SAVR and TAVI comparison across the globe based on current regional registry evidence - A meta-analysis of reconstructed time-to-event data.

Background: There is debate whether surgical aortic valve replacement (SAVR) or transcatheter implantation (TAVI) provide better results for treatment of aortic valve stenosis. While randomized clinical trials (RCTs) are considered to compare the average treatment effect of two methods in a selected patient population, registry data, although biased, reflect every day clinical practice and provide external validation of RCTs. We evaluated the impact of SAVR or TAVI on long-term survival based on local reports from all available regions in the world. Methods: We systematically searched three databases selecting risk-adjusted registry studies comparing outcomes for SAVR and TAVI with at least five years of follow-up. Reports without all-cause mortality were excluded. One time-to-event curve was reconstructed from survival curves. Cox regression model and sensitivity analysis were performed. Results: From 10,399 screened studies, 13 met the inclusion criteria with 28,344 patients in the final analysis (follow-up: 5-9 years). In ten studies, survival rates favored SAVR, three showed no difference and none favored TAVI. Hazard ratio (HR) for overall survival was 0.58 (95 %CI: 0.54-0.61, p < 0.01 - favors SAVR). A landmark analysis with a 6-months split showed no significant survival difference in the first 6 months (HR: 0.94, 95 %CI: 0.86-1.02, p = 0.14) and better survival for SAVR compared to TAVI thereafter (HR: 0.43, 95 %CI: 0.40-0.46, p < 0.01). All sensitivity analyses supported this outcome. Conclusions: This systematic regional registry-type comparison revealed that SAVR is associated with increased long-term survival compared to TAVI, which appears to be independent of the world region in which the study was performed.