Naltrexone-Bupropion for PCOS: The Trial Data and the Safety Conversation to Have First

What is naltrexone-bupropion, and how is its mechanism different from a GLP-1 drug?

Naltrexone-bupropion combines two drugs that already existed for other purposes — an opioid antagonist and an antidepressant — into a single tablet sold as Contrave in the US and Mysimba in the UK and EU, and its mechanism has nothing to do with the gut hormones a GLP-1 drug works on. Bupropion inhibits reuptake of dopamine and norepinephrine and acts on hypothalamic proopiomelanocortin (POMC) neurons, the same appetite-regulating circuit insulin and leptin signal into. Naltrexone blocks an opioid-receptor feedback loop that would otherwise dampen that POMC signal, so the combination is designed to sustain an appetite-suppressing effect that either drug alone loses within weeks. Semaglutide and tirzepatide work on a completely separate pathway — mimicking a gut hormone, slowing gastric emptying, and signaling fullness through a different hypothalamic route. Naltrexone-bupropion also has a documented effect on the brain’s mesolimbic reward circuit, the pathway underlying food-driven cravings rather than physical hunger, which is not something either GLP-1 drug is understood to target as directly.

That mechanistic difference is the actual reason this drug is worth a separate conversation rather than a footnote to the GLP-1 pages already on this site: someone whose eating pattern is driven more by reward-seeking than by hunger signaling, or who cannot tolerate the gastrointestinal slowdown a GLP-1 drug causes, is asking a genuinely different physiological question than someone considering semaglutide.

What do the pivotal trials actually show, and were any of them in PCOS?

Four large randomized trials, all sponsored by the drug’s manufacturer and all 56 weeks long, built the evidence base for naltrexone-bupropion — and none of the roughly 4,500 combined participants across them were selected for a PCOS diagnosis. The COR-I trial randomized 1,742 overweight or obese adults and found a mean weight loss of 6.1% on the full dose versus 1.3% on placebo, with 48% of the drug group losing 5% or more of body weight against 16% on placebo. COR-II , in 1,496 participants, found a nearly identical 6.4% versus 1.2% split at 56 weeks. COR-Diabetes enrolled 505 people specifically with type 2 diabetes and found smaller but still significant effects — 5.0% versus 1.8% weight loss, alongside an HbA1c reduction of 0.6 percentage points versus 0.1 on placebo. COR-BMOD added the drug on top of 28 sessions of intensive behavioral counseling in both arms and found the largest gap of the four: 9.3% versus 5.1% weight loss by week 56.

How does naltrexone-bupropion actually compare with a GLP-1 drug side by side?

The two drug classes are reasonable alternatives to raise in the same conversation precisely because they fail and succeed in different ways, not because one has replaced the other.

Who should not take naltrexone-bupropion — the contraindications that matter most here

Bupropion lowers the seizure threshold, and that risk is not theoretical: it is the reason the combination is absolutely contraindicated in anyone with a seizure disorder, and — specifically for this readership — in anyone with an eating disorder involving purging. The original controlled trial of bupropion in bulimia had to be stopped early after 4 of 55 participants (7.3%) experienced grand mal seizures, a rate the study’s own authors called far higher than anything seen with the drug in other populations — and that finding is the direct reason bupropion carries a bulimia/anorexia contraindication to this day. For comparison, a separate safety-surveillance study of the same sustained-release formulation in patients without an eating disorder or seizure history found a seizure rate of about 0.1% over a year at comparable doses — meaning the bulimia-population risk observed in the 1988 trial was on the order of 70 times higher. Eating disorders are more prevalent among people with PCOS than in the general population, which makes this specific contraindication unusually load-bearing for this readership rather than a rare edge case. The same seizure-threshold mechanism is why bupropion is also contraindicated during abrupt discontinuation of alcohol, benzodiazepines, or antiepileptic medication — withdrawal from any of those independently lowers the seizure threshold, and stacking that with bupropion compounds the risk.

Naltrexone is a full opioid-receptor antagonist, which creates two separate problems for anyone who might need an opioid. First, it blocks opioid pain relief outright — if you are taking naltrexone and are prescribed an opioid painkiller after surgery or an injury, it will not work as intended. Second, in anyone currently dependent on opioids, naltrexone precipitates acute, severe withdrawal rather than gently blocking a future dose. This makes both current opioid use and any anticipated need for opioid analgesia — a planned surgery, for instance — something to flag before starting, not after.

Blood pressure and heart rate are both monitored parameters on this drug, which matters directly given the cardiometabolic risk PCOS already carries. COR-I reported a transient increase of around 1.5 mm Hg in mean systolic and diastolic blood pressure in the weeks after starting, before settling to roughly 1 mm Hg below baseline by the end of the trial — a real early signal, not a lasting one, in that particular study. A dedicated cardiovascular outcomes trial in 8,910 overweight or obese participants with existing cardiovascular risk factors was designed to answer the larger safety question directly, but the trial’s own independent leadership recommended stopping it early after a confidential interim analysis was released without authorization — before the primary safety question could be fully answered. At the point of termination, major cardiovascular events had occurred in 2.0% of the drug group versus 2.3% on placebo, numerically reassuring, but the study’s published conclusion is explicit that “the cardiovascular safety of this treatment remains uncertain” because early stopping meant the planned statistical comparison was never completed. That is the honest state of the evidence — not a clean bill of health, and not a red flag either.

Pregnancy is a separate reason to avoid this drug rather than a dosing question: naltrexone-bupropion is not studied for safety in pregnancy, and it is not indicated for anyone who is pregnant or trying to conceive. That matters here specifically because many people reading a PCOS weight-loss article are actively trying to conceive, and unlike a GLP-1 drug — which at least has documented guidance about stopping before conception — naltrexone-bupropion has no comparable body of PCOS-specific fertility-adjacent data to draw on at all.

Bupropion carries an FDA boxed warning about increased risk of suicidal thoughts and behavior, a class-wide warning that applies to antidepressants generally and is not specific to this combination product or to the trials above. To state this without minimizing or overstating it: the pivotal weight-loss trials themselves did not find higher rates of depression or suicidality on the drug compared with placebo, and the boxed warning traces to a broader body of antidepressant-class data rather than a signal unique to naltrexone-bupropion in these specific trials. Both facts are true at once, and this is exactly the kind of thing worth naming directly for a prescriber rather than either dismissing or catastrophizing on your own.

What symptoms need same-day medical attention on this drug?

A seizure — any loss of consciousness with convulsive movement — is a medical emergency regardless of whether a risk factor was known beforehand, and it needs immediate emergency care, not a wait to see if it happens again. Sudden severe headache, chest pain, visual changes, or a resting heart rate that feels persistently and unusually fast are reasons to contact a prescriber the same day given the blood-pressure and heart-rate signal described above. Any new or worsening thoughts of self-harm, on this drug or any other, warrant contacting a prescriber or a crisis line immediately — this is the one symptom category on this page that should never wait for a scheduled appointment. And if you use opioids for any reason and start this medication, watch for the sudden onset of withdrawal symptoms — sweating, agitation, muscle aches, nausea — which signals the naltrexone component is doing exactly what it is designed to do, and needs urgent medical guidance on how to proceed safely.

Where this will not work for you

If you have any history of seizures, or any eating disorder involving purging, bingeing, or severe restriction, this drug class is off the table regardless of how well the weight-related numbers otherwise look — the seizure data above is not a relative risk to weigh against a benefit, it is an absolute contraindication. If you are pregnant, trying to conceive, or currently using opioids of any kind, the same applies. And if what you are actually looking for is a drug that works on gut hormones and gastric emptying — the mechanism behind semaglutide and tirzepatide — naltrexone-bupropion will not deliver that same physiological effect no matter the dose, because it is not acting on that pathway at all; it is a different tool for a different presentation, not a substitute for one.

You may see PCOS referred to as polyendocrine metabolic ovarian syndrome (PMOS), after a 2026 global consensus of more than 50 medical organizations renamed it. Nothing about the trial data or contraindications above changes under either name — this article uses PCOS because that is still what most readers search.

This is one entry in the full weight-loss guide, part of this site’s broader weight-loss coverage, which treats weight as a metabolic marker tied to insulin, androgens, and cardiometabolic risk — not a number a drug is meant to force down on its own.

Your next step

Bring four things to a prescribing conversation: your full seizure and eating-disorder history, your current medication list (specifically anything opioid, any MAOI, and any recent alcohol or benzodiazepine use), your current pregnancy plans, and a direct question — “Given the cardiovascular trial that was stopped early, what does the current safety picture actually say for someone with my risk factors?” That question gets you the most current, honest version of the LIGHT trial’s unresolved conclusion, rather than either a reflexive “it’s fine” or a reflexive “it’s dangerous.” If a reward-driven eating pattern rather than physical hunger is what’s driving your interest here, naming that specifically is more useful to a prescriber than describing the goal only as weight loss. And if the seizure or eating-disorder contraindications above rule this out entirely, phentermine is a different short-term option working through neither of this drug’s mechanisms, worth asking about instead.