Cerebrovascular disease is common in head and neck cancer patients, but it is unknown whether radiotherapy increases the cerebrovascular disease risk in this population.

We identified 6,862 patients (age > 65 years) from the Surveillance, Epidemiology, and End Results (SEER) -Medicare cohort diagnosed with nonmetastatic head and neck cancer between 1992 and 2002. Using proportional hazards regression, we compared risk of cerebrovascular events (stroke, carotid revascularization, or stroke death) after treatment with radiotherapy alone, surgery plus radiotherapy, or surgery alone. To further validate whether treatment groups had equivalent baseline risk of vascular disease, we compared the risks of developing a control diagnosis, cardiac events (myocardial infarction, percutaneous coronary intervention, coronary artery bypass graft, or cardiac death). Unlike cerebrovascular risk, no difference in cardiac risk was hypothesized.

Mean age was 76 +/- 7 years. Ten-year incidence of cerebrovascular events was 34% in patients treated with radiotherapy alone compared with 25% in patients treated with surgery plus radiotherapy and 26% in patients treated with surgery alone (P < .001). After adjusting for covariates, patients treated with radiotherapy alone had increased cerebrovascular risk compared with surgery plus radiotherapy (hazard ratio [HR] = 1.42; 95% CI, 1.14 to 1.77) and surgery alone (HR = 1.50; 95% CI, 1.18 to 1.90). However, no difference was found for surgery plus radiotherapy versus surgery alone (P = .60). As expected, patients treated with radiotherapy alone had no increased cardiac risk compared with the other treatment groups (P = .63 and P = .81).

Definitive radiotherapy for head and neck cancer, but not postoperative radiotherapy, was associated with excess cerebrovascular disease risk in older patients.