GAD65 Antibody-Associated Epilepsy

Tags

Top Highlights

• Antibodies involved in the development of seizures can be classified into two groups: antibodies against neuronal surface antigens that are implicated in the development of acute symptomatic seizures secondary to autoimmune encephalitis and antibodies against intracellular antigens (anti-GAD and onconeural antibodies) that are found in cases of AAE [5].

• Antibodies directed against neuronal surface antigens often target synaptic proteins and lead to the downregulation of receptors that alters synaptic transmission; therefore, they are directly pathogenic, whereas antibodies against intracellular antigens are not believed to be directly involved in the disruption of synaptic transmission and could be more of a reflection of an autoimmune process where not humoral but cellular immunity is the main factor leading to pathological changes [7].

• Both antibodies against neuronal surface antigens and antibodies against intracellular antigens are associated with neoplastic processes with the association being stronger in the latter group. Tumor type and presence strongly depend on the type of autoantibodies: for example, in anti-Leucine-rich glioma-inactivated 1 (LGI1) encephalitis, there is only a 5–10% chance of cancer (usually thymoma), whereas in anti-Hu encephalomyelitis, only 15% do not have cancer, with the rest usually diagnosed with small cell lung carcinoma, and in anti-NMDA receptor encephalitis, approximately 50% of females older than 12 years have an ovarian teratoma [8,13,14].

• Finally, if autoimmune encephalitis or AAE is suspected, it is essential to send samples for antibody panels. Both CSF and serum can be tested and the sensitivity and specificity of testing in both samples depend on the type of antibody; for example, in anti-NMDAR encephalitis, CSF testing is more sensitive than serum testing [15]. However, 6–16% of patients with anti-LGI1 encephalitis and 13–14% of those with Contactin-associated protein-like 2 (CASPR2) antibodies are detected only in the serum, whereas both CSF and serum testing for anti-Gamma-Aminobutyric Acid type B (GABAB) receptors have equal sensitivity. In rare cases of autoimmune encephalitis, antibodies may be undetected both in serum and CSF, but in the context of a typical clinical presentation, neuroimaging findings, and CSF results, the diagnosis should be made nevertheless [21].

• To aid the diagnostic process, a few diagnostic scales have been proposed, including the Antibody Prevalence in Epilepsy (APE) and the more recent Antibody Prevalence in Epilepsy and Encephalopathy (APE2) score [24,25]. A study that included patients with AAE or acute symptomatic seizures secondary to autoimmune encephalitis showed that an APE score ≥4 showed 77.9% specificity and 97.7% sensitivity to detect neural autoantibodies [24], and with its updated version, an APE2 score ≥4 was 99% sensitive and 93% specific for neural-specific antibodies [25]. In cases of recurrent seizures of unknown etiology, an APE2 <4 has been suggested to signify that an autoimmune etiology is unlikely, whereas in cases of APE2 ≥ 4, even in the absence of antibodies, an immunotherapy trial should be highly considered [26]. However, this scale was created when the term autoimmune epilepsy was still used and ILAE had not proposed the new definitions; therefore, it is problematic to apply in the current understanding of these disorders.

• 3.3. Treatment Contrary to acute symptomatic seizures secondary to autoimmune encephalitis, AAE usually has a poor response to immunotherapy and is also resistant to anti-epileptic medications as well [5]. In AE, first-line treatment typically is IVIG, intravenous corticosteroids, or plasmapheresis. These treatments can be given in combination or individually. Rituximab and cyclophosphamide are the most commonly used second-line options when there is no response to first-line treatment. Treatment usually involves two phases—the induction therapy phase followed by a maintenance phase of a few weeks or months [15]. There is no universally accepted treatment scheme, and each center typically has its own practice particularities; however, therapeutic algorithms have been proposed in review articles [29].

• There have been fewer than 200 cases of anti-GAD65 autoimmune-associated epilepsy reported in the literature; nonetheless, the incidence may be higher than was thought before due to low awareness of this disease and its diagnostic complexity [34].

• There are two existing isoforms of glutamic acid decarboxylase—GAD65 and GAD67. GAD65 is responsible for GABA synthesis, the major inhibitory neurotransmitter in CNS. Isoforms of GAD65 are found in two main places: in CNS inhibitory (GABAergic) neurons and pancreatic islet β cells [35]. GAD has three functional domains: an amino (N)-terminal domain, a middle pyridoxal-5′-phosphate binding domain containing the active catalytic site of the enzyme, and a carboxy (C-terminal) domain. The antibodies against glutamic acid decarboxylase in different diseases have different binding targets. That is why antibodies against GAD65 can be associated with a wide spectrum of neurological syndromes, such as cerebellar ataxia, stiff-person syndrome, limbic encephalitis, and temporal lobe epilepsy. In patients with epilepsy, antibodies are more likely to react against the C-terminal, and in cases of limbic encephalitis—against the N-terminal of GAD [34]. In addition, anti-GAD65 can be detected in ~87% of diabetes mellitus type 1 patients but only 0.8% have anti-GAD > 2000 U/mL, so in cases of a suspected anti-GAD-associated neurological syndrome in a patient with diabetes, the diagnosis of an anti-GAD neurological syndrome must be made with caution, especially if the antibody levels are on the lower end [34,35,36].

• Association with neoplastic processes is unusual in GAD autoimmunity and cancer screening is not routinely recommended unless there are additional factors: the clinical presentation is different from the typical GAD-associated epilepsy, the patient presents with a classic preneoplastic syndrome, or there is the presence of coexisting neuronal cell-surface antibodies [37]. Typically, anti-GAD65 AAE usually occurs in the second to third decade of life and is more common among females (70–80% of cases) and in those having other autoimmune diseases [34,35,36,38]. Seizures in the case of anti-GAD65 positivity can rarely present as part of an autoimmune encephalitis syndrome or, more frequently, as AAE; in both cases, seizures are usually of focal onset in temporal lobes [34]. Even though they are very rare, musicogenic reflex seizures may be more common in anti-GAD 65 AAE than in other epilepsies: one study showed that out of 1510 epilepsy patients (including 22 anti-GAD65), only three reported musicogenic reflex seizures and two of them had anti-GAD65 AAE, resulting in a prevalence of 9% [39,40].

• There are no anti-GAD-AAE-specific changes in EEG or brain MRI and they conform to those described in previous sections. In most anti-GAD AAE cases, the CSF white blood cell count is normal but sometimes can be mildly elevated, and oligoclonal bands may be present in some cases as well [34]. If available, anti-GAD65 antibodies should be tested both in the serum and in the CSF since CSF positivity helps to assess the significance of anti-GAD65 antibodies that were positive in the serum [15]. Nonetheless, there is no association between the seizure frequency, severity, or duration of epilepsy and anti-GAD65 antibody titers. It is noticed that after initiation of immunotherapy, persistently high antibody titers are associated with poor clinical response [34].