Performance of the pooled cohort equations in cancer survivors: the Atherosclerosis Risk in Communities study

Background: Cancer survivors may have elevated atherosclerotic cardiovascular disease (ASCVD) risk. Therefore, we tested how accurately the American College of Cardiology/American Heart Association 2013 pooled cohort equations (PCEs) predict 10-year ASCVD risk in cancer survivors. Objectives: To estimate the calibration and discrimination of the PCEs in cancer survivors compared to non-cancer participants in the Atherosclerosis Risk in Communities (ARIC) study. Methods: We evaluated the PCEs’ performance among 1244 cancer survivors and 3849 cancer-free participants who were free of ASCVD at the start of follow-up. Each cancer survivor was incidence-density matched with up to five controls by age, race, sex, and study center. Follow-up began at the first study visit at least 1 year after the diagnosis date of the cancer survivor and finished at the ASCVD event, death, or end of follow-up. Calibration and discrimination were assessed and compared between cancer survivors and cancer-free participants. Results: Cancer survivors had higher PCE-predicted risk, at 26.1%, compared with 23.1% for cancer-free participants. There were 110 ASCVD events in cancer survivors and 332 ASCVD events in cancer-free participants. The PCEs overestimated ASCVD risk in cancer survivors and cancer-free participants by 45.6% and 47.4%, respectively, with poor discrimination in both groups (C-statistic for cancer survivors = 0.623; for cancer-free participants, C = 0.671). Conclusions: The PCEs overestimated ASCVD risk in all participants. The performance of the PCEs was similar in cancer survivors and cancer-free participants. Implications for Cancer Survivors: Our findings suggest that ASCVD risk prediction tools tailored to survivors of adult cancers may not be needed.