What You May Have Missed: Critics Point to Problems With Celgene’s Crohn’s Drug Trial Design

Celgene Cuddles Up With Another Cambridge Biotech, This Time to Tackle Autoimmune Diseases

October 19, 2016
By Mark Terry, BioSpace.com Breaking News Staff

On October 16, Celgene International Sarl, a wholly owned subsidiary of Celgene Corporation , released data from a Phase Ib clinical trial that evaluated its GED-0301 (mongersen) in patients with active Crohn’s disease. Although the overall results appeared positive, after a closer look, investors and analysts noted that the company did not use a placebo control in the study.

As Celgene stated, “The study was designed to further enhance the understanding of GED-0301 activity in a difficult-to-treat, moderate to severe patient population. This population was more diverse than prior GED-0301 studies and included patients with endoscopically confirmed mucosal damage at entry and those who had previous surgeries. The study also included both biologic-exposed and biologic-naive patients, as well as patients with a diagnosis of Ileitis, Ileocolitis or colitis.”

Everything, it seems, but a placebo or normal control group.

Clinical improvement was observed by the second week. Patients who were eligible for endoscopies at week 12 had a 37 percent endoscopic response, “with no meaningful difference across treatment groups.” And in patients with greater endoscopic disease activity at the baseline, 63 percent had a decrease of 25 percent or more in SES-CD score, with 31 percent having a reduction of 50 percent or more.

“We are encouraged that oral GED-0301 showed both meaningful endoscopic improvement and clinical remission at an early time point in this study,” said Scott Smith, president of Celgene Inflammation & Immunology, in a statement on October 16. “The fact that this study included nearly 50 percent biologic-experienced patients further reflects the potential of GED-0301 as a novel approach for patients with Crohn’s disease searching for alternatives.”

Adam Feuerstein, however, writing for The Street, said, “Celgene wasted everyone’s time by running a Crohn’s disease study of GED-0301 without including a placebo arm. Absent a placebo arm for comparison, the GED-0301 Crohn’s data presented Tuesday (announced via press release on Monday) are largely uninterpretable.”

And Feuerstein isn’t the only critic. Josh Schimmer, an analyst with Piper Jaffray, wrote in a research note, “While response and remission rates seen with ‘301 are consistent with rates seen with active drug arms of placebo controlled trials, the rates with placebo in trials can be substantial … When [approximately] 60 patients are divided across three dose arms and subsequent subset analyses are performed, the error bars become extremely wide. We were concerned by how confident Celgene was in interpreting the small dataset overall but then fall back on the ‘small patient numbers’ argument to explain away any inconsistencies in the data.”

And by comparison, in a Phase II trial of filgotonib in Crohn’s disease run by Gilead Sciences and Galapagos , at the 50 percent cutoff point, the placebo group had a 14 percent endoscopic response rate. The endoscopic response rate for their compound, filgotonib, was 25 percent.

Celgene is arguing that GED-0301 works against Crohn’s disease without a placebo to compare it to, and warn that comparing their trial to the Gilead-Galapagos trial could be misleading.

Feuerstein writes, “Okay, that might certainly be true, but it’s also hard not to look at the 14 percent placebo response rate in the filgotinib study and wonder if GED-0301 (15 percent response rate) is also a placebo. It would certainly be more accurate and reassuring to have placebo data from the Celgene study, but it’s missing.”

Celgene won’t have more data available until sometime in 2018 when a Phase III study on GED-0301 in Crohn’s reads out. Sometime next year there will be some results from a Phase II trial of the drug in ulcerative colitis, but it won’t have a placebo either.

Celgene is currently trading for $100.06, down significantly from its November 2, 2015 price of $127.20 a year ago.

MORE ON THIS TOPIC