We thank Pollak et al for reiterating that anti-NMDA receptor encephalitis should be included as a differential diagnosis for patients presenting with acute psychosis. The association of anti-NMDA receptor encephalitis with psychosis is new, having been identified only as recently as 2008, ^{Reference Dalmau, Gleichman, Hughes, Rossi, Peng and Lai1} although the disorder has likely gone unrecognised and indeed untreated previously. Although to date there are no estimates regarding population prevalence rates of anti-NMDA receptor encephalitis, the California Encephalitis Project retrospectively screened 3000 patients with idiopathic encephalitis (with dyskinesia or movement disorders) and identified 10 (0.3%) anti-NMDA receptor-positive cases. ^{Reference Gable, Gavali, Radner, Tilley, Lee and Dyner2} Examining the incidence of catatonia in psychosis, Fink & Taylor estimate a prevalence of between 9 and 17% of patients in academic psychiatry in-patient units, ^{Reference Fink and Taylor3} while Peralta et al found that 31% of drug-naive patients with first-onset psychosis demonstrated at least one catatonic symptom, and found an interesting subgroup that showed a clear association with disorganisation and dyskinesia. ^{Reference Peralta, Campos, de Jalon and Cuesta4}

The neuropsychiatric presentation underlying NMDA receptor encephalitis has only recently been published in the psychiatric literature. ^{Reference Chapman and Vause5} Consequently, this clinical presentation involving psychiatric symptoms in approximately 77% of affected individuals has not been widely disseminated among psychiatrists. This was the driving force behind the publication of our case series.

Pollak et al restate our view that ‘there may be a pure psychiatric presentation associated with lower antibody titres’, and point to their own recent work showing that 3 out of 46 patients with first-episode psychosis had NMDA receptor antibodies. ^{Reference Zandi, Irani, Lang, Waters, Jones and McKenna6} This extremely important finding has profound implications for future differential diagnoses of first-onset psychosis, potentially involving relevant auto-antibody and, specifically, anti-NMDA receptor screening. Further, plasmapheresis may be required and in some cases may even be clinically indicated before a diagnosis of NMDA receptor encephalitis is confirmed. This will have implications for hospital resources and will require close liaison between psychiatry and neurology services.

N-methyl-D-aspartate receptor hypofunction, whether due to exposure to phencyclidine ingestion, NMDA receptor auto-antibody or altered NMDA receptor trafficking, ^{Reference Föcking, Dicker, English, Schubert, Dunn and Cotter7,Reference Schubert, Föcking, Prehn and Cotter8} is now implicated even more strongly in schizophrenia. Future studies focusing on this area may provide clues not only to the screening and management of NMDA receptor encephalitis among first-episode psychosis populations, but may also lead to a broader understanding of schizophrenia pathophysiology, with the potential for development of novel treatment strategies.