Controversies in medicine: redefining the diagnosis of type 1 diabetes

Diagnosis of autoimmune β-cell disorder before end-stage clinical type 1 diabetes is a key step towards the prevention of this disease

The incidence of type 1 diabetes has been rising by about 3% per year in children worldwide, with some plateauing since 2005 in Australia.1 Intensive management with technological advances in insulin delivery and glucose monitoring have improved the management and eased the burden for patients with type 1 diabetes. However, most patients still do not achieve optimal blood glucose control to minimise the risk of long term vascular complications.2 Prevention remains the ultimate goal.

The diagnosis of type 1 diabetes

The classical presentation of increasing polyuria, polydipsia and weight loss over days to weeks is familiar to practitioners and leads to the diagnosis of type 1 diabetes. However, clinical diagnosis is the end stage of subclinical pathology over months to years, during which time β‐cells in the islets of the pancreas undergo autoimmune destruction. This process begins early in life and is revealed by circulating autoantibodies to islet antigens, which denote pancreatic β‐cell or islet autoimmunity. Type 1 diabetes is clinically heterogeneous but most children who develop type 1 diabetes have detectable autoantibodies before the age of 5 years to at least two islet antigens: insulin, glutamic acid decarboxylase 65 (GAD65), insulinoma antigen 2 (IA2) and/or zinc transporter 8 (ZnT8).3

Sensitivity and specificity of enzyme‐linked immunosorbent assays for islet autoantibodies are 60–80% and 95–100%, respectively. The radiobinding assay for insulin autoantibodies — detected predominantly in young children — has a similar high specificity but lower sensitivity. These assays are being superseded by the development of multiplex, ultrasensitive and low cost islet autoantibody assays, without compromise of specificity. The risk of overdiagnosis is mitigated further by first screening for multiple antibodies and confirming all positive autoantibody results.4

Progression to diabetes is faster the younger the age when autoantibodies are first detected and in children with high risk human leucocyte antigen (HLA) class II susceptibility haplotypes DR3‐DQ2 and DR4‐DQ8. Among children with two or more islet autoantibodies before age 3 years, 57% and 75% progress to diabetes by age 6 and 10 years, respectively.3

The stages of type 1 diabetes

Children with two or more islet autoantibodies who are normoglycaemic and asymptomatic are defined as having stage 1 type 1 diabetes,5 encompassing the months to years of subclinical disease. Stage 2 is the intermediate stage that may last many months, when blood glucose rises transiently without symptoms, especially after food. Stage 3 is the classic presentation of type 1 diabetes with symptomatic fasting hyperglycaemia requiring insulin therapy (Box).

Islet autoantibodies are detected in several contexts: with stress hyperglycaemia associated with fever or treatment with sympathomimetic agents; in screening relatives of people with type 1 diabetes, 4% of whom have at least one islet autoantibody;6 and in screening pre‐school children in the general population, of whom 0.4% in Germany had more than one islet autoantibody.4 Screening studies remain in the research domain. Individuals with islet autoantibodies should be referred to paediatric or adult diabetologists for counselling and advice, including about current prevention trials.

Autoimmune β‐cell disorder

In other autoimmune diseases, the onset of symptoms and pathology is usually concomitant. For example, children with juvenile idiopathic arthritis or adults with rheumatoid arthritis present with painful, swollen joints when treatment with newer biological agents may favourably alter disease progression and preserve the joints. In type 1 diabetes, however, the pathology is latent until clinical presentation as a metabolic disorder in end‐stage disease. Treatment to counter autoimmunity and preserve residual β‐cell function at this stage stands less chance of success — akin to treating arthritis after it has destroyed the joints.

In order to advance our understanding of type 1 diabetes and the likelihood of its prevention, a paradigm shift would redefine type 1 diabetes as an autoimmune β‐cell disorder (ABCD) that begins with asymptomatic islet autoimmunity (stage 1) and not as a metabolic disorder resulting from end‐stage pathology (stage 3).7

The benefits of ABCD

The ABCD paradigm has important potential benefits.7 It may minimise the risk of diabetic ketoacidosis (DKA), decrease psychological stress in affected families, and lead to earlier initiation of treatment trials with a higher likelihood of preserving insulin production and preventing clinical disease.

Minimising the risk of diabetic ketoacidosis

Up to 30% of Australian children with newly diagnosed diabetes present with DKA.8,9 Prevention of DKA is an important target in the Australian National Diabetes Strategy 2016–2020.8 With access to intensive care facilities, it is now rare for a patient to have serious life‐threatening complications of DKA. However, DKA causes significant post‐traumatic stress for families. Emerging evidence indicates that children who present with DKA have worse metabolic control in the longer term.10 In an American study, glycated haemoglobin tracked 1.4% higher after severe DKA and 0.9% higher after mild or moderate DKA for up to 15 years of follow‐up after diagnosis.10 These increments are sufficient to increase the risk of long term vascular complications. Campaigns to alert primary care providers about the benefits of early diagnosis of type 1 diabetes in children have shown some impact.9 Children with islet autoimmunity followed prospectively in research studies consistently have a markedly lower risk of DKA.11,12 This is particularly apparent in young children in whom severe DKA at presentation is more common. These observations provide an impetus for general population screening for islet autoimmunity, currently underway in Europe4 and the United States.13 These research programs seek to provide evidence of cost‐effectiveness and facilitate intervention with research treatments in the asymptomatic stages of disease.

Decreasing psychological stress for families

Research studies that follow children at increased genetic risk of type 1 diabetes not surprisingly reveal heightened parental anxiety when a child is identified at increased risk by genetic or autoantibody screening. However, anxiety decreases over time and at clinical diagnosis is less, along with better coping ability, than in families in which the diagnosis was unheralded.14 General population screening in Europe includes measures of family and child anxiety,4 which may differ from those experienced by relatives of those with type 1 diabetes. Australian data are lacking but important to acquire because of cultural and regional differences.

Preventing type 1 diabetes

Interventions being investigated to preserve β‐cell function target different stages: pre‐stage 1 before islet autoimmunity (primary prevention), stages 1 and 2 after onset of islet autoimmunity and before the symptoms of diabetes (secondary prevention), and stage 3 soon after clinical diagnosis. Intervention at any stage may require combination treatments but earlier intervention is more likely to be successful.7 Screening for secondary prevention trials has been limited to first‐degree relatives but will extend to the general population. While there are yet no proven means to prevent type 1 diabetes, several agents, including antithymocyte globulin; the anti‐CD3 monoclonal antibodies, teplizumab and otelixizumab; abatacept (CTLA4‐Ig); and the anti‐CD20 monoclonal antibody, rituximab, show promise in preserving residual β‐cell function in recent‐onset clinical type 1 diabetes. Significantly, teplizumab delayed progression to type 1 diabetes in high risk participants in a recent landmark trial.15

The challenges of ABCD

Only 10% of children presenting with type 1 diabetes have an affected first‐degree relative. Therefore, most at‐risk individuals with ABCD would need to be identified by general population screening. However, unlike in relatives, the specificity, sensitivity and predictive value of islet autoantibody screening in the general population has to be established prospectively, although one study16 suggested that general population screening would be effective. This research will require consultation with public health experts, health economists and other professionals, and the diabetes and wider community to evaluate its cost–benefit ratio. At present, the costs of screening and follow‐up are likely to outweigh the savings from preventing DKA and potentially alleviating family stress. As many years may elapse before some children with multiple antibodies develop type 1 diabetes, there is a substantial obligation to provide adequate family support during follow‐up. The major personal and economic impact of screening will result when identifying ABCD in children allows for earlier intervention to prevent progression to clinical type 1 diabetes. While type 1 diabetes has been recognised as an autoimmune disease for several decades, clinical recognition of the first stage of disease as an autoimmune rather than a metabolic disorder would set the stage for expanded investment by academia, consumer bodies and industry in prevention trials, including with agents known to be efficacious in other autoimmune diseases.

Conclusion

Diagnosis of ABCD before end‐stage clinical type 1 diabetes and hyperglycaemia is a key step towards the prevention of type 1 diabetes. Screening of Australian children in the general population for islet autoimmunity as a research undertaking is likely to be well supported by advocacy consumer groups and would complement international efforts. Experience with screening at‐risk family members indicates general enthusiasm for early diagnosis and participation in intervention trials. Australia is fortunate in having robust research networks of type 1 diabetes clinicians and researchers that monitor the natural history of type 1 diabetes in genetically at‐risk families, screen for ABCD, and conduct intervention trials aiming to improve clinical outcomes. These networks — the Australasian Diabetes Data Network (www.addn.org.au), Environmental Determinants of Islet Autoimmunity (ENDIA) (www.endia.org.au), Diabetes Research Centre (www.diabetesresearchcentre.org.au) and Type 1 Diabetes TrialNet (www.trialnet.org) — are a strong foundation for population screening for ABCD and clinical trials to prevent type 1 diabetes.

Box – Autoimmune β‐cell disorder leading to clinical type 1 diabetes

---

Figure adapted from: Eisenbarth GS. Type I diabetes mellitus. A chronic autoimmune disease. N Engl J Med 1986; 314: 1360‐1368.