<i>ESR1</i> hotspot–specific chemotherapy resistance patterns in metastatic breast cancer identified by ctDNA analysis.

Abstract

e13009 Background: ESR 1 mutations are established drivers of endocrine resistance in ER+ metastatic breast cancer (MBC), yet their relevance to chemotherapy resistance remains unclear. Building on our prior work showing circulating tumor DNA (ctDNA) ESR1 monitoring is prognostic in MBC (ASCO 2025 #1042; AACR 2025 #3613), we investigated whether baseline ctDNA ESR 1 hotspots are associated with agent-specific and broad chemotherapy resistance to inform risk stratification and treatment sequencing. Methods: This study included 158 ER + MBC patients (IRB: NU16B06) treated at Robert H. Lurie Cancer Center. Baseline ctDNA was analyzed by Guardant 360 NGS panel. Two clinically motivated analyses were applied to patients with ESR 1 mutations: (1) ESR 1 hotspot mutations were categorized as classic hotspots (D538G, Y537S/Y537N, E380Q) versus non-classic hotspots (L536H, L536R, L536P, K362N, K520K, V392I), and their association with chemotherapy resistance; and (2) very broad resistance was defined a priori as resistance to ≥5 chemotherapeutic agents. Associations were tested using two-sided Fisher’s exact test with odds ratios reported. Results: ESR1 mutations were identified in 10 hotspots among 40 of 158 patients, including 8 E380Q (mean 6.3%), 12 Y537S (mean 5.6%), 13 D538G (mean 11.9%), 1 V392I (0.1%), 1 Y537N (1.7%), 1 L536P (3.4%), 1 L536R (11.0%), 1 L536H (23.9%), 1 K520K (0.6%) and 1 K362 (15.8%). The cases sensitive vs resistant to chemotherapy drugs are as follows: Doxorubicin (22 vs 18), Epirubicin (33 vs 7), Pegylated Liposomal doxorubicin (37 vs 3), Carboplatin (25 vs 15), Paclitaxel (21 vs 19), Docetaxel (23 vs 17), Vinorelbine (29 vs 11) , Ixabepilone (34 vs 6), Eribulin (32 vs 8), Capecitabine (17 vs 23), Cyclophosphamide (16 vs 24), Gemcitabine (36 vs 4), and Abraxane (35 vs 5). Among 40 patients, vinorelbine resistance was more frequent in those harboring non-classic ESR 1 hotspots compared with classic hotspots (66.7% [4/6] vs 20.6% [7/34]; OR 7.71, P = 0.0385). When chemotherapy resistance burden was evaluated, very broad resistance (≥5 agents) was observed in 8/40 (20%) patients, and was absent in D538G cases (0/13, 0%) compared with non-D538G cases (8/27, 29.6%; P = 0.0373), suggesting a distinct resistance spectrum. These findings suggest ESR 1 hotspot mutation information may help identify patients at risk for specific agent resistance and early multi-line chemotherapy failure. Conclusions: Our findings indicate that baseline ctDNA ESR 1 hotspot heterogeneity is associated with distinct chemotherapy resistance patterns. Non-classic ESR 1 hotspots were enriched for vinorelbine resistance, whereas D538G was inversely associated with an extreme multi-drug resistance phenotype (≥5 agents), suggesting ESR 1 hotspots may serve as a clinically relevant biomarker for chemotherapy resistance spectrum, supporting prospective validation and biomarker-guided treatment strategies.