Cardiac Events in Breast Cancer Patients Treated with Modern Radiotherapy

Abstract

<h3>Purpose/Objective(s)</h3> To characterize cardiac events in a large cohort of breast cancer patients treated with curative radiation therapy at our institution from 2006-2017. Recent reports of cardiac toxicity have primarily been limited to ischemic events. Our hypotheses were that (1) multiple classes of cardiac events would be observed, including ischemic, arrhythmia, cardiomyopathy, pericardial, and valvular events, and (2) toxicity would decrease as modern RT techniques evolved during the study period. <h3>Materials/Methods</h3> Baseline patient characteristics, comorbidities, tumor characteristics, surgical, RT, and systemic treatment were recorded. RT volumes included accelerated partial breast, whole breast, or breast/chest wall and nodal RT. Techniques included 3D-CRT, IMRT, brachytherapy, protons, with prone and deep inspiration breath hold techniques also introduced during the study period. Cardiac events were categorized by class (ischemic, arrhythmia, cardiomyopathy, pericardial, and valvular), as well as by grade (1-5) based on CTCAE and American College of Cardiology criteria. <h3>Results</h3> 3,435 breast cancer patients received adjuvant RT at our institution during this time period and were therefore included in the analysis. Median follow up for cardiac events was 96 months. Mean age at diagnosis was 57 years old and 70% of patients were postmenopausal. 71% were Caucasian, 26% African American, 2% Asian, 1% other. At baseline, 14% of patients had diabetes, 47% hypertension, 24% hyperlipidemia, and 11% had history of a cardiac event prior to treatment. Systemic therapy included endocrine therapy (66%), any chemotherapy (49%), anthracycline (25%) and Herceptin (11%). Incidence of cardiac events was 15% overall. Events by class were 4.5% cardiomyopathy, 3.6% ischemic, 7.0% arrhythmia, 1.8% pericardial, and 5.0% valvular. Patients with history of cardiac events prior to treatment had a greater risk of new events compared to those without pre-existing disease (p <0.01, RR 2.8). Incidence of Grade 1 events was 2.3%, Grade 2 was 7.2%, Grade 3 was 3.2%, and Grade 4 was 2.2%. Incidence of events by year of RT were 22% (2006-2007), 19% (2008-09), 17% (2010-11), 15% (2012-13), 12% (2014-15), and 12% (2016-17). Logistic regression found a significant decrease in cardiac events over time (p < 0.001). Median interval from RT to cardiac event was 45 months (all events), 41 mos (cardiomyopathy), 47 mos (ischemia), 53 mos (arrhythmia), and 45 mos (pericardial). <h3>Conclusion</h3> Incidence of cardiac events were overall fairly low, but not uncommon when all event classes were recorded. Incidence of events was lower in patients treated later in the study period, which may be due to introduction of further advances in RT for breast cancer patients over time. Although patients treated earlier have potential for longer follow up, the median interval to cardiac event was less than the median follow up for the cohort, so these findings will likely be observed with further follow up.