Skip to content
This GLP
← Research
Evidence

Seven arms, and the only one that worked was semaglutide by itself

Three combination doses, a second combination, and a single agent. The combinations all failed to beat placebo and the plain drug did not.

Dana Sullivan7 min read
Fibrosis improved, no worsening, at week 52semaglutide alone30%zalfermin 30 + semaglutide24%zalfermin 30 alone22%placebo16%Adding a second drug did not beat semaglutide by itself.

Combination therapy is the direction this field keeps pointing, and the CagriSema evidence on weight is covered on this site’s sister publication. This trial put several combinations against their own components in liver disease, and the result runs the other way[1].

The primary endpoint was fibrosis improving by at least one stage without the underlying steatohepatitis worsening, at 52 weeks. Zalfermin 30 mg plus semaglutide achieved it in 24% against 16% on placebo — an estimated difference of 7.98 percentage points with a confidence interval from -3.82 to 19.79 and a p value of 0.19. That is the trial’s answer, and it is negative — the shape of result that rarely travels, unlike the positive findings catalogued in the liver disease evidence.

That said, the semaglutide figure needs its qualifier kept. The authors describe it as nominally significant, which means it was not the trial’s protected primary comparison and carries no multiplicity correction. In a seven-arm trial, one arm clearing p = 0.024 outside the primary hierarchy is a finding to test, not a result to bank — the same distinction that separates prespecified from post hoc analysis in the HFpEF sex analysis.

The screening funnel is worth noting for anyone reading trial results in this disease. Of 2,420 people screened, 178 withdrew and 1,544 were disqualified, leaving 698 randomized. Biopsy-confirmed MASH trials select very heavily, which is part of why their populations differ from the people actually prescribed these drugs — the gap examined in the approval and the fibrosis stage.

Adverse events were mostly non-serious and mild to moderate, dominated by gastrointestinal complaints, which is the familiar profile. What the trial contributes is a caution against assuming that stacking mechanisms stacks benefit — a question that also sits behind the liver mediation analysis and the myostatin blocker trial, where a second drug was added for a different purpose.

Frequently asked

Did the combination work?
No. Zalfermin 30 mg plus semaglutide improved fibrosis in 24% against 16% on placebo, with a confidence interval crossing zero and a p value of 0.19.
What about semaglutide on its own?
It was the only arm to separate from placebo, at 30% against 16%. The authors describe that as nominally significant, meaning it sits outside the trial's protected primary comparison.
Does this affect CagriSema?
An exploratory arm combining cagrilintide with semaglutide also failed to differ substantially from placebo on this liver endpoint.

Sources

  1. [1] Loomba R, George J, Castera L, Francque S, Lawitz E, Shoeb A (2026). Efficacy and safety of zalfermin co-administered with semaglutide in participants with fibrosis and cirrhosis due to metabolic dysfunction-associated steatohepatitis: a phase 2, dose-ranging, double-blind, randomised controlled trial The Lancet Gastroenterology & Hepatology. PMID 42456707

Where to get it

Best GLP-1 injections

Every injectable seller we can verify, with the price each one publishes and an honest read of what the trials measured.

Compare providers →

More in Evidence