Identification of Optimal Measures of Human Abuse Potential: A Post Hoc Assessment of 19 Studies.

BACKGROUND: Human abuse potential studies include multiple measures to assess the subjective effects of central nervous system-active drugs. In this retrospective analysis, measurement properties of commonly used measures were assessed, and factor analysis was conducted to identify a core battery of measures. METHODS: Measures of positive, negative and other effects, for example, bipolar "at-the-moment" Drug Liking visual analog scale (VAS), were derived for active controls and placebo from 19 studies in recreational drug users (N = 570). Distribution, placebo response, variability, convergent/discriminant validity, parameter effect sizes (eg, maximum effect [Emax], time-averaged area under the effect curve), and predictive validity were evaluated. A factor analysis was conducted with 9 studies. RESULTS: Most parameters were not normally distributed. Bipolar VAS exhibited the lowest variability. Drug Liking VAS Emax was very sensitive, showed large effect sizes (>1.0), and was moderately to strongly correlated with Emax of other positive effects measures (r > 0.5), but weaker with less specific scales (eg, high, Any Effects VAS); time-averaged area under the effect curve showed higher variability and lower effect sizes. Maximum effect at any dose (EmaxD) was significantly correlated with Emax across all selected measures and showed higher effect sizes. In the overall factor analysis, factors could be categorized into positive effects/euphoria (77% of variance), negative effects (17.9%), and pharmacologic effects (5%). For predictive validity, effect sizes for Drug Liking VAS Emax/EmaxD were moderately correlated with postmarket adverse events related to abuse (R = 0.52). CONCLUSIONS: A core battery of 7 subjective measures was proposed, with additional measures added based on pharmacologic effects.

Abuse potential assessment of cannabidiol (CBD) in recreational polydrug users: A randomized, double-blind, controlled trial.

RATIONALE: Treatment with a highly purified oral solution of cannabidiol (CBD), derived from the plant Cannabis sativa L., demonstrated some evidence of central nervous system (CNS)-related adverse events in patients enrolled in phase 3 trials for treatment of childhood-onset epilepsy. Cannabidiol was categorized as a Schedule 1 substance by the United States Drug Enforcement Administration; therefore, it was important to test CBD for human abuse potential. METHODS: This was a single-dose, randomized, double-blind, double-dummy, placebo- and active-controlled crossover trial. The abuse potential of single oral doses of plant-derived pharmaceutical formulations of highly purified CBD (Epidiolex®; 750 mg, 1500 mg, and 4500 mg) was compared with that of single oral doses of alprazolam (2 mg), dronabinol (10 mg and 30 mg), and placebo in healthy recreational polydrug users. The primary endpoint to assess abuse potential was the maximum effect (Emax) on Drug-Liking visual analog scale (VAS). Other measurements included Emax on Overall Drug-Liking VAS, Take Drug Again VAS, positive and negative effects, other subjective effects, and Drug Similarity VAS. Cognitive and psychomotor functions were assessed using the Divided Attention Test, the Hopkins Verbal Learning Test-Revised, and the Digit-Symbol Substitution Task. Pharmacokinetic parameters were determined for CBD and its major metabolites. Standard safety measures and adverse events were assessed. PRINCIPAL RESULTS: Of 95 eligible subjects, 43 qualified for the treatment phase, received at least 1 dose of investigational medicinal product, and were included in safety assessments; 35 subjects were included in the pharmacodynamic analysis. Subjects receiving alprazolam and dronabinol had significantly higher Drug-Liking Emax (P < 0.0001) compared with those receiving placebo, confirming study validity. Compared with placebo, Drug-Liking was not significantly different for subjects taking 750-mg CBD (P = 0.51). Drug-Liking Emax values for 1500-mg and 4500-mg CBD were significantly different from placebo (P = 0.04 and 0.002, respectively); however, the mean differences were <10 points on VAS compared with >18-point differences between positive controls and placebo. Alprazolam and dronabinol had significantly higher Drug-Liking, Overall-Liking, and Take Drug Again VAS Emax values compared with all doses of CBD (P ≤ 0.004). In contrast to alprazolam, CBD administration had no observable effect on cognitive/psychomotor tests. Pharmacokinetic parameters for CBD in this trial were consistent with previous studies. The majority of adverse events reported during the trial were of mild or moderate severity; no serious adverse events or deaths were reported. CONCLUSION: Administration of a therapeutic dose of CBD (750 mg) showed significantly low abuse potential in a highly sensitive population of polydrug users. Although high and supratherapeutic doses of CBD (1500 mg and 4500 mg, respectively) had detectable subjective effects compared with placebo; the effects were significantly lower than those observed with alprazolam and dronabinol.

Assessment of the Abuse Potential of the Orexin Receptor Antagonist, Suvorexant, Compared With Zolpidem in a Randomized Crossover Study.

Suvorexant is a dual orexin receptor antagonist approved in the United States and Japan for the treatment of insomnia at a maximum dose of 20 mg. This randomized double-blind crossover study evaluated the abuse potential of suvorexant in 36 healthy recreational polydrug users with a history of sedative and psychedelic drug use. Single doses of suvorexant (40, 80, and 150 mg: 2-7.5 × maximum dose), zolpidem (15 and 30 mg: 1.5-3 × maximum dose), and placebo were administered, with a 10-day washout between treatments. Subjective and objective measures, including visual analog scales (VASs), Addiction Research Center Inventory, and cognitive/psychomotor tests, were evaluated for 24-hour postdose. Suvorexant had significantly greater peak effects on "drug liking" VAS (primary endpoint) than placebo. Although effects of suvorexant on abuse potential measures were generally similar to zolpidem, they remained constant across doses, whereas zolpidem often had greater effects at higher doses. Suvorexant (all doses) had significantly fewer effects than zolpidem 30 mg on secondary measures, such as "high" VAS, Bowdle VAS, and Addiction Research Center Inventory morphine-benzedrine group. The overall incidence of abuse-related adverse events, such as euphoric mood and hallucination, was numerically lower with suvorexant than zolpidem. In agreement with its classification as a schedule IV drug, suvorexant demonstrated abuse potential, compared with placebo. The abuse potential was similar to zolpidem using certain measures, but with a reduced incidence of abuse-related adverse events. Although this suggests that the overall abuse liability of suvorexant may be lower than zolpidem, the actual abuse rates will be assessed with the postmarketing experience.

Abuse liability assessment of eslicarbazepine acetate in healthy male and female recreational sedative users: A Phase I randomized controlled trial.

RATIONALE: Eslicarbazepine acetate (ESL) is a once-daily oral antiepileptic drug for the treatment of partial-onset seizures. Adverse events such as dizziness and somnolence reported in clinical studies suggest that ESL has detectable central nervous system (CNS) effects in addition to its antiepileptic effects. This Phase I study evaluated the abuse liability of ESL compared with that of alprazolam (ALP) and placebo (PBO) in recreational CNS depressant users. METHODS: In this single-dose, randomized, double-blind, PBO- and active-controlled crossover study, healthy recreational CNS depressant users who could discern between ALP 2mg and PBO received single oral doses of each of the following treatments with a washout interval of ≥7days between each treatment: ESL (800mg, 1600mg, 2000mg, and 2400mg); ALP (1.5mg and 3.0mg); and PBO. Subjective measures, including visual analog scales (VASs) e.g., Drug-Liking (primary endpoint), and Addiction Research Center Inventory (ARCI) Morphine-Benzedrine Group (MBG), Pentobarbital Chlorpromazine Alcohol Group (PCAG), and Lysergic Acid Diethylamide Group scales were evaluated at multiple time points up to 24h postdose. Cognitive effects were evaluated using the Choice Reaction Time (CRT), Divided Attention (DAT) and Hopkins Verbal Learning Task-Revised tests. PRINCIPAL RESULTS: Peak scores for Drug-Liking VAS (maximum effect [Emax]) were significantly higher for both ALP doses than for PBO (p<0.0001), thereby confirming study validity. Drug-Liking VAS Emax was significantly lower for all ESL doses than both ALP doses (p<0.0001). Drug-Liking VAS Emax for ESL 800mg was similar to that for PBO (least squares [LS] mean difference: 3.6; p=0.19). At the three higher ESL doses (1600mg and the supratherapeutic doses of 2000mg and 2400mg), Drug-Liking VAS Emax was significantly higher than for PBO, although the differences were minimal (LS mean difference: 9.3-13.3 out of 100). For most secondary subjective endpoints (i.e., Good Effects VAS and High VAS, ARCI-MBG, Take Drug Again VAS, Overall Drug-Liking VAS, and ARCI-PCAG; p<0.05), the effect of ESL (all doses) was significantly less than that of ALP (both doses). On most secondary measures, the dose-response relationship was relatively flat or showed saturation at higher ESL doses. Although significant differences were observed for ESL compared with those for PBO for some specific CRT and DAT endpoints (i.e., reaction time, manual tracking, hit latency), ALP demonstrated significant and dose-dependent impairment on the majority of cognitive endpoints when compared with PBO and ESL. Mean plasma concentrations of the active metabolite of ESL, eslicarbazepine, increased with increasing ESL dose. Pharmacokinetic parameters estimated for eslicarbazepine were generally comparable with results from previous studies in healthy volunteers. CONCLUSION: This study demonstrated that single doses of ESL may have less abuse liability than ALP in recreational sedative users. Although ESL had detectable subjective effects and showed some drug-'liking' at higher doses, the magnitude of these effects was small.

Estradiol and mental rotation: relation to dimensionality, difficulty, or angular disparity?

Several studies have reported that performance on spatial rotation tests is better at menses than at high estradiol phases of the menstrual cycle in women. These effects are debated because nearly all reports of menstrual cycle variability have relied on a single test, the Mental Rotations Test (MRT, Vandenberg and Kuse, 1978). In the present study, we investigated key features of the MRT that might be responsible for its association with estradiol levels. We hypothesized that associations could be demonstrated for other tasks that share the same characteristics. Forty-four women ages 20-38 years, matched on education and general ability, were assessed at low (n=24) or high (n=20) estradiol stages of the menstrual cycle on a set of spatial tests that varied in dimensionality, plane of rotation, angular disparity, and effortfulness. Saliva was used to quantify estradiol and progesterone. Low estradiol was found to be associated with significantly better accuracy on the MRT and also on a mental rotation task that required large angles of rotation but employed only two-dimensional object representations and rotations limited to the picture plane. In contrast, a task using identical stimuli that required only small angles of rotation did not show an estradiol effect. A group difference also was seen on a test of perceptual closure. The results confirm that the estradiol effect is not limited to the MRT, and identify the rotational element, but also aspects of figural perception, as possible processes that may be responsive to estrogens. These findings advance our understanding by showing an association between estradiol and discrete spatial processes. Implications for understanding the origins of the robust sex difference commonly observed on the MRT are discussed.