Peptide bioavailability primer: why route matters more than dose for half of these compounds, and what the published PK literature actually shows

What bioavailability is, and why it is not just absorption

Bioavailability (F) is the fraction of an administered dose that reaches systemic circulation as the active parent molecule. It is bounded by 100 percent (intravenous, by definition) and 0 percent (no drug ever reaches the bloodstream). It depends on three sequential processes: getting the molecule across the absorptive surface (the skin, the gut wall, the nasal mucosa, etc.); avoiding metabolism on the way (gut wall enzymes, first-pass hepatic metabolism); and arriving in active form (parent molecule, not a degradation product).

For peptides, the second and third steps dominate. Most peptides are protease substrates and have plasma half-lives of minutes once they reach circulation. The route therefore matters far more than for small-molecule drugs, because each route shapes both the absorbed fraction and the rate of arrival relative to the rate of clearance.

Subcutaneous injection: the default for therapeutic peptides, and why

Subcutaneous (SC) injection deposits the peptide into the loose connective tissue between the skin and underlying muscle. From there, the drug is absorbed primarily through local capillaries and lymphatic vessels into systemic circulation. SC bioavailability for therapeutic peptides typically ranges from 60 to 90 percent for compounds without targeted formulation tricks, with absorption time courses on the order of hours rather than minutes .

GLP-1 receptor agonists illustrate the SC pattern. Semaglutide SC has a Tmax (time to peak plasma concentration) of 1 to 3 days and a half-life of about 1 week, supporting once-weekly dosing. The half-life comes from albumin binding via a fatty-acid side chain, not from the SC depot per se . Tirzepatide is similar in profile . Insulin analogs use SC routinely; insulin lispro reaches peak in about 30 to 90 minutes, glargine and degludec reach effective concentrations more slowly, and the depot kinetics are tunable through formulation excipients (zinc, protamine, hexamer-monomer equilibria).

• Absorption rate is slowed by larger molecule size, hexameric or oligomeric self-association, albumin binding, and PEGylation.
• Site rotation matters because subcutaneous tissue remodels under repeated injection (lipohypertrophy), which changes absorption pharmacokinetics; the published diabetes injection literature documents this clearly.
• Cold or warm injection sites change local blood flow and therefore absorption rate; the effect is small for slow-release peptides and meaningful for fast-acting insulins.
• Inflammation at the site (recent trauma, dermatitis) can alter absorption unpredictably; the safer practice is to rotate to a healthy site.

Intramuscular: faster absorption, narrower indication set

Intramuscular (IM) injection deposits the drug deeper, into a muscle belly with substantially higher local blood flow than subcutaneous tissue. IM absorption is typically faster than SC and often produces higher peak concentrations for the same dose. The trade-off is that IM injection is more painful, has higher needle-stick risk, and carries different infection risk because of the deeper deposit.

For most peptide therapeutics, IM is not the labeled route. It is used for emergency-use products (intramuscular glucagon for severe hypoglycemia ) and for vaccines where rapid systemic distribution and immune-cell encounter are desirable. Routine IM administration of slow-release SC-labeled peptides changes the PK profile in ways the labeled-product trials did not study, which is why off-label IM use of SC-labeled GLP-1 products is not a recommended substitution.

Intranasal: high bioavailability for some, near-zero for others

The nasal mucosa is highly vascularized and avoids first-pass hepatic metabolism. For molecules that can cross the mucosa, intranasal (IN) bioavailability can be high, sometimes approaching IV. The catch is that the nasal epithelium is selective: small lipophilic molecules cross readily, large hydrophilic peptides cross poorly, and the mucosal proteases degrade some peptides in transit.

• Bremelanotide (PT-141): intranasal formulations were studied historically. The currently approved Vyleesi product is SC, not IN, because IN bioavailability and consistency were inadequate for a label .
• Oxytocin: intranasal oxytocin has substantial published research literature in social-cognition and bonding studies, but central nervous system penetration from IN dosing is debated and the systemic plasma rise is small. Effects observed in trials may be mediated centrally without large systemic concentrations .
• Calcitonin (salmon calcitonin): intranasal calcitonin is FDA-approved for postmenopausal osteoporosis and was for many years a rare example of a clinically used intranasal peptide .
• Desmopressin: intranasal desmopressin is FDA-labeled for diabetes insipidus and primary nocturnal enuresis .

The intranasal route is therefore not a generic substitute for injection. It works for specific molecules with the right physicochemical profile and an FDA-reviewed formulation. Off-label intranasal use of injectable peptides is not supported by published PK data for most compounds.

Oral: the hardest problem, and what has actually solved it

Oral peptide delivery has to clear three sequential barriers: gastric proteolysis, small-intestinal proteolysis, and crossing the gut epithelium. Solving any one barrier helps; solving all three is rare. Oral semaglutide (Rybelsus) is the only commercially available oral GLP-1 product. It uses a permeation enhancer (sodium N-(8-(2-hydroxybenzoyl)amino)caprylate, SNAC) to create a transient local pH bubble in the stomach where the molecule is protected from pepsin and pushed across the gastric epithelium .

The bioavailability of oral semaglutide is approximately 1 percent. The product works clinically only because the parent molecule is otherwise pharmacologically potent and because the SNAC formulation produces sufficient peak plasma concentrations to engage GLP-1 receptors. The general pattern: small percent of dose, but enough to be clinically useful for a high-affinity receptor agonist.

Most peptides do not have this combination. Insulin in particular has been the white whale of oral peptide delivery for over half a century. Multiple late-stage attempts (oral insulin programs in the 1990s and 2000s) failed for inconsistent bioavailability and inability to reproduce the meal-time PK profile that injectable insulin achieves.

Buccal, sublingual, transdermal, inhaled: niches that mostly do not apply

• Buccal and sublingual: bypass first-pass hepatic metabolism but require the molecule to cross oral mucosa. Small lipophilic molecules work; most peptides do not cross efficiently. Some clinical-stage products have used buccal tablets with permeation enhancers; bioavailability is typically <10 percent.
• Transdermal: peptides are too large and too hydrophilic to cross intact skin. Iontophoresis and microneedle arrays can help but are not generally used for systemic peptide delivery. The local cosmetic peptide market is a separate use case (topical only, not systemic).
• Inhaled: pulmonary bioavailability of small peptides can be substantial because the alveolar surface is large and absorptive. Inhaled insulin (Afrezza) is the surviving FDA-approved example; earlier products were withdrawn for commercial rather than safety reasons .
• Rectal: rarely used for peptide therapeutics. The rectal mucosa avoids first-pass hepatic metabolism for the lower rectum but has low absorptive capacity for large hydrophilic molecules.

What this changes for clinical and research decisions

When a vendor or clinician suggests using a peptide by a route different from the labeled or studied route, the burden of proof shifts onto the suggester. The published PK data for the labeled route do not transfer to other routes; bioavailability, time-to-peak, and degradation patterns can change by orders of magnitude. The right question is whether the new route has its own published PK profile in humans, not whether the molecule has been used by some route somewhere.