Medical Device Daily Washington Editor

GAITHERSBURG, Maryland — The weight of a practicing physician’s opinion tilts a lot of scales in U.S. healthcare, and this appeared to be the case at Wednesday’s advisory panel hearing for the premarket application for Repel CV, an anti-adhesion barrier for neonatal cardiac surgery.

The circulatory systems advisory panel expressed the usual range of concerns regarding product safety and efficacy, and some panelists were uncertain just what the clinical benefit of reduced severity of adhesions actually is. Still, the panel voted 8-3 in favor of approval, perhaps principally because the surgeons on the panel asserted that reducing adhesions is a benefit they said is clinically meaningful.

Repel CV, made by SyntheMed (Iselin, New Jersey), is a bioresorbable anti-adhesive film for use in patients undergoing cardiac surgery via sternotomy and likely to need a second procedure.

The patients in the pivotal trial, average 12 days old, underwent staged restructuring of the heart, due to congenital defects, but the firm hopes to get an indication without age restrictions. At present, there is no FDA-approved anti-adhesive product for this use, though the company obtained a CE mark in June 2006.

The FDA summary acknowledged that “finding a feasible population for study of this device in a reasonable time span is a major challenge,” in part because of the requirement for the patient needing a second surgery, but the primary driver behind the patient selection for the study.

Yunling Xu, an FDA statistical analyst, said that the pivotal study showed a 26% absolute reduction in adhesions, deemed statistically significant. But Xu said that because the device showed a large tendency toward allowing either many or very few adhesions, the validity of some of the statistical analyses was questionable.

“Several potential issues included center effect,” Xu said, a number of centers (of 13 total) demonstrating a 50% drop in adhesions, and two centers showing increases.

Also, about a quarter of the patients were evaluated by a surgeon who became aware of which arm of the trial the patient had been enrolled in. In some cases, this happened because the surgeon doing the first procedure was the only one available to do the second.

Wolf Sapirstein, MD, a medical officer at the Office of Device Evaluation, said the agency “acknowledges the reasons for enrollment of only pediatric patients,” because, in part, an “assurance of re-operation within a defined timeframe.” Going forward, the agency and SyntheMed assumed, he said, “that the study outcome could be extrapolated to other populations” because the “pathogenesis [of adhesions] is similar for all ages.”

Two violations of the protocol occurred, including one in which the physician decided not to implant Repel CV, the other occurring when the surgeon chose a Gore-Tex membrane rather than Repel CV. In both cases, the reasons were purged from the published document.

However, Sapirstein said, “the balanced protocol deviation did not alter the [trial data] analysis.”

Sapirstein said that adverse event rates were similar, 69.9% for study patients vs. 71% of controls. Six of the study subjects’ events were felt to be treatment-related, whereas only one event in the control group was felt to be treatment-related.

Synthemed made the case that the rate of adverse events was within parameters described in the literature. And Sapirstein agreed, saying that mortality rates “are consistent with the literature,” and that adverse events and serious adverse events were similar between the two study arms.

According to the FDA data, 38 patients in the study arm had no adhesions, and 15 in the control arm had no adhesions. However, the incidence of mediastinitis — inflammation of the mediastinum, the area bound laterally by the lungs and by the sternum and the spine in the front and back — was slightly higher in the study group.

Sapirstein said the only safety issue is “whether the mediastinitis was associated with the foreign body introduced into the operative field.”

In a review conducted on behalf of the panel, panelist Eugene Blackstone, MD, head of clinical research at the Cleveland Clinic, said, “clear from the data is that Repel is not a panacea” and that surgical times were consistently longer in study patients, regardless of whether adhesions occurred. And he asked why this was the case.

A company spokesman said that surgeons who had performed the procedure more than three times needed five fewer minutes to do it.

Blackstone’s concluded: “There is clear incremental benefit in the product,” although “the product does not perfectly protect.”

Panelist Richard Hopkins, MD, chief of the division of cardiothoracic surgery at Brown University School of Medicine (Providence, Rhode Island), said the study “doesn’t necessarily give us the perfect device of this type.” He said the study data was good, though it suggested foreign-body stimulus of the immune system.

“In fairness to the sponsor, this is a very difficult study to get your hands on,” he said, but that Repel CV “clearly does not reduce the problem to zero.”

Hopkins also posed the question of whether this device was appropriate for adult patients with heart disease inasmuch as the rate of re-operation for these populations is falling.

“I would probably use this device if I were convinced of its safety in patients whom I was convinced were going to come back in a relatively short period of time,” he said, but he was not keen on its use in patients with diabetes due to an increased risk of mediastinitis. Although neonates are immune-compromised, the rates of mediastinitis, he said, “are very, very low,” making difficulty any extrapolation to adult use.

“We’re talking about a device that clearly reduces the severity,” but not necessarily the incidence of adhesions between the heart and surrounding tissues, Blackstone said.

After considerable discussion, the panel voted for approval with four conditions. Among these were restricting its use to patients identified by FDA as pediatric — age 21 and younger — and for patients “at high likelihood of re-operation.”

The proposed label would have contraindicated for patients with a ventricular assist device. The panel recommended that this be dropped from the label, and that the label should say that the device is “indicated for reducing the severity, but not the incidence, of adhesions.”