Cardiotoxicity and related risk factors in long-term trastuzumab (T) administration for metastatic breast cancer (MBC): A retrospective analysis.

Abstract

e11616 Background: Continuing T beyond progression has become a common strategy in the treatment of human epidermal growth receptor 2- overexpressing (HER2) MBC. However, T administered for several years with concomitant chemotherapy elicits concern about cardiac safety especially in patients (pts) with risk factors. Methods: Cardiac events (CEs) and survival of HER2 MBC pts treated with T +/- chemotherapy at our institution from Dec 2003 to Jun 2012 were evaluated. CEs were graded by NCI-CTCAE v 3.0. Risk factors assessed for cardiotoxicity were: age, body mass index, antihypertensive therapy, history of cardiac disease, diabetes, hypothyroidism, smoking, prior radiotherapy on the chest wall, prior cumulative dose of anthracycline(A), interval between last A dose and first T dose, baseline LVEF, continued/interrupted T exposure, concomitant chemotherapy. Chi-square test was used to compare distribution of CEs over different times of T exposure (p≤ 0.05). Univariate and multivariate Cox regression analysis were used to assess the effect of risk predictors. Results: Sixty-two pts assessable. Median age 52 years (range, 29 to 76), median cumulative time receiving T 29.5 months (range, 3 to 99 months); 40 pts (64.5%) received T without interruption and 19 pts (30.6%) were treated for more than 36 months. CEs occurred in 11 out of all pts (17.7%): grade 1 in 3 pts (4.8%), grade 2 in 5 (8.1%) and grade 3 in 3 (4.8%). The rate of CEs showed no statistically significant difference in pts receiving T for up to 36 months and over: 7/43 (16.3%) and 4/19 (21%), respectively, (p =0.724). In univariate Cox regression analysis significant risk factors were: history of cardiac disease (HR 6,814, 95% CI: 1,384-33,542) and smoking (HR 5,228, 95% CI: 1,403-19,491). In multivariate analysis smoking was the only independent predictor (HR 5,886, 95% CI: 1,479-23,247). Median survival from MBC diagnosis was 50 months (range, 6 to 101 months). Conclusions: Despite the limited sample size, our analysis suggests that cardiotoxicity does not hamper a long-term use of T, since the rate of CEs did not increase in pts treated over 36 months. Moreover, smoking appears to be a predictive factor of T cardiotoxicity.