I'm not a big fan of new insulin pumps or its related technology (like continuous monitors) because they essentially keep patients as helpless addicts to the very costly products which support the system (closed-loop systems, for example, will keep us buying more insulin, infusion sets, not to mention blood glucose sensors which link to the system, as well as standard test strips necessary for calibration, glucose tabs for occasional lows which occur whenever insulin is delivered artificially in a non-physiological manner).
My thoughts on the "best" new technology is nanotechnology applied to address insulin's most critical failure, namely the fact that insulin (including all analogs) remains the FDA's top offender in terms of "drugs" with the most adverse effects in patients treated in emergency departments. According to a report in the October 18, 2006 edition of the Journal of the American Medical Association (JAMA), between 2004-2005, there was an estimated number of "adverse events" associated with insulin totalling 55,819 cases in the U.S. Hypoglycemia remains the top effect, although lipodystrophy and allergic reactions also exist. Regardless, analogs have not eliminated the possibility of hypoglycemia and in fact, intensive therapy has only increased the incidence, not reduced it. So what can be done?
Well, in 2003, several students from M.I.T. and Harvard won a prestigious prize from M.I.T. for their proposed business plan. Their idea was to develop nanotechnology to bond insulin molecules and sugar-sensitive proteins to a biodegradable polymer. The idea was to inject the nanoparticles into a repository under the skin similar to how patients with diabetes inject Lantus (insulin glargine) or other long-acting insulin formulations today. Using what then team spokesman Todd Zion called proprietary "clever biochemistry," the nanoparticles would theoretically detect a diabetic's glucose levels automatically and release only the appropriate amounts of insulin at precisely the right time to keep blood sugar levels steady, but not expose them to dangerous hypoglycemia. Those nanoparticles – dubbed “Smart Insulin” – contain nanoparticles that release insulin in proportion to blood-glucose levels, according to team member John Hebert, then a second-year student at M.I.T. Sloan School of Management. "These particles will start to slowly break down and release insulin into the bloodstream, regulating (the) blood-sugar level," Hebert said. "Once the blood sugar is at normal levels, the particles close back up, resolidify and then stop releasing insulin." They also secured a license from M.I.T. for the intellectual property rights to the technology disclosed in the U.S. patent application entitled “Stimuli-Responsive Systems For Controlled Drug Delivery”. This application covers work done by Dr. Todd Zion and others at the M.I.T. Chemical Engineering Department.
The company, called "
SmartCells, Inc. " is now based in Beverly, MA and has received considerable venture capital funding and NIH/NIDDK support. The product has successfully passed in pre-clinical (animal) trials, and is on the FDA's "fast-track" for future evaluation and review. Because insulin is highly similar conserved across the entire animal kingdom; the preclinical trials now being successfully undertaken in cats suggest it will also work in humans. To give an example of how true this can be true, consider that a nematode worm, uses insulin in very much the same way that vertebrates do. Within vertebrates, the simularity of insulins is remarkably close across all animal species. Bovine (cow) insulin differs from human in only three amino acid residues, and porcine (pig) insulin in only one. Even insulin from some species of fish is similiar enough to human to be effective in humans.
While the work remains fairly preliminary, theoretically, it would eliminate the need for testing, and all possibility of giving an inaccurate dosage, thus eliminating hypoglycemia, improving control and reducing the need for expensive test strips. Sounds like a winner to me!
What is the best new idea that you have heard of? 
Linear Mode
