Characterizing the genomic landscape of PIK3CA alterations from 121,221 adult patients with cancer: The next tissue-agnostic target?

Abstract

3088 Background: The PI3K–AKT–mTOR signaling pathway is frequently dysregulated in cancer and small molecule inhibitors targeting various nodes in the pathway have been pursued for decades. Activating mutations in PIK3CA are recognized potent drivers of oncogenesis, though the landscape of PIK3CA fusions and amplifications has yet to be well-defined. The field has been hampered by issues such as resistance and poor tolerance, however several isoform-specific PI3K inhibitors have now received regulatory approval and allosteric pan-mutant and specific mutant selective Inhibitors of PI3Kα are being investigated in early phase clinical trials. Here, we present a comprehensive analysis of PIK3CA alterations in pan-cancer adult malignancies. Methods: We analyzed 136096 samples from 121221 patients available from AACR Project GENIE v.11 database for the prevalence of PIK3CA mutations, fusions and copy number alterations in a range of cancer types. Results: 15712 alterations in PIK3CA were detected in profiled tumor samples (12%), in 14774 of patients profiled (12%), most frequently in endometrial (44% of 4135 cases), anal (35% of 286 cases), breast (35% of 14218 cases), cervical (27% of 761 cases) and colorectal cancer (20% of 3743 cases). PIK3CA alterations identified included 17740 mutations (11%) (including duplicate mutations in patients with multiple samples). 16685 missense mutations were identified (12%);14953 were identified as missense in driver mutation (likely oncogenic)(90% of missense); 1,732 were missense of variants of uncertain significance (VUS) using OncoKB database (11%). Missense mutations most frequently involved codons 545(3347, 20% of missense), codon 1047 (3644, 22%) and codon 542 (1881, 11%). 237 truncating mutations were identified (0.2%), 442 in-frame deletions, 29 in-frame insertion mutations. PIK3CA fusions were observed in 0.06% of tumor samples and were most identified in breast, colorectal, lung, GBM and ovarian cancer (24%, 17%, 13%, 7% and 6% of identified PIK3CA fusions respectively). PIK3CA fusions were most commonly intragenic fusions(36%);common fusion partners included TBL1XR1, FNDC3B and NAALADL2(17%, 7%, 5% of identified PIK3CA fusions, respectively). PIK3CA fusions were identified as VUS, aside from KMT2C-PIK3CA (2%) . PIK3CA amplification high level gain occurred in 0.5% of samples tested(662), deletion occurred in 0.16%(21). Conclusions: Activating PIK3CA mutations occur frequently across cancer types and could be considered for tissue-agnostic drug development. PIK3CA fusion and amplification events are extremely rare. Most PIK3CA missense mutation variants are described as oncogenic, while fusions are described as VUS, which may limit the impact of precision oncology for patients with this alteration. Further functional characterization of PIK3CA variants and basket trial enrollment are warranted.