A meta-analysis of valve-in-valve and valve-in-ring transcatheter mitral valve implantation.

Objectives We performed a meta-analysis of transcatheter mitral valve implantation (TMVI) for deteriorated bioprosthetic valves (valve-in-valve [VIV]-TMVI) and/or failed annuloplasty rings (valve-in-ring [VIR]-TMVI), comparing observed early (30-day) mortality with predicted operative mortality. Background It remains unclear whether VIV/VIR-TMVI reduces mortality as compared with redo MVS. Methods MEDLINE and EMBASE were searched current through 24 July 2018 using Web-based search engines (PubMed and OVID) to identify studies including >= 10 patients undergoing VIV/VIR-TMVI. For each study, data regarding observed 30-day mortality and predicted operative mortality (Society of Thoracic Surgeons Predicted Risk of Mortality [STS-PROM]) were used to generate risk ratios (RRs) and 95% confidence intervals (CIs). Study-specific estimates were combined using the inverse variance-weighted average of logarithmic RRs in the random-effects model. One-group meta-analyses of 30-day/late (including 30-day) mortality rates were also performed in the random-effects model. Results Of 270 potentially relevant articles screened initially, 17 eligible studies including a total of 1017 patients undergoing VIV/VIR-TMVI were identified. In all but four studies, the STS-PROM was available and varied from 7.7% to 22.0% (weighted mean, 11.5%). Pooled analyses of all VIV/VIR-TMVI studies demonstrated the 30-day mortality rate of 5.4% (95%CI, 4.0-6.8%), the midterm (1- to 5-year) mortality rate of 13.7% (95%CI, 9.0-18.5%), and significantly lower observed 30-day mortality than predicted operative mortality (RR, 0.67; 95%CI, 0.49-0.91; P = 0.01). Conclusions VIV/VIR-TMVI brought about relatively low early and midterm (1- to 5-year) mortality, and observed 30-day mortality was significantly lower than predicted operative mortality.