Abstract
OBJECTIVES: Mycoplasma pneumoniae pneumonia (MPP), particularly macrolide-resistant MPP has undergone a prolonged nonseasonal epidemic in China since the lifting of non-pharmaceutical interventions in 2023. This study aimed to identify novel biomarkers to predict disease severity in children with MPP and to develop a predictive model. METHODS: In this study, bronchoalveolar lavage fluid (BALF) samples were collected from 30 children, including 15 with mild and 15 with severe MPP, for quantitative proteomic analysis. The two groups were compared and differentially expressed proteins (DEPs) were identified. Core proteins associated with MPP severity were identified using least absolute shrinkage and selection operator (LASSO) analysis. Logistic regression analysis was used to develop a predictive model. RESULTS: A total of 154 DEPs were identified, of which 57 were upregulated in the severe group. Upregulated signaling was found to be mainly involved in the immune response and inflammatory signaling. Thirteen proteins were selected as core proteins associated with MPP severity. CD209, CHM, PBRM1, and SCAMP1 were the most influential predictors and a predictive model using these four proteins predicted MPP severity. CONCLUSION: A predictive model was developed to assess the potential of using the identified biomarkers to predict disease severity. This model provides insights into the pathogenesis of M. pneumoniae infection. IMPORTANCE: Differences in the proteomic characteristics of bronchoalveolar lavage fluid in children with mild and severe Mycoplasma pneumoniae pneumonia (MPP) were identified, revealing the role of lung-specific immunologic and inflammatory response in the pathogenesis of MPP. Given the increasing incidence of severe MPP in children in recent years and the emergence of macrolide-resistant M. pneumoniae infection in some regions, our findings provided valuable knowledge and insights into the pathogenesis of M. pneumoniae infection.