Event – Celgene’s chance to prove that Crohn’s bet was worth it
When in April Celgene splashed out $710m on Nogra Pharma’s lead project, GED-0301, a few eyebrows were raised. Next week the group could prove that it was worth it, as all-important phase II results of the Crohn’s disease agent are to be presented at a scientific meeting.
In its favour are stellar phase I data, a unique mechanism and oral dosing. The case for blockbuster sales will be tempered by the fact that the early results cannot possibly be reproduced in a larger, controlled study; but a smaller benefit can still be impressive, as long as it is not confounded by a high placebo response.
Five approved biological agents – Remicade, Humira, Cimzia, Tysabri and Entyvio – represent the cutting edge of Crohn’s disease treatment, and efficacy of GED-0301, an anti-Smad7 antisense DNA oligonucleotide, will have to be at least on a par. GED-0301 already has a key dosing advantage.
The only human data generated for it so far are impressive, showing a 100% response and 86% remission rate, with an average reduction in the Crohn’s disease activity index (CDAI) from 287 to 89 by day eight.
However, there are reasons for scepticism, hence the assumption that Celgene saw something else before handing over $710m up front (Celgene’s record-breaking deal hangs on phase II data, April 25, 2014). The phase I study was of short duration, in just 15 patients, in a single Italian hospital without a placebo control.
Because all patients were hospitalised some of the benefit can be attributed to a placebo effect, so in phase II investors must assume an effect that is lower albeit still impressive. The phase II data will be presented on October 21 at the United European Gastroenterology Week meeting in Vienna, with the abstract being unveiled the day before.
The study, at centres in Germany and Italy, randomised 166 patients to one of three 10-160mg doses or placebo on a double-blind basis. Its primary measure is efficacy after two weeks’ treatment, aiming for GED-0301 to show an advantage in terms of patients in remission, ie, those with a CDAI improvement of over 150 points.
Apart from Cimzia the biological agents are used in both induction and maintenance of Crohn's treatment, but in the former setting – which at this point is where GED-0301 is being tested – their response rates have varied from around 30% to 80%. Of the responding patients, 14-48% were in remission at one month.
A key point, however, is that these comparative studies have shown up to 40% response in placebo recipients. Still, the magnitude of GED-0301’s phase I benefit is far greater than any historical placebo data, greatly reducing the risk of a false positive, says ISI Group’s Mark Schoenebaum.
He expects a remission rate of around 10-20% in the phase II placebo cohort, and 40-60% in active groups. Clearly this will fall further in the even stricter phase III setting, though it could still be competitive against biologicals.
On top of this, of course, GED-0301 has the advantage of oral dosing, and should prove safer since it does not seem to be absorbed into the bloodstream. If phase II shows a dose-response this would further back the argument that GED-0301 is having a real effect.
EvaluatePharma consensus data only go up to 2020, at which point GED-0301 sales of $53m are expected, but the real potential lies beyond. Mr Schoenebaum, for instance, expects total peak sales of over $3bn, to which at present he applies a 50% risk adjustment.
Any longer-term signals will also be seized on given their potential to show a better picture of safety, and perhaps hint at GED-0301’s potential in the maintenance setting. On the debit side the phase II study could be criticised because it did not include colonoscopy – a relatively standard test – and did not measure biomarkers.
That said, these are all more advanced considerations that will no doubt be explored in phase III. For now Celgene must show that a trial that it did not design justifies its $710m bet.
|Phase II||166 Crohn's pts, 10mg, 40mg, 160mg or placebo, double-blind||2011-002640-27|