Background Cancer survivors have an increased risk of heart failure, but this is balanced by the risk of death from other causes. The results of this balance impact on the optimal time for guideline-recommended surveillance for cardiac dysfunction. This study aimed to investigate the association between cancer history and cardiac function at various times during follow-up. Methods This cross-sectional study included participants with documented cancer history from cancer registries and matched with non-cancer controls using propensity scoring based on age, sex, diabetes and blood pressure. Cardiac function, primarily left ventricular ejection fraction (LVEF), was assessed using cardiac magnetic resonance (CMR). Multivariable binomial regression analyses were conducted to analyse the association between cancer and cardiac function. Results Of 23 854 cancer survivors (aged 61.0±6.8 years, 60.9% female) and an equal number of matched controls, 1051 survivors and 1538 controls underwent CMR. Survivors from breast or haematological malignancies demonstrated minor differences in LVEF (59.5±6.4 vs 60.1±6.4, p<0.001) and global circumferential strain (−22.4±3.5 vs −22.6±3.5, p<0.001) compared with controls. Analysis stratified by time since cancer diagnosis revealed that both LVEF (p=0.014) and global circumferential strain (p=0.045) were less likely to be impaired with increasing time from diagnosis. Cancer survivors (prevalence ratio (PR)=1.19 (95% CI 1.05 to 1.35), p=0.006), particularly breast cancer (PR=1.39 (95% CI 1.18 to 1.64), p<0.001), were associated with low LVEF (≤55%) after adjusting for age, sex, years since cancer diagnoses and blood pressure medication. Conclusions Compared with people without cancer, cancer survivors have a higher risk of subclinical cardiac dysfunction. However, dysfunction is less common with increasing time since cancer diagnosis. These findings suggest prioritising cardiac monitoring early in survivorship, especially in breast cancer survivors.
Soh et al. (Tue,) studied this question.