Balloon-expandable or Self-expandable Transcatheter Heart Valves. Which Are Best?

John Webb; David Wood; Janarthanan Sathananthan; Uri Landes


Eur Heart J. 2020;41(20):1900-1902. 

So, which are best? Balloon-expandable or self-expandable valves?

The prototypic balloon-expandable Sapien-type valves look very similar to their imitators, although the materials, build, and delivery systems may differ dramatically. Similarly, the prototypic self-expanding CoreValve family of valves feature supracoronary fixation and supra-annular valve function, amongst other features that may be much more important than the mere mechanism of expansion. Other self-expanding valves (e.g. Jenavalve, Portico, Centera, Acurate, Allegra, Venus, and Vitaflow) differ dramatically in appearance and function. Amongst many differences, some do not utilize supracoronary fixation and many do not feature supra-annular leaflets. The point being that comparisons of 'self-expandable' vs. 'balloon-expandable' valves are becoming increasingly nuanced. Each valve design will need to stand scrutiny on its own.

To date, the majority of randomized controlled trials have focused on the comparison of transcatheter aortic valve implantation (TAVI) with medical or surgical therapy. TAVI has now been shown to have outcomes competitive with surgery, irrespective of surgical risk. The focus may now shift to assessing differences, if any, between different transcatheter heart valve designs. The largest randomized comparison of balloon- and self-expandable valves to date (SCOPE1) compared the Sapien 3 and Acurate neo valve, with the Sapien 3 demonstrating superior results.[1] However, few would conclude that this is proof that balloon-expandable valves are better. To a larger degree, each valve must stand on its own. It may be more helpful to focus on the current Sapien-type and CoreValve-type platforms.

For almost two decades, two iconic platforms have evolved in a stepwise fashion. The original Cribier–Edwards to the Sapien, Sapien XT, Sapien 3, and Sapien 3 Ultra. And the original Corevalve to the Evolut, Evolut R, and Evolut Pro. The delivery systems have evolved just as dramatically. Moreover, the knowledge base and skills associated with implanting each platform have also evolved dramatically. These two systems have been a moving target, and at different time points each platform might arguably lay claim to being 'best'. At least in terms of the main concerns that captured the attention of early investigators, such as procedural death, stroke, and vascular complications, outcomes have improved dramatically with both platforms.

In this issue of the European Heart Journal, Thiele et al. report non-inferiority for two contemporary devices (Sapien 3 and Evolut R) in terms of their primary efficacy composite endpoint (death, stroke, paravalvular regurgitation, and new pacemakers).[2] The experience of 447 patients is relatively small and there are limitations to the study, but it seems reasonable to conclude that both platforms can achieve excellent outcomes in the hands of experienced users.

This is not the first such comparison. The smaller 241 patient randomized CHOICE study compared the previous generation Sapien XT with CoreValve.[3] The primary endpoint favoured Sapien XT, mainly driven by less paravalvular regurgitation and a lesser need for implantation of a second valve. The larger 8192 patient CENTER collaboration performed a propensity analysis and found that Sapien-type valves had a statistically lower risk of stroke and new pacemakers.[4] The even larger 31 113 patient French national registry performed a matched analysis and reported a significantly lower all-cause mortality, cardiovascular mortality, new pacemakers, and rehospitalization for heart failure with Sapien 3 as compared with Evolut R.[5]

Both Sapien and CoreValve platforms have been separately evaluated in relatively comparable pivotal trials at various levels of surgical risk. In the high surgical risk trials, 30-day mortality was 3.4% vs. 3.4%,[6,7] respectively; in the intermediate risk trials, the figures were 3.9% vs. 2.2%;[8,9] and in the low risk trials they were 0.4% and 0.5%.[10,11]

Figure 1.

A sampling of transcatheter aortic valves. BE, balloon expandable; ME, mechanically expandable; SE, self-expandable.

Perhaps the best comparison is found in the two pivotal low surgical risk trials, PARTNER 3 and Evolut Low Risk, which separately evaluated contemporary Sapien 3 and Evolut R valves.[10,11] These trials were intentionally very similar; contemporaneous, identical regions, with almost identical entry criteria, and very rigorous oversight. As it turned out, comorbidities were very similar; in fact the Society of Thoracic Surgeons predicted risk of mortality (STS PROM) was identical at 1.9%. PARTNER 3 and Evolut Low Risk mortality at 30 days was similar at 0.4% vs. 0.5%, respectively. However, Sapien 3 had lower rates of stroke (0.6% vs. 2.1%), paravalvular leak ≥ moderate (0.8% vs. 3.4%), and new pacemakers (6.5% vs. 17.4%) while Evolut R was associated with a lower residual mean gradient (8.4 mmHg vs. 12.8 mmHg). Obviously the comparison suffers as it is not randomized, but the findings are impactful.

Interestingly, the SOLVE-TAVI study reported similar rates of new pacemakers for Evolut R and Sapien 3. There is a large body of evidence in documenting higher rates of new left bundle branch block and new pacemakers with the CoreValve platform.[12,13] In a 17 139 patient meta-analysis of 40 studies published in this journal, new pacemaker rate varied from 16% to 37% with CoreValve; from 14.7% to 26.7% with Evolut R; from 2.3% to 28.2% with Sapien/XT; and from 4% to 24% with Sapien 3.[13] This has been correlated with a longer hospital stay, left ventricular dysfunction, rehospitalization, procedural cost, and, although controversial, late mortality.[14] Lower pacemaker rates have been reported in small series with various procedural improvements, and it is clear that physician preference and local practice play major roles in determining pacemaker rates, but equivalence for the two platforms with regards to atrioventricular block remains to be proven.[15]

For now, we can probably agree that excellent early outcomes can be achieved with both the Sapien 3 and Evolut R valves available today. Both valves have a large and opinionated following. Rather than asking which valve is best, perhaps the more pertinent question is which valve is best for this particular patient. Patient comorbidities, anatomical factors, and procedural concerns may ultimately favour one valve design over another. Outcomes such as mortality, stroke, vascular injury, bleeding, and paravalvular regurgitation have captured our attention in the past. As these outcomes improve and plateau, the focus may move away the concerns of early adopters to newer issues such as the implications of atrioventricular injury, haemodynamic function, coronary access, durability, repeatability, and cost—areas of future research.