Longitudinal evaluation of left ventricular diastolic function in cancer patients

Abstract

Cancer therapy-induced left ventricular diastolic dysfunction (LVDD) is gaining research attention as a potential early marker of cardiotoxicity. Radiotherapy, chemotherapy, and targeted agents may cause cardiomyocyte energy depletion, cardiac fibrosis, and microvascular dysfunction—known mechanisms of diastolic dysfunction. Heart failure (HF) with preserved ejection fraction and LVDD remain understudied in oncological settings. We aimed to assess the changes in LV diastolic function caused by cancer treatment. We prospectively enrolled 60 adults planned for chemotherapy (CT) for breast (n = 50) or gastrointestinal (n = 10) cancers. Exclusion criteria were fewer than 3 visits, a poor acoustic window, or pre-existing cardiac or chronic pulmonary disease. We performed 2D echocardiography at baseline and 1, 3, 6, 9, 12, and 18 months after CT initiation. Left ventricular ejection fraction (LVEF) was considered abnormal if it dropped by over 10% to below 50%. Longitudinal LV systolic dysfunction was defined as a decrease of over 15% to below 7 cm/s in systolic tissue S′ velocity, averaged from septal and lateral S′. We assessed LVDD using the 2025 American Society of Echocardiography algorithm and the parameters: indexed left atrial volume (LAVi); early (E) and late (A) mitral inflow velocities; E/A ratio; isovolumic relaxation time (IVRT); pulmonary artery systolic pressure (PASP); early diastolic tissue e′ velocity (average of septal and lateral e′) (LVe′); and E/e′ ratio. The temporal parameters' changes were estimated using generalised estimating equations (GEE) regression. Correlations were examined using the Spearman and chi-squared tests and Goodman-Kruskal's gamma (γ) statistic. The mean patient age was 53 years (SD 11.6), with 91.7% female. Half (n = 30) received anthracycline-based CT, either with or without trastuzumab; the remainder received non-anthracycline CT, with or without trastuzumab. The median CT duration was 98 days (IQR: 64, 112). Twenty-seven patients (45%) had radiotherapy. Over a median follow-up of 381 days (IQR: 187, 464), LVE/e' and LVe' declined significantly in the third month. Both parameters showed moderate correlation with LVS’ (Figure 1). Six patients had baseline LVDD; in one, it progressed from grade 1 to grade 2. New-onset LVDD occurred in 6 patients (10%), all grade 1 (Figure 2). Among them, two had transient LVDD; in two cases, LVDD advanced to grade 2, and one patient showed symptoms of HF (NYHA class II). There was a trend toward LVDD at the first and sixth months (Figure 2). No new-onset LVDD patient developed abnormal LVEF, although two had concurrent longitudinal systolic dysfunction. LVDD and longitudinal systolic dysfunction showed a positive association (ꭓ² = 5.2, p = 0.022; Kruskal’s γ = 0.64). Cancer therapy-related LVDD may develop early and present as either a transient or persistent condition. Its clinical implications need further investigation.