Most pharmaceutical companies have been reticent to invest in addiction due in part to stigma (
79) as well as to perceptions that the market is small, that executing clinical trials in patients with substance use disorders is difficult (because of frequent comorbidities, criminal or legal problems, and poor adherence to treatment protocols), and that the regulatory bar required for FDA approval is too high (i.e., abstinence—discussed below [
80]). These factors represent a major challenge in medication development. Consequently, the National Institute on Drug Abuse (NIDA) has focused on partnering with industry to encourage drug development by identifying promising targets and funding research to lessen risks associated with drug development. Examples of such partnerships include helping in the development of medications to treat opioid use disorders, such as a buprenorphine/naloxone combination (Suboxone) (
81), a 1-month extended-release naltrexone (Vivitrol), a 6-month buprenorphine subdermal implant (Probuphine) (
82), and a “user-friendly” intranasal opioid overdose naloxone formulation (Narcan Nasal Spray) (
83). The current opioid crisis has further highlighted the urgency for greater participation of industry in medication development (
84).
Research is also ongoing to define endpoints other than abstinence for measuring treatment efficacy, such as clinically meaningful reductions in drug use associated with improvements in health outcomes (
95). Such adjustments could reduce the regulatory bar for obtaining FDA approval for new therapeutics. For example, research has indicated that reduced use of cocaine led to decreased endothelial dysfunction, a marker of heart-disease risk associated with chronic cocaine use (
96). Importantly, there is precedent for alternative endpoints: the FDA has approved the use of percentage of subjects with no heavy drinking days as an outcome for alcohol use disorder. This measurement is sensitive for detecting differences between medication and placebo, and allowing some days of consumption increases the effect size (
97). In the case of opioid use disorders, a relevant endpoint could be overdose prevention; recently, a study on parolees found reduced overdoses were associated with extended-release naltrexone (
98).
Promising Pharmacological Targets
The preclinical literature includes a wide range of promising strategies for substance use disorders that aim either at the main protein target of specific drugs, at modulators of the brain reward system, or at modulators of downstream circuits disrupted in addiction (
Figure 1).
Therapeutic strategies that target the mu-opioid receptor have been most effective for opioid use disorders and are being pursued in the development of nonaddictive analgesics, which could help reduce the risk of addiction as an unintended consequence of pain management (
84). In this respect, the recent identification of the structure of the mu-opioid receptor has provided novel insights into mechanisms of tolerance and is facilitating development of medications that target specific intracellular signaling pathways of the mu-opioid receptor, referred to as biased agonists. Biased opioid agonists that are developed as analgesics target the G-coupled protein “G
i” intracellular pathway, which is believed to underlie analgesia, while not engaging the β-arrestin pathway, which is associated with tolerance and the respiratory-depressing effects of opioid agonists (
99). Phase 3 clinical testing of a mu-opioid receptor biased agonist (TRV130) is under way (
100,
101). In addition, the orvinol analogue BU08028, a compound similar to buprenorphine, has been shown to be a safe opioid analgesic without abuse liability in nonhuman primates (
102). Research into these and other opioids (
103) is poised to lead to improved treatments for opioid use disorders and pain.
Novel pharmacological approaches for treating opioid use disorders that do not involve the mu-opioid receptor include strategies to modulate the reward circuit via antagonism of the neurokinin 1 receptor (
104) and to counteract withdrawal via antagonism of the kappa-opioid receptor (
89). Lofexidine, an α
2A-adrenergic receptor agonist originally developed as an antihypertensive, also decreases opioid withdrawal symptoms and is undergoing NIDA-funded trials (
105,
106). Another promising medication is lorcaserin, a selective 5-HT
2C receptor agonist already FDA-approved for weight loss that reduces opioid seeking in rodent models (
107).
Oxytocin, a neuropeptide known for its role in social bonding, is also of interest for substance use disorder. Oxytocin-expressing neurons project to brain regions implicated in reward (including the ventral tegmental area and nucleus accumbens) (
108,
109) and stress (including the amygdala and hippocampus) (
108,
109). Preclinical studies have shown that oxytocin decreases self-administration of heroin (
110), cocaine (
111,
112), methamphetamine (
113), and alcohol (
114–
116) and also alleviates nicotine withdrawal (
117). Oxytocin treatment during adolescence also reduced methamphetamine (
118) and alcohol (
119) seeking in adult rodents. In addition, oxytocin reduced reinstatement of drug seeking in rodents in response to triggers of drug craving for methamphetamine (
120–
122) and cocaine (
112,
123,
124). In humans, intranasal oxytocin reduces cue-induced craving for nicotine (
125), stress-induced craving for marijuana (
126), and withdrawal and anxiety symptoms.
Cannabinoids may also be useful for treating substance use disorders, and identifying medications that target the endocannabinoid system without producing cognitive impairment and rewarding effects could lead to new treatments for substance use disorders. For example, dronabinol (a synthetic tetrahydrocannabinol [THC] formulation approved for AIDS-related anorexia and chemotherapy-related nausea) reduced withdrawal symptoms associated with cannabis use disorder (
127), and nabilone (a synthetic cannabinoid similar to THC) reduced cannabis withdrawal- and relapse-related measures in human laboratory studies (
128). In addition, a study of nabiximol (an oral mucosal spray containing a THC-to-cannabidiol [CBD] ratio of 1:1) found reductions in severity and duration of cannabis withdrawal and increased retention in treatment (
129). Finally, an ongoing randomized clinical trial will evaluate whether CBD (Epidiolex), when added to medical management, can improve treatment outcomes for cannabis use disorders (
130).
The body’s endogenous cannabinoids (anandamide and 2-arachidonoylglycerol, or 2-AG, which interact with cannabinoid receptors CB1R and CB2R) optimize the inhibitory and excitatory balance in the brain in a state-dependent manner (
131), so side effects might occur with orthosteric ligands that either activate (e.g., dronabinol) or block (e.g., rimonabant) CB1R broadly. For example, rimonabant, which showed efficacy in treating obesity and inhibiting the rewarding effects of cannabis, was also associated with negative affect and suicidal ideation (
132). Instead, the manipulation of cannabinoid receptors (CB1R or CB2R) via allosteric modulators, which simply enhance or inhibit receptor responses to endocannabinoids, maintaining the state dependence of the endocannabinoid system, may be more promising for medication development. Although the development of CB1R allosteric modulator medications is in its infancy (
133), the negative allosteric modulator pregnenolone appears to protect against the intoxicating effects of THC (
134) and is being evaluated for the treatment of cannabis use disorder. In addition, the positive allosteric CB1R modulator ZCZ011 has antinociceptive effects without being reinforcing (
135) and thus holds potential as a nonaddictive analgesic.
Other endocannabinoid system modulators being evaluated for cannabis use disorder include specific inhibitors of fatty acid amide hydrolase or of monoglycerol lipase, which slow the breakdown of endocannabinoids, as these compounds may reduce withdrawal symptoms and/or may lead to new nonaddictive analgesics (
136,
137).
Stimulant use disorders have been among the most challenging for therapeutics development. The most investigated strategy has been the use of longer-acting stimulant medications (e.g., methylphenidate and amphetamines) for cocaine addiction. Many of these studies have failed to reach significant positive outcomes, except in individuals with comorbid attention deficit hyperactivity disorder. However, a meta-analysis reported evidence of mild benefit for the use of amphetamines in cocaine addiction (
138). Similarly, modafinil, a medication used for narcolepsy that has mild stimulant properties (
139), has been shown by some studies to be beneficial for the treatment of cocaine addiction (
140), though not by others (
141). A different approach, based on studies documenting that enhanced glutamatergic signaling from limbic and ventral prefrontal regions can drive cue-induced cravings and relapse (
142), targets medications that can help restore balance to these glutamatergic projections. For example,
N-acetylcysteine, which helps modulate glutamate signals by activating the cystine-glutamate exchange and thereby stimulating extrasynaptic metabotropic glutamate receptors (
143), decreases cocaine seeking in animal models (
144).
N-acetylcysteine is well tolerated in cocaine-dependent individuals and may reduce cocaine-related withdrawal symptoms and craving (
145). A clinical trial did not find
N-acetylcysteine to be effective for patients actively using cocaine; however,
N-acetylcysteine reduced cravings and prevented relapse in patients who had achieved abstinence (
146). The efficacy of
N-acetylcysteine is limited by low bioavailability and poor permeability of the blood-brain barrier. Similar compounds with greater potency and bioavailability, such as
N-acetylcysteine amide, may prove more effective (
147).
Strategies to deliver degradative enzymes have also been proposed for cocaine use disorders, and research is ongoing to develop stable, long-lasting forms of cocaine-degrading enzymes, including cocaine esterase, cocaine hydrolase (
148–
151), and butyrylcholinesterase (
152). Preclinical evidence has shown that these fusion proteins increase cocaine’s metabolism (
153,
154). However, clinical studies have failed to show efficacy (
152).
Antiepileptic drugs have also been evaluated in the management of substance use disorder. A proposed mechanism for their benefit is that in cocaine use disorders, and in other substance use disorders, dysfunction of GABA-ergic signaling contributes to drug taking (
155,
156). Gabapentin, a widely prescribed anticonvulsant and pain medication, has shown some benefits in the treatment of alcohol use disorders (
157) and might reduce cannabis use and withdrawal symptoms in cannabis use disorders. Topiramate is an antiepileptic medication (
158) for which there is preliminary clinical evidence of reduced cocaine use and improved outcomes in cocaine-dependent individuals (
159). In methamphetamine users, topiramate reduced the amount of drug taken by patients with an active use disorder and reduced relapse rates among those in recovery (
160).