This retrospective study examined how mitral regurgitation (MR) evolves after transcatheter aortic valve replacement (TAVR) in patients with low-flow aortic stenosis (LFAS). The authors evaluated MR trajectories across LFAS phenotypes, identified predictors of MR worsening, and assessed associations between MR trajectory, LFAS subtype, and clinical outcomes including all-cause mortality and heart failure hospitalization (HFH).
The cohort included 614 consecutive LFAS patients who underwent TAVR. Patients were classified into three LFAS phenotypes: low-flow high-gradient (LFHG; n=153, 24.9%), classical low-flow low-gradient (cLFLG; n=155, 25.2%), and paradoxical low-flow low-gradient (pLFLG; n=306, 49.8%). The dataset arose from a single institutional review board–approved retrospective analysis; informed consent was waived due to the study design.
MR severity was abstracted from routine clinical echocardiography reports and mapped to a six-level ordinal scale. Follow-up echocardiograms were available in 443 patients at approximately 30 days and in 290 patients at approximately 1 year after TAVR. The analysis used these timepoints to define early and intermediate MR trajectory.
The prespecified definition of MR worsening was a ≥1-grade increase in MR severity from baseline to follow-up at either ~30 days or ~1 year. The 30-day and 1-year assessments were treated as distinct endpoints to capture early and later changes in MR after valve replacement.
Multivariable logistic regression models were used to identify independent predictors of MR worsening at each timepoint. Kaplan–Meier estimates and Cox proportional hazards models evaluated the relationship of MR trajectory and LFAS phenotype with clinical endpoints: all-cause mortality, heart failure hospitalization (HFH), and a composite of both.
At 30 days, MR trajectory varied significantly across LFAS phenotypes. The classical low-flow low-gradient (cLFLG) group exhibited the highest rate of MR worsening, while the LFHG group exhibited the lowest rate. By 1 year, unadjusted distributions of MR trajectory did not differ significantly across phenotypes.
Follow-up completeness differed by timepoint: 443 patients had 30-day echocardiographic data and 290 had 1-year data, reflecting attrition in intermediate-term imaging.
After adjustment for covariates in multivariable logistic regression, cLFLG phenotype remained independently associated with MR worsening at both the 30‑day and 1‑year timepoints. The article reports that phenotype retained predictive value even when accounting for other measured factors; specific covariates and effect estimates are provided in the full manuscript.
Unadjusted analyses showed that MR worsening was associated with worse outcomes at 30 days. However, when multivariable adjustment was applied, MR worsening itself was not an independent predictor of the composite endpoint. Instead, LFAS phenotype—particularly cLFLG—emerged as the dominant predictor of adverse clinical outcomes, including higher event rates for death and HFH.
The study authors conclude that MR evolution after TAVR is phenotype-specific in patients with LFAS. Patients with cLFLG face the greatest risk of MR worsening and the poorest event-free survival among the analyzed phenotypes. These findings support a tailored, phenotype-informed approach to post‑TAVR echocardiographic surveillance and clinical follow-up, prioritizing closer monitoring of cLFLG patients.
Key study limitations include the retrospective design and incomplete echocardiographic follow-up at 1 year (290/614). MR severity was obtained from routine clinical reports rather than a centralized blinded core-lab adjudication, which may introduce measurement variability. The authors note that MR worsening was associated with worse unadjusted outcomes at 30 days but did not remain independently predictive after multivariable adjustment, with LFAS phenotype exerting a dominant effect.
The analysis is presented as a preprint that has not undergone peer review and therefore should not yet guide clinical practice. The underlying data are not publicly available due to patient privacy and institutional restrictions; de‑identified data may be requested from the corresponding author subject to institutional approvals.
In this single‑center retrospective cohort of 614 LFAS patients undergoing TAVR, MR trajectory following intervention differed by LFAS phenotype. cLFLG patients had the highest risk of MR worsening and worse event‑free survival, suggesting clinicians may consider phenotype when planning post‑TAVR monitoring. Because the work is a preprint and observational, findings require confirmation in peer‑reviewed studies before changing practice.