Conference Coverage

Recognizing anti-NMDA receptor encephalitis psychosis on the psych ward


 

– Prominent psychiatric symptoms are common in patients with anti-N-methyl-D-asparate receptor (NMDAR) encephalitis and often occur prior to onset of obvious neurologic symptoms, Maarten J. Titulaer, MD, PhD, said at the annual congress of the European College of Neuropsychopharmacology.

Moreover, occasionally the psychiatric symptoms occur in isolation without neurologic involvement, as was the case in 4% of a series of 501 patients with confirmed anti-NMDAR encephalitis reported by Dr. Titulaer and coinvestigators. The most prominent symptoms included delusional thinking, aggression, and mood disturbances, which were usually manic (JAMA Neurol. 2013 Sep 1;70[9]:1133-9).

Dr. Maarten J. Titulaer

Dr. Maarten J. Titulaer

The clinical implication is that it’s important to consider the possibility of anti-NMDAR encephalitis in all patients presenting with new-onset psychosis or mania, and to be cognizant of the red flags that warrant definitive diagnostic testing. This serious disease responds well to aggressive immunotherapy, especially if started early, which is impressive given that three-quarters of affected patients will need to spend a month or two in the ICU during the course of their illness, explained Dr. Titulaer, a neurologist at Erasmus University in Rotterdam, the Netherlands.

Anti-NMDAR encephalitis is an autoimmune disorder in which autoantibodies directed at NMDA receptors on neuronal plasma membranes induce severe neurologic and often psychiatric symptoms.

Red flags that raise the chance that a patient on the psychiatric ward with new-onset psychosis or mania might have primary anti-NMDAR encephalitis and should undergo diagnostic testing include autonomic disturbances such as tachycardia, fever, or hypertension, mild neurologic symptoms such as facial twitching, as well as catatonia, seizures, mutism, or development of extrapyramidal symptoms when placed on an antipsychotic agent. Anti-NMDAR encephalitis can have a relapsing course, so any behavioral change in a patient with a history of the disorder might signal relapse.

Certain cancers are strongly associated with anti-NMDAR encephalitis. New-onset psychotic or manic patients with a history of ovarian teratoma, small cell lung cancer, breast cancer, or thymoma should be tested for anti-NMDAR encephalitis. And conversely, screening for those tumors in occult form is warranted in patients with confirmed anti-NMDAR encephalitis, according to Dr. Titulaer.

Systematic screening for anti-NMDAR encephalitis should also be considered in women with severe acute psychosis during the postpartum period, particularly in the setting of extrapyramidal side effects of antipsychotic agents. Two of 96 consecutive women with acute-onset postpartum psychosis in a series reported by Dr. Titulaer and colleagues were antibody-positive for the disorder, and neither had an ovarian teratoma (Am J Psychiatry. 2015 Sep 1;172[9]:901-8).

If a patient hasn’t developed neurologic symptoms within 4 weeks after onset of psychiatric symptoms, anti-NMDAR psychosis becomes far less likely.

Some neurologists have suggested the presence of other autoimmune disorders in psychiatric patients is associated with increased likelihood that the psychiatric symptoms are secondary to anti-NMDAR encephalitis, but Dr. Titulaer doesn’t find the evidence to date persuasive.

The diagnosis of anti-NMDAR encephalitis hinges on the finding of IgG antibodies against the NR1 subunit of the NMDAR. But Dr. Titulaer and coinvestigators have shown there are testing pitfalls: The first-line commercially available cell-based serum assays have a sensitivity of roughly 75% along with 97%-99% specificity, so by relying solely on the cell-based assays a physician might miss one in four cases of anti-NMDAR encephalitis and wrongly diagnose the disease in 0.4%-3% of healthy individuals (Lancet Neurol. 2014 Feb;13[2]:167-77).

For this reason, a positive serum test should be confirmed by a cell-based assay of a cerebrospinal fluid (CSF) sample, which has 100% sensitivity and specificity. And if the serum assay is negative but anti-NMDAR is suspected based on clinical grounds or history, go ahead and test the CSF, the neurologist advised.

Other tools that can be helpful in making the diagnosis include the EEG, which is abnormal in 89% of patients with anti-NMDAR encephalitis. Thirty percent of affected patients will display a highly specific EEG abnormality called extreme delta brushes (Neurology. 2012 Sep 11;79[11]:1094-100).

Dr. Titulaer said that this extreme delta brushes pattern is not seen on the regular psychiatry ward, but only in the ICU, when the patient is severely ill. He has yet to see the first convincing extreme delta brushes pattern in a patient outside the ICU.

Brain MRI has proved “very disappointing,” as it’s abnormal in only one-third of patients with anti-NMDAR encephalitis, he continued.

First-line immunotherapy is corticosteroids, plasmapheresis, and/or intravenous immunoglobulin. In a series of 501 patients who received first-line immunotherapy or tumor removal, 53% improved within 4 weeks. Fifty-seven percent of those who didn’t then got second-line immunotherapy with rituximab (Rituxan) or cyclophosphamide. Outcomes continued to improve for up to 18 months following symptom onset. At 24 months of follow-up, just over 80% of patients in this observational study had a good outcome as defined by a modified Rankin scale score of 0-2, meaning they were living independently with no or minimal disability.

“Not bad, especially considering that the patients who didn’t improve on first-line therapy were in the ICU for a median of 6 weeks,” the neurologist observed.

“It’s important to diagnose patients with anti-NMDAR encephalitis,” he stressed. “Treatment might be difficult. You might need to be very aggressive. But in the end there are very good outcomes. It’s very rewarding to treat these patients.”

In multivariate analysis, Dr. Titulaer and coworkers identified earlier treatment and milder illness as reflected in no ICU admission as significant predictors of good outcome in the study population. Also, the use of second-line immunotherapy in nonresponders to first-line therapy was independently associated with a 2.69-fold increased likelihood of good outcome (Lancet Neurol. 2013 Feb;12[2]:157-65).

Twelve percent of patients experienced one or more relapses within 2 years.

In a separate study of 661 patients with anti-NMDAR encephalitis, only 31 were aged 45 years or older. They had less severe disease than the younger adults but a paradoxically worse outcome, possibly because their median time to diagnosis was twice as long. At 2 years, 60% of the patients aged 45 and up had full or substantial recovery (Neurology. 2013 Sep 17;81[12]:1058-63).

He stressed that treatment of anti-NMDAR encephalitis ought to be an interdisciplinary effort. Psychiatrists will typically not be the ones who administer the potent immunotherapy. But most patients will have behavioral problems in the very early and late phases that warrant psychiatric therapy. Dr. Titulaer suggested psychiatrists steer clear of haloperidol in these patients because it can exacerbate motor symptoms.

Asked if there are any specific patterns of movement disorders linked to anti-NMDAR encephalitis that might raise a psychiatrist’s index of suspicion, the neurologist replied no. Almost all the movement disorders have been seen in psychiatric patients with anti-NMDAR encephalitis. The one specific movement disorder that strongly suggests anti-NMDAR encephalitis is post–herpes simplex virus (HSV) encephalitis choreoathetosis. It appears that HSV encephalitis can trigger formation of NMDAR autoantibodies, resulting in onset of choreoathetosis 3-6 weeks after the HSV encephalopathy.

Dr. Titulaer reported having no financial conflicts of interest in regard to his presentation.

*This story was updated 1/26/2017.

Next Article: