FORTUNE — A recent article in the Wall Street Journal provided a fascinating description of a robotic pill designed to replace injectable drugs like insulin that are used to treat chronic conditions like diabetes.
The robotic pill hasn’t been tested in humans, and the Journal pointed out that it is at least a year away from even seeking federal approval, but a lot of people are rooting for it. They include InCube Labs, a Silicon Valley startup that’s doing the development work; Google Ventures (GOOG), which has provided early financing; and the more than 25 million Americans who currently suffer from diabetes.
The robot has competition. The shares of one would-be competitor — an Israeli company named Oramed Pharmaceuticals (ORMP) — soared from $6 to $32 in December when it reported positive results for an experimental insulin pill, citing a study that “clearly shows” that the drug “is safe and well-tolerated.” The stock gave back most of the gains when it turned out that results were based on a one-week test of 30 patients.
Here’s the thing. Everyone and his brother have been looking for a substitute for — or an alternate way to deliver — insulin, almost from the day it was discovered by a team of Canadian doctors more than 90 years ago. It does the job, but it has drawbacks.
MORE: America’s painful doctor shortage is threatening health care reform
Insulin is a protein, like hamburgers, and as such is digested and destroyed when swallowed. Back in the 1920s, it was injected using a glass syringe with a steel needle that had to be sterilized after every use. Early insulin was short-acting, reusable needles were never as sharp as during their first use, and injections were painful. It was a brutal, unsophisticated experience.
As early as 1925, attempts were made to find other ways to get insulin into the body. Then as now, the stakes were high: Complications from diabetes range from blindness and kidney failure to limb loss and cardiovascular disease. To make things worse, the lack of exercise and growing obesity are fueling a steady increase in diabetes. According to the Centers for Disease Control and Prevention, more than 40% of the U.S. population has either clinical diabetes or prediabetes, and the numbers are growing at what the CDC calls epidemic proportions.
Substitutes for insulin injections have regularly appeared during the past 90 years, but so far an acceptable answer — one that is effective, easily delivered and economically feasible — has not been developed.
The search for alternative delivery routes has involved the entire body. Although inhibitors at most sites preclude their use for insulin, every surface, membrane, and cavity has been explored as a possibility. In no special order, here are a few that have been studied over the years:
Vagina. Data show that insulin can be absorbed through the vaginal wall, but it’s an inefficient and socially unacceptable method.
Rectum. The rectal lining is a good vascular tissue and will absorb many drugs, but obvious limitations make this site unlikely for regular daily use.
MORE: Drone maker Titan could put Facebook in the cloud
Skin. The skin is a tough, dense tissue that protects our body by keeping foreign material out. Drugs can be delivered through the skin, but a large molecule like insulin needs a supplemental electrical force to pierce the skin. This is expensive and impractical. There have been attempts to use a patch with chemicals that open the pores, but over years it has not moved forward.
Nasal membranes. Insulin has been successfully delivered through the nasal membranes, but they are too fragile for regular use.
Lining of the mouth. This is a resilient membrane that also has absorptive capacity (think chewing tobacco). Insulin in solution has been sprayed into the mouth using an asthma-like inhaler that enables rapid absorption into the blood stream. This is a promising avenue that is undergoing tests in the U.S. and India.
Lungs. A few years ago delivery to and through the lungs was developed, but there are and were concerns about long-term usage and lung damage. The FDA approved it but only with regular periodic pulmonary function tests. The financial impracticality and limited prescribing led the producer to take it off the market. Another product has been under study for some time and will be reviewed by the FDA.
So if none of these really work, why not deliver insulin in a pill? Seems like a simple answer, but it’s anything but. For one thing, as anyone with recurrent heartburn, irritable bowel and the like knows, no two days in the GI tract are the same.
Then there’s the hamburger syndrome. When insulin is swallowed, it must pass through the vicious acid environment of the stomach, then through the equally vicious pool of digestive enzymes in the duodenum, then finally onto and through the lining of the intestines before it enters the bloodstream. Insulin must also pass through the liver, where it may be destroyed or sequestered before reaching the general blood flow and the body’s tissues.
MORE: Is Facebook overvalued?
Many companies have tried and failed to package insulin in a pill, and several new systems designed to protect insulin from the body are under development. I remain skeptical.
Which brings us back to the robotic device mentioned at the start, which is basically a capsule consisting of a polymer shell and a cargo of tiny, hollow needles made of sugar. Like a typical pill, the capsule is swallowed and makes its way through the GI tract. Once it reaches the small intestine, acids in the body dissolves the outer layer of the capsule, freeing a balloonlike structure that expands and gently pushes the drug-filled needles into the wall of the intestine.
It is an interesting concept that could have legs as a drug-delivery system, but it’s a long way from market. Its application to insulin also remains to be seen.
Dr. Gerald Bernstein is director of the Diabetes Management Program at Friedman Diabetes Institute at the Mount Sinai Beth Israel Medical Center in New York City and past president of the American Diabetes Association.
