You might call it 'smellevision.'

Imagine a microchip inside your TV set that picked up remote signals from the channel you're tuned to. They release scents or fragrances into your room, keyed to the commercial or program scene you're watching.

A variant of that far-out concept is microchip jewelry - rings, bracelets, necklaces - that respond fragrantly to changes in their wearer's mood, by emitting appropriate perfumes. In this version, the chip resides right inside the bauble, and its microprocessors activate appropriate scent-loaded reservoirs matching mood-triggered alterations in skin salinity.

These olfactory lifestyle amenities still lie somewhere in the future, according to biomaterials inventor Robert Langer, at the Massachusetts Institute of Technology (MIT), in Cambridge. Closer to the here and now are microchip devices for high-precision, remote-controlled delivery of therapeutic drugs in microgram doses. (See BioWorld Today, July 31, 1992, p. 2.)

Langer, an endowed professor of chemical and biomedical engineering at MIT, is senior author of a paper in today's Nature, dated Jan. 28, 1999, and titled "A controlled-release microchip." Its lead author is chemical-engineering graduate student John Santini Jr.

"In our current prototype device," Santini told BioWorld Today, "the types of drugs that you'd be able to release would have to be things that were very potent. Drugs that exist now in that dose range are hormones, steroids, some of the very potent pain medications, like the morphine derivatives. Because those are therapeutic down to the microgram level."

In their current dime-size, proof-of-principle model, the MIT co-authors have installed 34 drug reservoirs, "each about the size of a pinprick," Santini said, "and each capable of holding around 25 nanoliters of chemical, in solid, liquid or gel form."

Each of these drug-dispensing wells is shaped like a truncated pyramid. Its small end is sealed by a metallic gold film three-tenths of a micron thick, which serves as an electrical anode.

Into all 34 of those reservoirs' large (but still microscopic) 500-micron backsides, a laser writer's inkjet printer head pipettes drops of chemical solution simulating a therapeutic drug. This aperture is then closed with waterproof epoxy.

"Our chip," Santini said, "is very simple - no fancy electronics. The way it works is, we have this ultra-thin gold-film electrode that covers the reservoir containing the drug. And when we send an electrical one-volt current to it, that anode reacts. The gold dissolves and within 10 seconds the reservoir opens."

To illustrate the potential therapeutic application of this low-volume, high-potency drug release, Santini cites "for example, some infertility treatments that require the patient to wear a small pump that delivers pulses of certain hormones every 90 minutes for weeks at a time, via a catheter inserted through the skin. Those hormones could potentially be incorporated into a chip implanted under the skin, and [be] programmed to release the contents of specific reservoirs at specific times."

To control this dose-releasing administration, he proposes three alternative regimens. "One would be to have a microprocessor on the chip, preprogrammed to deliver the material at specific time intervals," he said. "Another [would be] some kind of remote-control unit, having either the physician or the patient, or some other person, able to send a signal to the device that activates release of material. Or thirdly, [scientists could] actually integrate a biosensor directly onto the chip, which could read concentrations in solution, and then react appropriately, delivering material when it's needed."

Besides implantable or ingestible drug-delivery microchip capsules or tablets, the MIT microchip has several other strings to its bow, in such fields as microbiology, combinatorial chemistry and diagnostics.

Immediate Bedside Analysis - Not Delayed Lab Tests

"Many diagnostic tests today," Santini said, "involve adding precise amounts of chemicals in a precise order to fluids like blood and saliva. As a result, samples must be sent to the lab, where results can take hours to days. A microchip preprogrammed to release the proper reagents at the right times and in the right order could be fitted to the end of a probe, swirled in a vial of body fluid at the bedside, and deliver the findings as the patient waits."

Of course, insulin is a drug that diabetics must inject chronically multiple times per day, for life. The outlook for adapting the MIT drug-releasing microchip to insulin points up both the device's present limits and future prospects.

"Insulin is a perfect example of the types of pulsatile delivery it would be useful for," Santini said, "for things you need to deliver on a frequent basis, not always at the same time. But the current method, using liquid solutions of insulin, probably would not work for diabetics, because the dosages are huge compared to what our chip probably could handle.

"Right now," he said, "the prototype chip is only 0.3 millimeters thick. We know that by increasing that thickness to even one millimeter, which is still extremely thin, we can augment the dose capacity per reservoir to 10 times its original value. Going to three or four mm thick, which is still thin, we're talking of getting a more-than-100-fold increase over what we're doing now. At that size, it could still be indwelling or implanted."

Insulin? Maybe Later; Not Yet

"Maybe if there were such an increase in microchip volume," Santini said, "that coincided somehow with the development of a different formulation of insulin, perhaps a solid form if possible, that might make it a feasible application."

Should that occur, he foresees that insulin release could be automatically regulated by the metabolic demand posed by the patient's ever-changing glucose levels. "I don't see why not," Santini observed. "People are using microfabrication technologies right now to make glucose sensors. There's no reason they couldn't be integrated directly onto the chip itself."

On August 25, 1998, the U. S. Patent Office issued patent No. 5,797,898 to inventors Santini, Langer and MIT's Michael Cima (the Nature paper's third author). The title is headed "Microchip Drug Delivery Devices." MIT currently has two other patent applications pending, one covering fabrication of the microchips, and a foreign filing that protects all aspects of the technology.