Why most peptide therapies stay injectable

Published 2026-06-155 min readBlogBy the Peptide Protocol editorial team · reviewed

The most common question from new GLP-1 patients is "why can't this just be a pill?" The answer is biology: peptides face three barriers in the GI tract that small-molecule drugs don't. Solving even one is hard. Solving all three takes either an absorption enhancer (SNAC) or replacing the peptide with a small molecule that's not a peptide at all.

TL;DR. Peptides face three barriers to oral absorption: stomach acid denatures them, gut proteases digest them, and the gut epithelium doesn't pass molecules larger than ~500 Da. Most peptide drugs (4,000–5,000 Da) fail all three. The two real workarounds: SNAC absorption enhancers (oral semaglutide) and replacing peptides with small molecules (orforglipron). Both have limits.

The three barriers

Barrier 1: Stomach acid (pH ~1.5)

The stomach is highly acidic — pH around 1.5–2 during digestion. Most peptides are folded into specific 3D structures stabilized by hydrogen bonds and electrostatic interactions. At pH 1.5, these bonds break: the peptide unfolds. An unfolded peptide is no longer biologically active, and it's also exposed to proteases that wouldn't reach the folded form.

Barrier 2: Gut proteases

Pepsin in the stomach, trypsin and chymotrypsin in the small bowel — all designed by evolution to cut peptide bonds. They don't distinguish between dietary protein and therapeutic peptide. A peptide drug landing in the stomach is being actively dismantled within minutes.

Barrier 3: Epithelial size cutoff

The gut epithelium's tight junctions don't let molecules above ~500 Daltons cross efficiently. Most therapeutic peptides are 1,000–5,000 Da:

BPC-157: ~1,419 Da
TB-500 (LKKTETQ): ~860 Da
Semaglutide: ~4,114 Da
Tirzepatide: ~4,813 Da
Insulin: ~5,808 Da
Human growth hormone: ~22,000 Da

Even if a peptide survived the acid and proteases, the epithelium wouldn't let it through. Some absorption occurs paracellularly (between cells) but the rate is too low for systemic effect.

Why injection works

Subcutaneous or intramuscular injection bypasses all three barriers. The drug enters the loose connective tissue or muscle, is absorbed into local capillaries, and reaches systemic circulation through the lymphatic and venous systems. No acid exposure. No protease exposure (the proteases that matter are in the gut, not the bloodstream). No epithelial barrier — capillary walls are much more permissive than the gut epithelium.

The two real workarounds for oral peptides

Workaround 1: SNAC and similar absorption enhancers

SNAC creates a localized "absorption pocket" in the stomach: buffers the pH temporarily, transiently increases epithelial permeability, and allows a small fraction (~1%) of the peptide to cross. See SNAC explained.

Limits:

~1–2% bioavailability ceiling. To deliver a therapeutic dose, the tablet contains 10–20× the equivalent injectable mass.
Empty stomach required. Food breaks the absorption pocket.
Works for semaglutide; the formulation engineering for other peptides is product-specific and difficult to generalize.
Tablet manufacturing is more complex than a vial of solution.

Workaround 2: Replace the peptide with a small molecule

Pharmaceutical chemistry has been steadily figuring out how to make small molecules (~500 Da or less) that fit and activate the same receptors as peptides. Once you have a small molecule, oral delivery becomes routine: standard drug development territory.

Examples in development:

Orforglipron: small-molecule GLP-1 agonist (~600 Da). See orforglipron explained.
Other small-molecule incretins: Pfizer, Roche, and others have programs.

Limits:

Designing a small molecule that mimics a peptide's receptor binding is hard. The receptor evolved to bind a peptide.
Selectivity is often imperfect — small-molecule analogs may activate the target receptor and several others.
Patent protection differs and pricing strategies vary.

Imperfect workarounds

Sublingual

Held under the tongue, peptides can cross the oral mucosa, which is more permissive than the gut epithelium. But still much less permissive than capillary walls. Sublingual peptide bioavailability is typically 5–25% — much worse than injection, much better than swallowed oral. Useful for some peptides (e.g., insulin in development); not a substitute for injection in most cases.

Nasal

The nasal mucosa is similarly more permissive than the gut. Calcitonin and a few other peptides have nasal formulations. Bioavailability is typically 10–30%. Limited by nasal-spray volume and dose ceiling.

Inhaled

Inhaled insulin (Afrezza) is FDA-approved and reaches systemic circulation through the lung alveoli. Niche use due to inconsistent dosing and pulmonary concerns.

What this means for the field

Injection isn't going away soon. Most peptide therapy will remain injectable for the foreseeable future, particularly for peptides above ~3,000 Da.
Small-molecule analogs are the medium-term future. For receptors where a small molecule can mimic a peptide, expect oral analogs over the next 5–10 years.
Absorption-enhancer technology will improve. SNAC is generation 1. Generation 2 enhancers may push oral bioavailability into the 5–10% range, enough to make oral peptide therapy practical for more drugs.
"Why isn't it a pill" is a temporary frustration. The answer is real biology, not pharma laziness.

FAQ

Why does insulin have to be injected after 100 years of research?

Three barriers, plus insulin's size (~5,800 Da) is well above the threshold. Several oral insulin programs have failed despite billions in investment.

Is the bioavailability problem the same for all peptides?

Roughly. Smaller peptides (under 1,000 Da) have somewhat better odds; very large peptides (above 5,000 Da) approach zero oral bioavailability. The shape of the molecule matters too — flexible peptides degrade faster than tightly folded ones.

What about peptide patches (transdermal)?

Skin is the toughest barrier of all for large molecules. Transdermal peptide delivery is in research but no FDA-approved product exists for systemic peptides via patch.

Why are some peptides effective sublingually if absorption is so limited?

Some peptides (notably BPC-157) appear to work partly locally — they don't need systemic absorption to produce some effects. For systemic effects, sublingual is much weaker than injection. See <a href="/blog/bpc-157-oral-vs-injectable-bioavailability/">BPC-157 oral vs injectable</a>.

Compare routes across peptides

Peptide Protocol shows which peptides have oral, sublingual, or injectable forms — and the bioavailability tradeoff for each.

Get the iPhone app →

Informational and educational only. Not medical advice. Consult a licensed clinician before starting, changing, or stopping any peptide protocol. Mentions of investigational, compounded, or research-use peptides are for informational purposes; many such substances are not FDA-approved for human use.