Robotic pills deliver gastrointestinal injections

The first insulin injection in 1922 marked a breakthrough in the therapeutics available for diabetes (1). However, the delivery route has served as a formidable barrier to the immediate adoption of this life-saving treatment. One study reported that people with diabetes delayed injecting insulin an average of 7.7 years in favor of less effective oral medications (2). “Often, both the patient as well as the healthcare provider have limitations in starting medications that involve an injection,” said Giovanni Traverso, a biomedical engineer at the Massachusetts Institute of Technology (MIT) and gastroenterologist at Brigham and Women’s Hospital. “That is a multifactorial challenge in that some people feel that injectables carry a stigma; some people are afraid of needles; some people find it difficult to get trained to give themselves an injection or to go for an infusion.” 

Finding ways to accommodate the preference for oral drug delivery over injections is not just a matter of patient comfort or convenience but also of medical care quality. It’s also challenging, as insulin and other protein- and nucleic acid-based biologics are rapidly degraded by digestive acids and enzymes in the gastrointestinal tract. Furthermore, many of these drugs are too big to diffuse through the tight cellular junctions and thick mucus layers of the stomach and small intestine. For these reasons, “oral delivery of biologics is considered to be the Holy Grail of drug delivery,” said Alex Abramson, a chemical and biomolecular engineer at the Georgia Institute of Technology. 

Amidst a sea of formulation strategies to shield fragile drugs from gastrointestinal fluids and enhance the diffusion of large biomolecules, scientists wondered whether they could simply inject biologics from inside the body. Researchers have developed robotic pills that respond to a physiological cue after ingestion, triggering the injection of a medication payload into the walls of gastrointestinal organs. Rather than relying on diffusion, the capsules insert the drug into blood vessel-rich tissue, enabling efficient absorption into the bloodstream. Importantly, as the gastrointestinal tract lacks sharp pain receptors, the injections are painless. With data demonstrating systemic uptake on par with subcutaneous injections and ongoing clinical development, these ingestible autoinjector devices are poised to provide a much needed oral alternative for a wide range of biologics. 

A spoonful of sugar glass 

As a graduate student at MIT, Abramson worked with Traverso to develop a device that delivered biologics into the highly vascularized submucosal layer of the stomach. The drug was enclosed inside a protective capsule and then directly injected into the tissue, allowing it to bypass the gastrointestinal environment.  

To create this type of device, the team needed to ensure that the injection reliably fired into the tissue wall rather than into the stomach’s large interior cavity — a challenge since the capsule could land in many possible orientations after being swallowed. One potential solution was to incorporate needles throughout the entire surface of the capsule, but with this design, most of the drug would be lost to the lumen. Instead, the researchers sought a way to control the orientation of the device so that a single needle always faced toward the tissue wall. They came across the work of mathematician Gábor Domokos on monostatic bodies, objects that possess only one stable resting position (3). Domokos discovered that the leopard tortoise could easily right itself after being tipped over by a predator due to its highly curved upper shell and low center of mass (4). The team used the tortoise’s shape and density distribution as a model for their device and mathematically optimized it. In this way, they designed a capsule made from a polymer material on top and stainless steel on the bottom that self-orients to a stable, tissue wall-facing position in the stomach. 

Oral delivery of biologics is considered to be the Holy Grail of drug delivery. 
- Alex Abramson, Georgia Institute of Technology 

Next, the researchers needed to engineer a system for automatically injecting the drug after the device settled against the stomach tissue wall. Recognizing that the capsule would be exposed to hydration once ingested, they searched for a material that could dissolve to release energy from a compressed spring. They decided to create sugar glass-like spring actuators from sucrose and isomalt. “In the humidity chamber, that sugar starts to dissolve, and as soon as it reaches a critical dimension, essentially, it fractures and the spring just gives way,” Traverso said. “This is on the order of milliseconds, so you're able to release that energy all of a sudden to support an injection event.”  

The capsule reaches the stomach within a minute of being swallowed and remains there for at least 30 minutes before passing into the small intestine, so the researchers timed the dissolution of the sugar so that that the injection would occur three to five minutes after ingestion. Targeting the stomach allowed them to define a consistent window during which to deliver the injection “while also making sure that almost as soon as the patient swallows the pill, they're able to get that drug delivered into their body,” Abramson said. 

The team first demonstrated that their self-orienting millimeter-scale applicator (SOMA) device could administer solid formulations of biologics (5). These solid drugs are more shelf-stable than their liquid counterparts and provide a more sustained release profile as they dissolve, making them well-suited for applications such as delivering long-acting insulin. As a model system, the researchers compressed insulin into a pointed structure that acts as both needle and drug. Using an endoscope, they placed insulin-loaded SOMA devices into the stomachs of swine. When gastrointestinal fluid entered the device through vents in the capsule, the sugar glass actuator disintegrated and released the spring. The injection inserted the insulin needle into the stomach tissue, where it dissolved. The researchers observed that the drug needle could penetrate through the mucus lining without damaging the outer muscular layer of the stomach. They found that injecting the drug via the SOMA yielded plasma concentrations of insulin and a decrease in blood glucose comparable to the effect seen when they placed the insulin needle under the skin. 

The researchers then adapted the device to inject liquid formulations. While the dosage of solid drugs (up to one milligram per capsule) is limited to the volume of the needle itself, liquid forms can spread throughout the stomach tissue, enabling higher doses up to four milligrams per capsule. Liquids also provide more rapid drug uptake, allowing the device to deliver fast-acting injection-based drugs such as mealtime insulin and epinephrine to treat allergic reactions. For this formulation, the team transitioned to a hollow hypodermic needle and redesigned the device’s injection mechanism to operate this syringe-like system. A dissolving sugar glass pellet on the shell of the device triggers a spring that inserts the needle into the tissue and then forces the liquid through the needle. The dissolution of the first pellet exposes a second pellet, which disintegrates to activate another spring. This spring retracts the needle back into the device after the injection, reducing the risk of damage as the capsule passes through the body and is excreted. 

The researchers loaded the liquid injecting device (termed L-SOMA) with either the antibody adalimumab, an analog of the peptide GLP-1, insulin, or epinephrine and endoscopically placed the capsules into swine stomachs (6). They observed that 28 out of 31 swine showed systemic drug uptake, with plasma concentrations of all four compounds comparable to the levels achieved via subcutaneous or intramuscular injections. The team calculated what fraction of the GLP-1 peptide and insulin dose they administered entered the bloodstream and found that these percent bioavailability values were similar for the L-SOMA and subcutaneous routes. Insulin gave a rapid drop in blood glucose, and epinephrine yielded a swift increase in blood glucose and heart rate when the drugs were delivered by either an L-SOMA or traditional injection. 

“It's very exciting that we're getting comparable uptake to subcutaneous injection,” Abramson said. “It does bring up one of the issues of the pills, that we're not getting 100 percent of the pills to deliver 100 percent of the time.” The researchers discovered that instances in which the device did not yield systemic drug uptake were due to insufficient depth of penetration of the needle. Using the pellet-based actuation mechanism, they can store enough energy in the spring so that the depth of penetration is limited by distance rather than force. By extending the length of the needle, the team aims to constantly reach the stomach’s submucosal layer. 

The researchers observed that the presence of a food layer in the stomach millimeters thick did not affect device performance, as the dense capsule sinks through food particles to settle against the tissue wall and perforates a thin food barrier with its forceful injection. 

The researchers administered L-SOMA injections to swine for three consecutive days and then euthanized the animals to analyze their stomach tissues. They found that only the most recent injection delivered on the day of euthanasia left any discernable mark, suggesting that the device may induce minimal, quick-healing tissue damage. While the safety of more invasive colonoscopies and repeated subcutaneous pokes bodes well for chronic SOMA injections, “we need to look into longer term dosing,” Abramson said. “So, what happens over the course of a year of injections into the stomach? That’s something that we’re not sure about yet.” The team also orally administered drug-free SOMA devices to dogs and, using radiograph imaging to track its journey through the body, observed that the capsule passed through the gastrointestinal tract without problems. 

Hopefully, by making something that people can self-administer, it will make it easier for them to be able to get the vaccine and will also, by removing the idea of an injection, make the idea of getting a vaccine more palatable to a greater audience. 
- Alex Abramson, Georgia Institute of Technology 

The SOMA device recently underwent a Phase 1 clinical trial led by the pharmaceutical company Novo Nordisk in which healthy participants swallowed the capsule once (7). The trial will report on activation of the injection, the integrity of the device after excretion, and any adverse effects such as obstruction. “We're eager to understand the early safety studies... so, we’re looking forward to those results,” Traverso said. He is also interested in identifying a clinical application for the SOMA technology that would be most helpful to people. “Is it something like a vaccine? Is it a protein? Is it a long-acting insulin? Is it something else?”    

A vaccine is certainly a possibility, as the researchers recently demonstrated that the device is also compatible with mRNA injections (8). They encapsulated mRNA inside polymer nanoparticles that shuttle the nucleic acid into the cell, where it is translated into a protein. When the researchers manually injected the nanoparticles into the stomach submucosa of mice, they observed evidence of mRNA delivery into stomach cells and liver cells, indicating both gastric and systemic uptake. In order to fit a clinically relevant dose of mRNA inside a capsule, the team freeze-dried the nanoparticle solution and resuspended it so that it was more highly concentrated. With this strategy, they could load 50 micrograms of mRNA into a single pill, which is similar to the quantities found in the mRNA vaccines for COVID-19. 

The team is now investigating which specific organs and cell types they can target with gastrointestinal delivery of mRNA drugs and how administering a vaccine in this way could influence the immune response. Abramson sees value in an oral option even for a vaccine that only requires a few doses. “Hopefully, by making something that people can self-administer, it will make it easier for them to be able to get the vaccine and will also, by removing the idea of an injection, make the idea of getting a vaccine more palatable to a greater audience,” he said. 

Robotic pills meet humans 

Other scientists have already demonstrated successful drug delivery in humans with an ingestible autoinjector device. Researchers at the clinical-stage biotechnology company Rani Therapeutics developed the RaniPill, a capsule that features an enteric coating that protects it as it traverses the acidic conditions of the stomach. When the capsule reaches the small intestine, the higher pH environment breaks down the pill’s coating and outer shell, exposing its interior to intestinal fluid. The fluid dissolves a valve separating two chemical reagents, allowing them to combine and react to generate carbon dioxide gas. The gas inflates a balloon, causing a microneedle on the balloon’s surface to inject into the intestinal wall. The biodegradable polymer needle and the solid drug formulation inside dissolve, allowing the drug to be absorbed into the vascularized tissue. Ejection of the needle leaves a hole in the balloon, causing it to deflate and pass through the body. 

In preclinical studies, the researchers observed that the RaniPill can deliver more than 12 biologics, including large proteins and antibodies, with similar bioavailability to subcutaneous injections. Their first-in-human clinical trial in 2019 demonstrated that the capsule was well tolerated and could intestinally administer octreotide, a peptide used to treat neuroendocrine tumors and acromegaly, with high bioavailability (9). In this study, the team tested multiple versions of the device with different balloon sizes, allowing them to improve the design so that the needle consistently reaches the intestinal wall without causing discomfort as the balloon expands. 

In a recent Phase 1 clinical trial, the researchers evaluated the pharmacokinetics and safety of the RaniPill’s delivery of teriparatide, an analog of the human parathyroid hormone (10). This drug is approved as a daily injection that is taken for up to two years to treat osteoporosis. In three treatment groups, healthy participants ingested one RaniPill containing either 20 or 80 micrograms of teriparatide or received one subcutaneous injection of 20 micrograms of the drug. The researchers collected participants’ blood samples over six hours and calculated the concentration of the drug in the blood as a function of time. Based on this measure of systemic absorption, they observed that relative to the subcutaneous injection, the lower dose RaniPill gave a three-fold increase in bioavailability, while the higher dose pill showed a four-fold increase. “You can get bioavailability as high as a sub[cutaneous] injection or even better because with the sub[cutaneous] injection, the absorption of the drug is a little bit slower compared to when it's delivered transenterically,” said Arvinder Dhalla, the vice president of clinical development at Rani Therapeutics. By detecting the presence of teriparatide in the blood, the researchers found that an optimized RaniPill capsule successfully delivered the drug in 20 out of 21 people. 

In a repeat dose portion of the study in which participants took the RaniPill once a day for one week, the researchers observed that food consumption did not affect device performance. In the small intestine, food exists as chyme, a chunky liquid-like substance through which the needle should easily penetrate. “It becomes a huge restriction for the patient if you cannot eat for so many hours, so we want to make sure that our pill would not carry that kind of restriction,” Dhalla said. 

The team did not observe any significant adverse effects related to the RaniPill in the single or repeat dose parts of the study. This is consistent with their preclinical finding that administering a daily pill to dogs for one week did not produce any damage to the small intestine, a rapidly self-healing organ. The researchers also incorporated an X-ray-based imaging marker into the device, allowing them to visualize the capsule and follow its passage through the body. They observed that in all subjects, device remnants were excreted as expected. So far, the team has given almost 200 RaniPills to almost 100 healthy volunteers without any safety concerns. 

Following these promising results, the team is interested in understanding the potential of the teriparatide-loaded RaniPill to treat osteoporosis. Using a rat model of the disease, they evaluated three drugs at equivalent doses: teriparatide injected intraperitoneally (which mimics the RaniPill’s delivery mode), teriparatide injected subcutaneously, and another osteoporosis drug called abaloparatide injected subcutaneously. After six weeks of treatment, the researchers measured the rats’ bone mineral densities and observed that the intraperitoneal teriparatide injection improved bone health to a similar degree as the subcutaneous injections, indicating that the administration route did not affect the drug’s activity. The team plans to conduct a Phase 2 clinical trial that will evaluate the efficacy of the teriparatide RaniPill capsule in osteoporosis patients.

As physicians and patients may delay the teriparatide injection regimen in the treatment of osteoporosis, the RaniPill could allow the drug to be adopted earlier in the progression of the disease. Even for osteoporosis patients seemingly tolerating the injections, “there are no symptoms. You don't feel any pain, but your bones are brittle, and now you have to take an injection every day for up to two years,” Dhalla said. “If we come up with an oral option, that's a huge advantage for patients. It's convenient for them to just take an oral pill compared to an injection and it reduces the burden of injections and improves their quality of life.” 

With an expanding portfolio of research programs and an upcoming Phase 1 clinical trial of a RaniPill containing the antibody ustekinumab (a drug for psoriasis and other inflammatory conditions), the team at Rani Therapeutics hopes to revolutionize the delivery of a broad spectrum of biologics (11). As Dhalla said, “For people who have to take lifelong injections for some of these chronic diseases, we hope to change their lives completely.” 

References

1. American Diabetes Association. The history of a wonderful thing we call insulin. 
2. Calvert, M.J., McManus, R.J., & Freemantle, N. Management of type 2 diabetes with multiple oral hypoglycaemic agents or insulin in primary care: retrospective cohort study. Br J Gen Pract  57, 455-460 (2007).  
3. Várkonyi P.L. & Domokos, G. Mono-monostatic bodies: the answer to Arnold's question. Math Intell  28, 34–38 (2006). 
4. Domokos, G. & Várkonyi, P.L. Geometry and self-righting of turtles. Proc Biol Sci  275,11–17 (2008).
5. Abramson, A. et al. An ingestible self-orienting system for oral delivery of macromolecules. Science  363, 611-615 (2019).  
6. Abramson, A. et al. Oral delivery of systemic monoclonal antibodies, peptides and small molecules using gastric auto-injectors. Nat Biotechnol  40, 103-109 (2022). 
7. Clinical Trials. A study investigating the safety and performance of DV3395, a new concept device for the delivery of medicine.
8. Abramson, A. et al. Oral mRNA delivery using capsule-mediated gastrointestinal tissue injections. Matter  5, 975-987 (2022). 
9. Clinical Trials. A first-in-human study of the RaniPill, an oral drug delivery platform.
10. Clinical Trials. Study evaluating PK of PTH administered orally via RaniPill capsule. 
11. Clinical Trials. Study evaluating PK of ustekinumab administered orally via RaniPill capsule.