Efficacy of Ketamine in the Treatment of Substance Use Disorders: A Systematic Review

Alcohol and illicit drug use represent an escalating global public health burden, with substantial prevalence and rising mortality driven in part by opioid-related overdoses. Substance use disorders (SUDs) are characterised by impaired control, craving, social impairment, risky use and withdrawal syndromes; severe withdrawal from alcohol or opioids can be life-threatening and is a common barrier to sustained treatment engagement. Existing pharmacotherapies are limited: approved medications modestly reduce relapse in alcohol use disorder and effective options for stimulant and cannabis use disorders are lacking, while opioid dependence relies on opioid agonists, partial agonists or antagonists. This review examines ketamine—a potent, non-competitive NMDA receptor antagonist with demonstrated rapid antidepressant effects—as a candidate pharmacotherapy for SUDs. Given preclinical and clinical evidence implicating glutamatergic dysregulation in addiction and ketamine’s putative synaptic and plasticity-promoting actions, the investigators set out to systematically summarise human studies published between 1996 and 2018 that evaluated ketamine for treatment of SUDs or withdrawal, and to outline ongoing trials and future research directions.

A systematic search of PubMed/MEDLINE, Scopus and ClinicalTrials.gov was conducted for human studies published from 1 January 1996 to 1 January 2018. PubMed searches combined the MeSH terms “Substance-Related Disorders” and “Ketamine/therapeutic use”; Scopus searches used “ketamine” with title keywords for specific substances (for example, cocaine, alcohol, cannabis, heroin, opi*); ClinicalTrials.gov was searched for “ketamine.” Results were limited to full-text human studies available in English and case reports were excluded. The inclusion criteria required that studies evaluated the efficacy of ketamine, with or without a behavioural intervention, for treating a SUD or substance withdrawal. After de-duplication and eligibility screening, seven completed clinical studies met the inclusion criteria and six ongoing trials were identified. The extracted text refers to a Figure and Table that summarise study characteristics and included studies, but these items were not present in the provided extraction; where specific methodological details appear only in those missing items, the extraction does not clearly report them. The investigators did not report using a formal risk-of-bias or quality-assessment tool in the extracted text, nor specify additional search limits beyond the database queries and language/human filters.

Seven completed clinical studies were identified, covering cocaine, alcohol, and opioid use disorders; no human studies were found for tobacco or stimulants other than cocaine. Six ongoing trials were also located on ClinicalTrials.gov addressing alcohol, cocaine, opioid and cannabis use disorders. Cocaine use disorder: Two small crossover/laboratory studies evaluated ketamine. In an eight-participant three-arm crossover trial, 0.41 mg/kg and 0.71 mg/kg IV ketamine were compared with an active lorazepam control (2 mg). Primary outcomes were motivation to quit (URICA) and cue-elicited craving (visual analogue scale, VAS). After the first infusion, ketamine increased motivation relative to lorazepam (reported median score change 0.15 vs 3.6, p = 0.012) and reduced craving by a mean of 168 mm on the VAS (approximately 60% change, p = 0.012). There was a significant reduction in self-reported frequency of use from 22/28 days at baseline to 5/28 days at 4-week follow-up (p = 0.012) and in money spent per use day (USD 149.30 to USD 10.50, p < 0.001). Dakwar and colleagues conducted a related inpatient crossover laboratory study of 20 non-treatment-seeking participants comparing a single 0.71 mg/kg ketamine infusion to 0.025 mg/kg midazolam. During 6 days of hospitalisation with “choice sessions” (self-administer 25 mg cocaine or receive USD 11), ketamine reduced cocaine self-administration rates by 66% relative to pre-infusion baseline (p < 0.0001), with no significant pre/post change in the control group. Opioid use disorder and withdrawal: Multiple studies evaluated ketamine in heroin-dependent populations and in opioid antagonist induction settings. One randomised trial of 70 heroin-dependent participants compared high-dose ketamine (2 mg/kg IM) versus low-dose ketamine (0.2 mg/kg IM), both combined with psychotherapy. Abstinence rates in the high-dose group were approximately 85% at 1 month versus 55% in the low-dose group (p < 0.01), and 24% versus 6% at 1 year (p < 0.05). Craving measured by VAS fell markedly in the high-dose group (mean VAS: 29.24 mm at baseline to 7.7 mm at 1 month and 5.4 mm at 3 months; p < 0.001). In a randomised comparison of single versus repeated ketamine-assisted psychotherapy (2 mg/kg IM given monthly), 50% of participants receiving multiple treatments were abstinent at 1 year versus 22% after a single session (p < 0.05), with larger craving reductions after repeated treatment. Jovaiša and colleagues randomised patients to saline or IV ketamine (0.5 mg/kg/h) prior to rapid opioid antagonist induction under general anaesthesia; ketamine suppressed physiological responses during induction (lower mean arterial pressure, heart rate and serum cortisol) but produced no between-group differences at 4 months on aftercare retention, abstinence or social outcomes, possibly influenced by both groups commencing opioid antagonist treatment and by administration during unconsciousness. A non-randomised comparison reported 1-year abstinence of 65.8% in a ketamine-assisted psychotherapy group versus 24% in a follow-up-as-usual group, but this study lacked randomisation and a concurrent control. Retrospective data from 23 patients given ketamine as an adjunct for severe alcohol withdrawal suggested a non-significant trend toward reduced benzodiazepine requirements at 12 and 24 hours (p = 0.11). Ongoing trials: Six registered trials were described. Three target alcohol use disorder: a 96-participant randomised placebo-controlled trial (NCT02649231) testing 0.8 mg/kg ketamine plus manualised relapse prevention versus saline plus education, with abstinence at 3 and 6 months as primary outcomes; a 40-participant trial (NCT02539511) assessing a single 0.71 mg/kg infusion plus 5 weeks of motivational enhancement therapy with self-reported alcohol use change at 5 weeks as the primary endpoint; and a 65-participant three-arm trial (NCT02461927) in persons with alcohol use disorder and major depressive disorder comparing weekly 0.5 mg/kg ketamine with or without IM naltrexone versus placebo over 4 weeks, with depression severity (MADRS) and complete abstinence at 4 weeks as primary outcomes. Dakwar is leading a 150-participant cocaine trial (NCT03344419) testing two 0.71 mg/kg infusions one month apart with adjunctive psychotherapy, and a 100-participant opioid detoxification trial (NCT03345173) is evaluating ketamine-facilitated initiation onto extended-release naltrexone using ketamine infusions during moderate withdrawal. A 15-participant proof-of-concept cannabis trial (NCT02946489) will test 0.71 mg/kg ketamine plus psychotherapy with abstinence at 6 weeks as an outcome.

Carrà and colleagues conclude that existing human studies collectively suggest ketamine may facilitate establishment and maintenance of abstinence across several SUDs, with consistent signals for reduced craving, increased motivation and decreased self-administration in cocaine studies and substantial long-term abstinence improvements reported after ketamine-assisted psychotherapy for alcohol and heroin in some trials. The opioid withdrawal literature includes data that ketamine can attenuate physiological responses during antagonist induction, and case series suggest possible benefit for withdrawal symptom reduction. The investigators emphasise the preliminary nature of these findings and enumerate important limitations. Key concerns include small sample sizes, short follow-up windows, and narrow or homogeneous participant samples in the cocaine trials. Several heroin and alcohol studies used a low-dose ketamine comparator rather than an inert placebo and did not fully control for adjunctive pharmacotherapies during follow-up; non-randomised designs in some reports further limit causal inference. The extraction also highlights differences in study populations—many heroin and alcohol studies recruited treatment-seeking individuals after extended residential programmes, whereas the cocaine trials involved non-treatment-seeking participants—raising uncertainty about generalisability. Unresolved questions identified by the authors include the role of baseline motivation or prior abstinence in moderating response, possible demographic or gender influences, optimal dosing and frequency (reported doses ranged from sub-anaesthetic 0.5–0.8 mg/kg IV to 2–2.5 mg/kg IM), and whether ketamine’s psychoactive (dissociative) effects mediate therapeutic benefit. The investigators argue for systematic assessment of psychoactive experiences in future trials. They also call for further study of ketamine’s effects on withdrawal states independent of general anaesthesia, greater evaluation of behavioural interventions as adjuvants, and exploration of alternative routes (for example, intranasal) that may broaden access. In sum, the authors position ketamine as a promising but still experimental adjunct in addiction treatment, and they recommend larger, rigorously controlled randomised trials with longer follow-up and careful assessment of concurrent treatments, mechanistic mediators and optimal dosing schedules to determine whether ketamine can meaningfully reduce addiction morbidity and mortality.