Clinical symptoms and psychosocial functioning in patients with schizophrenia spectrum disorders testing seropositive for anti-NMDAR antibodies: a case-control comparison with patients testing negative

Jurjen J Luykx,Robbert Visscher,Inge Winter-van Rossum,Patrick Waters,Lot D de Witte,W Wolfgang Fleischhacker,Bochao Danae Lin,Nini de Boer,Marte van der Horst,Ksenija Yeeles,Michael Davidson,Thomas A Pollak,Alkomiet Hasan,Belinda R Lennox

DOI: https://doi.org/10.1016/S2215-0366(24)00249-9

Abstract:Background: Antibodies against the N-methyl-D-aspartate receptor (NMDAR) have been described in the serum of people with schizophrenia spectrum disorders (schizophrenia). However, the prevalence and clinical relevance of these antibodies in schizophrenia is unclear. This knowledge gap includes the possibility of such antibodies being associated with a distinct clinical profile, which in turn might warrant a distinct treatment approach. We aimed to assess the seroprevalence of anti-NMDAR antibodies in schizophrenia, and compare symptoms and psychosocial functioning between patients with schizophrenia who were seropositive and seronegative for these antibodies. Methods: In this case-control comparison, by combining new and existing studies, we included patients diagnosed with schizophrenia from four independent cohorts for whom anti-NMDAR serostatus had been assessed (or could be assessed by us) with live cell-based assays. Included cohorts were from the EULAST study (a trial conducted across 15 European countries and Israel), the OPTiMiSE study (an interventional study in Europe), and the PPiP1 and PPiP2 studies (conducted in the UK). Patients from these cohorts were analysed if they had complete Positive and Negative Syndrome Scale (PANSS) data. No participant had been diagnosed with autoimmune encephalitis or received treatment for this condition. After calculating the prevalence of serum anti-NMDAR antibodies, we examined possible differences in PANSS scores (negative, positive, and general symptom subscales, and total score) between anti-NMDAR-seropositive and anti-NMDAR-seronegative patients. Psychosocial functioning as measured by Personal Social Performance (PSP) score was also compared. All analyses were exploratory and no adjustment was done for multiple testing. People with lived experience were not involved in the conduct of this study. Findings: We collected individual patient data from 1114 patients with schizophrenia across the four cohorts. The study population had a mean age of 28·6 years (SD 7·6) and comprised 382 (34·3%) women and 732 (65·7%) men, including patients of White (929 [83·4%]), Asian (54 [4·8%]), Black (68 [6·1%]), and other (62 [5·6%]) ethnicities. Overall, 41 (3·7%) participants (range 3·1-4·0% across cohorts) tested positive for serum anti-NMDAR antibodies. Lower symptom severity on the negative symptoms PANSS subscale was observed for anti-NMDAR-seropositive patients (mean score 15·8 [SD 6·4]) than for anti-NMDAR-seronegative patients (18·2 [6·8]; Cohen's d=0·36; p=0·026), as well as on the general symptoms PANSS subscale (32·9 [8·9] vs 36·1 [10·1]; d=0·33; p=0·029) and total PANSS score (65·5 [18·5] vs 72·6 [19·3]; d=0·37; p=0·013). Mean PSP score was better in anti-NMDAR-positive patients (62·0 [17·0]) than in anti-NMDAR-negative patients (53·5 [16·3]; d=0·52; p=0·014). Interpretation: Serum NMDAR antibodies are present in 3-4% of patients with schizophrenia and are associated with relatively low severity of negative symptoms and relatively good psychosocial functioning. Thus, although the findings await replication in cohorts from other geographical regions, serum anti-NMDAR antibodies might be associated with a different form of psychotic illness. These findings could inform future prognostic and interventional studies examining whether anti-NMDAR antibodies are associated with a specific course of illness or with treatment response. Funding: None.