Improving recognition and treatment of adult onset type 1 diabetes though though rational cost-effective classification biomarker testing

Research question: In adults with recent onset diabetes, when should NHS guidelines recommend biomarker tests to assist diagnosis of diabetes subtype, and which tests should be recommended? Background Over 5 million people live with type 1 diabetes (T1D) or type 2 diabetes (T2D) in the UK. These conditions have very different management but differentiating between them at diagnosis is challenging. Misclassification is therefore common: approximately 1 in 3 adults developing T1D are initially diagnosed and treated as T2D and 1 in 6 of those who are diagnosed and treated as T1D do not have this condition. In recent years multiple biomarker tests to assist diabetes classification and treatment have become routinely available. These include islet-autoantibodies (markers of the autoimmune process in T1D) and C-peptide (a measure of endogenous insulin secretion). However, as identified by a recent NICE evidence review, fundamental questions for the clinical use of these tests in those with recent onset diabetes have not been answered: we do not know which patients benefit from testing, or what tests should be measured. As a result practice varies hugely. Aim To establish the optimal cost-effective strategy for testing islet-autoantibodies and C-peptide in newly diagnosed adult-onset diabetes. Objectives To use existing cohorts to: Determine the diagnostic performance of commonly measured islet autoantibodies in adult-onset diabetes, alone and in combination, and determine whether novel assays improve performance. Determine whether early C-peptide measurement, within 1 year of diabetes diagnosis, improves diagnosis of diabetes subtype over and above islet autoantibodies. Compare performance of different clinical approaches to targeting biomarker testing at diagnosis using routinely available clinical features. Use decision analytic modelling to determine the optimal cost-effective approach to targeting classification biomarker testing at diabetes diagnosis. Methods The first part of this research (objectives 1-3) will use information from a recently completed prospective study that recruited over 1800 adults with new-onset diabetes, and assessed for loss of endogenous insulin secretion after >3 years diabetes duration, which confirms type 1 diabetes. We will determine which autoantibodies, or combinations of autoantibodies, are most helpful in identifying type 1 diabetes at diabetes diagnosis, and if early measurement of C-peptide is helpful. We will compare approaches to targeting testing based on routine clincial features, including assessing classification models developed by the investigators. In the second part of this research (objective 4) we will use information from objectives 1-3, together with other existing research studies and undertake Decision Analytic Modelling to determine the most cost-effective approach to classification biomarker testing in new-onset diabetes. Anticipated Impact The impact of the work will be to improve identification (and therefore treatment) of subtypes of diabetes at diagnosis and inform cost-effective use of classification biomarker testing. It will provide evidence for rational use and interpretation of classification tests in patient management, an area where there is considerable uncertainty and widely varying practice. This is likely to improve quality of life and outcomes for people living with diabetes, including reducing short- and long-term complications, while reducing healthcare costs.