Metagenomics reveals pathogenic diversity and temporal dynamics in severe pneumonia among patients in adult intensive care unit

Metagenomic next-generation sequencing (mNGS) emerging as a standout in the clinical setting. In this study, we harnessed the power of mNGS to explore the pathogenic spectrum and temporal variations in respiratory tract specimens collected from adult patients with severe pneumonia who were admitted to the Intensive Care Units (ICUs) of two hospitals in Guangxi, China. From December 2021 to July 2022, 44 respiratory tract samples (including sputum and bronchoalveolar lavage fluid) from 25 adult patients (comprising 18 males and 7 females) diagnosed with severe pneumonia and admitted to the ICUs of two hospitals in Guangxi. A customized mNGS detection protocol was developed and applied for analyzing the composition and temporal variations of pathogens within the respiratory tract samples. Among these patients, the bacteria, fungi, and viruses were markedly higher detected by mNGS compared to conventional microbial culture methods (P < 0.001). The most prevalent bacteria detected were Stenotrophomonas maltophilia (61.36%), Corynebacterium striatum (54.55%), and Escherichia coli (54.55%). The viruses with the highest detection rates were human herpesviruses(HSV-1, 31.82%;HCMV, 27.27%;HSV-2, 11.36%). The most frequently identified fungi were Candida albicans (50%) and Nakaseomyces glabratus (27.27%). Single-pathogen infections accounted for 64% (28/44) of the cases, while mixed-pathogen infections comprised 36% (16/44). Dynamic monitoring using mNGS in 8 patients uncovered diverse respiratory pathogenic spectra, with the majo Candida glabratarity of patients exhibiting dynamic changes that correlated with fluctuations in inflammatory markers such as leukocyte counts, procalcitonin levels, and C-reactive protein levels, alongside the clinical progression of the disease. mNGS exhibits superior performance in diagnosing mixed infections and real-time tracking of the pathogen spectrum, which provide a robust empirical basis for guiding clinical diagnosis and treatment strategies of patients in ICU. Not applicable.