Abstract
BACKGROUND: There is a strong correlation between lactylation, programmed cell death, and the progression of cancer. This study aims to identify prognostic genes associated with lactylation and programmed cell death in pancreatic ductal adenocarcinoma (PDAC), providing new insights for risk stratification and therapeutic strategies. METHODS: TCGA-PAAD, GSE62452, lactylation-related genes (LRGs), and programmed cell death-related genes (PCDRGs) were retrieved from relevant databases and references. Prognostic genes were identified through univariate Cox regression analysis, followed by random survival forest analysis for survival prediction. Subsequently, enrichment analysis, immune microenvironment analysis, drug sensitivity analysis, immunohistochemical analysis, and expression analysis of prognostic genes were conducted. Finally, the experimental verification was carried out in clinical samples. RESULTS: In this investigation, two prognostic genes (HMGA1 and KIF2C) linked to lactylation and programmed cell death were identified, and a robust prognostic risk model was developed. Enrichment analysis results included Cell cycle, G2M checkpoint, Myogenesis, and Angiogenesis. Moreover, immature B cells and activated B cells demonstrated the strongest positive correlation (cor = 0.97, P < 0.001), while neutrophils and activated B cells demonstrated the strongest negative correlation (cor = -0.68, P < 0.001). Furthermore, KIF2C and HMGA1 demonstrated the strongest negative relationships with mast cells (correlation coefficients = -0.36 and -0.53, P < 0.01). Drug sensitivity analysis revealed that Sapitinib was more effective in the high-risk group (HRG), while Doramapimod was more effective in the low-risk group (LRG) (P < 0.0001). Both immunohistochemical and expression analyses of prognostic genes showed that HMGA1 and KIF2C were upregulated in PDAC patients (P < 0.05). Finally, genes in the clinical samples also showed the same expression trend. CONCLUSION: In the present investigation, two prognostic genes were identified, and subsequently, a predictive risk model was established, which may serve as a valuable reference for the clinical management of PDAC.