Inhibitors of endocannabinoid breakdown for pain: Not so FA(AH)cile, after all

The analgesic properties of Δ9-tetrahydrocannabinol, the psychoactive ingredient of marijuana, have been difficult to demonstrate conclusively in humans [13]. This is because of limitations imposed by the CNS side effects that arise when high doses of this drug are used to fully activate one of its main targets, the cannabinoid CB1 receptor. Thus, the opportunity of activating cannabinoid receptors “from inside”, i.e. by prolonging the more “tissue- and time-specific” analgesic action of “on demand” biosynthesized endocannabinoids, through inhibition of their inactivation via catabolic enzymes, such as fatty acid amide hydrolase (FAAH) [4], is emerging as a possibly safer alternative strategy [10]. Indeed, pre-clinical studies in animal models of several types of pain, including inflammatory pain, carried out with either irreversible-covalent or reversible-competitive FAAH inhibitors belonging to different chemical classes, produced very promising results [12]. However, the outcome of the first phase II clinical trial of PF-04457845, an irreversible FAAH inhibitor developed by Pfizer in collaboration with The Scripps Institute [1], seems now to contradict these results. In fact, in this issue of Pain, Huggins and co-workers report that, in a randomized, placebo-controlled study, chronic administration of PF-04457845 to patients with pain due to osteoarthritis of the knee, whilst overall very well tolerated and strongly elevating the plasma levels of the endogenous substrates of the enzyme, i.e. the endocannabinoid anandamide and other N-acylethanolamines [10], failed to produce analgesia [9]. The reasons for this failure do not lie in the innovative as well as clever experimental design of the trial, which should have minimized both the chances of non-responders to affect the outcome of the study, and the bias of placebo effects, two frequent problems in pain trials. The study included a one week wash-in period, when all subjects took placebo, followed by a 4-arm treatment period of 2weeks, when patients were randomized into taking 4mg PF-04457845 once a day, 500mg naproxen twice a day, or placebo (two groups). After a 2-week wash-out and a 1week-wash-in period with placebo, the subjects who had taken the FAAH inhibitor or naproxen during the first phase were given placebo, and those who had taken placebo were instead given PF-04457845 or naproxen for another two weeks. The WOMAC (Western Ontario and McMaster Universities Osteoarthritis Index) pain score was used to assess the efficacy of the treatments. At the interim analysis, when only about half of the patients originally planned to be treated with PF-04457845 were actually administered the drug, there was no improvement of the pain score, whilst naproxen was instead effective. Therefore, based on the ineffectiveness of the drug, the study was interrupted. Although ethically necessary, because one of the initial premises of the study was that the FAAH inhibitor should have proved at least more effective than placebo, this premature termination did not allow the authors to fully assess the possible weight of non-responders and placebo in the final outcome of the study, nor to analyze results in patient subsets. In fact, the efficacy of FAAH inhibitors at reducing inflammation might depend on how much this pathological condition is due to up-regulation of the enzyme. Whilst loss-of-function FAAH polymorphisms are associated with other disorders [7], FAAH expression is regulated by several hormones and increases in some animal models of acute and chronic inflammation [3,10]. Therefore, with a higher number of treated subjects, it might have been possible to check if those patients with higher baseline FAAH levels were more responsive to the drug, at least in terms of lower need for rescue medication. Additionally, given the inflammatory nature of the pain condition investigated in this trial, it would have been important to check if the inhibitor had specific biochemical effects, e.g. on the plasma levels of inflammatory cytokines and mediators. It will be important, in future efficacy trials of FAAH inhibitors in inflammatory pain, to correlate response, or lack thereof, with individual baseline FAAH activity/levels and additional biomarker data. However, PF-04457845 was shown by the authors to inhibit leukocyte FAAH by more than 96%, and to elevate plasma anandamide levels in all subjects by at least 10-fold and for several days after the interruption of treatment [9]. Therefore, one would expect that if FAAH does play any role at all at limiting the putative analgesic effects of endogenous anandamide in subjects of osteoarthritic knee pain, some amelioration of pain with the drug should have been seen at the time of the interim analysis. One possible explanation for failure to translate data from animal models of pain to the clinic is that, in the former, one usually assesses objective physical endpoints (e.g. time and/or weight thresholds of resistance to innocuous heat and mechanical pressure), whereas in humans more subjective and mood-sensitive parameters are measured. Indeed, based on previous observations that, for analgesic responses to be observed, patients need to be more anxious than those in the present trial, the authors raised the possibility that the affective component of pain may have contributed to PF-04457845’s lack of efficacy [9]. This possibility might apply to FAAH inhibitors more than to other analgesics because of their anxiolytic-like and anti-depression-like effects in animal models [10]. Therefore, one should perhaps not be too discouraged by the present data. Future studies should investigate PF-04457845 analgesic actions in conditions where patients manifest a stronger affective component, as is the case for neuropathic pain. Indeed, a recent study showed that this condition is accompanied in mice by up-regulation of FAAH in the peri- and infra-limbic cortex, and that local inhibition of the enzyme in this brain area produces anti-allodynic effects [6]. On the other hand, since anticipatory mechanisms, including those characterizing non-opioid placebo analgesia [2], are already accompanied by an enhanced endocannabinoid tone, pain conditions with a strong affective component might also be less responsive to drugs that act by elevating endocannabinoid levels. The authors also suggest that selective and full FAAH inhibition might have unmasked alternative pathways of anandamide and N-acylethanolamine action (i.e. TRPV1 channel activation [14]) and breakdown (e.g. via COX-2-catalysed oxidation), which in turn could have counteracted the beneficial cannabinoid receptor-mediated effects of elevated anandamide levels on pain and inflammation [9]. This interesting hypothesis has been recently reviewed [11], and is supported by experimental evidence in rodents showing that: (1) a dual FAAH/TRPV1 blocker is more efficacious against inflammatory pain than are selective blockers of either target, and also devoid of the worrisome hyperthermic effect of some TRPV1 antagonists [3]; (2) the widely used pain killer and anti-fever agent, acetaminophen (paracetamol), is transformed into a metabolite capable of inhibiting both endocannabinoid inactivation and COX-2 [8]; and (3) inflammation of the knee is accompanied by the formation of prostamide F2α, a COX-2 metabolite of anandamide that exacerbates pain [5]. Therefore, I agree with the authors of this study when they suggest that future trials with FAAH inhibitors in pain should consider also their co-administration with other analgesic/anti-inflammatory drugs such as COX-2 inhibitors as this strategy could improve on the lowest effective dose and safety profile of the latter. Conflict of interest statement The author declares no conflict of interest related to this article.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.