By David N. Leff

Finland, a country of 1.1 million people and high hygienic standards, has the highest incidence of Type I diabetes in the world - about three times that of the U.S.

"There's a fair amount of evidence," observed diabetologist George Eisenbarth, "that in many countries throughout the world, the earliest of very young children are more often developing this form of the disease. In Finland it's doubled and tripled over the last 40 years. So there's almost an epidemic there of very young children developing Type I diabetes now, and we don't know exactly what has changed.

"It's almost impossible for so common a genetic disease as Type I diabetes to change that fast," he went on. "It's probably something in the environment - perhaps an environmental factor that prevents diabetes is decreasing. Maybe our society is becoming cleaner," Eisenbarth ventured. "but no one knows for sure."

Getting better known, but still far from sure, is predicting whether a newborn infant or toddler is fated to contract Type I diabetes, an autoimmune disease in which the body's T lymphocytes single out and slay the insulin hormone secreted by the pancreatic islet cells.

"Most importantly," Eisenbarth pointed out, "the autoantibodies appear very early in life in children who are developing diabetes - as well as in our best mouse models of Type I diabetes." Eisenbarth is director of the University of Colorado's Barbara Davis Center for Childhood Diabetes in Denver.

"The process that leads to childhood diabetes is locked in very early in life," he explained. "What we see now is that within their first year, often before nine months of age, children develop anti-insulin antibodies, and almost always go on to develop other autoantibodies. So they have a very high risk of becoming diabetic by three years of age."

Diabetologists have long prayed and striven for a metabolic crystal ball to help them foretell the autoimmune arrows of outrageous fortune that wound the very young.

"The first anti-insulin assay came almost 15 years ago," Eisenbarth recalled. "That was a very large test; you had to use almost a half-milliliter of blood. Then followed a major advance to make it a small-volume assay. These things develop in stages. The current version was developed by our laboratory. We have modified that assay now to where we can do everything in 96-well plates with a high throughput. "We take a drop of blood - mere microliters - from a mouse or a person, and we now do it even on filter paper, so it can work on a very large epidemiological scale."

Infant Mice, Humans Scored On Early Onset Type I

Eisenbarth is senior author of a paper on the subject in today's Proceedings of the National Academy of Sciences (PNAS), dated Feb. 15, 2000. It's title: "Early expression of antiinsulin autoantibodies of humans and the NOD mouse: Evidence for early determination of subsequent diabetes."

"The NOD [non-obese diabetic] mouse," Eisenbarth told BioWorld Today, "develops a form of diabetes very much like childhood diabetes, so NOD's one of our best animal models. It has some of the same genes that are involved in causing the disease." What he and his co-authors found "is that those mice, just like children, develop very high levels of insulin autoantibodies. The animals that get them early, by 4 to 8 weeks of age, become diabetic early - at about 16 weeks old.

"So there's a major effort using these mice to design new therapies to prevent their diabetes, which is caused by white blood cells - T lymphocytes - invading the islets of Langerhans cells that make insulin. And most of those T cells are actually targeting insulin, and killing the islets' beta cells." (See BioWorld Today, June 10, 1999, p. 1; and Sept. 20, 1999, p. 1.)

In the blood of human infants, the co-authors frequently found insulin autoantibodies (IAA) present as early as 9 months of age, at the time of first assay. In contrast, of 929 children who did not express persistent IAAs before age 1, only one went on to disease onset at 3 years of age, and expressed IAA at age 1.1 (13 months). Of five children with persistent IAAs before their first birthday, four progressed to diabetes by age 3.5, and the fifth is still under observation at age 2. "With such early determination of high risk of progression to diabetes," the PNAS paper concluded, "for maximal efficacy, immunologic therapies in humans may need to be treated in children before the development of IAA."

Different Ploys To Same Vaccine Goal

Apropos such therapies, Eisenbarth noted that "There are a number of immunologic vaccines being developed." He listed "at least three companies developing vaccines now - internationally - to try to prevent diabetes:

¿ "Neurocrine Biosciences Inc. has an altered peptide ligand for insulin B-chain peptide amino acids 9 through 23 - a molecule we've been very interested in;

¿ "[Another company] has a molecule, GAD - glutamic acid decarboxylase - another one of the target antigens;

¿ "Peptor, a research arm of the Weizmann Institute in Israel, is working with heat shock proteins.

So there's a lot of effort from the biotech industry to get molecules that might prevent this disease. Those would be eventual considerations for clinical trials to try to prevent the appearance of insulin autoantibodies.

"Meanwhile," Eisenbarth went on, "there are now clinical trials ongoing, trying to prevent Type I diabetes. A large one at the NIH is screening more than 60,000 first-degree relatives of diabetics for autoantibodies. Also studies following children from birth, to know who develops diabetes, such as our DAISY- [Diabetes Autoimmunity Study in the Young] project from Denver.

"The prediction of diabetes has gotten better over time," Eisenbarth commented. "We can now predict with pretty high accuracy who's at risk for Type I: about one in 300 children develop the disease in the U.S. Now there are these clinical trials to try to prevent it, and they will eventually use these types of assays either as an ancillary study or as the main entry criteria."