P1750 Association of high-sensitivity troponin elevation and LVEF decline in anthracycline-treated breast cancer patients

Abstract

Abstract Introduction Cancer therapies have been linked to a wide variety of side effects, with cardiotoxicity being the most significant one. Early detection of subclinical cardiac dysfunction in cancer patients (pts) is necessary in order to prevent unfavorable outcomes. High-sensitivity troponin I (hsTI) levels have been proposed as predictors of cancer therapy related cardiotoxicity and may guide cardioprotective therapy initiation. Purpose To describe the incidence of cancer treatment related cardiac dysfunction (CTRCD) in a population of breast cancer pts under chemotherapy in a single center cardio-oncology unit and to assess its relationship with hsTI levels. Methods We retrospectively evaluated 83 women on anthracycline therapy for breast cancer, with or without anti-HER2 therapy, followed-up between January 2017 and July 2018. CTRCD was defined as more than 10% absolute reduction of LV ejection function (LVEF) to a value below 50%. Elevation of hsTI was defined as at least one measurement above the 99th percentile upper reference limit during follow-up (>34 ng/L). Pts had an organized follow-up in our cardio-oncology unit, consisting of a clinical, laboratorial (with dosing of cardiac biomarkers) and echocardiographic assessment at 0, 3, 6 and 12 months (or more frequently in selected high-risk cases). Results A total of 83 women with a mean age of 49 years (26-76) were included. 4 pts (4.8%) developed CTRCD. 17 pts (20.5%) were considered at high risk of cardiac dysfunction due to hsTI elevation. During follow-up, the percent increase in the hsTI level (from basal level) correlated with CTRCD (p = 0.02). On the other hand, the absolute maximum value of hsTI did not (p = 0.159). In fact, pts who developed CTRCD had a significantly higher percent increase in the hsTI levels (142.9% +- 57.5%) vs those without CTRCD (14.29% +- 4.6%), p < 0.001. On ROC curve analysis, percent increase in troponin was a good identifier of CTRCD (AUC of 0.986; 95% CI 0.95-1.00; p = 0.022) and the best cut-off value was a 79.8% increase in hsTI (sensitivity: 100%; specificity: 97.2%). Conclusion In our population, the percent increase in the hsTI levels correlated with CTRCD. Larger studies are needed to prove this parameter as a predictor of CTRCD. Abstract P1750 Figure.