Co-occurrence of ESR1 and PIK3CA mutations in HR+/HER2- metastatic breast cancer: Incidence and outcomes with targeted therapy.

e13097 Background: ESR1 and PIK3CA mutations inform the use of targeted agents such as alpelisib (A), elacestrant (E), and capivasertib in patients (pts) with HR+/HER2- metastatic breast cancer (mBC). These mutations predict a worse prognosis and shorter response to endocrine therapy (ET). The prevalence of PIK3CA and ESR1 mutations detected via circulating tumor DNA (ctDNA) in mBC is about 35% and 33%, respectively, and the co-occurrence rate is 10-15%. Treatment outcomes with these targeted agents in pts with both mutations can inform sequencing decisions. Methods: We retrospectively analyzed pts treated at Miami Cancer Institute with HR+/HER2- mBC who underwent ctDNA testing (performed by Guardant360) between March 2019 and Jan 2024. The rate of co-occurring pathogenic ESR1 and PIK3CA mutations on a single sample was quantified. We report pt demographics, disease characteristics, and real world outcomes on A + ET and E in the co-mutated cohort. Results: Of the 626 pts who underwent ctDNA testing, 31.6% (n=198) had at least one PIK3CA mutation and 23.8% (n=149) had at least one ESR1 mutation. Co-mutations in both ESR1 and PIK3CA were found in 10.4% (n=65). In the co-mutated cohort, median age was 63, 84.6% were post-menopausal, and 58.5% were Hispanic. Most pts received prior CDK4/6 inhibitor (90.7%). A + ET was administered to 27 pts (41.5%) with a median of 2 prior lines of therapy (range 1-8). In these pts, 33.3% (n=9) achieved stable disease, 25.9% (n=7) partial response, 14.8% (n=4) progression, and 18.5% (n=5) discontinued due to toxicity prior to re-staging. E was administered to 12 pts (18.5%) with a median of 4 prior lines of therapy (range 1-13), with 4 pts still on treatment. None achieved stable disease or objective response, 66.7% (n=8) developed progression of disease as best response, and 33.3% (n=4) currently remain on treatment pending response evaluation. In the co-mutated cohort, only 4 received both A + ET and E, (3 received A first, 1 received E first). Pts treated with A + ET only (n=23) had a median time to next treatment (mTTNT) of 120 days (range 5-566) and those who received E only (n=8) had a mTTNT of 86 days (range 13-106) (Chi square=4.289, df = 1, p<0.05). The 4 pts who received both therapies were excluded from this mTNNT analysis. Conclusions: The rate of ESR1 and PIK3CA co-mutation in this cohort was consistent with the reported literature. mTTNT was significantly longer with A + ET compared to E, but real world outcomes on E remain immature with one-third of pts still on E at the time of analysis. A limitation of this study was the small sample size. Further analyses are warranted to clarify the optimal sequencing of these targeted therapies such as A, E, and capivasertib in pts with co-mutations in ESR1 and PIK3CA.