Individual differences in human opioid abuse potential as observed in a human laboratory study

Abstract

Background: Opioids have high abuse potential and pose a major public health concern. Yet, a large percentage of individuals exposed to opioids do not develop problematic use. Individual differences in opioid abuse potential are not well understood.

Methods: This within-subject (N = 16), double-blind, double-dummy, human laboratory study evaluated individual differences in response to dose (placebo, low, medium, high) following administration of heroin and hydromorphone through intravenous and subcutaneous routes, in opioid-experienced but non physically-dependent participants. Outcomes were self-reported visual analog scale (VAS) ratings (High, Liking, Drug Effect, Good Effect, Rush), pupil diameter change from baseline, and crossover point on the Drug vs. Money questionnaire. The degree to which results were consistent across measures within an individual was assessed using a mixed-effects model from which an intraclass correlation coefficient measure of between and within-subject variance was derived.

Results: The mixed effects model fit was significant (p < 0.0001) and revealed that 85.5% of the explainable variance was due to between-subject effects, suggesting the responses within an individual were highly consistent. Visual inspection reveals a myriad response pattern across participants, with some demonstrating classic dose-effect responses and others not differentiating any active doses from placebo.

Conclusions: Data suggest the abuse potential of opioids is significantly different between individuals but that the experience within an individual is highly consistent. Research to prospectively characterize and evaluate mechanisms underlying these differences is warranted and may provide a foundation to help identify persons at heightened risk of transitioning from opioid exposure to misuse and/or opioid use disorder.

Keywords: Heroin; Human individual differences; Hydromorphone; Opioid; Personalized medicine.

Figure 1.. Individual Outcomes Collapsed Across Drug and Route.

Values represent the peak rating during each drug administration session for five independent visual analog scales (VAS; Drug Effect, High, Good Effect, Drug Liking, Rush; gray-scale symbols), the drug vs. money choice crossover value (open square), and pupil diameter change from baseline in millimeters (mm; open circle). The left Y-axis represents peak ratings for the VAS (scale 0–100) and the drug vs. money choice procedure (scale 0–50). The right Y-axis represents pupil size in mm; since pupillary constriction is measured for evidence of an effect, the direction of the right Y-axis is descending so data will move in the same direction as the other outcomes. For all three outcomes, higher rankings represent greater effects. Results are collapsed across drug (heroin, hydromorphone) and route (intravenous, subcutaneous) and presented as a function of dose (placebo, low, medium, high) along the X-axis. Each graph represents a different participant (range 1–16), as designated by the number in the upper left corner. A mixed effects model (p<0.0001) that transformed all seven outcomes to the same scale for comparison yielded an intraclass correlation coefficient (ICC) of 0.855, indicating 85.5% of explainable variance in the model resulted from between-subject comparisons, suggesting a high degree of within-subject consistency in response patterns.