Invited commentary: the choice between mechanical versus biologic mitral valves is becoming clear, what next?

In this issue of EJCTS, Leviner et al. compare mid-term outcomes after biologic (BV) versus mechanical (MV) mitral valve replacement (MVR) in patients aged 50–70 years and conclude that patients aged 50–64 years have a survival benefit with mechanical valves, while those aged 65–70 years do not [1]. We congratulate the authors on contributing to the ongoing investigation of mitral valve prosthesis choice for patients in this middle-age ‘grey zone’. This multicentre retrospective study includes 796 patients from 4 medical centres in Israel, who received mechanical (N = 539) or bioprosthetic (N = 257) MVRs between 2005 and 2018. This sample did not include patients who underwent concurrent aortic valve, aortic, carotid artery or congenital heart disease operations. The authors performed propensity-adjusted analysis with a primary end-point of mid-term all-cause mortality. Initial analysis revealed 30-day mortality of 12.5% for BV patients and 8.3% for MV patients, a difference that was not statistically significant after propensity adjustment. However, subgroup analysis confirmed significantly lower mortality after MV compared to BV replacement in patients aged 50–64 years [Hazard Ratio (HR) 1.5], while mortality was similar between BV and MV subjects aged 65–70 years. In addition, they showed that male patients had higher mortality in the 50–64-year age group (HR 2.14). Consistent with prior observations in the USA, the authors noted a recent trend towards preferential use of BV for MVR in their institutions, from 10% in 2005 to 50% by 2019 [1–3]. This is similar to what was observed for aortic valve prostheses in the early 2000s in the USA and for both valves was likely secondary to the advent of transcatheter valve-in-valve procedures that could be accomplished following BV placement [4]. This study adds to the body of evidence that patients at the younger end of the ‘grey zone’ may benefit from MV prostheses. As the authors suggest, while these findings are in line with existing guidelines and large studies [2, 3, 5, 6], highlighting them is worthwhile to counteract the trend of more BV use among patients aged 50–64 years (when in line with patients’ preferences). This study utilized propensity score matching, similar to Chikwe et al. in their large 2015 study. Propensity score matching allows for 1:1 matching based on underlying factors such as comorbidities, to minimize confounding factors. However, inherent in this strategy is the exclusion of patients who would be indicated for only one of the valve types due to those factors. As such, we prefer inverse probability weighting, which balances patients without exclusions [7]. Regardless, the authors provide a sound statistical analysis that fortifies existing recommendations. The authors acknowledge several limitations to their study, which we agree could constrain the applicability and impact of this analysis. Specifically, the follow-up time of this study was short, at only 66 months overall and 86 months for patients who survived to the end of the study. While the first 5–7 years after MVR are informative, many adverse outcomes such as reoperation, bleeding events and death occur as patients’ valves deteriorate beyond 10+ years [4, 8]. In addition, the database utilized for this study did not include specific valve types. We have known since the beginning of prosthetic valve adoption that different valves provide different in vitro hydrodynamics, and we now know that even the orientation of a prosthetic valve can affect left ventricular flow dynamics [9, 10]. Looking forward, we encourage investigators to include granular valve type data, which would help refine decision-making regarding current valves and inform innovation of new technologies. To conclude, this study is a valuable addition to support current guidelines and inform surgical decision-making, and we are excited to see it appear in this edition of EJCTS.