Contributed by Alejandro Perez, MD, Gustavo C. Roman, MD, Suzanne Z. Powell, MD, Ron Fisher, MD, Andreana L. Rivera, MD, Joseph C. Masdeu, MD, and Matthew D. Cykowski, MD
FINAL DIAGNOSIS
Limbic Encephalitis due to Lgi1 Antibodies
DISCUSSION
Antibodies to neuronal antigens are recognized as a source of heterogeneous neurological syndromes that may respond to immunotherapy. Currently, the best characterized among these is N-methyl-D-aspartate receptor (NMDAR) encephalitis, which gives rise to psychosis, memory deficits, and dyskinesia, often occurring in young women with teratoma (2). In addition, several neurological syndromes are associated with antibodies directed against voltage-gated potassium channel (VGKC)-associated proteins (4). Encephalitis due to leucine-rich glioma inactivated 1 (Lgi1) antibodies, a subtype of VGKC-complex encephalitis, results in characteristic faciobrachial dystonic seizures (FBDS) (1). FBDS typically precede cognitive decline and are of short duration and characterized by ipsilateral face grimacing and upper extremity spasm and posturing. As the disease progresses, it is attended by episodes of confusion, behavioral changes, impaired memory, and temporal lobe seizures. Lgi1 antibody limbic encephalitis tends to respond well to immunotherapy and early treatment with immunotherapy after recognition of FBDS can reverse the symptoms (3). For these reasons, very few Lgi1 limbic encephalitis patients come to autopsy and little is known about the underlying neuropathology.
In this patient, VGKC/Lgi1 antibodies were identified in a premortem encephalopathy panel of serum autoantibodies. Unfortunately, despite premortem identification of VGKC/ Lgi1 autoantibodies, the patient's disease proved refractory to therapy. The neuropathologic findings at autopsy highlighted that Lgi1 limbic encephalitis is associated with a fairly circumscribed set of cellular pathologies centered on the amygdala and adjacent limbic areas, such as subiculum. How this pathology contributes to the distinctive faciobrachial dystonic seizures that characterize Lgi1 limbic encephalitis remains to be elucidated. Involvement of the amygdala and subiculum, as in this case, explains the recent memory loss that attends this diagnosis (1).
Provided the circumscribed set of reactive and inflammatory pathologies in the limbic region and amygdala, the main entities in the differential diagnosis are more common forms of limbic encephalitis and viral encephalitis in particular. In our case, NMDAR encephalitis was excluded by the autoimmune encephalopathy panel (NMDA autoantibodies were not identified). However, to our knowledge, there are too few cases of NMDAR and Lgi1 limbic encephalitis coming to autopsy to confidently exclude them on neuropathologic grounds alone. Viral encephalitis was also considered, although unlike typical viral encephalitis, our case lacked microglial nodules or neuronophagia and no viral inclusions were identified. In the absence of a premortem autoimmune encephalopathy panel, both immunohistochemistry and next-generation sequencing-based examinations of tissue may help to exclude a viral organism in appropriate cases. Conversely, when viral encephalitis is considered based on neuropathologic examination, especially if the pathologies are limited to the limbic region, then autoimmune encephalitis should also be considered in the appropriate clinical setting. In cases of possible limbic encephalitis without a premortem diagnosis (i.e., the question is raised by neuropathologic findings), sera and/or cerebrospinal fluid collected at autopsy can be submitted for an autoimmune encephalopathy panel to clarify the final diagnosis.
REFERENCES
Contributed by Alejandro Perez, MD, Gustavo C. Roman, MD, Suzanne Z. Powell, MD, Ron Fisher, MD, Andreana L. Rivera, MD, Joseph C. Masdeu, MD, and Matthew D. Cykowski, MD
International Society of Neuropathology