Since its introduction in 2002, the use of TAVI has increased exponentially. PVL remains one of the most common complications of the procedure and has traditionally been linked to reduced long-term survival, especially in cases of moderate or severe regurgitation. While new-generation transcatheter valves have reduced the incidence of moderate or severe PVL to rates similar to those observed after surgical valve replacement, the prevalence of mild PVL remains clearly higher in TAVI compared with surgical prostheses.

The long-term impact of mild PVL is not well established. Although some studies suggest that nearly half of PVLs tend to regress over time, the factors influencing this evolution remain uncertain.

The aim of this study was to evaluate both the incidence of PVL following TAVI and its impact on long-term survival. The authors retrospectively analyzed prospectively collected data from all patients undergoing TAVI at their institution over a 10-year period (November 2012 to January 2023). Patients were stratified into three groups according to residual PVL at 30 days: none/trace, mild, and moderate to severe. The primary endpoints were long-term survival and changes in PVL severity between 30 days and 1 year post-implantation.

During the study period, 3600 patients were reviewed. Of these, 2719 (75.5%) had no or trace PVL, 808 (22.5%) had mild PVL, and 73 (2.0%) had moderate to severe PVL at 30 days. A total of 2327 patients completed 1-year echocardiographic follow-up. Among those with mild or moderate to severe PVL, 52.7% showed regression in severity at 1 year. Conversely, 10.4% of patients initially classified as none/trace or mild experienced a worsening in PVL grade during this period. No variables were clearly associated with PVL progression, except for the year of implantation, with procedures performed after 2017 acting as a protective factor. The median follow-up duration for all groups was 3 years. Kaplan-Meier survival analysis demonstrated significant differences across groups, with reduced survival in the moderate to severe PVL cohort at both 1 and 5 years. However, there were no significant differences in survival between the none/trace and mild PVL groups at 1 year (p = .2). A divergence became evident from year 2 onwards, with worse survival in the mild PVL group (p = .03), though this association did not persist after multivariable adjustment.

COMMENTARY:

This study seeks to update long-term outcomes based on one of the most frequent complications after transcatheter aortic valve implantation: paravalvular leak. The authors reviewed a large cohort of 3600 patients, although only 2300 had sufficient follow-up imaging to allow evaluation of 30-day and 1-year outcomes. No explanation is provided for the loss of one-third of the initial cohort, which could introduce significant selection bias—particularly if patients with less PVL had more favorable early outcomes and were overrepresented among those completing follow-up.

The reported 30-day outcomes were acceptable, with mild PVL observed in 22.5% of patients and moderate to severe PVL in just 2%. However, the subsequent description of PVL at 1 year is not clearly detailed, particularly given the reduced number of patients with available follow-up. Based on the available data, 52% of patients with any PVL experienced regression in severity, though the extent of improvement was not clearly defined. Conversely, 10.4% of patients with initial none, trace, or mild PVL exhibited worsening, again without precise classification of the new severity.

Of note, in patients undergoing valve-in-valve procedures, the absence of PVL or presence of only trace PVL was significantly more frequent compared with those developing moderate to severe PVL. This is a favorable finding, as it aligns with the growing trend toward valve-in-valve implantation over surgical reoperation in appropriate candidates. Additional findings included a higher frequency of moderate to severe PVL with larger valve sizes and lower PVL rates with newer-generation prostheses. However, the study did not assess aortic annular calcification, annulus dimensions, or other anatomical predictors—thus precluding identification of modifiable risk factors or preprocedural considerations to mitigate PVL.

Regarding survival, despite inclusion of patients treated over a 10-year period, the mean follow-up was limited to 3 years. One-year follow-up was completed in 81% of the cohort, which could distort survival comparisons. Interestingly, while Kaplan-Meier curves did not show significant survival differences between the none/trace and mild PVL groups overall, the curves diverged after year 2, with lower survival in the mild PVL group. Although this was not confirmed in the multivariable model—likely due to limited statistical power—it warrants attention. As TAVI expands to younger, lower-risk populations with longer life expectancy, even mild and seemingly benign PVL may have deleterious consequences over time. Similar divergence beginning at 2 to 3 years has been reported in the PARTNER 2 trial. The current study does not clarify whether PVL progression correlates with mortality, or whether lack of regression itself is an independent predictor of worse outcomes.

Additionally, despite the thoughtful analysis, one key question arises: what is the actual incidence of PVL following surgical aortic valve replacement? A review of the literature shows that many recent studies still cite historical data from over 25 years ago, reporting surgical PVL rates ranging from 2% to 17%. In contrast, the 2023 PARTNER 3 trial reported a surgical PVL rate of 3.2%, compared with 20.8% in TAVI patients. Another recent article by Hayashi et al., published in The Journal of Thoracic Disease (March 2025), described an 8.1% incidence of PVL. However, that study included patients with infectious endocarditis and reoperations—clinical scenarios where annular integrity cannot be compared with that of naïve patients. A noteworthy strength of that article is its detailed classification of PVL type and location, such as annular segments or anatomical scenarios with higher leak risk, like extensive annular calcification—information that is highly useful for procedural planning. By contrast, the authors of the current study acknowledge that they were unable to identify specific predictors of PVL progression following TAVI.

All of this underscores that, although new-generation transcatheter valves have dramatically reduced PVL rates compared with earlier devices, surgical PVL rates have also declined in recent years—likely due to systematic use of intraoperative imaging and overall improvements driven by performance scrutiny. If future evidence confirms that mild PVL—more frequent and unpredictable after TAVI than surgery—is associated with increased mortality beyond 2 years, this should be a key consideration before extending TAVI indications based solely on age or risk level. Time will tell.

REFERENCE:

Warraich N, Brown JA, Ashwat E, Kliner D, Serna-Gallegos D, Toma C, et al. Paravalvular Leak After Transcatheter Aortic Valve Implantation: Results From 3600 Patients. Ann Thorac Surg. 2025 May;119(5):1037-44. DOI: 10.1016/j.athoracsur.2025.01.012

Hayashi Y, Russell JK, Dvorak CJ, Gebska MA, Hanada S, Singhal AK. Echocardiographic characteristics of paravalvular leak following surgical aortic valve replacement: a retrospective cohort study. J Thorac Dis. 2025 Mar 31;17(3):1249-58. DOI:10.21037/jtd-2024-1989.

Mack MJ, Leon MB, Thourani VH, Pibarot P, Hahn RT, Genereux P, et al. Transcatheter Aortic-Valve Replacement in Low-Risk Patients at Five Years. N Engl J Med. 2023 Nov 23;389(21):1949-60. DOI:10.1056/NEJMoa2307447.

Ahmad Y, Howard JP, Arnold AD, Madhavan MV, Cook CM, Alu M, et al. Transcatheter versus surgical aortic valve replacement in lower-risk and higher-risk patients: a meta-analysis of randomized trials. Eur Heart J. 2023 Mar 7;44(10):836-52. DOI:10.1093/eurheartj/ehac642.

Kananathan S, Perera LA, Mohanarajan M, Sherif M, Harky A. The management of paravalvular leaks post aortic valve replacement. J Card Surg. 2022 Sep;37(9):2786-98. DOI:10.1111/jocs.16672