ABSORB IV: End in Sight for BVS Stent Thrombosis Woes?

WASHINGTON -- Tip-top technique may be what it takes to bring down the rate of stent thrombosis with the use of the Absorb bioresorbable vascular scaffold, a first glimpse at data from the ABSORB IV trial suggested.

Although the study is still enrolling patients, pooled results from participants randomized so far to the Absorb bioresorbable vascular scaffold (BVS) or Xience everolimus-eluting stent (EES) show a markedly lower overall rate of stent/scaffold thrombosis at 30 days and 1 year compared with the ABSORB III experience: 0.3% in the new trial versus 0.9%, and 1.1% versus 0.5%, respectively. And improved operator technique is getting credit for the reduction.

"We don't know if it's Xience or BVS," presenter Gregg W. Stone, MD, of Columbia University Medical Center, told the audience at the Cardiovascular Research Technologies (CRT) meeting here, though he did conclude that "data is emerging that optimizing technique when implanting the 1st generation Absorb BVS can improve mid-term outcomes."

"In particular, avoiding [BVS] implantation in very small vessels (quantitative coronary angiography reference vessel diameter <2.25 mm), and routinely performing high pressure post-dilatation with a non-compliant balloon may reduce the rates of target lesion failure and scaffold thrombosis," Stone emphasized.

"No doubt you need good technique no matter which stents or scaffolds you use," agreed session moderator Ron Waksman, MD, of MedStar Washington Hospital Center. "I think the light at the end of this tunnel is the importance of technique and I think we got a glimpse of that in ABSORB IV with the overall lower rates of thrombosis. It is very comforting. No matter how much you split it, it looks like the right trend of lowering the rate of stent thrombosis."

But Signals Mixed

Importantly, the peek at ABSORB IV was an add-on revealed after Stone presented a separate analysis, this time a pooling of previous Absorb studies that showed no clinical benefit to that ideal BVS implantation technique: the PSP (prepare the lesion aggressively, size the scaffold correctly, and post-dilate all cases at high pressure).

Across the Absorb series of studies (n=2,973), only 12.4% of patients got optimal post-dilatation, Stone said, with scaffold thrombosis increased from 0.8% at 30 days to 2.7% at 3 years.

On multivariable analysis, vessel sizing was important to reducing scaffold thrombosis in the first year. The opposite was true for optimal pre-dilatation, though that technique became protective after the first year. Over 3-year follow-up, the effects cancelled each other out and pre-dilatation had a neutral impact on scaffold thrombosis.

Similarly, optimal pre-dilatation trended toward a 3-year overall reduction in stent thrombosis but without reaching statistical significance.

Vessel sizing was the only aspect of PSP that maintained a tie to fewer target lesion failures over follow-up (10.6% versus 11.6% without vessel sizing, P=0.03).

"Current follow-up is 1 year for ABSORB III, 2 years for ABSORB Japan and ABSORB China, and 3 years for ABSORB II and ABSORB Extend. Thus while the analysis is final for the 1-year results, the 1-3 year and 0-3 year analyses will not be complete until later this year," Stone noted.

He called out the potential for unmeasured confounders, since the reasons for technique choices were not collected.

Real-World Disconnect?

In a separate study, investigators found that Absorb was not disadvantaged for device thrombosis in the BVS Expand Registry -- but it was associated with more MIs in the long run.

Two-year data showed a 6.1% rate of MI with the BVS (versus 2.1% for EES, P=0.007), according to Robert Jan van Geuns, MD, PhD, of Erasmus Medical Center in Rotterdam, in another presentation here.

Yet at 2 years, the composite endpoint of all death, all MI, and all revascularization was about equally likely for the BVS group as their EES counterparts (15.8% versus 13.7%, log-rank P=0.408). There was no difference in definite device thrombosis, either (1.3% versus 0.9%, log-rank P=0.574).

Some baseline characteristics remained unbalanced despite propensity matching (more current smokers in the BVS group and more prior percutaneous coronary interventions in the EES arm), the presenter acknowledged, adding that the registry was underpowered to detect events like device thrombosis and death.

Like the rest of the Absorb trials, post-dilatation was rarely implemented, van Geuns said. Nonetheless, there were no cases of very late scaffold thrombosis.