Identification and characterization of lLF2 as a prognostic biomarker in HER2-positive breast cancer using Mendelian randomization and machine learning.

BACKGROUND: HER2-positive breast cancer is an aggressive molecular subtype characterized by high recurrence rates and poor prognosis. Identifying robust biomarkers for prognosis and therapeutic response is essential to improve individualized treatment strategies. METHODS: Based on genome-wide association study (GWAS) data, summary data-based Mendelian randomization (SMR) analysis was used to screen disease-related genes. The diagnostic model of HER2-positive breast cancer was constructed by combining LASSO, random forest and SVM-RFE and the characteristic genes were identified. The prognostic correlation was evaluated through Kaplan-Meier analysis. The SNP association, expression level, immune activity and drug sensitivity of ILF2 were analyzed emphatically. H&E staining and immunohistochemistry were used to verify the expression of ILF2 in tumor tissues. RESULTS: A total of 14 genes associated with HER2-positive breast cancer were identified through SMR analysis, with the most significant SNPs enriched on chromosome 1. Integrating three machine learning algorithms (LASSO, Random Forest, and SVM-RFE), four robust diagnostic feature genes were identified: SLC16A3, ILF2, ARRDC3, and SNAPIN. Among them, ILF2 emerged as a key gene of interest. Kaplan-Meier survival analysis revealed that high expression of ILF2, NEK10, LAMTOR5, and APOBEC3A was significantly associated with poorer prognosis. The top ILF2-associated SNP (rs10908848) showed a strong negative regulatory effect on disease risk. Further functional analyses indicated that high ILF2 expression was associated with suppressed antitumor immune activity and reduced sensitivity to multiple chemotherapeutic agents, suggesting its role in immune evasion and therapy resistance. Immunohistochemical staining confirmed significantly upregulated ILF2 protein expression in HER2-positive breast tumor tissues compared with adjacent normal tissues. CONCLUSION: ILF2 is biologically and clinically important in HER2-positive breast cancer, and its high expression is linked with tumor immune escape and increased resistance to chemotherapeutic agents. Future studies should further explore the mechanism and application of ILF2 as a potential therapeutic target.