The obstacles are structural and unforgiving. After ingestion, gastric pH below 2 denatures proteins, and in the small intestine trypsin, chymotrypsin, elastase, carboxypeptidases, and brush-border peptidases continue the attack. Between 90 and 99 percent of orally administered protein may be destroyed before it ever reaches the absorptive epithelium. Survivors then confront the mucus barrier, a viscoelastic hydrogel of the gel-forming mucin MUC2, roughly 100 to 500 micrometers thick in the small intestine, in which protein diffusivity is 10 to 100-fold lower than in aqueous buffer. Beneath the mucus lies the epithelial cell layer, where tight junctions assembled from claudins, occludin, and zonula occludens proteins enforce an effective pore radius measured in angstroms to a few nanometers, far smaller than any intact biologic. The hydrophobic lipid bilayer of enterocytes and limited receptor-mediated transcytosis complete the blockade. For most biologics, absolute oral bioavailability stays below 1 percent under physiological conditions.

Passive strategies have produced real but modest gains. Chemical permeation enhancers that transiently modulate tight junctions have reached the market: oral semaglutide under the Rybelsus brand and octreotide as MYCAPSSA are approved products, and the SNAC-based candidate MK-0616 is in Phase III development. Yet oral semaglutide, the most successful oral peptide formulation to date, achieves only about 0.4 to 1 percent bioavailability in human pharmacokinetic studies despite extensive optimization, and it requires strict fasting because its absorption window is narrow and fragile. Nanoparticle-based systems reach absolute bioavailabilities of 0.1 to 2 percent for most biologics. The review’s conclusion is blunt: the dominant barriers are structural and biological rather than diffusional, so formulation optimization cannot overcome them. That experience shifted attention toward active delivery systems that apply physical force.

The past decade brought a clear developmental arc of ingestible devices. In 2019, the SOMA self-orienting millimeter-scale applicator used asymmetric mass distribution to align with the gastric wall and deploy drug-loaded milli-posts into the gastric submucosa, while the LUMI luminal unfolding microneedle injector pressed microneedle patches against the small intestinal wall, with feasibility demonstrated in swine and ex vivo human tissue. In 2022, two platforms extended the approach: the L-SOMA liquid-injecting successor delivered full-length IgG antibodies into the gastric submucosa of swine with absolute bioavailability of up to 80 percent, the first evidence that an ingestible device could achieve systemic IgG levels approaching subcutaneous injection in a large-animal model, and the RoboCap robotic mucus-clearing capsule rotated to mechanically clear the mucus layer and deposit payload on exposed epithelium, improving bioavailability for insulin and vancomycin 20 to 40-fold over conventional oral delivery in rodent and swine models.

The third generation eliminated needles and rotating elements altogether by generating delivery force from pharmaceutical excipients. The OSPRAE oral self-pressurized aerosol capsule, developed in Mark Prausnitz’s laboratory at Georgia Tech, uses a sodium bicarbonate-citric acid effervescent reaction to build internal pressure. When pressure reaches the failure threshold of a calibrated orifice, a high-velocity gas and drug plume is released at 10 to 50 meters per second toward the intestinal epithelial surface. In ex vivo porcine intestine, pressurized delivery at 100 to 170 kilopascals drove nanoparticles deep into the mucus barrier, and the fraction of payload within 30 micrometers of the epithelial cell layer, the firmly adherent inner mucus layer that turns over slowly, increased by an order of magnitude compared with passive deposition. In preclinical rodent studies, pharmacodynamic outcomes for insulin matched subcutaneous injection. Critically, the mechanism is convective rather than ballistic: drug particles are carried in the moving gas stream and deposited at or near the epithelial surface, not projected as projectiles, so payload sizes from 20 to 1,000 nanometers deliver efficiently across the tested pressure range without platform redesign.

Independent validation of effervescent chemistry came from a different format. Zhang and colleagues demonstrated a rectally administered in situ bioadhesive foam for colonic delivery of an anti-TNF-alpha antibody fragment in inflammatory bowel disease models. Potassium bicarbonate and citric acid react on contact with rectal fluid, producing carbon dioxide that simultaneously aerates the formulation into foam and transiently disrupts the colonic mucus and epithelial barrier, improving macromolecule transepithelial flux by more than tenfold in a non-everted rat gut model. Rectal Fab foam produced higher plasma concentrations than a solution at a comparable dose and showed therapeutic activity across acute and chronic colitis mouse models. The same gas-generating reaction can thus drive capsule rupture, foam expansion, or mucoadhesive depot formation depending on system design, confirming carbon dioxide generation as a broadly applicable excipient-based force strategy.

The field’s clinical milestone arrived in 2026 with the RT-111 RaniPill, a compressed-gas-driven ingestible injector that deploys a needle within the intestinal wall. In a Phase 1 human study, the platform delivered a full-length ustekinumab biosimilar with 81 percent relative bioavailability versus subcutaneous injection in healthy volunteers, with no serious adverse events and device remnant clearance by day three in all participants. This is the first published human pharmacokinetic study demonstrating that device-based oral delivery of a full-length IgG can match subcutaneous injection. The pharmacokinetic profile carries broader implications: RT-111 produced a Tmax of three days versus ten days for subcutaneous injection, consistent with direct vascular entry following intestinal wall penetration rather than slow lymphatic transit from a hypodermal depot. The study did not include thoracic duct cannulation or lymphatic-versus-portal sampling, so the absorption pathway has not been experimentally confirmed, but the data establish a clinical benchmark for the field.

One underexplored pharmacokinetic question may shape the entire design trajectory. All active oral delivery platforms that deposit drug into the small intestinal wall produce portal venous absorption, meaning hepatic first-pass exposure equivalent to conventional oral dosing. Yet the small intestine contains an extensive lymphatic network, the lacteals, whose capillary junctions of 100 to 500 nanometers are far wider than blood capillary fenestrations of 6 to 12 nanometers. In subcutaneous pharmacokinetics, macromolecules above roughly 10 nanometers in effective diameter preferentially enter lymphatic over blood capillaries, and for subcutaneous IgG the dominant mechanism is established as lymphatic transport. Whether intramucosal deposition of appropriately formulated macromolecules can result in preferential lymphatic drainage from the jejunal lamina propria remains an open question, testable by standard thoracic duct cannulation. An affirmative answer would reduce or eliminate hepatic first-pass elimination and move the field from barrier bypass to deliberate pharmacokinetic control of the oral absorption route. For insulin, the portal-first route may itself be advantageous, resembling endogenous insulin secretion more closely than subcutaneous injection does.

Engineering and safety considerations now define what each platform must prove. Across active systems, histological evaluations consistently show tissue disruptions confined to areas smaller than approximately one square millimeter, with preservation of surrounding villi, crypt architecture, and submucosal integrity; epithelial restitution follows within minutes to hours through migration of adjacent cells, a proliferation-independent repair mechanism that operates continuously in healthy intestinal tissue. No persistent ulceration, crypt damage, or inflammatory infiltration has been reported following pressurized delivery. Device orientation at the time of force application remains an unsolved problem, addressed through asymmetric mass distribution, self-righting geometry inspired by the leopard tortoise, multi-orifice jet configurations, and mucoadhesive anchors, though rough estimates suggest a single fixed orifice achieves correct mucosal apposition in only one-quarter to one-third of delivery events. Critical quality attributes span gastric protection integrity, actuation trigger reliability, force generation reproducibility with burst pressure of 100 to 170 kilopascals at plus or minus 15 percent coefficient of variation, and delivered-dose fraction, which remains uncharacterized for all platforms under physiologically relevant conditions. Chronic-use indications will require repeat-dose gastrointestinal toxicology of at least six months in rodents and nine to twelve months in non-rodent species.

Regulatory and manufacturing pathways are emerging in parallel. All active oral delivery systems will be classified as drug-device combination products under 21 CFR Part 3, with lead center assignment by the FDA Office of Combination Products depending on primary mode of action; the RT-111 program has confirmed that the agency engages with ingestible combination products targeting the intestinal wall. In the European Union, such products fall under the EMA framework for integral drug-device combinations, with device components compliant with the Medical Device Regulation 2017/745. Self-pressurized convective capsules sit at the simple end of the manufacturing spectrum: gelatin shells, sodium bicarbonate, citric acid, and standard coating materials are individually conventional and available at commodity scale, with no micron-scale fabrication, electronics assembly, or sterile device handling. Active delivery is not appropriate for all drugs, however; poor candidates include molecules requiring daily doses above 50 to 100 milligrams, narrow therapeutic index agents vulnerable to delivery variability, and proteins susceptible to mechanical shear during ejection. The leading candidates are GLP-1 agonists, insulin, parathyroid hormone, selected cytokines, and full-length IgG antibodies, the class with the highest unmet need. By reframing oral biologic delivery as a controlled physical transport problem rather than a permeability problem, convective platforms extend the reach of oral dosing to biologics previously confined to needles and infusion lines, with particular promise for low-resource settings where injectable infrastructure is impractical.

Subject of Research: Oral delivery of biologics: from barrier-limited formulation to active convective transport

Article Title: Oral delivery of biologics: from barrier-limited formulation to active convective transport

Article References: Niazi, S. K. (2026). Oral delivery of biologics: from barrier-limited formulation to active convective transport. Frontiers in Drug Delivery, 6, Article 1783113. https://doi.org/10.3389/fddev.2026.1783113

Image Credits: AI Generated

DOI: 10.3389/fddev.2026.1783113

Keywords: Oral, delivery, biologics, barrier-limited, formulation, active, convective, transport, scientific research