A multicenter prospective study of comprehensive metagenomic and transcriptomic signatures for predicting outcomes of patients with severe community-acquired pneumonia.

BACKGROUND: Severe community-acquired pneumonia (SCAP) results in high mortality as well as massive economic burden worldwide, yet limited knowledge of the bio-signatures related to prognosis has hindered the improvement of clinical outcomes. Pathogen, microbes and host are three vital elements in inflammations and infections. This study aims to discover the specific and sensitive biomarkers to predict outcomes of SCAP patients. METHODS: In this study, we applied a combined metagenomic and transcriptomic screening approach to clinical specimens gathered from 275 SCAP patients of a multicentre, prospective study. FINDINGS: We found that 30-day mortality might be independent of pathogen category or microbial diversity, while significant difference in host gene expression pattern presented between 30-day mortality group and the survival group. Twelve outcome-related clinical characteristics were identified in our study. The underlying host response was evaluated and enrichment of genes related to cell activation, immune modulation, inflammatory and metabolism were identified. Notably, omics data, clinical features and parameters were integrated to develop a model with six signatures for predicting 30-day mortality, showing an AUC of 0.953 (95% CI: 0.92-0.98). INTERPRETATION: In summary, our study linked clinical characteristics and underlying multi-omics bio-signatures to the differential outcomes of patients with SCAP. The establishment of a comprehensive predictive model will be helpful for future improvement of treatment strategies and prognosis with SCAP. FUNDING: National Natural Science Foundation of China (No. 82161138018), Shanghai Municipal Key Clinical Specialty (shslczdzk02202), Shanghai Top-Priority Clinical Key Disciplines Construction Project (2017ZZ02014), Shanghai Key Laboratory of Emergency Prevention, Diagnosis and Treatment of Respiratory Infectious Diseases (20dz2261100).

Pseudorabies virus encephalitis in humans: a case series study.

Pseudorabies virus (PRV) is known to cause severe encephalitis in juvenile pigs and various non-native hosts; recent evidences suggest that PRV might cause encephalitis in humans. In a multicenter cohort study in China, next-generation sequencing of cerebrospinal fluid (CSF) was performed to detect pathogens in all patients with clinically suspected central nervous system infections. This study involved all the patients whose CSF samples were positive for PRV-DNA; their clinical features were evaluated, and species-specific PCR and serological tests were sequentially applied for validation. Among the 472 patients tested from June 1, 2016, to December 1, 2018, six were positive for PRV-DNA, which were partially validated by PCR and serological tests. Additionally, we retrospectively examined another case with similar clinical and neuroimaging appearance and detected the presence of PRV-DNA. These patients had similar clinical manifestations, including a rapid progression of panencephalitis, and similar neuroimaging features of symmetric lesions in the basal ganglia and bilateral hemispheres. Six of the patients were engaged in occupations connected with swine production. PRV infection should be suspected in patients with rapidly progressive panencephalitis and characteristic neuroimaging features, especially with exposure to swine.

Next-generation sequencing of the cerebrospinal fluid in the diagnosis of neurobrucellosis.

BACKGROUND: Brucellosis is the most common zoonotic infection in the world. Brucellosis with nervous system involvement is known as 'neurobrucellosis' (NB). The diagnosis of NB is difficult because its clinical manifestations are non-specific and the sensitivity of routine culture tests is low. METHODS: Next-generation sequencing (NGS) of cerebrospinal fluid (CSF) was used to detect pathogens in patients with clinically suspected central nervous system (CNS) infections at a tertiary referral center in China between June 1, 2016 and June 1, 2017. The clinical characteristics and NGS results of patients with the diagnosis of NB were reviewed in this study. RESULTS: Four patients were rapidly diagnosed with NB using NGS of the CSF in patients with clinically suspected CNS infections, although the clinical manifestations varied dramatically between these patients. NGS of the CSF revealed that the sequence reads identified that corresponded to Brucella species ranged from 11 to 104, with genomic coverage ranging from 0.043% to 0.4%. Rapid diagnosis led to prompt treatment with the appropriate antibiotics. CONCLUSIONS: This study demonstrates the power of NGS of the CSF coupled with a bioinformatic pipeline in the diagnosis of NB.