Scaffold Thrombosis ‘Unpredictable’ With Absorb: AIDA Analysis

Unlike in the ABSORB studies, even adhering to good implantation techniques failed to limit the risk of device thrombosis.

Scaffold Thrombosis ‘Unpredictable’ With Absorb: AIDA Analysis

PARIS, France—An in-depth analysis of a recent trial testing the Absorb bioresorbable vascular scaffold (BVS; Abbott Vascular) in an “all-comers” population suggests the risk of scaffold thrombosis is “unpredictable,” with investigators unable to identify any patient or procedure-related variables associated with the complication.

Adhering to the PSP protocol, which includes adequately preparing and sizing the vessel for scaffold implantation and using postdilatation, did not significantly lower the risk of clinical events, including definite/probable scaffold thrombosis, when compared with not following the protocol.  

Presenting the results of the Amsterdam Investigator-Initiated Absorb Strategy All-Comers Trial (AIDA) at EuroPCR 2017, lead investigator Joanna Wykrzykowska, MD, PhD (Academic Medical Center, Amsterdam, the Netherlands), cautioned that operator technique doesn’t tell the full story when it comes to explaining the risk of scaffold thrombosis with the bioresorbable device. 

“Technique can obviously improve things in the short term for early events, but for late events, I don’t think we understand the biology,” she said. “We don’t understand the interaction of the device and the biology of the vessel. It’s extremely complex.”

She suggested that physicians still using the device—it is restricted in Europe to ongoing clinical trials and open registries—“do the best job you can angiographically, which means predilatation, sizing, and postdilatation.” And while adhering to the PSP protocol can improve acute results, “the story is not that simple,” said Wykrzykowska.

Data Sweep Following Safety Concerns   

Following safety concerns raised by the 3-year results from ABSORB II, the AIDA investigators performed an early analysis of their data and published them in March in the New England Journal of Medicine. Those results, which were reported by TCTMD, showed that treating a wide range of patients with the Absorb BVS—55% of patients presented with an acute coronary syndrome—was associated with an increased risk of definite/probable scaffold thrombosis compared with the Xience everolimus-eluting stent (Abbott Vascular), as well as with a greater risk of target-vessel MI.

Proponents of Absorb have contended that the risks of scaffold thrombosis are mitigated with good implantation techniques. Gregg Stone, MD (Columbia University Medical Center, New York City, NY), previously presented pooled data from the ABSORB trials showing that optimal vessel sizing was associated with a significantly lower rate of target lesion failure at 3 years, while there were trends toward lower rates of scaffold thrombosis when operators followed the PSP protocol.

To TCTMD, Wykrzykowska said the AIDA patients are a complex, real-world population and that operators are usually aggressive in treating such patients, including the use of predilatation and postdilatation. In the analysis presented at EuroPCR, Wykrzykowska said that despite such aggressive treatment, they didn’t identify any predictors of device thrombosis.

For example, the risk of device thrombosis remained statistically significant with Absorb when investigators stratified their results by vessel size. In small and large vessels, those greater than 2.75 mm in diameter, Absorb was associated with a significant risk of device thrombosis. Similarly, ACS and non-ACS patients both had an increased risk of device thrombosis, as did patients treated early in the randomization process and those treated at later dates, when operators gained experience and understanding of the importance of good implantation techniques.     

The researchers also failed to identify any predictors of scaffold thrombosis when looking at the quantitative coronary angiography analysis for the Absorb-treated patients. Among those with definite scaffold thrombosis and those without, there was no difference in the use of predilatation and postdilatation and no difference in the reference vessel diameter. Overall, 220 patients received a scaffold in a vessel ≤ 2.25 mm—which is contraindicated for Absorb—but just three of these patients had definite scaffold thrombosis (P = 0.25 vs those without scaffold thrombosis).  

The researchers also examined clinical outcomes among patients with vessels 2.25 mm or smaller and those with larger vessels, but observed no significant differences. Device thrombosis occurred in 2.9% of patients with a scaffold in a vessel ≤ 2.25 mm and 3.3% of those with Absorb in larger vessels (P = 0.67).

“I think sizing is probably the most important aspect of PSP, but we looked at the PSP score, which has been put forward as a way to combine all the technical [implantation] aspects, but we didn’t see any predictors at the lesion level,” said Wykrzykowska. “Whether we followed the PSP model or not, the risk of thrombosis is still 3% in both groups. There’s a numerical difference when we look at the patient level, but that’s not significant. It gives you a little bit of a pause intellectually, because it should be even better at the lesion level, not the patient level, because that’s what you’re treating—the lesion.”

Investigators also stratified patients into three groups based on their SYNTAX risk score. For patients at low risk—those in the bottom tertile of SYNTAX scores—the risk of scaffold thrombosis was 1.1% among Absorb-treated patients. To TCTMD, Wykrzykowska said this number might “be acceptable, at least numerically,” but the risk of stent thrombosis was just 0.3% in the Xience-treated patients with low SYNTAX scores.

Future “Not Hopeless”   

Commenting on the findings during the clinical trials update session, Mahmoud Hashemian, MD (Day Hospital, Tehran, Iran), said the Absorb data to date are unable to compete with those seen with second-generation drug-eluting stents. In Iran, the device had been used “considerably” but after the presentation of the recent studies showing a risk of late and very late scaffold thrombosis, physicians are no longer using it, he said.

“But that doesn’t mean we have to be hopeless,” said Hashemian. “I’m not hopeless at all. I’m sure this will be the future of our stents, but companies need to work more on it. [Abbott] told me they are going to launch a newer [Absorb], maybe next year, with thinner struts and more expandability, and that’s good news.”

In a statement to TCTMD, Abbott Vascular stated that AIDA took place when the implantation techniques were not fully understood or developed. They argue that the postdilatation balloon pressures used in the study—15.4 atm in Absorb-treated patients—are lower than what is considered optimal. “Analyses of Absorb trials from around the world demonstrated that when implanted according to current instructions for use there was a less than 1 percent rate of stent thrombosis,” said Abbott Vascular’s spokesperson.

Michael O’Riordan is the Associate Managing Editor for TCTMD and a Senior Journalist. He completed his undergraduate degrees at Queen’s…

Read Full Bio
  • Wykrzykowska JJ, Kraak RP, Tijssen RYG. Amsterdam Investigator-Initiated Absorb Strategy All-Comers Trial. Presented at: EuroPCR 2017. May 19, 2017. Paris, France.

  • Wykrzykowska reports receiving grants/research support from Abbott vascular and consulting for Abbott Vascular and St. Jude Medical.



Alexander Kharlamov

6 years ago
I would remind that if calculate a sample size or estimate a statistical power from ABSORB trials, the AIDA trial was underpowered to evaluate a rate of the scaffold thrombosis. The second point is that Amsterdam treated truly real world population but "during the first year of enrollment, scaffolds were implanted according to the manufacturer’s instructions, which, at that time, did not include mandatory postdilation; postdilation was performed in 63% of the lesions in the scaffold group during the first year of enrollment." It means also that only selected patients were properly prepared (mostly without pre-dilation or advanced [with multimodal imaging] sizing) for implantation. The AIDA trial does not make that much difference for our understanding of the threats from BVS.