Eliza Ward: Hey — good to be back.
Brian Reed: Yeah, you too. What are we into today?
Eliza Ward: BBC just reported on a clinical trial — the CURB trial — testing whether semaglutide, the drug in Wegovy and Ozempic, can treat alcohol dependency. Two point seven million pounds, NIHR-funded, already running.
Brian Reed: Wait — Ozempic. The weight-loss drug.
Eliza Ward: The very same. Led by Prof Guru Aithal. Patients need a BMI of at least 30, a confirmed alcohol use disorder, and — this is the part that stopped me — evidence of liver scarring or cirrhosis.
Brian Reed: So they're targeting people who are already seriously ill. That's not a preventive trial, that's... hang on, what's the scale of the problem they're trying to fix here?
Eliza Ward: Alcohol misuse costs the NHS an estimated five billion pounds a year. That's the number that makes this more than a scientific curiosity — that's why the NIHR is writing a cheque.
Brian Reed: Five billion. And they're testing a drug that people already associate entirely with dropping dress sizes.
Eliza Ward: Right, and that's what makes the actual evidence interesting — because it isn't just one small trial someone ran in a basement. The phase 2 data: 48 adults, non-treatment-seeking, 71% female, mean age under 40. Semaglutide reduced both craving and consumption. Controlled setting.
Brian Reed: Non-treatment-seeking — meaning they weren't even trying to quit.
Eliza Ward: Exactly. Which is why the Lancet trial matters — that one extended the signal into people who actually showed up wanting help, with comorbid obesity. Randomised, double-blind, placebo-controlled.
Brian Reed: So what's the plain-language version of what the drug is actually doing? Because I keep seeing 'reward pathways' and — I mean, that phrase does a lot of heavy lifting without explaining much.
Eliza Ward: Dr Christian Hendershot's framing is the cleanest I've seen. Think of your brain's reward system like a volume knob. Semaglutide may be quietly turning that knob down — across the board. So the pull toward a drink feels quieter. Same way the pull toward a second slice of cake does.
Brian Reed: It's not killing the craving, it's just... lowering the signal.
Eliza Ward: That's the hypothesis. And the scale of the supporting data is actually — wait, this is the part that surprised me. A study of roughly 600,000 people suggesting GLP-1 drugs reduce problematic use across every major substance type. Simultaneously.
Brian Reed: Every major substance — so not just alcohol. And the animal data tracks that too, right? The vervet monkeys.
Eliza Ward: Rodents and green vervet monkeys — reduced alcohol intake, less cocaine self-administration, less nicotine interest. No apparent nausea in the primates. But here's where Hendershot's framing runs into a wall — researchers still haven't settled whether it's the metabolic pathway doing this or the brain-reward pathway. And that gap actually matters a lot for who it works for long-term.
Brian Reed: Right, but — that gap is actually the thing the optimistic take completely skips over. Because exenatide. Another GLP-1 receptor agonist, tested in AUD treatment-seekers, found no overall effect. Zero. And nobody's explaining why semaglutide would be different.
Eliza Ward: Hold on — that's the crack. Exenatide failing in treatment-seekers is the fact that should be in every headline right now and it's just... not.
Brian Reed: The take circulating is basically: Lancet trial plus 600,000-person study equals problem solved. But wait — the phase 2 trial was non-treatment-seekers. The Lancet trial filtered for comorbid obesity. Those are not the same people walking into an NHS clinic.
Eliza Ward: Think of it this way — imagine a 44-year-old woman, not trying to quit, just enrolled in a study. She gets the drug, craving drops, great. Now imagine someone who drove herself to a clinic on a Tuesday morning because she's scared. Different motivation, different stress load, different brain state when semaglutide hits. Exenatide didn't bridge that gap. We don't know yet if semaglutide does.
Brian Reed: No, I don't buy that the populations are close enough to just assume transfer.
Eliza Ward: And the scale of remaining uncertainty — there are more than a dozen additional GLP-1 addiction trials underway or enrolling right now. That's not a field that's reached consensus. That's a field that is still genuinely asking the question.
Brian Reed: Dr Ziyad Al-Aly put it well — no existing medication works across all addictive substances. So the cross-substance signal from the 600,000-person study is actually novel. But novel isn't the same as clinically proven, and I think that distinction is getting lost.
Eliza Ward: Which is exactly why the CURB trial exists. If the evidence already supported a treatment recommendation, you don't spend £2.7 million and years of NHS patient time running the trial — you just prescribe.
Brian Reed: And even if CURB lands cleanly — the part that I think makes this more complicated, we'll get to it — is whether semaglutide would actually reach patients when naltrexone's been approved since 1994 and fewer than 2% of people with AUD ever touch it.
Eliza Ward: That number — two percent — that's the thing. Disulfiram has been FDA-approved since the 1940s. Naltrexone oral since 1994. Vivitrol, the injectable version, 2006. Acamprosate, 2004. Four decades of available tools and the system still doesn't use them. So what exactly does a fifth drug fix?
Brian Reed: That's — yeah, that's the part that won't let go. Because the bottleneck isn't the molecule. The bottleneck is that a GP, or even an addiction specialist, looks at a patient and still doesn't reach for naltrexone first. Why would they suddenly reach for semaglutide?
Eliza Ward: Wait — there's one thing the semaglutide story has that naltrexone never had. The patient signal came first. GLP-1 users were already reporting reduced desire to drink before any trial was designed. That bypassed the prescribing gatekeepers entirely. Patients didn't wait for doctors to offer it.
Brian Reed: Hm — so the demand is coming from below rather than above. That's actually different.
Eliza Ward: It is different. But the reimbursement question is — I mean, that's where it collapses back to the same problem. The NIHR and Prof Guru Aithal need CURB to show cost-effectiveness, not just clinical effect. The NHS codes treatments by indication. Wegovy is coded for weight. Ozempic is coded for diabetes. Novo Nordisk would be asking the system to pay for a weight-loss drug inside addiction medicine — a speciality that has never reliably prescribed the drugs it already owns.
Brian Reed: And Eli Lilly just launched Foundayo — oral weight-loss pill, UK, August 23rd 2026. So Novo Nordisk is already feeling commercial pressure on the weight indication. A positive CURB result would be a lifeline, a new market. But that cuts both ways — does the company actually want semaglutide coded as an addiction treatment? That's a different sales infrastructure, different payer conversations.
Eliza Ward: Right — and addiction medicine specialists might just deflect it back. 'That's a metabolic drug, talk to endocrinology.' Which is exactly what happened with naltrexone in some pain clinics. Nobody claimed the territory.
Brian Reed: So the concrete things to actually watch — it's not CURB's outcome, it's what happens the day after a positive result. Does NIHR push an NHS reimbursement code for semaglutide in AUD patients specifically? Do addiction medicine teams pick it up or say it's not their lane? Those are the questions that determine whether the 98% stays at 98%.
Eliza Ward: The signal to watch is the coding decision. Not the trial result.
Brian Reed: And the people who actually prompted all of this research — the ones who just noticed, on their own, that they wanted to drink less while on Wegovy — they didn't wait for a coding decision. They didn't wait for a prescriber. That's what stands out.
Eliza Ward: Which is — I mean, that's genuinely new. But it doesn't transfer automatically. The CURB trial still has to answer whether semaglutide works for NHS patients who have AUD and liver scarring and obesity all at once. That's a different clinical reality than someone who just noticed a quieter pull toward wine.
Brian Reed: Right — and if CURB succeeds, the question isn't pharmacology anymore. It's actually: is this addiction medicine or metabolic medicine? Because the NHS has to put it somewhere.
Eliza Ward: That's the live question. And we don't know. A system that couldn't move naltrexone — approved drug, three decades on the shelf — to more than 2% of AUD patients... I'm not sure a positive trial result fixes that. I genuinely don't know what does.
Brian Reed: Neither do I. Let's leave it there.
