COMPARE-ABSORB: 3-Year Results Hint at Stabilization With Bioresorbable Scaffold

Three-year outcomes from COMPARE-ABSORB hint that the increased risks of thrombosis and target lesion failure with the now-discontinued Absorb bioresorbable scaffold (Abbott Vascular) do not continue to rise over time.

Rates of target-vessel MI and definite/probable scaffold thrombosis are significantly higher for the investigational device as compared with the metallic everolimus stent, but both risks appear to occur in the early period after device implantation, new data show. In a landmark analysis from 1 to 3 years, there was no significant difference in the risk of target lesion failure, target-vessel MI, or definite/probable scaffold thrombosis between the two devices.

“If you look closely at the event curves actually, already after the first month of follow-up, we don’t see any increases anymore in scaffold thrombosis and target-vessel MI with ABSORB as was seen in previous randomized controlled trials,” Pieter Smits, MD, PhD (Maasstad Hospital, Rotterdam, the Netherlands), who led the investigator-initiated study, told TCTMD. “The main reason for this, I believe, is the optimal implantation technique that was implemented from the start of the trial.”

In COMPARE-ABSORB, the “PSP” protocol, which includes predilatation, appropriate sizing (treatment of target vessels < 2.75 mm on angiography, with IVUS or OCT highly recommended), and high-pressure postdilatation (> 16 atm) and postdilatation with noncompliant balloons up to 0.5 mm larger than the scaffold was mandated.

The latest results from COMPARE-ABSORB trial were presented as a late-breaking clinical trial at CRT 2021. Smits said the landmark analyses support the rationale for further follow-up studies, which are planned out to at least 7 years, to see if there are any late advantages of the bioresorbable scaffold over time. “We know that with the current metal DES there is an ongoing linear incremental risk for TLF ranging between 1.5% and 2.5% per year,” he said. “A disappearing scaffold has the potential to flatten that line.”   

Well-known Travails of Absorb 

During his CRT presentation, Smits highlighted what is now well-known among the interventional community: use of Absorb has been consistently shown to be associated with an increased risk of MI and definite/probable scaffold thrombosis. Citing data from a 2018 meta-analysis, the risks of MI and scaffold thrombosis were still significantly higher out to 2 years, and while they attenuated between 2 and 3 years, event rates were still elevated with Absorb. 

One-year results from COMPARE-ABSORB were presented in 2018, and as was seen in other studies, Absorb was associated with nearly a twofold higher risk of target-vessel MI when compared with the Xience stent. Regarding thrombotic events, the rate of definite/probable scaffold thrombosis was more than three times higher in those treated with Absorb, with 76% of the reported definite/probable scaffold thromboses occurring within 10 days of the procedure.

COMPARE-ABSORB was unique in that investigators included patients and lesion sets not included in other trials, such as those with STEMI, NSTE ACS, bifurcation lesions, long lesions, and chronic total occlusions. During the enrollment, however, the 3-year results from ABSORB II, which was one of the first studies to raise a signal of worry, and the 2-year results from ABSORB III, which showed there was a higher risk of target-vessel MI and scaffold thrombosis with Absorb, were published. In March 2017, the US Food and Drug Administration warned of an increased risk of MACE with Absorb, while European regulators limited use of the scaffold to research settings. With these developments, the data safety and monitoring board recommended the COMPARE-ABSORB investigators put the trial on hold. A little over a week after this decision, Abbott Vascular stopped selling the scaffold.

Originally, the investigators intended to analyze whether Absorb was superior to Xience with respect to target lesion failure between 1 and 5 years. However, they extended the study’s follow-up to 7 years and revised the superiority analysis to study the period between 3 and 7 years, which captures the period after the scaffold has dissolved. This latest analysis includes data out to 3 years, including a landmark analysis performed from 1 to 3 years.

The rate of target lesion failure at 3 years was 8.9% and 7.4% with Absorb and Xience, respectively (HR 1.21; 95% CI 0.86-1.70). The risk of target-vessel MI was 61% higher for Absorb (P = 0.05), while the risk of scaffold thrombosis was more than doubled (2.4% vs 1.1%; P = 0.047).

In the landmark analysis, the risks of target-vessel MI and scaffold thrombosis were significantly increased from the time of implantation to 1 year, but there was no signal of harm between 1 and 3 years.

Gregg Stone, MD (Icahn School of Medicine at Mount Sinai, New York, NY), who was the chairman of the global ABSORB research program, said the 3-year outcomes from COMPARE-ABSORB look promising, particularly since this study included high-risk patients and those with complex lesions.

“There was an increased rate of complications in the first year, particularly in the first 30 days, but between 1 and 3 years, the event curves were parallel between the standard metallic drug-eluting stent and the Absorb bioresorbable scaffold,” he told TCTMD. “It suggests that if we could just improve the early outcomes, there would be a potential role for a disappearing scaffold, which presumably after 3 years, which is when the [poly-L-lactide acid scaffold] is essentially gone, would leave the coronary artery with its underlying architecture.” 

Technique Is Important

The results from COMPARE-ABSORB, specifically the landmark analyses, differ slightly from the findings of a meta-analysis of individual patient-level data from ABSORB II, Absorb Japan, Absorb China, and ABSORB III. In that study, which was led by Stone and Ziad Ali, MD (NewYork-Presbyterian/Columbia University Irving Medical Center, New York, NY), use of the bioresorbable scaffold was associated with an increased risk of target lesion failure and device thrombosis in midterm follow-up at 3 years, and these risks remained significantly higher in a landmark analysis from 1 to 3 years. However, in further follow-up at 5 years, the landmark analysis revealed there was no difference in the risk of target lesion failure, target-vessel MI, or device thrombosis between 3 and 5 years.   

“After 3 years, once the scaffold is completely resorbed, the events are certainly no greater, and may even be improved, with the bioresorbable scaffold technology,” said Stone, referring to their study.

As for why they saw a continued difference in events between 1 and 3 years, while the COMPARE-ABSORB trial did not, might be related to the PSP technique, he added. “Specifically, they used high-pressure postdilatation in more of their patients and they used postdilatation with an oversized balloon in the majority of their patients. We had previously shown that those techniques in particular are associated with better long-term outcomes, likely because they prevent acute malapposition, which should prevent intraluminal scaffold dismantling.” 

Both Stone and Smits remain bullish on the technology’s future. Smits noted Absorb was a first-generation scaffold with thick struts and it was up against Xience, which was one of the best newer-generation metallic DES. From a proof-of-concept standpoint, there isn’t a need for a permanent device to treat coronary artery stenoses, and some patients at high risk for restenosis—those with long lesions, bifurcations, and with diabetes—might benefit from a temporary scaffold, he added.

Going forward, Stone said bioresorbable scaffolds will need to have thinner struts so they endothelialize faster and have less side-branch coverage, which will result in fewer periprocedural MIs, and operators will need to avoid small vessels, and use optimal implantation techniques.