Transcatheter mitral valve replacement is increasingly being used as a treatment for high-risk patients who have native mitral valve disease; however, no comprehensive studies on its effectiveness have been reported. We therefore searched the literature for reports on patients with native mitral valve disease who underwent transcatheter access treatment.
We found 40 reports, published from September 2013 through April 2017, that described the cases of 66 patients (mean age, 71 ± 12 yr; 30 women; 30 patients with mitral stenosis, 34 with mitral regurgitation, and 2 mixed) who underwent transcatheter mitral valve replacement. We documented their baseline clinical characteristics, comorbidities, diagnostic imaging results, procedural details, and postprocedural results.
Access was transapical in 41 patients and transseptal in 25. The 30-day survival rate was 82.5%. The technical success rate (83.3% overall) was slightly but not significantly better in patients who had mitral regurgitation than in those who had mitral stenosis. Transapical access procedures resulted in fewer valve-in-valve implantations than did transseptal access procedures (P=0.026).
These current results indicate that transcatheter mitral valve replacement is feasible in treating native mitral disease. The slightly higher technical success rate in patients who had mitral regurgitation suggests that a valve with a specific anchoring system is needed when treating mitral stenosis. Our findings indicate that transapical access is more reliable than transseptal access and that securely anchoring the valve is still challenging in transseptal access.
Transcatheter aortic valve replacement (TAVR) has been a successful treatment for severe symptomatic aortic stenosis in older patients with comorbidities who are at high surgical risk.1 Consequently, the use of transcatheter mitral valve replacement (TMVR) for treating native mitral regurgitation (MR) and mitral stenosis (MS) has increased. Mitral valve (MV) repair or replacement is the gold standard for treating mitral disease, but approximately half of patients are at high surgical risk.2–4 Independent risk factors of 30-day postoperative death are New York Heart Association (NYHA) functional class IV status, diabetes mellitus, hypertension, renal insufficiency, rheumatic causes, and depressed left ventricular ejection fraction (<45%).5 When surgery is risky, TMVR may be an option. Because comprehensive data on current clinical outcomes of TMVR are not available, we reviewed the medical literature and gathered information on the clinical, anatomic, and periprocedural characteristics of TMVR cases. We then compared clinical outcomes when MR or MS was treated by means of transapical (TA) access or transseptal (TS) access.
Patients and Methods
We systematically searched all English-language articles from January 2000 through April 2017 in PubMed and Web of Science that described TMVR, using the search terms TMVI OR transcatheter mitral valve OR transcatheter mitral valve replacement OR transcatheter mitral valve implantation. Articles were excluded if they were not in English, focused on animal experiments, lacked relevant information on TMVR, had inadequate details on postoperative outcomes, involved valve-in-valve or valve-in-ring mitral implantation, or involved a thoracotomy approach (except conversion to thoracotomy intraoperatively).
We collected data on baseline clinical characteristics, relevant comorbidities, diagnostic imaging results, procedural details, and postprocedural outcomes. Technical success was defined in accordance with Mitral Valve Academic Research Consortium criteria: no procedural death; successful access, delivery, and retrieval of the device delivery system; successful deployment and correct positioning of the first intended device; and no emergency surgery or reintervention related to the device or access procedure.6 Continuous variables were described as mean ± SD; differences between them were analyzed by using t tests. Categorical variables were described as number and percentage, and the χ2 test was used to evaluate differences. Mean gradients were derived from <80% of the reports, and NYHA class from <50%; other variables were from >80%. Survival curves were estimated by using the Kaplan-Meier method. All data were analyzed with use of SPSS 23.0 for Windows (SPSS, an IBM company). P values <0.05 were considered statistically significant.
We found 40 reports, published from September 2013 through April 2017, with case descriptions of patients who had native-valve MR or MS and underwent TMVR (Table I).7–46 The 66 patients' mean age was 71 ± 12 years; 75% of those whose sex was specified were women (41 of 55); and 100% were in NYHA functional class III or IV. Mitral stenosis was predominant in 30 patients and MR in 34, and 2 patients had mixed native mitral disease. Forty-one procedures involved TA access, and 25 involved TS access. Twenty-nine patients with predominant MR were given one of the following transcatheter MV platforms: Edwards-CardiAQ (Edwards Lifesciences Corporation), Tendyne (Abbott Cardiovascular), Neovasc Tiara (Neovasc Inc.), or Edwards Fortis (Edwards Lifesciences). The 30 patients with predominant MS were given Sapien, Sapien XT, or Sapien 3 balloon-expandable transcatheter aortic valves (Edwards Lifesciences).
The median follow-up time was 2 months, and the longest was 20 months (Fig. 1). The overall mean survival time was 13.07 months (95% CI, 10.24–15.89 mo). The mean survival time for MR patients was 9.8 months (95% CI, 7.5–12.09 mo); and for MS patients, 13.17 months (95% CI, 9.34–17.01 mo). Median survival could not be calculated because of the limitation of reported follow-up time. For all patients, the technical success rate was 83.3% (55 of 66 cases), and the 30-day survival rate was 82.5% (47 of 57) (Table II).
Complications and Deaths
Six patients needed a second valve: 4 intraoperatively for severe regurgitation, paravalvular leak, or initial valve displacement; and 2 postoperatively after prosthesis migration. Of these 6, 5 underwent successful valve-in-valve implantation, and the remaining patient underwent urgent open surgical repair (Table III).
In the 23 patients who had postoperative complications, the most frequent was migration of the prosthetic MV. In 5 patients, this happened from 4 days to 8 months later, and one of these patients died. The remaining patients survived after open surgery, valve redeployment, or implantation of a second valve (Table IV).
Deaths. Two patients died intraoperatively, one of apical perforation from the delivery system's nose cone with consequent cardiac tamponade, and the other of cardiogenic shock. Of the 16 patients who died postoperatively (time range, 12 hr–9 mo), 8 died of cardiac causes. A patient who had no complications died of fractured cervical vertebrae; this death was not documented in Table III or IV.
Mitral Pathologic Conditions and Access Routes
The causes and pathophysiology of MS and MR differed. All 30 patients who had predominant MS had mitral annular calcification (MAC) visible on echocardiograms (100%), compared with 5 patients who had predominant MR (14.7%) (P <0.001). The mean mitral transvalvular gradient was significantly lower in the patients with MR (3.2 ± 1.3) than in those with MS (4.3 ± 1.6) (P=0.026). The results were otherwise similar between groups (Table II).
Fewer patients who had TA procedures (1 [2.4%]) needed valve-in-valve implantation than did those who had TS procedures (5 [20%]) (P=0.026). Otherwise, the results in regard to approach were similar (Table V).
The technical success rate of TMVR was 83.3% overall, the 30-day survival rate was 82.5%, and 23 patients had postoperative complications, chiefly valve migration. To date, the results of TMVR have not been as successful as those of MV surgery.47 Nevertheless, TMVR is a feasible option in high-risk patients who have native MV disease.
Puri and colleagues48 summarized the clinical, anatomic, periprocedural, and postprocedural characteristics of 11 patients with severely calcified MVs but discussed neither the role of TMVR in patients with varying types of MV disease nor novel devices specifically designed for treating noncalcific MR. Conversely, we compared the role of TMVR in patients with MR and MS, documenting the different causes, pathophysiology, and valve types. Severe MS was caused by MAC, a condition that can enable stable anchoring of balloon-expandable transcatheter aortic valves.49 In patients who do not have MAC, an anchoring system is needed. Accordingly, the patients with MS and the 5 with MR and MAC were given balloon-expandable transcatheter aortic valves, and transcatheter MV platforms were used in the patients who had MR but not MAC. Better technical success was achieved in patients with MR. Reasons for technical failure were incorrect valve positioning, valve migration, major bleeding, and apical perforation. Incorrect positioning and early migration caused valve embolism, left ventricular outflow tract obstruction, or perivalvular leakage. The 30-day survival rates of patients with MR (82.8%) and MS (80.8%) were similar, as were the results for the other outcome variables evaluated. Valve type and different baseline characteristics had no significant impact on outcome.
Transapical TMVR access is achieved through a minithoracotomy and has a shorter path. The direct access to the MV and the shorter distance between the introducer tip and the MV enable better control of the prosthesis during deployment. Transseptal access is much less invasive; however, stent anchorage occurs in a more complex geometric environment.50 Among transcatheter MV platforms, only the Edwards-CardiAQ valve has been implanted through both access routes; the others have been implanted only through TA access.45 Fewer patients treated by means of TA access needed a second valve implantation (P=0.026), possibly because TS access involves a longer path and anchoring is more difficult. Transcatheter mitral valve-in-valve implantation, performed when mitral bioprostheses degenerate, was successful in 5 of 6 patients; the 6th was converted to open surgery because of valve embolism. Technical success was slightly better in patients who underwent TA access, and postoperative complications were more frequent in patients who underwent TS access.
Because data consistency and completeness inevitably varied across reports, our main challenge was publication bias. Bias also resulted from the small number of cases, different standards in different centers, and different valves used. In addition, case reports and series mixed different valve types, pathologic conditions, and approaches. Cases were too few for subgroup analysis. Nevertheless, this review enabled objective conclusions about TMVR in treating native MV disease.
In high-risk patients who have MR and MS, TMVR is generally feasible. Less technical success in patients with MS implies that valves with specific anchoring systems are needed. The TA approach resulted in slightly better technical success and fewer postoperative complications. Comparatively more patients treated by means of TS access underwent a second valve implantation. Accurate valve fixation in TS access remains a challenge.
This work was supported by the Natural Science Foundation of China grant no. 81670463 and Shanghai Science & Technology Committee Foundation grant no. 16441907800.