Abstract
OBJECTIVE: In clinical practice, imaging manifestations of diffuse lung parenchymal lesions are common and indicative of various diseases, making differential diagnosis difficult. Some of these lesions are eventually diagnosed as lung cancer. METHODS: Because respiratory microorganisms play an important role in lung cancer development, we searched for microbial markers that could predict the risk of lung cancer by retrospectively analyzing the lower respiratory tract (LRT) microbiome of 158 patients who were hospitalized in the First Affiliated Hospital of Guangzhou Medical University (March 2021-March 2023) with diffuse lung parenchymal lesions. The final diagnosis was lung cancer in 21 cases, lung infection in 93 cases, and other conditions (other than malignancy and infections) in 44 cases. The patient's clinical characteristics and the results of metagenomic next-generation sequencing of bronchoalveolar lavage fluid (BALF) were analyzed. RESULTS: Body mass index (BMI) and LRT microbial diversity (Shannon, Simpson, species richness, and Choa1 index) were significantly lower (P< 0.001, respectively) and Lactobacillus acidophilus relative abundance in the LRT was significantly higher (P< 0.001) in patients with lung cancer. The relative abundance of L. acidophilus in BALF combined with BMI was a good predictor of lung cancer risk (area under the curve = 0.985, accuracy = 98.46%, sensitivity = 95.24%, and specificity = 100.00%; P< 0.001). CONCLUSION: Our study showed that an imbalance in the component ratio of the microbial community, diminished microbial diversity, and the presence of specific microbial markers in the LRT predicted lung cancer risk in patients with imaging manifestations of diffuse lung parenchymal lesions.