WASHINGTON – The FDA broke an embargoed release of data for Abbott Laboratories' Absorb GT1 scaffold, advising clinicians that the rates of major adverse cardiac events in a large study were higher than seen with the comparator, but leading cardiologists at ACC 2017 were nonetheless unwilling to recommend that clinicians abandon the device.

The FDA's safety alert preceded the embargo placed on the news from the Absorb III study – sponsored by the Abbott Park, Ill.-based company – by nearly 50 minutes, but contained no information that was not included in the presentation by Stephen Ellis, one of the investigators in Absorb III. Ellis reminded that the primary non-inferiority endpoint for Absorb III in regulatory terms was based on one-year data, and that the company had hoped to keep data from years two through five under wraps until those data could be blended with Absorb IV data for longer-term assessment of target-lesion failure (TLF).

However, Ellis said the superiority of the GT1 over the Xience drug-eluting stent for TLF would not likely emerge until after the GT1 scaffold had absorbed, which is typically expected to occur at three years post-implant. Consequently, the company and FDA have agreed, after consulting with clinical investigators, to use years three through seven to test the TLF superiority endpoint, thus allowing the unblinding of data for ACC 2017.

Abbott has two-year data from Absorb III for roughly 98 percent of the nearly 2,000 patients in both arms of the study, and Ellis pointed out that the study article had met the non-inferiority endpoint for TLF at one year. As has been previously described, clinicians have been cautioned about implanting the device in patients with reference vessel diameters smaller than 2.5 millimeters, a predicament that colored the latest release of data.

The Absorb had a difficult time keeping up with the Xience for target-lesion failure for the 13-to-25 month interval (the actual period of elapsed time used for the nominal one- and two-year data collection set points), which occurred in 3.7 percent of all enrollees on the study article regardless of lesion size, while only 2.5 percent of Xience patients experienced TLF. After dividing the data by RVD as assessed by quantitative coronary angiography with vessel diameter of 2.25 milimeters as the threshold, the GT1's rate of TLF dropped to 3.2 percent in the larger set of vessels, but Xience also benefitted from this analysis, with only 1.9 percent experiencing TLF. More than eight in 10 of enrollees in both arms had RVDs in excess of 2.25 mm, according to Ellis.

At 25 months, the difference in TLF for the two arms for the entire study population had widened in absolute terms (11 percent for GT1 to 7.9 percent for Xience), as was the case for the subset with larger vessels (9.4 percent to 7.0 percent). However, the relative differences in TLF had actually narrowed.

Ellis said the generalizability of Absorb III is limited by inclusion of those with stable ischemic disease, and that the study is underpowered to evaluate low-frequency events. "Importantly, implantation technique was still evolving" as the study commenced with enrollment, he said, adding that the combination of predilatation of the vessel, appropriate vessel size selection, and postdilatation – an approach known colloquially as PSP – might go a long way toward resolving some of the differences between the two devices.

Antonio Colombo of New York Presbyterian said of the new Absorb III data, "I think you can look at the glass [as] half empty or half full." Colombo stated further, "with the proper technique and selection, you can make it non-inferior" to Xience, suggesting that if clinicians stay in bounds on vessel size, frequently invoke postdilatation with a noncompliant balloon inflated to at least 18 atmospheres, and routinely use intravascular ultrasound (IVUS) to guide the implant procedure, the device's performance should shape up well in contrast to the Xience.

Roxana Mehran of Mt. Sinai Hospital in New York said she was "somewhat comforted that the event rates are not extremely high" for the GT1, but that the data leave her "waiting for the promise of the future; superiority of this device" over Xience for TLF.

Mehran said the question of whether a patient with the GT1 should stay on dual antiplatelet therapy until the scaffold disappears is still unanswered. She said most of the patients in the study were not treated via the PSP protocol, and that it is not entirely clear how IVUS influences outcomes.

Ellis responded, "I think, based on these results, as long as you follow the inclusion criteria in Absorb III" and deploy the device in a technically appropriate fashion, "the results are similar to what you see in Xience." Nonetheless, he said, clinicians "should be up front with the patient" about the uncertainty.

Regarding the broken embargo, the FDA told Medical Device Daily that it has "interacted with Abbott to ensure it has the most up-to-date information on which to base its recommendations. As new information becomes available, the FDA will update its communications and recommendations as necessary."