Assessment of left ventricular function by CMR versus MUGA scans in breast cancer patients receiving trastuzumab: a prospective observational study.

ABSTRACT

Little is known about the comparison of multiple-gated acquisition (MUGA) scanning with cardiovascular magnetic resonance (CMR) for serial monitoring of HER2+ breast cancer patients receiving trastuzumab. The association of cardiac biomarkers with CMR left ventricular (LV) function and volume is also not well studied. Our objectives were to compare CMR and MUGA for left ventricular ejection fraction (LVEF) assessment, and to examine the association between changes in brain natriuretic peptide (NT-BNP) and troponin-I and changes in CMR LV function and volume. This prospective longitudinal two-centre cohort study recruited HER2+ breast cancer patients between January 2010 and December 2013. MUGA, CMR, NT-BNP and troponin-I were performed at baseline, 6, 12, and 18 months after trastuzumab initiation. In total, 41 patients (age 51.7 ± 10.8 years) were enrolled. LVEF comparison between MUGA and CMR demonstrated weak agreement (Lin's correlation coefficient r = 0.46, baseline; r = 0.29, 6 months; r = 0.42, 12 months; r = 0.39, 18 months; all p < 0.05). Bland-Altman plots demonstrated wide LVEF agreement limits (pooled agreement limits 3.0 ± 6.2). Both modalities demonstrated significant LVEF decline at 6 and 12 months from baseline, concomitant with increased LV volumes on CMR. Changes in NT-BNP correlated with changes in LV diastolic volume at 12 and 18 months (p < 0.05), and LV systolic volume at 18 months (p < 0.05). Changes in troponin-I did not correlate with changes in LV function or volume at any timepoint. In conclusion, CMR and MUGA LVEF are not interchangeable, warranting selection and utility of one modality for serial monitoring. CMR is useful due to less radiation exposure and accuracy of LV volume measurements. Changes in NT-BNP correlated with changes in LV volumes.