June 14, 2006 | Is there one unified view of the adaptive trial paradigm at the FDA? No, says Robert Powell, the FDA's senior pharmacist in the Office of Clinical Pharmacology and Biopharmaceutics at CDER. "There are many FDAs," says Powell. "Making some sort of blanket statement about the drug area would be over-reaching. What folks [at the FDA] are looking for are situations where they can get experience. People who haven't done that [adaptive] work are thinking about how am I going to deal with this sort of a problem, particularly if this becomes a pivotal trial." No additional regulations or guidances on the topic are planned, but some scientists at FDA will be publishing ideas on the topic in the medical literature.

By themselves, adaptive designs are nothing new. But the joining of distinct Phase II and III studies -- a "seamless" strategy -- is novel. In seamless trials, he says, carryover effects may loom as an area of regulatory concern. With drugs that need time to take effect in the body, he says, an adaptive trial could mean that patients still have an old dose in their bodies as a new one is introduced. That could complicate the analysis of the drug's safety and efficacy. "It all depends on how the rules are set up for making the adaptation," he says. "When do you move from dose X to dose Y to dose Z?"

The agency is also worried about operational bias in adaptive designs. Briefly, the concern is that a sponsor's midcourse tweaks to a crucial study might not be statistically kosher or correct. "That's definitely a hot button," Powell says. "That goes to the heart of protecting the blind, protecting against bias, which is probably one of the biggest concerns."

'Hot Button'
Powell says FDA scientists understand the potential benefits of adaptive trials, but are still trying to determine which trials are best suited for the approach. Take a trial testing a medical device for spinal injuries. Such patients might be exceedingly rare, and an adaptive design might be especially potent. Says Powell:  "It would be hard to imagine getting thousands of people together to go through dose-ranging and two pivotal Phase III trials."

How should the industry avoid problems? Come in as early as possible. "If a company wants to go down this path," Powell says, " they are going to have to begin planning these trials earlier and having discussions earlier. If someone waited until their end of Phase II meeting to discuss their adaptive trial design, they'd be taking a risk and getting a lot of questions that might slow them down."
 
Powell notes that a private conference between the FDA and the Pharmaceutical Research and Manufacturers Association (PhRMA) in November 2006 could be a watershed event. Speakers from both industry and the FDA will privately review the recent history of adaptive designs, and the proper taxonomy and scope of adaptive trials, and explore their potential and operational needs. A summary of that conference by the FDA will be presented in public at a DIA interest group meeting in early 2007. --M.D.U.

Back to main article.