Limbic Encephalitis: An Update

Abstract & Commentary

Synopsis: The diagnosis of limbic encephalitis should be considered in patients who present with progressive memory deficits, confusion, and seizures, as it is a treatable entity in many cases and may suggest the presence of an underlying malignancy.

Source: Asztely F, Kumlien E. The diagnosis and treatment of limbic encephalitis. Acta Neurol Scand DOI: 10.1111/j/1600-0404-2012-01691.x.

Limbic encephalitis initially was described more than 40 years ago as a subacute progressive condition manifested by confusion, memory loss, and seizures with postmortem inflammatory pathological changes in the temporal lobes. An association with cancer was hypothesized at that time, but it was not until the late 1980s and early 1990s that the concept of paraneoplastic neurological disorders was more formally recognized and understood. Over the ensuing decade, multiple cancer-associated autoantibodies were associated with the development of limbic encephalitis. These included antibodies associated with a wide variety of primary tumors including small cell lung cancer (Hu, CV2, Ri), breast cancer (ampiphysin, Ri, Yo), thymoma (CV2), ovarian cancer (Yo), and testicular cancer (Ma2). The recognition of the relationship between onconeuronal antibodies and limbic encephalitis sometimes allowed for the diagnosis of a systemic malignancy at the time of presentation with the neurological syndrome.

Over the past 10-15 years, the medical community has recognized that many cases of classic limbic encephalitis were not associated with any of the traditional onconeuronal antibodies, nor was an underlying tumor found despite extensive investigations. These cases of "non-paraneoplastic limbic encephalitis" were poorly understood until more recent descriptions of novel auto-antibodies causing the disorder. These antibodies include those directed against the NMDA receptor, AMPA receptor, GABA-B receptor, glycine receptor, voltage gated potassium channels, and glutamic-acid decarboxylase. There is often no underlying neoplasm associated with these antibodies, although NMDA receptor-associated limbic encephalitis sometimes is associated with underlying ovarian teratoma. There are still cases of limbic encephalitis with no identified auto-antibody, suggesting that there are antibodies yet to be discovered. This is a rapidly evolving area in neurology.

The diagnosis of limbic encephalitis should be considered in a patient who has days to weeks of confusion, short term-memory loss, disorientation, agitation, and occasionally seizures. Some patients present more slowly with psychiatric manifestations. The MRI may be normal or may demonstrate medial temporal T2 hyperintensity. The CSF usually demonstrates a modest lymphocytic pleocytosis, elevated protein, and the presence of oligoclonal banding. It is important to recognize that the differential diagnosis of limbic encephalitis is quite broad and other diagnoses should be excluded. These include acute viral encephalitis (especially the encephalitis caused by HSV-1), acute disseminated encephalomyelitis, Hashimoto encephalopathy, central nervous system vasculitis, Creutzfeldt-Jakob disease, metabolic delirium, neoplasms, and alcohol-related neurological disorders like Wernicke encephalopathy. These diagnoses can be differentiated from limbic encephalitis based on serum, CSF, and imaging findings.

Once the diagnosis of limbic encephalitis is being considered, and alternative diagnoses are being excluded, the workup of the patient should include testing serum and CSF for the presence of the classic onconeural antibodies (Hu, CV2, ampiphysin, Ri, Yo, Ma2) as well as the newer antibodies, especially NMDA and AMPA receptor antibodies. An extensive search for an underlying malignancy should be undertaken, guided by the type of antibody discovered on the workup. Most authorities recommend screening with whole body FDG-PET/CT. However, ovarian teratoma may be missed on this modality and MRI of the pelvis may be necessary. Additionally, serum tumor markers for testicular cancer should be checked in men (AFP, beta-HCG). If an underlying malignancy is not found, recommendations are to screen again in 3-6 months if there is a positive paraneoplastic antibody in the serum or CSF as the tumor may be too small to detect on initial presentation.

Treatment of limbic encephalitis depends on the cause. Certainly, the underlying tumor should be treated if identified. Treatment of the neurologic syndrome is also important in most cases. There are no well-established therapies for limbic encephalitis. However, since the condition is autoimmune by nature, most patients are treated initially with high-dose intravenous methylprednisolone, sometimes followed by intravenous immunoglobulin (IVIg) or plasma exchange if the steroids are not successful. Steroids may also be co-administered with IVIg as well. Second-line therapies include rituximab and/or cyclophosphamide. In general, earlier treatment is better than delayed treatment because the immune activation may cause permanent neuronal injury.

The literature on limbic encephalitis is rapidly evolving. Every year, new antibodies are discovered and more is added to the literature regarding therapy. It is important to consider this diagnosis early in the presentation of the illness and to thoroughly assess the patient for an underlying malignancy. Neurologists are often called to see patients who are confused for no obvious reason, and limbic encephalitis should always be on the differential diagnosis in those cases.