Editorial Comment: Lessons learned with tricuspid valve replacement

Abstract

van Slooten et al. [1] have demonstrated prosthetic valve-related complications to be more common in patients with congenital heart disease and to be unrelated to prosthesis material at late follow-up. In addition, their data suggested that patient–prosthesis mismatch may have had a negative effect on event-free survival. While repair of the tricuspid valve is preferred in almost all circumstances, there are situations when repair may not be feasible, particularly in the congenital population because of significant anatomical tricuspid abnormalities, or because tricuspid repair(s) has already been performed in the past. Numerous tricuspid valvuloplasty techniques have been described in the congenital literature, largely because of the wide variety of abnormalities seen with Ebstein malformation (EM) [2, 3] and other congenital tricuspid anomalies [4]. These various repair techniques can be applied in a ‘re-repair or re-re-repair’ situation as well as other challenging anatomical aetiologies in an effort to avoid valve replacement. However, when tricuspid valve replacement (TVR) is required, there are a number of surgical pearls that deserve comment. These ‘lessons learned’ are from our institutional experience of more than 1500 TVRs over the last 20 years. Anatomical considerations are critically important when planning TVR. Preoperative analysis of tricuspid valve anatomy and right ventricular (RV) function is essential. In general, regurgitant lesions typically result in annular dilatation ... sometimes severe annular dilatation. Annular dilatation is common with EM. Consequently, it is uncommon to place a prosthesis that is too small, i.e. patient–prosthesis mismatch is rare in the EM population. However, congenital (non-EM) tricuspid regurgitation is more likely to have a normal-sized or minimally dilated annulus, therefore placement of a small prosthesis resulting in potential patient–prosthesis mismatch is more likely. The determination of RV function is also important. Echocardiography provides the qualitative assessment of RV dysfunction—mild, moderate or severe, and MR imaging provides a more accurate determination of RV size and ejection fraction. Poor RV function is a potential setup for prosthetic tricuspid valve dysfunction—both bioprosthetic and mechanical. The low right atrial pressure and low RV pressure with additional poor RV contractility result in poor leaflet(s) or disc(s) motion predisposing to valve thrombosis. In addition, if a very large prosthesis is utilized, then bioprosthetic leaflet or mechanical disc motion may be further compromised because a disc or leaflet may not need to open if the gradient across the large prosthesis is very low (1–2 mmHg). This emphasizes the importance of not over sizing the prosthesis too much when significant annular dilatation is present. Furthermore, if a bidirectional cavopulmonary anastomosis is being performed because of depressed RV function (EM most commonly), the size of the bioprosthesis should be appropriately downsized so all leaflets move normally. We avoid mechanical TVR in the setting of a 1.5 ventricle repair because of reduced venous inflow across the prosthesis and poor RV function that predisposes to poor disc motion and potential thrombosis. For all of the issues outlined above, we use warfarin anticoagulation for tissue valves for the first 3 months after TVR and aspirin indefinitely. In general, we prefer tissue valves in the tricuspid position because maze surgery is also frequently performed to eliminate atrial tachyarrhythmias and percutaneous ‘valve-in-valve’ therapy should facilitate subsequent tricuspid replacement down the road. A mechanical TVR is used less frequently in the congenital population and is considered when RV function is normal or only mildly depressed, or when pulmonary hypertension is present, as this may optimize disc motion. A mechanical TVR is also considered when a left-sided mechanical prosthesis is present and when RV function is preserved. When a mechanical valve is used, we aim for an International Normalized Ratio of 3.0– 3.5 in addition to aspirin 81 mg daily. Other anatomical considerations relevant to TVR include the following. In the setting of EM, the prosthesis is mounted in a supra-annular position (i.e. an intra-atrial position) in order to avoid injury to the conduction tissue and the right coronary artery. The struts of the bioprosthesis are positioned so they straddle the conduction tissue, which also helps to avoid heart block. Resection of the displaced anterior leaflet tissue towards the RV outflow tract (RVOT) is done in order to avoid potential RVOT obstruction from a prosthetic valve strut pushing the displaced anterior leaflet tissue up towards the pulmonary valve. In congenital (non-EM) tricuspid regurgitation, the native tricuspid leaflet tissue is generally preserved and suture placement is performed in native tricuspid leaflet tissue. Finally, prosthesis selection is of paramount importance when performing TVR. In general, we use a porcine bioprosthesis for TVR. This is because right atrial and RV (and pulmonary artery) pressures are usually low and RV function is often depressed and sometimes poor. Leaflets of a porcine bioprosthesis are thin and pliable. Specifically, we avoid a pericardial bioprosthesis in the tricuspid position because the relatively thicker and stiffer