Predicting Lymphoma Prognosis Using Machine Learning-Based Genes Associated with Lactylation

Background: Lactylation, a newly discovered PTM involving lactic acid, is linked to solid tumor proliferation and metastasis. Lymphoma patients exhibit high lactic acid levels, yet lactylation's role in lymphoma is underexplored. This study aimed to identify lactylation-related genes in lymphoma using tumor databases and assess their predictive value in patient prognosis through cell experiments and clinical specimens. Methods: Using TCGA and GEO datasets, we analyzed the expression levels of lactylation-related genes in diffuse large B-cell lymphoma patients. We also evaluated the prognostic significance of lactylation gene risk scores, exploring their impact on drug sensitivity and tumor immune function. Key lactylation-affecting genes were identified and functionally validated through cell experiments and mouse in vivo experiments. Additionally, the relationship between lactylation and lymphoma prognosis was examined in clinical specimens. Results: We identified 70 genes linked to diffuse large B-cell lymphoma prognosis from the lactylation-related gene set. Using clinical data and a COX regression algorithm, we developed an optimized lactylation Riskscore model. This model significantly correlated with prognosis and showed differences in immune cell infiltration, particularly macrophages. High-risk patients showed resistance to chemotherapy drugs but responded well to immunotherapy. HNRNPH1, a lactylation-related gene, influenced patient prognosis, apoptosis, cell cycle distribution in lymphoma cells, and tumor volume in mice. In lymphoma specimens, lactylation levels correlated with Bcl-2, C-myc, and P53 levels. Conclusions: Lactylation impacts diffuse large B-cell lymphoma prognosis, tumor immune function, and drug resistance. Our lactylation-based Riskscore model aids in patient stratification and treatment selection. HNRNPH1 regulates lactylation, thereby affecting patient prognosis.