Background: There is little information on cardiovascular (CV) outcomes and survival of breast cancer (BC) pts who had cardiotoxicity during adjuvant trastuzumab (Tmab) treatment requiring discontinuation in the real world. Methods: This was a retrospective population-based study of early-stage (I-III) BC pts diagnosed before 2010 and treated with Tmab in Ontario. Patients were stratified based on Tmab doses received: early discontinuation as 1-8 or 9-15, and ≥16 as completion. Time-dependent multivariable Cox models (adjusting for known cardiac risk factors, comorbidities, stage and chemotherapies received) were used to analyze the primary endpoint overall survival (OS), and the following composite endpoints: hospitalization/emergency room visit for heart failure (HF) or death; non-HF CV event (myocardial infarction, stroke) or death; clinically significant relapse (initiation of palliative systemic therapy > 90 days after last Tmab dose) or death. Results: Of the 3134 women (83% <65 years), 6%, 10%, and 85% received 1-8, 9-15, and ≥16 doses, respectively. Over 5-year median follow-up, early discontinuation was significantly associated with worse outcomes (Table) and lower OS (77%, 80%, and 93%; p<0.001). Early discontinuation (Table) and clinically significant relapse (HR 34.0, 95% CI 24.9-46.6) were both independent predictors of mortality. These results were confirmed using propensity score methods. Conclusions: Early Tmab discontinuation, a marker of cardiotoxicity, is a powerful independent predictor of adverse cardiac events and clinically significant relapse, both likely contributing to poor overall survival. Both optimal cancer and cardiovascular treatment strategies are needed for to improve the outcome of these high-risk patients.

† Modeled as time-dependent covariate (≥16 doses as reference).