Managing mild paravalvular leak after TAVI


Paravalvular leak after transcatheter aortic valve implantation (TAVI) continues to be a concern, and the question of whether mild paravalvular leak is important remains unanswered. Susheel Kodali (director, Interventional Cardiology Fellowship Program, co-director, NYP Columbia Heart Valve Center, Columbia University Medical Center, New York, USA) speaks to Cardiovascular News about the identification and management of mild paravalvular leak. 

Overall, what percentage of patients experience paravalvular leak after TAVI?

Having some degree of paravalvular leak is very common after TAVI. Different studies have shown that mild paravalvular leak is present in 40–60% of patients, and moderate or severe leak is present in 10–20% of patients. A recent meta-analysis of 45 studies, by Athappan et al, demonstrated that moderate or severe aortic regurgitation was present in 11.7% of patients. In the PARTNER trial, in which all the echos were analysed by a core lab, moderate or severe aortic regurgitation was present in about 12% and mild in about 40% of patients. 

In your view, what is the definition of mild paravalvular leak (as opposed to moderate/severe)?

The grading of paravalvular leak can be a challenge. The studies to date have had variable methods for grading it—some have used angiography while others have used echocardiography. The gold standard method is echocardiography. Even with the use of echo, there is variability from reader to reader. There have been efforts to standardise the grading through the Valve Academic Research Consortium (VARC 2) document, which advocates for a semi-quantitative grading of aortic regurgitation based on echo. In reality, it is fairly easy to define moderate and severe aortic regurgitation as well as none and trace aortic regurgitation. The challenges are in the grey zone between mild and moderate. Efforts should be made by all operators to integrate all of the information available in making the assessment of aortic regurgitation. In addition to echo, this may include haemodynamic and angiographic data in limited cases. 

What do data indicate for the effects of mild paravalvular leak (ie. does it affect mortality/morbidity)?

Multiple studies have shown that moderate or severe paravalvular leak is clearly associated with higher one-year mortality. In the meta-analysis by Athappan et al, the hazard ratio for moderate or severe paravalvular and one year mortality was 2.27 with a highly significant p value. In the same meta-analysis, mild paravalvular was also associated with higher one year mortality. However, the hazard ratio was 1.83 with a p value of 0.048. We recently presented an analysis from the PARTNER trial with over 2,200 patients. In this analysis, there was a stepwise increase in mortality with both mild (hazard ratio 1.47, p value 0.003) and moderate/severe paravalvular leak (hazard ratio 2.38, p value <0.001) on multivariable analysis. It is important to note that there are significant differences in the baseline characteristics of the patients that had none-trace paravalvular leak vs. mild paravalvular leak vs. moderate/severe paravalvular leak. These differences in baseline characteristics may confound the analysis. Nevertheless, the fact that despite accounting for these differences on multivariable analysis, both mild and moderate/severe paravalvular leak is associated with increased one-year mortality is concerning.

How should a patient with mild paravalvular leak be managed?

The goal is to minimise paravalvular leak to start. We have learnt a lot about how to size TAVI devices over the last decade. Careful pre-procedural planning and appropriate valve size selection have reduced paravalvular leak rates. Also, the devices and procedural techniques have improved allowing for a more predictable valve deployment in the correct intra-annular position that has also helped reduce paravalvular leak rates. However, due to anatomic issues and limitations of the current devices, some patients will have mild paravalvular leak post implant. Our strategy is to immediately post-dilate the valve after implantation if we think the risks are low and doing so will improve the paravalvular leak. We have shown in our analysis that post-dilatation results in improved stent expansion and reduced regurgitant volume. There are risks with post-dilatation so careful thought to the risk-benefit ratio should be given before proceeding with post-dilatation. Otherwise, if patients are left with mild paravalvular leak, they should be managed as any other patient with optimal medical therapy of their heart failure. 

Do you think, with improvement and development of TAVI devices (eg. Sapien 3 or the Lotus valve), paravalvular leak will be less of a concern in the future?

The hope is that these newer devices, which have various mechanisms to seal the gap between the prosthesis and annulus, will reduce paravalvular leak in the future. At recent meetings, there has been significant excitement regarding them as initial feasibility studies have demonstrated much less paravalvular leak. However, these are early stage investigations and we need to continue to study them carefully.