The drugs that made semaglutide ↗ and tirzepatide ↗ household names are being studied hard as liver treatments. This class, drugs that mimic the gut hormone GLP-1, lowers fat in the liver, and the newer dual and triple agonists are in trials for the fatty-liver disease now called MASH. So it is worth pausing on a review that points the other way. A paper in Clinical Therapeutics ↗ collects 31 published case reports of suspected liver injury from these same drugs, alongside 232 reports of drug-induced liver injury filed to the FDA's adverse-event reporting system (FAERS), and notes that three GLP-1 drug programs have been halted over liver safety.
None of this is on the label. The review's central observation is that the liver has not been named as an organ of safety concern for the class. The prescribing information for these drugs does not flag it, even as the case literature builds. That gap, between a drug developed partly to help the liver and a small but real set of reports that it can injure the liver, is the whole story.
The cases have a consistent shape. Patients ranged in age from 17 to 79, with more women than men. The injury usually announced itself as abdominal pain, nausea, and vomiting, which are also the drugs' most common and expected side effects. That overlap is part of why it can be missed, because the liver signal hides inside the gut signal. Onset ran from about two weeks to six months after starting, and often followed a dose increase or a switch from one drug to another. The pattern was mostly hepatocellular, meaning the injury showed up as damage to liver cells rather than to the bile ducts, and most cases were graded severe. A minority progressed to acute liver failure or decompensation, the point at which the liver can no longer keep up. The reassuring part is that most cases resolved once the drug was stopped.
Numbers this small come with heavy caveats, and the review is candid about them. Case reports and voluntary adverse-event filings establish that something can happen, not how often. The 232 FAERS reports, set against the very large number of people now taking these drugs, describe a rate that could be genuinely tiny, and both sources are shaped by who chooses to report. The people flagged as higher risk, meaning women, those with existing fatty-liver disease, and those losing weight fast or escalating dose quickly, also describe much of the typical user base, which makes any single risk factor hard to isolate. The authors are clear that the true incidence is unknown and probably rare, and that a reported association is not proof the drug caused the injury.
Set against the development story, the signal is easy to misread in either direction. These drugs genuinely lower liver fat, and one glucagon and GLP-1 agonist, survodutide ↗, has posted real results against liver scarring (fibrosis) in MASH. A rare injury signal does not overturn that. But "protective for the liver on average" and "never harmful to any patient's liver" are different claims, and the label currently supports the first while staying silent on the second.
The practical takeaway the review lands on is modest and sensible. When a patient on one of these drugs develops abdominal pain, nausea, or vomiting, the reflex is to call it the expected gastrointestinal effect and wait it out. The case series is a reminder to check liver enzymes before assuming, particularly after a dose increase, and to treat unexplained, severe, or persistent symptoms as a reason to look rather than a reason to reassure.