Psychiatric clinical decision making in anti-NMDA receptor encephalitis

Abstract

BackgroundAnti-N-Methyl-D-Aspartate Receptor (NMDAR) encephalitis is an autoimmune disorder caused by directly pathogenic antibodies affecting the central nervous system. The most common initial presentation of this disorder is an acute onset of psychiatric symptoms, severe enough to require hospitalisation. Within weeks, the majority of cases deteriorate neurologically with the development of movement disorders, seizures and autonomic disturbance. Early commencement of immunomodulatory treatment is associated with better outcomes. Therefore, rapid identification of this disorder within psychiatric services is vital but also complicated as, thus far, the psychiatric symptoms have been minimally described. There have been screening and clinical diagnostic criteria proposed but their utility has not been assessed within psychiatric populations. Additionally, there is minimal guidance about psychiatric treatment for anti-NMDAR encephalitis symptoms. AimsThe aim of this thesis is to explore and guide psychiatric clinical decision making when considering a diagnosis of anti-NMDAR encephalitis in adults. Specifically, this thesis aims to:1) Refine the psychiatric symptoms associated with this disorder and determine if there is a difference between psychiatric presentations associated with anti-NMDAR antibodies found in the cerebral spinal fluid (CSF), the gold standard diagnosis, as compared to solely in the serum. 2) Assess the utility of two proposed screening criteria (Herken and Pruss, Scott et al), particularly to determine their ability to identify which adult psychiatric cases, on initial presentation to hospital, require antibody testing. 3) Assess the utility of proposed clinical diagnostic criteria (Graus et al) which may be beneficial to services without timely access to lumbar puncture or laboratory antibody analysis. 4) Identify the current psychiatric treatments being utilised in anti-NMDAR encephalitis cases, including pharmacology, psychological therapy and electroconvulsive therapy (ECT). Evaluate the risks and side effect profile associated with these treatments as well as noted benefits. MethodsA systematic review of all published cases of anti-NMDAR encephalitis was used to elucidate the specific psychiatric symptoms associated with this disorder. Data gained from the published cases, informed variables collected from a state-wide (Queensland, Australia) cohort of anti-NMDAR antibody positive cases presenting to psychiatric services. The psychiatric presentation and outcomes of CSF antibody positive cases were compared to serum positive/CSF negative antibody cases, using the Queensland cohort. Descriptive statistics were utilise for analysis. The two proposed screening criteria were assessed using anti-NMDAR antibody positive cases which presented with first episode psychosis (FEP) (subset of literature cohort and Queensland cohort) and a control cohort of FEP anti-NMDAR antibody negative cases. Area under a receiver operator characteristic curve, sensitivity and specificity were used as a measure of test accuracy. The proposed diagnostic criteria were also assessed using these anti-NMDAR antibody positive and negative cases, updated for recency. Sensitivity and specificity were reported at two timepoints: day seven of admission and at discharge from hospital. The spectrum, benefits and side effects of psychiatric treatments were assessed in an expanded Queensland cohort. Given initial indications of the benefit of ECT for catatonic symptoms, a systematic review of the use of ECT in anti-NMDAR encephalitis was conducted. ResultsIn the systematic review of over 600 cases, behavioural symptoms, such as severe agitation, speech, sleep and mood disturbance, were the most common initial symptoms of anti-NMDAR encephalitis, occurring in up to 80%. Psychotic symptoms were present in under half of the cases. There were four symptoms - prodromal cognitive deficits, catatonia, speech disturbance and anti-psychotic sensitivity which differentiated the majority of CSF antibody positive from serum positive/CSF negative antibody cases, in the Queensland cohort. These psychiatric symptoms were all present within the first few days of admission. The two proposed screening criteria demonstrated high accuracy, both with a sensitivity of over 97%. There was no significant difference in accuracy of the criteria when neurological symptoms were excluded, indicating utility of the screening criteria at initial psychiatric admission. Additionally, although the criteria were designed for use in FEP, there was maintained sensitivity in identifying all psychiatric presentations of anti-NMDAR encephalitis. Proposed diagnostic criteria, based on clinical symptoms, were highly specific (100%) but demonstrated low sensitivity, 19% by day seven of admission and 49% by the point of discharge, compared to the gold standard diagnosis (CSF anti-NMDAR antibody detection). The clinical diagnostic criteria did detect cases on average 16 days earlier than antibody testing. The majority (80%) of anti-NMDAR encephalitis cases in Queensland received antipsychotics, prior to immunotherapy, but psychotropic choice was variable and polypharmacy common. There was no significant difference in the rate of extrapyramidal or neuroleptic malignant syndrome (NMS) like symptoms between those prescribed antipsychotics and those not. Aggression and agitation reduced or resolved with antipsychotics, although there was minimal improvement in other symptoms such as psychosis, anxiety or catatonia. Systematic review of 30 cases using ECT, primarily for catatonia, in anti-NMDAR encephalitis, demonstrated that this was both an effective (65% improved) and safe (87% no safety concerns) treatment modality.