Typical cardiovascular risk factors like age and comorbidities overshadow cancer-specific treatments in predicting late cardiovascular disease among older long-term cancer survivors, new research shows.
The retrospective cohort study involved 95,100 survivors of breast, prostate, colon and rectal cancers in the US aged 66 and older at the time of diagnosis, who were followed for 5-15 years post diagnosis using the linked SEER-Medicare database.
Late CVD – defined as myocardial infarction, stroke, congestive heart failure, or cardiomyopathy five or more years after cancer diagnosis – occurred in 23.2% of survivors.
Findings published in the Journal of the National Cancer Institute [link here] showed the risk of late CVD was strongly associated with non-cancer-related factors, including a prior acute cardiovascular event, advanced age, and comorbidities such as diabetes and hypertension present at cancer diagnosis.
Cancer-specific exposures were not associated with late CVD, except for stage III breast cancer, and radiation plus androgen deprivation therapy for prostate cancer.
The researchers created risk scores that included age, comorbidities, and prior CVD for all survivor cohorts; with the addition of high Gleason grade, higher area-level poverty, radiation late treatment, and marital status for the risk score for prostate cancer; stage III disease and oestrogen receptor–positive disease for breast cancer; and male sex (versus female) for colon and rectal cancers.
Risks scores suggested 3-4 fold differences in CVD incidence, ranging from 13.4% for breast cancer and 22.0% for rectal cancer across low-risk groups versus 43.3% for prostate cancer and 53.3% for colon cancer across high-risk groups.
In the high-risk stratum of all survivor cohorts, the risk of late CVD outweighed that of non-CVD mortality, the researchers noted.
“Surprisingly, despite their known cardiotoxic effects, cancer treatments such as chemotherapy, radiotherapy, and hormone therapy were not independently associated with late CVD events in this cohort, irrespective of an early CVD event. This suggests that beyond five years from diagnosis, cardiovascular risk prognostication in older, long-term cancer survivors may be best served by cancer agnostic models,” said the Yale University-led researchers.
They explained that other work had suggested a peak in CVD risk related to treatment within 1-2 years of a cancer diagnosis, and that differences in annual incidence of CVD by treatment might converge around five years after diagnosis.
“Placing our findings into context with prior literature, the lack of association between cancer-specific risk factors and CVD points to risks specific to an older population,” they said.
Writing in an accompanying editorial [link here], US cardio-oncology researchers said while the study did not negate the relevance of treatment-related toxicity, particularly in younger populations, it reframed long-term survivorship care.
“The 5-year milestone represents a critical transition point, when oncology follow-up diminishes and responsibility shifts toward primary care. Embedding structured cardiovascular assessments—including evaluation of hypertension, diabetes, and lipid status—at this juncture may improve outcomes,” they said.
“High-risk groups, defined by age, prior CVD, and comorbidity burden, should be prioritised for cardio-oncology or cardiology co-management. Health in low-risk survivors can be effectively managed in primary care, easing workforce pressures.”
The editorial also stressed that the study had not included modern targeted cancer therapies such as immune checkpoint inhibitors or small molecule inhibitors, for which the data on long-term cardiovascular consequences were limited.