‘Worrisome’ 3-Year Data From ABSORB II Raise Concerns About First-Generation Bioresorbable Stent

WASHINGTON, DC—Three-year outcomes from the ABSORB II trial, a time point in which the Absorb GT1 bioresorbable vascular scaffold should be degraded, did not result in an improvement in vasomotor tone and was associated with an increase in late lumen loss when compared with the Xience everolimus-eluting metallic stent.  

The study also showed that treatment with Absorb (Abbott Vascular) was associated with a two-fold increased risk of device-oriented clinical events, specifically an increased risk of target-vessel MI (7% vs 1%; P = 0.006), as well as an increased risk of late scaffold thrombosis compared with Xience (Abbott Vascular). There were six incidents of definite scaffold thrombosis occurring beyond 365 days among patients who received the Absorb stent compared with no reported cases of definite or probable stent thrombosis for patients who received the Xience stent. Approximately one-third of patients in both treatment arms were on dual antiplatelet therapy at 3 years.

Presenting the ABSORB II results at TCT 2016, lead investigator Patrick Serruys, MD (Imperial College, London, England) said he was disappointed in the findings. “It’s not what we were expecting,” he said, noting the results don’t line up with 5-year outcomes from 101 patients participating in the first-in-human trials with the Absorb stent. “That’s the reality,” he told TCTMD. “Of course, we’re still in the first generation. I think, in general, physicians and patients would like to have something that disappears. Conceptually, if there was no setback or tradeoff, they would [choose] that. From what I’ve seen over the past few years, there is room for improvement.”

That said, though, Serruys appeared to acknowledge the new results as a setback for the current device, particularly the emergence of “devastating and unpredictable events” in a group of patients with relatively simple lesions. “I mean, the patient is doing well, they have exercise testing every year to prove that they were OK, and then suddenly chest pain and it’s STEMI,” he said. “It’s a full-fledged myocardial infarction. That’s quite frightening.”

Paul Teirstein, MD (Scripps Clinic, La Jolla, CA), who was not involved in the trial, told TCTMD he was also taken aback by the 3-year data. “Look, I think we need to have some perspective on the device,” he said. “You don’t want to give up on it, but this particular device, [the results] are worrisome.”

Like others, Teirstein said he was surprised and disappointed, adding it will take time to digest the findings and process how they will impact his clinical practice. “Different physicians have different reactions to the bioresorbable stent, in general,” he said. “I have been fairly reserved about it, personally. I have a patient population referred to me that is very complex, doesn’t fit in [the ABSORB] trials—a lot of calcium, a lot of bifurcations, a lot of [chronic total occlusions]. I haven’t been using that many. How will I use this information today going forward? It certainly won’t increase my use.”

Also speaking with TCTMD, Tommaso Gori, MD (University Medical Center, Mainz, Germany), who was not involved in the trial, echoed Teirstein, saying the extended data from ABSORB II are a letdown. “I guess a lot of people will be unhappy with the findings,” he said. “It is very disappointing.”

Gori noted that even in the ABSORB III study, the large-scale, randomized trial showing Absorb was noninferior to Xience regarding the primary endpoint of target lesion failure at 1 year, there had been some discussion, including commentary at the US Food and Drug Administration advisory panel meeting, that event curves were going in the wrong direction. Although ABSORB II was underpowered for clinical outcomes, the event curves again go in the wrong direction, said Gori.  

“I mean, six [definite scaffold thromboses] is 2% and it’s not a huge number, but it is disturbing,” he said. “It shouldn’t have happened.”

The Data From ABSORB II

ABSORB II, a prospective, randomized, multicenter study, included 501 patients with one or two de novo coronary lesions randomized 2:1 to treatment with the Absorb bioresorbable stent or the Xience metallic stent. The 3-year results, which are published in the Lancet to coincide with the TCT presentation today, included data on 335 patients treated with Absorb and 166 patients treated with Xience.  

Regarding the coprimary endpoint, vasomotion assessed by the change in mean lumen diameter after the administration of intracoronary nitroglycerine, Absorb was not superior to the Xience stent, an “unexpected” finding given the rate of resorption, according to the researchers. Late lumen loss, the other coprimary endpoint, was larger among patients in the Absorb arm (0.37 mm vs 0.25 mm; P = 0.78 for noninferiority). There was also no significant improvement in angina status or exercise performance among patients treated with the two stents.

In terms of device-oriented clinical events, a composite of cardiac death, target-vessel MI, and clinically indicated target lesion revascularization, occurred in 10% of Absorb patients compared with 5% of Xience patients (HR 2.17; 95% CI 1.01-4.69). The difference was driven largely by a significantly increased risk of target-vessel MI among the Absorb-treated group. Clinically indicated target lesion revascularization occurred in 6% of the Absorb-treated patients and 2% of those treated with Xience (P = 0.04).

In addition to ABSORB II, lead investigator Runlin Gao, MD (Fu Wai Hospital, Beijing, China), presented 2-year data from ABSORB China, a study that included 480 subjects enrolled at 24 sites. The results showed no difference in device-oriented clinical events between the Absorb- and Xience-treated patients. In the Absorb arm, there was one patient who had a very late stent thrombosis between 1 and 2 years, as well as one patient with an early definite scaffold thrombosis (< 30 days).

To TCTMD, Gori said that late clinical events, those beyond 1 year, are unlikely related to implantation technique—Absorb requires careful implantation with imaging and vessel preparation, an issue that has been documented by experts—although vessel inflammation after resorption might remain an issue. Serruys said acute and subacute thrombosis is definitely associated with implantation technique, but what happens 3 years later is “somewhat mysterious.”

For their part, Abbott Vascular, contend the adverse outcomes are related to problems with operator technique. Optimal technique includes a combination of predilation, appropriate vessel sizing, and postdilatation. Speaking with TCTMD, Abbott suggested in their preliminary analysis of ABSORB II, the adverse events (target lesion failure and stent thrombosis) occurred in patients who did not receive optimal stent implantation.

Teirstein, who spoke during the morning press conference, said he doesn’t want to lose confidence in the Absorb stent, which was approved by the US Food and Drug Administration in July 2016 and has been available in Europe since 2011. That said, he will likely use the device in young patients with large vessels, “and I’m going to be very, very careful with my implantation technique,” he stressed. He openly questioned whether patients might not need dual antiplatelet therapy for 3 or 4 years after receiving the Absorb stent.

Don’t Want to Lose Confidence

For Robert Byrne, MD, and Adnan Kastrati, MD (Deutsches Herzzentrum, Munich, Germany), who wrote an editorial accompanying the study, the findings add to earlier concerns about the risk of stent thrombosis with Absorb, including other published reports showing a two-fold increase risk of stent thrombosis. They caution that while the clinical outcomes data are concerning, even worrisome, ABSORB II was not powered for clinical endpoints and “it remains to be seen whether the adverse safety signal is a real finding.”  

The editorialists point out that the hypothesized benefits of the bioresorbable scaffold, which would include reestablishing normal vasomotion, should have been “readily detectable.”

To TCTMD, Gori noted the study suggested there is a degree of vasomotion among the patients treated with Absorb, roughly the same observed in the nonrandomized trials, but the surprise is that there was also vasomotion observed in the Xience-treated patients. As a result, there was no detectable between-group differences for the two arms. However, the more important findings are those showing a higher risk of binary restenosis, MI, and scaffold thrombosis with the bioresorbable stent.

In his practice, Gori said he has followed patients for 4 years and has not observed late clinical events. “Very late thrombosis is something I don’t see,” he said. “Should I be reluctant? Of course, randomized studies are better done than simply following up with patients that come on a clinical basis, and any bad news from a clinical trial is bad news. It is disturbing—disturbing, for sure.”

Going forward, Byrne and Kastrati say ongoing clinical trials with bioresorbable scaffolds should schedule additional evaluations to report 3-year data as that is the “best chance to assuage concerns arising from the observations of Serruys and colleagues.” There also remains a need to redefine the optimal intensity and duration of dual antiplatelet therapy in patients treated with the devices.

To TCTMD, Serruys noted that all new technologies, including balloon angioplasty, bare-metal stents, and drug-eluting stents, went through early rough patches. However, he believes in the need for a disappearing stent that will allow for repeat procedures, if needed. He said there are currently more than two dozen companies developing their own bioresorbable vascular scaffolds. Future bioresorbable scaffolds, he said, will be engineered with thinner yet stronger struts and will likely have a faster rate of absorption.

“We’re currently in 2016, and I suspect, give us another 4 or 6 years and we’ll see some serious improvements,” he said. 

Photo credit: Dr. Ziad Ali.