Thromboembolic events after aortic valve replacement in elderly patients with a Carpentier-Edwards Perimount pericardial bioprosthesis.

OBJECTIVES

Thromboembolic events after aortic valve replacement with a bioprosthesis were the most frequently occurring complications in elderly patients. Whether this was valve related or dependent on other factors needed further exploration.

METHODS

Five hundred patients with a median age of 73 years were followed retrospectively after aortic valve replacement with a pericardial prosthesis for occurrence of thromboembolism. Of these, 348 also underwent coronary artery bypass grafting. Twenty-five factors were investigated for their potential effect by using univariate and multivariate analysis.

RESULTS

Univariate analysis revealed 6 significant factors: preoperative endocarditis (P =.0001), preoperative cerebrovascular accident (P =.002), use of postoperative warfarin sodium (Coumadin, DuPont Merck; P =.006), arterial hypertension (P =.023), size of valve prosthesis of 27 mm or larger (P =.023), and hospital thromboembolism (P =.040). There was a trend toward increased fatal thromboembolism in patients without medication. With a multivariate analysis, 4 factors remained significant: preoperative cerebrovascular accident (risk ratio, 4.8; P =.0016), warfarin sodium (risk ratio, 3.0; P =.0028), preoperative endocarditis (risk ratio, 5.6; P =.006), and hospital thromboembolism (risk ratio, 6.1; P =.016). Hypertension had a borderline effect. Age, sex, diabetes, 4 coronary artery factors, 3 other valvular factors, atrial fibrillation, and carotid artery disease had no significant effect.

CONCLUSIONS

Some emboli seemed triggered by the valve prosthesis. A proper anticoagulant protocol but also a treatment of hypertension is important in the prevention of thromboembolism after aortic valve replacement with a bioprosthesis. We did not find a significant role of atrial fibrillation and carotid artery disease.