Efficacy of oral selective estrogen receptor degraders (SERD)s in the treatment of estrogen receptor positive (ER+), HER2-negative metastatic breast cancer (MBC): A stratified analysis of the ESR1 wild type (wt) and mutant (mt) subgroups.

1096 Background: Oral SERDs are a novel drug class developed to counteract endocrine resistance due to ESR1 mutations. Several SERDs have emerged from phase 2 and 3 trials, and the FDA has most recently limited approval for Elacestrant to patients with ESR1mt tumours despite PFS benefit in the overall population. Questions remain on whether patients with ESR1wt tumours stand to benefit from SERDs. We conducted an IPD meta-analysis of the trials EMERALD, SERENA2, ACELERA and AMEERA3 to compare survival outcomes of SERDs in the 2nd or 3rd line setting against endocrine therapy (ET), using KMSubtraction to unveil HRs of unreported ESR1wt subgroups from AMEERA3 and EMERALD. Meta-analysis results were stratified by ESR1 status. Methods: RCTs investigating the efficacy of novel SERDs versus ET for ER+, HER2- MBC, and which reported the Kaplan Meier (KM) plots of progression free survival (PFS) were selected after a systematic search of Embase and PubMed from inception until January 21,2023. A graphical reconstructive algorithm was applied to estimate time-to-event outcomes from reported KM plots in all overall and reported subgroup cohorts, deriving IPD. A bipartite matching algorithm, KMSubtraction was used to subtract patients in the reported subgroup from the overall cohort, to derive survival data for unreported ESR1wt subgroups. An IPD meta-analysis was then performed, pooling data by ESR1 mutation. Results: A total of 1216 patients were included in this analysis. In the pooled analysis of the overall cohort, PFS benefit was observed with SERDs compared to treatment of physician's choice (TPC) (HR 0.787, 95%CI 0.683-0.908, p<0.001). In the ESR1mt subgroup SERDs demonstrated improved PFS (HR 0.557, 95%CI 0.440-0.705, p<0.001) compared with TPC. In the ESR1wt subgroup, SERDs demonstrated no significant PFS benefit (HR 0.948, 95%CI 0.790-1.137, p=0.565) when compared to TPC. KMSubtraction was implemented on AMEERA3 to reveal survival data in ESR1wt subgroup. Amcenestrant revealed no significant PFS difference in the ESR1wt population (HR 1.197, 95% CI 0.857-1.670, p=0.291) when compared to ET. Additionally, a separate analysis of EMERALD was performed using KMSubtraction to derive survival data of the ESR1wt subgroup comparing Elacestrant vs Fulvestrant. No significant PFS difference was observed between the 2 agents in the ESR1wt population (HR 0.832, 95% CI 0.587-1.178, p=0.299). Conclusions: Our results suggest that PFS benefit in the overall population is mainly driven by the ESR1mt subgroup. These findings are in line with the recent FDA approval of Elacestrant for patients with ESR1mt tumours. [Table: see text]