Abstract / Summary
Viral encephalitis requires rapid etiological assessment, yet targeted polymerase chain reaction, serology, culture, and multiplex panels leave many cases unresolved because they depend on prespecified pathogens, sampling time, and specimen biology. Metagenomic next-generation sequencing (mNGS) can interrogate DNA and RNA in cerebrospinal fluid without a predefined target and may detect unexpected, rare, opportunistic, or emerging viruses. Its clinical performance, however, is not uniform. In published cohorts, diagnostic yield varies with case selection, reference standards, pathogen burden, specimen volume, pretreatment, laboratory workflow, sequencing depth, bioinformatic thresholds, and contamination control. Short-read Illumina workflows currently have the largest clinical evidence base, whereas real-time nanopore sequencing offers speed and portability but still requires site-specific validation. The strongest role for mNGS is therefore not routine first-line replacement of polymerase chain reaction or serology, but early complementary use in severe, progressive, unexplained, atypical, or immunocompromised cases after urgent targeted testing and empirical treatment have begun. A positive result requires assessment of read distribution, genome coverage, controls, clinical plausibility, and orthogonal confirmation; a negative result cannot exclude a low-titer, serology-defined, or tissue-restricted infection. This minireview critically compares the available evidence, identifies causes of between-study variation, and proposes a practical decision pathway for integrating mNGS into the diagnosis of viral encephalitis.