Contributed by Julia C. Kuehn MD, Angelika Scheuerle MD, Jan Bauer MD, Albert J. Becker MD, Rainer Wirtz MD, Jan Lewerenz MDFINAL DIAGNOSIS
Anti-Leucine Rich Glioma Inactivated 1 (LGI1) - Autoimmune-mediated Limbic Encephalitis
DISCUSSION
Lesions of the hippocampus are frequently associated with symptomatic epileptic seizures. However, a diffuse swelling hippocampus with contrast-enhancement upon MRI may indicate a higher grade glioma or inflammatory changes in the context of an limbic encephalitis(4). Histopathological examination of the resected hippocampus in our case showed increased cellularity, in part due to prominent astroglial proliferation. However, molecular pathology and marker expression patterns made the diagnosis of a glioma, suggested by the imaging findings, highly unlikely. In contrast, neuronal loss in association with complement deposition and CD8+cytotoxic T lymphocytes was found. The lack of lymphoblasts and presence of different populations of immunocompetent cells rules out primary CNS lymphoma. Screening for antineuronal antibodies revealed anti-LGI1 antibodies in serum (1:100) and CSF (1:10). Anti-LGI1 encephalitis is a disease predominantly affecting elderly men(1). Epileptic seizures, cognitive decline, mesiotemporal T2 hyperintensities and hyponatremia are common findings. CSF findings are mostly normal(1). Multiple mechanisms may play a role in the pathogenesis of anti-LGI1 encephalitis, a disease formerly subsumed in the group of autoimmune encephalitides with voltage-gated potassium channel-complex (VGKC) antibodies, as LGI1 interacts with VGKC and thus anti-LGI1 antibodies can be detected indirectly by the VGKC radioimmunoprecipitation assay (1). Firstly, the autoantibodies seem to disrupt the LGI1-ADAM22 interaction followed by AMPA receptor downregulation(10). In addition, activation of the complement system and CD8+cytotoxic T lymphocytes probably contributing to neuronal death found in our case have been previously observed in a single patient with anti-LGI1 encephalitis and one additional patient with VGKC antibodies of unknown specificity, presumably LGI1 (2). Both mechanisms are absent in anti-NMDA receptor encephalitis, the most common form of autoimmune-mediated encephalitis(2, 8).
In our case, we observed prominent astrogliosis in the hippocampus, at the first sight reminiscent of a glioma, the differential diagnosis initially made based on the imaging findings. In addition, strong microglial activation was observed. These findings are in line with the astrocytic and microglial activation reported for the autoptic findings in a male patient with limbic encephalitis, a very high VGKC titer and hyponatremia, most likely representing a case with anti-LGI1 associated limbic encephalitis(6).
It seems plausible that the neuronal loss observed in LGI1 antibody associated limbic encephalitis is mediated by the observed complement activation and T cell invasion. However, as the disease is also associated by sometimes frequent epileptic seizures(1), excitotoxicity secondary to an initial dysbalance of excitation and inhibition might contribute to neurodegeneration that then might be subfield specific in line with the neuroanatomical architecture and known differential susceptibility of the hippocampal subfield in hippocampal sclerosis associated with temporal lobe epilepsy(11) or excitotoxicity (7). In our case, neuronal loss was most pronounced in CA4. However, using ultra high-field MRI region-specific atrophy was identified as most pronounced in CA3 following the acute phase of the disease (9). This is associated with the severe amnesia for autobiographical events that is typical for LGI1 antibody-associated limbic encephalitis. The rare histopathological reports indicate that in anti-LGI1 encephalitis the neuronal loss on the cellular level might show a contrasting distribution, indicating that the CA3 subfield atrophy reported in MRI might not directly reflect neuronal loss. Of note, LGI1 shows strong expression in the mossy fibers in the molecular layer of the CA3 subfield (5). The one case with high-titer VGKC antibodies and hyponatremia was reported to show a relative sparing of CA1, 2 and 3, while neuronal loss in CA4 was, as in our case, most pronounced(6). In the other reported autopsy case with presumably LGI1 antibody-associated VGKC antibody-positive encephalitis, also TUNEL-positivity of numerous neurons in CA4 was reported. However, in this case additional neuronal death in the dentate gyrus was described(2). Of note, the emerging pattern of hippocampal neuronal loss seems to be entirely different from the pattern found in hippocampal sclerosis.(11)
Our case demonstrates that anti-LGI1 encephalitis can mimic high-grade glioma in imaging studies. Therefore, the awareness of this differential diagnosis and autoantibody testing are crucial. Due to good response to immunotherapy chronic stages with atrophy and hippocampal sclerosis are avertable.
Acknowledgement
We thank the patient and his family for the permission to present the case history to the medical public. J.L. thanks Martina Leis for the excellent technical assistence. J.C.K. is supported by the Else Kröner-Fresenius Stiftung with a grant in aid (EKFS-Promotionskolleg 'Neuroimmu-nology').
REFERENCES
Contributed by Julia C. Kuehn MD, Angelika Scheuerle MD, Jan Bauer MD, Albert J. Becker MD, Rainer Wirtz MD, Jan Lewerenz MD
  |
