Discussing paravalvular leakage after transcatheter aortic valve implantation (TAVI) requires an overview of the procedural evolution and outcomes over its 21-year history. Because TAVI involves implanting a device while the native aortic valve remains in place, there is an increased risk of incomplete remodeling of calcifications around the annulus and leaflets. This could potentially lead to paravalvular leaks from gaps between native tissue and the prosthetic stent. In response, successive generations of these prostheses have incorporated technical modifications tailored to different implantation approaches: balloon-expandable, self-expanding, or mechanically-expandable.

In balloon-expandable models, like the Edwards® Sapien®, the primary advantage is greater radial force during implantation, which enables more effective remodeling of calcified tissue. Additionally, the inclusion of skirt-like adaptors that adjust the stent to irregular tissue contours has been the proposed solution to reduce paravalvular leakage. Self-expanding valves, such as the Medtronic® Evolut®, Boston® ACURATE neo®, Abbott® Portico®, and Navitor®, exert less radial force initially but offer greater adaptability, improving fit over subsequent days through the martensitic phase of the nitinol stent. These devices also feature skirts, albeit with a lower profile than balloon-expandable options. The most effective enhancement, however, has been dynamic sealing, achieved through a reverse valve mechanism in the Abbott® Navitor® prosthesis, which effectively blocks regurgitant flow. Lastly, mechanically-expandable valves like the Boston® LOTUS Edge® use high-density mesh stents to replicate the radial force of balloon-expandable models. This stent structure provides tissue remodeling, a feature termed “adaptive seal,” within which the valvular mechanism is anchored, bypassing the tissue remodeling phase seen in balloon-expandable models and reducing structural damage risk to the leaflets.

The emphasis on and clinical significance of paravalvular leakage has evolved over time, increasingly considering a risk-benefit assessment of the procedure. Initially, even moderate leaks were deemed acceptable, given that most TAVI patients were inoperable or at high surgical risk, with priority given to resolving stenosis (PARTNER I). It was also argued that, given that aortic valve disease often presents as mixed lesions, patients would better tolerate some regurgitation. However, as TAVI has been extended to younger, lower-risk patients, tolerance for complications, including paravalvular leakage, has declined. In intermediate-risk populations (SURTAVI, PARTNER 2), a slight reduction in survival due to mild paravalvular leakage was observed in medium-term follow-up (5 years, Makkar et al., PARTNER 2).

This finding accelerated the establishment of assessment and grading criteria, culminating in the current VARC-2 guidelines based on three echocardiographic levels (none/trivial, mild, moderate, severe), which are employed in this study. It remains to be seen how paravalvular leakage will impact future studies on low-risk groups (PARTNER 3, Evolut Low-Risk), particularly with the inclusion of newer prosthetic designs. However, the trend now favors real-world registries over extended clinical trial follow-ups due to the excessive biases and follow-up losses highlighted by Barili, previously discussed on this blog.

To this end, this study reviews outcomes from the TAVI Registry, drawing data from two centers, one in Japan and the other in Sweden, covering the years 2008 to 2019. Various prosthesis models were used, with balloon-expandable valves being the most prevalent (58%) and transfemoral approach the most common (94%). Post-dilatation was performed in 6% of patients without residual leakage and in 14% of those with mild leakage (p = 0.004, adjusted groups). The analysis included patient characteristics, morbidity and mortality rates at follow-up, and 5-year survival among patients with absent/trivial paravalvular leakage versus those with mild leakage. Higher grades of leakage were excluded from analysis due to their known impact on short- to medium-term survival. Raw data from 1,404 patients were adjusted using propensity score matching, resulting in two homogenous groups of 332 patients each. Only patients with complete follow-up were included. In the unadjusted analysis, patients with mild residual paravalvular leakage were older, had more comorbidities, a lower body mass index, and more severe stenosis. These differences were mitigated by the adjustment, with findings indicating that patients with mild paravalvular leakage had a 41% higher mortality rate at 5 years, independent of periprocedural mortality or complications.

The authors conclude that patients with mild residual paravalvular leakage after TAVI experience increased mortality at medium-term follow-up compared to those without residual leakage, underscoring the prognostic significance of even mild leakage.

COMMENTARY:

Divergent findings on the impact of mild paravalvular leakage observed across clinical trials have led to it being viewed as an acceptable complication, sometimes considered a satisfactory technical outcome. However, this study confirms the reduction in survival over follow-up, as suggested in prior studies. Therefore, mild paravalvular leakage, along with new-onset left bundle branch block, the need for permanent pacemaker implantation, subclinical vascular injuries, asymptomatic thrombosis, or limitations in percutaneous access to the coronary tree, should no longer be considered minor complications. These factors must be weighed in patient selection for surgery versus intervention by the Heart Team, especially in low-risk and asymptomatic cases.

Several hypotheses have been proposed to explain the association between mild paravalvular leakage and increased mortality. First, in patients with aortic stenosis, left ventricular compliance is often low, meaning volume overload might persist despite mild residual regurgitation. Indeed, this adverse effect has been described with mild mitral regurgitation post-repair and the poorer prognosis in patients with combined aortic stenosis and regurgitation versus those with pure stenosis. Second, mild paravalvular leakage may lead to gastrointestinal bleeding episodes due to an acquired von Willebrand factor deficiency caused by the mechanical stress from regurgitant flow, resembling Heyde syndrome pathophysiology. Third, it is hypothesized that mild leakage might gradually worsen over time, exacerbating these adverse effects.

In summary, after 21 years of development, TAVI continues its pursuit of technical perfection through advancements in device technology and procedural aspects. Meanwhile, patient selection will continue to rely on the discernment of the Heart Team, particularly in intermediate- to low-risk and asymptomatic cases. Further research is needed to determine which preprocedural factors can predict the development of these so-called minor complications that significantly impact patient survival. Assessment will need to be increasingly thorough, moving beyond evaluations based solely on age, surgical risk, vascular access diameters, and predictors of major complications. Comprehensive assessment of the burden and distribution of calcification in aortic root structures, potentially through simulation and modeling, may help better align patients with the most appropriate therapeutic option, including the specific transcatheter prosthesis. Despite the favorable outcomes observed with TAVI, there may remain a subset of patients for whom resection of calcified leaflets and implantation of a bioprosthesis on a decalcified annulus will continue to be necessary.

REFERENCE:

Yokoyama H, Sugiyama Y, Miyashita H, Jalanko M, Ochiai T, Shishido K, et al. Impact of mild paravalvular regurgitation on long-term clinical outcomes after transcatheter aortic valve implantation. Am J Cardiol. 2023 Mar 15;191:14–22. doi: 10.1016/j.amjcard.2022.12.002.