Measurement of autoantibodies to insulin informs diagnosis of diabetes in a childhood population negative for other autoantibodies

Aims: Some childhood cases of type 1 diabetes are islet autoantibody negative at diagnosis. Potential explanations include misdiagnosis of genetic forms of diabetes or an insufficient screening panel of islet autoantibodies. Most NHS laboratories test for three markers (GADA/IA-2A/ZnT8A). Autoantibodies to endogenous insulin (IAA) are used less routinely because high-performance assays are not widely available, and IAA must be measured within two weeks of insulin treatment. Testing of islet cell antibodies (ICA) is sometimes performed but is associated with reduced specificity. We sought to determine the added value of testing for IAA using robust and highly sensitive radiobinding assays (RBAs) close to diagnosis.
Methods: RBAs were used to test for GADA/IA-2A/ZnT8A/IAA in individuals with newly-diagnosed type 1 diabetes from the BOX study [n = 419; median age 10.3 years (range 0.7-18.0); median duration 0 days (range -61.0-14.0)]. Where available, ICA data were also considered.
Results: Using GADA/IA-2A/ZnT8A RBAs, 19 of 419 (4.5%) children were autoantibody negative. Testing for IAA classified a further 8/19 (42.1%) children. Of the remaining 11 children negative for GADA/IA-2A/ZnT8A/IAA, ICA data were available for seven and were confirmed negative; 10 were positive for at least one high-risk HLA haplotype and one was positive for the protective variant DQB1*0602.
Conclusion: Testing for GADA/IA-2A/ZnT8A using highly sensitive tests fails in detecting autoantibodies in 1/20 children diagnosed with type 1 diabetes. Testing for IAA proved an autoimmune basis in a further 42.1%. Experiments are ongoing to establish the added value of testing for IAA and Tspan7A in both children and adults at diagnosis.