Exploring the clinical spectrum of glutamic acid decarboxylase antibodies in neurological patients, through the validation of an enzyme-linked immunosorbent assay

Glutamic Acid Decarboxylase (GAD) antibodies are markers of type 1 diabetes mellitus (T1DM) and autoimmune neurological syndromes. The main neurological syndromes linked to GAD abs are stiff-person syndrome (SPS), cerebellar ataxia, limbic encephalitis and epilepsy. Titers of GAD abs in neurological patients are much higher than in diabetes. The meaning of low titres in neurological syndromes is still uncertain and the cut off for the relevance of GAD abs is debated. Moreover, commercial kits for detection of GAD abs are usually validated only in patients with T1DM. Our work aims: - to perform a clinical validation of a commercial ELISA assay for GAD abs detection in serum and cerebrospinal fluid (CSF) of patients with suspected neurological autoimmune diseases; - to set up a clinically relevant cut-off for serum GAD abs; - to explore the meaning of serum low-titre GAD abs in neurological patients; - to investigate the presence of coexisting antineuronal antibodies in GAD abs seropositive patients; We tested 330 patients with suspected autoimmune neurological syndromes for GAD abs by a modified commercial ELISA kit. Sera and CSF were titrated by multiple dilutions and the final values extrapolated by a mathematical model. When paired serum and CSF were available, intrathecal synthesis was calculated. Specimens were also tested for other antineuronal abs (onconeural and surface antigens) by a cell-based assay and immunoblot; GAD abs were searched also by immunoistochemistry on frozen rat brain and immunoblot. Clinical features were reviewed by one observer blinded for GAD titers; he noted if a typical GAD-related autoimmune neurological syndrome was present; then a ROC curve was generated to find the best cut off that distinguishes clinically relevant from clinically not relevant titres. Forty-three patients tested positive by optimized ELISA (serum GAD abs, range 6 – 2019681 UI/ml), 35 with available clinical information were included (14 males and 21 females, age range 15-82 years old). In 26/35 patients (74%) the final diagnosis belonged to the already reported spectrum of GAD-abs related diseases, specifically 10 patients diagnosed with encephalitis, 9 with SPS, 5 with ataxia and 2 with epilepsy. Among these, 12/26 failed to fit the criteria described in the methods and were noted as “Atypical”. 14 patients showed typical features. ROC analysis identified a cut off at 13000 IU/ml. Intrathecal synthesis of GAD abs were detected in 9 patients with serum high-titre GAD abs. One patient with a high GAD-abs titre showed co-reactivity with anti-Amphyphisin abs; 4/36 (11%) with low-titre GAD abs had a relevant neuronal surface antibodies positivity for anti-LGI1 (3) and anti-GlyR antibodies (1). We confirmed clinical spectrum of GAD autoimmunity. The study validates a modified ELISA in routine detection of GAD abs in neurological patients, setting the threshold of 13000 IU/ml as a discriminant serum cut off to identify clinically relevant GAD abs.

Gli anticorpi diretti contro l’enzima glutammato decarbossilasi (GAD) furono descritti per la prima volta nel 1988 in un paziente affetto da stiff-person syndrome (SPS) – malattia caratterizzata da rigidità e spasmi dolorosi assiali e degli arti - epilessia e diabete mellito. Alcuni studi successivi identificarono gli anticorpi anti-GAD come marcatori specifici del diabete autoimmune. In ambito neurologico, dopo la SPS, gli anti-GAD sono stati descritti in quadri di atassia cerebellare progressiva, in una forma di epilessia temporale e, più raramente, in pazienti con encefalite limbica e disfunzioni del tronco encefalico. Attualmente la determinazione degli anticorpi anti-GAD nei pazienti affetti da malattie neurologiche è effettuata tramite immunoistochimica indiretta (IHC) su cervello di ratto, test radioimmunologici (RIA), immunoenzimatici (ELISA) ed immunoblot. Viste le difficoltà tecniche dell’IHC e il perfezionamento dei test ELISA, i kit commerciali RIA ed ELISA sono i più diffusamente utilizzati, sebbene essi siano progettati allo scopo di individuare i bassi titoli di anti-GAD che sono tipici del diabete autoimmune. Nella maggior parte dei casi, tali kit non sono stati validati clinicamente per l’utilizzo su siero e liquor nei pazienti affetti da malattie neurologiche, che presentano anticorpi a titoli molto più elevati rispetto ai pazienti diabetici. Scopo del lavoro di tesi è descrivere retrospettivamente una serie di pazienti con malattie neurologiche associate al riscontro di anticorpi anti-GAD individuati con un test ELISA modificato, validando tale metodica sul siero e liquor dei pazienti affetti.