By Jeffrey Zimmet, MD, PhD

Associate Professor of Medicine, University of California, San Francisco; Director, Cardiac Catheterization Laboratory San Francisco VA Medical Center

SYNOPSIS: This analysis of mitral surgery after failed transcatheter edge-to-edge repair demonstrates high rates of valve replacement as opposed to repair. Surgical mortality was higher than predicted but was significantly lower in high-volume centers.

SOURCE: Chikwe J, O’Gara P, Fremes S, et al. Mitral surgery after transcatheter edge-to-edge repair: Society of Thoracic Surgeons database analysis. J Am Coll Cardiol 2021;78:1-9.

In 2013, the FDA approved transcatheter edge-to-edge repair (TEER) to treat severe degenerative mitral regurgitation in patients at high and prohibitive risk for surgery. In 2019, this approval expanded to include a subset of patients with heart failure and significant functional mitral regurgitation despite optimal medical therapy. The use of the MitraClip system has grown; to date, approximately 15,000 of these procedures have been performed in the United States. As a minimally invasive approach, TEER can be remarkably effective in the treatment of mitral disease. Yet, recent real-world retrospective analyses have shown between 20% and 30% of patients are left with residual or recurrent moderate or severe mitral regurgitation within a year of the procedure. In addition, some fraction of patients trade regurgitation for stenosis.

Some patients with unsatisfactory results ultimately go to surgery. Considering patients gain entry to TEER partially by being judged as high risk for surgery, one would assume this would remain true for surgery after a failed transcatheter approach. Chikwe et al mined the Society of Thoracic Surgeons (STS) Adult Cardiac Surgery Database to identify and characterize patients who have undergone surgical reintervention after TEER. During the period from June 2014 to July 2020, 524 such patients were confirmed. After exclusion of emergencies and patients with prior mitral surgery, a cohort of 463 patients was available for analysis.

The average age of patients was 76 years, and half were women. The annual number of such surgeries increased year over year, with 32 procedures in 2015 and 126 in 2019. One of the most striking results was that only 22 patients underwent mitral valve repair; the remainder underwent valve replacement, with a bioprostheses in most cases. Just over half of patients underwent extra procedures in addition to mitral valve surgery. The remainder (48.8%) underwent isolated mitral valve surgery. Add-on procedures included concomitant tricuspid repair or replacement in 152 patients for moderate or severe tricuspid regurgitation, and coronary artery bypass grafting in 57 patients. In the overall cohort of patients undergoing reintervention for failed MitraClip, the mortality rate within 30 days was 10.6%. Among the subset of patients undergoing isolated mitral valve surgery, the median STS predicted risk of mortality was 6.5%, while the observed mortality was 10.2%.

A total of 357 different surgeons performed the included procedures at 227 unique hospitals. An analysis of mortality by operative case volume revealed only centers in the highest quintile of volume (> 10 cases) recorded a significantly lower operative mortality rate (2.6%, or two of 76). Operative mortality was 12.4% (n = 64 of 515) in centers that performed fewer procedures.

Notably, 177 patients were labeled with underlying degenerative mitral valve disease. Successful mitral valve repair in this group was similar (n = 12) to the entire cohort. Although the predicted risk of mortality was 7.3% in this group, the observed mortality was somewhat lower at 6.2% (n = 11 of 175). Mortality was significantly lower in those surgeries that were labeled as elective (n = 3 of 108, 2.8%) as opposed to urgent. The authors concluded surgery after failed MitraClip was uncommon but increased steadily during the study period and was associated with low rates of successful mitral repair. The highest-volume centers achieved significantly lower mortality than others.


It should come as no surprise that observed mortality was high after surgical reintervention for patients who, in theory, were at high risk for surgery in the first place. The data collection from the STS database leaves many gaps that would be pertinent to this analysis. For example, no data are available on the timing of surgery relative to the TEER procedures, the number of clips involved, or whether the surgical treatment was for mitral regurgitation or stenosis. Most patients who were classified with non-degenerative mitral regurgitation had no further information as to the etiology of the regurgitation. Importantly, only patients who underwent surgery were included, so no conclusions can be drawn about the presumably larger population of patients with failed TEER who were turned down for surgery.

Also notable was the rate of successful valve repair, as opposed to replacement, was extremely low. This was true in patients whose initial valve pathology was listed as degenerative, as well as in the entire population. Surgery as the upfront treatment (not after unsuccessful TEER) for degenerative mitral regurgitation should be expected, in experienced centers, to be successful about 95% of the time. This is considered the gold standard for treatment of this subset of mitral regurgitation patients. The relatively low operative mortality (2.8%) seen in the subset of degenerative regurgitation patients undergoing elective procedures certainly calls into question whether these patients were characterized correctly as high surgical risk to begin with. The low rate of subsequent repair reminds us that even successful surgery after failed TEER results in a second-best outcome of valve replacement. It is in this context that the current data can inform patient selection for TEER and may play a role in patient consent. For patients with failed TEER who do require surgery, strong consideration should be given to referral to experienced centers.