Abuse Liability Assessment Conference Highlights Role of Science in Regulatory Aspects of Medication Development

Scientists working in animal and human psychopharmacology serve an important role in the medications development process by providing valid methods for abuse liability assessment in both animals and humans.

Maxine L. Stitzer, PhD
Professor, Dept of Psychiatry and Behavioral Sciences
Behavioral Pharmacology Research Unit
Johns Hopkins University School of Medicine

Scientists working in animal and human psychopharmacology serve an important role in the medications development process by providing valid methods for abuse liability assessment in both animals and humans. A two-day meeting sponsored by the College on Problems of Drug Abuse (CPDD) was held in Bethesda, MD on October 28-29, 2002.   Attending were representatives from academic research, from the pharmaceutical industry and from federal agencies. The purpose of the meeting was to make recommendations on abuse liability testing methods that will enable pharmaceutical developers and NIH to meet regulatory requirements and that will foster and guide future research in this area. 

Drug development is spearheaded by the pharmaceutical industry in its search for new efficacious medications to treat neurological and psychiatric disorders.  Specific types of drugs include those for treating ADHD, anxiety, pain, obesity, cognitive disorders and drug dependence. Another important area of application includes new formulations of drugs such as those that might reduce the abuse liability of potentially highly abusable chemicals (e.g., novel oral and transdermal technologies) and those that might increase the abuse liability of presently nonscheduled drugs (e.g., nicotine lung inhaler).  Federal regulatory agencies oversee of the development process and grant approval for marketing of new medicines once efficacy and safety have been established.  Abuse liability testing is required for all CNS active drugs as a means of public health protection, and to aid in scheduling recommendations.  The scheduling of drugs is an important issue, as it is thought to influence physician prescribing practices and patient access to new medications.  That is, those medications with very restrictive scheduling (e.g. narcotics) will be relatively more difficult to access.  Scheduling decisions are made by DEA and the Department of Justice with advice from FDA.  These decisions are based on examination of a broad array of data ranging from neuroscience profiling to animal toxicology and behavioral testing to human abuse liability testing and including epidemiology data from overseas if available.  However, specific animal and human testing methods that have been designed to assess abuse liability provide data that are central to these decisions.  

The opening session featured comments by the meeting’s organizer, Charles Schuster, PhD, from Wayne State University.  Dr. Schuster, emphasized as a central theme of the meeting the need to balance access to useful drugs while protecting the public from the consequences of ready access to abusable substances. Glen Hanson, PhD, acting director of the National Institute of Drug Abuse lent a visionary perspective on the important role of medications in the future of drug abuse treatment. Subsequent speakers presented papers designed to provide an overview of the current state of the art in abuse liability testing.  Papers were followed by in-depth discussions among members of the multi disciplinary audience that formed the basis for final meeting recommendations.  Meeting chairs were Lou Harris, PhD, president of the CPDD, and Marty Adler, PhD, Executive Officer of CPDD, and Dr. Charles Schuster.  Speakers included Robert Balster, PhD, from Medical College of Virginia, Nancy Ator, PhD, and Roland Griffiths, PhD, from Johns Hopkins School of Medicine, Ron Mansbach, PhD, from Pfizer Global Research and Development, Kathleen Brady, PhD, from Medical University of South Carolina and Ted Cicero, PhD, from Washington University School of Medicine.  The conclusions and recommendations were developed by an Expert Panel chaired by Edward Sellers, PhD, and summarized and presented by Chris-Ellyn Johanson, PhD, from Wayne State University and Jack Henningfield, PhD, from Pinney Associates. 

Conference participants were in agreement that abuse liability testing is important for public health protection and that all centrally active drugs should undergo such testing.  Further, such testing is required by regulatory agencies.  However, type and extent of evaluation may vary depending on target therapeutic population, context of use, exposure of special at risk populations, intrinsic properties of the drug and dosage delivery form.

Valid methods have been established in both animal and human research to screen abuse liability of centrally active compounds.  In pre-clinical work, techniques used include drug discrimination testing, self-administration and physical dependence testing.  In humans, subjective report profiles, drug discrimination, self-administration and preference testing is used.  Currently employed methods have validity for testing abuse liability of stimulants, opioids and sedatives.  Animal and human studies yield similar findings and both agree with population profiles of abuse. The same methods can form the framework for testing potential abuse liability of novel drugs (e.g. cannabinoids; GHB) and novel drug delivery formulations, although they may need to be modified. 

Despite the availability of excellent and widely used procedures, there is ample room for new research, and this theme was also emphasized at the conference.

For example, Dr. Brady suggested that some initial information relevant to abuse liability could be gleaned from clinical trials.  At present, positive subjective effects are generally not collected systematically in clinical trials, although adverse events are carefully tracked.  This is a wasted opportunity, since all drugs must go through extensive efficacy testing on their way to marketing.  Development of brief scales to assess both positive and negative drug-induced symptoms was recommended.  These assessments could be conducted both during chronic administration and discontinuation phases of the trial to identify reinforcing effects and physical dependence.  Medication adherence data (active vs. placebo) may also be of considerable use for assessing both efficacy and potential for abuse liability.

As emphasized by Dr. Cicero, post-marketing surveillance is an underdeveloped area that provides an excellent opportunity to obtain additional real world data on abuse liability.  Assessment must be based on rates (relative to exposure) of adverse events not absolute number of cases.  Number of patients exposed is the ideal denominator, and risk benefit ratio criteria need to be established.  A distinction must also be made between abuse and misuse (off-label use for therapeutic purposes).  New proactive systems are needed to identify emerging drugs of abuse in a timely manner and taking into account both geographic variation and target/vulnerable populations.

Recommendations of the meeting highlighted consensus development and adoption of standardized abuse liability testing methods, both human and pre clinical, while still leaving flexibility for methodology development to encompass valid testing of new drug entities. There is a need to develop guidelines for evaluating drug effect profiles across multiple measures, and scientific organizations such as CPDD, APA, SRNT and RSA could serve a useful role by helping with guidelines for data interpretation in their area of expertise.

There is a consensus that current methods for testing of opiates, sedatives and stimulants are valid and can be used to test novel compounds with appropriate modification.  However, chronic dosing and drug interaction studies would be useful.  In addition, development of new subjective effects scales is needed to capture qualitative aspects of drug effects including effects of novel compounds outside traditional drug classes.

In the area of animal self-administration testing, it would be useful to develop and validate methods to capture relative abuse liability of weak vs. stronger reinforcers.  Animal models should be extended to account for factors that influence abuse such as genetic vulnerability, drug exposure history, social context of use, alternative reinforcers, and presence of stress or depression. 

Similarly, human testing in therapeutically exposed and other potentially vulnerable populations (e.g. those with history of addiction, recovering alcoholics, people with chronic pain, and people vulnerable to stress or depression) would be important for a more complete understanding of abuse and misuse liability in the general population.  Researchers should consider incorporating into clinical trials measures of positive drug effects (both therapeutic and nontherapeutic), medication adherence and aversive effects during discontinuation.  Such measures would shed additional light on efficacy of the medication and could potentially give an early signal about potential for misuse.

It will be important to identify a battery of core testing methods that are used in all abuse liability testing studies with animals and humans.  Scientific organizations such as CPDD, APA, RSA and SRNT could be instrumental in formulating consensus on such a battery. There is also a critical need to develop post-marketing research methodology, an area that is currently underdeveloped. Finally, it would be highly useful and important to obtain more accurate information about the impact of scheduling regulations on medical practice and on abuse.

Background papers as well as conclusions and recommendations of the meeting will be published in a special issue of the CPDD journal, Drug and Alcohol Dependence. Videotapes and CD’s containing the full meeting discussions can be obtained from the FDA website.