Two oral GLP-1 drugs were dropped after liver concerns.
What does that mean for the next one? At first glance, the two failures point to the same risk. One drug raised liver enzymes across three studies. The other was stopped after a single potential drug-related injury. The details were scattered across trial updates, later papers and FDA reviews. We pulled them together to show how much weight each failure deserves when investors assess the next drug.
The headline hid an important difference.
Stop at “two oral GLP-1s, two liver-related failures” and the takeaway looks obvious. Read the underlying studies and it becomes narrower: one drug showed a repeated signal across three studies; the other was stopped after one potential injury. The public record has not established that they failed for the same reason.
What happened in each program
Lotiglipron: a pattern repeated across studies
Lotiglipron was stopped in 2023 after liver-enzyme elevations appeared in two early drug-interaction studies and a Phase 2 trial. The public update reported no liver-related symptoms, liver failure or need for treatment. The repeated laboratory pattern was still sufficient to end development. June 2023 program update ↗
Liver-enzyme elevations in the interaction studies
View figure data
| Study | Participants with elevations | Observed proportion | Calculated 95% CI |
|---|---|---|---|
| Study 1, Cohort 1 | 5 / 16 | 31.3% | 14.2–55.6% |
| Study 2 | 3 / 16 | 18.8% | 6.6–43.0% |
Danuglipron: the record changed over time
Danuglipron first faced a tolerability problem. Its twice-daily Phase 2b trial met the weight-loss endpoint, but discontinuation exceeded 50% across doses, compared with approximately 40% on placebo. Those figures covered all reasons for stopping and cannot be read as liver-event rates. Development continued with a once-daily formulation. December 2023 trial update ↗
A July 2024 update reported no liver-enzyme elevations among more than 1,400 participants studied to that point. In April 2025, a later dose-optimization study produced one potential drug-induced liver injury. It resolved after treatment stopped, and development ended after review of the full dataset and regulator input. The participant’s drug levels and the specific regulatory concern were not disclosed. July 2024 formulation update ↗ · April 2025 program update ↗
The category label identified a safety question. The program histories determined how much weight the question deserved for another asset.
What investors learned later
A 2026 lotiglipron publication reported that concentrations of several drugs rose after the liver-enzyme elevations, with no substantial differences before them. The authors linked the pattern to reduced liver transport or metabolism. The chronology helps explain the interaction concern while leaving the initiating cause unresolved. Clinical pharmacology publication ↗
Exposure ratios when liver enzymes were elevated
View figure data
| Reported group | Peak-concentration ratio | Type of value |
|---|---|---|
| Without enzyme elevations | 0.15–0.41 | Reported range |
| With enzyme elevations | 2.58 | Individual reported value |
| With enzyme elevations | 14.83 | Individual reported value |
A separate oral GLP-1 program, TERN-601, added a related chemical design to the record. Three participants had concerning liver findings: two were consistent with drug-induced liver injury, while a third had a possible alternative cause. The authors identified structural features shared with lotiglipron and danuglipron. Those features form a specific hypothesis for investigation; their causal role remains unproven. The findings also began at week eight or later, beyond the program’s earlier 28-day repeated-dose study. TERN-601 Phase 2 publication ↗
Orforglipron supplied a later-stage comparison. FDA liver specialists reviewed three Phase 3 trials with more than 3,000 treated participants and found no probable idiosyncratic liver injury. They considered structural differences and lower dose plausible contributors to the more reassuring profile while retaining concern about possible indirect injury through gallstone disease. Approval in April 2026 included required additional safety data and a request for five years of enhanced liver-injury reporting. Those actions applied to that product. FDA hepatology assessment ↗ · FDA approval letter ↗
Aleniglipron offered an earlier-stage comparison. Its 36-week Phase 2b study reported no drug-induced liver injury or persistent liver-enzyme elevations, although temporary elevations occurred. That finding applies within the study’s size and duration. Phase 2b publication ↗
What this means for the next drug
The two failures should not be carried forward as one class-wide verdict. The record supports three separate conclusions, each with a different consequence for the investment case.
| Question | What the record supports | How to use it |
|---|---|---|
| Do the two failures predict the outcome of another oral GLP-1 drug? | Being in the same category is not enough. | Start with the new drug’s own clinical record and keep its remaining liver risks explicit. |
| Should investors worry about a drug with similar chemical features? | Yes, when it shares the feature tied to the concern. Whether that feature caused the injuries is still unproven. | Check the chemistry, drug exposure over time, patient population and treatment regimen. |
| Did the discontinued drugs fail for the same reason? | The public record does not establish a shared cause. | Keep the proposed explanation out of the base case until stronger evidence supports it. |
Evidence establishing a causal feature shared with the asset would increase the weight of the precedent. New, clinically significant liver findings under the asset’s intended regimen would also reopen the assessment. Until then, the review can preserve the unresolved mechanism without presenting it as the explanation.
This case does not supply a numerical probability of success or a valuation adjustment. It shows the work required before an earlier failure becomes an assumption in either one.
This is a retrospective analysis of the linked public sources, including evidence published after the original April 2025 cutoff. It is not an exhaustive safety review, clinical advice or a recommendation concerning a security.