Most peptides die in the gut, where acid, pepsin, mucus, and tight junctions keep protein-sized molecules out of blood. Oral semaglutide (Rybelsus) is the rare tablet that made it through the FDA, and not because it became a small molecule. It did it with sodium N-[8-(2-hydroxybenzoyl)aminocaprylate] (SNAC), a permeation enhancer that opens a brief, local window in the stomach.

TL;DR

  • Swallowed peptides fail because the GI tract digests protein rather than absorbing it intact (Drucker, 2020).
  • SNAC buffers a patch of gastric mucosa, limits pepsin, and lets semaglutide cross transcellularly next to the tablet (Buckley et al., 2018).
  • Absolute bioavailability on the approved R1 tablets (3, 7, and 14 mg) is still only about 0.4 to 1% (FDA label).
  • PIONEER 1: oral 14 mg cut HbA1c 1.4% vs 0.3% on placebo over 26 weeks. OASIS 1: oral 50 mg (not the approved tablet) cut weight 15.1% vs 2.4% at 68 weeks.
  • A research capsule labeled “oral” is not SNAC-semaglutide. The formulation is the drug.

Educational only. Not medical advice, not a protocol, and not a reason to swallow a research vial.

Why do swallowed peptides fail so reliably?

A peptide is a short amino-acid chain. The stomach and small intestine exist to cut those chains into food. Pepsin at low pH, then trypsin downstream, will hydrolyze most unprotected sequences in minutes.

Even if a few molecules survive, they still have to cross mucus and an epithelium whose tight junctions are sized for ions, not for a ~4 kDa GLP-1 analog (Drucker, 2020). That is why almost every therapeutic peptide is injected.

Native GLP-1 already lasts about two minutes in blood. The oral problem is upstream of that clock. For half-life see the DAC, PEGylation, and lipidation guide, for the receptor see GLP-1 agonists, and for the chemistry primer see Peptides 101.

What is SNAC actually doing in the stomach?

SNAC is a small, lipophilic salicylic-acid derivative. Buckley’s dog and clinical studies showed semaglutide did not absorb along the intestine. Absorption happened in the stomach, next to the tablet, and only when SNAC was co-formulated with the peptide (Buckley et al., 2018).

Three local jobs, then it is gone. SNAC raises local pH enough to blunt pepsin and favors semaglutide monomers over oligomers. It also fluidizes epithelial membranes so the peptide can take a transcellular path.

The same paper found no evidence that tight junctions opened. The enhancement was transient and compound-specific. You cannot assume SNAC will carry an arbitrary research peptide the same way.

Single-dose work then showed that more SNAC is not better. Across 150 to 600 mg, 300 mg produced the highest semaglutide exposure and was locked into later development (Granhall et al., 2019). The FDA label still describes absorption as predominantly gastric, with food and extra water cutting exposure.

What did the oral semaglutide trials actually measure?

FDA approved Rybelsus on September 20, 2019 as the first GLP-1 receptor agonist that does not have to be injected. The molecule is the same lipidated semaglutide used in weekly pens. SNAC and gastric conditions are part of the dose form (FDA label).

Study or source What was tested Design Main finding
Buckley 2018 Site and path of absorption Clinical PK plus dog imaging Stomach, tablet-adjacent, transcellular, SNAC-dependent
Granhall 2019 SNAC amount vs exposure Single and multiple ascending dose 300 mg SNAC beat 150 mg and 600 mg
FDA label, R1 tablets Absolute bioavailability Population PK About 0.4 to 1% for 3, 7, and 14 mg
PIONEER 1, 2019 Oral 14 mg vs placebo in type 2 diabetes 26 weeks, monotherapy HbA1c -1.4% vs -0.3% from an ~8% baseline
OASIS 1, 2023 Oral 50 mg vs placebo, no diabetes 667 adults, 68 weeks Mean weight -15.1% vs -2.4%

PIONEER 1 is the clean monotherapy study. Oral 14 mg dropped HbA1c 1.4 percentage points versus 0.3 on placebo over 26 weeks (Aroda et al., 2019). A ~1% oral fraction still works because tablet milligrams are larger than the weekly injection and a week-long half-life integrates the daily trickle.

OASIS 1 pushed a 50 mg oral dose, which is not the approved Rybelsus tablet. In 667 adults with overweight or obesity and no diabetes, mean weight change at 68 weeks was -15.1% versus -2.4% on placebo (Knop et al., 2023). Eighty-five percent versus 26% lost at least 5%.

The clinician takeaway is not a 50 mg protocol. It is that once SNAC gets intact semaglutide into plasma, the pharmacology looks like the injectable peptide.

Why is 1 percent bioavailability still a success?

Because the alternative for most peptides is zero. A 0.4 to 1% fraction sounds like a failed formulation until you remember that unprotected proteins are food. The newer R2 tablets (1.5, 4, and 9 mg) list about 1 to 2% absolute bioavailability, which is why those milligram strengths are not interchangeable with R1 milligram-for-milligram (FDA label).

Variability is the clinical tax. Water volume, residual food, and the wait before the next oral drug all move AUC. The label’s empty-stomach, limited-water, 30-minute gap exists because gastric absorption is a local, saturable event. Miss the window and exposure falls.

That is why catalog “oral peptides” are not this product. SNAC-semaglutide is an FDA-reviewed tablet with measured gastric absorption. A research peptide in a gelatin capsule has none of that engineering. For injectables, use the reconstitution guide and the calculator.

What should clinicians take from this?

Oral delivery is no longer a fantasy for one GLP-1 analog. It is still a fantasy for most of the research list. The Buckley mechanism does not generalize to BPC-157 or GH secretagogues just because a vendor printed “oral” on a bottle.

When you evaluate a peptide product, ask which barrier was solved. Half-life extension, albumin hitchhiking, and SNAC gastric absorption are different inventions. Confusing them is how a 1% oral fraction gets sold as if it were a high-bioavailability injection.

For listed compounds and who carries them, use the price index and the provider directory. For the people who actually run these trials, see experts.

Bottom line

Oral semaglutide works because SNAC buys a few minutes of gastric, transcellular absorption, not because peptides became swallowable. A 0.4 to 1% bioavailability with a week-long half-life is a pharmaceutical trick. It is not a property of peptides as a class.

Educational only. Not medical advice.