Abstract P3-03-13: Long-term heart failure risk of trastuzumab with or without anthracyclines in early stage breast cancer: A SEER-Medicare Database Analysis

Background: Cardiotoxicity associated with use of trastuzumab (T) in breast cancer has been assessed in multiple studies with short-term follow up. However, long-term risk of clinical heart failure (CHF) with the use of T, either with or without anthracyclines (A), is poorly understood. Since the approval of T for use in early-stage breast cancer (ESBC) in 2006, long-term follow up data is now available in the SEER-Medicare database. Methods: We performed a retrospective cohort study in patients (pts) with ESBC identified in the SEER-Medicare database, diagnosed between 2005-2016, excluding pts with in-situ or metastatic disease at time of diagnosis or a history of CHF prior to diagnosis. Primary exposure variables were identified using HCPCS codes and included receipt of T or A. The reference baseline population used for univariate logistic regression of the primary exposure variables was pts with ESBC who received neither T nor A. For multivariate analyses, to avoid co-linearity in the final model, the exposure variables were categorized as either having or not having the exposure (i.e. with or without T, with or without A). The primary outcome variable was CHF, identified using codes 428.0-9, 402.11, 402.91 (ICD9) and I50.0-9, I110 (ICD10). Covariates used in multivariate logistic regression (MLR) included hypertension (HTN), valve disease, hyperthyroidism, diabetes (DM), emphysema, coronary artery disease (CAD), left sided radiation, age, and race. MLR models were constructed using a stepwise-up approach and confirmed using the stepwise-down method with a cutoff of p< 0.05 for inclusion in the final model. Results: A total of 244,129 pts were identified with a median follow up of 6.7 years (7.7 years in the T cohort). Median age was 68. Other demographics are listed in Table 1. Of these, 10,603 were HER2-positive (HER2+) pts who all received T and all of whom also received some form of chemotherapy. Of the HER2+ pts, 19.1% (2,026) also received A in addition to T as part of their treatment regimen. Pts who received T had significantly higher baseline rates of HTN, valve disease, hyperthyroidism, DM, emphysema, and CAD as compared to pts who did not receive T. Compared to pts who received T without A, pts who received T with A had significantly higher baseline rates of HTN, valve disease, hyperthyroidism, emphysema, and CAD with no difference in baseline rate of DM. Risk of CHF with exposure variables are summarized in Table 2. The overall rate of CHF for the entire 244,129 population was 21.3% with a rate of 20.8% in pts who did not receive any T or A (the baseline population). The rate of CHF in pts who received T without A was 19.1% (OR 0.90, 95% CI 0.85-0.95), and 29.0% for those who received T with A (OR 1.55, 95% CI 1.41-1.71). Associations of covariates of interest with risk of CHF as well as MLR models are summarized in Table 3. After adjusting for all baseline cardiac comorbidities as well as other covariates, receipt of T was associated with a reduced risk of CHF (OR 0.74, 95% CI 0.70-0.79, p< 0.001), and receipt of A remained associated with an increased risk of CHF (OR 1.19, 95% CI 1.15-1.24, p< 0.001). Conclusions: With long-term follow up, the addition of A to T significantly increased the risk of CHF over T without A (from 19.1% to 29.0%) in pts with ESBC. Additionally, pts who received T without A had a similar risk of CHF compared to the baseline population who received neither T nor A. After adjusting for potential confounders in the MLR model, T was associated with a slightly decreased overall risk of CHF, while A remained associated with an increased risk of CHF. Possible explanations for these findings include potential increased cardiac monitoring in the T population which may have led to optimization of cardiac comorbidities, which were higher in the T population at baseline. Taken together, these data indicate concern with use of T plus A in ESBC regarding the long-term development of clinical CHF; especially when non-A regimens with similar efficacy are available. Table 1: Demographics Table 2: Risk of Heart Failure with Primary Exposure Variables Table 3: Univariate and Multivariate Analyses Citation Format: Nicholas P. McAndrew, Karissa Britten, Xiaoyan Wang, Eric Yang, Sara Hurvitz, Dennis Slamon. Long-term heart failure risk of trastuzumab with or without anthracyclines in early stage breast cancer: A SEER-Medicare Database Analysis [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P3-03-13.