Abstract
BACKGROUND: With a growing population of breast cancer survivors, it is important to acknowledge long-term consequences of breast cancer treatment, including left ventricular systolic dysfunction (LVSD). Although echocardiography is a reliable technique to diagnose LVSD, its limited accessibility in primary care poses challenges. METHODS: A cross-sectional diagnostic accuracy study among 350 long-term breast cancer survivors, at least 5 years after breast cancer diagnosis, comparing the diagnostic performance of index tests ECG and N-terminal pro B-type natriuretic peptide (NT-proBNP) to the reference test echocardiography. LVSD was defined as left ventricular ejection fraction (LVEF) less than 54% or LVEF less than 50% on echocardiography. RESULTS: The median age at time of investigation was 63 years (IQR 57-68), with a median follow-up duration since breast cancer diagnosis of 10 years (IQR 7-14). An abnormal ECG demonstrated a sensitivity of 63.0% (IQR 48.7-75.7), a corresponding specificity of 51.7 (IQR 45.8-57.6) and a negative likelihood ratio of 0.7 (IQR 0.5-1.0) for detecting a LVEF less than 54%. An abnormal ECG showed a sensitivity of 75.0 (IQR 47.6-92.7), a corresponding specificity of 50.6 (IQR 45.1-56.2) and a negative likelihood ratio of 0.5 (0.2-1.2) for detecting LVSD defined as LVEF less than 50%. The area under the curve for NT-proBNP was 0.59 (95% confidence interval: 0.50-0.68) for detecting LVEF less than 54% and 0.56 (95% confidence interval: 0.39-0.74) for detecting LVEF less than 50%. DISCUSSION: ECG and NT-proBNP are inadequate diagnostic tools to screen for LVSD among asymptomatic long-term breast cancer survivors.