Alpelisib (ALP) + endocrine therapy (ET) in patients (pts) with hormone receptor-positive (HR+), human epidermal growth factor receptor 2-negative (HER2–), PIK3CA-mutated (mut) advanced breast cancer (ABC): Baseline biomarker analysis and progression-free survival (PFS) by duration of prior cyclin-dependent kinase 4/6 inhibitor (CDK4/6i) therapy in the BYLieve study.

Abstract

1018 Background: ALP (PI3K-α selective inhibitor and degrader) + fulvestrant (FUL) is approved for pts with HR+, HER2– ABC and a tumor mutation in PIK3CA (̃ 40% of these pts). Primary analyses from the Phase 2 BYLieve study demonstrated efficacy and safety of ALP + ET in pts with PIK3CA-mut, HR+, HER2– ABC in the post-CDK4/6i setting. Post hoc analyses, including pts with disease progression within 6 mo of CDK4/6i + ET treatment (Tx), confirmed ALP benefit regardless of duration of prior CDK4/6i. Here we assess baseline biomarkers in circulating tumor DNA (ctDNA) by duration of prior CDK4/6i Tx and PFS in pts from BYLieve Cohorts A and B. Methods: In the BYLieve study, pts with PIK3CA-mut, HR+, HER2– ABC had CDK4/6i + aromatase inhibitor (Cohort A) or + FUL (Cohort B) as immediate prior Tx to receiving ALP + FUL and ALP + letrozole (LET), respectively. At data cutoff dates, pts had ≥ 18-mo follow-up in Cohort A and ≥ 6-mo in Cohort B. In each cohort, pts were grouped based on duration of prior CDK4/6i Tx (≤ 6 mo or >6 mo). Alterations were detected on ctDNA using next-generation sequencing (PanCancer V2 Panel). PFS was assessed in each cohort and by duration of prior CDK4/6i Tx. Results: Of 127 and 126 pts enrolled in Cohorts A and B, respectively, 98 (≤ 6-mo: 24; >6-mo: 74) and 94 (≤ 6-mo: 28; >6-mo: 66) were included in this analysis based on availability of ctDNA samples, data on duration of prior CDK4/6i, and centrally confirmed PIK3CA-mut disease. In this population, median (m) PFS (95% CI) was 8.2 mo (5.6 - 9.5) and 5.6 mo (3.7 - 7.1) in Cohorts A and B, respectively. In Cohort A, mPFS (95% CI) was 12.0 mo (5.5-non estimable) and 6.2 mo (5.4 - 8.5) in the ≤ 6-mo and >6-mo groups, respectively. The OncoPrint genomic profiles showed that pts in the ≤ 6-mo vs >6-mo group had a lower median ctDNA fraction and fewer detected gene alterations, including in genes associated with ET and/or CDK4/6i resistance, and fewer chromosomes 8/11 amplifications (linked to early relapse). In Cohort B, mPFS was 5.9 mo (3.5 - 11.0) and 5.6 mo (3.7 - 7.1) in the ≤ 6-mo and >6-mo groups, respectively. Both groups had high median ctDNA fractions and complex tumor mutation profiles reflecting more extensive treatment history. Conclusions: Lower median ctDNA fraction and lower mutational complexity observed in Cohort A ≤ 6-mo vs >6-mo group was associated with numerically longer mPFS, potentially indicating increased dependence on the mutant PI3K-α. In Cohort B, both ≤ 6-mo and >6-mo groups had high median ctDNA fractions and similar tumor mutation profiles. Additional ctDNA and tissue analyses are needed to elucidate the correlation between ALP + ET efficacy and treatment timing and baseline genomic complexity.