Immune intervention for type 1 diabetes mellitus


  • J. S. Skyler

    1. Division of Endocrinology, Diabetes and Metabolism; and Diabetes Research Institute, University of Miami Miller School of Medicine, Miami, FL, USA
    Search for more papers by this author

  • Disclosures: JSS is Chairman of Type 1 Diabetes TrialNet, a clinical trials network conducting multiple studies designed to interrupt the type 1 diabetes disease process. Also, recipient of grant support from Bayhill Therapeutics and Osiris Therapeutics. The author serves on the Board of Directors of Amylin Pharmaceuticals, and chairs the Type 1 Diabetes Advisory Board for sanofi-aventis.

  • Endorsed by the International Conference on ATTD organized by Kenes International.

Jay S. Skyler,
Division of Endocrinology, Diabetes, and Metabolism; and Diabetes Research Institute, University of Miami Miller School of Medicine, Miami, FL, USA
Tel.: 305-243-6146
Fax: 305-243-4484


The major form of type 1 diabetes (T1D) is characterised by immune-mediated pancreatic islet β-cell destruction, and has also been called type 1A diabetes to distinguish it from idiopathic forms of islet β-cell loss. Since the first demonstration of islet cell antibodies in 1974, the concept has been that this form of diabetes is autoimmune in nature. The commonly accepted concept is that antibodies (representing the humoral arm of the immune system) do not mediate the β-cell destruction but rather serve as markers of that destruction, while the cellular arm of the immune system, specifically T-lymphocytes, mediate the β-cell destruction. Yet, the T-lymphocytes do not act alone. They receive help in initiating the response from antigen-presenting cells such as dendritic cells and macrophages, and appear to receive help also from B-lymphocytes. In addition, the initial immune response engenders secondary and tertiary responses – involving the whole immunological army – which collectively result in impairment of β-cell function, progressive destruction of β-cells, and consequent development of type 1A diabetes. The process is insidious and may evolve over many years, with the overt expression of clinical symptoms becoming apparent only when most β-cells have been destroyed. Yet, the process clearly evolves at different speeds – much more rapidly in young children, much more slowly in older individuals. And, although it has been thought that ultimately there is complete β-cell destruction, several studies have now demonstrated some degree of persistent β-cell function or existence (at autopsy) in long-standing T1D. A major focus of investigation in T1D is the preservation of β-cell function (and, it is hoped, of β-cells themselves), in the expectation that continuing endogenous insulin secretion will contribute towards better glycaemic control, reduce episodes of severe hypoglycaemia, and slow the development of complications such as retinopathy and nephropathy. Thus, there have been many studies designed to interdict the T1D disease process, mostly by altering the immune system, both during the stage of evolution of the disease and at the time of disease onset. This chapter of the Yearbook of Advanced Technology and Treatments in Diabetes reviews the key papers that have appeared in this field between July 2009 and June 2010. Articles selected were confined to studies in human beings. All immune intervention studies reported in this time frame were included. In addition, the author selected other relevant articles dealing with mechanisms, markers, triggers, and pathology of human type 1 diabetes.

Immune interventions studies in human type 1 diabetes

The first group of papers involves reports of immune intervention studies in human type 1 diabetes, mostly in recent-onset T1D.

Four-year metabolic outcome of a randomised controlled CD3-antibody trial in recent-onset type 1 diabetic patients depends on their age and baseline residual β-cell mass

B. Keymeulen,1M. Walter,2C. Mathieu,3L. Kaufman,4F. Gorus,1R. Hilbrands,1E. Vandemeulebroucke,1U. Van de Velde,1L. Crenier,5C. De Block,6S. Candon,7H. Waldmann,8A. G. Ziegler,2L. Chatenoud,7D. Pipeleers1

1Diabetes Research Center and University Hospital, Brussels Free University-VUB, Brussels, Belgium,2Hospital München-Schwabing, Munich, Germany,3Department of Endocrinology, Katolieke Universiteit Leuven-KUL, Leuven, Belgium,4Department of Biostatistics and Medical Informatics, Brussels Free University-VUB, Brussels, Belgium,5Hôpital Erasme, Université Libre de Bruxelles-ULB, Brussels, Belgium,6Department of Diabetology, University Hospital Antwerp, Edegem, Belgium,7INSERM U580-IRNEM, Hôpital Necker, Paris, France, and8Sir William Dunn School of Pathology, Oxford, UK

Diabetologia 2010; 53: 614–23

Background: The authors previously reported the results of their initial randomised, controlled 18-month trial in which a short course of the humanised Fc-mutated recombinant aglycosylated anti-human CD3 antibody, ChAglyCD3 (now called otelixizumab), was given for six consecutive days shortly after diagnosis of T1D. In their earlier report, they demonstrated preservation of β-cell function – as measured by C-peptide – and equivalent glycaemic control with lower insulin doses. The current paper extends follow-up to 4 years.

Methods: The original study included 80 subjects (40 treated with anti-CD3, 40 treated with placebo), ages 12–39. Four-year follow-up data are reported on 64 subjects (33 of the anti-CD3 subjects, 31 of the placebo subjects). Change in insulin dose over 48 months was the primary end-point.

Results: Treatment with anti-CD3 delayed the rise in insulin requirements of subjects with recent-onset T1D and reduced the magnitude of that rise over 48 months of follow-up. Using multivariate analysis, this effect was correlated with higher baseline β-cell function and younger age.

Conclusions: Anti-CD3 therapy shortly after onset of T1D has sustained effects for 48 months.

Treatment of patients with new-onset type 1 diabetes with a single course of anti-CD3 mAb Teplizumab preserves insulin production for up to 5 years

K. C. Herold,1,2S. Gitelman,3C. Greenbaum,6J. Puck,3W. Hagopian,7P. Gottlieb,8P. Sayre,10P. Bianchine,9E. Wong,4V. Seyfert-Margolis,10K. Bourcier,10J. A. Bluestone,5,10Group Immune Tolerance Network ITN007AI Study

1Department of Immunobiology, Yale University, New Haven, CT, USA,2Department of Medicine, Yale University, New Haven, CT, USA,3Department of Pediatrics, University of California at San Francisco, CA, USA,4Department of Statistics, University of California at San Francisco, CA; PPD Inc., Wilmington, NC, USA,5Department of Medicine, University of California at San Francisco, CA, USA,6Benaroya Research Institute, Seattle, WA, USA,7Department of Medicine, Pacific Northwest Research Institute and University of Washington, USA,8Department of Medicine, University of Colorado, Denver, CO, USA,9National Institute for Allergy Immunology and Infectious Diseases, Bethesda, MD, USA, and10Immune Tolerance Network, Bethesda, MD, USA

Clin Immunol 2009; 132: 166–73

Background: The authors previously reported the 1- and 2-year results of an open-label trial in which a short course of the humanised Fc-mutated CD3-specific monoclonal antibody, hOKT3γ1 (Ala-Ala) (now called teplizumab), was given for 14 consecutive days shortly after diagnosis of T1D. In the reports of that study, they demonstrated preservation of β-cell function – as measured by C-peptide – and both better glycaemic control and lower insulin doses. The current paper reports on a smaller second trial that had enrolment interrupted due to adverse events.

Methods: The study enrolled 10 subjects before being halted for adverse events that seemed to be a consequence of higher doses of the drug used in this trial. The subjects were followed, nonetheless.

Results: At 2 years, the treated subjects (n = 6) had lower insulin doses than the comparison subjects (n = 4), and a trend towards better maintained C-peptide. The C-peptide levels in the treated subjects remained fairly constant between 2 and 5 years of follow-up. The authors concluded that the higher doses resulted in more adverse events without increased efficacy.

Conclusion: This small study suggests that anti-CD3 therapy shortly after onset of T1D may have prolonged effects with fairly constant C-peptide levels between 2 and 5 years.

  • Comment: These two studies demonstrate that relatively short courses of treatment (6 or 14 days) with an anti-CD3 monoclonal antibody can have sustained effects on β-cell function – as measured by C-peptide – for as long as 4 to 5 years. Both of the anti-CD3 antibodies used – otelixizumab and teplizumab – are now being studied in full-scale phase 3 trials for potential commercialisation for use in recent-onset T1D. Thus, the fact that there are long-term beneficial effects is important. Nonetheless, in the original trials with both of these antibodies, there was progressive decline in β-cell function, suggesting that there may be a need for repeated courses of administration. Alternatively, the efficacy of these antibodies might be improved if used in combination therapy with another agent (or agents) acting synergistically to improve β-cell function.

Cyclosporin and methotrexate therapy induces remission in type 1 diabetes mellitus

D. O. Sobel, A. Henzke, V. Abbassi

Department of Pediatrics, Georgetown University Medical Center, Washington, DC, USA

Acta Diabetol 2010; 47: 243–50

Background: Studies reported in the late 1980s and early 1990s had shown that cyclosporine appears to have beneficial effects in recent-onset T1D, as long as treatment is maintained, but at the price of significant adverse effects, particularly induction of renal damage. The current study was designed as a pilot study to determine whether lower doses of cyclosporine could avert the adverse effects, and if combined with methotrexate might still demonstrate the beneficial effects.

Methods: The study was an open non-randomised pilot study, in which seven subjects with recent-onset T1D were treated with cyclosporine plus methotrexate, and compared to 10 subjects not so treated. Only the clinical course was evaluated, and β-cell function was not reported.

Results: During the 1 year of follow-up, during which treatment was maintained in the intervention group, that group had lower insulin doses, greater likelihood of discontinuing insulin, and at the 12-month time point lower A1c. Adverse effects were said to be less than in earlier cyclosporine studies.

Conclusion: The combination of low-dose cyclosporine plus methotrexate may improve the course of T1D over 1 year.

  • Comment: This study attempts to take advantage of an oft utilised medical principle that using lower doses of two drugs may obviate the adverse effects of higher doses, whilst still achieving efficacy. The study was a very small pilot study in which only seven subjects were treated. That small number makes it impossible to assess the risk of adverse effects. Also, the absence of measurement of β-cell function makes it impossible to compare the outcome with that in other contemporary studies. Finally, most current studies in recent-onset T1D do not intentionally attempt to reduce insulin doses to achieve ‘remission’, but rather do so only to avert hypoglycaemia, maintaining the highest insulin dose that does not result in significant hypoglycaemia. This makes the claim of achieving ‘remission’ because insulin was able to be discontinued difficult to interpret.

Failure to preserve β-cell function with mycophenolate mofetil and daclizumab combined therapy in patients with new-onset type 1 diabetes

P. A. Gottlieb,1S. Quinlan,2H. Krause-Steinrauf,2C. J. Greenbaum,3D. M. Wilson,4H. Rodriguez,5D. A. Schatz,6A. M. Moran,7J. M. Lachin,2J. S. Skyler,8Type 1 Diabetes TrialNet MMF/DZB Study Group

1Barbara Davis Center for Childhood Diabetes, University of Colorado at Denver, Aurora, CO, USA,2Biostatistics Center, George Washington University, Rockville, MD, USA,3Benaroya Research Institute, Diabetes Clinical Research, Seattle, WA, USA,4Pediatric Endocrinology Department, Stanford University, Stanford, CA, USA,5Department of Pediatrics, Indiana University, Indianapolis, IN, USA,6Department of Pediatrics, University of Florida, Gainesville, FL, USA,7Department of Pediatrics, University of Minnesota, Minneapolis, MN, USA, and8Diabetes Research Institute, University of Miami, Miami, FL, USA

Diabetes Care 2010; 33: 826–32

Background: The authors evaluated the use of mycophenolate mofetil (MMF), either alone or in combination with daclizumab (DZB), in recent-onset T1D. MMF has potent cytostatic effects on lymphocytes, and has been effective in other autoimmune diseases and in diabetic animal models. DZB is a humanised monoclonal antibody that binds to CD25, the α-subunit of the interleukin-2 (IL-2) receptor expressed on the surface of activated lymphocytes, and has been effective in other autoimmune diseases as well as in islet transplant protocols. The combination of MMF and DZB had a beneficial synergistic effect in diabetic animal models.

Methods: The study included 126 subjects, ages 8–45, randomised to MMF alone, MMF plus DZB, or placebo. Two DZB (or placebo) infusions were administered – one initially and one 2 weeks later. MMF (or placebo) was given continuously, with a planned study end-point at 2 years.

Results: The Data Safety and Monitoring Board halted the study early due to the projected futility of achieving a beneficial result. There was a similar decline in β-cell function in all three groups – MMF alone, MMF plus DZB, and placebo. In addition, A1c and insulin dosage were similar in all three groups.

Conclusions: MMF, either alone or in combination with DZB, failed to prevent decline in β-cell function in recent-onset T1D.

  • Comment: This large, well-powered study failed to demonstrate preservation of β-cell function in recent-onset T1D. Although many interventions have shown benefit in recent-onset T1D, this study reminds us that not all interventions which show promise in animal models will work in human beings. The study does represent one of the first, if not the first, fully powered study using combination therapy in recent-onset T1D. There will no doubt be other combination therapy trials. In designing them, one might note one feature of this study – half of the treated subjects received an infusion of DZB, and half of the placebo subjects received an infusion of DZB. This maintained masking, while not requiring all subjects to receive infusions. This is a design that can be used if a single agent is being compared to its use in combination. Another approach would be a factorial design. Attention to design of combination studies will be an important consideration in future T1D trials.

Rituximab, B-lymphocyte depletion, and preservation of β-cell function

M. D. Pescovitz,1C. J. Greenbaum,2H. Krause-Steinrauf,3D. J. Becker,4S. E. Gitelman,5R. Goland,6P. A. Gottlieb,7J. B. Marks,8P. F. McGee,3A. M. Moran,9P. Raskin,10H. Rodriguez,1D. A. Schatz,11D. Wherrett,12D. M. Wilson,13J. M. Lachin,3J. S. Skyler,8Type 1 Diabetes TrialNet Anti-CD20 Study Group

1Indiana University School of Medicine, Indianapolis, IN, USA,2Benaroya Research Institute, Seattle, WA, USA,3George Washington University Biostatistics Center, Rockville, MD, USA,4University of Pittsburgh, Pittsburgh, PA, USA,5University of California, San Francisco, CA, USA,6Columbia University, New York, NY, USA,7University of Colorado Barbara Davis Center for Childhood Diabetes, Aurora, CO, USA,8University of Miami Diabetes Institute, Miami, FL, USA,9University of Minnesota, Minneapolis, MN, USA,10University of Texas Southwestern Medical School, Dallas, TX, USA,11University of Florida, Gainesville, FL, USA,12Hospital for Sick Children, University of Toronto, Toronto, ON, Canada, and13Stanford University, Stanford, CA, USA

N Engl J Med 2009; 361: 2143–52

Background: There is growing evidence that B-lymphocytes play a role in many T-lymphocyte mediated diseases, by serving as antigen-presenting cells or generating cryptic peptides to which T-lymphocytes are not tolerant. This study used the anti-CD20 antibody, rituximab, to selectively deplete B-lymphocytes in recent-onset T1D.

Methods: The study randomised 87 subjects (57 assigned to anti-CD20, 30 assigned to placebo), ages 8–45. Intervention consisted of four weekly infusions. The primary end-point was β-cell function – as measured by C-peptide – at 1 year, with 78 subjects included in the analysis.

Results: At 1 year, the mean level of C-peptide was significantly higher in the rituximab group than in the placebo group, and declined at a slower rate. The rituximab group also had significantly lower A1c levels and required less insulin. Adverse effects were minimal and mostly occurred during the first rituximab infusion.

Conclusions: The anti-CD20 antibody rituximab showed beneficial effects on β-cell function in recent-onset T1D, suggesting a potential role for B-lymphocytes in T1D pathogenesis.

  • Comment: This study demonstrates that a relatively short course of treatment (weekly infusions over 4 weeks) with an anti-CD20 monoclonal antibody can have beneficial effects on β-cell function – as measured by C-peptide – at 1 year. Nonetheless, there is progressive decline in β-cell function, as in virtually all intervention trials in recent-onset T1D, again suggesting that there may be a need for repeated courses of administration, or, alternatively, for combination therapy with another agent. An important feature of this study is the demonstration that B-lymphocytes contribute to the pathogenesis of T1D. This may have important implications for the design of combination therapy regimens for evaluation in T1D.

Six months of diazoxide treatment at bedtime in newly diagnosed subjects with type 1 diabetes does not influence parameters of β-cell function and autoimmunity but improves glycaemic control

M. A. Radtke,1,2I. Nermoen,3M. Kollind,4S. Skeie,5J. I. Sørheim,6J. Svartberg,7,8I. Hals,1T. Moen,9G. H. Dørflinger,1V. Grill1,2

1Department of Cancer Research and Molecular Medicine, Norwegian University of Science and Technology, Trondheim, Norway,2Department of Endocrinology, St Olavs Hospital/University Hospital of Trondheim, Trondheim, Norway,3Department of Endocrinology, Akershus University Hospital, Lørenskog, Norway,4Endocrinology Unit, Department of Internal Medicine, Levanger Hospital, Levanger, Norway,5Section of Endocrinology, Division of Medicine, Stavanger University Hospital, Stavanger, Norway,6Section of Endocrinology, Department of Medicine, Haukeland University Hospital, Bergen, Norway,7Section of Endocrinology, Division of Internal Medicine, University Hospital of North Norway, Tromsø, Norway,8Institute of Clinical Medicine, University of Tromsø, Tromsø, Norway, and9Department of Laboratory Medicine, Children’s and Women’s Health, Faculty of Medicine, Norwegian University of Science and Technology, Trondheim, Norway

Diabetes Care 2010; 33: 589–94

Background: Diazoxide provides β-cell rest by reversibly suppressing glucose-induced insulin secretion. Previous studies have shown that diazoxide may preserve β-cell function in recent-onset T1D, but with frequent adverse effects consisting of lanugo hair growth, oedema and hypotension. The current study was designed to test whether a lower dosage of diazoxide would eliminate side effects and still exert beneficial effects.

Methods: The study randomised 41 subjects (22 to diazoxide, 19 to placebo), ages 18–40, with recent-onset T1D. Subjects received 6 months of treatment with placebo or 100 mg diazoxide at bedtime, and were followed for a subsequent 6 months. Outcome measures were C-peptide (fasting and glucagon stimulated), A1c and insulin dose.

Results: A1c levels at both 6 months and 1 year were lower in the diazoxide group, but there was no difference in insulin dose, C-peptide levels (either actual levels or C-peptide/glucose ratios), or in proportion of Tregs (regulatory T-lymphocytes). There was more weight gain in the placebo group. Insulin sensitivity, as assessed by HOMA-S% (homeostatic model assessment for insulin sensitivity), increased in the diazoxide group but remained stable in the placebo group.

Conclusions: Low dose diazoxide averted adverse effects but failed to improve outcome in recent-onset T1D.

  • Comment: Diazoxide, at higher doses, has previously been shown to have beneficial effects in preserving β-cell function in recent-onset T1D. However, its use has been limited due to intolerability from side effects. In the current randomised controlled trial, a lower dose was better tolerated but failed to demonstrate preservation of β-cell function. Interestingly, there were differences in A1c, which was lower in the diazoxide group. This may have been a consequence of greater insulin sensitivity and less weight gain in the diazoxide group. The authors propose that low-dose diazoxide could serve as a component in a combination therapy regimen. Indeed, that may be worth exploring.

No effect of the altered peptide ligand NBI-6024 on β-cell residual function and insulin needs in new-onset type 1 diabetes

M. Walter,1A. Philotheou,2F. Bonnici,2A. G. Ziegler,1R. Jimenez,3NBI-6024 Study Group

1Diabetes Research Institute, Forschergruppe Diabetes e.V., Munich, Germany,2Diabetes Clinical Trials Unit, Faculty of Health Sciences, University of Cape Town, Cape Town, South Africa, and3Clinical Development, Neurocrine Biosciences, San Diego, CA, USA

Diabetes Care 2009; 32: 2036–40

Background: Insulin B (9–23) peptide is thought to be an important antigen of T-lymphocytes in autoimmune diabetes in animals and in human beings. It is possible to target specific autoreactive T-lymphocyte activation using a soluble altered peptide ligand (APL) to block or change that activation. NBI-6024 is an APL that contains two amino acid substitutions in the (9–23) sequence of the B-chain of insulin: alanine is substituted for tyrosine at position 16, which is a key contact site at the T-lymphocyte receptor; and alanine also is substituted for cysteine at position 19. The resulting APL (Ala16,19) (also known as NBI-6024) does not activate insulin B (9–23) reactive murine or human T-lymphocytes.

Methods: The study randomised 188 subjects, ages 10–35, to three different doses of the APL (50 subjects to 0.1 mg, 48 subjects to 0.5 mg, 43 subjects to 1.0 mg) or to placebo (47 subjects). Intervention consisted of subcutaneous injections at randomisation, 2 weeks, 4 weeks and monthly until 24 months. The primary end-point was β-cell function – as measured by C-peptide – at 2 years.

Results: At 2 years, the mean peak C-peptide concentration showed no significant difference between any of the treated groups and the placebo group. C-peptide declined linearly in all groups.

Conclusions: The insulin B (9–23) APL (Ala16,19) (NBI-6024) failed to show benefit in any of three doses tested in a large dose-ranging study.

  • Comment: Antigen-specific therapy is thought to be a highly desirable strategy to interrupt the immune processes that result in T1D. Such therapies are generally quite safe, are specific for T1D, and are not expected to alter generalised immune responses. The insulin B (9-23) APL (Ala16,19) (NBI-6024) was shown to alter the course of diabetes in the NOD mouse and prevented activation of both murine and human T-lymphocytes in the laboratory. Preliminary studies demonstrated that there were no safety issues. Thus, the current dose-ranging study was awaited with considerable enthusiasm. The study was well designed and carefully conducted. Unfortunately, the APL had no impact on β-cell function. A critical question is whether an adequate dose was given. It is impossible to measure blood levels of the APL or a marker that would indicate that a given dose was altering some biological response. This is a vexing question that has hampered other studies with antigen-specific interventions. Extrapolation of effective doses from mice to human beings is always a game of roulette. It is unlikely that this particular APL will be further pursued. That is unfortunate.

Autoantigen-specific regulatory T cells induced in patients with type 1 diabetes mellitus by insulin B-chain immunotherapy

T. Orban,1K. Farkas,1H. Jalahej,1J. Kis,1,2A. Treszl,1,3B. Falk,4H. Reijonen,4J. Wolfsdorf,5A. Ricker,1J. B. Matthews,6N. Tchao,6P. Sayre,6P. Bianchine7

1Joslin Diabetes Center, Boston, MA, USA,2Polyclinic of the Hospitaller Brothers, Budapest, Hungary,3Zentrum für Experimentelle Medizin Institut für Medizinische Biometrie und Epidemiologie, Hamburg, Germany,4Benaroya Research Institute at Virginia Mason, Seattle, WA, USA,5Children’s Hospital Boston, MA, USA,6Immune Tolerance Network, University of California, San Francisco, CA, USA, and7National Institute of Allergy and Infectious Diseases, Bethesda, MD, USA

J Autoimmun 2010; 34: 408–15

Background: Insulin B-chain contains the insulin B (9–23) peptide, which is a major epitope recognised by the immune system and thought to be a diabetes-specific antigen. The insulin B-chain fragment is metabolically inactive. Incomplete Freund’s adjuvant (IFA) has been used in a number of vaccines as a delivery system that promotes regulatory immune responses. The current report describes a pilot study testing insulin B-chain administered with IFA in T1D.

Methods: The study randomised 12 subjects (six to 2 mg insulin B-chain in IFA, six to placebo, IFA alone), ages 18–35, with recent-onset T1D. Subjects received a single intramuscular injection of test substance, and were followed for 2 years. Outcome measures were β-cell function – as measured by C-peptide – and insulin-specific humoral and T-lymphocyte responses.

Results: There was no statistical difference in β-cell function, measured every 6 months, between arms. All patients in the experimental group who received insulin B-chain, but none who received placebo, developed robust insulin-specific humoral and T-lymphocyte responses, including insulin B-chain-specific CD4+ T-lymphocytes that were cloned and showed characteristics of regulatory T-lymphocytes.

Conclusions: Insulin B-chain, given together with IFA, may beneficially impact immune response in T1D.

  • Comment: This is another test of an antigen-specific therapy in T1D. It was only a very small pilot study involving 12 subjects, six who received a single injection of insulin B-chain in IFA, and six who received IFA alone. Although claimed to be safe, it is impossible to assess safety in such a small study. Also impossible to assess in a small study is β-cell function, which here was carefully measured with meal tolerance tests. Although there was no statistical difference between groups, the C-peptide levels were persistently higher in the placebo group. Had the results been reversed, it would not be surprising for the authors to claim a beneficial ‘trend’. This illustrates the difficulty in interpretation of small pilot studies. Intriguingly, the cloned T-lymphocytes from the experimental group had characteristics of regulatory T-lymphocytes. But, what if there was also activation of effector T-lymphocytes that was not demonstrated. Could that have resulted in a subtle greater loss of β-cell function? Despite these questions, the results are sufficiently provocative to support further studies with insulin B-chain. Questions that can be raised, however, are whether 2 mg is the right dose and whether a single intramuscular vaccination is the right number of vaccinations. Only full-scale dose-ranging studies can answer these questions. These preliminary data suggest that such studies are warranted.

No effect of the 1α,25-dihydroxyvitamin D3 on β-cell residual function and insulin requirement in adults with new-onset type 1 diabetes

M. Walter,1T. Kaupper,1K. Adler,1J. Foersch,1E. Bonifacio,2A. G. Ziegler1

1Diabetes Research Institute, Forschergruppe Diabetes e.V., Munich, Germany, and2Deutsche Forschungsgemeinschaft Center for Regenerative Therapies, Dresden, Germany

Diabetes Care 2010; 33: 1443–8

Background: Mechanistic studies show that 1α,25-dihydroxyvitamin D3 [1,25(OH)2D3] modulates dendritic cell maturation in vitro and in vivo and facilitates a shift from a Th1 to a Th2 immune response. Studies in the NOD mouse show that 1,25(OH)2D3 reduces the incidence of insulitis and diabetes. Thus, the current study was conducted to determine whether 1,25(OH)2D3 is safe and improves β-cell function in patients with recent-onset T1D.

Methods: Two studies were conducted – a safety study in 25 subjects, all treated with 1,25(OH)2D3 and followed for 18 months, and an efficacy study in which 40 subjects, ages 18–39, with recent-onset T1D, were randomised either to 1,25(OH)2D3 (n = 22) or to placebo (n = 18). The primary end-point was β-cell function – as measured by C-peptide – at 18 months. Careful safety assessments were performed.

Results: There was a similar decline in β-cell function in both groups – 1,25(OH)2D3 and placebo. In addition, A1c and insulin dosage were similar in both groups. The treatment was safe.

Conclusion: 1,25(OH)2D3 was shown to be safe, but did not have a beneficial effect in adults with recent-onset T1D.

Maternal intake of vitamin D during pregnancy and risk of advanced β-cell autoimmunity and type 1 diabetes in offspring

L. Marjamäki,1S. Niinistö,2,3M. G. Kenward,4L. Uusitalo,1,2U. Uusitalo,2M. L. Ovaskainen,2C. Kronberg-Kippilä,2O. Simell,5R. Veijola,6J. Ilonen,7,8M. Knip,9,10S. M. Virtanen1,2,9,11

1Tampere School of Public Health, University of Tampere, Tampere, Finland,2Department of Lifestyle and Inclusion, National Institute for Health and Welfare, Helsinki, Finland,3Department of Public Health, University of Helsinki, Helsinki, Finland,4Department of Epidemiology and Population Health, Medical Statistics Unit, London School of Hygiene and Tropical Medicine, London, UK,5Department of Pediatrics, University of Turku, Turku, Finland,6Department of Pediatrics, University of Oulu, Oulu, Finland,7Immunogenetics Laboratory, University of Turku, Turku, Finland,8Department of Clinical Microbiology, University of Eastern Finland, Kuopio, Finland,9Department of Pediatrics, Tampere University Hospital, Tampere, Finland,10Hospital for Children and Adolescents and Folkhälsan Research Center, University of Helsinki, Helsinki, Finland, and11Research Unit, Tampere University Hospital, Tampere, Finland

Diabetologia 2010; 53: 1599–607

Background: Maternal intake of vitamin D from food or use of vitamin-D-containing supplements during pregnancy was weakly associated with decreased risk of early β-cell autoimmunity in two cohort studies and with clinical T1D in one case–control study.

Methods: Mothers of 3723 infants born between 1997 and 2002 completed a 181-item food frequency questionnaire, including questions on dietary supplements. Offspring were observed at 3–12 month intervals for appearance of T1D-associated autoantibodies and for the development of T1D.

Results: Maternal mean daily intake of vitamin D was 5.1 μg from food and 1.3 μg from supplements. Maternal intake of vitamin D, either from food or from supplements, was not associated with the risk of β-cell autoimmunity or with the risk of T1D in offspring.

Conclusions: Maternal intake of vitamin D did not appear to influence the T1D process in offspring.

  • Comment: For some time, the potential role of vitamin D as a preventative intervention for T1D has been raised. There are clearly effects of vitamin D in vitro and some beneficial effects in animal models, particularly in vitamin-D-deficient animals. In human beings, cohort studies and meta-analyses of studies done in infants have suggested potential benefit. Thus, the use of vitamin D remains an attractive hypothesis for prevention of T1D, the best case being for its use in infants. In the two studies described above, there was no support for this hypothesis. The intervention study with 1,25(OH)2D3 was conducted in adults after the development of T1D. Although a small study, it failed to show a beneficial effect. Intervention with vitamin D at that stage might not be expected to have beneficial effects. The Finnish study evaluated maternal intake of vitamin D and whether it correlated with β-cell autoimmunity or with the risk of T1D in offspring. There are two important points to note, however. First, this was not an intervention study, which might better have permitted assessment of the impact of vitamin D. Second, although the food questionnaire has been validated, it is difficult to assess how well it reflected vitamin D intake, and whether the range of vitamin D intake was sufficient to allow assessment of its impact.

Regeneration of insulin production by autologous bone marrow blood autotransplantation in patients with type 1 diabetes

E. Esmatjes,1,6,7X. Montaña,2M. I. Real,2J. Blanco,1I. Conget,1,6,7R. Casamitjana,3,6,7M. Rovira,4,6R. Gomis,1,6,7P. Marin5,6

1Diabetes Unit, Hospital Clinic Universitari, Barcelona, Spain,2Interventional Angioradiology Unit, Hospital Clinic Universitari, Barcelona, Spain,3Biochemistry and Molecular Genetics Department, Hospital Clinic Universitari, Barcelona, Spain,4Bone Marrow Transplant Unit, Hospital Clinic Universitari, Barcelona, Spain,5Hemotherapy Unit, Hospital Clinic Universitari, Barcelona, Spain,6Institut d’Investigacions Biomèdiques August Pii Sunyer (IDIBAPS), Barcelona, Spain, and7CIBER de Diabetes y Enfermedades Metabólicas Asociadas (CIBERdem), Barcelona, Spain

Diabetologia 2010; 53: 786–9

Background: Some experimental reports suggest that the bone marrow harbours cells capable of differentiation into β-cells, or as a facilitator of β-cell regeneration, or as an immune modulator, and thus use of bone marrow cells has been proposed as treatment for T1D.

Methods: A pilot study was conducted in three subjects with established T1D and absent β-cell function. Autologous bone marrow was harvested and subsequently injected via selective arteriography as an intrapancreatic infusion.

Results: Data were available for two of the three subjects. There was no evidence of β-cell function during 6 months of follow-up.

Conclusions: Intrapancreatic autologous bone marrow infusion had no impact in two subjects with long-standing T1D.

Autologous umbilical cord blood transfusion in very young children with type 1 diabetes

M. J. Haller,1C. H. Wasserfall,2K. M. McGrail,2M. Cintron,1T. M. Brusko,3J. R. Wingard,4S. S. Kelly,5J. J. Shuster,6M. A. Atkinson,2D. A. Schatz1

1Department of Pediatrics, University of Florida, Gainesville, FL, USA,2Department of Pathology, University of Florida, Gainesville, FL, USA,3Diabetes Center, University of California at San Francisco, San Francisco, CA, USA,4Department of Medicine, University of Florida, Gainesville, FL, USA,5Department of Pediatrics, University of Texas, Houston, TX, USA, and6Department of Epidemiology and Health Policy Research and the General Clinical Research Center, University of Florida, Gainesville, FL, USA

Diabetes Care 2009; 32: 2041–6

Background: Umbilical cord blood contains a population of immature unprimed functional regulatory T-lymphocytes. These cells could, in theory, limit inflammatory cytokine responses and anergize effector T-lymphocytes, which are thought to mediate cellular autoimmune processes.

Methods: Children aged > 1 year who developed T1D and had banked umbilical cord blood at an approved centre were recruited into this study. Fifteen subjects enrolled in an open-label phase 1 study using autologous umbilical cord blood infusion. They were followed for 1 year.

Results: There was no evidence at 1 year of maintenance of preservation of β-cell function – as measured by C-peptide – while insulin doses increased over time. No changes were observed in autoantibody titres, regulatory T-lymphocyte numbers, or other immune parameters. There were no significant adverse events.

Conclusions: Autologous umbilical cord blood infusion was safe, but had no beneficial effect in children with T1D.

  • Comment: Many investigators believe that cellular therapies for T1D hold substantial promise. A study from Brazil (1) discussed in last year’s Yearbook evaluated the effects of non-myeloablative autologous haematopoietic stem cell transplantation (AHSCT) in recent-onset T1D. In that study, 20 of 23 subjects became insulin free, some for protracted periods of time. Although many have attributed the results to the infused cells, it is not possible to distinguish whether the putative effects of AHSCT were due to immune reconstitution or otherwise altering the immune-mediated β-cell destruction that eventuates in T1D, to regeneration of β-cells, or more probably to the effects of the profound immunosuppression given at the outset with AHSCT serving to rescue the patient from death from destruction of their immune system. Thus, in human beings there is no clear evidence yet of beneficial effects of any cellular treatment per se in T1D (other than pancreatic or islet transplantation). The two studies described above also failed to show beneficial effects. These negative outcomes, however, should not be cited as a reason to halt research of cellular therapies. In animal models and laboratory experiments, there is clearly much promise for cellular based therapies. Eventually, success will be achieved.

Developing combination immunotherapies for type 1 diabetes: recommendations from the ITN-JDRF Type 1 Diabetes Combination Therapy Assessment Group

J. B. Matthews,1T. P. Staeva,2P. L. Bernstein,1M. Peakman,4,5M. von Herrath,3and ITN-JDRF Type 1 Diabetes Combination Therapy Assessment Group

1Immune Tolerance Network, San Francisco, CA, USA,2Juvenile Diabetes Research Foundation International, New York, NY, USA,3La Jolla Institute for Allergy and Immunology, La Jolla, CA, USA,4Department of Immunology, King’s College London, UK, and5NIHR Biomedical Research Center at Guy’s and St Thomas’ NHS Foundation Trust and King’s College London, London, UK

Clin Exp Immunol 2010; 160: 176–84

  • Comment: This paper describes the results of a workshop conducted by the Immune Tolerance Network (ITN) and the Juvenile Diabetes Research Foundation (JDRF) to review strategies for developing combination therapies for T1D. The advantages of a combination strategy include the ability to minimise toxicities and realise synergies to enhance and prolong efficacy. The paper notes that a clear framework must be developed that specifies the type and quality of preclinical data, including which animal models are acceptable, as well as toxicology and pharmacodynamic data expectations that will be required for a combination to meet acceptable safety standards to justify human trials. It points out that there is a major need for surrogate end-points based upon beneficial immunological responses that manifest soon after treatment, which would facilitate rapid assessment and prioritisation of any individual T1D therapeutic and would be particularly important for selection of potential combinations. The paper notes that biomarkers could speed clinical assessments by providing surrogate end-points, permit stratification of analysis of trial data, and facilitate personalised medicine by informing treatment decisions. The authors gave priority rankings of combination immune-based therapies for T1D, favouring approved agents, favouring combination of immune modulators with antigen-specific therapies, and limiting initial proposals to two agents in combination. The paper is cognizant of regulatory hurdles and the need for negotiations with commercial sponsors. It is a solid paper worth reading by anyone conducting investigations in the field.

Mechanisms, markers, triggers, pathology

The second group of papers includes reports of studies in human T1D that focus on mechanisms, markers, triggers and pathology.

Progression to diabetes in relatives of type 1 diabetic patients: mechanisms and mode of onset

E. Ferrannini,1A. Mari,2V. Nofrate,2J. M. Sosenko,3J. S. Skyler,3DPT-1 Study Group

1Department of Medicine, University of Pisa School of Medicine, Pisa, Italy,2C.N.R. Institute of Biomedical Engineering, Padua, Italy, and3Division of Endocrinology, University of Miami, Miami, FL, USA

Diabetes 2010; 59: 679–85

Background: The Diabetes Prevention Trial – Type 1 (DPT-1) studied relatives of T1D patients at enhanced risk of developing T1D. This study explored the mode of onset of hyperglycaemia and how insulin sensitivity and β-cell function contribute to the progression of the disease.

Methods: The study analysed results from 328 non-diabetic relatives from the control arms of the DPT-1 studies. Subjects had sequential oral glucose tolerance tests performed at baseline, every 6 months, and an average of 2.7 years later, when 115 subjects became diabetic. β-cell glucose sensitivity (slope of the insulin secretion/plasma glucose dose–response function) and insulin sensitivity were obtained by mathematical modelling of the oral glucose tolerance test glucose/C-peptide responses.

Results: Progressors to T1D and non-progressors had similar baseline insulin sensitivity, fasting insulin secretion and total post-glucose insulin output. In contrast, β-cell glucose sensitivity was impaired at baseline and predicted development of diabetes. Moreover, glucose sensitivity began to decline significantly before any changes in plasma glucose, while insulin secretion and insulin sensitivity remained stable.

Conclusions: A defect in β-cell glucose sensitivity is detectable in at-risk subjects years before diagnosis, and anticipates plasma glucose increments.

  • Comment: The failure of the β-cell to adequately respond to glucose –β-cell ‘blindness’ to glucose – is a characteristic pathophysiological abnormality in both T1D and type 2 diabetes. It results in a lack of adequate early insulin secretory response that eventuates in significant postprandial hyperglycaemia. Here it is demonstrated that this defect is present very early in the disease process, before other metabolic abnormalities. Teleologically, it may be a way the injured β-cell manifests itself. Presumably, in T1D, the β-cell has been injured via immunological attack. Indeed, if humoral anti-islet antibodies are a marker of the immune attack, all of the subjects included in DPT-1 already had that under way, as an entry criterion was presence of antibodies. It would be interesting to do similar modelling in studies involving genetically high risk individuals who do not yet have anti-islet antibodies, as it is possible that a defect in β-cell glucose sensitivity is genetically determined and may indicate that a particular subject is more vulnerable to immune-mediated damage.

Pancreatic islet autoantibodies as predictors of type 1 diabetes in the Diabetes Prevention Trial – Type 1

T. Orban,1J. M. Sosenko,2D. Cuthbertson,3J. P. Krischer,4J. S. Skyler,2R. Jackson,1L. Yu,5J. P. Palmer,6D. Schatz,7G. Eisenbarth,5Diabetes Prevention Trial – Type 1 Study Group

1Joslin Diabetes Center, Boston, MA, USA,2Division of Endocrinology, University of Miami, Miami, FL, USA,3Pediatrics Epidemiology Center, University of South Florida, Tampa, FL, USA,4Division of Informatics and Biostatistics, University of South Florida, Tampa, FL, USA,5Barbara Davis Center for Childhood Diabetes, Denver, CO, USA,6Division of Endocrinology/Metabolism, University of Washington, Seattle, WA, USA, and7Division of Endocrinology, University of Florida, Gainesville, FL, USA

Diabetes Care 2009; 32: 2269–74

Background: In another report from the DPT-1 studies, the prediction of T1D by specific types of pancreatic islet autoantibodies, either alone or in combination, was explored.

Methods: The study included two cohorts from DPT-1, both derived from initial screening for islet cell autoantibodies (ICAs). Also measured were antibodies to GAD65, ICA512 and insulin. Participants were followed for the occurrence of T1D. One cohort, called ‘Trials’, included 528 DPT-1 participants (83.3% ICA+) who were randomised in the DPT-1 studies. The other cohort, called ‘Questionnaire’, included 28,507 individuals (2.4% ICA+) who did not enter the DPT-1 trials but responded to questionnaires as to T1D status.

Results: In both cohorts autoantibody number was highly predictive of T1D. In the Questionnaire cohort, as single autoantibodies, ICA (3.9%), GAD65 (4.4%) and ICA512 (4.6%) were similarly predictive of T1D, whereas no subjects with only insulin autoantibodies developed T1D.

Conclusions: The number of autoantibodies is predictive of T1D. However, there are differences in prediction based on antibody type and titre.

  • Comment: The finding that autoantibody number predicts T1D is consistent with other studies. However, this represents the largest prospective study of relatives yet to be reported. It confirms that an individual with only a single antibody being positive is far less likely to develop T1D than those with two or more antibodies. It also finds that the particular type and titre of an autoantibody can influence prediction. Over the years, our tools for prediction of T1D have continued to improve. This has been demonstrated in the DPT-1 studies, in the ENDIT (European Nicotinamide Diabetes Intervention Trial) study and many others. Yet, the focus of many studies has been on relatives of individuals with T1D. It is probably time to advance screening and prediction efforts more widely into the general population. This might best be accomplished with screening at birth for high risk HLA genotypes, as has been done in studies such as DIPP (Diabetes Prediction and Prevention), DAISY (Diabetes Autoimmunity Study in the Young) and TEDDY (The Environmental Determinants of Diabetes in the Young).

Preservation of β-cell function in autoantibody-positive youth with diabetes

C. J. Greenbaum,1A. M. Anderson,2L. M. Dolan,3E. J. Mayer-Davis,4D. Dabelea,5G. Imperatore,6S. Marcovina,7C. Pihoker,8SEARCH Study Group

1Diabetes Research Program, Benaroya Research Institute, Seattle, WA, USA,2Wake Forest University School of Medicine, Winston-Salem, NC, USA,3Cincinnati Children’s Hospital and Medical Center, Cincinnati, OH, USA,4Department of Nutrition, University of North Carolina at Chapel Hill, Chapel Hill, NC, and the Department of Epidemiology and Biostatistics, University of South Carolina, Columbia, SC, USA,5Department of Epidemiology, Colorado School of Public Health, University of Colorado, Denver, CO, USA,6Division of Diabetes Translation, National Center for Chronic Disease Prevention and the Health Promotion Centers for Disease Control and Prevention, Atlanta, GA, USA,7Department of Medicine, University of Washington, Seattle, WA, USA, and8Department of Pediatrics, University of Washington, Seattle, WA, USA

Diabetes Care 2009; 32: 1839–44

Background: SEARCH is a population-based study conducted at six centres in the USA, including existing (prevalent) and newly diagnosed (incident) cases of diabetes in youth. This report describes the extent of β-cell function in youth with diabetes who have GAD65 and/or IA2 autoantibodies.

Methods: Fasting C-peptide levels were obtained from 2789 GAD65 and/or IA2 autoantibody-positive youth aged 1–23 years. Preserved β-cell function was defined on the basis of cut points derived from the Diabetes Control and Complications Trial (DCCT) (fasting C-peptide ≥ 0.23 ng/ml) and from the adolescent population of the National Health and Nutrition Examination Survey (NHANES) 5th percentile for fasting C-peptide (≥ 1.0 ng/ml). The clinical characteristics between those with and without preserved β-cell function were compared.

Results: Within the first year of diagnosis, 82.9% of youth had a fasting C-peptide ≥ 0.23 ng/ml and 31.2% had values ≥ 1.0 ng/ml. Among those with 5 or more years of diabetes, 10.7% had preserved β-cell function based on the DCCT cutoff and 1.0% were above the NHANES 5th percentile.

Conclusions: Early on, the vast majority of antibody-positive youth with diabetes have significant residual β-cell function, but by 5 years of diabetes, only 10% do.

  • Comment: The SEARCH study data, as well as the control groups in the intervention studies discussed earlier, clearly demonstrate that β-cell function is not rapidly lost after diagnosis of diabetes, but rather that over 80% of youth have significant residual β-cell function within the first year of diagnosis, and 10% continue to have significant residual β-cell function after 5 years. This may have important implications as successful intervention strategies evolve, as those with significant residual β-cell function may potentially benefit from intervention, which in most trials is currently being studied only in recent-onset T1D.

Breastfeeding patterns of mothers with type 1 diabetes: results from an infant feeding trial

S. Sorkio,1D. Cuthbertson,2S. Bärlund,1A. Reunanen,3A. M. Nucci,4C. L. Berseth,5K. Koski,6A. Ormisson,7E. Savilahti,8U. Uusitalo,2J. Ludvigsson,9D. J. Becker,10J Dupré,11J. P. Krischer,2M. Knip,8,12,13H. K. Akerblom,6S. M. Virtanen,1,14,15TRIGR Study Group

1Nutrition Unit, National Institute for Health and Welfare, Helsinki, Finland,2Pediatrics Epidemiology Center, University of South Florida, Tampa, FL, USA,3National Institute for Health and Welfare, Helsinki, Finland,4Division of Nutrition, School of Health Professions, College of Health and Human Sciences, Georgia State University, Atlanta, GA, USA,5Bristol-Myers Squibb, Evansville, IN, USA,6Institute of Clinical Medicine, University of Helsinki, Helsinki, Finland,7Department of Paediatrics, University of Tartu, Tartu, Estonia,8Hospital for Children and Adolescents, University of Helsinki, Helsinki, Finland,9Division of Pediatrics, Department of Health and Environment, Faculty of Health Sciences, University of Linköping, Linköping, Sweden,10Children’s Hospital of Pittsburgh, Pittsburgh, PA, USA,11Robarts Research Institute, London, ON, Canada,12Folkhlsan Research Center, University of Helsinki, Helsinki, Finland,13Department of Pediatrics, Tampere University Hospital, Tampere, Finland,14Tampere School of Public Health, University of Tampere, Tampere, Finland, and15Research Unit, Tampere University Hospital, Tampere, Finland

Diabetes Metab Res Rev 2010; 26: 206–11

Background: It had previously been reported that mothers with T1D are less likely to breastfeed their children or breastfeed for a shorter period of time than non-diabetic mothers. The aim of this study was to prospectively examine the breastfeeding patterns among mothers with and without T1D.

Methods: Families from an infant feeding study (TRIGR) included 2160 babies; 1096 were born to women with T1D and 1064 to unaffected women. Information on infant feeding was acquired by frequent prospective dietary interviews.

Results: Most (> 90%) of the infants were initially breastfed, regardless of the mother’s T1D status. However, breastfeeding rates declined more steeply among mothers with T1D than without T1D, being 50% and 72% at 6 months, respectively. After adjusting for age, educational status, timing of delivery and caesarean section rate, all factors associated with the termination of breastfeeding, there was no difference in the duration of breastfeeding among mothers with and without T1D.

Conclusions: Maternal diabetes status was not itself associated with shorter breastfeeding. Shorter duration of breastfeeding in mothers with T1D could be explained by their more frequent caesarean sections, earlier delivery and lower age and education.

  • Comment: Although mothers with T1D breastfed for a shorter duration, this did not seem to be related to T1D itself, but rather to other confounding factors, i.e. differences in the mode of delivery, length of gestation, parental age and education. Breastfeeding has substantial health advantages and thus is desirable for all infants, and should be encouraged. It also may impact the frequency of development of T1D. Yet, the TRIGR study (2), from which this study is derived, is evaluating the content of infant formula – cow’s milk based versus casein hydrolysate – in order to ascertain whether early introduction of cow’s milk may serve as an environmental trigger for T1D. It will be several years before the results of this important study become available. In the meantime, breastfeeding defers the introduction of any formula, and prolonged breastfeeding may be beneficial in terms of forestalling T1D. It would be unfortunate if the presence of maternal T1D directly led to shorter duration of breastfeeding. This study also highlights a separate issue. Namely, one has to wonder whether greater efforts should be made to have women with T1D progress to spontaneous labour and vaginal delivery in contrast to the not uncommon practice of early delivery by caesarean section.

Prevalence of enteroviral capsid protein vp1 immunostaining in pancreatic islets in human type 1 diabetes

S. J. Richardson,1A. Willcox,1A. J. Bone,2A. K. Foulis,3N. G. Morgan1

1Institute of Biomedical and Clinical Sciences, Peninsula Medical School, Plymouth, UK,2School of Pharmacy and Biomolecular Sciences, University of Brighton, Brighton, UK, and3Department of Pathology, Royal Infirmary, Glasgow, UK

Diabetologia 2009; 52: 1143–51

Background: Much evidence suggests that enterovirus infection may be an important environmental trigger of T1D. This study examined pancreatic islets in autopsy specimens for evidence of enteroviral infection.

Methods: Pancreatic autopsy specimens were examined, including 72 from recent-onset T1D (mean age 12.65 ± 1.1 years, range 1–42 years, and a mean time since diagnosis of 8.2 ± 4.1 months, range 0–6 years). A large number of control specimens were examined, including 25 pancreases from adult patients with type 2 diabetes. Serial sections were stained for insulin, glucagon, enteroviral capsid protein vp1, protein kinase R (PKR) (which is unregulated in islets with enterovirus infection), MHC class 1 and CD45. Double immunofluorescence staining was performed. Islets were classified as insulin-containing islets (ICIs), presumably residual β-cells, or insulin-deficient islets (IDIs). (In the T1D pancreases, 60% of islets are IDIs.)

Results: The majority (44 of 72, 61%) of recent-onset T1D cases were positive for enteroviral vp1 antigen. Enteroviral capsid protein vp1 immunostaining was restricted to ICIs, and double immunofluorescence demonstrated that this is restricted to β-cells. Other markers of infection (PKR, MHC-1) were also present. Enteroviral capsid protein vp1 was detected in a few cells in the pancreases of three of 39 (7.7%) non-diabetic paediatric cases. A total of 40% of type 2 diabetic pancreases also stained for enteroviral capsid protein vp1.

Conclusions: Enterovirus staining is common in pancreases from recent-onset T1D. Interestingly, it is also seen in some pancreases from individuals with type 2 diabetes.

  • Comment: These findings provide further support for a potential role of enterovirus infection as an environmental factor leading to T1D. There is a growing body of evidence supporting this concept, which is being studied in detail by the TEDDY Study Group (3). Identification of environmental factors that are important in the aetiology or pathogenesis of T1D may facilitate the development of intervention strategies to reduce or eliminate the impact of such factors, and thus may be important in the eventual prevention of T1D.

Recurrence of type 1 diabetes after simultaneous pancreas–kidney transplantation, despite immunosuppression, is associated with autoantibodies and pathogenic autoreactive CD4 T-cells

F. Vendrame,1A. Pileggi,1,2E. Laughlin,3G. Allende,1A. Martin-Pagola,1R. D. Molano,1S. Diamantopoulos,1N. Standifer,3,4K. Geubtner,3B. A. Falk,3H. Ichii,1,2H. Takahashi,2I. Snowhite,1Z. Chen,5A. Mendez,1,6L. Chen,2J. Sageshima,2P. Ruiz,2G. Ciancio,2C. Ricordi,1,2,5,6H. Reijonen,3G. T. Nepom,3G. W. Burke,1,2A. Pugliese1,5,6

1Diabetes Research Institute, Leonard Miller School of Medicine, University of Miami, Miami, FL, USA,2Department of Surgery, Division of Transplantation, Leonard Miller School of Medicine, University of Miami, Miami, FL, USA,3Benaroya Research Institute, Seattle, WA, USA,4Clinical Immunology, Amgen Inc., Seattle, WA, USA,5Department of Microbiology and Immunology, Leonard Miller School of Medicine, University of Miami, Miami, FL, USA,6Department of Medicine, Division of Endocrinology and Metabolism, Leonard Miller School of Medicine, University of Miami, Miami, FL, USA

Diabetes 2010; 59: 947–57

Background: Recurrent autoimmunity was first described in pancreas transplants from identical twin donors, who did not need immunosuppression to prevent graft rejection. The current study explores the nature of recurrent diabetes in pancreas–kidney transplant recipients who were receiving immunosuppression therapy.

Methods: In three patients with recurrent diabetes after pancreas–kidney transplantation, there was serial monitoring of antibodies and autoreactive T-lymphocytes, as well as pancreas biopsy.

Results: Pancreas transplant biopsies were taken after hyperglycaemia recurrence and revealed β-cell loss and insulitis, and no evidence of graft rejection. From the time of biopsy and on further follow-up, both antibodies and autoreactive T-cells were repeatedly demonstrated. Immune monitoring during immunosuppressive therapy showed that autoimmunity was not resolved by the agents used.

Conclusions: Recurrent autoimmunity may be responsible for hyperglycaemia recurrence in pancreatic transplant recipients, despite immunosuppressive therapy.

  • Comment: Loss of function after time is a common feature in recipients of pancreas or islet transplants. There are many possible reasons for this, one of which is the recurrence of autoimmunity. Yet, it might be expected that the immunosuppressive agents used to prevent allograft rejection would also prevent recurrent autoimmunity. In the patients described here, that is clearly not the case, as there was no evidence of rejection either in the pancreas biopsy or in the function of the transplanted kidney. Yet, biopsies revealed β-cell loss and insulitis, and both autoantibodies and autoreactive T-lymphocytes were demonstrable. This causes one to conclude that the mechanisms responsible for allograft rejection and for autoimmunity are different, and may require different forms of immune intervention to control each. This has profound importance in transplantation of these organs, particularly because simultaneous pancreas and kidney transplantation is the procedure of choice for treatment of end stage renal disease in individuals with T1D. It also offers important insights into evolving appropriate immune modulation strategies to interrupt the T1D disease process, either for prevention or to sustain β-cell function in recent-onset T1D.

Dimorphic histopathology of long-standing childhood-onset diabetes

R. Gianani,1M. Campbell-Thompson,2S. A. Sarkar,1C. Wasserfall,2A. Pugliese,3J. M. Solis,1S. C. Kent,4B. J. Hering,5E. West,1A. Steck,1S. Bonner-Weir,6M. A. Atkinson,2K. Coppieters,7M. von Herrath,7G. S. Eisenbarth1

1Barbara Davis Center for Childhood Diabetes, University of Colorado Denver, Aurora, CO, USA,2Department of Pathology, University of Florida at Gainesville, Gainesville, FL, USA,3Diabetes Research Institute, Miami, FL, USA,4Center for Neurologic Diseases, Brigham, and Women’s Hospital Harvard Medical School, Boston, MA, USA,5Schulze Diabetes Institute, University of Minnesota, Minneapolis, MN, USA,6Joslin Diabetes Center, Harvard Medical School, Boston, MA, USA, and7La Jolla Institute for Allergy and Immunology, La Jolla, CA, USA

Diabetologia 2010; 53: 690–8

Background: Analysis of C-peptide in children characterised at diabetes onset for autoantibodies shows heterogeneous preservation of insulin secretion in long-standing diabetes. This study sought to characterise pancreases of childhood-onset diabetes in order to define the pathological basis of that heterogeneity.

Methods: Pancreases were evaluated from 20 cadaveric organ pancreases from individuals with childhood-onset (mean age of onset 11.2 years, range 3–18 years) long-term (mean duration 14 years, range 1–35 years) T1D. Pancreatic histology, islet autoantibodies and C-peptide of patients were analysed.

Results: Most (70%) of the pancreases had only IDIs. C-peptide was not present in patients lacking histological evidence of β-cells. Of six patients with extant β-cells, these could be divided into two patterns. In pattern A, there were mostly insulin-deficient islets (IDIs) and lobular retention of areas with ‘abnormal’β-cells; islets retaining β-cells contained survivin and enhanced HLA class 1, had low or no C-peptide, and were positive for anti-islet autoantibody at time of death. In pattern B, none had IDIs, with all islets containing normal-appearing but quantitatively reduced β-cells, without survivin or HLA class 1; they were antibody-negative with relatively high C-peptide, and lacked high risk HLA alleles.

Conclusions: The data suggest that C-peptide secretion in long-standing diabetic patients can be explained by two different patterns of β-cell survival, one (pattern A) that probably reflects autoimmune type 1A diabetes, and the other possibly reflecting a different subset of T1D.

  • Comment: The investigators used pancreases collected by the JDRF nPOD (Network for Pancreatic Organ Donors with Diabetes) consortium to evaluate the histology of patients with long-standing childhood diabetes. Although most patients had the expected lack of β-cells and lack of C-peptide, 30% showed persistent C-peptide, and these could be divided into two patterns. One pattern was consistent with classical type 1A diabetes presumably on an immune basis. The other pattern appeared to be totally different, with normal-appearing islets, relatively high C-peptide, absence of high risk HLA, and lack of antibodies at the time of death. All were said to have had the typical clinical appearance of T1D at the time of diagnosis. This suggests that we still have a lot to learn about the evolution of T1D, and that we need better tools at diagnosis to better characterise individuals. This may prove to be particularly important if we are to appropriately select patients with recent-onset T1D for newer immunological interventions, either in the context of clinical trials or for treatment when such agents are approved by regulatory agencies.

Overall commentary

This year has been marked by a stunning number of papers relating to immune intervention of T1D. Longer term observations suggest that there may be sustained effects of a short course of anti-CD3 monoclonal antibody, which depletes T-lymphocytes and on recovery seems to favour regulatory T-lymphocytes over effector T-lymphocytes. That B-lymphocytes might be involved in T1D pathogenesis was suggested by positive results from a trial with an anti-CD20 monoclonal antibody, rituximab. There were also provocative observations from an extraordinarily tiny study with insulin B-chain given with IFA. Yet, most intervention studies were negative – mycophenolate mofetil given alone or in combination with daclizumab, low-dose diazoxide, vitamin D, an altered peptide ligand of insulin B (9–23), autologous bone marrow blood injected into the pancreatic arterial system, and umbilical cord blood transfusion. The combination of low-dose cyclosporine and methotrexate in a very small pilot study was impossible to interpret, and is unlikely to be pursued further anyway. That negative studies continue to dominate the field, and that the positive ones still show decline in β-cell function over time, has led to more calls for combination approaches. When I have advanced such prospects at meetings of paediatric diabetologists, I hear groans. Yet when I have advanced these prospects at meetings of immunologists and transplant surgeons, I hear cheers. Hopefully, preclinical studies will provide further guidance to support combination therapies, and hopefully regulatory agencies will offer a path for evaluation of these in human beings.

Meanwhile, a number of interesting studies have offered further insights into the pathology and pathogenesis of T1D, the pathways by which it may unfold, and how better to quantify who is at risk of developing T1D. These studies have also suggested that T1D is not the same in everybody who appears to develop it. And, perhaps most importantly, that autoimmunity may not be able to be addressed with the same therapeutic agents that are used to prevent transplant rejection.

It has been a very exciting year, indeed.