The Cross-Company Abuse Liability Council (CCALC) meeting held in Washington, DC, covered several themes related to the evaluation of abuse potential, including behavioral economics, physical dependence, and adverse events (AEs); the discussion convened experts from FDA’s Controlled Substances Staff (CSS), academia, and industry. Dr. Vince Clinical Research (DVCR) hosted a webinar that reviewed CCALC discussions related to the challenges in evaluating abuse potential data and recent regulatory guidelines related to the abuse potential of new molecular entities. Speakers included Dr. Naama Levy-Cooperman and Dr. Kerry Schoedel from Altreos Research Partners, and Dr. Lynn Webster, Executive VP, Scientific Affairs, from DVCR.
Dr. Levy-Cooperman’s section focused on two primary themes: 1) addressing discordant findings between clinical and non-clinical data; and 2) improving statistical methods for analyzing human abuse potential (HAP) studies. She began by reviewing the basic design of HAP studies, noting that these studies are typically randomized, double-blind, crossover designs1,2. These studies compare the investigational drug to placebo and an active control drug (i.e., controlled substances that have similar effects or pharmacology as the investigational products), using various subjective measures, such as drug liking, take drug again, good and bad effects, and alertness or drowsiness. These measures—administered as computerized visual analog scales—are collected at multiple time points to capture the full-time course of a drug’s effects. In addition, special focus is given to the inclusion of an enriched population of non-dependent recreational drug users, which helps to reduce variability in responses.
Dr. Levy-Cooperman discussed a presentation given by Dr. Silvia Calderon (FDA CSS) that summarized potential discordance between animal models and human trials by presenting cases in which drugs showed minimal abuse signals in animal studies, but that demonstrated positive signals on subjective measures predictive of abuse collected in HAP studies. These measures include abuse-related adverse AEs, such as euphoria, that have been demonstrated in other clinical trials, highlighting that animal models cannot always predict human responses accurately. Data were presented from recent studies of anti-epileptic drugs and dual-receptor antagonists, neither set of results showing potential for abuse in animal studies, yet both showing positive signals in subsequent human trials. This discrepancy underscores the importance of HAP studies in drug development.
The second area of focus discussed was the statistical methods used in these studies, particularly the challenges posed by non-responders or outliers (i.e., subjects who respond in an inappropriate manner to the positive control or placebo during the Treatment Phase). Dr. Levy-Cooperman summarized a talk given by Dr. Ling Chen (FDA statistician), in which improvements were proposed to subject qualification (i.e., a phase of the study prior to the treatment phase that ensures subjects can reliably distinguish between placebo and an active control). Dr. Chen recommended implementing stricter criteria for subject inclusion to improve study reliability, such as excluding non-responders to control drugs, which would enhance the validity of the study’s findings.
in an inappropriate manner to the positive control or placebo during the Treatment Phase). Dr. Levy-Cooperman summarized a talk given by Dr. Ling Chen (FDA statistician), in which improvements were proposed to subject qualification (i.e., a phase of the study prior to the treatment phase that ensures subjects can reliably distinguish between placebo and an active control). Dr. Chen recommended implementing stricter criteria for subject inclusion to improve study reliability, such as excluding non-responders to control drugs, which would enhance the validity of the study’s findings.
Despite the advancements, the methodology for HAP studies is still evolving. Dr. Levy-Cooperman discussed new recommendations, such as improving the qualification phase to ensure inclusion of more reliable responders. Dr. Chen suggested the use of a modified completer population approach that would exclude non-responders from the analysis, as this modification has been shown to improve the accuracy of the results. However, this approach also poses challenges, such as reducing the sample size, which could affect study power.
Overall, it is clear that HAP studies are essential for informing scheduling decisions of controlled substances. However, the webinar emphasized that these studies are just one part of a larger decision-making process. The totality of data, including clinical trials, non-clinical data, and post-marketing surveillance, ultimately determine a drug’s schedule. As new drugs are developed, the methods for assessing abuse potential continue to evolve, aiming to improve study design, analysis, and the accuracy of conclusions regarding a drug’s potential for abuse.
The second speaker at the webinar, Dr. Kerri Schoedel, focused on the collection and analysis of abuse and dependence-related data from clinical studies, particularly those from Phase 1 to Phase 3 clinical trials. Dr. Schoedel addressed two primary topics: abuse-related AEs and drug-discontinuation-emergent AEs.
Abuse-related AEs are defined as special interest events that require systematic categorization, tabulation, and analysis to detect signals of potential for abuse. These AEs provide critical information for evaluating the abuse potential of drugs during development, helping to inform regulatory submissions and determine drug scheduling. The analysis primarily looks for subjective effects such as euphoric mood, sedation, stimulation, and hallucinations, which are directly identified in the Controlled Substances Act (CSA)3. A particular challenge in collecting these data is ensuring accurate reporting by investigators, as these events are often subjective. For example, euphoria might be misinterpreted or missed as a non-adverse effect, leading to gaps in understanding the pharmacological effects of the drug. Misreporting and miscoding are common issues, as shown in examples where a patient’s description of feeling “hungover” was incorrectly coded as “drunk,” an AE of particular interest.
missed as a non-adverse effect, leading to gaps in understanding the pharmacological effects of the drug. Misreporting and miscoding are common issues, as shown in examples where a patient’s description of feeling “hungover” was incorrectly coded as “drunk,” an AE of particular interest.
Drug-discontinuation-emergent AEs occur after stopping chronic medication and provide insights into physical dependence potential. These events are important not only for assessing the potential for dependence in humans, but also for determining proper drug scheduling. The complexity of interpreting drug-discontinuation-emergent AEs arises from confounding factors, such as drug half-life and re-emergence of underlying disease symptoms. Therefore, comparing the incidence of these events with active drug relative to placebo, and comparing on-treatment versus off-treatment AEs, are essential for accurate assessment. Additionally, analyzing clusters of symptoms that occur together can help identify withdrawal syndromes or signs of physical dependence.
In conclusion, Dr. Schoedel emphasized that while abuse-related and drug-discontinuation-emergent AEs do not definitively signal abuse potential or dependence, they form crucial components of a broader assessment. Proper training for investigators and thorough data analysis are key to collecting meaningful data that can guide clinical decisions and assist with regulatory evaluations regarding drug safety and abuse potential.
Dr. Lynn Webster’s talk focused on the complexities of withdrawal symptoms in relation to HAP studies, particularly their implications for drug labeling and understanding acute versus chronic withdrawal symptoms. One of the primary concerns in these studies is evaluating the effects of discontinuing a drug, which can lead to physical dependence and withdrawal symptoms. Withdrawal can manifest both physiologically and psychologically, with short-term symptoms typically appearing within hours to days and psychological symptoms sometimes persisting for weeks or months. These symptoms can vary in severity depending on the drug class and how the withdrawal is managed (e.g., abrupt versus tapered discontinuation).
In terms of assessing withdrawal, various scales are employed, such as the visual analog scale for acute symptoms and specific scales for different drugs, like the Clinical Opiate Withdrawal Scale (COWS) for opioids. These scales are crucial in understanding both the onset and duration of withdrawal symptoms, which can vary widely. For instance, cannabis withdrawal can last for weeks, while symptoms from drugs like benzodiazepines can extend up to a year. Such prolonged withdrawal effects are sometimes missed in traditional clinical trials, which typically focus on short-term evaluations.
Dr. Webster’s presentation also emphasized that the severity and duration of withdrawal symptoms can be influenced by various factors, including the patient’s health condition, treatment adherence, and history of polysubstance abuse, or the drug’s mechanism of action. Additionally, mental health conditions and comorbidities can exacerbate withdrawal symptoms. It was also noted that some drugs, such as beta blockers, may not result in significant withdrawal effects despite physical dependence.
The conversation underscored the need for comprehensive drug labeling that includes information about withdrawal symptoms, as understanding these effects is crucial for clinicians. For example, abrupt withdrawal from opioids could trigger intense symptoms, leading patients to relapse into substance use disorder. As such, proper management strategies are critical to mitigate harm.
Dr. Webster concluded by pointing to advancements in telehealth and mobile health, which provide real-time data collection and intervention opportunities. These technologies allow for continuous monitoring and more effective management of long-term withdrawal symptoms, offering potential solutions for addressing the often-overlooked chronic withdrawal effects that can extend for months, or even years. Overall, Dr. Webster’s presentation highlighted the importance of thorough withdrawal symptom evaluations in drug studies and their application to patient care, emphasizing the need for accurate labeling and personalized treatment approaches in the real world.
In summary, Dr. Levy-Cooperman, Dr. Schoedel, and Dr. Webster provided additional insights into some of the complexities of evaluating human abuse potential, from study design and statistical methods to the collection and analysis of abuse-related AEs and withdrawal symptoms. As new drugs continue to emerge, the methodologies and elements of the regulatory guidelines for assessing abuse potential are adapting to ensure more reliable and accurate evaluations. Ultimately, the advancements discussed in the webinar underscore the need for ongoing collaboration between key opinion leaders in industry, academia, and regulatory bodies, to ensure that drug development aligns with public health and safety priorities.
1Chen L, Tsong Y. Design and Analysis for Drug Abuse Potential Studies: Issues and Strategies for Implementing a Crossover Design. Drug Information Journal. 2007;41(4):481-489.
2Schoedel, K. and Sellers, E. (2008), Assessing Abuse Liability During Drug Development: Changing Standards and Expectations. Clinical Pharmacology & Therapeutics, 83: 622-626.
3“Title 21 United States Code (USC) Controlled Substances Act”. Drug Enforcement Administration: Office of Diversion Control. United States Department of Justice. Retrieved January 31, 2025.
