Evaluation of glutamic acid decarboxylase (GAD) 65 antibody detection methods for neurological and diabetic investigation in an Australian diagnostic laboratory

The role of anti-glutamic acid decarboxylase (GAD) 65 autoantibodies in autoimmune neurological conditions is evolving, but testing recommendations remain unchanged in Australia with GAD enzyme-linked immunosorbent assay (ELISA) and immunoblot as the only two Therapeutic Goods Administration approved testing methods available in Australia. Common practice is for use of ELISA in diagnosis of type 1 diabetes mellitus (T1DM) and use of immunoblot for diagnosis of GAD65-associated neurological disease. We observed a cohort of patients with negative immunoblot results and positive ELISA in the context of GAD-associated neurological disease without T1DM. In the absence of robust consensus guidelines on preferred testing modalities, we sought to determine if ELISA could have a superior role in the diagnosis of GAD-associated neurological disease when compared to immunoblot in this paper. We tested for anti-GAD65 autoantibodies on 55 patient samples, 40 samples requested for neurological disease and 15 type 1 diabetes samples with detectable anti-GAD65, using two testing platforms: Euroimmun anti-GAD enzyme-linked immunosorbent assay (ELISA) and. Euroimmun EuroLine immunoblot for paraneoplastic neurologic syndromes. These results were correlated against the clinical scenario. Positive ELISA results had a sensitivity of 100% and specificity of 91% for GAD65-related neurological disease. Immunoblot showed sensitivity of 43% and specificity of 76% for GAD65-related neurological disease. ELISA proved more sensitive and specific for GAD65-related neurological disease compared to immunoblot, raising questions about the role of this testing modality in neurological disease. We propose that ELISA should be used as a sole diagnostic method for all GAD65 antibody-related neurological disease over immunoblot. The presence of anti-GAD65 antibody on immunoblot is of doubtful clinical significance.