Individuals with psychosis more likely to have certain antibodies

Individuals with psychosis were more likely to have a higher prevalence of a particular antibody vs. controls, according to results of a meta-analysis published in The Lancet Psychiatry.

“The discovery of an encephalitic syndrome driven by antibodies targeting the N-methyl-D aspartate receptor (NMDAR; specifically the NR1 subunit), in which psychotic symptoms feature prominently, has prompted questions as to whether these neuronal surface antibodies are also present in psychosis patients with no accompanying encephalopathic illness,” Alexis E Cullen, PhD, of the Institute of Psychiatry, Psychology & Neuroscience at King’s College London, and colleagues wrote. “Although the detection of these antibodies in cerebrospinal fluid is the gold standard required for a diagnosis of NMDAR autoimmune encephalitis, lumbar puncture is rarely undertaken in routine psychiatric care because it is impractical to do so.”

Thus, research in this area has largely focused on NMDAR antibodies in serum. Prior studies that measured the immunoglobulin g subclass in serum demonstrated variable prevalence estimates, and it is not known whether individuals with psychosis are more likely than controls to have these antibodies. Cullen and colleagues sought to evaluate the effects of these factors on heterogeneity, since prior inconsistencies could be linked to patient characteristics and methodological approaches. They searched two databases for cross-sectional and case-control studies published between January 2000 and May 5, 2019. They included 14 cross-sectional and 14 case-control studies that reported NMDAR IgG antibody seropositivity among individuals with psychosis.

Results of the analysis of cross-sectional studies showed NMDAR IgG antibodies detected among 0.73% (95% CI, 0.09-1.38) of individuals with psychosis. However, those with psychosis were not significantly more likely to be seropositive than healthy individuals (OR = 1.57; 95% CI, 0.78-3.16) in case-control studies. According to meta-regression analyses, heterogeneity was significantly linked to assay type across both study designs, illness stage in cross-sectional studies and study quality in case-control studies. Cross-sectional and case-control studies that used the live method had higher pooled prevalence estimates and higher ORs vs. those that used a fixed cell-based assay. Cross-sectional studies had higher prevalence in exclusively first-episode samples vs. multi-episode or mixed samples. Case-controlled studies had higher ORs in low-quality studies vs. high-quality studies.

“We strongly recommend that future studies make efforts to obtain large, representative patient samples, in which illness phase, stage and mode of onset are accurately classified, which might necessitate the development of national registries,” Cullen and colleagues wrote. “We also suggest using a live [cell-based assay] for serum testing and doing [cerebrospinal fluid] testing (including cell count, serum-paired oligoclonal bands and IgG index where possible) for all seropositive cases and a random sample of seronegative cases, reporting the sensitivity, specificity and positive predictive value of serum results for [cerebrospinal fluid] abnormalities as a minimum. Finally, we recommend that future studies provide mean absolute titer values for patient and control groups, in addition to positive or negative test status, to facilitate comparison of mean differences.”

In a related editorial, Carsten Finke, MD, of the department of neurology at Charité University Medicine Berlin, highlighted the implications of this meta-analysis for future research.

“This meta-analysis identifies the pain points in the analysis of serum NMDAR antibodies in patients with psychosis, namely assay types, disease stage and reporting quality,” Finke wrote. “Further studies are now needed that compare NMDAR antibody frequencies using different assays (live vs fixed [cell-based assays]) in both patients and controls, preferably in serum and [cerebrospinal fluid]. Such studies should also assess differences in the clinical presentation between antibody-positive and antibody-negative patients, for example by regarding the clinical spectrum of psychotic symptoms and additional symptoms such as cognitive impairment, and they should follow the temporal dynamics of symptoms in longitudinal study designs.”