Comparison of Multi-Planar Acquisition Scan and Echocardiography Strategies for Detecting Cancer-Treatment-Related Cardiac Dysfunction in a Prospective Cohort

Previous studies suggest cardiac imaging in cancer patients for detection of cancer-therapy-related dysfunction (CTRCD) is discordant with international guidelines, clinical implications unknown. Cancer were recruited prior to either anthracyclines or trastuzumab administration and underwent concurrent strategy Multigated Acquisition Scan (MUGA)-based timing determined by treating oncologist, echocardiographic-based advised European Society Medical Oncology (ESMO) guidelines. Outcomes included: (a) adherence timepoints; (b) limits-of-agreement (LoA) between MUGA echocardiographic LVEF measurements; (c) CTRCD; (d) global longitudinal strain accuracy predicting dysfunction. 35 (15 breast cancer, 19 hematological malignancy 1 gastric patient). 14 treated 21 anthracyclines. Adherence guideline-advised was 37.1% MUGA-guided approach, 83.8% echocardiographic-guided approach (p<0.001 difference). 2D-LVEF MUGA-LVEF correlated poorly (r=0.51, 95%CI [-0.09 0.37], p=0.22), poor agreement (bias=0.2%, LoA=37.0%, [-18.45 +18.7%]). Significant correlation seen 3D-LVEF (r=0.33, [0.1 0.53], p=0.005) (bias=1.8%, LoA=29.9%, 95% CI [-13.1 16.8%]). Only case (2.9%) CTRCD diagnosed MUGA, compared 6 (17.1%) 3D-LVEF. Four (11%) had >10% decrement not detected MUGA. GLS at 2 months displayed significant ability predict as defined (AUC 0.75,95%CI [0.55 – 0.94]). Cardiac risk diverges from recommended guidelines the frequently-chosen modality correlates measurements.