BACKGROUND: Cerebral vascular events (CVA) have been documented following clinical trials of patients with advanced hormone-refractory prostate cancer, and safety warnings about increased risk of stroke and other cardiovascular diseases in advanced prostate cancer patients were recently added to labels of gonadotropin-releasing hormone agonists. Published data on the incidence of CVAs in a general population of men with various stages of prostate cancer is lacking.
OBJECTIVES: To estimate the incidence of new onset CVAs after diagnosis of prostate cancer among male US Medicare enrollees aged 65 years and in similarly-aged men without prostate cancer.
METHODS: Men with prostate cancer diagnosed from 1999– 2005 were identified from the Surveillance, Epidemiology, and End Results (SEER)-Medicare linked database (n¼77,110). CVA events were ascertained from Medicare data through 2007 and defined as first hospitalization with a primary diagnosis of stroke (ICD-9-CM 430, 431, 432.9, 433.x1, 434.x1, or 436) or death from stroke after diagnosis of prostate cancer. Deaths could not be ascertained in nonprostate cancer comparison subjects. The incidence rate (IR) per 1,000 person-years (PY) was estimated overall and by baseline age, comorbidity score, stage, and castration status (surgery or hormone therapy) in follow-up.
RESULTS: The IR of CVA (hospital events only) per 1,000 PY in men with prostate cancer was 8.7 (95% CI 8.4–9.0) and in age-matched comparison subject was 9.1 (95% CI 8.9–9.3). The IR Ratio, adjusted for race, geographic region, and comorbidity score, was 0.95 (95% CI 0.91–0.99). The overall IR of CVA (hospital events and deaths) per 1,000 PY in men with prostate cancer was 9.5 (95% CI 9.2, 9.9), and was higher in stage IV patients, 13.8 (95% CI 12.2–15.4) than in stages I–III combined, 9.2 (95% CI 8.9–9.6). The IR per 1,000 PY was greater for men with a high baseline comorbidity score (18.7, 95% CI 17.1–20.6), and for those on castration therapy (15.1, 95% CI 13.5–16.9).
CONCLUSIONS: Advanced prostate cancer patients have an increased risk of CVA. These data can inform baseline rates and aid in prevention efforts.
