To the Editor: Mitral valve repair is the preferred treatment for mitral regurgitation (MR). However, failure or recurrence of MR might occur. Re-operation might carry a high risk in patients with comorbidities. A few reports have suggested the feasibility of transcatheter mitral valve replacement after failure of surgical ring annuloplasty (1). We report here the first series of transvenous mitral valve-in-ring implantation (TVIR).

After evaluation and decision by a heart team, procedures were performed under general anesthesia and 2-dimensional and 3-dimensional transesophageal echocardiographic (TEE) guidance. Transseptal puncture was done at the high and posterior part of the fossa ovalis. An Inoue wire (Toray Medical Company, Chiba, Japan) was placed in the left atrium, and septal dilation was performed with Wanda balloons from 12 to 14 mm diameters (Boston Scientific, Galway, Ireland). Then, the mitral valve was crossed with either a Critikon balloon wedge pressure catheter (Arrow International, Inc., Reading, Pennsylvania) or a JR 4 catheter advanced on a 0.035-inch J wire. After placing a J pre-shaped 0.035-inch Amplatz ExtraStiff wire (Cook Medical, Bloomington, Indiana) at the apex of the left ventricle, a pre-dilation of the mitral valve was performed if judged necessary. A 23- or 26-mm SAPIEN XT (Edwards Lifesciences, Inc., Irvine, California) valve was mounted, upside down, on a Novaflex catheter and advanced to the mitral ring. The prosthesis was directed toward the mitral valve by full flexion of the catheter. Then, the SAPIEN XT valve was placed within the mitral ring, its position checked by fluoroscopy and TEE, and deployed by progressive balloon inflation under rapid pacing (160 to 200 beats/min). Post-operatively, patients received a combination of aspirin and clopidogrel for 3 months (then, aspirin alone), or aspirin and oral anticoagulation if indicated. Prosthetic function was assessed according to recommendations (5).

From March 2011 to February 2012, 6 high-risk patients underwent TVIR (Table 1). Pre-dilation of the mitral valve was performed in 2 patients with mitral stenosis and commissural fusion. In all the cases, prostheses were implanted in an adequate position, with no major complication. Mean hospital stay duration was 9 ± 5 days. No patient had more than mild paravalvular regurgitation. One patient had a “significant” residual stenosis, and 2 had a “possible” stenosis (5). In 3 cases, a dynamic gradient was observed in the left ventricular outflow tract, due to the displacement of the subvalvular apparatus. No significant inter-atrial shunt was observed. Although the rings became more circular after prosthesis implantation, their shape was not fully circumferential. At discharge, all patients were in New York Heart Association functional class II. Patient 1 was readmitted 1 year after TVIR for the progression of aortic stenosis and successfully underwent transcatheter aortic valve implantation. Patients 2 and 6 had congestive heart failure due to multifactorial causes and improved with medical therapy. At last follow-up (mean: 5 ± 4 months), 5 patients were in New York Heart Association functional class II. There was no dysfunction of the mitral prosthesis. No hemolysis occurred.

There is limited experience of percutaneous treatment after failure of mitral valve repair. Percutaneous mitral commissurotomy has been successfully performed in selected patients with restenosis. Two cases have been reported with the Edge-to-Edge repair in patients with MR.

Several aspects of the TVIR should be highlighted. Sizing of the prosthesis is difficult, because surgical rings are not circular; fluoroscopy provides measurements close to those of multi-slice computed tomography. The available prosthesis sizes can be used for the treatment of a large number of patients with ischemic MR and certain patients with rheumatic valve disease; however, they are not large enough for the majority of degenerative MR. Incomplete rings won't provide the necessary support for implantation of the prosthesis. Our last patient suggests that TVIR might also be done in flexible rings.

Transseptal approach was successful in all cases; however, it is a demanding technique that requires specific training. Positioning of the valve prosthesis is challenging; fluoroscopic guidance in a view perpendicular to the mitral ring is crucial; 2-dimensional and 3-dimensional TEE guidance is also helpful. In this aspect, the surgical transapical or direct transatrial approach might provide a straighter entry and easier stabilization.

The radial force of the balloon-expandable prosthesis allows a solid fixation of the valve in the ring, as demonstrated experimentally (1). Paravalvular MR might occur in case of undersizing or if the prosthesis is “too atrial” where the pericardial skirt of the prosthesis does not cover the ring. Intra-prosthetic gradients observed at discharge seem slightly higher than those observed with current bioprostheses; these figures are similar to those observed after Valve-in-Valve implantation with the same prosthesis. After mitral valve replacement, intraventricular gradients might be due to the displacement of the anterior leaflet and subvalvular apparatus that are preserved, which is not avoidable in TVIR or small ventricular cavity. During follow-up, non-valvular factors accounted for the secondary clinical failures, similarly to what is observed after transcatheter aortic valve implantation.

This preliminary series suggests that TVIR is feasible and safe and might provide short-term improvement in valve function and clinical condition in selected high-risk patients with failure of surgical ring annuloplasty. The lessons learned from TVIR will be useful for developing transcatheter mitral valve replacement.