Meta-analysis of valve hemodynamics and left ventricular mass regression for stentless versus stented aortic valves.

BACKGROUND Stentless aortic bioprostheses have been advocated as being superior to conventional bioprosthetic valves, with benefits including superior left ventricular mass regression and larger effective orifice area. Several high-quality randomized studies now exist on this topic, and we sought to summarize them by meta-analysis. METHODS The literature was searched from 1995 to 2006, in MEDLINE, EMBASE, CRISP, metaRegister of Controlled Trials, and the Cochrane database. Experts were also contacted and reference lists searched. Studies were combined using the inverse variance fixed-effects model. Heterogeneity was assessed and a sensitivity analysis performed. Publication bias was also investigated. RESULTS Ten studies were identified that included 919 patients in which the Freedom (Sorin Biomedica Cardio, Via Crescentino, Italy), Freestyle (Medtronic, Minneapolis, MN), Prima Plus (Edwards Life Sciences, Irvine, CA) and the Toronto and Biocor (St Jude Medical, St. Paul, MN) valves were used. The mean aortic valve gradient was lower in the stentless groups, with a weighted mean difference (WMD) of -3.57 mm Hg (95% confidence interval [CI], -4.36 to -2.78; p < 0.01). The left ventricular mass index was significantly lower in the stentless groups at 6 months (WMD, -6.42; 95% CI, -11.63 to -1.21; p = 0.02), but this improvement disappeared after 12 months (WMD, 1.19; 95% CI, -4.15 to 6.53; p = 0.66). The weighted mean increase in cross-clamp time was 23 minutes, and the increase in bypass time was 29 minutes with a stentless valve. CONCLUSIONS This meta-analysis showed that stentless aortic valves provide an improved level of left ventricular mass regression at 6 months, reduced aortic gradients, and an improved effective orifice area index, at the expense of a 23-minute longer cross-clamp time and a 29-minute longer bypass time.