LAUSR

Simard et al. have summarized their single‐center experience comparing the 5‐year outcomes of transcatheter mitral valve replacement (TMVR) and surgical mitral valve replacement (SMVR). They found TMVR to have significantly lower 30‐day mortality but over the long‐term, SMVR did have better outcomes. As the authors mentioned, the long‐term outcomes were not unexpected since the TMVR cohort did have more advanced age and higher rates of comorbidities and the short‐ term outcomes were a remarkable demonstration of the impact of the transseptal delivery for TMVR. Despite, censoring patients to achieve age matched cohorts in the TMVR and SMVR groups, the TMVR cohort's clinical profiles were still significantly worse except for significant tricuspid regurgitation (TR). Severe TR is a significant determinant of survival in mitral valve interventions and its long‐term impact may have shaped the long‐ term survival of the TMVR cohort. Although details are not shown, assuming that any significant tricuspid valve regurgitation was addressed during surgery, it may in part explain improved long‐ term outcomes for SMVR. Certainly, selection bias towards healthier patients in SMVR likely plays a significant role, but the role of significant TR does and should impact choice of therapy. Until transcatheter TR therapies are more mature and have been disseminated, multivalvular heart disease should still lean towards surgery to address compound problems. A notable observation is the high rate of anticoagulation in the long‐term follow up for TMVR. There is no consensus guideline for duration of anticoagulation post‐TMVR but some late cases of valve thrombosis or leaflet thickening have been observed in other series. Patient selection for tolerance of anticoagulation has not been central to TMVR discussions as much of the focus has been on anatomic screening for left ventricular outflow tract obstruction. Nevertheless, the importance of anticoagulation should not be overlooked to prolong prosthesis durability. There is very little long‐term data to support their practice but given the frail nature of TMVR patients, there is little tolerance for jeopardizing the patient's prothesis. Transcatheter heart valve size was not found to have a significant impact of mortality in this series. Although not statistically significant, it was noted in the Kaplan–Meier analysis that 23mm valve had the worst survival and it has been long been suspected given the worse hemodynamic performance of this group. Although statistical significance was not demonstrated in Simard's, the small sample size hampers the ability to independently identify predictors of mortality. While the Valve‐in‐Valve International Data Registry did confirm the high proportion of gradient >10mmHg in this cohort, they could not identify high gradients as a predictor of mortality. Despite the lack of confirmatory evidence, operators should avoid small prostheses as often as possible. Simard and his coinvestigators are to be congratulated on their publishing their long‐term data. Just as the authors admit to the limitations of a retrospective, single‐center study with limited sample size, long‐term data is lacking and should be reported. Clinicians need to better understand the advantages and drawbacks of TMVR as the field is exiting the “early” experience of TMVR, therefore, a more refined patient selection process should be developed based on available data.