Reviews in Endocrine and Metabolic Disorders

, Volume 7, Issue 3, pp 215–224

Prevention strategies for type 1 diabetes


    • The Barbara Davis Center for Childhood Diabetes
  • H. Peter Chase
    • The Barbara Davis Center for Childhood Diabetes
  • Jennifer M. Barker
    • The Barbara Davis Center for Childhood Diabetes

DOI: 10.1007/s11154-006-9015-z

Cite this article as:
Kishiyama, C.M., Chase, H.P. & Barker, J.M. Rev Endocr Metab Disord (2006) 7: 215. doi:10.1007/s11154-006-9015-z


Type 1 diabetes (T1D) is a common chronic disease of childhood. Patients with T1D are at significant risk for developing serious health complications. Understanding of the genetics, environmental factors, and natural history of diabetes has lead to greater understanding of the etiology and epidemiology of T1D. Furthermore, technology has greatly improved glycemic control and reduction of complications. However, prevention of the development of diabetes remains elusive. This review article describes the past, current and upcoming strategies for diabetes prevention for patients at risk for developing autoimmunity, after antibody production, and patients with new onset diabetes.


Type 1 diabetesPreventionGeneticsImmunologyAutoimmunityTherapy

1 Introduction

Type 1 diabetes (T1D) is one of the most common chronic diseases in childhood and adolescents. It affects approximately 150,000 people under the age of 18 in the United States, and more than 13,000 new pediatric patients are diagnosed annually. The incidence of T1D is increasing at a rate of approximately 3%/year [1]. Patients with diabetes are at significant risk for developing serious health complications over time including cardiovascular disease, diabetic retinopathy, and diabetic nephropathy. Although therapies directed at maintaining good metabolic control and reducing diabetic complications are improving, prevention or cure of diabetes is the ultimate goal for treatment.

Over the years, researchers have learned a considerable amount about the natural history of diabetes. T1D is an autoimmune disease directed at the pancreatic β-cells. Figure 1 shows a model for the development of T1D. People at increased genetic risk for diabetes either by family history of T1D or through known diabetes risk genes at the major histocompatibility complex (MHC) and other genetic loci are shown at the upper left. A small subset of these subjects initiate autoimmune responses against the β-cells. The mediators of the initiation of the autoimmune response continue to be elucidated but are thought to be environmental factors such as diet or infection. There is strong evidence that T cells mediate this attack and the process is marked by the production of diabetes related autoantibodies. As ongoing loss of β-cells occurs, the pancreas’ ability to maintain normoglycemia declines and metabolic abnormalities are detected, progressing from altered first phase insulin response (FPIR) on intravenous glucose tolerance testing to impaired glucose tolerance on oral glucose tolerance testing to overt diabetes. A key feature of this model is the observation that at onset of diabetes, subjects continue to exhibit production of C-peptide (evidence of pancreatic β-cell function). Studies have shown that subjects who have continued C-peptide production have lower hemoglobin A1c’s (HbA1c’s), fewer hypoglycemic events, and are at less risk for complications.
Fig. 1

Model for the development of T1D. J. Skylar (From Eisenbarth et al. Type 1 Diabetes: Molecular, Cellular, and Clinical Immunology.

Diabetes prevention can be targeted at each phase of the development of diabetes: at determination of increased genetic risk prior to the development of autoimmunity, at the appearance of diabetes related autoimmunity, and at clinical diagnosis of diabetes. Several past and current studies using this strategy are discussed next.

2 Interventions prior to autoimmunity

Over the past several years, the understanding of the genetics of diabetes has increased severalfold. As with other autoimmune diseases, T1D is associated with increased frequency of certain higher risk Human Leukocyte Antigen (HLA) genotypes [2]. Increased risk of diabetes is associated with HLA—DR3 and DR4 if carried with CBB1* 0302 in particular [3]. Identifying subjects with an increased risk of developing diabetes has allowed researchers to target this population of subjects for potential preventative treatments prior to the development of autoimmunity. Intervention would therefore prevent the development of diabetes and any pathological sequelae.

The preschool population is of particular interest, because this group has the greatest increase in the incidence of T1D. The Euro-Diab ACE Study Group found an increase of 6.3% for children ≤4 years, 3.1% for ages 5–9 years and 2.4% for ages 10 to 14 years [4]. This increase cannot be explained by genetics, and most current investigators believe the cause to be environmental. With the increase in this very young population, the environmental exposure must occur during pregnancy or the first few years after birth. Evidence suggests that most people develop islet cell antibodies prior to 5 years of age [5]. Thus, if prevention is to begin prior to the development of islet cell autoimmunity, it will need to be at a very young age.

There are many possible explanations for the increase in T1D in this young population. One hypothesis is the “hygiene” or “clean-environment” etiology [6]. This theory arose in part from the observation that non-obese diabetic (NOD) mice exposed to multiple infections have a low incidence of diabetes [6]. In contrast, the incidence of diabetes in NOD female mice kept in a clean environment is approximately 70%. A suggestion that this theory is applicable for humans relates to day care attendance during the first year of life has been inversely associated with T1D [7]. These findings suggest early infectious exposure may play a role in the development of immunoregulatory mechanisms which protect against diabetes. However, current data are from retrospective analysis of case control data and prospective data are needed. Also, immunizations would be suspected to result in fewer infections, but do not increase the incidence of T1D [8].

A second environmental factor that might influence the development of diabetes are viral infections. In some ways this is the opposite of the hygiene theory with viral infections causing the initial autoimmune reaction. Although more than a dozen different viruses have been implicated [9], congenital rubella [10] and Coxsackie [11] were the first to be described. It is likely the viral infection must remain persistent (as with congenital rubella) or require multiple infections. The picornaviruses (including human enteroviruses and rhinoviruses) are currently the leading suspects. Prospective studies from Finland have found strong correlations between enteroviral infections using PCR, and the development of islet cell autoantibodies [12]. Honeyman et al. found a specific and highly significant association between rotaviral sero-conversion and increases in the IA-2, IAA and GAD antibodies [13]. A review of the possible role of viruses in initiating the initial islet cell autoimmune reaction and the problems in identifying this role can be found elsewhere [14]. To account for the increasing incidence of T1D, one would have to propose that this group of viruses also increased in incidence in the general population, or that the incidence has increased specifically in pregnant women or in infants. To date there is no evidence for this.

A third area which might be related to the increased incidence in young children is nutrition. At this time, nutrition-related trials to prevent the early autoimmune reaction of T1D can be safely done and two such trials have already begun.

3 Nutrition-related trials

3.1 Cow’s milk

The first involves examination of whether infant dietary exposure to cow’s milk is associated with islet cell autoimmunity and/or T1D. Cow’s milk is an attractive culprit for initiating the autoimmune reaction, as it is often the first or sole food fed to infants. Elliott et al. [15] showed that an amino acid-based formula fed to weanling BB rats prevented diabetes. Karges et al. [16] found similar results with the NOD mouse. There have been numerous studies of the effects of cow’s milk in human infants—summarized in a metaanalysis by Gerstein [17]. He concluded that “Ecologic and time-series studies consistently showed a relationship between type 1 diabetes and either cow’s milk exposure or diminished breast feeding.” Unfortunately, many of the studies had been retrospective in design and are adversely affected by duration of the recall period as well as by the already presence of a disease. At least one study with minimal periods of recall found no association between early exposure to cow’s milk and β-cell autoimmunity [18]. Couper et al. prospectively followed 317 children with a first degree relative with T1D for onset of islet cell autoantibodies [19]. They found no relation for the appearance of autoantibodies to duration of exclusive breast feeding or the introduction of cow’s milk or dairy products. In spite of these findings, there is a clear need for a prospective randomized clinical trial.


The Trial to Reduce IDDM in the Genetically at Risk (TRIGR) was preceded by pilot trials in Finland in which 234 HLA high risk infants having a first degree relative with T1D were randomized to receive a cow’s milk based formula or a hydrolyzed casein formula, usually after breast feeding for 6 to 8 months [20]. The investigators had previously hypothesized that the insulin transmitted to the infant in cow’s milk might be an etiologic factor in initiating the autoimmune reaction of T1D. This would explain why insulin autoantibodies (IAA) are the most frequent autoantibodies initially found in offspring of diabetic parents [21]. The findings of the pilot trial showed a higher response of cellular and humoral factors to bovine insulin in infants fed the cow’s milk based formula. This is of interest but does not prove a relationship with islet cell autoimmunity and/or T1D.

The current TRIGR study will be similar to the pilot trial and is underway in 17 countries with the sample size of 2,730 infants expected to be reached in 2006. The study is comparing a hydrolyzed casein formula (Nutramigen™; Mead Johnson Nutritional [MJN]) to a cow’s milk formula (Enfamil™, [MJN]) supplemented with Nutramigen (to preserve blinding). Although it will likely be 2009 before the results begin to appear, the study is of enormous importance and a monumental undertaking.

3.3 NIP-diabetes

Nutritional Intervention to Prevent Diabetes (NIP-Diabetes) is the second early prevention trial currently underway. It is a randomized clinical trial of docosahexanoic acid (DHA), a 22:6 n 3 fatty acid, to be taken in late pregnancy and early infancy to try to prevent the development of islet cell autoimmunity. The intake of DHA has decreased dramatically in the past century [22], fitting with the increase in T1D. In addition, obstetricians have generally recommended against fish intake during pregnancy due to possible contaminants such as methyl mercury and polychlorinated biphenyl compounds. Fortunately, the DHA to be used in the NIP-Diabetes study is grown from algae in vats and does not contain contaminants.

It had been previously observed in a case-control epidemiologic study that cod liver oil taken during pregnancy or by infants in the first year of life was associated with a decreased risk of developing T1D [23, 24]. Similarly, children born to women from Norwegian fishing villages had a significantly decreased risk of getting diabetes compared with children of women who lived in cities away from the coast [25]. It is also known that children with diabetes have significantly lower levels of DHA and lower n-3/n-6 ratios than do control children [26].

The mechanism of action of DHA in prevention might be related to reduction of the production of inflammatory cytokines [27] and/or of inflammatory prostaglandins [28, 29]. Both children who are islet-antibody positive (at risk for T1D) and those with T1D have been shown to have lower levels of the anti-inflammatory cytokines, IL4 and TGF-beta [30]. C-reactive protein (CRP) is a marker associated with inflammation. In the DAISY study of genetically at risk infants, those who were antibody positive in the preschool years were more likely to have had elevated CRP levels during periods >6 months prior to the development of T1D, than for infants who did not progress to T1D [31].

Numerous studies have shown the safety of DHA both in pregnancy and in infants. Studies in pregnancy have shown it may help to prevent pre-term births [32]. Many infant formulas now contain small amounts of DHA.

The NIP pilot trial is admitting pregnant mothers and/or infants under age 5 months who have a first degree relative with T1D. Supplementation with DHA or placebo can start in the last trimester of pregnancy or in the first 5 months after birth. To continue in the trial after birth, or to begin in the trial following birth (to age 5 months), the infant must also have a high risk HLA type. Nursing mothers will receive either DHA or placebo, and formulas are available for infants of non-nursing mothers. Supplementation with oil/capsules will continue until age 3 years—with the infants followed until the end-point of the development of two positive islet cell antibodies (on two separate tests) or until the development of T1D or reaching his/her ninth birthday.

Progression of the pilot trial to a full trial will depend on: (1) adequacy of recruitment, (2) treatment compliance, (3) ability to increase plasma DHA levels in test subjects, and (4) ability to decrease the major inflammatory cytokine, IL1-beta, in test subjects. It is believed that the pilot trial of 90 subjects will require approximately two years to complete. The decision will then be made whether to progress to a full international trial.

4 Interventions with autoimmunity prior to diabetes

Studies on the natural history of T1D have demonstrated that diabetes related autoantibodies are a sensitive and specific marker of risk for T1D [33]. Investigators have employed these autoantibodies to identify groups of people with evidence for an autoimmune process against the beta cells of the pancreas for potential intervention trials [34].

Several large scale prevention trials have targeted subjects after the development of diabetes related autoimmunity and prior to the development of diabetes. The ability to screen for diabetes related autoimmunity (autoantibodies to insulin, GAD65 and IA-2 and islet cell autoantibodies (ICA)) has allowed the identification of subjects at risk for diabetes prior to the development of hyperglycemia. While the majority of diabetes develops in the general population, the over all risk for diabetes in the general population is relatively small (prevalence of 1/300 children [35, 36]). Therefore, these studies have screened a group with an increased a priori risk for diabetes, those with a family history of T1D for screening and potential intervention [37].

4.1 DPT-1

The use of insulin for the prevention of T1D has been suggested by animal models of diabetes (particularly the non-obese diabetic mouse (NOD) [38, 39]. Therefore, the Diabetes Prevention Trial-type 1 (DPT-1) was initiated to test both parenteral [40] and oral insulin [41] for the prevention or delay of the development of T1D. DPT-1 screened first and second degree relatives for islet autoimmunity using ICA [34]. Subjects with positive ICA were enrolled into the staging study in which they received an oral and intravenous glucose tolerance test, DNA was obtained to identify the protective diabetes HLA allele and insulin autoantibodies were obtained. A subset of subjects was diagnosed with diabetes immediately and therefore not eligible for the prevention study. The remainder were either identified as having impaired or normal glucose tolerance and normal or low first phase insulin response on IVGTT. With these studies, subjects were identified as having a greater than 50% 5-year risk for diabetes or a 25–50% risk for type 1 diabetes. Subjects with the higher risk were offered enrollment into the parenteral insulin trial, subjects with the lower risk and positive insulin autoantibodies were offered enrollment into the oral insulin trial. Subjects with the protective HLA allele were not eligible for enrollment.

DPT-1 screened approximately 100,000 relatives of subjects with T1D and identified approximately 3,000 subjects with ICA. From these subjects a total of 711 were randomized to the parenteral (339) [40] and oral (372) [41] insulin prevention trials. The results of the parenteral trial were reported in 2002, there was no effect of parenteral insulin in the incidence of diabetes. Of note, subjects with β-cell autoimmunity and evidence for metabolic abnormalities (either impaired glucose tolerance of low first phase insulin response) were predicted to have a risk for diabetes of more than 50% risk in 5-years. It also identified a 15% risk/year of diabetes development, suggesting that algorithms that we have to predict risk for disease are highly accurate in the population with a relative with T1D.

The oral insulin trial results were more recently reported with intervention and observation arms having a similar incidence of diabetes [41]. In the group as a whole, there was no delay in diabetes development in the oral insulin treated subjects compared with placebo treated subjects. However, during the trial the entry criteria declined from an insulin autoantibody level greater than 80 nU/ml to greater than 40. Post-hoc analysis revealed an approximately 4 year delay in the development of diabetes in subjects treated with oral insulin with a randomization insulin autoantibody level greater than 80 nU/ml, suggesting an effect of oral insulin in subjects with higher levels of insulin autoantibody. These findings are the basis for proposed studies of oral insulin in insulin autoantibody positive subjects with higher levels of insulin autoantibodies.

4.2 Nicotinamide

Nicotinamide is a vitamin B3 derivative that affects ADP ribosylation in immune cells and beta cells [42]. Data from the NOD mouse showed that nicotinamide could protect these mice from spontaneous diabetes. Based on trials in patients with new onset T1D, in which there appeared to be preservation in the C-peptide responses in subjects treated with nicotinamide compared with placebo [43, 44], it was hypothesized that it might delay the development of diabetes in subjects already autoantibody positive. Because of these observations, the decision was made to use nicotinamide in subjects with evidence for beta cell autoimmunity prior to the development of diabetes in the European Nicotinamide Diabetes Intervention Trial (ENDIT) [45]. This trial also screened first-degree relatives of patients with T1D for ICA. They screened 29,718 relatives for ICA and randomized 552 to placebo vs. nicotinamide. There was no decrease in diabetes incidence in the nicotinamide vs. placebo group over 5 years of follow-up. Subgroup analysis comparing response by age, gender, oral glucose tolerance status and first-phase insulin response was unable to identify a subgroup that responded to therapy.

While the results from ENDIT and DPT-1 parenteral and oral insulin prevention trials have been disappointing in terms of preventing diabetes, these trials demonstrate that we have a simple and accurate way of identifying subjects at risk for diabetes and that these subjects (at least among relatives with T1D) are willing to participate in intervention trials. Therefore, despite the fact that they were unable to identify a treatment that delayed diabetes onset, these trials have been viewed as an organizational success and serve as the basis for TrialNet. Through these experiences, TrialNet for TID has been established and is currently enrolling subjects into its natural history study and will be initiating prevention and treatment trials for T1D in the years to come.

One of the major constraints of prevention strategies such as these is that subjects are already experiencing beta cell autoimmunity. This may suggest that they have already initiated an immune response that will be very difficult to stop or slow down. Therefore strategies implemented prior to the development of autoimmunity may be more effective. However, these strategies must be very safe and relatively easy to implement because the majority of subjects treated will not develop diabetes.

5 Interventions after diagnosis

The final stage for intervention for patients with T1D is at the onset of diabetes Most of these patients have lost approximately 80–90% of their beta cell mass [46]. The remaining 10–20% of islet cells fail to compensate, and insulin production is overwhelmed by the body’s insulin requirement, thus presenting with hyperglycemia and diabetes.

The goal of intervention at the time of diagnosis is preservation of the remaining beta cell function. Although prolongation of beta cell function may not reverse the need for exogenous insulin therapy, preserving even minimal function has been shown to lower HbA1c levels, lower the incidence of hypoglycemia, and decrease the amount of complications [47, 48]. Currently, C-peptide measurement is the most useful test to evaluate islet cell function. The proinsulin molecule is cleaved to form insulin and C-peptide in normal individuals. In patients with T1D, as the beta cells are destroyed, the ability to produce both insulin and C-peptide decline. Therefore, C-peptide measurement is currently the best available test to quantitatively assess remaining islet cell function and endogenous insulin production.

The majority of diabetes intervention trials have occurred after the development of diabetes. There are several studies aimed at immune modulation in animal models (Table 1). Therapies are targeted at stopping the autoimmunity, or development of immune tolerance. Several trials were effective in animal models, and a number of these potential therapies transitioned to clinical trials in humans (Table 2). Below is a description of past and present trials aimed at delaying or preventing the further destruction of islet cells.
Table 1

Immune modulation in animal models





Immune suppression


NOD and BB

Nonspecific suppression of cell-mediated immunity prevents type 1 diabetes, but long-term immunosuppression is unacceptable


CD4 T cells


Anti-CD4 monoclonals prevent diabetes


CD3 T cells


Anti-CD3 monclonals reverse diabetes at onset with long-duration of effect

IL-15 antagonist + rapamycin

Activated T cells


Tolerance induction



NOD and BB

Grafts of marrow, dendritic cells, fetal liver and thymus protective

Immune stimulation



Immune activation by agents such as BCG protective

Immunologic vaccination



Multiple mechanisms, potential activation T regulatory cells.

Oral tolerance



Immune deviation delays diabetes




Radical scavenging, nutritional specific (e.g., lack peptides) are protective


NK T cells


Activation CD1 restricted cells


IL-10, TNF, IL-4


Complex effect dependent often upon timing of Rx

Gene therapy



Multiple potential targets


Beta cell


Beta cell proliferation/inhibition apoptosis

NOD = nonobese diabetic mouse; BB = BioBreeding rat

(From Eisenbarth GS, et al. Type 1 Diabetes: Molecular, Cellular, and Clinical Immunology.

Table 2

Prevention trials in type 1 diabetes

Name of the study


Type of study





Primary prevention

Randomized, placebo-controlled, double-blind

Hydrolyzed cow’s milk formula

FDR at high genetic risk



Secondary prevention

Randomized, placebo-controlled, double-blind


High-risk FDR: ICA+

No prevention

DPT-1 parenteral

Secondary prevention

Randomized, no placebo

Parenteral insulin

High-risk relatives

No prevention

DPT-1 oral

Secondary prevention

Randomized, placebo-controlled, double-blind

Oral insulin

Moderate-risk relatives

Reduced incidence of T1DM in relatives with IAA ≥80 U


Tertiary prevention

Randomized, placebo-controlled, double-blind


New-onset T1DM patients

Temporary remission of T1DM


Tertiary prevention

Randomized, placebo-controlled, double-blind


New-onset T1DM patients

No effect

Azathioprine, prednisone

Tertiary prevention

Randomized, unblind

Azathioprine, prednisone

New-onset T1DM patients

Partial, temporary remission of T1DM

Anti-CD3 treatment

Tertiary prevention

Randomized, placebo-controlled


New-onset T1DM patients

Prevention of loss of C-peptide


Tertiary prevention

Randomized, placebo-controlled

Anti-CD20 monoclonal antibodies

New-onset T1DM patients

To be started


Tertiary prevention

Randomized, placebo-controlled

Antithymocyte polyclonal antibodies

New-onset T1DM patients

To be started


Secondary prevention

Randomized, placebo-controlled, double-blind

Nasal insulin

High-risk subjects


NBI-6024 neurocrine

Tertiary prevention

Randomized, placebo-controlled, double-blind

Altered peptide ligand insulin B:9–23

New-onset T1DM patients


Peptor HSP60

Tertiary prevention

Randomized, placebo-controlled, double-blind

Heat shock protein 60

New-onset T1DM adult patients

Ongoing; prevention of loss of C-peptide

BCG vaccination

Tertiary prevention

Randomized, placebo-controlled, double-blind

BCG vaccine

New-onset T1DM adult patients

No effect

BCG, bacillus Calmette-Guerin; FDR, first degree relatives; IAA, insulin antibodies; ICA, islet cell antibodies; T1DM, type 1 diabetes mellitus.

Table 2 adapted from: Casu A, Trucco M, Pieotropaolo M. A look in the future: prediction, prevention, and cure including islet transplantation and stem cell therapy. Reprinted from Pediatric Clinics of North America, Vol 52, Casu A, A Look to the Future: Prediction, Prevention, and Cure Including Islet Transplantation and Stem Cell Therapy, Page 1787, Copyright (2005), with permision from Elsevier.

5.1 Cyclosporine A

Cyclosporine A is an immunosuppressive drug which interferes with T-cell cytokine production, particularly interferon γ (IFN-γ) and tumor necrosis factor α (TNF-α). IFN-γ and TNF-α are important mediators in T-cell expansion, and ultimately β-cell destruction [49, 50]. Stiller et al. conducted the first pilot trial in human subjects, and 16 of 30 patients (53%) with new onset T1D became insulin-independent, normalized their C-peptide concentration, and decreasing titers of ICA antibodies [51]. Bougneres et al. confirmed these findings in the first trial in pediatric patients. Out of 40 subjects with newly diagnosed T1D, 27 (67%) were able to discontinue insulin therapy within 48 days after initiation, and 75% of these 27 patients remained off of insulin at 1 year [52].

Unfortunately, despite the dramatic beneficial effects for patients with T1D, the side effects from cyclosporine A became a major concern. In a study of 21 patients who received cyclosporine A, 4 of the 21 patients developed microalbuminuria compared to none of the 17 diabetic controls [53]. Although no further studies have duplicated these results, the risk for increased renal toxicity remains a concern for future use despite the C-peptide improvement and decline of ICA antibodies.

5.2 Glucocorticoids and azathioprine

Another immunosuppressive approach was a combination therapy of glucocorticoids in combination with azathiprine. Silverstein conducted a study in 46 new-onset pediatric patients. Half were randomized to glucocorticoid treatment for 10 weeks, followed by azathioprine for 1 year, the rest to placebo. The treatment group had a greater improvement in their metabolic outcomes including HbA1c, stimulated C-peptide, and lower insulin requirement [54]. Despite the improvement in beta cell function, side effects were the reason for discontinuation in this therapy as well. Intolerable side effects included hair loss, gastrointestinal upset, Cushingoid facies, and lymphopenia [54].

5.3 Antithymocyte globulin (ATG)

ATG has been used in organ transplantation, but has not yet been shown to be effective in inducing immune tolerance. ATG is produced by isolating T-lymphocytes, and injecting them into an animal such as a rabbit or horse. The animal then sees the T-cell surface markers as foreign, and makes several antibodies to the T-cell antigens. The serum is then purified of all material except the antibodies, thus making ATG. ATG is then injected back into the subject, and the ATG acts to attach to T cells, causing the host to see the T cells with antibodies attached as foreign, thereby neutralizing the potentially harmful T cells. One clinical trial showed a reduction of HbA1c levels and lower insulin requirements [55]. However, two patients developed severe thrombocytopenia. Upcoming trials of ATG will need to carefully examine these risks, particularly in children.

5.4 Anti-CD3 (hOKT3)

OKT3 is another immunosuppressant medication that targets T-cell function. It is a monoclonal antibody to the CD3 receptor. The CD3 receptor is important for T-cell activation and immunologic destruction. OKT3 binds to the CD3 receptor, causing transient activation, cytokine release, and ultimately blocks T-cell proliferation and differentiation. Modifications made in humanized or hOKT3 have dampened the cytokine release syndrome, (massive cytokine release leading to fever, hypotension, myalgias, and arthralgias [56, 57]) observed with the native monoclonal antibody, making hOKT3 better tolerated. Herold and colleagues conducted a randomized placebo-controlled phase I–II trial with hOKT3 in 12 patients. They demonstrated a decreased decline of stimulated C-peptide, lower HbA1c levels, and lower insulin requirements in 75% (9/12) of patients receiving hOKT3 compared to 17% (2/12) controls [58]. Maintenance of C-peptide levels persisted for approximately 18 months. The next phase is a randomized, 2-arm, open-label phase II trial with 81 patients in which subjects will receive repeated courses of therapy [59].

5.5 Mycophenolate mofetil-daclizumab (MMF-DZB)

MMF is an immunosuppressant that inhibits the synthesis of the purine guanosine monophosphate. Because T- and B-lymphocytes are dependent upon de novo synthesis of purines, MMF inhibits proliferation of lymphocytes. DZB is another immunosuppressive medication which competitively binds the CD25 receptor. IL-2 typically binds this receptor. Therefore, DZB prevents IL-2 activation of lymphocytes. Because these two medications work at different targets, the combination therapy potentially works synergistically while minimizing toxicity. As toxicity has precluded the implementation of many of the above immunosuppressant therapies, this is an important consideration. Gottlieb and colleagues have initiated a multi-center, three-arm, randomized, double-masked, placebo-controlled trial investigating the efficacy of this combination therapy [58].

5.6 Anti-CD20

Rituximab, or anti-CD20 is a monoclonal antibody which targets the CD20 receptor which is unique to B cells. B cells have two functions, first as an antigen-presenting cell and second, antibody production against non-self-antigens. By targeting the CD20 receptor, rituximab inhibits the B-cell function, thus reducing presentation of free antigen to T cells, and theoretically secondarily preventing B-cell expansion and antibody production. This medication has shown success in treatment of patients with another organ specific autoimmune disease rheumatoid arthritis. Therefore, it may be employed in the future as a treatment for new onset T1D. There are no current clinical research studies on anti-CD20 in patients with diabetes, but this has shown promise in other autoimmune disorders.

5.7 BCG

BCG has been shown to prevent diabetes in animal models. It is administered as a vaccine which is thought to provide antigenic stimulation, leading to a preferential shift of the T-helper cell away from the destructive TH1 response towards a protective TH2 destiny. Theoretically, a robust TH2 lymphocytic response could potentially halt the pathologic immune destruction of the beta cells. Two independent studies with BCG prevented the development of T1D in non-obese diabetic (NOD) mouse models [59, 60].

BCG has now been studied in human populations without success. Two randomized placebo-controlled trials were completed in patients with newly diagnosed diabetes. There was no difference in preservation of C-peptide levels. Granted, these were patients with known diabetes, and possibly BCG would be more appropriate for preventing the development of diabetes in subjects at risk rather than patients with known diabetes.

5.8 Heat shock protein (HSP) 60 and peptide 277 (p277)

HSP-60 is thought to be an autoantigen involved in T-cell autoimmune destruction. HSP-60 has been shown to lead to T-cell reactivity on the NOD mouse [61]. Peptide 277 which arises from HSP-60 can potentially reduce beta cell destruction in mice presenting with clinical diabetes [62, 63]. There have been studies investigating the effect of p277 in humans, of which one has shown some effect. Raz et al. demonstrated in a small study that stimulated C-peptide remained higher in treated subjects compared to controls, and it was thought that there may be a shift from TH1 to TH2 cytokine production [64]. Recently, scientists have also postulated that the p227 peptide is an antigenic portion buried in the HSP 60 or 65, and may stimulate a toll receptor which acts as a negative feedback, to establish homeostasis and downregulate HSP-60 reactive effects. Whichever theory holds true, p277 effects may offer another avenue for therapy.

6 Conclusions

Prevention of diabetes remains a significant area of interest. A challenging treatment paradox exists for diabetes (Fig. 2). Targeting subjects early provides the potential of preventing the initiation of the autoimmune process, thereby preventing the development of diabetes. However, a number of problems exist when targeting this group of subjects. Because we do not yet have an inexpensive, precise predictive tool to determine who will develop diabetes, a large group of subjects would need to receive prophylactic treatment to prevent a small amount of diabetes. This would require many subjects, who may never develop diabetes, to receive potentially harmful therapies while only a few will benefit from prevention. As patients develop autoimmunity, our ability to diagnose patients at risk of developing diabetes improves. However, the later the diagnosis, the lesser the number of the beta cells remaining, resulting in potentially decreased therapeutic benefit of intervention regimens. Additionally, once the autoimmune process has begun it might be quite difficult to alter.
Fig. 2

Treatment paradox Copyright© 2005 American Diabetes Association from Diabetes, vol. 54, 2005; 1252–1263. Reprinted with permission from the American Diabetes Association

The optimistic response to this quandary lies in continued research. Diabetes prevention research is an extremely dynamic field. Understanding the genetics of diabetes and determining which patients are at greatest risk will aid in early diagnosis and allow appropriate prevention strategies. Therapies are becoming more efficacious and simultaneously more specific, which will decrease the potential side effects of immunomodulatory treatments. Combination therapies allow for synergy and further reduction of toxicity.

The history of diabetes is filled with many groundbreaking research revelations. Currently, advancement in the understanding of diabetes and the advent of advanced technology is allowing health care providers, patients, and families to better manage diabetes. The future may hold the promise of prevention of diabetes.

Copyright information

© Springer Science+Business Media, LLC 2006