GLP-1 And Fatty Liver: The Placebo Column

MASLD And MASH Are Not The Same Thing

The naming changed recently, which has made an already confusing area worse. The current terms:

• MASLD (metabolic dysfunction-associated steatotic liver disease, formerly NAFLD) — fat accumulated in the liver. Very common. Often found incidentally on an ultrasound done for something else.
• MASH (metabolic dysfunction-associated steatohepatitis, formerly NASH) — fat plus inflammation and liver cell injury. This is the version that can progress to fibrosis, cirrhosis and liver failure.

The distinction is not academic. It decides whether the trial below has anything to do with you. MASH is diagnosed by biopsy, or increasingly by validated non-invasive tests — not by a radiographer writing “fatty liver” on an ultrasound report.

Fibrosis is then staged 0 to 4, where 4 is cirrhosis. The trial enrolled stage 2 and 3 only.

What The ESSENCE Trial Found

ESSENCE is a phase 3, multicentre, randomised, double-blind, placebo-controlled trial. 1,197 patients with biopsy-defined MASH and fibrosis stage 2 or 3 were assigned 2:1 to weekly semaglutide 2.4 mg or placebo, planned to run 240 weeks. What has been published is the week-72 interim analysis of the first 800 patients.

The two primary endpoints, both assessed on repeat biopsy:

• Resolution of steatohepatitis without worsening fibrosis: 62.9% on semaglutide vs 34.3% on placebo. Difference 28.7 percentage points (95% CI 21.1 to 36.2), p<0.001.
• Reduction in fibrosis without worsening steatohepatitis: 36.8% vs 22.4%. Difference 14.4 points (95% CI 7.5 to 21.3), p<0.001.

Both together occurred in 32.7% vs 16.1%. Mean weight change was −10.5% vs −2.0%.

Take the design seriously for a moment. This is histology — actual liver tissue read under a microscope, before and after, with the pathologist blinded. That is a substantially harder endpoint than the liver enzyme panels and imaging scores most claims in this area rest on. The effect is real and it is large.

A Third Of The Placebo Group Also Got Better

34.3%. More than one in three people who received a placebo injection for 72 weeks had resolution of steatohepatitis on repeat biopsy.

That is an extraordinary control-group response, and essentially no popular coverage of this trial mentions it. It is worth sitting with, because it tells you several things at once:

• MASH is genuinely reversible. Not in everybody, not reliably — but a third of people got there without the drug.
• Trial participation is itself an intervention. People in the placebo arm were seen regularly, monitored, and given lifestyle advice. They lost 2% of body weight on placebo injections.
• The drug is not the only route to the outcome. It is a faster and more reliable one in this population — 62.9% against 34.3% is a real and large advantage — but the destination is reachable by other means.

Read the same way as any other trial: the drug roughly doubled the resolution rate. It did not create a possibility that did not otherwise exist.

Who The Trial Was Run In

Same discipline as with any trial result. Entry required:

• Biopsy-defined MASH — confirmed on tissue, not suspected on imaging
• Fibrosis stage 2 or 3 — moderate to advanced scarring, but not cirrhosis

So this is a trial in people with established, confirmed, moderately advanced liver disease. Which is exactly right: those are the people for whom the question matters most, and their event rate makes the study feasible.

It also means the trial says nothing directly about the far larger group with simple fatty liver and no inflammation, no fibrosis, and no biopsy. That group is not evidence-free — weight loss is well established to reduce liver fat regardless of method — but ESSENCE is not their trial.

Stage 4 — cirrhosis — was also excluded. Results here should not be read as applying to established cirrhosis.

Histology Is Not Yet An Outcome

The endpoints in ESSENCE are histological: what the tissue looks like. The endpoints that ultimately matter are clinical: progression to cirrhosis, liver failure, liver cancer, transplant, death.

Improved histology is a well-accepted surrogate and a regulatory pathway for this disease, and it is far closer to the real outcome than a liver enzyme reading. But it is still a surrogate. The trial is planned to run 240 weeks; what has been reported is week 72 of that. The long-term clinical outcomes are, at the time of writing, not yet published.

The other open question is the familiar one: mean weight loss was 10.5% against 2.0%, and weight loss by any route improves liver histology. ESSENCE has no arm that lost equivalent weight without the drug, so it cannot separate a liver-specific effect from the effect of losing a tenth of body weight. The same limitation applies to the cardiovascular data.

Gastrointestinal adverse events were more common on semaglutide, consistent with the rest of the trial programme — covered in our side effects guide.

If Your Scan Said Fatty Liver

This is the most common way readers arrive at this topic, so it is worth addressing directly.

An incidental note of hepatic steatosis on an ultrasound is a finding, not a diagnosis of MASH. The appropriate next step is a conversation with the doctor who ordered the scan, who can decide whether fibrosis assessment is warranted — there are validated non-invasive scores and elastography for exactly this, well short of a biopsy.

What is not in doubt, and does not require any of the above: weight loss reduces liver fat, and it does so whatever produces the weight loss. That is the part of this that has been established far longer than any of the drugs.

If you are already on a GLP-1 for weight, the liver effect is a plausible additional benefit rather than a reason to start. And it does not change the problem the rest of this silo is about: roughly a quarter to a third of what you lose is lean mass unless you do something specific about it — which is the part you can control.