The third leg of Roche’s triumvirate of important clinical events–and by far the most important–is thus far not unfolding as well as did the first two. The Phase III program for bitopertin, Roche’s first-in-class glycine site transporter inhibitor (GlyTi), began four years ago, with six trials enrolling about 4000 patients all told, using a planned 52 week duration of treatment. Three of the trials focused on the hoped-for effect of bitopertin (as an adjunct to antipsychotics) upon the negative symptoms of schizophrenia, and the first two of these just had their results disclosed–they failed to hit their endpoints. They are cited as failing at 24 weeks, thus this must have been a planned interim analysis. Roche had unblinded those two trials in mid-December, but kept the results quiet until now for fear of hindering/skewing enrollment in the trials still recruiting patients. We wonder if the three trials out of the six still cited as ‘recruiting’ have stopped doing so–certainly the third trial addressing negative symptoms would most probably have stopped bringing in new patients. There was no mention of any subgroup analyses or trends in Roche’s terse press release, details will await formal presentation.
Now, the best that Roche can reasonably hope for is demonstration of some benefit in the three trials testing bitopertin as an adjunct in the treatment of schizophrenia patients whose positive symptoms are ‘sub-optimally controlled’ by their current antipsychotic regimen. One of those trials (‘TwiLyte’) appears to be completing, as had been expected. The other two trials, also enrolling approximately 600 patients each, are cited as likely to finish in late 2014 and early 2015, though it appears that there is a six month interim analysis stage, thus news of failure is likely to come earlier, whereas a ‘go’ signal from the interim analysis is almost certainly not going to be announced, and those trials would then continue. From that point of view, no news might be good news. This presumes that Roche continues enrollment, and that sites are able to recruit patients for a trial using a compound featuring two failures thus far. Our best guess is that the first of the trials focused on positive symptoms may report results before midyear, the other two, between mid-2014 and late-2015.
What can be expected from the positive symptom trials? The failure of the negative symptom trials thus far points to the nonpredictive nature of Roche’s initial Phase II, which had focused on negative symptoms. In that 323 pt trial, which used an eight-week duration of treatment, the effect size on negative symptoms for the 10mg dose was 0.37, it was 0.40 for the 30mg dose. Unfortunately, the longer, larger Phase IIIs did not replicate this. The Phase II effect on positive symptoms is a bit harder to tease out, because Roche did a subgroup analysis of approximately 130 patients whose baseline positive symptom subscale score (on the PANSS) reached an undisclosed minimum criterion. Roche reported a trend towards efficacy, citing an effect size of 0.37 for the 30mg dose, the 10mg dose appeared to be quite similar. Roche has not publicly disclosed how they are parsing out this potential responder population (those with elevated positive symptoms) in these three trials; since patients had to be ‘clinically stable’ for four months pre-enrollment, their positive symptoms had to hit some intermediate range of severity– salient, but not unstable. Hopefully, the trend towards efficacy seen in the 130pt subgroup in Phase II may reach significance in these larger studies, but the negative symptom trial results show that this cannot be assumed.
Thus, the bitopertin story is not necessarily over, but the larger-scale opportunity, the hope of impacting the recalcitrant negative symptoms of schizophrenia, appears to have vaporized. The market opportunity from an antipsychotic ‘enhancer’ vis-a-vis positive symptoms is not to be completely dismissed–after all, one-third of the Phase II sample met Roche’s criterion for ‘suboptimal’ response–but would be unlikely to constitute the blockbuster product that bitopertin would have been had the full array of Phase III trials been successful. Now, we will have to wait to see if a partial success can be salvaged.