Real-world estrogen receptor 1 ( <i>ESR1</i> ) testing patterns and results in U.S. patients with metastatic breast cancer, 2018-2024.

579 Background: Estrogen receptor alpha 1 ( ESR1 ) mutations (m) are a significant mechanism of resistance to endocrine therapy (ET) in estrogen receptor-positive (ER+) advanced breast cancer (BC), leading to shorter progression-free and overall survival. Although multiple clinical studies, including PADA-1 and SERENA-6, have highlighted the importance of ESR1 testing during first-line (1L) therapy, little is known about testing in real-world practice. This study aims to characterize real-world ESR1 testing patterns and the timing of emergence of ESR1 m in 1L. Methods: This retrospective, observational cohort study used the US-based, deidentified Flatiron Health-Foundation Medicine metastatic BC (mBC) Clinico-Genomic Database (CGDB) and Flatiron Health Research Database (FHRD). Adult patients with a confirmed diagnosis of ER+/human epidermal growth factor receptor 2-negative (HER2-) mBC from 1/1/2018 to 9/30/2024 were included. ESR1 testing patterns and the prevalence of ESR1 m were descriptively analyzed. Results: Among 7772 patients with mBC in the FHRD who initiated 1L therapy, tumor ESR1 m status was evaluated for 222 (2.9%) patients at baseline (within 90 days before initiating 1L) and 1355 (17%) during 1L. The percentage of patients whose tumor was tested during 1L increased over time (11% among those diagnosed in 2018-19 to 19% in 2020-21 and 22% in 2022-23), even before accounting for shorter follow-up in recent years. While most patients (88%) did not provide a specimen for ESR1 testing during 1L, among the 963 patients that did, 90% provided 1 specimen and 10% provided 2+ specimens. The majority (60%) of tests during 1L were performed using tissue specimens. Median time from 1L initiation to first ESR1 test was 7.5 months (mos) among tested patients. A positive test result was reported for 29/222 (13%) patients tested at baseline and 240/1355 (18%) tested during 1L. Among the subset who initiated 2L, 185/993 (19%) tested positive before 2L start date. Focusing on time periods wherein specimens were provided, ESR1 m test positivity was 6.7% (76/1,127) for specimens provided at baseline, 23% (15/65) for specimens provided 9 to 12 mos into 1L therapy, 38% (26/69) for those provided 15 to 18 mos into 1L, and 40% (38/94) for those provided from 18 to 24 mos into 1L. Among 419 patients in the CGDB (which provides deeper genomic data on an analogous population) whose tumors were found to harbor ESR1 m, the most common ESR1 variants were D538G (201 [48%]), Y537S (142 [34%]), and E380Q (68 [16%]). Conclusions: These findings indicate that ESR1 m can be detected at any time interval during 1L, with the probability of ESR1 m detection increasing over time. The low rates of testing during 1L highlight an opportunity for improved testing surveillance to better guide patient management and improve outcomes during 1L.