Risk-corrected impact of mechanical versus bioprosthetic valves on long-term mortality after aortic valve replacement

Choice of a mechanical or biologic valve in aortic valve replacement remains controversial and rotates around different complications with different time-related incidence rates. Because serious complications will always "spill over" into mortality, our aim was to perform a meta-analysis on overall mortality after aortic valve replacement from series with a maximum follow-up of at least 10 years to determine the age- and risk factor-corrected impact of currently available mechanical versus stented bioprosthetic valves.

Following a formal study protocol, we performed a dedicated literature search of publications during 1989 to 2004 and included articles on adult aortic valve replacement with a mechanical or stented bioprosthetic valve if age, mortality statistics, and prevalences of well-known risk factors could be extracted. We used standard and robust regression analyses of the case series data with valve type as a fixed variable.

We could include 32 articles with 15 mechanical and 23 biologic valve series totaling 17,439 patients and 101,819 patient-years. The mechanical and biologic valve series differed in regard to mean age (58 vs 69 years), mean follow-up (6.4 vs 5.3 years), coronary artery bypass grafting (16% vs 34%), endocarditis (7% vs 2%), and overall death rate (3.99 vs 6.33 %/patient-year). Mean age of the valve series was directly related to death rate with no interaction with valve type. Death rate corrected for age, New York Heart Association classes III and IV, aortic regurgitation, and coronary artery bypass grafting left valve type with no effect. Included articles that abided by current guidelines and compared a mechanical and biologic valve found no differences in rates of thromboembolism.

There was no difference in risk factor-corrected overall death rate between mechanical or bioprosthetic aortic valves irrespective of age. Choice of prosthetic valve should therefore not be rigorously based on age alone. Risk of bioprosthetic valve degeneration in young and middle-aged patients and in the elderly and old with a long life expectancy would be an important factor because risk of stroke may primarily be related to patient factors.