Yuecui Li,1,* Chenghang Li,1,* Yu Fang,2,* LiLi Zhang,1 Xiaoyan Ying,1 Ruotong Ren,3 Yinghui Zang,1 Dandan Ying,1 Shengwei Zhu,1 Jiao Liu,1 Xuefang Cao3 

1Department of Infectious Disease, The First People’s Hospital of Yongkang, affiliated to Hangzhou Medical College, Jinhua, People’s Republic of China; 2Zhejiang Key Laboratory of Digital Technology in Medical Diagnostics, Hangzhou, People’s Republic of China; 3MatriDx Biotechnology Co., Ltd, Hangzhou, People’s Republic of China

*These authors contributed equally to this work

Correspondence: Yuecui Li, Department of Infectious Disease, The First People’s Hospital of Yongkang, affiliated to Hangzhou Medical College, 599 Jinshan West Road, Jinhua, 321300, People’s Republic of China, Email yklycwh@126.com Xuefang Cao, MatriDx Biotechnology Co., Ltd, Hangzhou, 311112, People’s Republic of China, Email xfczju@163.com

Background: This study aimed to investigate the co-infecting pathogens and lung microbiomes in patients with clinically confirmed pulmonary tuberculosis (TB) and explore potential diagnostic biomarkers to differentiate between varied infection patterns.
Methods: We conducted a retrospective cohort study by analyzing 198 bronchoalveolar lavage fluid (BALF) samples collected from patients with suspected pulmonary TB. All BALF samples were sequenced using metagenomic next-generation sequencing (mNGS).
Results: A total of 63 pathogens were detected in all samples. The TB group exhibited a higher diversity of pathogens (n=51) than the Non-TB group (n=37). The analysis revealed that TB patients had significantly higher pathogen counts (P=0.014), and specific microorganisms, such as Mycobacterium tuberculosis complex (MTBC), MTB, Streptococcus infantis, and Campylobacter curvus, were significantly enriched. Furthermore, the abundance of MTBC was negatively correlated with hemoglobin levels (R=− 0.17, P=0.015) and positive correlated with C-reactive protein (CRP) levels (R=0.16, P=0.029). The random forest model combined eight differential microbes and five clinical parameters, yielding an area under the curve (AUC) of 0.86 for differentiating TB from Non-TB cohorts, whereas subgroup differentiation yielded an AUC of 0.571, demonstrating the potential for targeted diagnostics in pulmonary infections.
Conclusion: Our findings highlight the complexity of co-infection patterns in pulmonary TB and emphasize the potential of integrating microbial and clinical markers to improve diagnostic accuracy. This study provides valuable insights into the role of the lung microbiome in TB and informs future research on targeted therapies for this disease.

Keywords: tuberculosis, bronchoalveolar lavage fluid, lung microbiome, biomarkers, metagenomic next-generation sequencing