Phenotypes of Myocardial Dysfunction on Serial Echocardiography and CMR in Women With Early-Stage Breast Cancer.

BACKGROUND

Understanding phenotypic variations in left ventricular (LV) dysfunction during cancer therapy may allow for tailored surveillance and prevention.

OBJECTIVES

This study sought to determine LV dysfunction phenotypes during cancer therapy and their interrelated-ness and association with cancer therapy-related cardiac dysfunction (CTRCD), myocardial tissue changes, and blood biomarkers.

METHODS

This is a secondary analysis of the EMBRACE-MRI (Evaluation of Myocardial Changes During BReast Adenocarcinoma Therapy to Detect Cardiotoxicity Earlier With MRI) study in which women with early-stage HER2+ breast cancer were recruited prospectively. High-sensitivity troponin I, B-type natriuretic peptide, and echocardiography were obtained pre-anthracycline and every 3 months with measurement of systolic and diastolic function and left atrial reservoir strain (LARS). Cardiac magnetic resonance (CMR) was performed at baseline and follow-ups with quantification of myocardial T1, T2, and extracellular volume (ECV). Diastolic dysfunction (DD) was graded using American Society of Echocardiography guidelines ("conventional") and regraded by replacing left atrial volume with LARS <24% ("modified"). Relative reduction in global longitudinal strain (GLS) >15% was considered "worsening GLS," and CTRCD was defined using CMR-derived left ventricular ejection fraction.

RESULTS

Among 136 women (age 51.1 ± 9.2 years), CTRCD developed in 37 of 136 (27%) and worsening GLS in 53 of 126 (42%) with analyzable studies. Incident DD occurred in 25 (19.4%) of 129 and 19 (14.4%) of 132 patients by conventional and modified grading, respectively. Using LARS improved the ability to classify DD. Transition state analysis demonstrated that the first abnormal state during cancer therapy could be worsening GLS, DD, CTRCD, or a combination. A greater proportion of patients who first transition to DD vs worsening GLS developed subsequent CTRCD (5 of 8 [63%]) vs 7 of 39 [18%]). Worsening DD was associated with higher odds of subsequent CTRCD (OR: 20.9, 95% CI: 3.4-129.5) vs worsening GLS (OR: 4.9, 95% CI: 2.6-9.4). DD was significantly associated with radiation dose and ECV but not with blood biomarkers.

CONCLUSIONS

Patients receiving breast cancer therapy can develop significant GLS change, DD, or CMR-defined CTRCD that can occur in isolation, concurrently, or sequentially. Development of DD is associated with ECV and higher risk for subsequent CTRCD. (Evaluation of Myocardial Changes During BReast Adenocarcinoma Therapy to Detect Cardiotoxicity Earlier With MRI [EMBRACE-MRI]; NCT02306538).