Orforglipron ATTAIN trials — oral non-peptide GLP-1 agonist

The manufacturing story is what excites me most. Oral semaglutide (Rybelsus) requires the SNAC enhancer, has ~1% bioavailability, and needs 14 mg oral to approximate 0.5 mg injectable. That's incredibly wasteful from a manufacturing standpoint and limits cost reduction potential.

Orforglipron is a small molecule made by standard pharmaceutical chemistry — no peptide synthesis, no complex formulation. If it works, it could break the supply and cost bottleneck that has plagued GLP-1 RAs. Lilly has said it expects manufacturing costs to be "substantially lower" than injectable GLP-1 RAs.

Of course, "if it works" is doing a lot of heavy lifting. Phase 2 was small (n=272) and short (36 weeks). We need Phase 3 to confirm efficacy, assess longer-term safety, and understand the full dose-response. ATTAIN-1 results are expected in mid-2025.

Let me do a cross-compound comparison at similar timepoints. At 36-40 weeks:

• Orforglipron 45 mg oral daily: -14.7% (Phase 2, 36 wks)
• Semaglutide 2.4 mg SC weekly: ~-12.5% (STEP 1 at week 36 extrapolated from the published curve)[2]
• Oral semaglutide 50 mg daily: -15.1% (OASIS-1, 68 weeks, though not directly comparable)[3]
• Tirzepatide 15 mg SC weekly: ~-16% (SURMOUNT-1 at week 36, extrapolated)

Orforglipron is competitive but likely won't match tirzepatide's dual-agonist efficacy. Its advantage is the oral route, small molecule manufacturing, and potential for lower cost. For many patients, a daily pill that achieves 12-15% weight loss is preferable to a weekly injection that achieves 15-21%.

I have a pharmacology question. If orforglipron is a non-peptide small molecule, does it bind to the GLP-1 receptor in the same way as native GLP-1 and semaglutide? And does that matter for efficacy or safety?

I'm wondering if different binding modes could lead to different signaling profiles — what pharmacologists call "biased agonism." Could orforglipron have a different balance of G-protein vs beta-arrestin signaling compared to peptide GLP-1 RAs?

Excellent pharmacology question. You're right that binding mode matters. Orforglipron binds to an allosteric site on the GLP-1 receptor, distinct from the orthosteric site where native GLP-1 and semaglutide bind[4]. Despite this, it appears to be a full agonist with a signaling profile that closely mimics peptide GLP-1 RAs in terms of cAMP generation and insulin secretion.

There are subtle differences in receptor internalization and recycling kinetics. Preclinical data suggests orforglipron may cause less beta-arrestin-mediated receptor internalization than peptide agonists. In theory, this could result in less tachyphylaxis (receptor desensitization) over time — but this is speculative and unproven in humans.

From a safety perspective, the Phase 2 data showed a very similar adverse event profile to peptide GLP-1 RAs (nausea, vomiting, diarrhea as the most common). No novel safety signals. This suggests the clinical effects are mediated through the same downstream pathways.