GI Dynamics (ASX:GID) said today that it still plans to pursue FDA approval for its EndoBarrier weight loss device, despite missing safety and efficacy endpoints in a pivotal U.S. trial of the gut liner.
Lexington, Mass.-based GI Dynamics shut down the trial last July after the FDA ordered it to cease enrolling patients due to a higher-than-expected rate of hepatic abscess. The EndoBarrier is a plastic sleeve designed to promote weight loss and relieve the symptoms of diabetes by slowing the uptake of nutrients from food.
There were 7 cases of hepatic abscess in the 325-patient trial, for an incidence rate of roughly 3.5% – nearly double the 2% safety threshold and well above the 0.73% rate for patients outside the U.S. Today the company said that because the trial only enrolled about ⅔ of the 500 patients it was designed for, the sample size provides less statistical power.
Still, GI Dynamics said, the data is slated to be submitted to the FDA next month. The company plans to seek an investigational device exemption for another trial, using a new treatment algorithm it said helped reduce the risk of hepatic abscess with the EndoBarrier device.
“Although the Endo trial did not achieve its primary safety endpoint, evaluation of the safety data has revealed an opportunity to improve the clinical algorithm for the use of EndoBarrier therapy. This optimized treatment algorithm is specifically designed to reduce the incidence of hepatic abscess and improve overall safety without compromising device efficacy, resulting in an improved benefit:risk profile,” the company said.
The Endo study showed an 11.7% rate of device-related serious adverse events requiring early device removal, GI Dynamics said. Although the primary safety endpoint was a statistical test to demonstrate the incidence of primary safety events with EndoBarrier Therapy was less than 15%, the Endo trial still missed the safety mark of 96.5% probability, coming in at 90.5% due to the high rate of liver infections.
The primary efficacy endpoint was another statistical test, designed to show that the EndoBarrier demonstrates a more than 96.5% probability that the change in HbA1c was at least 0.4% greater in the EndoBarrier arm than sham control arm. The Endo study showed a 92.8% probability, GI Dynamics said.
“These results are nonetheless encouraging, given that (1) the trial only enrolled ⅔ of the intended subjects; and (2) because of early suspension of the trial, data for nearly ⅓ of the enrolled and randomized subjects were not available for the protocol-specified primary efficacy analysis,” the company said.