Population-Based Long-Term Cardiac-Specific Mortality Among 34 489 Five-Year Survivors of Childhood Cancer in Great Britain

Abstract

Background: Increased risks of cardiac morbidity and mortality among childhood cancer survivors have been described previously. However, little is known about the very long-term risks of cardiac mortality and whether the risk has decreased among those more recently diagnosed. We investigated the risk of long-term cardiac mortality among survivors within the recently extended British Childhood Cancer Survivor Study.

Methods: The British Childhood Cancer Survivor Study is a population-based cohort of 34 489 five-year survivors of childhood cancer diagnosed from 1940 to 2006 and followed up until February 28, 2014, and is the largest cohort to date to assess late cardiac mortality. Standardized mortality ratios and absolute excess risks were used to quantify cardiac mortality excess risk. Multivariable Poisson regression models were used to evaluate the simultaneous effect of risk factors. Likelihood ratio tests were used to test for heterogeneity and trends.

Results: Overall, 181 cardiac deaths were observed, which was 3.4 times that expected. Survivors were 2.5 times and 5.9 times more at risk of ischemic heart disease and cardiomyopathy/heart failure death, respectively, than expected. Among those >60 years of age, subsequent primary neoplasms, cardiac disease, and other circulatory conditions accounted for 31%, 22%, and 15% of all excess deaths, respectively, providing clear focus for preventive interventions. The risk of both overall cardiac and cardiomyopathy/heart failure mortality was greatest among those diagnosed from 1980 to 1989. Specifically, for cardiomyopathy/heart failure deaths, survivors diagnosed from 1980 to 1989 had 28.9 times the excess number of deaths observed for survivors diagnosed either before 1970 or from 1990 on.

Conclusions: Excess cardiac mortality among 5-year survivors of childhood cancer remains increased beyond 50 years of age and has clear messages in terms of prevention strategies. However, the fact that the risk was greatest in those diagnosed from 1980 to 1989 suggests that initiatives to reduce cardiotoxicity among those treated more recently may be having a measurable impact.

Keywords: mortality; neoplasms; pediatrics.

Figure 1.

Cumulative mortality for cardiac death by first primary neoplasm (FPN) diagnostic groups, by follow-up (years since diagnosis), compared with that expected. HL indicates Hodgkin lymphoma.

Figure 2.

Absolute excess risk per 10 000 person-years for cardiomyopathy/heart failure (CM/HF) deaths for each calendar year of diagnosis. The curve was produced by fitting a restricted cubic spline with the number of CM/HF deaths per 1 calendar year band as a weight. Point estimates of the absolute excess risk with corresponding 95% confidence intervals are plotted at the mean of the relevant treatment era (1940–1969, 1970–1979, 1980–1989, and 1990–2006).