Autoantibodies to N-terminally Truncated GAD(65)(96-585): HLA Associations and Predictive Value for Type 1 Diabetes

OBJECTIVE: To evaluate the role of autoantibodies to N-terminally truncated glutamic acid decarboxylase GAD(65)(96-585) (t-GADA) as a marker for type 1 diabetes (T1D) and to assess the potential human leukocyte antigen (HLA) associations with such autoantibodies. DESIGN: In this cross-sectional study combining data from the Finnish Pediatric Diabetes Register, the Type 1 Diabetes Prediction and Prevention study, the DIABIMMUNE study, and the Early Dietary Intervention and Later Signs of Beta-Cell Autoimmunity study, venous blood samples from 760 individuals (53.7% males) were analyzed for t-GADA, autoantibodies to full-length GAD(65) (f-GADA), and islet cell antibodies. Epitope-specific GAD autoantibodies were analyzed from 189 study participants. RESULTS: T1D had been diagnosed in 174 (23%) participants. Altogether 631 (83%) individuals tested positive for f-GADA and 451 (59%) for t-GADA at a median age of 9.0 (range 0.2-61.5) years. t-GADA demonstrated higher specificity (46%) and positive predictive value (30%) for T1D than positivity for f-GADA alone (15% and 21%, respectively). Among participants positive for f-GADA, those who tested positive for t-GADA carried more frequently HLA genotypes conferring increased risk for T1D than those who tested negative for t-GADA (77% vs 53%; P < 0.001). CONCLUSIONS: Autoantibodies to N-terminally truncated GAD improve the screening for T1D compared to f-GADA and may facilitate the selection of participants for clinical trials. HLA class II-mediated antigen presentation of GAD(96–585)-derived or structurally similar peptides might comprise an important pathomechanism in T1D.