TAVR With Sapien XT Noninferior to Surgery in Intermediate-Risk Patients, With a Win for Transfemoral

CHICAGO, IL—Transcatheter aortic valve replacement using a second-generation balloon-expandable device is at least as good as surgery in intermediate-risk patients with severe symptomatic aortic stenosis. What’s more, for those who can be treated via a transfemoral approach, TAVR appears to be the winning strategy. 

Results from the PARTNER 2, Cohort A trial showed that death or disabling stroke at 2 years—the trial’s primary endpoint—were not significantly different for patients randomized to TAVR using the Sapien XT (Edwards Lifesciences) or open-heart surgery. That observation held up both in the intention-to-treat population (19.3% vs 21.1%, respectively) and in an as-treated analysis (18.9% vs 21.0%).

More than three-quarters of the patients enrolled were eligible for treatment using a transfemoral approach, and in this group, TAVR was associated with a 21% relative reduction in the primary endpoint, a difference that just met statistical significance (HR 0.79; 95% CI 0.62-1.00; P = 0.05) in the intention-to-treat analysis and was more robust in the as-treated analysis (HR 0.78; 95% CI 0.61-0.99; P =0.04). TAVR-treated patients had more vascular complications and more paravalvular leaks post-procedure; by contrast, surgery-treated patients had more acute kidney injury, severe bleeding, and new-onset atrial fibrillation.

Craig Smith, MD (NewYork-Presbyterian/Columbia University Medical Center, New York, NY),  coprincipal investigator for the study, presented the full results here this morning in the opening late-breaking clinical trial session of the American College of Cardiology 2016 Scientific Sessions. They were published simultaneously in the New England Journal of Medicine.

PARTNER 2: Intermediate-Risk

PARTNER 2A randomized 2,032 intermediate-risk patients with severe symptomatic aortic stenosis at one of 57 centers to either open surgical valve replacement or transcatheter aortic valve implantation with the Sapien XT. The decision to treat via a transfemoral approach or a transthoracic approach (either transapical or transthoracic) was made prior to randomization. 

“With this second-generation valve, TAVR, certainly from a statistical standpoint was noninferior [to surgery], and in every one of the analyses, the point estimate favored TAVR vs surgery,” coprincipal investigator Martin Leon, MD (NewYork-Presbyterian/Columbia University, New York, NY), observed. “And when you do the subgroup analyses in the transfemoral cohort, which was 76% of the patients, it was significantly better than surgery for the primary endpoint of death and significant stroke.”

Smith, for his part, characterized the difference in the transfemoral group as “borderline significant” and pointed out that while the analysis was prespecified, the trial was not powered to look at this endpoint. All the same, “in patients who are candidates for transfemoral access, TAVR may result in additional clinical advantages,” Smith noted. That said, he continued, “long-term durability assessments of transcatheter bioprosthetic valves are still lacking and extrapolation of these findings to low-risk patients requires further clinical trial evaluation.”

No difference in the primary endpoint was seen between patients treated surgically and those who underwent a transthoracic TAVR procedure.

Questioned about this point during the late-breaking clinical trial session, Smith commented: “My own guess is that, as time goes on, there will be still be some patients who can’t have transfemoral TAVR and shouldn’t have surgery and still need some other access route to have TAVR. But at the moment you’d have to say if they have to have either a transthoracic TAVR vs conventional surgery and they are reason candidates for conventional surgery, [surgery is the preferred option].”

Secondary Endpoints

A number of secondary endpoints were also notable. Rates of disabling stroke were not significantly different between groups at 2 years, despite the fact that in PARTNER 2, the US Food and Drug Administration (FDA) mandated neurology oversight.

“That meant that every patient in both groups had to be seen by a neurologist before the procedure, after the procedure, before discharge, and at follow-up, and the ascertainment of stroke is always much higher when you do that,” Leon explained. “So the fact that the absolute stroke rates were actually numerically less than the earlier PARTNER trial, even in the setting of neurology oversight, suggests that the frequency is probably going down.” 

Early data suggest that stroke rates are even lower with the lower-profile Sapien 3 device (Edwards Lifesciences), which is the subject of another late-breaking clinical trial presentation tomorrow, also in intermediate-risk patients. “We think that the lower profile device, with better positioning and with less trauma to the valve and more experience operators the stroke rates [with TAVR] may continue to decline,” he said.

None of the patients in PARTNER 2A underwent procedures involving embolic protection devices, the first trials for which are only now wrapping up. These studies may ultimately demonstrate an opportunity to reduce stroke rates still further.

 

Importantly, patients in PARTNER 2A were lower-risk than patients treated in previous trials, with STS scores in the range of 6. In the original PARTNER study, for example, the mean STS score was just under 12 for both groups. Speaking with TCTMD, however, Leon stressed that patients in this latest randomized trial were “not low-risk by any stretch.” Their average age was over 80, more than three-quarters were in NYHA class III or IV heart failure, and nearly half were determined to be in “frail condition” on the basis of a 5-meter walk test.

During the late breaker, Patrick O’Gara, MD (Brigham and Women’s Hospital, Boston, MA), challenged Smith on the use of the term “intermediate-risk” in this trial. “I look at an STS risk score of about 6%,” he said, “and it seems to me that 6% is in a very is a high-risk category for patients undergoing surgery.”

Smith, in reply, noted that he himself would have preferred an “all-comers trial” but that no one else shared his enthusiasm. As a proof-of-concept, however, he pointed to the fact that there was a doubling of patients available to undergo surgery with the criteria used, as compared to the high-risk patients studied in the first PARTNER trial. Moreover, 30-day mortality for surgery patients was cut by half from previous studies. “To me that’s proof that these were—if not intermediate-risk patients—substantially lower risk than patients at high-risk,” he commented.

Indeed, the trial design had anticipated a primary event rate of 30%, but ultimately events in both arms hovered at around 20%. Asked about this, Smith pointed to the fact that the trial had been designed based on 2010/2011 data and that evolution of the tools and advancing TAVR skills led to better results. Panel member Pieter Kappetein, MD (Erasmus Medical Center, Rotterdam, the Netherlands), himself a surgeon, then pointed out that this improvement was also seen in the surgical arm, to which Smith quipped: “Fair enough. We’re getting better too.”

The 2-year primary endpoint—a first for a TAVR randomized controlled trial—was requested by the FDA. Both the intention-to-treat and as-treated analyses were prespecified, anticipating a higher drop-out pre-procedure among patients randomized to surgery. And indeed, 17 patients randomized to TAVR did not go on to have the intervention, compared with 77 randomized to receive open surgery.

“This has happened to us again and again, and it happened also to the CoreValve study,” Leon commented to TCTMD. More patients after being randomized to surgery “back out” of the trial. To balance against the disproportionate withdrawals in the surgery arm, both as-treated and intention-to-treat analyses were built into the PARTNER 2 trial design. 

Asked by TCTMD what happens to the surgery-arm dropouts, given that the device is already approved for commercially approved, albeit for a high-risk population, Smith conceded that “a lot” or “most” may go down the street to find someone else to implant a transcatheter valve. “We’re not allowed to follow them unless they volunteer that information,” he said, “but we presume in many cases that that’s exactly what they do.”

Paravalvular leaks were significantly higher in TAVR-treated patients at all study time-points. Severe paravalvular leak rate was 3.7% in TAVR treated patients at 30 days, and this was associated with a significantly higher risk of mortality at 2 years than was seen in patients with no or trace regurgitation. Mild-to-moderate paravalvular aortic regurgitation was much more common, seen in 22.5% of TAVR patients at 30 days, but was not associated with decreased survival—a finding in contrast with other studies. “There’s no question that we can do better with paravalvular leaks, which is one reason why we think Sapien 3 is a much better device,” Leon said. 

A Consistent Message?

Neil Moat, MD (Royal Brompton Hospital, London, England), who wrote an editorial accompanying the NEJM paper, notes that there is now a “consistent message emerging” from the trials that have investigated TAVR in “high-intermediate” and very high risk- patients. “Transcatheter aortic-valve replacement is noninferior to surgery in terms of early and midterm mortality and is likely to be superior if the patient has vascular anatomy and vessels that are healthy enough to be treated with the use of a transfemoral approach.”

There is an important message in these results for patients outside of this group, however, given what Moat refers to as “an inherent excess mortality . . . associated with transthoracic access,” compared with surgical replacement.

Leon put it differently. “If you are a candidate for transfemoral, that’s good. If you are not a candidate, we can’t discern any benefit [of TAVR] other than [it being a] lower risk, beating-heart procedure. We can’t discern any particular hard-endpoint, clinical benefit compared to open surgery.” That said, the overall lower rate of acute kidney injury, major bleeding, and atrial fibrillation and the shorter ICU stay are “are meaningful,” Leon added. 

TAVR was also associated with greater post-procedure valve area, a finding Moat zeroed in on in his editorial, noting that surgeons may be continuing to favor smaller-size prostheses. The use of smaller surgical devices, leading to a smaller valve area, may not only reduce the benefits of valve replacement, they may also limit the options for subsequent valve-in-valve TAVR. 

“While the use of such small stented bioprostheses might be deemed to be acceptable in high-risk populations, it would not be appropriate in younger or lower-risk populations,” Moat writes.

Asked about valve undersizing, Smith told TCTMD that he “personally” felt that “a lot of small valves were put in.” Investigators plan to look more deeply at this issue to see if this and any differences between surgeons had any effect on outcomes, he said. Smith was adamant, however, that on the best yardstick for available for judging the quality of surgery in PARTNER 2A—actual versus predicted mortality—“this was high-quality surgery.”

Moat also raises the question of costs, noting that TAVR has proven itself cost-effective in high-risk patients, but whether this holds up in a lower-risk population “in whom the cost of conventional surgery is substantially less,” remains to be established, particularly given the high cost of current TAVR devices.

Responses to the PARTNER 2A results here at ACC 2016 were largely positive. Valentine Fuster, speaking to the media following the trial’s formal presentation, called it “unbelievably fantastic.”

Panel member Pamela Morris, MD (Medical University of South Carolina, Charleston, SC) also offered begrudging praise. “I must say, as a ‘preventionalist’ myself, I have always been skeptical of mechanical approaches to cardiovascular disease, and decades of research are really proving me wrong. I was skeptical of balloon angioplasty, I was skeptical of stents, I was skeptical of TAVR; I’ve always believed in intervening at the pathophysiologic level, but . . . congratulations.”

Smith’s response: “I’m glad you were never skeptical of surgery.”

Earlier this year, the FDA granted the go-ahead to both Edwards Lifesciences and Medtronic for trials testing TAVR in low-risk patients.