Treatment patterns in HR+/HER2- metastatic breast cancer (MBC) with co-occurring <i>PIK3CA</i> and <i>ESR1</i> mutations.

Abstract

1082 Background: PIK3CA and ESR1 mutations co-occur in 12-15% of patients with HR+/HER2- MBC. While FDA-approved PI3K inhibitors (PI3Ki) and selective estrogen receptor degraders (SERDs) have advanced care for patients with biomarkers, treatment outcomes and optimal sequencing in those with co-occurring mutations remain unclear. Methods: We conducted a retrospective analysis of patients with HR+/HER2- MBC and co-occurring PIK3CA and ESR1 mutations treated at Memorial Sloan Kettering (MSK) between 2010 and 2024. Mutations were identified via MSK-IMPACT (tumor tissue) and MSK-ACCESS (ctDNA), with additional genomic data integrated from Guardant and Foundation One. Clinical and treatment data, including PI3Ki (alpelisib, inavolisib or investigational agents) and/or SERDs (fulvestrant or oral agents) were abstracted. Median progression-free survival (mPFS) was assessed in patients with mutations identified prior to therapy initiation. Results: 3,166 patients with HR+/HER2- MBC were identified, including 1,444 (46%) with PIK3CA mutations, 664 (21%) with ESR1 mutations and 243 (8%) with co-occurring mutations. After excluding 26 patients with incomplete records or concurrent malignancies, the final cohort included 217 patients (7%), with 206 having MSK-IMPACT, 58 MSK-ACCESS, 49 Guardant, and 3 FoundationOne data. Of 217 patients, 77 (36%) received a PI3Ki after a median of 4 prior lines (range: 1–16), with 68 (88%) having prior CDK 4/6i. Single-agent SERD was administered to 46 patients (21%) after a median of 3 prior lines (range: 2–18), including 34 (74%) with prior CDK 4/6i. 8 patients received PI3Ki followed by SERD, and 7 received SERD followed by PI3Ki, either consecutively or with intervening treatments. The mPFS was 7.1 m with PI3Ki (95% CI: 4.6–9.3) and 4.0 m with single-agent SERD (95% CI: 3.4–9.8). In the SERD cohort, earlier line of treatment (1–2 vs. 3+; HR 0.29, 95% CI 0.09–0.90) and liver metastasis (HR 3.42, 95% CI 1.10–10.6) were independently associated with PFS. CDK 4/6i duration ≥12 m in the SERD cohort was associated with improved mPFS in stratified analysis (9.8 vs. 2.7 m; log-rank p = 0.002). For patients treated with PI3Ki followed by SERD (n = 8), mPFS1 was 8.9 m (95% CI: 5.6, NE) and mPFS2 was 6.0 m (95% CI: 2.3, NE). SERD to PI3Ki (n = 7) yielded a mPFS1 of 3.4 m (95% CI: 1.8, NE) and mPFS2 of 10.0 m (95% CI: 1.6, NE). Conclusions: Prior CDK4/6i ≥ 12 m in patients with co-occurring mutations treated with a SERD was associated with significantly improved PFS, potentially reflecting a conditioning effect of prior therapy. The total PFS (PFS1 + PFS2) in patients treated with sequential targeted therapies was similar; however, small sample sizes and potential confounders in this retrospective cohort limit definitive interpretations. Larger prospective studies are needed to determine optimal sequencing strategies.