I've talked about a lot of difficult therapeutic areas, but here's another boulevard of broken dreams: schizophrenia drugs. I was working on follow-ups to a promising clinical candidate, which has since been promising a number of times without ever delivering. It certainly missed its endpoints in schizophrenia by a mile in Phase II. That was actually my introduction to the drug industry back in 1989 - I followed that up with several years working on Alzheimer's, another notorious graveyard of good ideas, which makes me wonder why I didn't just quit at some point and open that chain of all-you-can-eat catfish restaurants that the Northeast so desperately needs.
Of course, once in a while a drug for dementia actually works a bit, and since there's a huge underserved market out there, it's a prize worth seeking (ask Lilly (NYSE:LLY) or J&J (NYSE:JNJ)). But clinical success rates are absolutely horrific in the whole CNS area, and the latest company to demonstrate this is Vanda Pharmaceuticals (NASDAQ:VNDA) in Maryland (I've always wondered if they're named after a spectacular, and spectacularly finicky, genus of orchid).
Vanda's drug iloperidone has been kicking around for years now. Hoechst Marion Roussel (now Aventis) seems to have discovered it in the early 1990s, and they, Novartis (NYSE:NVS), and Titan (NYSE:TTP) have all handed it off to someone else over the years. Vanda was the last in line, but they got the dreaded "Not Approvable" letter from the FDA Monday, and the company's stock was blitzed, down 73 percent at the close. And the thing is, this drug got a lot closer than anything I used to work on. Vanda did hit their endpoints against placebo and against haloperidol, but the problem is, these are not necessarily the standard of care in schizophrenia:
"The FDA stated that Vanda had demonstrated the effectiveness of iloperidone at 24 mg/day in the 3101 study for which the company reported results in December, 2006, and that the efficacy was similar to the active comparator, ziprasidone (Geodon[R], Pfizer Inc.). In addition, the FDA also stated that iloperidone was superior to placebo in patients with schizophrenia at doses of 12-16 mg/day and 20-24 mg/day in a prior study. However, the FDA expressed concern about the efficacy of iloperidone in patients with schizophrenia relative to the active comparator, risperidone (Risperdal[R], Johnson & Johnson), used in prior studies. The FDA indicated that it would require an additional trial comparing iloperidone to placebo and including an active comparator such as olanzapine (Zyprexa[R], Eli Lilly & Company) or risperidone in patients with schizophrenia to demonstrate the compound's efficacy further. The FDA also stated that it would require Vanda to obtain additional safety data for patients at a dose range of 20 to 24 mg/day."
So iloperidone works, but quite possibly not well enough compared to what's already on the market. That alone won't quite sink your drug - you can always hunt for a patient cohort that benefits from a new compound, and you'll quite likely be able to find one if you have the resources. But as that last line mentions, there are additional safety concerns.
Reading between the lines, it would appear that iloperidone had the best chance of distinguishing itself in efficacy at the higher doses, but that the FDA wanted to make sure that side effects didn't start kicking in up there. This paper makes you wonder if one problem is the (dreaded) QT interval prolongation. Many other factors have looked relatively clean in some of the reported trials.
I greatly doubt if we'll see iloperidone surface again. Vanda wouldn't seem to have the resources, and too many other organizations have passed on it. At this point, it's hard to see why more money would be put into the compound...