With that in mind, a large amount of work in my lab in the past 20 years — since I’ve been a PI — has been focused on studying this neuroendocrine pathways related to the gut-brain axis. For example, we have done work on insulin and leptin, primarily; we had done work on ghrelin, and since 2015 on the GLP-1 RAs.
With that in mind, the framework we are working on, which is also substantiated by many studies done by our team and other teams in the neuroscience field, kind of supports the idea that, similar to what we see in obesity, these medications may work by affecting what we call reward processing, or the seeking for addictive drugs, such as alcohol, and also the drugs such as the stimulants, opioids, nicotine, and so on.
The idea is that the mechanism is driven by the ability of the medication — semaglutide and all the GLP-1 RAs — to reduce the rewarding properties of alcohol and drugs. To maybe make the example more pragmatic, what does that mean? It means, for example, that a patient who typically has 10 drinks per day in the afternoon and night, while they are on the medication they may feel the lack of need to drink up to 10 to feel the same reward.
They may be able to stop after two or three drinks, which means a significant harm reduction and a beneficial outcome. This also brings us to another mechanism, which may be related to society. We don’t fully understand how much the society mechanism, including society mechanism related to GI motility, may also play a role.
With that said, we don’t think that the effect of the GLP-1 RAs is merely due to alcohol being a calorie-based nutrient because, in fact, we see alcohol as an addictive drug, not as a nutrient. Also, the GLP-1 RAs, at least in animal models, seem to work on other addictive drugs that don’t have calories, such as nicotine, and possibly with cannabis, opioids, and stimulants.
Then on the molecular level, our team recently showed, in collaboration with Dr. Marisa Roberto from Scripps in La Jolla, California, that semaglutide may in fact change the GABA transmission at the level of some brain regions, such as the amygdala and the prefrontal cortex. These are brain regions that are well-established hubs that play a role in the mechanism underlying addiction.
There are also some very exciting recent data showing how these medications may work not just on GABA or just on dopamine, which is the canonical way we conceive of reward processing, but by working on both by modulating GABA transmission — for example, at the ventral tegmental area and dopamine transmission at the nucleus accumbens.
Bottom line, if I summarize all of this, is that the mechanism is not fully understood, but there is definitely a contribution of this medication to effect and reward processing, possibly by altering the balance between GABA and dopamine. There are still some unknown questions, such as, are these mechanisms all brain driven or are they signaling from the periphery to the brain, or maybe both?
Also, as we all know, there are many differences across all these GLP-1 analogs in brain penetrance. Whether the drug needs to go to the brain to have an effect on alcohol drinking, cocaine seeking, or smoking is really an open question.
Dr. Jain: This is so thought-provoking. I guess the more we uncover, the more mesmerized we get with all the potential crosstalk. There is a large amount of overlap in the brain with each of these different things and how it all interplays with each other.
Speaking of interplay, I’m thinking about how many people prone to having alcohol use disorder can potentially develop complications, one of these being chronic pancreatitis. This is a well-known complication that can occur in people having alcohol addiction. Along that same line, we know that previous history of pancreatitis is considered a use-with-caution, or we don’t want to use GLP-1 RA therapy in people who have had pancreatitis.
Now it becomes this quagmire where people may have chronic pancreatitis, but we may want to consider GLP-1 RA therapy for management of alcohol use disorder. What are your thoughts about this, and the safety, potentially, in using it in these patients?
Dr. Leggio: This is another wonderful question. That’s definitely a top priority in our mind, to address these kinds of questions. For example, our RCT does have, as core primary outcomes, not only the efficacy defined as a reduction in alcohol drinking, but also safety.
The reason is exactly what you just explained. There are many unanswered questions, including whether giving a GLP-1 RA and alcohol together may have synergistic effects and increase the likelihood of having pancreatitis.
The good news is that, so far, based on the published literature, including the RCT done with exenatide in Denmark and published in 2022 and also the ongoing clinical trials — including my own clinical trial, but of course we are blind — pancreatitis has not been coming out as an adverse event.
However, it’s also true that it often happens in clinical medication development. Of course, we screen and select our population well. For example, we do exclude people who have a history of pancreatitis. We exclude people with high lipase or with any of the clinical symptomatology that makes us concerned about these people having pancreatitis.
As often happens when you move a medication from clinical trials to clinical practice, we still need to understand whether this medication works in patients. I’m just speculating, but even if the clinical trials do not raise red flags in terms of increased risk for some side effects such as pancreatitis, I think it will be very important for practitioners to keep a close eye on the death risk regardless.
It’s very interesting that it’s similar to alcohol liver disease. With pancreatitis, not every single patient with alcohol addiction has pancreatitis. We don’t really fully understand why some people develop pancreatitis and some people do not. The point being that there are many patients with alcohol addiction who don’t have pancreatitis and may benefit from these medications if they work. Again, we have to prove that in patients.
On the other side, as we all know, pancreatitis is a potentially life-threatening condition for those people who either have it or are at risk for it. I think we have to be very careful before we consider giving them a GLP-1 RA.
One could argue that alcohol is the leading cause of mortality and morbidity in the world. For example, right now, alcohol is the leading cause of liver disease. It’s the main reason for liver transplantation in our country. Alcohol is affecting thousands of people in terms of death and emergency room visits.
You could argue that the downside is not treating these people and they die because of alcohol addiction. A GLP-1 RA is not going to be for everybody. I will remind everybody that (1) we do have FDA-approved medications for alcohol addiction; and (2) there are also other medications not approved by the FDA, but with a proven efficacy in some clinical trials — for example, topiramate and gabapentin — and they’ve been endorsed by the American Psychiatric Association.
There is also some evidence for another medication, baclofen, which has been endorsed by the American College of Gastroenterology for patients with alcohol addiction and liver disease.
The point I’m making is that it’s not that either we use the GLP-1 RAs or we have no other tools. We have other tools. I think we have to personalize the treatment based on the patient’s profile from a safety standpoint and from a phenotypic standpoint.
Dr. Jain: I love that thought. I think individualization is the key here.
We know that people with diabetes have a higher risk for pancreatitis by virtue of having diabetes. People with obesity also have a higher risk for pancreatitis by virtue of having obesity. These are the two conditions where we are using a large amount of GLP-1 RA therapy. Again, the idea is looking at the person in front of us and then deciding, based on their past medical history and their current risk, whether or not a medication is a right fit for them.
I think more individualization here will come as we start using these medications that might be having potential effects on different organ systems. You mentioned a little bit about the liver, so a thought came in my mind. We know that people with diabetes who have alcohol use disorder are at a higher risk for potential hypoglycemia. If they have events when they have increased consumption of alcohol, there can be more hypoglycemia.
We now could potentially be using semaglutide or other GLP-1 RA therapy for management of alcohol use disorder. In your own experience in the studies that you’ve done or the literature that’s out there, has that been associated with an even higher risk for hypoglycemia?
Dr. Leggio: It’s a wonderful question. I’m not aware of any formal and published report of that association. That said, your thinking from a physiopathologist standpoint makes total sense. I could not agree more. The fact that nothing has been published, at least to my knowledge, doesn’t mean that the death risk doesn’t exist. In fact, I agree with you that it does exist.
Alcohol use disorder is interesting and tricky clinically because chronically, alcohol addiction or alcohol use disorder is associated with an increased risk for diabetes. Acutely, as you point out; and this could be with or without alcohol use disorder. An episode of a high volume of binge drinking may lead to hypoglycemia.
This is one of the reasons why people may show up to the emergency room with intoxication, and one of the symptoms detected at the emergency room is that they also have hypoglycemia in addition to vomiting, nausea, and everything else that we see in patients with acute intoxication.
Similar to the discussion about pancreatitis, as we work on understanding the possible role of GLP-1 RA in patients with alcohol use disorder, we do have to keep a close eye on the risk for hypoglycemia. The short answer is that this is not well established, but based on the simple concept of “first, do no harm,” I think we need to track that very carefully.
In the ongoing clinical trial we’re doing in Maryland in my program at the NIH, we do just that. We are tracking glucose levels. Of course, patients come to clinic weekly, so unless they have symptoms, typically we don’t see anything at the time.