Abstract
BACKGROUND: Pediatric acute myeloid leukemia (AML) is characterized by poor prognosis and low survival rates following recurrence. While mitochondria and programmed cell death (PCD) are implicated in various diseases, their role in pediatric AML remains poorly understood. Identifying prognostic genes associated with PCD and mitochondrial function could enhance therapeutic approaches. METHODS: Transcriptomic data and gene sets were sourced from public databases. Differentially expressed genes (DEGs) that intersected with PCD-related genes (PCD-RGs) and mitochondrial-related genes (mito-RGs) were selected as candidate genes. Regression analyses were performed to identify prognostic genes, which were then used to develop and validate a prognostic model. A nomogram was constructed, followed by functional analysis, immune microenvironment assessment, molecular regulatory network investigation, drug sensitivity profiling, and clinical validation through RT-qPCR. RESULTS: Twenty-six candidate genes were identified, with three-PDHA1, OGG1, and OPA1-confirmed as potential prognostic markers through regression analyses. The prognostic model demonstrated robustness in both internal and external validations, and the nomogram exhibited good predictive power. Pathway enrichment analysis highlighted the involvement of DNA replication and epithelial-mesenchymal transition, alongside 14 differentially abundant immune cells (p < 0.05). Molecular network analysis indicated that hsa-miR-199a-5p regulates PDHA1 and OGG1. Drug sensitivity profiling identified potential therapeutic agents, including SB505124_1194. RT-qPCR validation confirmed consistent expression patterns for the prognostic genes. CONCLUSIONS: PDHA1, OGG1, and OPA1 were identified as potential prognostic markers for pediatric AML, providing valuable insights for the development of targeted therapeutic strategies. However, further validation in larger and more diverse clinical cohorts is still required to confirm its clinical applicability.