Impact of cyclooxygenase-2 inhibition on cannabis withdrawal and circulating endocannabinoids in daily cannabis smokers

Abstract

Attenuating enzymatic degradation of endocannabinoids (eCBs) by fatty acid amide hydrolase (FAAH) reduces cannabis withdrawal symptoms in preclinical and clinical studies. In mice, blocking cyclooxygenase-2 (COX-2) activity increases central eCB levels by inhibiting fatty acid degradation. This placebo-controlled study examined the effects of the FDA-approved COX-2 selective inhibitor, celecoxib, on cannabis withdrawal, 'relapse', and circulating eCBs in a human laboratory model of cannabis use disorder. Daily, nontreatment-seeking cannabis smokers (12M, 3F) completed a crossover study comprising two 11-day study phases (separated by >14 days for medication clearance). In each phase, the effects of daily BID placebo (0 mg) or celecoxib (200 mg) on cannabis (5.3% THC) intoxication, withdrawal symptoms (4 days of inactive cannabis self-administration) and 'relapse' (3 days of active cannabis self-administration following abstinence) were assessed. Outcome measures included mood, cannabis self-administration, sleep, food intake, cognitive performance, tobacco cigarette use and circulating eCBs and related lipids. Under placebo maintenance, cannabis abstinence produced characteristic withdrawal symptoms (negative mood, anorexia and dreaming) relative to cannabis administration and was associated with increased OEA (a substrate of FAAH) and oleic acid (metabolite of OEA), with no change in eCB levels. Compared to placebo, celecoxib improved subjective (but not objective) measures of sleep and did not affect mood or plasma levels of eCBs or associated lipids and increased cannabis craving. The overall absence of effects on cannabis withdrawal symptoms, self-administration or circulating eCBs relative to placebo, combined with an increase in cannabis craving, suggests celecoxib does not show promise as a potential pharmacotherapy for CUD.

Keywords: cannabis use disorder; marijuana; self-administration.

Figure 1:

Selective mean peak cluster ratings and caloric intake during cannabis administration and across 4 days of cannabis abstinence as a function of celecoxib dose. Maximum score for ratings = 100 mm. Each graph represents data collected in 15 participants. Error bars represent ± SEM.

Figure 2:

Mean peak ratings of cannabis craving during cannabis administration and across 4 days of cannabis abstinence as a function of celecoxib dose (left panel) and mean number of cannabis puffs purchased for self-administration across 3 days following cannabis abstinence. Error bars represent ± SEM. Asterisks indicate a significant difference between celecoxib and placebo (** p < 0.01).

Figure 3:

Mean plasma levels of eCBs and associated lipids as a function of time since last cannabis use (11 hours vs 4.5 days) and celecoxib dose. Eleven hours after the last cannabis cigarette reflects non-abstinent conditions and 4.5 days after the last cannabis cigarette presents reflects cannabis abstinence. Error bars represent ± SEM. Number signs indicate a significant difference during cannabis administration and during cannabis abstinence under placebo medication conditions (# p < 0.05; ## p < 0.01).