This Article Full Text (PDF) Submit a related Letter to the Editor Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Becker, D. J. Articles by Dosch, H.-M. Search for Related Content PubMed Articles by Becker, D. J. Articles by Dosch, H.-M.

Prevention of Type 1 Diabetes: Is Now the Time?¹

Division of Endocrinology Children’s Hospital Pittsburgh Pittsburgh, Pennsylvania 15213-2583

	Introduction

Top Introduction What do we know?... What do we know?... Caveats from animal research IDDM prevention trials in... Current IDDM prevention trials Nicotinamide trial The DPT-1 Trial The Nutritional Prevention of... Human intervention studies: if... References

 
THE PREVENTION of any disease has three major prerequisites: 1) identification of subjects at risk to develop that disease or disorder; 2) identification of the cause of the disease and/or its precipitators: and 3) some understanding of the pathogenesis of the disease. Research into the prediction and prevention of Type 1 diabetes has been directed over the years toward all these prerequisites, and we approach the new millennium with an increasingly large body of new knowledge of the disease. The issues to be debated here are whether we are ready for new, large intervention trials at this time and whether the current ongoing intervention trials were helpful, timely, or possibly premature. When considering an intervention study we must determine: 1) the characteristics and the size of the target population to be studied; 2) when is the right time to intervene in subjects at risk? 3) what duration of the study should be planned? and 4) what should the intervention be?

	What do we know? Identification of individuals at risk

Top Introduction What do we know?... What do we know?... Caveats from animal research IDDM prevention trials in... Current IDDM prevention trials Nicotinamide trial The DPT-1 Trial The Nutritional Prevention of... Human intervention studies: if... References

 
Type 1 diabetes is an autoimmune disorder with T cell-mediated destruction of the insulin-producing ß cells of the islets of Langerhans in both humans and animal models of disease (1). Peri-insulitis is well described in animal models, but does not cause insulin deficiency until a precipitating event, or a genetic program, induces islet invasion with macrophages/dentritic cells, CD4+ and CD8+ T cells, with the latter appearing very early in the process (1, 2, 3). It is likely, although not sequentially proven, that this course of events also occurs in humans (4). In humans, the phase prior to the onset of insulin deficiency is heralded by the development of circulating autoantibodies against multiple ß cell antigens. The discovery of islet cell antibodies (ICAs), initially in patients with multiple autoimmune endocrinopathies (5), gave the impetus to studies like our own, aimed at the identification of subjects at risk for the development of insulin-dependent presumably Type 1 diabetes. The obvious, ultimate goal of these studies was to identify a population with high disease risk in whom intervention efforts might prevent total ß cell destruction.

The Pittsburgh Epidemiology and Etiology Study, initiated in February 1979 (6, 7), has now confirmed our initial impressions that ICA alone (even in high titer) is not a sufficiently specific marker of rapid progression to insulin deficiency (8). Whereas it is true that the majority of Caucasian children and adolescents (over 90%) who develop insulin-dependent diabetes mellitus (IDDM) under 18 yr of age do have ICAs detected on either human and/or rat pancreas substrates, only 60% of their relatives who later develop IDDM, have ICAs in their first blood sample (9).

Our most recent analysis is the IDDM conversion rates in 2164 euglycemic parents and siblings of children diagnosed with clinical Type 1 diabetes recruited during the first 5 yr of our study (i.e. 1979–1984). All subjects were contacted in 1996 and 1997 (i.e. after 13–18 yr of actual follow-up) (Fig. 1). Life table analysis shows that only 15% of ICA-positive subjects on human pancreas (5 JDF) had developed IDDM after 5 yr and 34% after 10 yr of actual follow-up. As can be seen in Fig. 1, even subjects with relatively high titer ICAs (i.e. above 20 JDF units) had only a 30% risk of developing insulin dependency after 5 yr and 62% after 10 yr. These conversion rates to IDDM are very similar to the Barts-Oxford Study (10) and the large, but shorter, ICARUS Data Set (11). In the ICARUS data set, a 50% risk of developing IDDM after 5 yr based on the presence of ICAs alone was only seen in those individuals with ICAs of more than 80 JDF units (11).

View larger version (16K): [in this window] [in a new window]   Figure 1. Life table analysis showing conversion to IDDM in all normoglycemic first-degree relatives recruited from February 1979 to January 1984 (the first 5 yr) with absolute follow-up in 1996 and 1997.

Results from early studies initiated in Boston, Massachusetts, have emphasized a role for the presence of insulin autoantibodies (IAAs) in the prediction of IDDM in younger subjects (18). In addition, metabolic studies using first-phase insulin response (FPIR) to iv glucose stimulation predict fairly rapid progression to insulin requirement (19). The presence of IAAs in the ICARUS data set gave similar risks to ICAs when analyzed alone, but were clearly age dependent with an inverse correlation of IAAs with age among these relatives (11), as is seen in newly diagnosed patients (20). The ICARUS Study confirmed the previously reported importance of age at the time of detection of ICAs in predicting outcome (21). For example, relatives under the age of 10 yr at the first blood draw had a 54% risk of conversion to IDDM after 5 yr compared with only 10% of those 40 yr of age or older (11). Thus, the presence of IAAs or age could be used interchangeably when calculating risks (11). The importance of age, whether it be an independent risk factor or a marker of the presence of IAAs, should be taken into account when designing intervention studies that use conversion to clinical IDDM after 5 yr as an end point.

The development of assays that can be used universally that detect specific autoantibodies other than those directed at insulin has revealed that combined analyses of autoantibodies to GAD, IA2, insulin, and ICA improve prediction of IDDM in ICA+ relatives (11, 22). These initial reports have now been confirmed in other large studies, including our own (9, 23, 24). Our current data show that the presence of three antibodies (ICA, GAD, and IA2AA) identifies 55% of subjects who develop IDDM by 5 yr by life table analysis (Fig. 2). Our preliminary findings in a small data set (n = 21) have shown that if we included low FPIR and high-risk genotypes without these antibodies we could improve the prediction to only 57% after a prolonged follow-up (24), but to 80% with the addition of GAD65 and IA2AA (25). Thus, the combinatorial analysis of multiple autoantibodies and histocompatibility leucocyte antigen and/or FPIR diabetes family members can identify subjects with considerable, long-term disease risk. But these markers also identify a number of individuals who will not develop IDDM and also miss a subset of future converters (15% with 0 to 1 autoantibody at 10 yr of follow-up).

View larger version (16K): [in this window] [in a new window]   Figure 2. Life table analysis showing conversion to IDDM in a subset of first-degree relatives according to the presence of 0, 2, or 3AA at the first blood draw (25 ).

	What do we know? Intervention in animal models of IDDM

Top Introduction What do we know?... What do we know?... Caveats from animal research IDDM prevention trials in... Current IDDM prevention trials Nicotinamide trial The DPT-1 Trial The Nutritional Prevention of... Human intervention studies: if... References

Because of the reported success of oral tolerization in animal models of autoimmune encephalitis, oral feeding of insulin has been tried with some success in a number of different IDDM mouse models (40, 41, 42, 43, 44). However, oral insulin does not prevent diabetes in BB rats (45). It is likely that this form of peripheral tolerance induces peripheral CD4+ regulatory T cells, which have been shown to confer suppression of the development of IDDM in NOD mice after adoptive transfer of these cells (46). It is probable that these CD4 regulatory T cells secrete a number of anti-inflammatory cytokines, including interleukin-4, interleukin-10, and transforming growth factor ß. Although somewhat indirect, a large body of evidence from rodents, and to some extent humans, supports the hypothesis that IDDM immunoprotection is conferred by changing the prevalent bias from Th1 to Th2 phenotype (2, 46). The intranasal route of insulin delivery has also been investigated as a form of immunotherapy in NOD mice, both at weaning and at 7 weeks of age, after the autoimmune process has been established (47, 48, 49). It is suggested that the mechanism of tolerance is similar to that induced by the oral route.

Intravenous administration of both insulin and soluble GAD have been reported to decrease immune responses, reducing but not preventing insulitis and diabetes (50, 51). The continued debate as to whether this delivery route of insulin has its effect by decreasing the need for insulin synthesis (ß cell rest), or whether it only has an immune mechanism, is not yet formally resolved. The former mechanism is supported by the effectiveness of diazoxide, which inhibits insulin secretion in BB rats (52). However, the fact that nonbioactive insulin peptides also provide a degree of protection supports the latter mechanism (34, 53, 54).

Other forms of vaccination have also been attempted in experimental animals. Whole T-cell vaccination is reported to be successful in murine experimental autoimmune models (29). Vaccination with T cells specific for a peptide of the 65-kDa heat shock protein was the first to be attempted, with success in prevention of diabetes in NOD mice (55, 56). Subsequently, reports have been published showing delay in the onset and reduced incidence of NOD diabetes using lymphocytes obtained from NOD mouse spleens, with similar results (57). This type of research, as well as vaccination with TCR peptide vaccines and DNA vaccination, are in their infancy, but very exciting data are accumulating (33, 53).

Another form of therapy that was initially reported to be effective in NOD mice is the vitamin B compound nicotinamide (59, 60). The protective effect of nicotinamide reported in an alloxan- and streptozotocin-induced diabetes, as well as in NOD mice, has led to a number of intervention trials with this substance in humans (60, 61). It is interesting that nicotinamide apparently delayed rather than prevented both insulitis and IDDM onset and that it was less, if at all effective, in a number of studies in the BB rat model (60).

A very different form of intervention developed from studies first in BB rats and later NOD mice, demonstrated very robust protection from IDDM by manipulating the weaning diet (62). Such studies delineated variable degrees of diabetogenicity for a variety of foods, including cow milk proteins (62, 63). Although mechanisms remain uncertain (39, 64, 65), there is good consensus that a nonantigenic, hydrolyzed casein diet reduces insulitis and prevents IDDM in BB rats (66) and in NOD mice (67).

The difference in the effect of nicotinamide in oral insulin in these animal studies raises the issue as to whether data obtained in autoimmune or chemically induced models of IDDM in mice or rats or transgenic or adoptive transfer models of diabetes are directly analogous to the development of Type 1 diabetes in humans. If not, strategies that are effective in these animals may fail in humans or even act as immunogens and accelerate disease. Some work in animals raises concerns regarding the right dose of antigens that are effective, the exact timing, and the potential for acceleration of disease. Thus, although it seems rational to test and prove intervention therapies in relevant animal models, this does not translate into immediate success (or promise of success).

	Caveats from animal research

Top Introduction What do we know?... What do we know?... Caveats from animal research IDDM prevention trials in... Current IDDM prevention trials Nicotinamide trial The DPT-1 Trial The Nutritional Prevention of... Human intervention studies: if... References

 
The first difficulty is related to the timing of the intervention. The majority of interventions in animal models have been initiated shortly after weaning around the time of very early peri-insulitis and well before the beginning of invasive insulitis, which in NOD mice is signaled by the appearance of autoantibodies, such as ICA (68). Indeed, nearly all the experimental strategies that can prevent diabetes halt the prediabetic process at the peri-insulitis stage and prevent islet invasion (30, 69). Effective later interventions are, at best, rare (70).

Presumably, the presence of circulating ICAs reflects destructive islet invasion. Could immune intervention at this stage accelerate (immunize) rather than prevent (tolerize) disease? The first such model to be described was the induction of diabetes in rabbits by the injection of insulin in Freund’s complete adjuvant (71). A number of different rodent strains did not show this reaction, and, in striking contrast to the animal models for multiple sclerosis (72), the precipitation of diabetes by autoantigen is very unusual. However, accelerated IDDM was observed in one of several published studies in NOD mice carrying GAD65 transgenes (73), and iv administration of an immunodominant peptide from ICA69 accelerated disease in adoptive transfer and cyclophosphamide-induced diabetes models (74). Similarly, protocols of oral insulin administration effective in NOD mice (and considered for human use), exacerbate diabetes development in BB rats (75). Mice that express transgenic ovalbumin in the islet, develop cytolytic T cells and diabetes on autoantigen feeding with ovalbumin (76). Immunotherapy can also have beneficial effects that fade over time and lead to exacerbated recurrence of autoimmunity (77). It should, perhaps, not come as a surprise that immunotherapy is a two-sided sword with beneficial, as well as adverse potential, outcome. Therefore, the understanding of detailed mechanism should be a prerequisite for diabetes intervention therapies.

A major area of concern is the autoantigen dose response and its transfer from rodent to human. For example, in one NOD mouse study, 1 mg oral insulin conferred protection from disease, whereas 5 mg resulted in acceleration (40). In a slowly progressive, virus-dependent IDDM model, the expression of IDDM was decreased by 50% in mice receiving 1 mg insulin twice a week for 2 months, whereas a dose of 0.1 mg had no effect. Larger doses were not examined (44). A more recent report showed no effect of 2 mg oral insulin in protecting BB rats from developing diabetes (75). On a body mass basis, this dose is equivalent to only a quarter of that protective in NOD mice in the hands of the same investigators (78). Such examples fail to differentiate species differences from autoantigen dose effects. The importance of optimal doses and delivery schedules is emphasized by the data showing that only a high dose of insulin equivalent to 21 U/kg (about 0.88 mg) given sc 5 days a week, resulted in some diabetes protection, whereas a dose of 0.3 U/kg had no effect. This latter dose is similar to that used in current human intervention studies (79). It is reasonable to assume that if autoantigen therapy is aimed to ablate autoreactive T-cell pools through peripheral tolerance mechanisms, then high doses will target high- and low-affinity clones, whereas lower doses will spare T cells with lower affinity. Also, such treatments will preferentially target CD4+ T cells, but be much less efficient in targeting CD8+ T cells, which clearly are critical elements of progressive prediabetes (3, 80).

The pharmacokinetics and bioavalability of administered autoantigen must be considered. Even small amounts of insulin conjugated to cholera toxin B subunit were able to prevent clinical diabetes in NOD mice (81). However, caution was raised by the exacerbation of disease in BB rats when oral insulin was administered with a bacterial adjuvant. This effect using a similar oral immunization protocol was not seen in NOD mice, emphasizing the potential differences in responses among different species (75).

Exacerbation of disease has been reported among other models of autoimmune disease and diabetes. Autoantigen treatment in a marmoset model of multiple sclerosis induced an initial delay in disease expression, followed by exacerbation (77). The oral administration of myelin basic protein, when fed in low doses, also resulted in the exacerbation of experimental allergic encephalomyelitis in mice (82).

These and other studies demonstrate the critical role of antigen dose and timing of delivery in determining whether tolerization or immunization occurs. This is of major relevance in transferring knowledge obtained from research in a variety of animal models to humans with heterogeneous genetic, immunologic, and age characteristics. A major challenge will be the calculation of an appropriate antigen dose for a 5, 25, or 75 kg human from that which is either effective or dangerous in a 25-g mouse or 250-g rat. Additional research is needed to optimize autoantigen delivery, particularly in terms of relatively small doses, and also to expand our mechanistic insight. Specifically, we need to understand deviation from beneficial effects and to compare the two available diabetes models as a prelude for the big step into the human system.

	IDDM prevention trials in humans

Top Introduction What do we know?... What do we know?... Caveats from animal research IDDM prevention trials in... Current IDDM prevention trials Nicotinamide trial The DPT-1 Trial The Nutritional Prevention of... Human intervention studies: if... References

 
Prevention trials can be classified into three categories. First, primary prevention trials aim to forestall disease in high-risk populations, with initiation well before any measurable evidence of prediabetic autoimmunity appears. The identification of subjects will rely on family history and/or immunogenetics. Primary prevention probably targets the processes preceding and/or coinciding with peri-insulitis. Second, secondary prevention is aimed at delay and possibly suppression of progressive ß-cell destruction in euglycemic subjects who have clear signs of autoimmunity. This coincides with invasive insulitis, and actual tissue damage is signaled by the presence of increasing numbers of autoantibodies. Subjects may be classified into those with very high, intermediate, or low risk for development of IDDM. Although not yet tested prospectively, and still technically challenging, routine analysis of autoreactive T cells may well become an important part of the identification and monitoring of such subjects, and relatively short-lived T cells may allow detection of intervention success (or failure) well before overt disease. Third, tertiary prevention is initiated after the clinical onset of insulin deficiency, aimed at inducing prolonged remission or allowing ß cell regeneration in the face of suppressed autoimmunity. Preservation of residual ß cell function is measured by circulating C-peptide levels.

Tertiary intervention studies are the logical starting point to test the safety, and perhaps efficacy, of an intervention strategy, particularly if it has potential toxic effects. Because autoimmunity continues unabated after disease onset (83, 84), if started early enough, there is some optimism that such efforts may be effective. However, immunotherapy with potent immunosuppressants, including cyclosporin, showed only transient efficacy in human tertiary intervention trials in older subjects and minimal, if any, in children, especially with delay in diagnosis (85, 86). Side effects limit the use of such immunosuppressants, although some ministudies have been initiated in high-risk children (87). Other strategies, such as the use of nicotinamide in newly diagnosed patients, were only marginally effective, although a meta-analysis did show some preservation of C-peptide secretion (88). Results were not sufficiently compelling to encourage wide-spread use of nicotinamide after the onset of clinical Type 1 diabetes. However, these results, together with a pilot study in ICA-positive first-degree relatives (89), prompted current multinational secondary prevention studies using nicotinamide (16, 17, 61, 90, 91). The positive effect of intensive insulin therapy on preservation of C-peptide levels in both new onset children with Type 1 diabetes (iv insulin infusions) and the long-term DCCT study (92, 93) led to two small pilot studies of insulin therapy as secondary prevention in subjects thought to be at very high risk for the development of Type 1 diabetes. The first study was nonrandomized, comparing high-risk volunteers as assessed by a dual parameter predictor model (94) for the study with refusers (95, 96). One of five children, aged 8–14 yr, treated with insulin (iv and sc), developed IDDM in 2.3–3.3 yr. Seven controls (children as well as adults) all developed diabetes within 2.5 yr (95). However, the life table analysis of the control group showed much more rapid progression to diabetes than has been our experience (24). In an update of this study, all eight nontreated relatives became diabetic within 3 yr compared with two of nine insulin-treated patients, with the longest follow-up of 9 yr (28). A protective effect of a similar treatment regimen in high-risk first-degree relatives has been reported in a small randomized study in Germany (96). These studies were the prelude to the current, large American Diabetes Prevention Trial–Type 1 diabetes (DPT-1) Intervention Trial (15, 16, 97).

	Current IDDM prevention trials

Top Introduction What do we know?... What do we know?... Caveats from animal research IDDM prevention trials in... Current IDDM prevention trials Nicotinamide trial The DPT-1 Trial The Nutritional Prevention of... Human intervention studies: if... References

 
Current Type 1 diabetes prevention trials with their design and inclusion criteria have recently been reviewed (16, 17, 28, 97). There were pros and cons to the initiation of each of these studies. One relatively small randomized study using oral nicotinamide ended early because of lack of effect (91): this DENIS study was designed to detect an 80% reduction in the incidence of IDDM (from 30–6%). Thus, the failure to achieve this goal, while showing that nicotinamide is not a panacea, does not exclude possibility of success by less stringent criteria such as a 50% reduction in incidence or even a delay in the onset of insulin requirement. Statistical power is the coin of kings in clinical trials inevitably run by commoners pressed for time.

	Nicotinamide trial

Top Introduction What do we know?... What do we know?... Caveats from animal research IDDM prevention trials in... Current IDDM prevention trials Nicotinamide trial The DPT-1 Trial The Nutritional Prevention of... Human intervention studies: if... References

 
Like the above DENIS study (91), the first population-based nicotinamide intervention trial in New Zealand has suffered criticism because of unconventional study design that randomized children in schools rather than as individuals (90). The design of the study is similar to that frequently used in population testing of vaccines. Recent results from this trial suggest a 56% protective effect of nicotinamide after an average follow-up of 7 yr.

Data from this study and apparent successful delay in IDDM in the NOD mouse, but not the BB rat, led to the design and implementation of the ENDIT study (16, 17, 61). This admirably large multicenter secondary intervention trial in ICA-positive (20 JDF units) first-degree relatives is a randomized, double-blind, placebo-controlled attempt to decrease the development of IDDM by 35% in 5 yr. The study will be unblinded in the year 2002. The major questions regarding this trial relate to the adequacy of relatively low nicotinamide dosing in humans compared to NOD mice. The second potential concern is that nicotinamide contains nicotinic acid, which has been shown to cause insulin resistance in individuals with islet autoimmunity (98), although increases in first-phase insulin secretion were not seen in the doses used in this study (99).

The ENDIT study was initiated before the establishment of current GAD65 and IA2 autoantibody assays and without using immunogenetic risk markers that may have helped to identify the population at highest risk for developing IDDM within 5 yr. Our decision not to participate in this study at the time was based on our conviction that we needed to develop better markers of rapid progression to clinical IDDM among ICA-positive individuals before embarking on the very long journey of a significant intervention trial: this is the crux—the haste to find prevention that we all share, is mitigated by the long-term commitment to a given course.

	The DPT-1 Trial

Top Introduction What do we know?... What do we know?... Caveats from animal research IDDM prevention trials in... Current IDDM prevention trials Nicotinamide trial The DPT-1 Trial The Nutritional Prevention of... Human intervention studies: if... References

 
This trial started in 1994 and is the largest, most ambitious, multicenter intervention study initiated, to date (15, 17, 97). There are two arms to this study of first- and second-degree relatives, which both aim at a reduction of IDDM by 50% within 4 yr. Very high-risk subjects who are 3–45 yr of age with ICA 10 JDF units of or greater, low FPIR, and do not have a DQB1*0602 HLA allele are randomized to receive either iv and sc insulin or no therapy. In another arm, oral insulin is administered in a randomized, controlled, double-blind manner to subjects that have an FPIR above the 10th percentile.

In 1994, there was a great deal of excitement generated by significant insulin-mediated diabetes protection in four BB rat (52, 100, 101, 102) and two NOD mouse studies (103, 104). In addition, the first publications of similar success with oral insulin administration were reported in experiments in NOD mice from two groups (27, 40). Although the possibility of some untoward effects of oral tolerization were entertained, many of the experiments cited above were published only after the study began.

In Pittsburgh, our concerns with the study, then and now, were that insulin injections (which most children hate) would be administered twice a day to a group of people, half of whom were destined not to develop diabetes during the time of the study without any therapy. Our studies emphasize clinically significant decrements in mental efficiency during even mild hypoglycemia (60 mg/dL) in children with Type 1 diabetes (105). Despite the relatively small dose of insulin to be administered sc (0.25 u/kg·day of ultralente insulin), clinical experience has demonstrated that even this dose can induce significant, if not severe, hypoglycemia in newly diagnosed children with IDDM. Thus, it was possible that the first-degree relatives in this study with altered insulin secretary dynamics, may well suffer from frequent episodes of mild, either symptomatic or asymptomatic hypoglycemia, which could be, at best, uncomfortable and, at worst, affect school performance. In 1994, we were beginning to assess the effect of mild hypoglycemia using insulin glucose clamps in nondiabetic adolescents, but data were not yet available. Our most recent analysis suggests that decrements in mental efficiency during mild hypoglycemia (60 mg/dL) in normal adolescents does not cause significant impairment of mental efficiency, unlike in their diabetic counterparts (106). Young children, who have a greater degree of hypoglycemia susceptibility, were not included in our study. Early data from the parental insulin intervention studies suggest that mild hypoglycemia does occur in these subjects, but does not seem to have major consequences, although there was no formal testing (97).

A criticism of DPT-1 is that the insulin doses used are miniscule in comparison with those required in animal models to obtain a significant effect. In the oral insulin arm of DPT-1 (7.5 mg/day) in a 25-kg child is less than 1% of the dose effective in NOD mice. The fear in these studies is that the perceived immunological manipulation may actually accelerate disease because its timing of the intervention is late in the prediabetes phase when few, if any, interventions have been successful in animals. Whereas insulin (3) or proinsulin (107, 108, 109) are definite targets of prediabetic autoimmunity, our understanding of their role in progressive prediabetes is incomplete (75). We hope to gather mechanistic insights from this late prediabetes trial that could guide future efforts.

	The Nutritional Prevention of IDDM Trial (TRIGR)

Top Introduction What do we know?... What do we know?... Caveats from animal research IDDM prevention trials in... Current IDDM prevention trials Nicotinamide trial The DPT-1 Trial The Nutritional Prevention of... Human intervention studies: if... References

 
This decade-old effort has generated observations of a protective effect of nonantigenic hydrolyzed weaning formulas in BB rats (66) and NOD mice (67), as well as a large body of epidemiologic evidence which, cumulatively suggests a critical, potentially diabetogenic role of the weaning diet in infants with genetic diabetes risk (17, 110, 111, 112). The accumulated data in multiple epidemiologic studies in children led to the design of the TRIGR cow milk avoidance protocol (17, 113). Spearheaded in Finland, pilot studies tested whether strict avoidance of complex protein diets, such as common cow milk-based formulas in the first 6 months of life, would prevent the development of diabetic autoimmunity and/or overt disease in genetically susceptible newborn infants. This is the first primary intervention study, to date. A first pilot study established the study feasibility (114, 115) and led to a second nationwide pilot study with the development of antibodies to islet cell antigens over the first 2 yr of life as an end point.

This blinded, placebo-controlled prospective trial randomizes newborn first-degree relatives with high genetic risks to a weaning diet of either standard formula or a hydrolyzed casein formula. The preliminary results presented this year (116) show significantly decreased autoantibodies in the group fed hydrolyzed formulas. These results indicate that a large sufficiently powered multicenter study is warranted to prove or disprove whether avoidance of complex weaning diets, such as cow milk formula, has an effect on the development of IDDM in humans.

The attraction of this study is that it is safe and uses an intervention that has been in general pediatric use for decades. As a primary intervention strategy, it could be applied to the general population (117). It is important to recognize that the intervention is a hydrolyzed formula rather than a soy-based formula, which also contains complex antigens. The criticism of the study is that the epidemiological data are controversial (17). Also, the study is difficult to implement, requiring cooperation of newborn nursery staff, trial personnel, and the families involved. However, the Finnish group has clearly overcome many of these obstacles and has shown that such a study is possible. The fact that there is no association between the emergence of autoantibodies and exposure to cow milk formulas in current prospective epidemiological studies should not negate the possible detrimental role of cow milk protein (118, 119). The experience with the discovery of the association of celiac disease with gluten intake serves as a lesson that, when an environmental agent is ubiquitous, it will have to be removed to prove a role.

	Human intervention studies: if not now, when and how?

Top Introduction What do we know?... What do we know?... Caveats from animal research IDDM prevention trials in... Current IDDM prevention trials Nicotinamide trial The DPT-1 Trial The Nutritional Prevention of... Human intervention studies: if... References

 
The fact that our group has not joined any of the current secondary intervention efforts was based on the many concerns outlined above. We feel strongly that continued longitudinal epidemiological studies in both high-risk individuals and the general population are mandatory to delineate specific risk markers that will identify those individuals with signs of autoimmunity who will or who will not progress to insulin deficiency over a defined period of time. Although ICAs are very powerful predictors of Type 1 diabetes, they are not sufficiently specific. Major strides have been made in identifying high-risk and protective genotypes, a role for the presence of three autoantibodies and risks associated with low FPIR. However, with all these tools, about half the subjects enrolled in current secondary prevention studies would not have developed diabetes, even without an intervention. In addition, about 10% of subjects thought to be at low risk will develop IDDM, and they represent a sizable cohort.

There is no doubt that we will learn from post hoc analyses of the ENDIT and DPT-1 studies. Additional studies should be based on lessons learned once all autoantibodies are assayed in the baseline samples of these participants and genotyping is available. Although there will be insufficient statistical power to assess effectiveness of the interventions in subgroups of these studies, there is a lot to be learned from the additional data that would have included or excluded participants using more stringent selection criteria. Only then will we know whether the initiation of these studies should have been delayed until newer assays were available to assist in selection. Would other markers allow identification of a group of first-degree relatives with at least a 50% risk of developing IDDM without decreased insulin secretion, which may be a stage that is too late for intervention? In the year 2002, we will find out whether or not these secondary intervention trials in progress at this time should have been delayed a few years. Also, it is possible that the observation period in both studies should be prolonged to assess the effect of the interventions.

A subject of much debate is the application of knowledge we have obtained in relatives of Type 1 diabetic patients to the general population. Because 90% of the patients with Type 1 diabetes do not have a family history, we will only make inroads into the incidence of the disease if the general population is studied. There is some degree of controversy as to whether progression to IDDM occurs in the same manner in ICA-positive individuals from the general population as has been shown in first-degree relatives (120). The general population studies currently underway, therefore, are of major significance. Any intervention applied to the general population of apparently healthy individuals, needs to be safe, as well as effective. The same criteria should be used in any study of children whose decisions are made for them by their parents. It has been said that if an intervention was as safe as giving an aspirin it should be used to prevent diabetes, even if it means giving this to people without major risk for disease. However, aspirin is a good example of a drug that is usually safe in adults, but can be very dangerous in children. It still is implicated in the pathogenesis of Reye’s syndrome. Another example is the effect of hypoglycemia on the developing brain, which is likely to be major in children under 5 years of age (121, 122). Only sufficiently large studies have the power to detect relatively rare, but possibly dangerous, side effects of any therapy (123). The younger the subjects included in the study, the more cautious one should be in introducing an intervention and following both beneficial and detrimental effects. The fact that interventions are safe in animals or adults does not ensure that this is the case in children.

There is a lot to be learned regarding dosing and timing of prevention therapies that are currently being investigated in both animal models and humans. Data are accumulating steadily, and, with time, direction for future interventions will become apparent. The ability to identify individuals with high genetic risk for the development of IDDM among first-degree relatives and probably in the general population, makes early intervention increasingly feasible. As most Caucasians develop autoantibodies, these may serve as reasonable surrogate markers of the effectiveness of a primary intervention strategy. The potential protective effect of weaning from breast feeding to a hydrolyzed cow milk formula currently seems to be the most attractive next step in designing an intervention trial. There is abundant animal data, there is suggested epidemiological data, and the intervention is safe.

Primary or secondary intervention strategies in newborns or young children should be undertaken with great caution. Results of current trials should be available before drug therapy is attempted in any future trials in any other population. The risks of the intervention should be weighed clearly against the risks and discomfort of the treatment of diabetes, as we approach the next millennium. The ultimate prognosis of a child with diabetes has vastly improved over the last two decades and can improve further with the availability of adequate health care dollars. But this is not an easy life, and a safe prevention should be the ultimate goal for all of us. We should move ahead with determination, but with steady evaluation of the advantages and disadvantages of every step we take. New and exciting animal studies are ongoing that include gene therapy, DNA vaccination, and a variety of forms of immunomodulation. The day will come when there is sufficient evidence for new therapies to be applied, first in consenting adults, but the time is not yet here. We must hurry—but slowly and carefully.

	Footnotes

 
¹ Supported by NIH Grants RO1-DK24021, RO1–46864, and 5MO1-RR00084 and the Renziehausen Fund.

	References

Top Introduction What do we know?... What do we know?... Caveats from animal research IDDM prevention trials in... Current IDDM prevention trials Nicotinamide trial The DPT-1 Trial The Nutritional Prevention of... Human intervention studies: if... References

 

1. Bach JF. 1994 Insulin-dependent diabetes mellitus as an autoimmune disease. Endocr Rev. 15:516–542.[Abstract]
2. Fox CJ, Danska JS. 1997 IL-4 expression at the onset of islet inflammation predicts nondestructive insulitis in nonobese diabetic mice. J Immunol. 158:2414–2424.[Abstract]
3. Wong FS, Karttunen J, Dumont C, et al. 1999 Identification of an MHC class I-restricted autoantigen in type 1 diabetes by screening an organ-specific cDNA library (see comments). Nat Med. 5:1026–1031.[CrossRef][Medline]
4. Shimada A, Imazu Y, Morinaga S, et al. 1999 T-cell insulitis found in anti-GAD65+ diabetes with residual ß-cell function. Diabetes Care. 22:615–617.[Abstract]
5. Bottazzo GF, Florin-Christensen A, Doniach D. 1974 Islet cell antibodies in diabetes mellitus with autoimmune polyendocrine deficiencies. Lancet. 2:1279–1283.[Medline]
6. Lipton RB, Atchison J, Dorman JS, et al. 1992 Genetic, immunological, and metabolic determinants of risk for type 1 diabetes mellitus in families. Diabetes Med. 9:224–232.[Medline]
7. Kuller LH, Becker DJ, Cruickshanks KJ, et al. 1990 Evolution of the Pittsburgh studies of the epidemiology of insulin-dependent diabetes mellitus. Pittsburgh Diabetes Epidemiology and Etiology Research Group. Genet Epidemiol. 7:105–119.[CrossRef][Medline]
8. Lipton RB, Kocova M, LaPorte RE, et al. 1992 Autoimmunity and genetics contribute to the risk of insulin-dependent diabetes mellitus in families: islet cell antibodies and HLA DQ heterodimers. Am J Epidemiol. 136:503–512.[Abstract/Free Full Text]
9. Pietropaolo M, Becker DJ, Dorman JS, et al. 1999 Are GAD65 and IA-2 autoantibodies sufficient to predict type 1 diabetes? Diabetes. 48:A45.
10. Bingley PJ, Christie MR, Bonifacio E, et al. 1994 Combined analysis of autoantibodies improves prediction of IDDM in islet cell antibody-positive relatives. Diabetes. 43:1304–1310.[Abstract]
11. Bingley PJ. 1996 Interactions of age, islet cell antibodies, insulin autoantibodies, and first-phase insulin response in predicting risk of progression to IDDM in ICA+ relatives: the ICARUS data set. Islet Cell Antibody Register Users Study. Diabetes. 45:1720–1728.[Abstract]
12. Nepom GT. 1993 Immunogenetics and IDDM. Diabetes Rev. 1:93–103.
13. Sheehy MJ. 1992 HLA and insulin-dependent diabetes. A protective perspective. Diabetes. 41:123–129.[Abstract]
14. Ettinger RA, Liu AW, Nepom GT, Kwok WW. 1998 Exceptional stability of the HLA-DQA1*0102/DQB1*0602 ß protein dimer, the class II MHC molecule associated with protection from insulin-dependent diabetes mellitus. J Immunol. 161:6439–6445.[Abstract/Free Full Text]
15. DPT-1 Study Group. 1995 The Diabetes Prevention Trial–type 1 diabetes (DPT-1): implementation of screening and staging of relatives. Transplant Proc. 27:337.
16. Muir A, Schatz D. 1995 Prevention of insulin-dependent diabetes–1995. Trends Endocrinol Metab. 6:312–317.[Medline]
17. Knip M, Akerblom HK. 1998 IDDM prevention trials in progress—a critical assessment. J Pediatr Endocrinol Metab. 11:371–377.
18. Vardi P, Brik R, Barzilai D. 1991 Insulin autoantibodies: reflection of disturbed self-identification and their use in the prediction of type I diabetes. Diabetes Metab Rev. 7:209–222.[Medline]
19. Vardi P, Crisa L, Jackson RA, et al. 1991 Predictive value of intravenous glucose tolerance test insulin secretion less than or greater than the first percentile in islet cell antibody positive relatives of type 1 (insulin-dependent) diabetic patients. Diabetologia. 34:93–102:1991.[CrossRef][Medline]
20. Arslanian S, Becker DJ, Rabin B, et al. 1985 Correlates of insulin antibodies in newly diagnosed children with insulin-dependent diabetes prior to therapy. Diabetes. 34:926–930.[Abstract]
21. Riley WJ, Maclaren NK, Krischer J, et al. 1990 A prospective study of the development of diabetes in relatives of patients with insulin-dependent diabetes. N Eng J Med. 232:1167–1172.
22. Verge CF, Gianani R, Kawasaki E, et al. 1996 Prediction of type 1 diabetes mellitus in first degree relatives using a combination of insulin, glutamic acid decarboxylase and ICA 512bdc/IA-2 autoantibodies. Diabetes. 45:926–933.[Abstract]
23. Leslie RDG, Atkinson MA, Notkins AL. 1999 Autoantigens IA-2 and GAD in Type 1 (insulin-dependent) diabetes. Diabetologia. 42:3–14.[CrossRef][Medline]
24. Becker D, Cakan N, LaPorte R, Drash A, Trucco M. 1995 High risk DQ alleles improve IDDM prediction in ICA+ve first-degree relatives with decreased insulin secretion. Diabetologia. 38(Suppl 1):A42.
25. Pietropaolo M, Trucco M, Cakan N, Drash AL, LaPorte RE, Becker DJ. 1996 Autoantibodies to GAD65 and ICA512/IA-2 improve IDDM prediction in first-degree relatives with decreased first phase insulin response and ICA positivity. Autoimmunity. 24(Suppl 1):51.
26. Bach JF. 1994 Predictive medicine in autoimmune diseases: from the identification of genetic predisposition and environmental influence to precocious immunotherapy. Clin Immunol Immunopathol. 72:156–161.[CrossRef][Medline]
27. Muir A, Ramiya V. 1996 New strategies in oral immunotherapy for diabetes prevention. Diabetes Metab Rev. 12:1–14.[CrossRef][Medline]
28. Simone EA, Wegmann DR, Eisenbarth GS. 1999 Immunologic "vaccination" for the prevention of autoimmune diabetes (type 1A). Diabetes Care. 22(Suppl 2):B7–B15.
29. Atkinson MA, Leiter EH. 1999 The NOD mouse model of type 1 diabetes: as good as it gets? Nat Med. 5:601–604.[CrossRef][Medline]
30. Sadelain MW, Qin HY, Sumoski W, Parfrey N, Singh B, Rabinovitch A. 1990 Prevention of diabetes in the BB rat by early immunotherapy using Freund’s adjuvant. J Autoimmun. 3:671–680.[CrossRef][Medline]
31. Sadelain MW, Qin HY, Lauzon J, Singh B. 1990 Prevention of type 1 diabetes in NOD mice by adjuvant immunotherapy. Diabetes. 39:583–589.[Abstract]
32. Harada M, Kishimoto Y, Makino S. 1990 Prevention of overt diabetes and insulitis in NOD mice by a single BCG vaccination. Diabetes. 8:85–90.
33. Elias D, Reshef T, Birk OS, van der Zee R, Walker MD, Cohen IR. 1991 Vaccination against autoimmune mouse diabetes with a T-cell epitope of the human 65-kDa heat shock protein. Proc Natl Acad Sci USA. 88:3088–3091.[Abstract/Free Full Text]
34. Muir A, Peck A, Clare-Salzler M, et al. 1995 Insulin immunization of nonobese diabetic mice induces a protective insulitis characterized by diminished intraislet interferon-gamma transcription. J Clin Invest. 95:628–634.
35. Daniel D, Wegmann DR. 1996 Protection of nonobese diabetic mice from diabetes by intranasal or subcutaneous administration of insulin peptide B-(9–23). Proc Natl Acad Sci USA. 93:956–960.[Abstract/Free Full Text]
36. Tisch R, Liblau RS, Yang XD, Liblau P, McDevitt HO. 1998 Induction of GAD65-specific regulatory T-cells inhibits ongoing autoimmune diabetes in nonobese diabetic mice. Diabetes. 47:894–899.[Abstract]
37. Elliot J, Qin H, Bhatti S, et al. 1994 Immunization with the larger form of mouse glutamic acid decarboxylase (GAD67) prevents autoimmune diabetes in NOD mice. Diabetes. 43:1494–1499.[Abstract]
38. Ramiya VK, Lan MS, Wasserfall CH, Notkins AL, Maclaren NK. 1997 Immunization therapies in the prevention of diabetes. J Autoimmun. 10:287–292.[CrossRef][Medline]
39. Karges W, Hammond-McKibben D, Gaedigk R, Shibuya N, Cheung R, Dosch H-M. 1997 Loss of self-tolerance to ICA69 in non-obese diabetic mice. Diabetes. 46:1548–1556.[Abstract]
40. Zhang ZJ, Davidson L, Eisenbarth G, Weiner HL. 1991 Suppression of diabetes in nonobese diabetic mice by oral administration of porcine insulin. Proc Natl Acad Sci USA. 88:10252–10256.[Abstract/Free Full Text]
41. Bergerot I, Fabien N, Maguer V, Thiovolet C. 1994 Oral administration of human insulin to NOD mice generates CD4+ T cells that suppress adoptive transfer of diabetes. J Autoimmun. 7:655–663.[CrossRef][Medline]
42. Ploix C, Bergerot I, Fabien N, Perche S, Moulin V, Thivolet C. 1998 Protection against autoimmune diabetes with oral insulin is associated with the presence of IL-4 type 2 T-cells in the pancreas and pancreatic lymph nodes. Diabetes. 47:39–44.[Abstract]
43. Sai P, Damage C, Rivereau AS, Hoeltzel A, Gouin E. 1996 Prophylactic oral administration of metabolically active insulin entrapped in isobutylcyanoacrylate nanocapsules reduces the incidence of diabetes in nonobese diabetic mice. J Autoimmun. 9:713–722.[CrossRef][Medline]
44. vonHerrath MG, Dyrberg T, Goldstone MBA. 1996 Oral insulin treatment suppresses viral-induced antigen-specific destruction of beta cells and prevent autoimmune diabetes in transgenic mice. J Clin Invest. 98:1324–1331.[Medline]
45. Mordes JP, Schirf B, Roipko D, et al. 1996 Oral insulin does not prevent insulin-dependent diabetes mellitus in the BB rat. Ann NY Acad Sci. 778:418–421.[CrossRef][Medline]
46. Bergerot I, Fabien N, Mayer A, Thivolet C. 1996 Active suppression of diabetes after oral administration of insulin is determined by antigen dosage. Ann N Y Acad Sci. 778:362–367.[Abstract]
47. Daniel D, Wegmann DR. 1996 Intranasal administration of insulin peptide B: 9–23 protects NOD mice from diabetes. Ann N Y Acad Sci. 778:371–372.[Medline]
48. Maron R, Melican NS, Weiner HL. 1999 Regulatory Th2-type T cell lines against insulin and GAD peptides derived from orally- and nasally-treated NOD mice suppress diabetes. J Autoimmun. 12:251–258.[CrossRef][Medline]
49. Harrison LC, Dempsey-Collier M, Kramer DR, Takahashi K. 1996 Aerosol insulin induces regulatory CD8 T-cells that prevent murine insulin-dependent diabetes. J Exp Med. 184:2167–2174.[Abstract/Free Full Text]
50. Hutchings PR, Cooke A. 1995 Comparative study of the protective effect afforded by intravenous administration of bovine or ovine insulin to young NOD mice. Diabetes. 44:906–910.[Abstract]
51. Kaufman DL, Clare-Salzler M, Tian J, et al. 1993 Spontaneous loss of T cell tolerance to glutamic acid decarboxylase in murine insulin dependent diabetes. Nature. 366:69–72.[CrossRef][Medline]
52. Vlahos WD, Seemayer TA, Yale JF. 1991 Diabetes prevention in BB rats by inhibition of endogenous insulin secretion. Metabolism. 40:825–829.[CrossRef][Medline]
53. Daniel C, Grakoui A, Allen PM. 1998 Inhibition of an in vitro CD4+ T cell alloresponse using altered peptide ligands. J Immunol. 160:3244–3250.[Abstract/Free Full Text]
54. Karounos DG, Bryson JS, Colhen DA. 1997 Metabolically inactive insulin analogue prevents type 1 diabetes in prediabetic NOD mice. J Clin Invest. 100:1344–1348.[Medline]
55. Elias D, Marcus H, Reshef T, Ablamunits V, Cohen IR. 1995 Induction of diabetes in standard mice by immunization with the p277 peptide of a 60-kDa heat shock protein. Eur J Immunol. 25:2851–2857.[Medline]
56. Elias D, Cohen IR. 1995 Treatment of autoimmune diabetes and insulitis in NOD mice with heat shock protein 60 peptide p277. Diabetes. 44:1132–1138.[Abstract]
57. Smerdon RA, Peakman M, Hussain MJ, Vergani D. 1993 Lymphocyte vaccination prevents spontaneous diabetes in the non-obese diabetic mouse. Immunology. 80:498–501.[Medline]
58. Coon B, An LL, Lindsay WJ, vonHerrath MG. 1999 DNA immunization to prevent autoimmune diabetes. J Clin Invest. 104:189–194.[Medline]
59. Reddy S, Bibby NJ, Elliott RB. 1990 Early nicotinamide treatment in the NOD mouse: effects on diabetes and insulitis suppression and autoantibody levels. Diabetes Res. 15:95–102.[Medline]
60. Kolb H, Burkhart V. 1999 Nicotinamide in type 1 diabetes: mechanism of action revisited. Diabetes Care. 22(Suppl 2):B16–B20.
61. Gale EA. 1996 Practice of nicotinamide trials in pre-type 1 diabetes. J Pediatr Endocrinol Metab. 9:375–379.[Medline]
62. Karges W, Dosch H-M. 1996 Environmental factors: cow milk and others. In: Palmer JP, ed. Diabetes prediction, prevention and genetic counselling in IDDM. Chichester, UK: John Wiley and Sons Ltd.; 167–180.
63. Scott FW, Norris JM, Kolb H. 1996 Milk and type I diabetes: examining the evidence and broadening the focus. Diabetes Care. 19:379–383.[Medline]
64. Scott FW, Cloutier HE, Kleemann R, et al. 1997 Potential mechanisms by which certain foods promote or inhibit the development of spontaneous diabetes in BB rats: dose, timing, early effect on islet area, and switch in infiltrate from Th1 to Th2 cells. Diabetes. 46:589–598.[Abstract]
65. Malkani S, Nompleggi D, Hansen JW, Greiner DL, Mordes JP, Rossini AA. 1997 Dietary cow’s milk protein does not alter the frequency of diabetes in the BB rat. Diabetes. 46:1133–1140.[Abstract]
66. Elliott RB, Martin JM. 1984 Dietary protein: a trigger of insulin-dependent diabetes in the BB rat? Diabetologia. 26:297–299.[Medline]
67. Karges W, Hammond-McKibben D, Cheung RK, et al. 1997 Immunological aspects of nutritional diabetes prevention in NOD mice. A Pilot study for the cow’s milk-based IDDM Prevention Trial. Diabetes. 46:557–564.[Abstract]
68. Reddy S, Bibby N, Elliott RB. 1990 Longitudinal study of islet cell antibodies and insulin autoantibodies and development of diabetes in non-obese diabetic (NOD) mice. Clin Exp Immunol. 81:400–405.[Medline]
69. Karges W, Ilonen J, Robinson BH, Dosch H-M. 1995 Self and non-self antigen in diabetic autoimmunity: molecules and mechanisms. Mol Aspects Med. 16:79–213.[CrossRef][Medline]
70. Elias D, Cohen IR. 1994 Peptide therapy for diabetes in NOD mice (see comments). Lancet. 343:704–706.[CrossRef][Medline]
71. Titlbach M, Nouza K, Mullerova M, Pekarek J. 1988 Immunodiabetes in rabbits. Endocrinol Exp. 22:3–18.[Medline]
72. Owens T, Sriram S. 1995 The immunology of multiple sclerosis and its animal model, experimental allergic encephalomyelitis (Review; 170 refs.). Neurol Clin. 13:51–73.[Medline]
73. Geng L, Solimena M, Flavell RA, Sherwin RS, Hayday AC. 1998 Widespread expression of an autoantigen-GAD65 transgene does not tolerize non-obese diabetic mice and can exacerbate disease. Proc Natl Acad Sci USA. 95:10055–10060.[Abstract/Free Full Text]
74. Gunaratnam L, Wiener S, Hammond-McKibben D, Cheung R, Dosch H-M. 1998 Precipitation of IDDM by the ICA69 T-cell epitope peptide Tep69. Diabetes. 47(Suppl 1):A34.
75. Bellmann K, Kolb H, Rastegar S, Jee P, Scott FW. 1998 Potential risk of oral insulin with adjuvant for the prevention of type 1 diabetes: a protocol effective in NOD mice may exacerbate disease in BB rats. Diabetologia. 41:844–847.[CrossRef][Medline]
76. Blanas E, Carbone FR, Allison J, Miller JFAP, Heath WR. 1996 Induction of autoimmune diabetes by oral administration of autoantigen. Science. 274:1707–1709.[Abstract/Free Full Text]
77. Genain CP, Abel K, Belmar N, et al. 1996 Late complications of immune deviation therapy in a non-human primate. Science. 274:2054–2056.[Abstract/Free Full Text]
78. Hartmann B, Bellmann K, Ghiea I, Kleemann R, Kolb H. 1997 Oral insulin for diabetes prevention in non-obese mice: potentiation by enhancing Th2 cytokine expression in the gut through bacterial adjuvant. Diabetologia. 40:902–909.[CrossRef][Medline]
79. Petersen JS, Karounos D, Dyrberg T. 1999 Dose dependent effect of prophylactic insulin: high dose is critical for prevention of diabetes. Diabetologia. 42(Suppl 1):A8.
80. DiLorenzo TP, Graser RT, Ono T, et al. 1998 Major histocompatibility complex plus one-restricted T-cells are required for all but the end stages of diabetes development in nonobese diabetic mice and use a prevalent T-cell receptor alpha chain gene rearrangement. Proc Natl Acad Sci USA. 95:12538–12543.[Abstract/Free Full Text]
81. Bergerot I, Ploix C, Petersen J, et al. 1997 A Cholera toxoid-insulin conjugate as an oral vaccine against spontaneous autoimmune diabetes. Proc Natl Acad Sci USA. 94:4610–4614.[Abstract/Free Full Text]
82. Meyer AL, Benson JM, Gienapp IE, Cox KL, Whitacre CC. 1996 Suppression of murine chronic relapsing experimental autoimmune encephalomyelitis by the oral administration of myelin basic protein. J Immunol. 157:4230–4238.[Abstract]
83. Savola K, Sabbah E, Kulmala P, et al. 1998 Autoantibodies associated with type 1 diabetes mellitus persist after diagnosis in children. Diabetologia. 41:1293–1297.[CrossRef][Medline]
84. Dosch H-M, Cheung RK, Karges W, Pietropaolo M, Becker DJ. 1999 Anergic and non-anergic autoreactive T cells persist past onset of autoimmune diabetes. J Immunol. In press.
85. Martin S, Schernthaner G, Nerup J, et al. 1991 Followup of cyclosporine A treatment in type 1 (insulin dependent) diabetes mellitus: lack of long-term effects. Diabetologia. 34:429–434.[CrossRef][Medline]
86. De Filippo G, Carel JC, Boitard C, Bougneres PF. 1996 Long-term results of early cyclosporine therapy in juvenile IDDM. Diabetes. 45:101–104.[Abstract]
87. Carel JC, Boitard C, Eisenbarth G, Bach JF, Bougneres PF. 1996 Cyclosporine delays but does not prevent clinical onset in glucose intolerant pre-type 1 diabetic children. J Autoimmun. 9:739–745.[CrossRef][Medline]
88. Pozzilli P, Browne PD, Kolb H, The Nicotinamide Trialists. 1996 Correct analysis of nicotinamide in patients with recent onset insulin dependent diabetes. Diabetes Care. 19:1356–1363.
89. Elliott RB, Chase HP. 1991 Prevention or delay of Type 1 (insulin-dependent) diabetes mellitus in children using nicotinamide. Diabetologia. 34:362–365.[CrossRef][Medline]
90. Elliott RB, Pilcher CC, Ferguson DM, Stuart AW. 1996 A population based strategy to prevent insulin-dependent diabetes using nicotinamide. J Pediatr Endocrinol Metab. 9:501–509.[Medline]
91. Lampeter EF, Klinghammer A, Scherbaum WA, et al. 1998 The Deutsche Nicotinamide Intervention Study: an attempt to prevent type 1 diabetes. DENIS Group. Diabetes. 47:980–984.[Abstract]
92. Shah SC, Malone JI, Simpson NE. 1989 A randomized trial of intensive insulin therapy in newly diagnosed insulin-dependent diabetes mellitus. N Engl J Med. 320:550–554.[Abstract]
93. The Diabetes Control, and Complications Trial Research Group. 1994 Effect of intensive diabetes treatment on the development and progression of long-term complications in adolescents with insulin-dependent diabetes mellitus. Diabetes Control and Complications Trial. J Pediatr. 125:177–188.[CrossRef][Medline]
94. Eisenbarth GS, Gianani R, Yu L, et al. 1998 Dual-parameter model for prediction of type 1 diabetes mellitus. Proc Assoc Am Phys. 110:126–135.[Medline]
95. Keller RJ, Eisenbarth GS, Jackson RA. 1993 Insulin prophylaxis in individuals at high risk of Type 1 diabetes. Lancet. 341:927–928.[CrossRef][Medline]
96. Fruchtenbusch M, Rabl W, Grassl B, Bachmann W, Standl E, Ziegler AG. 1998 Delay of type 1 diabetes in high-risk first-degree relatives by parenteral antigen administration: The Schwabing insulin prophylaxis pilot trial. Diabetologia. 41:436–441.
97. Schatz DA, Rogers DG, Brouchard BH. 1996 Prevention of insulin-dependent diabetes mellitus: An overview of three trials. Cleve Clin J Med. 63:270–274.[Medline]
98. Greenbaum CJ, Kahn SE, Palmer JP. 1996 Nicotinamide’s effect on glucose metabolism in subjects at risk for IDDM. Diabetes. 45:1631–1634.[Abstract]
99. Bingley PJ, Caldas G, Bonfanti R, Gale EA. 1993 nicotinamide and insulin secretion in normal subjects. Diabetologia. 36:675–677.[CrossRef][Medline]
100. Gotfredsen GF, Buschard K, Frandsen EK. 1985 Reduction of diabetes incidence of BB Wistar rats by early prophylactic insulin treatment of diabetes-proned animals. Diabetologia. 28:933–935.[CrossRef][Medline]
101. Like AA. 1989 Morphology and mechanisms of autoimmune diabetes as revealed by studies of the BB/Wor rat. In: Hanahan D, McDevitt HO, Cahill GJ, eds. Perspectives on the molecular biology and immunology of the pancreatic ß cells. Cold Spring Harbor, NY: Current Communications in Molecular Biology; 81–91.
102. Gottlieb P, Handler E, Appel M, et al. 1991 Insulin treatment prevents diabetes but not thyroiditis in RT6-depleted resistant BB/Wor rats. Diabetologia. 34:296–300.[CrossRef]