Epileptic Disorders


Authors’ response Volume 17, numéro 1, March 2015

To the Editor,

We thank Dr. Simabukuro for his interest in our case report (Kojima et al., 2014) and appreciate his insightful comment. We agree that the classification of autoimmune encephalopathy associated with antibodies to the cell surface or synaptic proteins is evolving as new antibodies are discovered.

As Dr. Simabukuro points out, anti-GABA-A receptor antibody was not discovered at the time of our patient's presentation and, therefore, was not measured. We agree that there are similarities between our case and recently reported cases of encephalitis with GABA-A receptor antibodies (Petit-Pedrol et al., 2014; Ohkawa et al., 2014). Notably, the MRI findings of extensive multifocal abnormalities on FLAIR and T2 imaging, with cortical and subcortical involvement without contrast enhancement (Petit-Pedrol et al., 2014; figure 4 [Index Patient 1]), are strikingly similar to those of our patient (Kojima et al., 2014). Petit-Pedro et al. divides the patients into two groups: (1) patients with high serum and CSF titers of anti-GABA-A receptor antibodies (six patients) and (2) patients with a low serum titer of anti-GABA-A receptor antibodies (12 patients). The clinical outcome and subsequent CSF, EEG, and MRI results of the latter group are relatively variable. Many patients in both of these groups had co-existing antibodies, including thyroid auto-antibodies and anti-GAD, anti-GABA-B receptor, and anti-NMDAR antibodies, among which anti-GAD antibodies were most commonly detected in 6 of 18 patients (Petit-Pedrol et al., 2014). Another shared clinical presentation is neurological symptoms which precede seizure episodes; our patient had auditory hallucination for three days before the onset of seizures (Kojima et al., 2014) and all of six patients with high titers had epileptic symptoms preceded by, or associated with, a change in behaviour or level of cognition (Petit-Pedrol et al., 2014). In addition, hemiparesis prior to seizure was observed both in our patient (before the second seizure attack) (Kojima et al., 2014) and in one of the six patients (Petit-Pedrol et al., 2014).

Whether the low-titer group can be attributed to anti-GABA-A receptor antibody-related encephalitis, or whether anti-GABA-A receptor antibody is a marker of autoimmunity in these cases needs further study (Suleiman and Dale, 2014). Of the two patients reported by Ohkawa et al., one patient did not have seizures (Ohkawa et al., 2014). Both of them had invasive thymomas, but our patient's thymoma (benign) was resected years ago, and the patient was in remission at the time of presentation. While these two patients had co-existing anti-LGI1 or anti-CASPR2 antibodies, anti-GAD antibody and thyroid auto-antibody were not mentioned. It would be interesting to determine what the PET scan and pathology would have shown for these patients with anti-GABA-A receptor antibody-associated encephalitis. If there is a discrepancy, as in our case (PET positive but very little inflammation on pathology), this may support the notion that synaptic, but not intracellular, antibodies are causative and that they induce transient neuronal dysfunction, but not overt inflammation and cell death. However, because of the novelty and a very small number of anti-GABA-A receptor antibody-associated encephalitis cases, it is still too soon to attribute the syndrome to anti-GABA-A receptor antibodies. More studies and immunological understanding of the pathogenesis of autoantibody-induced encephalitis will be needed to finalize how we would classify these various syndromes associated with neuronal autoantibodies. It is possible that some of the previous cases of anti-GAD antibody-associated encephalitis with extralimbic involvements may have been associated with anti-GABA-A receptor antibody.

In summary, as in any case report, we acknowledge that we cannot provide final answers on every aspect of this patient, but we still feel our case provides valuable clinical information on pathology, PET imaging, and MRI. We do acknowledge that anti-GAD antibody may not have been pathogenic in our case, and may have simply been a marker of autoimmunity. However, this information was still valuable, because it led to the successful treatment of our patient. As this field of neuroimmunology and neuronal autoantibodies evolves, the nomenclature and classification will be clearer. We do agree that our case may indeed harbour anti-GABA-A receptor antibody and we would like to pursue this possibility in our patient. We thank Dr. Simabukuro for initiating a very fruitful correspondence.