Final Diagnosis -- Anti-NMDA receptor encephalitis



This patient exhibits a severe form of encephalitis associated with the formation of autoantibodies targeted towards N-methyl-D-aspartate receptors (NMDAR) in the setting of an ovarian teratoma, also known as anti-NMDAR encephalitis. This disease was originally described in 2007 by Dalmau et al. when they reported on a cohort of twelve young women that demonstrated various severe neuropsychiatric symptoms, ovarian teratomata, and autoantibodies targeting the NMDA receptors. Due to the association of ovarian teratoma and the formation of these autoantibodies, this disease is frequently classified as a paraneoplastic syndrome, although there have been reports in the literature of this disease occurring in the absence of a tumor. Regardless, since its original description in 2007, more than 600 patients with this disease have been reported, and its discovery has impacted the diagnostic approach to many diverse neurologic and psychiatric complaints. In addition, the pathogenesis of this disease has led to the discovery of other autoimmune encephalitides associated with the formation of autoantibodies targeted towards neurotransmitter receptors, and in turn, the specificity of the antibodies associated with these diseases have provided insights into the function of the targeted receptors.

NMDA receptors are a specific type of glutamate neurotransmitter receptor that is localized in the membrane of post-synaptic neurons throughout the CNS, but predominantly in the hippocampus, forebrain, basal ganglia, spinal cord and cerebellum. These receptors function as ligand-gated ion channels and play a major role in synaptic transmission and plasticity, a mechanism which underlies learning, memory, and cognition. These receptors are composed of specific heteromeric combinations of two subunits, a glycine-binding NR1 subunit and a glutamate-binding NR2 subunit. In anti-NMDAR encephalitis, autoantibodies are produced that are characteristically targeted to an epitope in the extracellular domain of the NR1 subunits, and the presence of these autoantibodies results in a dose-dependent decrease of NMDARs within the CNS. It is hypothesized that the presence of a tumor containing neural components which express NMDA receptors, most often a teratoma, represents the inciting event by which immunologic tolerance may be broken in patients with paraneoplastic disease. However, the immunologic stimulus for the generation of these autoantibodies in patients without an identifiable tumor remains unclear.

The number of reported cases of anti-NMDAR encephalitis continues to increase exponentially, implying that this disorder is not rare and that there is an increased awareness of the disease. The majority of cases of anti-NMDAR encephalitis occur in young women with ovarian teratoma, with a female predominance of approximately 8.5:1.5 and a median age of 19 years. However, there have been reports of this disease occurring in older patients, men, and children; some with teratoma, some without. Some studies report that the detection of an underlying tumor is dependent on age, sex, and ethnic background. According to Dalmau et al. in 2011, the analysis of 400 patients with anti-NMDAR encephalitis determined that the younger the patient, the less likely a tumor will be detected, and in female patients older than 18 years, the frequency of an underlying ovarian teratoma approximates up to 60%. In addition, African American women are more likely to have an underlying ovarian teratoma than patients of other ethnic groups. Only about 5% of male patients older than 18 years will have an underlying tumor. The detection of tumors other than ovarian teratoma has been described in the literature, but this occurrence is rare, approximately 2% of patients. Some of these reported "other" tumors include breast carcinoma, ovarian neuroendocrine tumor, sex cord stromal tumor, pseudopapillary neoplasm of the pancreas, small cell lung cancer, neuroblastoma and Hodgkin lymphoma. According to some reports, the ovarian teratomata removed from patients with anti-NMDAR encephalitis universally contain neural tissue expressing NMDARs. The expression of NMDARs in other types of tumors has been examined in only one patient with breast cancer and it was positive. However, it remains unclear whether tumors other than teratomata are directly associated with the development of anti-NMDAR encephalitis or if they represent incidental findings.

Clinically, anti-NMDAR encephalitis demonstrates symptoms that characteristically occur in phases. The early phase of the disease includes a viral-like prodrome followed by a period of psychiatric symptoms. The prodrome usually consists of non-specific symptoms such as headache, fever, nausea, vomiting, and diarrhea. Within two weeks of these symptoms, severe psychiatric symptoms begin to develop. These symptoms can be quite variable but can include anxiety, agitation, disorientation, mania, delusional behavior, psychosis, and aphasia. Short-term memory loss is also a common feature. However, memory loss can be difficult to discern in these patients due to the overlying psychiatric symptoms and speech problems. Due to the nature of many of these symptoms, these patients will initially present in psychiatric treatment facilities, often with the presumed diagnosis of drug-induced psychosis. By the second to third week of symptoms, a period of decreased responsiveness begins to develop. During this phase, the patient can alternate between periods of agitation and catatonia. In addition, seizures, abnormal movements, and autonomic instability are usual manifestations. The seizures are typically complex and generalized seizures, but the severity can escalate quickly to status epilepticus. The abnormal movements are mostly semi-repetitive oro-lingual-facial dyskinesias but may also include choreoathetosis in the extremities, oculogyric crisis, rigidity, dystonic limb posturing, ataxia, and opisthotonus. The most frequent autonomic manifestations include hypo- or hypertension, tachy- or bradycardia, cardiac arrhythmias, hyperthermia, urinary incontinence, and hypoventilation. These symptoms may be so severe as to require strong sedation and mechanical ventilation, thus many patients require treatment in intensive care units for long periods of time. In addition, it is not uncommon for these patients to also demonstrate dissociative responses to stimuli, such as resistance to eye opening or little to no response to all stimuli. This bizarre constellation of phasic symptoms may seem baffling at first glance. However, considering the fact that the majority of NMDA receptors are present in areas of the brain responsible for memory, personality, movement, cognition, and autonomic control, it does not seem surprising that the increment autoantibody-mediated loss of these receptors and their synaptic signal could result in these manifestations. This concept is further corroborated by studies that demonstrate a similar profile of phasic symptoms in animals given progressively increased doses of NMDAR antagonists, such as MK801, ketamine, and phencyclidine, which are clinically referred to as dissociative anesthetics.

The diagnosis of anti-NMDAR encephalitis relies strongly on the clinical presentation and the detection of the disease-defining anti-NMDAR autoantibodies. However, the clinical presentation can be initially misleading and may require ruling out other more common diseases, such as an infectious process. An MRI of the brain is unremarkable in approximately 50% of patients while the other 50% of patients demonstrate non-specific changes that are usually mild and transient. Electroencephalogram (EEG) results are abnormal in up to 90% of patients, and usually demonstrate non-specific, slow, and disorganized activity with occasional electrographic seizures. This activity does not correspond with abnormal movements and does not respond to antiepileptic therapy. A lumbar puncture with investigation of the cerebrospinal fluid (CSF) will typically demonstrate a moderate lymphocytic pleocytosis, normal to mildly increased intrathecal protein, and, in about 60% of patients, CSF-specific oligoclonal bands. The final diagnosis is based on the presence of anti-NMDAR autoantibodies in the serum and CSF. The detection of these antibodies is performed using a semi-quantitative cell-binding assay which detects these IgG antibodies by indirect immunofluorescence. In this test, a small amount of the patient's CSF or serum is applied to a composite slide containing human embryonic kidney (HEK) cells that have been transfected with NR1/NR2 subunits, also known as HEK293 cells. Following incubation of the sample with these cells, a wash is performed and a fluoroscein-conjugated anti-human IgG antibody is added to detect any bound patient IgG autoantibody. The patient sample is classified as positive or negative based on the intensity of the detected immunofluorescence. If positive, a titer is reported. In 2011, Wandinger et al. attempted to evaluate the efficacy of this test in the diagnosis of anti-NMDAR encephalitis. They performed the assay on CSF and serum samples from 66 patients with clinically characterized anti-NMDAR encephalitis, samples from 31 patients with other autoimmune encephalopathies, samples from 100 patients with multiple sclerosis, samples from 50 patients with systemic lupus erythematosus, and samples from 200 healthy blood donors. As a result, they determined that the assay had up to 100% sensitivity and 100% specificity for detecting the anti-NMDAR antibody in patients with anti-NMDAR encephalitis. Following the detection of this autoantibody, some studies suggest that monitoring of the CSF and serum anti-NMDAR antibody titers during the course of the disease may be of interest since antibody concentrations have been shown to correlate with the patient's clinical state and response to treatment, i.e higher titers of antibody in the CSF or serum correlate with more severe disease and/or poor treatment response and vice versa. Yet, once the diagnosis of anti-NMDAR encephalitis is suspected or confirmed by the presence of anti-NMDAR antibodies, the first task in female patients is to screen for ovarian teratoma. This is best performed with an MRI, CT scan, and pelvic and/or transvaginal ultrasound. Ovarian tumor markers, such as CA125, ?-HCG, ?-fetoprotein, or testosterone, are usually negative. Finally, despite the presence of severe encephalitis, brain biopsies in these patients are not considered helpful in the diagnosis of anti-NMDAR encephalitis because they usually demonstrate normal or non-specific findings.

There is current clinical and scientific evidence that the anti-NMDAR autoantibodies are pivotal to the pathogenesis of anti-NMDAR encephalitis. As mentioned previously, it is thought that the generation of these antibodies is triggered by the presence of a tumor containing neural components that express NMDARs. Unfortunately, these autoantibodies cross-react with NMDARs normally expressed within the CNS, but the process by which these antibodies access and expand within the CNS remains unclear. Some speculate that the blood brain barrier is disrupted by a non-specific infection, such as the infectious prodrome seen in all of these patients, but none of this has been proven. Regardless, in the pathogenesis of anti-NMDAR encephalitis, the immune response eventually predominates in the CNS with intrathecal synthesis of anti-NMDAR antibody, which is clinically indicated by CSF lymphocytic pleocytosis and oligoclonal bands. The antibody produced within the CSF results in the selective decrease of surface NMDAR clusters on post-synaptic neurons in a reversible fashion that corresponds with antibody titers, as demonstrated by numerous in vitro and in vivo studies conducted by Dalmau et al. and Hughes et al. In their studies, patient anti-NMDAR antibodies that were incubated with rat hippocampal neuronal cultures resulted in a concentration-dependent decrease of cell surface NMDA receptors. In addition, this effect was proven to be reversible by comparing neurons incubated with patient antibodies and neurons incubated with normal control CSF. Neurons incubated with patient CSF for three days and seven days exhibited fewer clusters of NMDA receptors than neurons incubated with normal control CSF. On the contrary, neurons incubated with patient antibodies for three days followed by incubation with normal control CSF had similar numbers of NMDAR clusters at the end of the incubation period than neurons treated only with control CSF. These experiments also demonstrated the selectivity of the antibodies since they only affected the concentration of NMDAR clusters and not the concentration of the control post-synaptic protein, PSD-95. Once these antibodies were implicated in the selective, but reversible decrease of NMDARs, Hughes et al. sought to define the mechanism by which these antibodies actually decrease NMDAR cluster density. It was previously determined that these anti-NMDAR antibodies are of the IgG1 and IgG3 subclass, and therefore have the ability to activate complement. However, Hughes et al. determined that complement was unlikely to play a role in the destruction of the NMDARs by these antibodies. They performed immunohistochemical staining on paraffin-embedded sections of hippocampus from two patients with anti-NMDAR encephalitis and the hippocampus of three age-matched normal patients. In the patients with anti-NMDAR encephalitis, there were several areas with significantly decreased staining intensity for NR1 when compared with the normal controls, and within these same areas, deposits of IgG were detected, but not complement components. They also noted that when the Fc domain of the IgG anti-NMDAR antibodies was removed and the remaining Fab fragments were incubated with neurons for 1 day, the NMDAR cluster density was the same as neurons incubated with control IgG. However, when anti-Fab secondary antibodies (which link two Fab fragments in a conformation similar to normal antibody) were added, the neurons demonstrated significantly lower NMDAR cluster density compared with neurons incubated with control IgG. These results confirmed that these antibodies mediate the loss of surface NMDARs by binding, capping, and cross-linking NMDARs, resulting in internalization of the receptors. As mentioned previously, the degree of the decrease in NMDAR cluster density correlates with the patient's anti-NMDAR titer in the CSF. To prove this, Hughes et al. incubated rat hippocampal cultures with CSF samples from two patients with anti-NMDAR encephalitis at two different time points - the initial CSF was collected at the time of symptom onset and the second sample was collected at the time of symptom improvement in one patient and at symptom worsening in the other patient. Using the semi-quantitative anti-NMDAR antibody assay (described previously), they determined that the CSF antibody titer was higher at symptom presentation than in symptom improvement in one patient, but in the other patient the CSF antibody titer during symptom worsening was higher than the titer at symptom presentation. In addition, Dalmau et al. demonstrated that patients who exhibited clinical improvement during the course of their illness had a parallel decrease in CSF antibody titers. Those who did not improve clinically or died maintained elevated CSF titers throughout the duration of their illness.

The treatment of anti-NMDAR encephalitis can be frustrating because there is conflicting data in the literature and there is no established standard of care. However, since it is known that the presence of autoantibodies results in disease and that the autoantibody titers correlate with disease severity, the main aim of therapy is to reduce/eliminate anti-NMDAR antibody levels. The general themes for performing this task entail removal of the stimulus for autoantibody production (i.e teratoma excision, if present), removal of the offending autoantibodies from circulation through methods such as therapeutic plasma exchange (TPE) or intravenous immunoglobulin (IVIG), and suppression of antibody production with various immunotherapeutic agents. For patients that do not have an identifiable tumor, therapy is focused on removal of the autoantibodies from circulation and aggressive immunotherapy. In 2011, Dalmau et al. proposed a treatment algorithm for paraneoplastic disease that consisted of teratoma removal with corticosteroid and IVIG or TPE as first line treatment, followed by additional rituximab and cyclophosphosphamide therapy if the patient does not demonstrate an adequate clinical response after 10 days of first line treatment (Figure 1). These treatments are gradually discontinued when patients demonstrate substantial clinical recovery, which is also usually accompanied by a decrease in CSF and serum anti-NMDAR antibody concentrations. Following clinical improvement, these patients do not usually require continued antiepileptic therapy. However, they should be maintained on continued immunosuppression for at least one year to reduce the risk of relapse. Relapse occurs in approximately 20 to 25% of patients, and these patients usually have undetected tumors, recurrent tumors, or no identifiable tumors. Besides monitoring patients for symptoms of relapse, patients that have recovered from anti-NMDAR encephalitis should also undergo regular tumor surveillance.

The acute stage of anti-NMDAR encephalitis characterized by seizures, abnormal movements, autonomic dysfunction, and hypotension may require patients to be hospitalized in an ICU setting for at least three to four months. Approximately 75% of patients will recover or have only mild sequelae, and the remaining 25% usually remain severely disabled or succumb to death. Unfortunately, the causes of death reported in the literature are complications from the disease process, such as cardiac arrest, refractory status epilepticus, or tumor progression, or complications from prolonged ICU admissions, such as sepsis or acute respiratory distress. The estimated mortality of anti-NMDAR encephalitis has been reported to be approximately 4%, with the mean time from disease onset to death being approximately 3.5 months (range of 1 to 8 months). On the contrary, if a patient does recover, the process typically occurs in a multistage fashion that is the reverse of the order that the symptoms presented. In total, the gradual progress of clinical recovery may take weeks to months to years, and upon recovery, most patients exhibit amnesia for the duration of the illness, again reflecting how this disease leads to the disruption of normal NMDAR-mediated plasticity.

The prognosis of patients with anti-NMDAR encephalitis absolutely depends on early recognition, prompt immunotherapy, and complete tumor removal in patients that have paraneoplastic disease. The best outcomes are observed in patients with early and complete tumor removal combined with immunotherapy because these interventions remove the inciting "antigen" and provide the most substantial reduction in anti-NMDAR antibody levels. In general, patients with paraneoplastic disease tend to have better outcomes than patients with non-paraneoplastic disease. However, the response of the antibody titer in the CSF to treatment corresponds with the best prognosis for both disease cohorts. In fact, a decrease in CSF antibody titers is considered to be associated with a good response to immunomodulatory therapy and clinical remission. On the other hand, patients do not do as well clinically if treatment is unable to result in a rapid and sustained control of the immune response within the CNS. Therefore, it may seem practical to evaluate the effectiveness of therapy by regularly monitoring anti-NMDAR antibody titers in the CSF during treatment. However, this practice is not always performed since the assay is costly, usually performed at an outside reference laboratory, and is both uncomfortable and inconvenient for the patient that has to undergo frequent lumbar punctures. Unfortunately, CSF anti-NMDAR antibody titers were not trended in the patient described in this case report, but would have likely been of great interest. The reason for this patient's poor outcome is unclear, but may be related to a combination of delayed diagnosis, delayed tumor detection, or poor response to therapy.

In conclusion, anti-NMDAR encephalitis represents a fairly new category of immune-mediated disease that is frequently paraneoplastic in nature, treatable, and able to be diagnosed serologically. There is a need for future studies to clarify the mechanism as to how the disease is expanded in the CNS, the best treatment approach, strategies that accelerate and optimize the recovery process, and more precise molecular effects of these antibodies at the cellular level. However, as a clinician, anti-NMDAR encephalitis should be suspected in any individual, especially an adolescent, teenager, or young adult that develops a rapid change in behavior, psychosis, abnormal postures or movements, seizures, and variable signs of autonomic instability. Some authors even suggest considering the diagnosis whenever words such as "encephalitis of unknown origin," "drug-induced psychosis," or "new onset epilepsy" are used to describe admitted patients.


Dalmau J, Erdem T, Hai-yan W, et al. Paraneoplastic Anti-N-methyl-D-aspartate Receptor Encephalitis Associated with Ovarian Teratoma. Ann Neurol. 2007 Jan; 61(1):25-36.

Dalmau J, Lancaster E, Martinez-Hernandez E, et al. Clinical experience and laboratory investigations in patients with anti-NMDAR encephalitis. Lancet Neurol. 2011 Jan;10(1):63-74.

Wandinger K, Saschenbrecker S, Stoecker W, and Dalmau J. Anti-NMDA-receptor encephalitis: A severe, multistage, treatable disorder presenting with psychosis. J Neuroimmunol. 2011 Feb;231(1-2):86-91.

Dalmau J, Gleichman A, Hughes E, et al. Anti-NMDA-receptor encephalitis: case series and analysis of the effects of antibodies. Lancet Neurol. 2008 Dec;7(12):1091-8.

Hughes EG, Peng X, Gleichman AJ, et al. Cellular and Synaptic Mechanisms of Anti-NMDA Receptor Encephalitis. J Neurosci. 2010 Apr 28;30(17):5866-75

Irani SR and Vincent A. NMDA Receptor Antibody Encephalitis. Curr Neurol Neurosci Rep. 2011 Jun;11(3):298-304.

Day GS, High SM, Cot B, and Tang-Wai DF. Anti-NMDA-Receptor Encephalitis: Case Report and Literature Review of an Under-Recognized Condition. J Gen Intern Med. 2011 Jul;26(7):811-6.

Contributed by Jessica Dwyer, MD and Bruce Rabin, MD, PhD

Case IndexCME Case StudiesFeedbackHome