Late incidence and determinants of stroke after aortic and mitral valve replacement

Background. Stroke is a devastating complication in patients with prosthetic valves, but characterization of its late occurrence from a large cohort is lacking. Methods. Three thousand one hundred eighty-nine adult patients who underwent a total of 3,576 operations for left-heart valve replacement were managed with contemporary anticoagulation. guidelines and prospectively followed in a dedicated clinic. Total follow-up was 20,096 patient years. Bootstrapped survival analysis was used to determine the impact of patient and valve related factors on the incidence of stroke. Results. Most strokes were embolic. Linearized embolic stroke rates were 1.3% +/- 0.2% per year for aortic bioprostheses, 1.4% +/- 0.2% per year for aortic mechanical valves, 1.3% +/- 0.3% per year for mitral bioprostheses, and 2.3% +/- 0.4% per year for mitral mechanical valves (p = 0.002, vs other implant types). Age more than 75 years, female gender, and smoking were independent risk factors after aortic and mitral valve replacement. Atrial fibrillation, coronary disease, and tilting-disc mechanical prostheses were independent predictors of embolic stroke after aortic valve replacement. Preoperative left ventricular (LV) dysfunction was an independent risk factor in patients with mitral prostheses. Primary operative indication, diabetes, redo status, or the presence of two prosthetic valves were not associated with an increased hazard. The addition of acetyl salicylic or dipyridamole to warfarin anticoagulation did not significantly lower embolic stroke risk in patients with mechanical prostheses. Conclusions. Approximately 20% of patients with valve prostheses have an embolic stroke by 15 years after valve replacement. Some risk factors such as the avoidance of smoking, mitral mechanical prostheses, aortic tilting-disc valves, and proceeding to mitral surgery before LV dysfunction occurs are potentially modifiable. (C) 2004 by The Society of Thoracic Surgeons.