An autoimmune condition
There are many pointers to an immune-mediated basis for human type 1 diabetes. These include the association of type 1 diabetes with other autoimmune disorders and with the HLA system, which modulates immune responses; lymphocytic infiltration of the islets in post mortem human pancreas (insulitis); evidence of cell-mediated and humoral immunity directed against islet constituents; transfer experiments in animals, evidence of transmission of diabetes from marrow donors to recipients; recrudescence of autoimmunity in pancreatic grafts between identical twins; and preservation of beta cell mass and function by immune interventions.
Evidence that type 1 diabetes is immune-mediated
Figure 1. Presence of other auto-immune antibodies in 491 children with type 1 diabetes. One of the earliest pointers to an autoimmune aetiology was the observation that other autoimmune conditions were over-represented in juvenile cases of diabetes and their relatives, but not in late-onset cases. Type 1 diabetes overlaps with autoimmune thyroid disease, coeliac disease, Addison’s disease and pernicious anaemia among other organ-specific autoimmune conditions (see Figure 1). Autoantibodies directed against islet constituents precede the onset of clinical disease by many years and can be used to predict it; they are present in 95% of newly presenting patients. Moreover, they are also more prevalent in first degree relatives of patients than in the general population (Figure 2).
Figure 2. Presence of various-auto-antibodies in patients with recent onset type 1 diabetes, in their first-degree relatives and in healthy controls. Several islet antigens have been characterised, and these include insulin itself, the enzyme glutamic acid decarboxylase (GAD), protein tyrosine phosphatase (IA-2) and the cation transporter ZnT8. Animal models of autoimmune diabetes such as the non-obese diabetic (NOD) mouse have been extensively studied. The observation that treatment with immunosuppressive agents such as ciclosporin prolongs beta cell survival in newly diagnosed patients has confirmed that the disease is immune-mediated.
Autopsies of younger patients who died shortly after diagnosis of diabetes show variable degrees of lymphocytic infiltration, resembling that seen in autoimmune thyroiditis. Willy Gepts was the first to show that insulitis is largely confined to recently diagnosed cases in the young and suggested its significance as a marker of autoimmunity.
Nerup observed in 1971 that leucocyte migration was inhibited by the presence of islet tissue in juvenile but not late-onset cases. Cell transfer experiments in the NOD mouse have established that autoimmune diabetes is a T cell mediated disease requiring the presence of both CD4 and CD8 cells.
The identification of islet cell antibodies (ICA) in 1974 was a key step towards the recognition of type 1 diabetes as an immune-mediated disease. Autoantibodies directed against islet constituents precede the onset of clinical disease, often by many years, and can be used to predict it; they are present in up to 95% of newly presenting patients. Several islet antigens have been characterised, and these include insulin itself, the enzyme glutamic acid decarboxylase (GAD), protein tyrosine phosphatase (IA-2) and the cation transporter ZnT8.
A viral aetiology?
Although there have been occasional case reports of diabetes developing rapidly after fulminant viral infections, the typical long latent period suggests that diabetes is rarely the direct of consequence of direct damage by viral infection. Viruses do however show tissue tropism, mouse models of virally-induced diabetes are available, and the phenomenon of MHC-restriction (susceptibility to specific viruses associated with certain HLA types) is well established. Some investigators therefore conceive of type 1 diabetes as an condition in which beta cells subject to viral infection are capable of provoking a secondary immune-mediated response resulting in diabetes. The concept of viral causation is thus consistent with an immune-mediated aetiology of diabetes, but this is by no means established beyond doubt.
Further evidence of autoimmunity in type 1 diabetes has accumulated over the past 25 years. The establishment of animal models represented a major step forward. The Wistar-derived rat from the Bio Breeding Laboratories of Canada Ltd – the BB rat – was first described in 1976. These non-obese animals developed a spontaneous insulin-deficient, ketosis-prone form of diabetes associated with histological features of insulitis. The most widely-used model of type 1 diabetes, the non-obese diabetic (NOD) mouse, was described in 1980.
Transfer of diabetes in man
Evidence in humans is more anecdotal, and comes from bone marrow transplantation from donors with type 1 diabetes to non-diabetic recipients. Analysis of the International Bone Marrow Transplant Registry showed nine cases in which this is known to have occurred; six recipients died within 2 years of marrow donation without developing diabetes, but two of three long-term survivors developed the disease.
Twin-to-twin pancreas transplants
Further support comes from twin-to-twin pancreas transplants. Unaffected monozygotic twins in long-term discordant pairs are considered unlikely to develop type 1 diabetes, and can therefore donate the tail of their pancreas to their affected co-twin. Attempts to graft from one twin to another were, however, foiled by recrudescence of the type 1 disease process in the transplanted organ. Insulitis developed within weeks of transplantation and progressed rapidly to diabetes, presumably because of reactivation of memory T cells and re-enactment of autoimmune destruction of the islets.
Final proof that immune processes mediate human type 1 diabetes comes from trials of immune intervention. Most convincing among these were the trials of ciclosporin A in the newly diagnosed. Two major trials showed that immune intervention in recently diagnosed patients prolonged beta cell survival and allowed a proportion of cases to survive without insulin for more than a year.  The effect is lost as soon as the drug is withdrawn, however, and the risk:benefit ratio for this form of intervention is unfavourable. Many other forms of immunotherapy have been tried since then, with (as yet) limited success.
^ Gepts W, LaCompte PM. The pancreatic islets in diabetes. Am J Med 1981;70:105–15.
^ Gale EAM. The discovery of type 1 diabetes. Diabetes 2001;50:217–26
^ Lampeter EF, McCann SR, Kolb H. Transfer of diabetes type 1 by bone-marrow transplantation. Lancet 1998;351(9102): 568–9
^ Sutherland DE, Goetz FC, Sibley RK. Recurrence of disease in pancreas transplants. Diabetes 1989;38 (suppl 1):85–7
^ Feutren G et al. Cyclosporin increases the rate and length of remissions in insulin-dependent diabetes of recent onset. Lancet 1986;2:119–23
^ Canadian-European Randomized Control Trial Group. Cyclosporin-induced remission of IDDM after early intervention. Association of 1 year of cyclosporin treatment with enhanced insulin secretion. Diabetes 1988;39:1574–82