LAUSR

BACKGROUND The association between stroke and cancer is well-known but insufficiently investigated. Aim of this multicentre retrospective cohort study was to estimate the prevalence of cancer-associated ischemic stroke, describe clinical outcomes in patients with cancer-associated ischemic stroke and investigate independent factors associated with active cancer. METHODS Consecutive adult patients admitted for acute ischemic stroke were included. Included patients were admitted in the Stroke Unit of the Hospital of Perugia, Italy, from March 2005 to March 2015, and in a medical unit of the Hospital of Varese, Italy, from January 2010 till December 2011. Clinical and laboratory data of patients with and without active cancer were collected. Multivariate logistic regression analysis was performed to identify independent factors associated with active cancer. RESULTS A total of 2209 patients admitted with acute ischemic stroke were included with a median hospital stay of 9 days (interquartile range 5.75-14). Mean age was 72.7 years (standard deviation +/- 13); 55% patients were male and 4.4% had active cancer. Factors significantly associated with the presence of active cancer were age > 65 years (odds ratio [OR] 3.34; 95% confidence interval [CI] 1.64-6.81), occurrence of venous thromboembolism [VTE] (OR 2.84; 95% CI 1.12-7.19), low-density lipoprotein (LDL) cholesterol level < 70 mg/dL (OR 1.92; 95% CI 1.06-3.47), cryptogenic stroke subtype (OR 1.93; 95% CI 1.22-3.04). Overall mortality rate during hospital stay was greater in patients with active cancer (21.5% vs. 10% P < 0.05). CONCLUSIONS Older age, occurrence of VTE, low LDL level, and cryptogenic stroke subtype, are independently associated with active cancer. Overall, our findings suggest a possible prevalent role of hypercoagulability in the pathogenesis of cancer-associated ischemic stroke.