Currently, all of the 1 million type I diabetics in America and almost half of the 17 million type II diabetics need a dose of insulin two to four times a day. Injections were the only option until the recent development of a compact insulin pump with its rigid tube piercing the skin. Either method has drawbacks. Injections hurt and are bothersome to administer, and the skin opening for the pump’s tube is a welcome mat for infection.

Research motivated in part by a huge market potential is seeking alternative delivery routes. Among the devices rising toward the clinical forefront are breath inhalers shooting aerolized insulin directly into the lungs, new pills fashioned to protect insulin until reaching the intestine, and a sophisticated insulin patch, the favorite of Whitaker investigator Nadine Smith, Ph.D., at Pennsylvania State University. Smith’s patch uses ultrasound to penetrate the skin, offering diabetics a noninvasive and pain-free alternative to injections.

Delivering drugs through the skin with the help of ultrasound is not new, but most devices are too heavy, too bulky, or too expensive for diabetics. Smith’s device is uniquely compact—current prototypes are about the size of a matchbook—and easy to use. It sticks to the skin like a bandage.

To make such a small patch, Smith, an assistant professor of bioengineering and acoustics, relied on small cymbal transducers to produce the needed low-frequency ultrasound. Low-frequency ultrasound waves, around 20 kilohertz, safely enhance drug transport through the skin 1,000 times more than high-frequency ultrasound, between one and three megahertz. Exactly why this occurs is unknown.

Smith’s four cymbal transducers, named for their shape, measure just less than the diameter of a dime and are arranged in a two-by-two array. Think of them as four small, high-powered speakers producing imperceptible sound waves. When placed on the skin, the ultrasound blasts open microscopic pores, or channels, in the skin with the slightest tingling sensation. These channels allow insulin contained in reservoirs in front of the transducers to pass through the skin—insulin molecules are too large to pass through normal skin pores—then enter capillaries near the skin’s surface and into the blood.

In experiments with rats, Smith and colleagues demonstrated that the patch can deliver therapeutically effective doses of insulin and may be far more efficient than anticipated. Tuning the frequency reduced the exposure time from 60 minutes to only 20 minutes while still delivering the same effective dose of insulin. “We are hopeful that, eventually, we may be able to tune the system so that one to five minutes of exposure may be enough,” Smith says.

More recently, Smith has demonstrated the device’s effectiveness in rabbits. In an experiment published in the Proceedings of the IEEE Ultrasonics Symposium in October, Smith stuck the patch on three different groups of hyperglycemic rabbits. One group received the ultrasound patch loaded with insulin; another group had the patch loaded with a saline solution. In the third group, insulin was loaded, but the patch was never turned on. Over a period of one hour, only the rabbits using the active ultrasound patch with insulin experienced a significant decrease in their blood glucose to a safe level. The other two groups showed no decrease at all.

The next experiments plan to involve sheep and possibly other large animals before a human prototype is attempted. Smith envisions a prototype either hooked to a walkman-like power supply device or supplied with a self-contained battery that might make the patch disposable. Another feature could pair the patch with a continuous glucose monitor, essentially creating an automatic closed-loop system.

Already Smith is working toward this with another former Whitaker investigator, Michael Pishko, Ph.D., an associate professor in Penn State’s Department of Chemical Engineering. Pishko’s glucose sensor also uses ultrasound to open channels in the skin, but his device instead draws out fluid from which glucose levels are then taken. Ideally, the monitor would tell the patch when to turn on and off to maintain salutary glucose levels. Such a system would greatly benefit elderly and young diabetics, who could remain active without the constant watch of family or medical staff.

Smith says the ultrasound patch might be modified to deliver some pain relievers, asthma drugs, hormones, or AIDS medications in addition to insulin and other drugs that cannot now be taken by mouth. “There are different directions we can take this,” says Smith. “We think there’s a lot of potential.”

Smith received a Whitaker Foundation Biomedical Engineering Research Grant in 2000 for ultrasound research.