DRUG ABUSE LIABILITY SHOULD BE ASSESSED BY END OF PHASE II -- FDA's HARTER; ABUSE TREATMENT DEVELOPMENT DRAFT GUIDELINES MAY BE READY BY SEPTEMBER

Drug abuse liability should be assessed by the end of Phase II trials for all new drug applications, FDA Pilot Drug Evaluation Staff Director John Harter, MD, said at the Feb. 5 meeting of FDA's Drug Abuse Advisory Committee. "Before proceeding to large scale trials, the company should have made some sort of decision about whether or not [drug abuse] testing is necessary or not, and it could be part of the end of Phase II deliberations with FDA," Harter suggested. The committee met to consider the "Guidelines for Abuse Liability Assessment." Committee members concurred with Harter that the guidelines should require an assessment of abuse liability by the end of Phase II. The draft guideline carries the provision that "all new drug applications should include a specific section addressing abuse liability issues." "If you wait until the NDA to look at what the applicant thinks about their drug's abuse liability, you are going to be faced with a much narrower database," Harter contended. "You need a checkpoint, when they decide nothing more needs to be done, to see whether you agree with it," Harter remarked. FDA and the company would discuss at that point what kind of additional clinical trials would be necessary if abuse potential was detected, Harter said. The draft guideline states that manufacturers "should provide an overview of relevant issues including pharmacological properties of the substance and a discussion of the characteristics and data specifically relevant to the abuse liability assessment" in an abuse liability section of the NDA. "They should not simply state their conclusion, but should comment on what they thought of in reaching that conclusion," committee member George Bigelow, PhD, Johns Hopkins' Key Medical Center, suggested. Bigelow also recommended that the guidelines carry a "cautionary note against data dredging, [i.e.] taking all the variables and putting them in and selectively presenting only those that show what you want." FDA Pilot Drug Evaluation Staff Medical Officer Curtis Wright, MD, added that "the statistical section of the IND [abuse liability] studies should specifically address in greater detail than usual . . . a good description of how dropouts will be handled, how multiplicity will be handled, what kind of considerations will be given to power, and whether there is a positive and a no-effect comparator." The committee also discussed whether or not good laboratory practices [GLPs] should be applied to abuse liability studies. GLPs are currently required for toxicology studies, but not for pharmacology studies. "The requirement of the toxicology GLP standards may be excessive for this," remarked Bigelow. Harter agreed that GLPs are too "stringent" for these kinds of studies; however, he insisted that the committee devise a set of high standards for the abuse liability studies that would be more rigorous than those for pharmacology studies, but less stringent than GLPs for toxicological studies. FDA's Wright explained that "what will happen when we start asking for [abuse liability studies] in the course of working up a drug, [is that] the same contract labs, which had been [involved] in the early toxicology testing will want to enter the field." Those contract labs will "breathe a sigh of relief, because they can make a real low-ball bid [if] they don't have to do it under GLPs," Wright asserted. "The regulatory issue is how shall we assure the quality of these kinds of studies when they are done outside," he said. The committee briefly discussed new "Guidelines for the Development of Medications for Treating Drug Dependence and Abuse." Committee member and chief coordinator for the guidelines George Woody, MD, Veterans Administration Medical Center in Philadelphia, reported that a draft guideline would be available by the beginning of September. The guidelines will address three key issues, Woody said: the need for rapid and efficient drug development; complications such as other drug interactions, co-existing need problems (AIDS and hepatitis B), and pregnant women and babies; and efficacy standards, which have been difficult to define. Woody reported that there will be four target situations for drug development: detoxification; substitution for another drug, such as methadone; relapse prevention; and reversal of toxic drug effects. The different drug classes to be addressed will be opiates, stimulants, sedatives/alcohol/anti-anxiety drugs, hallucinogens, solvents (glue), and nicotine.