49P A ctDNA tool for identifying ESR1-mutated biology in ESR1-non-mutated hormone receptor-positive (HR+) metastatic breast cancer (MBC)

The efficacy of Selective Estrogen Receptor Degraders (SERDs) in ESR1-non-mutated (ESR1-non-mut) tumors is uncertain. To address this gap, we used DNADX, a DNA NGS phenotypic profiling tool that identifies 150 distinct signatures and categorizes tumors into 5 subtypes (Nat Commun 2023) using DNA copy number (CN) profiles from ctDNA. Our research focuses on comparing ESR1 mutated (ESR1mut) and ESR1-non-mut tumors using genomic features. This study used 126 patients (pts) with advanced HR+ MBC after progression on CDK4/6 inhibitors (CDK4/6i) plus endocrine therapy (ET). Plasma ctDNA samples were characterized for ESR1mut status by Guardant-360 liquid biopsy assay and profiled by the DNADX assay (cohort 1). In this cohort, we developed an ESR1mut genomic predictor using the DNADX assay. Validation of the ESR1mut predictor was conducted in three independent cohorts: a plasma dataset of 132 HR+/HER2- BCs before starting CDK4/6i plus ET (cohort 2), and a tumor tissue dataset of 381 pts (MSKCC, cohort 3). The DNADX ESR1mut genomic predictor was also compared with the DNADX 5-subtype classification. ESR1mut tumors had higher DNADX plasma tumor fraction than ESR1-non-mut tumors (p=0.008). In plasma, ESR1mut tumors displayed increased activity in DNA luminal-related signatures and decreased activity in Basal-like-related signatures, with no notable differences in proliferation-based signatures compared to ESR1-non-mut tumors. DNADX ESR1mut genomic predictor demonstrated an Area Under the Curve of 0.764 in the training dataset and 0.710 in the validation dataset for ESR1mut identification. Notably, 30-40% of ESR1-non-mut tumors had high DNADX ESR1mut genomic scores, indicating biological similarity to ESR1mut tumors. ESR1mut tumors distribution across the DNADX's 5-subtype classification varied significantly, with the lowest prevalence in the aggressive Cluster-3 subtype (4% vs. 20.4%; p<0.001). Metastatic ESR1mut and non-mut tumors show biological heterogeneity in ctDNA CN landscapes. A subset of ESR1-non-mut tumors share molecular features with ESR1-mut tumors. DNADX profiling could be crucial in guiding SERDs development beyond ESR1mut focus.