During the first day, speakers at the workshop reviewed studies on virus-induced diabetes in animals and on immune system interactions that might lead to the disease. Another panel discussed the evidence that viral infections, including retroviruses, may be implicated in human diabetes. A third group looked into the evidence of the association between viruses, such as rubella and Coxsackie B4, with the development of diabetes. A fourth group discussed various non-viral environmental factors that might be involved in the etiology of diabetes such as dietary proteins and toxins. This group also looked into possible ways of exploiting these factors in intervention studies.

The second day was devoted to a broad discussion on how to proceed with developing strategies for detecting diabetogenic agents such as viruses; on ways to develop studies of the interaction between viruses and the immune system; and to exploring strategies for identifying non-viral environmental factors.

There has accumulated a substantial amount of evidence that Type 1 diabetes is the result of an autoimmune response. Autoantibodies directed at the beta cells have been identified, and studies in the relatives of those with already diagnosed diabetes have shown that these autoantibody markers are present many years before the disease becomes apparent.

The group was particularly interested in a report from Arthur Like, MD, University of Massachusetts Medical School, Worcester, MA, who noted that under some circumstances autoimmunity could be induced, modulated, or triggered by viruses, and several speakers suggested that more animal models of this sort should be investigated.

While there is a growing body of evidence that immune mechanisms are involved in the development of Type 1 diabetes, it is more than likely that the disease is not the result of some single acute infection or other

insult. Rather diabetes develops over the years, albeit more readily in some individuals than in others, and is due to the interaction of several factors acting in concert. This made the task of those charged with suggesting strategies for looking for and identifying potential diabetogenic agents particularly difficult.

"We’re looking for things that will jump start the field, novel approaches and risky projects that might not be funded under the normal system of investigator-initiated grant proposals. We’re looking for new people, for example virologists who haven’t studied diabetes, people with innovative ideas," said Richard Eastman, MD, Director, Division of Diabetes, Endocrinology, and Metabolic Diseases, National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK).

Essentially diabetes is the result of pancreatic beta-cell destruction and the resulting failure of the body to secrete insulin. The insulin lack can be replaced artificially but the long-term consequences of the disease remain. They include poor circulation often resulting in kidney failure and blindness. Diabetes carries a major burden on those affected and makes a better understanding of the disease an urgent matter both for improvements in treatment and for developing effective prevention steps.

There is a definite genetic component to the development of Type 1 diabetes and some of these genes are known, although it is possible there are other "diabetogenic" genes to be discovered. But genetic predisposition to diabetes is far from the whole story. For one thing there is a high prevalence of these so-called susceptibility genes in the general population, yet this is not reflected in the incidence of the disease. Even in identical twins, one of which may develop diabetes, in about half the cases the other twin, who carries the same genes, does not develop diabetes. Furthermore, in recent years the disease has been increasing in the general population at about 2.5% a year.

At the end of their two-day meeting Thomas Merigan, MD, Stanford University Medical Center, Stanford, CA, drew up a short list of suggestions that emerged from the discussions. In essence they were:

1. There is a need for a US survey to identify those at high risk of diabetes by evaluating genetic markers and measuring autoantibodies as well as to make a major effort to obtain pancreatic tissues from persons who die of diabetes.

2. More emphasis is needed on the epidemiology of diabetes. There are a whole set of questions that need to be answered regarding studies on environmental factors, migrant populations and indeed the entire natural history of the disease. The epidemiology of the disease has to be in place in order to deal adequately with any good idea about controlling the disease, Merigan said.

3. The possibility of doing intervention trials involving modification of diets, the use of antiviral agents, or immunosuppressive measures, should be explored. Dietary factors and infectious agents have been implicated in the development of diabetes, but definitive proof is still lacking.

4. There is a need to study relevant animal models and tissue culture systems so as to look at the production of disease, at interactions between the islet cells of the pancreas and immune system mechanisms. Data, such as those obtained from animal studies of viral infections, would come from these studies and would have a bearing on any intervention trials.

5. There is a need to encourage new scientists into the field by providing training and opportunities for career advancement.

Only a minority of genetically susceptible individuals develops Type 1 diabetes. "Why is this?" asked Jaaka Tuomilehto of the National Diabetes and Genetic Epidemiology Unit, National Public Health Institute, Helsinki, Finland. He recommended examining the role of environmental and genetic factors in the development of diabetes. The issue is: How to go about identifying these factors and teasing out the role they play?

Franco Bottazzo, London Hospital Medical College, London, England pointed out that, since 1960, Finland and Sardinia have been experiencing an increasing incidence of diabetes that now exceeds that of other nations. "Yet there is no concordance between them. Something has happened to cause this increase," he said. He said it would be worth while trying to elucidate the factors involved.

Besides these issues, the group discussed ways that might be used to study the possible roles of virus infections, both acute and chronic; and the influence of diet and toxins and other possible agents on the development of diabetes. Michael B. A. Oldstone, MD, Department of Neuro-pharmacology, Scripps Research Institute, La Jolla, CA, said the time was ripe to take advantage of the emerging technologies that enable one to measure genes that get turned on or off. "These new technologies

offer incredible opportunities," he said recommending that their potentiality be fully investigated.

A number of viruses such as Coxsackie B4 and rubella have been identified as possibly playing a role in the development of some cases of diabetes. Animal studies have shown that virus infections can result in diabetes. "The association between congenital rubella and diabetes became clear following the 1963-65 rubella pandemic," noted Paul Parkman, MD, a consultant from Kensington, MD and a developer of the rubella vaccine. He estimated the risk of Type 1 diabetes in congenital rubella was 100 times that of the general population. However, diabetes is not associated with either the natural disease or vaccine-induced infection.

"Rubella provides the strongest evidence of the viral etiology of diabetes, but we have no idea how the virus initiates the destruction of beta cells. We should take a much closer look at how this virus is involved," said Audrey J. Tingle, MD, British Columbia Research Institute for Children’s and Women’s Health, Vancouver, Canada.

On the basis of a paper published in the journal Cell that appeared in 1997 there was considerable enthusiasm expressed that retroviruses might be involved in the etiology of diabetes. However, by the time the workshop was held several groups had reported they were unable to reproduce the findings of the original report. Consequently this issue was not discussed in any detail at the workshop.

But as the discussions progressed it became clear that one of the major roadblocks to an organized effort in the study of diabetes was ignorance of the natural history and pathogenesis of the disease.

"I know of no other major disease where we know so little about the natural history and pathogenesis of the disease, in terms of what happens within the target organ, i.e. the pancreas, as the disease develops," said Abner Notkins, MD, the workshop’s chairman and a researcher at the National Institute of Dental and Craniofacial Research. This Institute and the NIDDK along with the National Institute of Allergy and Infectious Diseases were involved in the planning and support of the workshop. The Juvenile Diabetes Foundation International also supported the workshop.

Notkins said the single most important issue is to obtain and study pancreatic tissue from those identified as susceptible to the disease.

Apart from the safety and ethical aspects, involved in obtaining a biopsy of the pancreas in otherwise healthy individuals, biopsy specimens might not be informative. Pancreatic islet cells are not uniformly affected. "Islet cell destruction is patchy," said Alan Foulis, MD of the Royal Infirmary,

Glasgow, Scotland, so a particular specimen may not present with a true picture.

Nevertheless, Notkins said, this issue needs to be fully explored. "We must obtain pancreata from individuals who die of diabetes but may still have some residual beta cells. But," he noted, "so far, obtaining such tissue has been extraordinarily difficult." He suggested that this was one of the reasons why so much diabetes research has concentrated on animal studies. "We don’t have human pancreatic tissue with which to work." He noted that the National Disease Research Interchange had been successful in obtaining pancreatic tissue, but mostly from normal individuals. "Few, if any, came from newly onset diabetic individuals."

Notkins proposed establishing an International Pancreas Tracking System that would enable scientists to obtain pancreata of people who die of diabetes and examine them in an organized coordinated way by a consortium of scientists with expertise in histopathology, immunology, and infectious diseases. "If we succeed in getting hold of a couple of dozen pancreata, study them systematically and in depth, we will obtain an enormous amount of information. Hopefully, from this information we will be able to design experiments which will help us determine whether

environmental factors are really involved in the pathogenesis of diabetes."

One possible source for such pancreata would be the participants in the diabetes prevention trials that are currently going on. In these long-term studies there might be deaths among participants from causes other than diabetes, so pancreatic tissues could be obtained from these individuals.

Merigan suggested that a program similar to one conducted by the Centers for Disease Control and Prevention (CDC) to investigate unexplained deaths might be a way to study pancreatic tissue specimens in an organized manner and to monitor the high risk individuals. He explained that the CDC picks up cases of unexplained deaths, gathers the specimens and distributes them to investigators each with a special skill or expertise who look for possible causes. "It starts with the foot soldiers who go out and get the specimens and ends with the data coming out of the lab," Merigan said, suggesting that a consortium of scientists could get together to form such a program.

Over the years there has been much attention to the view that dietary factors, such as early nutrition, the ingestion of breast versus cows’ milk is associated with the risk of developing Type 1 diabetes. "This issue needs to be clarified," said several of the participants.

Regarding dietary influences, Fraser Scott, Ph.D., of the Nutrition Research Division, Banting Research Center in Ottawa, Canada said that he didn’t think that just infant diet is likely to be a factor. "I think that there needs to be a longer term exposure to dietary components to produce diabetes." This is true in animals, he said. But in respect to studies in humans he noted that there are a number of studies going on that might be helpful.

He cited the US National Health and Nutrition Examination Survey (NHANES) which collects detailed diet histories and which looks at the relation of diet and the development of chronic diseases. There are a number of ways to do this, he said. One would be to develop new molecular biomarkers of nutrient status and of disease status, that would detect the expression of genes required for nutrient uptake, for infectious agents, and for toxins. This could be done in humans, he said. Commenting on this, Samuel Baron, MD, of the University of Texas Medical Branch in Galveston, TX said that a systematic approach along these lines could narrow the field down making dietary intervention studies in humans feasible.

Regarding diabetes intervention studies in general, Helsinki’s Tuomilehto had a warning. "Most of the data we have regarding the dietary components tell us that the population at risk is about 10% to 15%. That’s the fraction in which we may be able to prevent disease by dietary interventions. So the studies will not be very powerful and they will be extremely costly."

He recommended that a group review the ethical and scientific aspects of any large-scale intervention studies before they are started. "Otherwise we may be wasting a lot of time and money not to mention putting those high risk families that would be involved to some considerable inconvenience and perhaps even risk." But even if an intervention program is effective in only 10% of the population at risk it would be very cost effective, said Keith Taylor, MD, University of London, Rye, Essex, England.

At one point Stanford’s Merigan reminded the group of the reason they were discussing these issues at all. "We’re having this meeting because Congress has specifically authorized funds for implementing our

suggestions." He noted that diabetes activists have convinced Congress that we must try and do something to improve the care of those with diabetes. "This is a very practical approach, a request for the NIH to spend this money on practical solutions. We’re talking about trying to develop a strategy that makes sense to the average person."

NIDDK’s Richard Eastman noted the various suggestions that had been put forward and asked how they might be prioritized. Since resources are limited, he indicated that the most attractive suggestions would be those that would be most likely to make the biggest change. With that in mind, he said, the group’s suggestions would be carefully reviewed. "We will need to think in some detail how these proposals can be implemented."

At the end of the workshop’s two days the group concluded that, based on experimental animal findings and on provocative human epidemiologic studies, it is possible that viruses might be involved in the etiology of diabetes. However, the group concluded that at this stage the data are far from conclusive.