Rethinking tricuspid ring size: one size does not fit all?

Dear Editor,

With great interest, we read the article by Ono et al. [1] entitled ‘Optimizing ring selection for secondary tricuspid regurgitation: the role of body size’. It addresses an interesting and rather understudied topic in cardiothoracic surgery. The tricuspid valve has long been referred to as the ‘forgotten valve’; however, it has recently gained more attention with more studies being performed to further understand its epidemiology of disease and to get insight into treatment possibilities. Herewith, we would like to address several points of concern related to the methodology and generalizability of this study.

Besides the limitations already mentioned by the authors, we would like to highlight some methodological concerns. The mean body surface area (BSA) in this study was 1.53 m², which raises concerns regarding its representativeness for the European population where the average BSA is markedly higher (i.e. 1.86 m² from a German study [2]). In addition, despite the propensity score matching, more patients in the small tricuspid annuloplasty ring index (TARI) group underwent a mitral valve replacement than the large TARI group (72.9% versus 58.3%, respectively) and, accordingly, less patients underwent a mitral valve repair. Both the severity and mechanism of disease might be substantially different behind these 2 procedures, which might have skewed the results towards 1 end. Last, the preoperative tricuspid regurgitation (TR) grade was moderate in 43.1% and severe in 14.6% of the patients; however, no information is provided regarding the indication for a concomitant tricuspid annuloplasty in the other 42.3%.

Four important statistical issues should also be considered. First, only 3 years of follow-up were reported despite data availability for up to 8 years, indicating possible reporting bias. Second, the investigators do not report on missing data or lost to follow-up potentially impacting the reliability of findings. Third, the maximum selected log-rank statistic necessitates multiple testing to determine the TARI threshold increasing the risk of type I error. Hence, external validation in a separate population is deemed necessary to secure the validity of results. Lastly, only 96 patients were matched, significantly reducing the power of the study and possibly introducing selection bias. Failure of matching may indicate substantial differences in baseline characteristics, meaning that the selected patients might not be representative of the population.

Several factors have been associated with recurrence of TR [3–5]. The leaflet tenting height, indexed tricuspid annulus diameter, recurrence of atrial fibrillation and mitral regurgitation, right ventricular function, right ventricular systolic and pulmonary artery pressures, preoperative severity of mitral valve disease and postoperative medication are important confounders that should have been accounted for. This is an important limitation that could have resulted in both selection and attrition bias potentially compromising the validity of the observed differences. Moreover, the lack of standardized criteria for ring type determination (surgeon’s discretion) may have influenced the outcomes.

In conclusion, while the study focuses on an important clinical question, its methodology limits its generalizability. Future studies should consider all relevant confounders and prolong follow-up time to provide robust insights into the effect of BSA on TR after annuloplasty.

FUNDING

Not applicable.

Conflict of interest: none declared.

ACKNOWLEDGEMENTS

Not applicable.

REFERENCES