Metagenomic next-generation sequencing for lung cancer low respiratory tract infections diagnosis and characterizing microbiome features.

Abstract

Background: The capability of mNGS in diagnosing suspected LRTIs and characterizing the respiratory microbiome in lung cancer patients requires further evaluation.

Methods: This study evaluated mNGS diagnostic performance and utilized background microbial sequences to characterize LRT microbiome in these patients. GSVA was used to analyze the potential functions of identified genera.

Results: Bacteria were the most common pathogens (n=74) in LRTIs of lung cancer patients, and polymicrobial infections predominated compared to monomicrobial infections (p<0.001). In diagnosing LRTIs in lung cancer patients, the pathogen detection rate of mNGS (83.3%, 70/84) was significantly higher than that of sputum culture (34.5%, 29/84) (p<0.001). This result was consistent with that of non-lung cancer patients (p<0.001). Furthermore, in the specific detection of bacteria (95.7% vs. 22.6%) and fungi (96.0% vs. 22.2%), the detection rate of mNGS was also significantly higher than that of CMTs mainly based on culture (p<0.001, p<0.001). However, in the detection of CMV/EBV viruses, there was no significant difference between the detection rate of mNGS and that of viral DNA quantification (p = 1.000 and 0.152). mNGS analysis revealed Prevotella, Streptococcus, Veillonella, Rothia, and Capnocytophaga as the most prevalent genera in the LRT of lung cancer patients. GSVA revealed significant correlations between these genera and tumor metabolic pathways as well as various signaling pathways including PI3K, Hippo, and p53.

Conclusion: mNGS showed a higher pathogen detection rate than culture-based CMTs in lung cancer patients with LRTIs, and also characterizing LRT microbiome composition and revealing potential microbial functions linked to lung carcinogenesis.