Department of Pediatrics Childrens Medical Services Center Gainesville, Florida 32608
![]() |
Introduction |
---|
![]() ![]() ![]() |
---|
Schatz et al. (1) are key players in the United States diabetes prevention trial, DPT-1, whereas Becker et al. (2) in Pittsburgh are perhaps the only large group of academic pediatric diabetologists not participating in a secondary prevention trial. The prominence of the latter group in studies of the etiology and natural history of diabetes makes it important to consider the reasons for this position.
The Pittsburgh brief begins with the assertion that any disease prevention has three major prerequisites: identification of those at risk, identification of the etiology or precipitating factor, and some understanding of pathogenesis. The fundamental prerequisites of the DPT-1 group, on the other hand, are that the disease be a demonstrated burden, that high-risk individuals be accurately identified, and that treatment is available that is safe, rational, and with the potential to be effective. These sets of prerequisites only agree on the point of identification of those at risk. The prerequisites put forth by Schatz et al. (1) are those usually associated with justification for screening a population; they are also more broadly applicable to our experience with successful disease prevention efforts. For example, smallpox was effectively prevented in a number of populations long before development of the germ theory (i.e. the understanding of etiology and pathogenesis of the disease; Ref. 3). It is conceivable that a preventive such as avoidance of bovine milk protein might work and be broadly applied before there is a coherent explanation for its role in the pathogenesis of diabetes.
Schatz et al. (1) outline the burden of diabetes, the inadequacy of treatment, and the profound importance of prevention, consistent with their first prerequisite, and this importance is inarguable. They consider that reducing prevalence rates by as little as 1015% would be sufficient justification for intervention, in view of the gravity of diabetes. Becker et al. (2) would seem to not agree with this, because of their concern that immune markers, at best, connote a 50% risk of developing diabetes over the subsequent 5 yr, and that 10% of relatives identified as low-risk also develop diabetes. This is an issue of statistical power and informed consent. Despite these limitations, substantial numbers of families have chosen to be a part of these pilot studies.
A recent report from Finland suggests considerably better predictability with available immune and metabolic markers than suggested in the review by Becker et al. (2). Siblings of children with recent onset type 1 diabetes mellitus were classified into four stages depending on whether there were no antibodies ("no prediabetes"), one antibody specificity ("early prediabetes"), two antibodies ("advanced prediabetes"), and more than three antibodies ("late prediabetes"). Six hundred sixty-one siblings were classified thusly and followed, and another 712 siblings were classified using first phase insulin response (FPIR) to iv glucose; that classification was no (no antibodies), early (one antibody specificity, normal FPIR), and late prediabetes (one or more antibodies, reduced FPIR). Followup was an average of 9 yr, during which only 0.9% in both groups who were classified as no prediabetes went on to develop the disease. The progressive classifications without FPIR were associated with rates of 6%, 23%, and 66%, whereas those including FPIR were 7%, 26%, and 92%. The authors considered staging of preclinical diabetes an important tool for eventual prevention therapy (4).
In reviewing the DPT-1, Schatz et al. (1) note that there have been no major safety concerns, an issue that is particularly important to the Pittsburgh group, who have generated considerable data about the risks of hypoglycemia to cognitive function, especially in young children.
Becker et al. (2) have provided a useful review of what is known about immune pathogenesis and prevention of diabetes in animal models and the background for studies in humans. Substantial detail is provided about the exciting primary prevention study in Finland, with the worlds highest incidence of type 1 diabetes (2.5 times that of United States white children), a single payor health system, and a homogeneous population. This is a model that will be difficult to emulate in the United States.
Concern is also expressed by the Pittsburgh authors about the applicability of trials in relatively high-risk relatives to the general population from which 90% of incident cases of type 1 diabetes come. Schatz et al. (5) have addressed this issue elsewhere with data indicating comparable significance of autoimmune markers for diabetes in the general population as in relatives. The logic of using relatives in initial studies is obvious, encompassing costs, commitment, and compliance. Fewer people need to be examined, and families of children with diabetes are extremely anxious to participate in research, especially if it has a possibility of a direct effect on other children in the family. This eagerness, of course, makes ethical considerations particularly delicate. I am reassured by knowledge of the stringency of institutional review board requirements, particularly for studies in healthy children, and the multiplicity of centers involved in the DPT-1. Before prevention efforts can be directed to the general population, strong evidence will be needed of efficacy and safety from studies of relatives.
Even as the Pittsburgh and DPT-1 group respond to the question, "Is now the time?" with opposite answers, both agree that the DPT-1 and other contemporary prevention studies are unlikely to provide the definitive approach, but that useful information will be derived about selection criteria and intervention strategies. Although Becker et al. (2) are convinced that now is not the time, they do not provide specific criteria that will assure them when that time has come. They are comfortable with the primary prevention trial, avoiding cow milk, because of its safety and the benign nature of the intervention. It would be interesting to know their take on the intranasal insulin study in the Finnish population, or the aerosolized insulin study in Australia, considering that one of their principal concerns was the administration of painful insulin injections twice daily to a group of children, half of whom are not expected to develop diabetes during the study period.
Becker et al. (2) have presented a thoughtful and understandable explanation for their decision to refrain from participation in current secondary prevention trials. Nonetheless, they recognize that the pilot studies will provide guidance for further steps in this endeavor. We are all eager to see the results of the ongoing efforts.
![]() |
Footnotes |
---|
Received October 20, 1999.
Accepted November 12, 1999.
![]() |
References |
---|
![]() ![]() ![]() |
---|