Absorb BVS: Better Technique, Better Outcomes? More Data—and More Debate

WASHINGTON, DC—A large analysis of clinical trials testing the Absorb bioresorbable vascular scaffold (BVS; Abbott Vascular) provides new evidence that operator technique is critically important to achieving good clinical outcomes. These new details, however, come as many in the field are struggling with unanswered questions raised in studies of this technology over the last 18 months.

In five studies testing the Absorb scaffold—ABSORB II, ABSORB-Japan, ABSORB-China, ABSORB III, and ABSORB-Extend—optimal vessel sizing was associated with a significantly lower rate of target lesion failure (TLF) at 3 years, while there were trends toward lower rates of scaffold thrombosis when operators adequately prepared the lesion, treated only appropriately sized vessels, and provided adequate postdilatation after the stent was implanted.  

“The first five studies were done before the learning curve,” lead investigator Gregg Stone, MD (Columbia University Medical Center, New York, NY), told TCTMD. “And that’s reflected by the fact that a lot of very small vessels were treated and there were very low rates of postdilatation. In ABSORB IV, that study is going to reflect [what physicians have learned] about the stent.”

Pooled Data and Importance of “PSP”

Presenting the data at CRT 2017 here, Stone noted that the Absorb BVS is a first-generation technology, 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.

In fact, Absorb has been likened to a first-generation DES, with operators stating that technique is critical to achieving good clinical outcomes.

In their analysis, Stone and colleagues specifically assessed the effect of predilatation, appropriate sizing (QCA reference vessel diameter ≥ 2.25 mm to ≤ 3.75 mm), and postdilatation with a high-pressure noncompliant balloon. In recent years, such an implantation strategy has been referred to as “PSP” (prepare the lesion, size, and postdilatation). The five pooled studies included 2,973 patients treated with the Absorb scaffold.

The first five studies were done before the learning curve. Gregg Stone

Across the combined cohort, Stone noted that operators treated a sizeable number of patients with very small vessels. Indeed, overall, just 82.3% of patients in the five studies had an optimally sized vessel for Absorb BVS. Furthermore, just 60.1% received predilatation and, depending on the definition of postdilatation, as few as 12.4% of patients received adequate high-pressure postdilatation with a noncompliant balloon.  

Stone and colleagues also looked at the relative impact of different implantation techniques across three different periods of time. In the first year, Stone said, optimal vessel sizing was significantly associated with a reduced risk of TLF and scaffold thrombosis. In contrast, predilatation was associated with an increased risk of TLF and scaffold thrombosis. Between 1 and 3 years, optimal sizing, predilatation, and postdilatation were all associated with trends toward lower TLF, whereas predilatation was associated with a significantly reduced risk of scaffold thrombosis. Finally, when looking at 3-year follow-up in its entirety, optimal sizing was associated with a lower risk of TLF, while all three components of PSP were associated with trends toward a lower risk of scaffold thrombosis.

At 3 years, the rates of TLF and definite/probable scaffold thrombosis were 10.8% and 2.7%, respectively.

Speaking during the late-breaking clinical trials session, Stone said the goal after implantation is to obtain a large minimal lumen diameter. Even though the data only shows trends toward a lower rate of scaffold thrombosis and TLF with postdilatation after implantation, “I strongly believe it’s true,” said Stone.

Lessons Learned?

There are early signs that operators have learned their lessons from ABSORB III and are applying these in ABSORB IV. As part of his CRT presentation, Stone provided a snapshot of pooled, blinded 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 (Abbott Vascular). According to Stone, just 4% of patients randomized in the trial had a reference vessel diameter less than 2.25 mm on quantitative coronary angiography (QCA) and 83% of the Absorb patients received optimal high-pressure postdilatation.  

Most importantly, the pooled Absorb/Xience data showed the rates of scaffold/stent thrombosis were 0.3% at 30 days and 0.5% at 1 year.

Although the ABSORB IV data are blinded—meaning it is unknown whether rates were lower with one device than the other—the results appear lower than those seen in ABSORB III. In that trial, the pooled stent/scaffold thrombosis rates among Xience- and Absorb BVS-treated patients were 0.9% at 30 days and 1.1% at 1 year. Notably, 19% of patients in ABSORB III had the scaffold implanted in a very small vessel (< 2.25 mm).

“These [ABSORB IV] data are very reassuring,” said Stone. “I mean, it’s a two-thirds reduction in 30-day thrombosis rates and more than a 50% reduction in thrombosis at 1 year. Of course, it’s blinded and pooled so we don’t know which device [improved], but maybe it’s both.”

Looking Forward, Looking Back

Other investigators are also looking to previous studies, hoping to glean additional insights that will help move the technology forward. In another session devoted to the ABSORB trials, Patrick Serruys, MD (Imperial College, London, England), presented a new analysis from ABSORB II, zeroing in on predictors of expansive remodeling in that trial.

As reported by TCTMD, 3-year results from ABSORB II garnered a lot of attention at TCT 2016 when the longer-term follow-up showed a larger increase in late lumen loss and adverse safety signal with the Absorb BVS as compared with Xience. In this analysis, said Serruys, implantation with the Absorb stent, as well as female sex, previous PCI, higher LDL cholesterol levels, a larger preprocedural necrotic core, as well as some procedural and postprocedural factors, were all associated with expansive remodeling.

Speaking during a session devoted to the Absorb technology, Serruys said expansive remodeling with lumen area enlargement following the loss of mechanical support has been documented previously with the stent. The analysis of ABSORB III was intended to confirm that the Absorb BVS has the potential for more vessel remodeling when compared with Xience, said Serruys.

The results, he said, show that “lesions treated with Absorb BVS exhibit frequently larger increases in vessel area and lumen area than lesions treated with Xience.” The clinical impact of the observed lumen and vessel remodeling will need to be studied in future studies, particularly when it comes to optimizing patient outcomes, he added.

Not There Yet  

Despite the comforting signs that improvements to technique may lead to better results with the Absorb BVS, much of the focus at CRT 2017 was on the safety signals that have arisen in the last 18 months.

In one session, Robert Byrne, MD (Deutsches Herzzentrum, Munich, Germany), cited data from several meta-analyses published in 2016 showing the Absorb scaffold is associated with early TLR rates that are comparable to Xience, but also with an approximate two-fold increased risk of scaffold thrombosis.

“What we can see is that when we’re looking at the first year, there is definitely a time-dependent element, a strong interaction with time, such that most of the risk is in the early period at 30 days,” said Byrne. “This tells us that this is clearly related to the procedure and offers hope that it’s something we can modify with our implantation techniques.”

What is less clear is the impact optimal technique—the adherence to the PSP protocol—will have on late clinical events. Byrne said that optimists, a group to which he belongs, expect to see advantages for patients when the stent disappears, but there are reports of late and very late scaffold thrombosis, events that occur after dual antiplatelet therapy has been stopped.

Byrne also pointed to the 3-year data from ABSORB II showing a two-fold increased risk of device-oriented clinical events, specifically an increased risk of target-vessel MI and clinically-indicated TLR. Additionally, there was an increased risk of very late scaffold thrombosis compared with Xience, including six cases of definite scaffold thrombosis occurring beyond 365 days with Absorb (and none with Xience).    

Until concerns . . . can be assuaged, the current generation of bioresorbable scaffolds should not be preferred to DES in routine clinical practice. Robert Byrne

Regarding late events, Byrne presented unpublished data from five clinical trials with 2 years or more clinical follow-up (ABSORB-China, ABSORB II, ABSORB-Japan, EVERBIO II, and TROFI II) and showed the rate of TLR was significantly higher with Absorb compared with Xience (OR 1.73; 95% CI 1.09-2.76), a finding that suggests a long-term “erosion in TLR” over time. In those trials, the rate of definite/probable scaffold thrombosis was nearly three-fold higher with Absorb (OR 3.08; 95% CI 1.36-7.04).

“My impression is that until concerns, particularly the late performance of [Absorb] BVS, can be assuaged, the current generation of bioresorbable scaffolds should not be preferred to DES in routine clinical practice,” said Byrne. “We should also think seriously about patients who have already received the device in that they might be considered for prolonged DAPT, out to 3 years.”   

What Happens in the Meantime?

Ron Waksman, MD (MedStar Heart and Vascular Institute, Washington, DC), who chaired the session highlighting data from several ABSORB clinical trials, asked Byrne what data he would want to see to ease his anxiety about the Absorb BVS. In an ideal world, Byrne said he’d like to see a randomized controlled clinical trial with operators routinely employing optimal implantation techniques (PSP), even though he acknowledged such a trial is unlikely.

 “I think what we’ll really need to see is the influence of PSP data prospectively collected from a randomized trial, maybe ABSORB IV, to show us if this is a problem that has been dealt with,” said Byrne. “With regard to the later outcomes, of course, it’s going to require some time. It may well be that scaffolds that are better implanted, more deeply implanted into the growing neointima, are protected from mechanical discontinuity when they break down between two and three years and perhaps this signal will go away. But for me, that’s a lot of ‘ifs’ at the moment.”  

In the meantime, Byrne agrees with the recommendations of the UK National Institute for Health and Care Excellence, which advises physicians to ensure patients receiving the bioresorbable scaffold “understand the uncertainty about the procedure’s safety and efficacy in the longer term and provide them with clear written information.”

Dean Kereiakes, MD (The Christ Hospital, Cincinnati, OH), a principal investigator for ABSORB III, however, stressed that the rationale for optimal sizing of the stent and high-pressure postdilatation is to embed the scaffold and obtain optimal endothelial coverage of the device. These, he predicts, will have an impact on late events.

Also speaking during the ABSORB session, Ted Bass, MD (University of Florida, Jacksonville), called the first-generation device a great advance still in its infancy. “The more I’m hearing, the more I’m convinced that’s the case,” said Bass. “We’re still working out procedural aspects, some of the adjunctive therapies, but there seems to be a clear and consistent signal from meta-analyses and many of the trials that there is a problem early and there may be a problem late.”

That said, Bass noted that as more and more observational studies are published, it appears physicians were using the device in patients who were not good candidates, or who were not studied in clinical trials. These include patients with bifurcation, calcified, and left-main lesions.

“It’s not fair to the patients to push the envelope when we still have a lot of unanswered questions about this,” said Bass.  

It’s not fair to the patients to push the envelope when we still have a lot of unanswered questions. Ted Bass

Nick West, MD (Papworth Hospital, Cambridge, England) raised some concerns about prolonged DAPT out to 3 years, a strategy suggested by Byrne, saying that until researchers truly understand the cause of late clinical adverse events with Absorb BVS, “we really need to be cautious about exposing patients to increased bleeding risks, as we saw in the DAPT study, rather than the risk of scaffold thrombosis.” While prolonged DAPT is an interesting idea, there are no data showing that prolonged treatment would mitigate the risk of late scaffold thrombosis events with Absorb BVS, said West.  

Despite some of the concerns about the Absorb device, other registry data provided some reassurance. In the late-breaking clinical trials session, Robert Jan van Geuns, MD (Erasmus Medical Center, Rotterdam, the Netherlands), presented 2-year data from a propensity-matched analysis of Absorb BVS and Xience in the BVS-Expand registry. In that analysis, the rate of definite scaffold thrombosis with Absorb was 1.3% at 2 years compared with 0.9% with Xience (P=0.574), while a patient-oriented composite endpoint of death, MI, and revascularization showed equivalent rates between the two stent arms.

Earlier at CRT, Alexandre Abizaid, MD (Institute Dante Pazzanese de Cardiologia, Sao Paulo, Brazil), presented similar results from 812 patients enrolled in the ABSORB-Extend registry. At 2 years, after propensity-score matching, there was no difference in MACE, TLR, target vessel failure, and scaffold/stent thrombosis rates between the Absorb- and Xience-treated patients.