Prediction of Ischemic Heart Disease and Stroke in Survivors of Childhood Cancer
The Childhood Cancer Survivor Study (CCSS) led to the largest cancer survivor follow-up cohort. It is based on data extracted from medical records and questionnaires that were sent to childhood cancer survivors (CCS). CCS have been shown to have a >10-fold increased risk of ischemic heart disease and stroke compared with siblings. This study developed a risk-prediction model using end-of-treatment data to risk stratify CCS and externally validate the model.
This manuscript used data from the CCSS from children diagnosed between 1970 and 1986 at 26 institutions in the United States and Canada (n=13,060). A random sample of siblings (n=4,023) served as comparison cohort. Two validation datasets were used: the SJLIFE cohort from St. Jude Children’s Research Hospital (led by the same team as the CCSS cohort) and the EKZ/ AMC cohort (unrelated group, n= 1,300 - 1,800 CCS). A set of variables with demographic and treatment data were tested for their association with ischemic heart disease and stroke.
Predictors of ischemic heart disease were male sex, neck or chest radiation; predictors for stroke were alkylator therapy, platinum agents, cranial and chest radiation. Two models were created: a simple model where cancer therapy–related exposures were categorized as yes or no only; and a standard model where clinical dose information was known. Risk scores were attributed and summed to create low-, moderate-, and high-risk groups. Low-risk groups had cumulative incidence risks of =15% of ischemic heart disease or stroke, respectively.
The cohorts used in this study were diagnosed three or more decades ago (training data set) or at least 15 years ago (validation data sets). Therefore, these findings might not be valid for modern treatment protocols. Particularly radiation was greatly reduced and application techniques changed within the last decades.