Genetic Heterogeneity and Tissue-specific Patterns of Tumors with Multiple PIK3CA Mutations

Purpose: To comprehensively characterize tissue-specific and molecular subclasses of multiple PIK3CA (multi-PIK3CA) mutations and assess their impact on potential therapeutic outcomes. Experimental Design: We profiled a pan-cancer cohort com-prised of 352,392 samples across 66 tumor types using a targeted hybrid capture-based next-generation sequencing panel covering at least 324 cancer-related genes. Molecularly defined subgroups, allelic configuration, clonality, and mutational signatures were identified and tested for association with PI3K inhibitor therapeutic response. Results: Multi-PIK3CA mutations are found in 11% of all PIK3CA-mutant tumors, including 9% of low tumor mutational burden (TMB) PIK3CA-mutant tumors, and are enriched in breast and gynecologic cancers. Multi-PIK3CA mutations are frequently clonal and in cis on the same allele and occur at characteristic positions across tumor types. These mutations tend to be mutually exclusive of mutations in other driver genes, and of genes in the PI3K pathway. Among PIK3CA-mutant tumors with a high TMB, 18% are multi-PIK3CA mutant and often harbor an apolipoprotein B mRNA-editing enzyme, catalytic polypeptide (APOBEC) muta-tional signature. Despite large differences in specific allele combi-nations comprising multi-PIK3CA mutant tumors, especially across cancer types, patients with different classes of multi-PIK3CA mutant estrogen receptor-positive, HER2-negative breast cancers respond similarly to PI3K inhibition.Conclusions: Our pan-tumor study provides biological insights into the genetic heterogeneity and tissue specificities of multi-PIK3CA mutations, with potential clinical utility to guide PI3K inhibition strategies.