A phase 2 trial published in 2026 in The Lancet Gastroenterology & Hepatology tested a combination of zalfermin and semaglutide in people with MASH and clinically significant fibrosis, and the headline is a study in how to read negative and positive results honestly. The combination did not beat placebo on the primary endpoint, but semaglutide alone showed a nominally significant benefit - a signal that is being read carefully, not celebrated.
The trial, led by Rohit Loomba and colleagues, enrolled 698 adults across 187 sites in 22 countries. Every participant had biopsy-confirmed MASH with fibrosis ranging from stage F2 up to compensated cirrhosis (F4c). That is the harder end of the disease - the group that most needs an effective therapy - and it is why the results are worth attention even when the combination failed.
The numbers, plainly stated
At week 52, 24% of participants on zalfermin 30 mg plus semaglutide 2.4 mg met the primary endpoint, versus 16% on placebo - a difference that was not statistically significant (p=0.19). Neither zalfermin 30 mg alone nor the lower-dose combinations moved meaningfully away from placebo. The contrast was semaglutide 2.4 mg alone: 30% met the primary endpoint versus 16% on placebo, a nominally significant difference at p=0.024.
The safety story was broadly consistent with what is known about GLP-1 medicines: adverse events were mostly mild to moderate and gastrointestinal, and serious adverse events were spread across groups. Five deaths occurred during the trial, and one - heart failure in a zalfermin group - was considered possibly related to the study drug. That kind of detail is why trials report more than just the endpoint.
What the trial does and does not settle
The honest read is twofold. First, adding zalfermin to semaglutide did not improve fibrosis outcomes beyond placebo in this population - a clear negative for that combination. Second, semaglutide alone produced a signal worth pursuing, and the authors conclude it warrants further assessment as a potential disease-modifying therapy even in compensated cirrhosis. The word “nominally” matters: this was not the pre-specified winning arm in a clean two-way test, so the signal is encouraging rather than definitive.
For a reader, the useful distinction is between what is settled and what is open. What is settled is that semaglutide is a prescription medicine with weight-loss and metabolic effects, and that its role in liver disease is a clinical question for a clinician. What is open is whether it will become a routine fibrosis therapy - that needs the larger, longer trials already underway.
Where diet fits around a medicine
This is the part this site exists to explain. A medicine like semaglutide does not replace the plate; it changes the context around it. The trial’s weight-loss and metabolic effects - appetite suppression, insulin sensitivity - work on the same insulin-resistance engine that the food pattern addresses from the other direction. When a clinician prescribes a GLP-1 medicine, the food guidance does not go away; it becomes the framework the medicine slots into.
The database gives the per-food version of that framework. The Recommended column - fish, legumes, whole grains, vegetables, olive oil, fruit - keeps saturated share low and fiber high, which supports the metabolic improvements a GLP-1 medicine is trying to achieve. Protein adequacy matters especially during weight loss on these medicines, and the database rates lean proteins like chicken breast, fish, and legumes in Recommended. The pattern and the prescription are complementary, not competing.
How our editorial team read this
Drug trials are where accuracy matters most, because a wrong read can travel fast. Our data editor verified the trial size, the endpoint, and the p-values against the published report, and we kept the “nominally significant” qualifier on semaglutide exactly as the authors did. We did not present the combination failure as a reason to distrust semaglutide, and we did not present the semaglutide signal as a green light for off-label use. Both would be the kind of overstatement this site is built to avoid.
What we did do is keep the food guidance steady. Nothing in this trial changes the Recommended column or the saturated-share rule; it adds to the medicine side of the conversation, which belongs entirely with a clinician. The reader-facing line is simple: the plate is the foundation, the medicine is a clinician’s decision, and the trial is one more data point in that decision.
The practical takeaway
For someone with MASH and fibrosis, this trial is a reason for cautious optimism about semaglutide, not a reason to change the kitchen. The pattern stays: Recommended column, minimal added sugar, regular activity, and a clinician who can interpret the trial evidence and the bloodwork together. The database covers the plate; the medicine question stays with the doctor.