Medical Device Daily Washington Editor
GAITHERSBURG, Maryland – The FDA Obstetrics and Gynecology Devices Panel's review of study data for the premarket approval (PMA) application for Adept, an adhesion barrier product marketed by Innovata (Nottingham, UK), resulted in a stamp of approval. But several potential tripwires surfaced during the day-long hearing Monday.
In the end, Innovata won the advisory panel's recommendation to approve, but not without some discussion of modification of the language used in the indication, and not without a short-lived threat of substantial postmarket liability.
One of the advisory committee's discussions centered on the possible recommendation that the FDA require the firm to undertake a postmarket study of any relationship between the adhesion-barrier device and pregnancy.
However, Nancy Brogdon, director of the agency's Division of Reproductive, Abdominal and Radiological Devices, informed the committee that this suggested a different indication, and would hence constitute “a different PMA.“
Innovata seemed to almost invite the requirement by including a pair of secondary endpoints that addressed fertility scores in the study, but the language of the product's indications and limitations in the so-called AFS scoring system drained the idea of any substantial momentum.
Adept is a different application of a product that Innovata has marketed as a peritoneal dialysis solution under the brand name Extraneal. Extraneal, a 7.5% solution of icodextrin (sodium chloride, sodium lactate, calcium chloride, magnesium chloride), was cleared for marketing in the UK in 1992, but did not make it past the FDA until 10 years later.
Adept, which is a 4% solution of icodextrin and is marketed worldwide by Baxter Healthcare (Deerfield, Illinois), began appearing in European operating rooms in June 2000.
According to Michael Kuchinski, microbiologist and lead reviewer at the Office of Device Evaluation, Adept was handled by the Center for Devices and Radiological Health (CDRH) because of its principal mode of action, that of a barrier product. However, Extraneal was reviewed and is regulated as a drug.
One of the most important characteristics of icodextrin is that it is metabolized primarily by amylase, which is not present in the peritoneal cavity in meaningful quantities. This allows the substance to persist for up to four days, long enough to more than cover the first three days after surgery, which are the most crucial for adhesion formation.
According to Colin Pollard, chief of the Obstetrics and Gynecology Devices Branch at CDRH, the FDA has cleared three other anti-adhesives for surgical use, but none for laparoscopic gynecological surgery.
The company's proposed indication for Adept is “as an adjunct to good surgical technique for the reduction of post-surgical adhesions in patients undergoing gynecological laparoscopic surgery which may include adhesiolysis.“ The proposed use of Adept is as “both an intra-operative irrigant and post-operative instillate.“ The company recommends that surgeons use one liter of Adept for both irrigant and instillate purposes.
Innovata's trial for Adept was a randomized, double-blind, open-label study at 16 centers. Total patients enrolled in the study for both Adept and the control article, which was lactated Ringer's solution (LRS), was 449. The analysis of outcomes required a second laparoscopic examination of study subjects to evaluate the efficacy of the study article vs. placebo.
One of the unintended consequences of the study is that the LRS did nearly as well as Adept on a number of measures. On the first primary endpoint, a reduction by a total of three adhesions – or by 30% of adhesions when 11 or more are observed – 45% of the Adept patients' outcomes qualified and just over 35% of the controls reached that threshold. The unexpected show of efficacy of the Ringer's solution, which consists of sodium chloride, sodium lactate, calcium chloride and potassium chloride, led some on the panel to speculate as to whether this product might also find use in such procedures.
A source of some confusion for the mathematically challenged in the audience, and of consternation for the sponsor, was the agency's requirement that the clinical trial data be subjected to a somewhat stringent standard of performance as indicated by an altered use of the confidence interval.
Innovata's documents indicated that the lower end of the confidence interval for the first primary endpoint was 0.7%. However, given that the lower bound of the confidence interval was required to come in at or above 5% the confidence interval for the placebo data, the data was said to have failed to support the first endpoint.
Essentially the same fate befell the third endpoint, that of “percentage of patients with fewer sites with dense adhesions.“ The study protocol required that the difference between the lower end of the confidence interval here simply be greater than that for the LRS. The lower end of the confidence interval here was -8.13%, but the lower end for the CI for LRS was 10.37%.
The second primary endpoint managed to clear the statistical hurdle because the upper end of the test article's confidence interval (for number of adhesions) was less than zero. This number rang in at -1.62.
After his discussion on behalf of Innovata, biostatistician Dr. Steven Piantadosi of Johns Hopkins University (Baltimore) said that such a stringent threshold for confidence intervals increased the probability of making a Type I statistical error. This type of error occurs when one either incorrectly rejects a true null hypothesis – which would be to reject the notion that the Ringer's solution was as effective as the Adept when it was as effective – or to fail to reject a false null hypothesis, meaning that one takes the position that the Ringer's was as effective as Adept when such was not the case. He stated at one point that “we don't do that for any other studies.“
In a phone interview, Pollard told Medical Device Daily it is not “that unusual that we will assign a higher mark for the confidence interval to ensure that the effect that we are seeing was a robust effect.“ He declined to address whether this was applied to the Adept study because it was a pivotal study, which the panel made reference to on several occasions.
During the afternoon discussions, the advisory committee also took up the notion of whether the indications might be broadened. However, the consumer representative on the committee, Diana Romero, PhD, of Columbia University (New York), suggested that the FDA examine postmarketing data to evaluate this question.
Jonathan Weeks, MD, of Norton Suburban Hospital (Louisville, Kentucky), remarked that “given that we're struggling with efficacy as it is“ and that physicians may engage in off-label use, he was reluctant to agree to broaden the stated indication.