Follow-up data from low-risk patients through 6 years of the NOTION trial continue to favor transcatheter aortic valve replacement (TAVR) with the self-expanding CoreValve (Medtronic) over surgical bioprostheses for valve hemodynamic performance and structural deterioration.

Rates of bioprosthetic failure and endocarditis were similar for both strategies, and there was no sign of valve thrombosis in either group, Lars Søndergaard, reported in a hotline session at the Congress of the European Association of Percutaneous Cardiovascular Interventions (EuroPCR) 2018.

Analysis of TAVR durability is important in light of its increasing use in patients with longer life expectancy who will survive their bioprosthetic valve, he said. There have also been troubling reports that TAVR valves may not be as durable as their surgical counterparts.

The post hoc analysis examined valve durability in the all-comers NOTION trial using extended definitions of structural valve deterioration and valve failure standardized in a 2017 European consensus statement. The trial enrolled 280 patients (mean age, 79 years) with severe aortic stenosis and randomly assigned 145 patients to TAVR and 135 patients to SAVR. More than 80% of patients had a Society of Thoracic Surgeons score less than 4.

All-cause mortality through 6 years was 42.5% with TAVR and 37.7% with SAVR (P = .58). Bioprosthetic valve dysfunction, defined as structural valve deterioration, nonstructural valve deterioration, bioprosthetic valve thrombosis, or endocarditis, trended lower with TAVR compared with SAVR (56.1% vs 66.7%; P= .07).

The individual component of structural valve deterioration, defined as having a mean gradient of at least 20 mm Hg, a change in mean gradient of at least 10 mm Hg from 3 months, or moderate/severe aortic regurgitation, was significantly less common with TAVR than with SAVR (4.3% vs 23.7%; P <  .0001).

Søndergaard noted that patients were enrolled early in the TAVR era from 2009 to 2013 when aortic annulus sizing was based on echocardiography instead of CT imaging and no echocardiography core lab was used. After 6 years, the effective orifice area was larger for TAVR than for SAVR (1.53 cm2 vs 1.16 cm2), although mean gradient was lower (9.9 mm Hg vs 14.7 mm Hg; P <  .001 for both).

After the presentation, an audience member questioned why the rate of prosthesis-patient mismatch, although lower than in the surgical arm, was higher than expected with the CoreValve despite its excellent hemodynamics.

The panel also asked how much of the hemodynamic difference between TAVR and SAVR is due to prosthesis mismatch or smaller valve size and whether the data can be used to support TAVR as the preferred option in smaller women, regardless of risk.

Søndergaard observed that there was no default strategy for these patients to receive transcatheter treatment outside of a clinical trial when NOTION launched and that the self-expanding CoreValve prosthesis was believed to be the optimal treatment for patients with the small annulus. Panelist Corrado Tamburino, asked whether the high rate of paravalvular leak after TAVR might have played a role in an uptick in mortality at 6 years in this group.

Valve Failure

Rates of valve failure were low and similar for the transcatheter and surgical valves (7.5% vs 6.7%; = .89). The outcome was defined by three components: valve-related death (5.0% vs 3.7%), valve re-intervention (2.2% vs 0.7%), or severe hemodynamic structural valve deterioration (0.7% vs 3.0%; all comparisons nonsignificant).

Session co-chair Ran Kornowski asked how many patients who initially had valve deterioration went on to have a failure or whether they were a priori with failure vs deterioration. Søndergaard replied that it is a very good point and worth following up on, but at the time of the analysis, the consensus statement was only just published.