GLP-1 receptor agonists are promising but unproven treatments for alcohol and substance use disorders

Abstract

Preclinical and initial human studies suggest that glucagon-like peptide-1 receptor agonists may be promising treatments for alcohol use disorder, but existing approved treatments should be used until safety and efficacy is demonstrated in clinical trials.

The development and rapid clinical adoption of potent and long-lasting glucagon-like peptide-1 receptor agonists (GLP-1RAs) is quickly changing the landscape of diabetes and obesity treatment. In particular, semaglutide (marketed as Ozempic, Wegovy and Rybelsus) has attracted attention among the general public for its remarkable effects on weight loss. The explosive growth of its use in clinical practice has been accompanied by an important ancillary finding: frequent anecdotal reports of reductions in alcohol use and other addictive behaviors. These observations, which we have heard ourselves from patients and providers, are being amplified by media and public attention and have led to discussions about potential clinical applications of semaglutide and other GLP-1RAs for addictive disorders, especially alcohol use disorder (AUD). We write as leaders of clinical trials currently underway in North America and Europe to evaluate the efficacy of semaglutide for AUD to encourage clinicians and patients to be prudent and avoid placing supposition before science.

Initial evidence from animal and human studies

Preclinical work on the effects of GLP-1RAs on the consumption of alcohol and other substances started more than a decade ago. Research on mice, rats and non-human primates carried out by independent laboratories and using several alcohol-related models (which reflect key phenotypic aspects of excessive alcohol consumption and AUD) has produced consistent evidence that GLP-1RAs reduce the rewarding properties of alcohol and other addictive drugs such as nicotine, opioids and cocaine, leading to reduced motivation for and intake of these drugs (reviewed in ref. 1). Recent studies by Chuong et al.² and Aranäs et al.³ in mice and rats independently provide support for an effect of semaglutide on reducing alcohol consumption and modulating other phenotypes and behaviors relevant to AUD.

Although research in humans has also suggested a relationship between alcohol consumption, AUD and the endogenous GLP-1 system^4–6, there is a lack of randomized controlled trials (RCTs). A recent study using information from nationwide registers of the Danish population reported that patients taking GLP-1RAs had lower incidence of alcohol-related events compared with those taking dipeptidyl peptidase IV inhibitors⁷.

The only RCT including patients with AUD showed that the first-generation GLP-1RA exenatide was not superior to placebo on the primary alcohol consumption outcomes in the full sample, although a secondary analysis showed that exenatide reduced alcohol use in a subgroup who had AUD and comorbid obesity⁸. Furthermore, activity induced by alcohol cues in brain areas related to reward processing was reduced in the exenatide group⁸. Compared with more recently developed GLP-1RAs such as semaglutide, exenatide has reduced sequence homology to endogenous GLP-1, and has been shown to have less of an effect on blood glucose and weight in diabetes and obesity studies. This inconclusive research in humans emphasizes the need for prospective RCTs with GLP-1RAs — including semaglutide — in patients with AUD. Such RCTs are underway, and human studies with GLP-1RAs are expanding to include patients with other substance use disorders^9,10.

Repurposing drugs

The surge in clinical and anecdotal reports of decreased alcohol use in people taking semaglutide for diabetes and obesity should not be ignored. AUD is a leading cause of mortality and morbidity¹¹, but compared with other chronic diseases such as cancers, hypertension, diabetes and depression, there is a very small number of US Food and Drug administration (FDA)-approved medications for AUD (acamprosate, disulfiram and naltrexone). These drugs also have low uptake (less than 2% in the USA¹²) due to several factors, including patients’ reluctance to seek treatment, lack of addiction training among healthcare providers and considerable stigma surrounding AUD and addiction¹¹.

Developing new medications is a difficult, expensive and time-consuming process; drug repurposing, therefore, has the potential to expedite much-needed treatment options for AUD. There are precedents for drug repurposing, including in addiction; for example, incidental reports of spontaneous reductions in cigarette smoking in patients taking bupropion for depression prompted the evaluation and eventual FDA approval of bupropion for smoking cessation. Repurposing requires practicing clinicians and clinical scientists conducting RCTs to listen to their patients, as they are a source of vital and unique information.

Forty percent of US adults have obesity and eleven percent of US adults met the diagnostic criteria for AUD in the past year. These conditions commonly co-occur and have shared genetic influences¹³. With the rapid increase in semaglutide prescriptions for weight loss, population-level uptake of semaglutide and other GLP-1RAs will vastly outstrip that of existing FDA-approved medications for AUD. This trend in semaglutide prescriptions is likely to expand even further, given the 2023 announcement of an RCT showing reduced risk of major cardiovascular events in adults who are overweight or obese and taking semaglutide. If the efficacy of GLP-1RAs for AUD is demonstrated in future RCTs, even without formal FDA approval for AUD, net reductions in alcohol consumption are still likely to have substantial health implications for diabetic and/or overweight populations, considering the sheer number of patients who are expected to be treated with these medications and the high prevalence of alcohol-related medical comorbidities, including liver disease, hypertension and cardiomyopathy.

Many healthcare providers outside the addiction field are prescribing and managing GLP-1RAs for patients with diabetes and/or obesity and are likely to feel comfortable expanding the use of these medications for other indications such as addiction. If proven effective, these medications have the potential to address the frustratingly low clinical utilization of approved medications for AUD.

The role of clinical trials

Semaglutide and other GLP-1RAs hold promise for treating AUD, but scientific rigor will be needed to assess their safety and efficacy in target populations. There are emerging anecdotal reports of patients seeking (and physicians prescribing) semaglutide with expectations of reduction in alcohol or substance use, despite the lack of controlled clinical data on semaglutide in patients with addiction. The practice of evidence-based medicine is critically important, and although the preclinical data and anecdotal clinical reports on semaglutide are promising, it is premature to prescribe off-label semaglutide to these patients. Instead, patients seeking help for AUD should be reminded that there are other effective and available psychosocial, behavioral and pharmacological treatments for AUD¹¹. Healthcare providers should offer these already approved and available treatments, including the FDA-approved acamprosate and naltrexone, both of which have shown clear efficacy for treating AUD¹⁴.

RCTs involving patients with AUD and other addictive disorders are now needed to assess the safety, tolerability and efficacy of the latest generation of GLP-1RAs, such as semaglutide. These RCTs are critically important for at least three reasons. First, they will provide insight on the safety of GLP-1RAs in patients with addictive disorders. AUD is often comorbid with metabolic disorders¹⁵, thus it is crucial to understand the effects of GLP-1RAs on the metabolic profile of patients with AUD. Some GLP-1RAs promote more weight loss than others, and so may be contraindicated for patients with normal or low weight. For example, special attention may be needed in patients with stimulant use disorders, given the known anorexigenic effects of psychostimulants.

Second, as discussed above, despite emerging results on the potential role of semaglutide and other GLP-1RAs in treating addictive disorders, prospective RCTs are needed to assess their potential efficacy. Although anecdotal reports that a medication could be repurposed for a new indication, rigorous clinical trials may subsequently fail to demonstrate efficacy.

Third, it is unlikely that a medication works for all patients, which is exactly why precision medicine is so important, and addiction medicine is not an exception¹¹. Even if RCTs support the efficacy of semaglutide and/or other GLP-1RAs in patients with addictive disorders, it will be critical to identify subgroups who are more or less likely to tolerate and respond to GLP-1RAs. The circumstances under which semaglutide might prove safe, tolerable and efficacious in these patients remain unclear and could be affected by weight range, dosage, treatment duration, potential rebound effects upon treatment cessation, addiction severity, and other factors.

Health equity issues

The emergence and growing popularity of GLP-1RAs such as semaglutide, coupled with the increased attention of the media and the public, is increasing awareness that they may be a potential treatment for AUD and other addictions. Although we believe that off-label use of GLP-1RAs for AUD is premature, the increased attention to this topic may have the benefit of increasing awareness of AUD, its impact on health, and the existence and availability of approved medications to treat AUD.

Clinical decisions should follow from controlled studies, triangulated with other clinical and preclinical evidence. Well-controlled RCTs with scientific rigor are needed before any claim of efficacy for AUD and other substance use disorders can be made. Clinical trials will assess the tolerability, safety and potential subgroups who may best respond to these treatments.

The high costs of newer GLP-1RAs pose barriers to treatment access, and so it will also be important to address the health equity issues that have already arisen in the implementation of GLP-1RAs for diabetes and obesity, and that may pose similar (if not greater) challenges for patients with alcohol and substance use disorders.