ABSORB III: Two-Year Results Show a Higher MACE Rate Compared With Xience

WASHINGTON, DC—Two-year results from the ABSORB III study presented today at the American College of Cardiology 2017 Scientific Session show that the Absorb GT1 bioresorbable vascular scaffold (BVS, Abbott Vascular) is associated with an increased risk of target lesion failure (TLF) when compared with a conventional metal stent.

At 25 months, TLF occurred in 11.0% of patients who received the Absorb scaffold compared with 7.9% of patients who received the everolimus-eluting stent (Xience, Abbott Vascular), a difference that was statistically significant. The increased risk of TLF was driven by an increased risk of target-vessel myocardial infarction (7.3% versus 4.9%; P = 0.04).

Additionally, the rate of definite/probable scaffold/stent thrombosis (ST) occurred in 1.9% of patients treated with Absorb and 0.8% of patients who received the Xience stent, a nonsignificant difference. Between years 1 and 2, the rate of ST in the Xience-treated patients was zero, but four ST events (0.3%) occurred in patients who received the Absorb BVS.

In advance of the late-breaking clinical trial session, the US Food and Drug Administration (FDA) issued a letter to physicians informing them of the increased risk of major adverse cardiac events with the Absorb stent compared with Xience. The FDA reminded physicians to adhere to the approved indication and not implant the device in small vessels.

“I think the clinical community should be aware not only of the pros but also the cons,” Stephen Ellis, MD (Cleveland Clinic, OH), co-principal investigator for ABSORB III, told TCTMD. “We have learned from the careful analysis of the 1-year data that one of the roadblocks [to optimal performance] is vessel size, and it’s pretty easy to skirt around. There is accumulating data that technique makes a big, big difference.”

Like all previous discussions about the Absorb BVS technology, Ellis emphasized the device is a first-generation scaffold, with thicker struts and different expansion and structural characteristics than contemporary drug-eluting stents. Since its European approval in 2011, investigators have increasingly recognized the importance of adequate lesion preparation, treating appropriately sized vessels, and postdilatation to achieve success with the device.

To TCTMD, David Kandzari, MD (Piedmont Heart Institute, Atlanta, GA), said the FDA “Dear Doctor” letter is an appropriate response given the data from ABSORB III. “Now that we have statistical significance for an endpoint like target lesion failure, it’s a cautionary note to raise awareness of the difference [with Xience] and to ensure that practitioners adhere to appropriate techniques, standards, and protocols, including patient selection,” he commented.

Kandzari noted that when the Cypher sirolimus-eluting stent was first introduced to US physicians, there were widespread reports of acute thrombosis, largely attributable to implantation technique and lack of appropriately sized stents. At the time, the FDA issued a number of safety reports outlining the potential risks.

“In many ways, there is a great deal of similarity to this as well,” he said. “In comparison to the Cypher experience, the application in clinical practice of BVS is still very, very low. Its use is not as widespread as the first drug-eluting stent.”

Brent Muhlestein, MD (Intermountain Medical Center, Salt Lake City, UT), who was not involved in the study, said the increased risk of TLF at 2 years did not take him by surprise given that the 1-year ABSORB III results suggested differences in the rate of TLF between Absorb BVS and Xience (7.8% in the Absorb arm vs 6.0% in the Xience arm; P = 0.15). He also said the 2-year data are “not terrible,” since the rate of TLF remains low and there was no significant effect on cardiac mortality.

“Still, I would also certainly say I’m not ready to start putting it into my patients, even though it is available at the present time,” said Muhlestein. “I’m willing to wait.”

Analyzing Data by Vessel Size

In 2016, the FDA approved the Absorb stent for coronary artery disease patients with a reference vessel diameter (RVD) ≥ 2.5 mm and ≤ 3.75 mm. In ABSORB III, approximately one in five patients who received Absorb had an RVD less than 2.5 mm.

During the late-breaking clinical trial session, Ellis attempted to really drive home critical differences in clinical outcomes when physicians implanted the scaffold in patients with an appropriately sized vessel. He noted that while the TLF rate was significantly higher among the Absorb-treated patients in the entire cohort, when they analyzed data from patients with an RVD ≥ 2.5 mm, there was no such signal.

Among patients with these larger vessels, the TLF rate was 9.4% among the Absorb-treated patients and 7.0% among the Xience-treated patients, a difference that was not statistically significant (HR 1.35; 95% CI 0.93-1.96). Similarly, in those with an RVD ≥ 2.25 mm, the 2-year rate of definite/probable ST was 1.3% and 0.6% in the Absorb- and Xience-treated patients, respectively. Again, this difference was not statistically significant.

An additional analysis of the data showed that when physicians followed the PSP protocol (predilatation, appropriate sizing, and postdilatation), the rates of TLF and ST in the Absorb BVS arm were much closer to rates observed with the Xience stent.

Commenting on the 2-year ABSORB III results, Dipti Itchhaporia, MD (Newport Coast Cardiology, Newport Beach, CA), said that with any revolutionary technology, there is a period of education for early adopters. In this case, that includes learning about avoiding small vessels and the use of appropriate implantation techniques. However, Itchhaporia said, “it’s hard to argue against the data,” noting the extremely low rate of ST with Xience, including the complete absence of ST between 1 and 2 years.

Not only that, Absorb BVS is costlier, takes longer to implant, and requires the additional expense of intravascular ultrasound, she said.

“As a clinician, am I going to run out and be implanting more Absorb stents?” she asked. “Probably not. But I also think this is in its infancy and these are the growing pains we have with any new technology where we learn lessons. It’ll get better, but unfortunately we have to go through these growing pains to learn these lessons.”

To TCTMD, Kandzari said that when the FDA issues warnings via letters, there typically tends to be an increase in reporting of adverse events to the agency and an overall decline in utilization of the drug or device.    

FDA Indication is Strict

During the presentation, Ellis presented a snapshot of blinded, pooled interim data from the ongoing ABSORB IV study, in which 2,494 patients were treated with the Absorb BVS or the Xience everolimus-eluting stent. In this analysis, said Ellis, just 4% of patients in the trial had an RVD less than 2.25 mm on QCA and 84% received optimal high-pressure postdilatation.  

These pooled Absorb/Xience data also showed the overall rates of scaffold/stent thrombosis for the cohort were 0.4% at 30 days and 0.5% at 1 year.

“The obstacles are surmountable,” said Ellis. “I’ll be the first to admit, though, that we don’t know for sure if this device will do better over the long haul. It’s one of those things in medicine where we don’t have all the data, it’s a bit of a moving target, and I think the physician needs to have a careful and somewhat nuanced discussion with the patient before they decide what type of stent or scaffold to put in.” 

During the afternoon press conference, Spencer King III, MD (St. Joseph’s Health System, Atlanta, GA), wondered why the 2-year ABSORB III data were making such a splash, stating the scaffold is not widely used in clinical practice and many of the concerns about the technology were already known. Even Ellis appeared taken aback by some of the media’s questions.

“To be honest with you, I think the ABSORB II and ABSORB-Japan data raised some concerns, but if anything, this calms the water,” said Ellis.

Specifically, he said the rate of ST between 1 and 2 years in ABSORB III—0.3%—is the lowest ever reported. In ABSORB II, ABSORB-Japan, and ABSORB-China, the rate of ST between 1 and 2 years was 0.6%, 1.6%, and 0.4%, respectively. “I understand there are concerns, and I think the concerns are legitimate, but I think we have enough data to suggest this is not a dangerous device,” Ellis said. “We need more data, and we’re only going to get more data if we use it.”

Antonio Colombo, MD (Columbus Hospital/San Raffaele Hospital, Milan, Italy), a panelist during the Main Tent presentation, said the Absorb scaffold began with “problems,” but investigators have learned the importance of good implantation in appropriately sized vessels. Although he views the glass as half full, he said, the device is still not as good as the best-in-class drug-eluting stents.

“It’s not appropriate to pretend this device is noninferior to Xience, which is the gold standard,” said Colombo.

Roxana Mehran, MD (Icahn School of Medicine at Mount Sinai, New York, NY), also emphasized that Absorb is a breakthrough technology, but added the clinical promise of a disappearing stent has yet to be realized. Disappointing 3-year data from ABSORB II, previously reported by TCTMD, showed treatment with Absorb was associated with a twofold increased risk of device-oriented clinical events, specifically an increased risk of target-vessel MI, as well as an increased risk of late ST compared with Xience.

Given the ABSORB II data, and the data from Absorb III showing an inferior result, Mehran questioned what physicians should do with patients who currently have an Absorb scaffold. For example, how long should they stay on dual antiplatelet therapy?

Ellis, however, didn’t have an answer and emphasized physicians should personalize care. He said he’d be more comfortable stopping dual antiplatelet therapy in patients if they had an excellent implantation, where physicians adhered to the PSP protocol, or if the patient underwent confirmatory IVUS imaging.

Photo Credit: Cleveland Clinic