Metagenomic detection of central nervous system infections missedby conventional testing

Abstract

Community-acquired infectious meningoencephalitis is associated with high rates of mortality and morbidity, compounded by limited access to diagnostic resources. The current study assessed acute central nervous system (CNS) infections in patients with meningoencephalitis enrolled in a hospital-based diagnostic surveillance study in São Paulo, Brazil. Cerebrospinal fluid (CSF) was collected from 600 patients between March 2018 and November 2019 and initially screened for a broad range of pathogens according to a local diagnostic algorithm. Standard microbiological and molecular diagnostic methods were applied. Metagenomic sequencing was used as a complementary approach to investigating etiology in instances where no pathogen was initially identified. Standard testing identified infectious etiologies in 292 patients (48.6%), with 227 (77.7%) confirmed as viral infections, predominantly caused by enteroviruses (n = 144) and herpesviruses (n = 40). Nonviral agents were identified in 65 patients (22.3%). Metagenomic sequencing (mNGS) of 277 of 308 undiagnosed patients revealed several additional potential etiologies, including Parvovirus B19, Toxoplasma gondii, Picobirnavirus, other enterovirus species and Vesivirus, the latter being associated with CNS infection for the first time. These findings underscore the complexity of CNS infections and highlight the potential of metagenomics to improve diagnostic accuracy, inform treatment strategies, and support efforts to address future pandemics.

Keywords: Clinical Research; Infectious disease; Molecular diagnosis.

Figure 1. Flowchart of the study protocol in patients with acute meningitis or meningoencephalitis.

The chart details patient enrollment and exclusion criteria, the distribution of diagnostic etiologies following analysis with standard microbiological tests and real-time PCR, and subsequent metagenomic analysis. PNBM, postneurosurgical bacterial meningitis; EM, epidemic meningitis; MTB, Mycobacterium tuberculosis; and T. pallidum, Treponema pallidum.

Figure 2. Distribution of the microorganisms detected by qPCR among 279 samples from patients with acute meningitis or meningoencephalitis.

(A) Distribution of known etiologies including bacteria, viruses, and other microorganisms among the samples. (B) Detailed percentage distribution of various viruses identified. (C) Distribution of bacterial pathogens identified. (D) Composition of other microorganisms found. Each segment of the charts corresponds to the proportion of each pathogen in relation to the total identified in each category.

Figure 3. Distribution of viral genera detected by mNGS in cerebrospinal fluid (CSF) samples.

Distribution of viral genera identified in cerebrospinal fluid (CSF) samples, based on reads per million (n = 103). Each genus is represented by different colors, indicating whether it represents a complete or near-complete genome (color) or fragments of genome (gray). The color coding for each genus is displayed on the right side of the histogram, showing the identity of the samples associated with each viral genus. No. max of RPM indicates the maximum number of reads per million (shown on the left), while the number of positive samples is indicated on the right.

Figure 4. Phylogenetic analysis and genomic fragment details of picobirnaviruses and vesiviruses.

(A) Maximum Likelihood tree for Picobirnavirus based on the RdRp region, with samples annotated and colored according to geographical location. The genome from this study is highlighted in red. (B) Maximum likelihood tree of vesiviruses based on a partial genome sequence, annotated with virus species colors, and with this study’s genome highlighted in red. (C) Detailed distribution of 12 genomic fragments of Vesivirus detected in cerebrospinal fluid samples, including key genomic details and sample history spanning from 1968 to 2014 across the USA and Japan.