Sequence of similar-payload ADCs: Does it make sense in real-world HER2-low advanced breast cancer?—The Portuguese national study.

Abstract

1140 Background: New-generation Antibody-Drug Conjugates (ADC) have expanded treatment options for HER2-low advanced breast cancer (ABC), regardless of hormone receptor (HR) status. Sacituzumab govitecan (SG) and trastuzumab deruxtecan (T-DXd) are used sequentially in clinical practice, but real-world data on efficacy, sequencing and potential cross-resistance are limited. We aimed to evaluate efficacy, sequencing patterns and predictors of outcome in a national real-world cohort of HER2-low ABC. Methods: We conducted a multicenter, retrospective real-world cohort study including patients with HER2-low ABC treated sequentially with SG and T-DXd, regardless of sequence, across 14 Portuguese centers. Eligible patients initiated the first ADC between 2023 and 2025 and received ≥1 cycle of both ADCs. The primary endpoint was progression-free survival (PFS), evaluated for the first (PFS1) and second ADC (PFS2). Secondary endpoints included objective response rate (ORR), primary resistance to the second ADC (PFS2 ≤3 months) and safety. PFS was analyzed using time-to-event methods. For exploratory analyses, PFS1 and PFS2 were dichotomized at the median and associations with longer PFS were assessed using chi-square tests. Predictors of PFS2 were explored in an exploratory multivariable Cox regression model. Results: Sixty-four patients were included; 36 (56%) had HR-positive disease and 28 (44%) were HR-negative. Median age at ADC1 initiation was 51 years (27–79). T-DXd was used as the first ADC (ADC1) in 34 (53%) patients and was more frequently administered as ADC1 in HR-positive disease. A total of 22 (34%) patients received ≥1 intermediate line, mainly chemotherapy (91%). Median PFS1 was 8.9 months (95% CI 5.8–9.7) and median PFS2 was 2.7 months (95% CI 2.1–3.2). PFS2 was consistently shorter than PFS1 across subgroups, independent of treatment line, with 43 (70%) patients presenting primary resistance to the second ADC. ORR to ADC2 was 3%. No significant differences in PFS1 or PFS2 were observed by ADC sequencing. Longer PFS1 was not associated with improved PFS2 (p=0.44). In the multivariable model, none of the evaluated variables, including the use of ≥1 intermediate line between ADCs, were significantly associated with longer PFS2 (all p>0.05). No new safety signals were observed; rates of dose reductions and interruptions were consistent with known safety profiles of SG and T-DXd. Conclusions: In this national real-world cohort of HER2-low ABC, clinical benefit from a second ADC was limited, with shorter PFS and frequent early progression, irrespective of sequencing order. Importantly, no baseline clinical factors predicted PFS2. These findings may suggest clinically relevant cross-resistance between ADCs with similar payloads and highlight shared resistance mechanisms as a key challenge in optimizing ADC sequencing.