Systematic review and meta-analysis of chordal replacement versus leaflet resection for posterior mitral leaflet prolapse

Objectives To compare outcomes of chordal replacement versus leaflet resection techniques for repair of isolated posterior mitral leaflet prolapse. Methods We searched MEDLINE and EMBASE databases for studies that compared chordal replacement (“neo-chord” group) versus leaflet resection (“resection” group) techniques for the treatment of posterior mitral leaflet prolapse. Data were extracted by 2 independent investigators and subjected to a meta-analysis using a random-effects model. Results One randomized controlled trial (RCT), 1 propensity-matched study, and 6 unadjusted observational studies, with a total of 1926 patients, met our inclusion criteria. Two studies reported only perioperative outcomes; mean follow-up ranged from 1.0 to 5.9 years in the remaining studies. In pooled data from unadjusted observational studies, annuloplasty ring diameter was higher in the neo-chord group (+1.5 mm; P =.0003), but with high heterogeneity (I² = 91%). Based on limited data, postprocedural left ventricular ejection fraction may be greater in the neo-chord group, but this difference reached statistical significance only in the RCT (+3.4%; P =.03), and not in 2 observational studies that reported this outcome (+2.7%; P =.10). There was no difference in recurrent mitral regurgitation at follow-up between the resection and neo-chord groups; however, patients in the neo-chord group had a lower rate of mitral valve reoperation at follow-up in the unadjusted observational studies (incidence rate ratio, 0.22; P =.0008 [I² = 0%; 4 studies, 1331 patients]). Conclusions Chordal replacement may be associated with greater freedom from reoperation and may lead to improved postoperative left ventricular function compared with leaflet resection. However, these conclusions are supported primarily by data from unadjusted observational studies, and high-quality RCTs of chordal replacement versus leaflet resection are needed.