Abstract PS1-03-22: Predictors of cardiotoxicity with HER2-directed antibody therapy in a real-world setting across a large, Western US hospital system

Abstract Background: Incidences of cardiotoxicity related to HER2-directed antibody therapy (anti-HER2) vary considerably from 0.5% to 30% depending on comorbidities and chemotherapy regimen. Recent studies have posited whether low-risk cohorts could decrease or eliminate surveillance echocardiography given low incidences of treatment-related heart failure (HF). Objectives: The aim of this study was to characterize the incidence of cardiotoxicity in patients receiving HER2-therapies, evaluate predictors of cardiotoxicity, and examine the incidence of LVEF decline and new HF among a low-risk subgroup. Methods: We performed a retrospective cohort study of early stage (I-III) breast cancer patients who received anti-HER2 between 1/1/2016-3/20/2024 and completed baseline and surveillance echocardiography in the 7-state Providence healthcare network. The primary outcome was development of LVEF ≤50% during anti-HER2. Secondary outcomes included new HF diagnosis (based on ICD-10 codes I50.x), any hospitalization, and all-cause mortality. Incidences of LVEF ≤50% and new HF were also obtained in a low-risk subgroup. Logistic regression models accounting for demographics, comorbidities, HER2 treatment regimen, and preceding cardiac medications were performed. Results: The study included 1,215 patients (Table). Fifty-two patients (4%) developed LVEF ≤50% during anti-HER2. Among this subset, 52% were hospitalized and 42% had a new HF diagnosis during the treatment period. No patients died while receiving anti-HER2 or during 1-year follow up. Univariate modeling revealed that CAD (aOR [95% CI] = 3.22 [1.21, 8.61]) and prior HF (aOR [95% CI] = 3.67 [1.23-10.99]) were associated with treatment-related LVEF ≤50%, though no comorbidities were predictive on multivariate analysis. After risk adjustment, HER2 treatment regimen was the only predictor of LVEF ≤50% during therapy (taxanes + platinum vs. anthracyclines: aOR [95% CI] = 0.22 [0.10, 0.48]; taxanes vs. anthracyclines: aOR [95% CI] = 0.29 [0.13, 0.65]. New HF was observed in 11% of patients receiving anthracyclines compared to 4-5% in other treatment arms (p = 0.024). Patients deemed low-risk (non-anthracycline regimen, no prior CAD or HF) developed an LVEF ≤50% in 3.3% of patients and new HF in 2.6% of patients. Conclusions: In a real-world clinical setting, anti-HER2-associated LVEF declines were uncommon but were associated with increased hospitalizations and HF diagnoses. Exposure to anthracyclines was the only predictive risk factor for LVEF decline. Interim analysis of a low-risk cohort (non-anthracycline regimen, no prior CAD or HF) still demonstrated a modest incidence of LVEF decline and HF, suggesting that enhanced risk stratification should be considered before deferring echocardiographic surveillance in these subgroups. Citation Format: C. Smith, M. Layoun, H. Li, E. Koltner, M. T. Imboden, Z. Taylor, K. J. Spinelli, D. Page. Predictors of cardiotoxicity with HER2-directed antibody therapy in a real-world setting across a large, Western US hospital system [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS1-03-22.