A Medical Device Daily

The circulatory systems advisory committee has been the busiest of all FDA device panels this year, and it is convening again, today and tomorrow, to examine the premarket applications (PMAs) for two cardiac devices that are no strangers to FDA reviewers.

But as the stent meeting at the end of last year demonstrated, FDA reviewers are nothing if not cautious about putting a device to market that hints at any problems with safety or efficacy.

Today's session will examine the PMA for the Xience drug-eluting stent (DES), made by Abbott Cardiovascular (Abbott Park, Illinois). The panel on Friday will review the PMA for the Heartmate II, a left-ventricular assist device made by Thoratec (Pleasanton, California).

According to FDA documents, the preliminary indication for the Xience DES would be for coronary artery lesions up to 4.25 mm in diameter, larger than any approved DES.

The PMA for the Xience, which Abbott picked up in the carve-up of Guidant in 2006, is based on three studies. The Spirit First was a first-in-man study that randomized 60 patients to a bare-metal control group, and SPIRIT II was a study of 300 patients that randomized patients to the Taxus at a variety of sites in Europe and Asia.

According to Abbott's web site, the Xience "demonstrated superiority in the primary endpoint of in-stent late loss at 6 months, and clinical superiority on MACE with a 71 percent reduction compared to TAXUS at one year."

The third supporting study, the Spirit III, enrolled roughly 1,000 at 80 sites. Spirit III was said to have demonstrated that Xience outperformed Taxus for late vessel lumen loss, the primary endpoint, at eight months and equaled Taxus in target vessel failure at nine months. Exhaustive 24-month data from both the Spirit II and III trials was not available in time for the meeting, but Abbott had performed an ad hoc analysis "on a subset of combined...subjects with a two-year follow-up" as of Oct. 30.

Second-year data from the Spirit II and III showed that two of 379 patients in the study arm were lost to cardiac death, but none of the 153 patients in the Taxus arm passed away. No Q-wave infarction was seen in either arm from this preliminary analysis, but four of the study-arm subjects experienced non-Q wave infarction versus one of the 153 in the control arm.

Two patients with Xience experienced very late stent thrombosis (VLST) as defined by the protocol, but a definition of VLST provided by the Academic Research Consortium showed only one incident of VLST. The FDA document did not compare the two definitions.

FDA said that the ad hoc analysis "should not be interpreted as an interim analysis" and that an exhaustive data set might depict different outcomes. Also potentially problematic was the fact that different adjudicators were used to determine the classification of adverse events.

FDA will ask the panel whether the product label adequately prescribes dual anti-platelet therapy (DAPT), an issue for all such products. As currently written, the label recommends aspirin and clopidogrel (Plavix) for six months and continuation of aspirin therapy for another 54 months. The label also warns the patient that premature termination of anti-platelet medication "could result in higher stent thrombosis, myocardial infarction or death risk." FDA's question is whether "the language...adequately conveys a recommended course" of DAPT.

According to the agency's documents, the Spirit II trial showed that Xience was non-inferior to Taxus at 180 days for in-stent loss of lumen (blood vessel diameter) and at 240 days for loss of lumen for the vessel. Perhaps primarily because the data sets are not exhaustive, FDA intends to ask the panel whether the data "provide a reasonable assurance of effectiveness."

Thoratec has high hopes for its second-generation HeartMate device to drive its continued dominance in the VAD sector. One reason for sending this device to panel was that seven patients failed to meet an inclusion criterion of a BSA greater than 1.5 square meters.

The underlying rationale was that a set of performance goals for devices of this class was developed for patients with BSAs greater than 1.5 square meters.

According to FDA's draft questions for the panel, 72 of the enrolled 126 patients who received the Heartmate II obtained a transplanted heart within 180 days while another four recovered and left the hospital with no further intervention. Another 13 were still awaiting transplants or other destination therapy/devices at 180 days.

FDA will ask whether the Heartmate II data provide a reasonable assurance of safety and efficacy "even though the data did not meet the performance goal" of helping between 65 and 70% of patients to transplant within 180 days. The numbers for the trial crunched out to about 57% getting a transplant by day 180.

FDA's interest in this device probably reflects the dire condition of most such patients. The 13 patients in the study arm awaiting transplant were listed as transplant status 1A or 1B. 1A status includes a probability of death within a week and the use of a ventricular assist device. Status 1B is the use of a VAD for more than 30 days, but not likely to be a fatality within a week.