Zenagamtide Worked as a Pill and a Weekly Injection. The Trade-Off Was Tolerability
Two Lancet phase 2 studies tested zenagamtide as a daily pill and weekly injection. Both lowered HbA1c, but high-dose tolerability complicates the headline.
One peptide, two routes, and a useful reality check
Zenagamtide just produced something the obesity-drug pipeline rarely gives us: two peer-reviewed phase 2 studies from the same program, published side by side, testing the same experimental peptide as a daily tablet and a weekly injection. Both versions lowered blood sugar in adults with type 2 diabetes. The injection produced the stronger headline numbers. The tablet showed that the molecule can work orally. Neither result makes zenagamtide an approved treatment, and the discontinuation pattern keeps the story from being a simple victory lap.
The studies appeared online in The Lancet on July 30, 2026. Zenagamtide, previously called amycretin, is a single peptide designed to activate GLP-1, amylin, and calcitonin receptors. That matters because it packages complementary appetite and glucose signals into one molecule rather than combining two separate drugs. GLP-1 signaling can improve glucose-dependent insulin secretion and reduce appetite. Amylin signaling helps bring meals to an end, slows gastric emptying, and suppresses glucagon. The calcitonin-receptor activity is part of zenagamtide’s pharmacology as reported by the investigators, although its clinical contribution is not yet neatly separable.
The interesting question is no longer whether the concept can move a biomarker. It can. The harder question is whether enough patients can stay on a dose that delivers the benefit, and whether the oral route can eventually offer convenience without giving away too much efficacy.
What the phase 2 program actually tested
NCT06542874 was a 36-week, randomized, placebo-controlled, dose-finding study run at 83 hospital and clinic sites across 11 countries. Participants were 18 to 75 years old, had type 2 diabetes with an HbA1c between 7.0% and 10.0%, and were taking stable metformin with or without an SGLT2 inhibitor. The registry lists 448 participants across the full oral and subcutaneous program. The two Lancet papers report the routes separately.
In the oral study, 186 participants were randomized to once-daily zenagamtide at maintenance doses of 6, 25, or 50 mg, or to matched placebo. All zenagamtide groups began at 1.5 mg and escalated every four weeks. In the injection study, 262 participants were randomized to once-weekly doses ranging from 0.4 to 40 mg or placebo. The injected groups began at 0.2 mg and also escalated every four weeks.
The primary endpoint in both studies was the change in HbA1c at week 36. This was a diabetes trial, not a head-to-head obesity trial. Weight change was an important secondary outcome, but the studies were not designed to tell us that zenagamtide beats semaglutide, tirzepatide, or CagriSema. There was no active comparator.
The pill lowered HbA1c, but the dose response was not dramatic
Starting HbA1c averaged roughly 7.9% to 8.1% across the oral groups. At week 36, estimated HbA1c fell by 0.9 percentage points with 6 mg, 1.3 points with 25 mg, and 1.4 points with 50 mg. Compared with placebo, the treatment differences were 0.5, 0.99, and 1.09 percentage points, respectively. Each comparison met statistical significance; the 6 mg result was the least certain, with a 95% confidence interval that ran from a 1.04-point advantage to a 0.04-point advantage.
That is clinically relevant glucose lowering, especially for an oral peptide. It is not evidence that a pill is equivalent to the weekly injection. Cross-trial comparisons are imperfect even inside a shared program, and the oral abstract does not establish a formal non-inferiority comparison between routes.
The oral route is scientifically important because peptides are usually degraded in the digestive tract and cross the intestinal wall poorly. Getting a complex peptide to work as a tablet is a formulation achievement. For patients, though, route convenience depends on more than avoiding a needle. Daily dosing conditions, food restrictions, titration, gastrointestinal symptoms, manufacturing cost, and real-world adherence will determine whether an oral product is genuinely easier to use. Those questions are not answered by this 36-week study.
The weekly injection delivered the larger efficacy signal
The subcutaneous study produced a clearer dose-related range. From a mean starting HbA1c of 7.8%, estimated reductions ran from 0.9 percentage points at 0.4 mg to 1.7 points at 40 mg. The highest-dose difference versus placebo was 1.56 percentage points, with a 95% confidence interval from 1.07 to 2.05 points and a p value below 0.0001.
Weight supplied the rehook. At 40 mg, mean body weight fell 14.6% by week 36, compared with 2.1% on placebo. The company’s ADA poster reported no apparent plateau at the higher doses. It also noted an important design detail: participants assigned to 20 mg received that maintenance dose for only eight weeks, while those assigned to 40 mg received the full maintenance dose for only four weeks because of the long escalation schedule.
That detail cuts both ways. It suggests weight loss might have continued with longer treatment. It also means the highest dose had very little time to demonstrate stable, long-term tolerability. A curve that has not plateaued is not a forecast. Phase 3 needs to show what happens when more people remain at therapeutic doses for much longer.
The tolerability cost is the part worth watching
Gastrointestinal events were the most common adverse effects with both formulations. In the oral study, gastrointestinal events occurred in 26% of participants at 6 mg, 41% at 25 mg, 47% at 50 mg, and 23% with placebo. Seven participants across the active oral groups reported serious adverse events; none were reported in the oral placebo group. No deaths occurred.
The injection poster adds more texture than the abstract. Nausea affected 50% of participants at 20 mg and 37% at 40 mg. Vomiting occurred in 24% and 29%, while diarrhea occurred in 16% and 37%. Across the injected study, 16.8% of all participants discontinued treatment because of adverse events, mainly nausea and vomiting. Treatment withdrawal reached 34% in each of the 20 mg and 40 mg groups, compared with 6% in pooled placebo.
Those are not footnote numbers. A dose can look excellent in an efficacy chart and still be difficult to use if one in three participants assigned to the highest levels stops treatment. The trial used fixed escalation: when a planned dose was not tolerated, treatment was permanently discontinued rather than maintained at a lower personalized dose. That design may exaggerate discontinuation relative to flexible clinical practice, but it also exposes how demanding the chosen escalation schedule was.
Serious adverse events in the injected study were distributed across active and placebo groups, with 21 of 261 treated participants reporting them and no deaths. The small dose groups were not large enough to characterize uncommon risks. Phase 2 can reveal frequent nausea; it cannot rule out rare safety problems.
What the study can and cannot tell us
The strengths are real: randomized allocation, placebo control, masking within dose levels, international enrollment, prespecified endpoints, and publication of both routes in a major peer-reviewed journal. The results are also linked to a completed public trial record, which makes the design easier to audit.
The limits are equally important. Each individual dose group was small. Follow-up lasted 36 treatment weeks, with only four additional weeks of observation. Participants were taking metformin with or without an SGLT2 inhibitor, so the findings do not automatically transfer to every person with diabetes or to people seeking obesity treatment without diabetes. The trial was funded and sponsored by Novo Nordisk, and several authors were company employees and shareholders. Industry sponsorship does not invalidate the results, but independent replication and regulator review still matter.
Most importantly, placebo is a low bar for positioning a next-generation metabolic drug. Clinicians will eventually need direct comparisons on HbA1c, weight, treatment persistence, quality of life, cardiovascular outcomes, and cost. Until those data exist, claims that zenagamtide is better than established incretin therapy are speculation.
The Oria take
Zenagamtide’s real novelty is not the 14.6% number by itself. It is the demonstration that one multi-receptor peptide can produce meaningful glucose lowering in both a daily oral formulation and a weekly injection. That creates optionality. The injection may become the higher-efficacy route; the tablet may suit people who prioritize avoiding injections. Or later trials may show that one route has a clearer safety, adherence, or manufacturing advantage.
The current data also sharpen a lesson that applies across the peptide pipeline: receptor ambition has to be matched by dose practicality. Activating more pathways can increase efficacy, but it can also narrow the space between an effective dose and an intolerable one. The 20 mg and 40 mg injection groups make that tension visible.
Zenagamtide is investigational. It is not a supplement, not an approved prescription, and not a compound with a clinically established self-directed protocol. Products sold online under the amycretin or zenagamtide name are not validated by these trials. The studies used controlled formulations, structured escalation, eligibility screening, and medical oversight; an unverified vial or tablet is not the same intervention.
The next useful evidence will come from larger phase 3 programs with active comparators, longer maintenance periods, flexible dose management, and enough participants to assess less common harms. The oral program also needs fuller reporting of weight outcomes and practical dosing requirements. Until then, zenagamtide belongs in the promising-but-unproven category.
Sources
Evidence grade
Evidence Grade: B. Two randomized, double-blind, placebo-controlled phase 2 reports show meaningful HbA1c improvement, with strong secondary weight-loss data for the injection. Confidence is limited by small dose groups, short duration, high discontinuation at the upper injected doses, no active comparator, and sponsor involvement.
Medical disclaimer
This article is for educational purposes only and is not medical advice. Zenagamtide is investigational and is not approved for self-directed use. Do not buy, compound, inject, or take products sold under this name outside a lawful clinical trial. Decisions about diabetes or weight-management treatment should be made with a licensed clinician who can assess your medical history, medications, and individual risks. Oria exists to make peptide evidence easier to understand, not to replace medical care.
Related Compounds
Build Your Personalised Protocol
Turn your research into a personalised supplementation protocol with the BioStack Generator.