1034 Background: While the antibody-drug conjugates (ADCs) SG and T-DXd have transformed MBC care, predictive biomarkers of ADC response are lacking. This retrospective study evaluates real-world outcomes of MBC patients treated with SG or T-DXd and explores potential genomic and transcriptomic correlates of response. Methods: MBC patients treated with SG or T-DXd at BC Cancer who underwent tumor whole genome and transcriptome sequencing (WGTS) between January 2020 - May 2025 were included. Biopsy for WGTS was obtained either before (ADC-naïve) or after ADC exposure (ADC-exposed). Progression free survival (PFS) was clinically evaluated and assessed using Kaplan-Meier analysis. Treatment response was classified into 2 groups using PFS benchmarks from ASCENT (Bardia et al 2021, SG: expected ≤6 months (mo), favorable >6 mo) and DESTINY-Breast04 (Modi et al 2022, T-DXd: expected ≤12 mo, favorable >12 mo). RNA-seq data were assessed with DESeq2 differential expression and gene pathway enrichment analyses. Immune cell deconvolution was performed using CIBERSORT, with group comparisons by Wilcoxon rank-sum tests. Results: Forty-three patients met eligibility for inclusion (17 SG, 24 T-DXd, 2 treated with both agents). Median PFS was 4.0 mo (95% CI 2.0-6.0) for those treated with SG and 7.4 mo (95% CI 4.6-10.7) for those treated with T-DXd. Among T-DXd-naïve patients, ESR1 expression was higher in those with poor ADC response (69.8 vs 7.8 transcripts per million, p = 0.049). Oxidative phosphorylation gene cluster expression was reduced in T-DXd-exposed vs T-DXd-naïve patients (normalized enrichment score -1.9, p = 0.005). ADC target expression did not differ by response in either treatment cohort ( TACSTD2 p = 0.86, ERBB2 p = 0.67). Whole genome sequencing identified 12 tumors with FGFR1 amplifications, all with poor ADC response, and demonstrated homologous recombination deficiency in half of those with favorable SG response. While not statistically significant, the post-ADC tumor microenvironment of favorable responders trended toward enrichment with naïve B-cells and T-cell infiltrate in T-DXd-treated patients (p = 0.28, 0.33) and both monocytes and NK cells in SG-treated patients (p = 0.1, 0.27). Conclusions: This study highlights the heterogeneity of outcomes among SG- and T-DXd-treated MBC patients and identifies potential predictive and prognostic biomarkers warranting exploration in prospective trials of first-line ADC (NCT07299409) and ADC sequencing (NCT06665178). SG-treated patients (n=19) T-DXd-treated patients (n=26) HR status:PositiveNegative 118 224 HER2 status:Negative/UltralowLowPositive 8101 11114 Timing of biopsy for WGTS:Pre-ADCPost-ADC 127 1511 Favorable ADC response 4 3 Median lines of therapy for MBC pre-ADC (range) 2 (0-6) 4 (0-8) Median lines of endocrine therapy for HR + MBC pre-ADC (range) 0 (N/A) 1 (0-5)
Arthur et al. (Wed,) studied this question.