Background: Prior retrospective data has shown ESR1 mutation is an acquired resistant mechanism to AI. However, little is known about ist prognostic and predictive value to endocrine-based therapy and chemotherapy. PEARL study compared PAL+ET vs CAP in AI resistant patients. Here we wxplored prospectively the ESR1 mutational status based on updated PFS and OS. Methods: PEARL is a phase 3 study with 2 subsequent cohorts: cohort 1 with 296 pts randomized to PAL+Exemestane (EXE) vs CAP and cohort 2 with 305 pts randomized to PAL+ Fulvestran (FUL) vs CAP. ESR1 hotspot mutations were analyzed in circulating free DNA at baseline by digital PCR. Cox regresion analysis for PFS and OS were performed. Adjusted hazard ratios (aHR) and interaction test between treatment and ESR1 status were calculated. Results:ESR1 mutational status was assessed in 557 pts, 91% and 94% treated with CAP and PAL+ET, respectively; 164 (29%) had ESR1 mutations. Characteristics between ESR1 wild-type (WT) and mutated pts were balanced except for ECOG, prior sensitivity to ET, treatment line, prior AI for metastatic disease and time from first metastatic diagnosis to randomization. Median follow-up was 22.5 months (m). Median PFS was 9.6 vs. 7.5 m (HR: 1.06 [0.88 - 1.28], p = 0.522) and median OS was 30.2 vs 30.3 (HR: 0.99 [0.78 - 1.26] p = 0.934), for CAP vs. PAL+ET, respectively. No interaction was seen between treatment arm and ESR1 status either for PFS (p = 0.538) or OS (p = 0.957). Conclusions: In luminal MBC, ESR1-mutated pts had poorer OS than ESR1-WT pts regardless of treatment received. Clinical trial information: NCT02028507.

* adjusted by age, treatment arm (for the global population), disease site (visceral yes/no), prior sensitivity to ET, prior chemotherapy for MBC, prior AI for MBC, number of involved sites and time from first metastatic diagnosis to randomization.