Orforglipron

Every GLP-1 receptor agonist before orforglipron was a peptide. That has two consequences: it has to be injected, or swallowed with an absorption enhancer and an elaborate fasting ritual, and it has to be manufactured by fermentation and solid-phase synthesis, which is why supply has repeatedly failed to meet demand. Orforglipron is a small molecule designed to activate the same receptor, made by ordinary chemical synthesis and taken as an ordinary tablet [1].

Phase 2. In 272 adults with obesity, 36 weeks of orforglipron gave weight changes ranging from -9.4% to -14.7% across doses against -2.3% on placebo [3]. In 383 adults with type 2 diabetes, 26 weeks gave HbA1c reductions up to 2.10% and weight loss up to 10.1 kg, against 0.43% and 2.2 kg on placebo and 1.10% on dulaglutide [4]. Those numbers looked like they might match the injectables.

Phase 3 brought them down. ATTAIN-1 randomised 3127 adults with obesity. At 72 weeks, mean weight change was -7.5% at 6 mg, -8.4% at 12 mg and -11.2% at 36 mg, against -2.1% on placebo; in the 36 mg group, 54.6% lost at least 10%, 36.0% at least 15% and 18.4% at least 20% [1]. ATTAIN-2, in 1613 people with obesity and type 2 diabetes, gave -9.6% at 36 mg against -2.5% on placebo [2]. Weight loss is reliably smaller in people with type 2 diabetes across this whole drug class, so that drop is expected.

Where that leaves it. Around 11% is below injected semaglutide's 15% [5] and well below tirzepatide. It sits close to oral semaglutide's territory - an indirect comparison of oral semaglutide 25 mg against orforglipron 36 mg has been published, though indirect comparisons are weak evidence [6]. The argument for orforglipron is not that it works better. It is that a scalable tablet with no dosing ritual reaches people that injections and fasting rules do not.

What is missing. No cardiovascular outcome trial has reported. Improvements in risk biomarkers were seen in phase 2 [7], which is not the same thing at all - semaglutide's standing rests on SELECT, not on biomarkers [8].

Orforglipron binds and activates the GLP-1 receptor as a non-peptide agonist. The downstream biology is the class biology: glucose-dependent insulin secretion, glucagon suppression, slowed gastric emptying and central appetite suppression [9].

The interesting question is whether a small molecule engages the receptor in exactly the same way a peptide does. GLP-1 is a class B G-protein-coupled receptor, which normally binds its long peptide ligand across a large extracellular domain; getting a small molecule to reproduce that was the hard medicinal chemistry problem. The phase 3 efficacy - real, but consistently a few percentage points below injected semaglutide - is consistent with a partial or differently biased activation, although the trials were not designed to answer that.

The practical mechanism that matters most is pharmaceutical rather than pharmacological. Oral semaglutide needs SNAC to get any peptide across the gastric epithelium, and its absorption is destroyed by food or water [10]. A small molecule is absorbed the way small molecules are, which removes the ritual and much of the variability.

An ordinary tablet. No cold chain, no injection device, no absorption enhancer, no fasting window [1]. Manufacturing is conventional small-molecule chemistry, which is the supply-side argument for the drug: peptide manufacturing capacity has been the binding constraint on GLP-1 availability worldwide.

The adverse-effect profile is the class profile, and it is dose-related. In ATTAIN-1, adverse events led to discontinuation in 5.3% to 10.3% across the orforglipron groups against 2.7% on placebo [1]. In phase 2 in diabetes, treatment-emergent adverse events ran from 61.8% to 88.9% across doses against 61.8% on placebo and 56.0% on dulaglutide, the majority gastrointestinal and mostly mild to moderate [4].

Three things are genuinely unknown. There is no cardiovascular outcome trial. There is no long-term exposure data beyond the trial durations. And as a small molecule it has a metabolic and drug-interaction profile that peptides do not have - hepatic metabolism creates interaction possibilities that injected peptides largely avoid - which will only be characterised properly with wider use.

The class-wide perioperative concern about delayed gastric emptying and residual gastric content should be assumed to apply [12][13].

No interaction data of note have been published. Unlike the peptide agonists, orforglipron is a small molecule subject to hepatic metabolism, so pharmacokinetic drug interactions are possible in a way they are not for semaglutide.

Delayed gastric emptying can slow the absorption of other oral drugs, and combining with insulin or a sulfonylurea raises hypoglycaemia risk, as for the rest of the class [4].