At first, I was very perplexed by last week's FDA decision on Sparlon, which, according to what I'm reading on the web (reported largely from the perspective of Cephalon and its CEO), was related to FDA's concerns surrounding a single case of Steven Johnson Sydrome [SJS] purportedly associated with Sparlon during clinical studies of the drug in children with ADHD (see FDA's March Advisory Committee briefing doc for details of this case).
The active ingredient in Sparlon, modafinil, is widely sold as Provigil for use in adults with sleep disorders. The proposed usage of modafinil for ADHD would be primarily in children. As described in a prior news story, at least four cases of SJS in adults taking modafinil had been reported to FDA's AERS database at the time an advisory committee recommended non-approval for ADHD in April of this year, whereas no cases had been reported in children, although an estimated 36,000 children had been prescribed the drug (27,000 patient-years exposure as of August 2005, according to FDA). SJS is serious enough, and often enough associated with use of a newly started drug (~50% of cases) that it's reasonable to expect that had there been several cases of SJS in children taking Provigil, at least one would have been reported to FDA or another drug agency. That none were reported does not mean that modafinil does not increase the risk of SJS in children. Indeed, as the incidence of SJS is reported to be 1-3/million/year, even a ca. 100-fold increase in risk of SJS due to a drug could be missed in a population the size of the reported pediatric population using Provigil off-label.
So why was I perplexed at first by the non-approval of Sparlon? Well, I hadn't yet read FDA's advisory committee transcript or its briefing document when I read the news this morning. So I was left with just my logical reasoning, which went something like this: The drug is sold as Provigil to adults with narcolepsy and excessive daytime sleepiness and is widely used off-label for a variety of reasons, but mostly to keep people awake when they're sleepy. If modafinil is believed to cause SJS in kids there should also be a concern about its use in adults with sleep disorders (since the mechanism of SJS wouldn't be expected to differ dramatically between children and adults nor between ADHD and sleep disorders). So, the issue of approvability is likely not really just an issue of an overwhelming concern over patient safety (i.e. risk) based on a single observed adverse event--as reported today--but rather an issue of relative safety given the target effects of the drug (i.e. its benefit to risk balance) in children versus adults. To my reckoning, then, FDA must have determined that the benefit to risk equation was imbalanced in kids towards the risk side but not so in adults. If I make the assumption that FDA considered the drug's risks in kids and adults to be roughly equivalent (and what evidence would FDA have to make an alternative assumption?) then I would have to conclude that the relative benefits to kids with ADHD were judged by FDA to be not as profound as the benefits to adults with narcolepsy/excessive sleepiness.
Now, forget about my logical inferences above and look again at the briefing document and at the transcript of the advisory committee meeting in March. Unfortunately for Cephalon, what actually appears to have happened with this drug today is that FDA somehow concluded in concordance with the committee's majority opinion that the risk of modafinil is either too great or too uncertain to approve it for ADHD. However, FDA also appears to have concluded that the risks are reasonably well known and acceptable for modafinil's use by adults for sleep disorders. I was confused no more. Cephalon told it like it really is today. FDA appears to have reversed its course after climbing out on an unstable branch of its decision tree and falling into a pit of rotting logic. I feel Cephalon's pain. Every company struggling to get drugs approved in the U.S. deserves to deal with a regulator that fosters a transparent, predictable and rational approval process. It sure seems like Cephalon didn't get that opportunity with Sparlon.