AUTHOR=Zong Shenghua , Vinke Anita M. , Du Peng , Hoffmann Carolin , Mané-Damas Marina , Molenaar Peter C. , Damoiseaux Jan G. M. C. , Losen Mario , Rouhl Rob P. W. , Martinez-Martinez Pilar 

TITLE=Anti-GAD65 autoantibody levels measured by ELISA and alternative types of immunoassays in relation to neuropsychiatric diseases versus diabetes mellitus type 1

JOURNAL=Frontiers in Neurology

VOLUME=Volume 14 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2023.1111063

DOI=10.3389/fneur.2023.1111063

ISSN=1664-2295

ABSTRACT=Abstract (Maximum 350 words)
Background: Anti-GAD65 autoantibodies (GAD65-Abs) may occur in patients with epilepsy and other neurological disorders, but the clinical significance is not clear-cut. Whereas high levels of GAD65-Abs are considered pathogenic in neuropsychiatric disorders, low or moderate levels are only considered as mere bystanders in e.g. diabetes mellitus type 1 (DM1). The value of cell-based assays (CBA) and immunohistochemistry (IHC) for GAD65-Abs detection has not been clearly evaluated in this context. 
Objective: To re-evaluate the assumption that high levels of GAD65-Abs are related to neuropsychiatric disorders and lower levels only to DM1 and to compare ELISA results with CBA and IHC to determine the additional value of these tests.
Methods: 111 sera previously assessed for GAD65-Abs by ELISA in routine clinical practice were studied. Clinical indications for testing were e.g. suspected autoimmune encephalitis or epilepsy (neuropsychiatric cohort; n=71, 7 cases were initially tested positive for GAD65-Abs by ELISA), and DM1 or latent autoimmune diabetes in adults (DM1/LADA cohort (n=40, all were initially tested positive). Sera were re-tested for GAD65-Abs by ELISA, CBA, and IHC. Also, we examined the possible presence of GAD67-Abs by CBA and of other neuronal autoantibodies by IHC. Samples that showed IHC patterns different from GAD65 were further tested by selected CBAs.
Results: ELISA retested GAD65-Abs level in patients with neuropsychiatric diseases was higher than in patients with DM1/LADA (only retested positive samples were compared; 6 vs 38; median 47092 U/mL vs 581 U/mL; p=0.02). GAD-Abs showed positive both by CBA and IHC only if antibody levels were above 10000 U/mL, without a difference in prevalence between the studied cohorts. 
We found other neuronal antibodies in one patient with epilepsy (mGluR1-Abs, GAD-Abs negative), and in a patient with encephalitis, and two patients with LADA. 
Conclusion: GAD65-Abs levels are significantly higher in patients with neuropsychiatric disease than in patients with DM1/LADA, however, positivity in CBA and IHC only correlates with high levels of GAD65-Abs, and not with the underlying diseases.