Pathway-level mutational signatures predict breast cancer outcomes and reveal therapeutic targets.

Background and purpose: In order to significantly improve the therapeutic treatment of breast cancer, the exploration of underlying genetic and molecular differences is absolutely necessary. Here, our goal was to integrate mutational status of entire pathways to reveal molecular pathway interactions determining survival.

Experimental approach: A comprehensive analysis of breast cancer mutations was conducted by integrating data from three distinct databases with a total of 4586 samples encompassing over 25,000 genes. For each gene, we filtered mutations that disruptively affect the protein structure. Cox proportional hazard regression was employed to link altered pathways to outcome. We also identified the co-occurring and mutually exclusive disruptive mutations.

Key results: We identified 17 genes, the mutation status of which alone seriously affects relapse-free survival. The three most significant genes were TP53 (HR: 2.04, p: 4.65 × 10^-33), CARD11 (HR: 2.59, p: 1.54 × 10^-5) and PIK3R1 (HR: 2.27, p: 3.66 × 10^-5). The five most significant biological processes and KEGG pathways affecting relapse-free survival include negative regulation of cell population proliferation, positive regulation of DNA-templated transcription, protein stabilisation, and MicroRNAs in cancer, hepatocellular carcinoma, and breast cancer. Co-mutation and mutual exclusivity analysis identified significant enrichment in 241 gene pairs. Finally, we also established an online platform to enable future analysis of the established cohort for any selected pathway.

Conclusions and implications: We assembled a comprehensive database of breast cancer samples and used this cohort to identify cancer-specific disruptive mutation signatures linked to altered survival outcomes.