COMMENTARIES -IDS
not we have sufficient information at this time to achieve this goal. I believe that the answer to this question is still 'no'. Two major pieces of information are needed. The first is reliable markers to unequivocably predict who is at high risk of developing IDDM. The recent demonstration that autoantibodies to insulin, GAD, and IA-2/IA-2b are good markers goes a long way in that direction and it now appears that the recombinant form of these antoantigens will soon replace the more cumbersome immunofluorescence islet cell autoantibody assay for predicting IDDM. 1 Ongoing studies from several laboratories indicate that the individuals at greatest risk are those who have autoantibodies not to one, but to more than one, of these autoantigens. The identification of other still unrecognized autoantigens may increase the reliability of prediction even further. Thus, it would be prudent to improve prediction before embarking on large scale population screening for neonatal intervention trials that might result in serious side-effects.
The second major piece of information that is needed for intervention is a better understanding of the pathogenesis of IDDM which is though to be autoimmune in nature. 2,3 Present thinking favours the use of islet cell autoantigens to produce immunological tolerance and to switch the immune response from Th1 to the less deleterious Th2. Although approaches along these lines have been successful in NOD mice, the NOD model differs from human IDDM in several important respects and to what extent one can extrapolate from the mouse model to human IDDM is unclear. It also is unclear as to whether islet cell autoantibodies cause, contribute, or are simply the result of the disease process. The majority of the investigators in this field currently favour the idea that it is really the cellmediated immune response, rather than the humoral immune response, that is responsible for b-cell destruction. Here too, the data on which this argument is based comes to a very large extent from animal models. Also of great concern is the lack of information as to the identity of the factors that trigger the autoimmune response in humans. Information as to whether the triggering factors are genetic; environmental, or both is needed to devise a rational plan for determining the direction and nature of therapeutic intervention.
Until more precise information on the reliability of autoantibody markers in predicting IDDM is obtained and the nature of the pathogenic process at the human level is better understood, intervention to prevent diabetes will be difficult to achieve.