuberculous meningitis (definite or probable) was optimal (AUC=0.619, 95% CI=0.516–0.721); the positivity rate was 27.3%. At SSRNs ≥5 or 10, the diagnostic performance was optimal for definite bacterial meningitis (AUC=0.846, 95% CI = 0.711–0.981); the sensitivity was 73.3%. The sensitivities of mNGS (at SSRN ≥2) in the diagnosis of cryptococcal meningitis and cerebral aspergillosis were 76.92 and 80%, respectively. Conclusion: mNGS of cerebrospinal fluid effectively identifies pathogens causing infectious CNS diseases. mNGS should be used in conjunction with conventional microbiological testing. Trial Registration: Chinese Clinical Trial Registry, ChiCTR1800020442.

Introduction. In recent years, metagenomic next-generation sequencing (mNGS) has become widely used in medical microbiology to detect pathogen infection. Aim. We aimed to assess the diagnostic performance of mNGS of cerebrospinal fluid (CSF) for prediction of cryptococcal meningitis (CM). Methodology. A comparative evaluation of mNGS (performed on CSF samples) and conventional methods, including India ink staining, culture for fungi and cryptococcal-antigen (CrAg) detection by enzyme immunoassay, was performed on 12 consecutive non-HIV-infected patients with chronic or subacute CM. Results. India ink staining and culture of the CSF were positive for Cryptococcus in 83.33 % (10/12) of the samples; 100 % (11/11) were positive via CrAg EIA. The mNGS results of the CSF identified DNA sequences corresponding to Cryptococcus in 75 % of samples (9/12). However, the DNA of both C. neoformans s.l. and C. gattii s.l. was detected concurrently in 33.33 % (4/12). Conclusion. mNGS is helpful for identifying Cryptococcus species. The application of mNGS, together with India ink staining, culture methods, and CrAg, may significantly improve the diagnostic precision in CM, thereby informing choice of appropriate antifungal treatment courses.