SCHRES-05858; No of Pages 10 Schizophrenia Research xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Schizophrenia Research journal homepage: www.elsevier.com/locate/schres

Meta-analysis of the association between N-methyl-D-aspartate receptor antibodies and schizophrenia, schizoaffective disorder, bipolar disorder, and major depressive disorder Daniel M. Pearlman a,b,1, Souhel Najjar a,⁎,1 a b

Neuroinflammation Research Group, Epilepsy Center Division, Department of Neurology, NYU School of Medicine, New York, NY, USA The Dartmouth Institute for Health Policy and Clinical Practice, Audrey and Theodor Geisel School of Medicine at Dartmouth, Lebanon, NH, USA

a r t i c l e

i n f o

Article history: Received 5 February 2014 Received in revised form 30 April 2014 Accepted 1 May 2014 Available online xxxx Keywords: N-methyl-D-aspartate receptor Schizophrenia Schizoaffective disorder Bipolar disorder Major depressive disorder Antibodies Immunoglobulin Meta-analysis Systematic review Psychoimmunology

a b s t r a c t Background: N-methyl-D-aspartate receptor (NMDAR) antibodies have been documented in the serum of individuals with primary psychiatric disorders from several independent cohorts, but these findings have not been systematically assessed in aggregate or in relation to methodological covariates. Methods: We searched MEDLINE, EMBASE, and PsycINFO for studies in any language that provided data on NMDAR antibody seropositivity or absolute serum titers in schizophrenia or schizoaffective, bipolar, or major depressive disorders. We used a random effects model to pool estimates across studies. Results: Nine studies met the eligibility criteria. Five studies (3387 participants) provided data on NMDAR antibody seropositivity in psychiatric versus control groups based on high-specificity seropositivity thresholds (cell-based assays [CBAs]: 1:320 dilution, 1:200 dilution, visual score N 1; enzyme-linked immunosorbent assay [ELISA]: 90th percentile of control titers). Meta-analysis showed significantly higher odds of NMDAR antibody seropositivity among those with schizophrenia or schizoaffective, bipolar, or major depressive disorders compared with healthy controls (odds ratio [OR], 3.10; 95% confidence interval [CI], 1.04–9.27; P = .043; I2 = 68%). Four studies (3194 participants) provided outcome data for these groups based on low-specificity seropositivity thresholds (CBAs 1:10 dilution; ELISA: 75th percentile of control titers). Meta-analysis showed greater heterogeneity and no significant between-group difference (OR, 2.31; 95% CI, 0.55–9.73; P = .25; I2 = 90%). Seropositive participants in psychiatric groups had various combinations of IgG, IgM, and IgA class antibodies against NR1, NR1/NR2B, and NR2A/NR2B subunits. Subgroup analysis revealed significantly higher odds of seropositivity among all participants based on 1:10 versus 1:320 dilution seropositivity thresholds (OR, 4.56; 95% CI, 2.41–8.62; P b .001; I2 = 0%; studies = 2, n = 2920), but no apparent difference between first-episode and chronic schizophrenia or schizoaffective disorder (OR, 1.15; 95% CI, 0.19–7.24; P = .88, I2 = 43%, studies = 2, n = 1108). Average NR2A/NR2B antibody titers determined by ELISA were significantly higher among participants with first-episode schizophrenia (P b .0001) and acute mania (P b .01) compared with healthy controls. Levels decreased by 58% at 8 weeks in first-episode schizophrenia, and by about 13% at 4 days in acute mania. Conclusions: Individuals with schizophrenia or schizoaffective, bipolar, or major depressive disorders are collectively about three times more likely to have elevated NMDAR antibody titers compared with healthy controls based on high-specificity, but not low-specificity, seropositivity thresholds, though considerable methodological and statistical heterogeneity exists. Evidence concerning the effect of disease state and time of serum acquisition is varied and consistent, respectively. Adequately powered longitudinal studies employing standardized assay methods and seropositivity threshold definitions, and quantifying NMDAR antibodies in both sera and cerebrospinal fluid are needed to further elucidate the clinical and pathophysiological implications of this association. © 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

1. Introduction

⁎ Corresponding author at: Epilepsy Center Division, Department of Neurology, NYU School of Medicine, 223 East 34th Street, New York, NY 10016-4852, USA. Tel.: +1 646 558 0807; fax: +1 646 385 7166. E-mail address: [email protected] (S. Najjar). 1 Both authors contributed equally to this work.

N-methyl-D-aspartate receptors (NMDARs) are glutamate-gated ion channels that mediate higher cognitive function and synaptic plasticity (recently reviewed by Paoletti et al., 2013). Each NMDAR consists of two NR1 subunits, and two non-NR1 subunits (NR2A, NR2B, NR2C, NR2D, NR3A, and/or NR3B). Each subunit, in turn,

http://dx.doi.org/10.1016/j.schres.2014.05.001 0920-9964/© 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

Please cite this article as: Pearlman, D.M., Najjar, S., Meta-analysis of the association between N-methyl-D-aspartate receptor antibodies and schizophrenia, schizoaffective disorder, bipolar disorder, and ..., Schizophr. Res. (2014), http://dx.doi.org/10.1016/j.schres.2014.05.001

2

D.M. Pearlman, S. Najjar / Schizophrenia Research xxx (2014) xxx–xxx

consists of an extracellular N-terminal domain, an extracellular agonist-binding domain, a transmembrane domain, and an intracellular C-terminal domain. Activation of NMDARs requires that coagonist glycine or D-serine binds the agonist-binding domain of the NR1 subunits, and that glutamate binds the agonist-binding domain of the two non-NR1 subunits. Antibodies directed against NMDARs found in selected autoimmune diseases are associated with acute and first-episode neuropsychiatric sequelae that resemble primary psychotic and mood disorders. The pathophysiology of anti-NMDAR encephalitis — the prototype of these disorders — is characterized by IgG class NMDAR antibodies binding the N368/G369 region of the N-terminal domain of the NR1 subunit on dendritic clusters (Gleichman et al., 2012), resulting in crosslinking and internalization of these subunits in a dose-dependent manner (Hughes et al., 2010). Up to two-thirds of patients with anti-NMDAR encephalitis, initially present to psychiatric, rather than neurological, services (Dalmau et al., 2008). Kayser and colleagues found that the prevalence of isolated psychiatric presentations of anti-NMDAR encephalitis was 4% (23 of 571)—including five cases presenting at first-episode and eight cases presenting during acute relapses (Kayser et al., 2013). NMDAR antibodies are also implicated in neuropsychiatric systemic lupus erythematosus, whereby IgG class NMDAR antibodies are thought to interact with the allosteric zinc-binding site on the agonist-binding domain of NR2A and NR2B subunits (Omdal et al., 2005; Lapteva et al., 2006; Arinuma et al., 2008; Gono et al., 2011). Once bound, these antibodies can promote NMDAR-mediated excitotoxicity via upregulation of neuronal mitochondrial permeability (Faust et al., 2010) and intracellular Ca2 + influx (Gono et al., 2011). Further, they can increase blood–brain permeability by upregulating endothelial inflammatory and vascular alterations (Yoshio et al., 2013). Among those with neuropsychiatric systemic lupus erythematosus, NMDAR antibodies have been specifically associated with depressive mood and psychosis as opposed to focal neurological sequelae (Lauvsnes and Omdal, 2012). NMDAR antibodies have also been associated with progressive cognitive impairment (Pruss et al., 2012b), primary Sjögren's syndrome (Lauvsnes et al., 2013) and herpes simplex encephalitis (Prüss et al., 2012; Armangue et al., 2014). Although NMDAR antibodies have been linked to psychiatric sequelae secondary to autoimmune diseases, and both NMDAR dysfunction (Olney and Farber, 1995; Jentsch and Roth, 1999) and comorbid autoimmune disease (Eaton et al., 2010; Benros et al., 2013; Benros et al., 2014a; Benros et al., 2014b) have been consistently linked to primary psychiatric disorders, the relationship between NMDAR antibodies and primary psychiatric disorders is less clear. Over the last 4 years, several studies have yielded mixed findings concerning this relationship (Ezeoke et al., 2013; Pollak et al., 2013). There are no systematic reviews to date evaluating the relationship between NMDAR antibodies and bipolar or major depressive disorders, or addressing the effects of covariates such as NMDAR-subunit/immunoglobulin-class combinations assessed versus identified, timing of serum acquisition, disease state severity, medication status, or assay methods, including the dilution cut-off points used to define seropositivity. Here, we aimed to address these remaining gaps in the literature by performing a systematic review with a meta-analysis of the association between NMDAR antibodies and schizophrenia and schizoaffective disorder, as well as bipolar and major depressive disorders. We also aimed to isolate sources of heterogeneity underlying the mixed findings in the extant literature. We discuss our findings in light of the implications for clinical practice and future etiologic research in psychoimmunology. 2. Methods and materials We devised the methods of this systematic review in accordance with the MOOSE guidelines (Stroup et al., 2000).

2.1. Search strategy We sought potentially eligible studies for inclusion through a series of electronic database searches within Ovid MEDLINE (1947–2014), EMBASE (1974–2014), and PsycINFO (1808–2014); see Table S1 in the Appendix for a copy of our full-length search strategy. We looked for peer-reviewed journal articles (published data) and meeting abstracts (unpublished data) written in any language. Searches were regularly updated through 26 April 2014. We supplemented these searches by handsearching the reference lists of included studies and of other content-relevant articles. We also asked experts in the field if they were aware of any additional studies (12th Psychoimmunology Expert Meeting; 6–9 March 2014; Günzburg, Germany). 2.2. Study selection We defined eligible studies as those using case–control, prospective cohort, or cross-sectional designs and providing categorical (seropositive/seronegative) and/or continuous data (antibody titers) on NMDAR antibodies in individuals with schizophrenia, schizoaffective disorder, bipolar disorder, and/or major depressive disorder. We included studies that involved participants enrolled at first-episode presentations who were later diagnosed with one of these psychiatric disorders after establishing symptom persistence (as required by diagnostic criteria). To preserve intragroup homogeneity, we excluded participants with psychotic disorder not otherwise specified, delusional disorder, borderline personality disorder, comorbid autoimmune disease, and non-healthy controls, as well as those whose initial diagnosis was one of the eligible psychiatric disorders when the diagnosis was subsequently revised to anti-NMDAR encephalitis after obtaining antibody test results (Table S2). This ensured that studies addressed the focus of our research question regarding individuals with psychiatric disorders, as the prevalence of isolated psychiatric manifestations of antiNMDAR encephalitis has been assessed directly elsewhere (Kayser et al., 2013). Study eligibility was determined by consensus. 2.3. Data collection We extracted data from the full-texts of included studies on predefined variables using a standardized and piloted data collection form. To characterize included studies, we extracted data for eligibility criteria, citation information, country of publication, participant recruitment strategy, and methods used for immunostaining assays. Data with regard to the latter included assay technique, NMDAR subunits assessed, immunoglobulin class antibodies assessed, blinding of primary outcome assessment, and dilution cut-off points or other criteria used to define NMDAR antibody seropositivity. To characterize included study participants, we extracted data for psychiatric and control groups on age, gender, positive psychotic symptoms, and disease state inclusion criteria. For the initial data extraction, we preserved any distinctions regarding psychiatric disorders and disease states. For example, wherever possible, we extracted data on participants with schizophrenia alone, rather than on participants with either schizophrenia or schizoaffective disorder. Likewise, we extracted data for participants with first-episode schizophrenia, rather than on those with either first-episode or chronic schizophrenia. Last, we preserved any distinctions with regard to the study time point of serum acquisition and in the criteria constituting the operational definitions used to define the threshold of NMDAR antibody immunoreactivity/titers indicative of a seropositive versus seronegative test result (henceforth “seropositivity thresholds”). We defined our primary outcome, NMDAR antibody seropositivity, according to included study investigators' operational definitions as provided in the original reports. We summarized primary outcome data for each participant group, disease state, time of serum acquisition, and seropositivity threshold as the odds of NMDAR antibody

Please cite this article as: Pearlman, D.M., Najjar, S., Meta-analysis of the association between N-methyl-D-aspartate receptor antibodies and schizophrenia, schizoaffective disorder, bipolar disorder, and ..., Schizophr. Res. (2014), http://dx.doi.org/10.1016/j.schres.2014.05.001

D.M. Pearlman, S. Najjar / Schizophrenia Research xxx (2014) xxx–xxx

seropositivity equal to the number of seropositive participants (events) divided by the number of seronegative participants (non-events). Our secondary outcome was serum titers of NMDAR antibodies in nanograms per milliliter. We summarized secondary outcome data for each study group, disease state, and time of serum acquisition, as mean ± standard deviation antibody titers. 2.4. Statistical analysis For studies that provided primary outcome data on psychiatric and healthy control groups, we calculated point estimate odds ratios (ORs), with corresponding 95% confidence intervals (CIs) equal to the odds of NMDAR antibody seropositivity among participants with psychiatric disorders divided by those of the healthy control group. We repeated these calculations for each study according to psychiatric disorder, disease state, time of serum acquisition, and seropositivity threshold. We performed meta-analysis on individual study point estimates of NMDAR antibody seropositivity for psychiatric versus healthy control groups based on high- and low-specificity seropositivity thresholds. We performed subgroup analyses comparing distinct disease states and seropositivity thresholds. For studies providing secondary outcome data on psychiatric and healthy control groups, we calculated the weighted mean difference with 95% CIs. Likewise, we repeated this analysis for each psychiatric disorder, disease state, and time of serum acquisition. We also extracted data on analysis of variance tests of trend for the interaction of antibody titers and time. We used the DerSimonian and Laird random effects model to undertake meta-analysis (DerSimonian and Laird, 1986). We chose a random effects model because it generally yields more conservative estimates than do fixed-effect models such as the Mantel–Haenszel method. Moreover, it is a more appropriate statistical test to address this specific research question given the extent of methodological variability across studies. Secondary outcome data were not pooled for the metaanalysis because of insufficient data. We tested for heterogeneity across point estimates pooled for the meta-analysis by computing the χ2 statistic (Cochrane's Q), and applied the standard of P values less than 0.10 as indicative of statistically significant heterogeneity. We quantified any observed heterogeneity by computing the I2 statistic. We considered values of I2 ≤ 25%, ≤ 50%, and ≤ 75%, as indicative of low, moderate, and high degrees of heterogeneity, respectively (Higgins et al., 2003). We attempted to obtain any missing data on primary and secondary outcome measures by emailing corresponding authors. With the exception of the above-mentioned heterogeneity estimates, we considered P values less than 0.05 statistically significant. We undertook all statistical analyses using Stata version 12.1 (StataCorp, LP. College Station, TX, USA). 3. Results 3.1. Search results Results of the study search and selection processes are shown in Fig. S1. A total of nine studies met all eligibility criteria (Dickerson et al., 2013; Govorin and Vasil'eva, 2011; Hammer et al., 2013; Haussleiter et al., 2012; Masdeu et al., 2012; Rhoads et al., 2011; Steiner et al., 2013; Tsutsui et al., 2012; Zandi et al., 2011; Table 1). Most excluded studies were discernibly ineligible; one exception was a retrospective cohort briefly mentioned within a case report article, wherein the authors found 3 IgM class NMDAR antibody (unspecified subunit(s)) seropositive participants among 70 patients with a clinical diagnosis of psychosis not otherwise specified (Choe et al., 2013). Among our included studies, two reported data across more than one report: Dickerson et al. (2012) and Dickerson et al. (2013), as well as Zandi et al. (2011) and Lennox et al. (2009). We obtained an English translation for Govorin and Vasil'eva (2011), which was written in Russian. We also obtained data on NMDAR

3

antibody seropositivity by disease state severity subgroups for Hammer et al. (2013) from a published article wherein the authors acquired these data via a personal communication (Pollak et al., 2013). All other data were obtained directly from the original published sources. Seven studies used HEK293 cell-based assays, and two studies used enzyme-linked immunosorbent assays (ELISAs); studies tested for varied combinations of immunoglobulin classes and NMDAR subunits (Table 1). Characteristics of included study participants are summarized in Table 2. 3.2. Primary outcome: N-methyl-D-aspartate receptor antibody seropositivity Eight studies provided primary outcome data for one or more psychiatric groups (Fig. 1); the exception was Govorin and Vasil'eva (2011). Six of these studies also provided primary outcome data for a healthy control group (Rhoads et al., 2011; Zandi et al., 2011; Masdeu et al., 2012; Dickerson et al., 2013; Hammer et al., 2013; Steiner et al., 2013). Five studies (3387 participants) provided data on NMDAR antibody seropositivity based on high-specificity seropositivity thresholds. Dickerson et al. (2013) used the 90th percentile of control titers, Masdeu et al. (2012) used 1:200 dilution, and Zandi et al. (2011) used a visual score of immunoreactivity greater than one (as described by Irani et al., 2010). Steiner et al. (2013) and Hammer et al. (2013) provided seropositivity data based on serial titrations. For these studies, we chose 1:320 dilution as the high-specificity threshold because it was the lowest common denominator between them that exceeded 1:10 dilution and the clinically validated standard of 1:200 dilution used in the context of diagnosing anti-NMDAR encephalitis (Dalmau et al., 2008). Meta-analysis showed significantly higher odds of NMDAR antibody seropositivity among participants with schizophrenia, schizoaffective disorder, bipolar disorder, or major depressive disorder, as compared with healthy controls (OR, 3.10; 95% CI, 1.04–9.27, P = .043; Fig. 2A). There was a moderate degree of statistically significant heterogeneity across the individual study point estimates contributing to this summary estimate (I2 = 68%, P = .025). Four studies (3194 participants) provided data on NMDAR antibody seropositivity based on low-specificity thresholds. Dickerson et al. (2013) used the 75th percentile of control titers, whereas the remaining three studies used 1:10 dilution (Rhoads et al., 2011; Hammer et al., 2013; Steiner et al., 2013). The summary estimate showed a statistically insignificant increase in the odds of NMDAR antibody seropositivity among participants with schizophrenia or schizoaffective, bipolar, or major depressive disorders, as compared with healthy controls (OR, 2.31; 95% CI, 0.55–9.73; P = .25; Fig. 2B). There was a high degree of statistically significant heterogeneity across the individual study point estimates contributing to this summary estimate (I2 = 90%, P b .001). Two studies (1108 participants) provided data on NMDAR antibody seropositivity for two or more distinct disease states (Zandi et al., 2011; Hammer et al., 2013). Meta-analysis showed no statistically significant difference between the odds of NMDAR antibody seropositivity among 132 participants with first-episode schizophrenia or schizoaffective disorder as compared with 976 participants with chronic schizophrenia or schizoaffective disorder (OR, 1.15; 95% CI, 0.19–7.24; P = .88; Fig. 3A). There was a moderate degree of heterogeneity across studies, which was not statistically significant (I2 = 43%, P = .19). Two studies (2920 participants) provided data on NMDAR antibody seropositivity for two or more distinct seropositivity thresholds (Hammer et al., 2013; Steiner et al., 2013). The summary estimate showed that all participants, including those in psychiatric or healthy control groups, were four and one-half times more likely to test seropositive for NMDAR antibodies based on a 1:320 dilution seropositivity threshold as compared with a 1:10 dilution seropositivity threshold (OR, 4.56; 95% CI, 2.41–8.62; P b .001; Fig. 3B). There was a moderate degree of statistically insignificant heterogeneity across point estimates (I2 = 44%, P = .18). For the study by

Please cite this article as: Pearlman, D.M., Najjar, S., Meta-analysis of the association between N-methyl-D-aspartate receptor antibodies and schizophrenia, schizoaffective disorder, bipolar disorder, and ..., Schizophr. Res. (2014), http://dx.doi.org/10.1016/j.schres.2014.05.001

4

D.M. Pearlman, S. Najjar / Schizophrenia Research xxx (2014) xxx–xxx

Table 1 Included studies.

Hammer et al. (2013)a Steiner et al. (2013)a Dickerson et al. (2013) Haussleiter et al. (2012) Tsutsui et al. (2012) Masdeu et al. (2012) Zandi et al. (2011) Rhoads et al. (2011) Govorin and Vasil'eva (2011)

Consecutive

Blinded

CBA b

ELISA c

NR1

NR1/NR2B

NR2A/NR2B

IgG

IgM

IgA

? ● ● ? ? ? ● ? ?

● ● ? ? ? ● ● ? ?

● ● ○ ● ● ● ● ● ○

○ ○ ● ○ ○ ○ ○ ○ ●

● ● ● ● ○ ● ○ ● ○

● ● ○ ● ● ○ ● ○ ○

○ ○ ● ○ ○ ○ ○ ○ ●

● ● ● ? ● ● ● ● ?

● ● ○ ? ○ ○ ○ ○ ?

● ● ○ ? ○ ○ ○ ○ ?

Included study participant enrollment strategy (consecutive series vs. convenience sample), masking of outcome assessment (blinded vs. un-blinded to case/control status), type of assay (cell-based assay vs. enzyme-linked immunosorbent assay), N-methyl-D-aspartate receptor antibody subunits assessed (NR1, NR1/NR2B, NRA/NR2B), and immunoglobulin classes assessed (IgG, IgM, IgA). ● = yes; ? = unclear/not reported; ○ = no; CBA = cell-based assay; ELISA = enzyme-linked immunosorbent assay; Ig = immunoglobulin. a Refers to studies that reported the maximum dilution at which immunoreactivity was observed for each participant's sample. b Human embryonic kidney (HEK293) cell-based assays. c Gold Dot Test, CIS Biotech, Inc., Atlanta, GA, USA. Table 2 Study participants.a N

Age, years

Male sex

PANSS P1–P7

1081 101 954 26 148 1272

39 (13) .. .. .. .. .. .. 50 (16) 37 (13)

723 (67) .. .. .. .. .. .. 70 (47) 780 (61)

14 (6) .. .. .. .. .. .. .. .. .. ..

Steiner et al. (2013) Schizophrenia First episode Acute (unmedicated for ≥6 weeks) Major depressive disorder (acute/first episode) First episode Acute (unmedicated for ≥6 weeks) Healthy controls

119 45 74 70 35 35 230

36 (11) .. .. .. .. 40 (12) .. .. .. .. 36 (12)

77 (65) .. .. .. .. 26 (37) .. .. .. .. 159 (69)

22 (6) .. .. .. .. 9 (2) .. .. .. .. .. ..

Dickerson et al. (2013) Acute maniab Schizophrenia or schizoaffective disorder (multi-episode, chronic) Bipolar disorder (acute depressive state) Healthy controls

57 234 28 207

35 (13) 42 (12) 38 (14) 33 (11)

16 (28) 137 (59) 6 (21) 80 (39)

.. .. .. .. .. .. .. ..

Haussleiter et al. (2012) Schizophrenia (NR) Schizoaffective disorder (NR)

44 4

43 (13) 45 (7)

15 (34) 1 (25)

21 (7) 13 (3)

Tsutsui et al. (2012) Schizophrenia, Schizoaffective disorder, or brief psychosis (NR) Schizoaffective disorderd

51 3

.. (15–72)c 32 (6)

10 (20) 1 (33)

.. .. .. ..

Masdeu et al. (2012) Schizophrenia (first episode) Healthy controls

80 40

29 (10) 31 (9)

58 (73) 25 (62)

.. .. .. ..

Zandi et al. (2011) Schizophrenia First episode Chronice Bipolar or Major depressive disorders (first episode)f Healthy controlse

53 31 22 7 23

.. .. .. .. .. .. .. .. .. ..

.. .. .. .. .. .. .. .. .. ..

.. .. .. .. .. .. .. .. .. ..

7 3

39 (10) 23 (1)

3 (43) 1 (33)

.. .. .. ..

23 10

24 (6) 24 (3)

12 (52) 4 (40)

14 (2) .. ..

Hammer et al. (2013) Schizophrenia (n = 881) or schizoaffective (n = 200) First episode Chronic Unspecified Bipolar disorder or major depressive disorder (NR) Healthy controls

Rhoads et al. (2011) Schizophrenia (NR) Healthy controls Govorin and Vasil'eva (2011) Schizophrenia (first episode) Healthy controls

NR = not reported; PANSS = Positive and Negative Syndrome Scale; P1–P7 = positive subscale. a Values are mean (standard deviation) for continuous data, and number (percent) for categorical data. b Bipolar 1 disorder manic episode, bipolar 2 disorder hypomanic episode, or schizoaffective disorder (bipolar type) manic, hypomanic, or mixed episode. c Range. d Comorbid narcolepsy. e Primary outcome data was extracted from a meeting abstract reporting results from the same study. f The participants with major depressive disorder were also observed as having psychotic features.

Please cite this article as: Pearlman, D.M., Najjar, S., Meta-analysis of the association between N-methyl-D-aspartate receptor antibodies and schizophrenia, schizoaffective disorder, bipolar disorder, and ..., Schizophr. Res. (2014), http://dx.doi.org/10.1016/j.schres.2014.05.001

D.M. Pearlman, S. Najjar / Schizophrenia Research xxx (2014) xxx–xxx

Please cite this article as: Pearlman, D.M., Najjar, S., Meta-analysis of the association between N-methyl-D-aspartate receptor antibodies and schizophrenia, schizoaffective disorder, bipolar disorder, and ..., Schizophr. Res. (2014), http://dx.doi.org/10.1016/j.schres.2014.05.001

Fig. 1. Odds of N-methyl-D-aspartate receptor antibody seropositivity in participants with schizophrenia, schizoaffective disorder, bipolar disorder, and major depressive disorder versus healthy controls, by disease state severity and dilution cut-off point seropositivity thresholds. A. Admitted during a manic or hypomanic episode and an admitting diagnosis of Bipolar I disorder (single manic episode, most recent episode manic, or most recent episode mixed); Bipolar II disorder (most recent episode hypomanic); or Schizoaffective disorder, bipolar type (manic, hypomanic, or mixed state). B. Community sample, recruited during chronic or multi-episode disease states. C. Admitted with symptoms of depression with an admission diagnosis of bipolar disorder. Note that for a given study, the same participants often contribute to the data on more than one row. BEP = brief episode psychosis; CI = confidence interval; First ep. = First episode; Multi ep. = Multi-episode; NR = not reported. 5

6

D.M. Pearlman, S. Najjar / Schizophrenia Research xxx (2014) xxx–xxx

A High-specificity

Psychiatric group

Healthy controls

Events

Total

%

28

1229

5

189

NR

Zandi et al. (2011), visual score > 1

% Weight

Odds Ratio (95% CI)

Events

Total

%

2.3

22

1272

1.7

41.37

1.32

2.6

0

230

0.0

10.89

13.74

(0.76–250.14)

57

..

NR

207

..

37.40

5.27

(2.39–11.62)

3

60

5.0

0

23

0.0

10.34

2.86

(0.14–57.57)

Masdeu et al. (2012), 1:200 dilution

0

80

0.0

0

40

0.0

0.00

Total

..

1615

..

..

1772

..

100.00

3.10

(1.04–9.27)

Hammer et al. (2013), 1:320 dilution Steiner et al. (2013), 1:320 dilution Dickerson et al. (2013), 90th percentile

Heterogeneity:

2

(0.75–2.33)

Not estimable

= 9.35, df = 3, I2 = 67.9%, P = .025 0.01

Test of overall effect: z = 2.03, P = .043

0.1

1

1

Odds higher in Healthy controls

10

100

Odds higher in Psychiatric group

B Low-specificity

Psychiatric group Events

Events

Total

%

93

1229

7.6

137

1272

10.8

40.52

0.68

Steiner et al. (2013), 1:10 dilution

12

189

6.3

1

230

0.4

22.40

15.53

NR

57

..

NR

207

..

37.08

2.79

(1.33–5.85)

Rhoads et al. (2011), 1:10 dilution

0

7

0.0

0

3

0.0

0.00

Total

..

1482

..

..

1712

..

100.00

2.31

(0.55–9.73)

Heterogeneity:

2

%

Odds Ratio (95% CI)

% Weight

Hammer et al. (2013), 1:10 dilution

Dickerson et al. (2013), 75th percentile

Total

Healthy controls

(0.51–0.89) (2.00–120.53)

Not estimable

= 20.18, df = 2, I2 = 90.1%, P < .001

Test of overall effect: z = 1.14, P = .25

0.01

0.1

Odds higher in Healthy controls

1

1

10

100

Odds higher in Psychiatric group

Fig. 2. Odds of N-methyl-D-aspartate receptor antibody seropositivity for comparisons between schizophrenia, schizoaffective disorder, bipolar disorder, and major depressive disorder versus healthy controls, based on (A) high-specificity and (B) low-specificity seropositivity thresholds. Summary estimates and heterogeneity were computed using DerSimonian and Laird's random effects model. Square data markers and horizontal error bars indicate individual study point estimates of the odds ratio and the corresponding 95% confidence intervals, respectively. The size of each square is directly proportional to the weight the point estimate contributes to the pooled summary estimate (diamond). The solid vertical line reflects the null hypothesis (odds ratio = 1.00); point estimates with error bars that overlap with this line, are statistically insignificant at α = 0.05. CI = confidence interval; NR = not reported.

Hammer et al. (2013), changing the dilution seropositivity threshold from 1:320 to 1:10 had the effect of decreasing the pretest odds of NMDAR antibody seropositivity more than four-fold among those with schizophrenia or schizoaffective disorder as well as among those with bipolar or major depressive disorders. In contrast, doing so resulted in a nearly seven-fold decrease in seropositivity among healthy controls. All three of these intra-group comparisons were statistically significant. A similar but less pronounced trend was observed for the study by Steiner et al. (2013). A third study also provided outcome data for more than one seropositivity threshold (90th and 75th percentiles of healthy control titers; Dickerson et al., 2013). However, it could not be included in these analyses because of missing event data for the number and proportion of seropositive participants underlying the reported odds ratios and 95% CIs (we were unable to obtain these data from the authors). Although the odds ratio for NMDAR antibody seropositivity for the comparison between acute mania and healthy control groups was higher with the 90th percentile cut-off point, the change in seropositivity within each group accounting for the ratio change is unclear. The overall distribution of immunoglobulin class NMDAR antibodies among seropositive participants in the psychiatric group, by disorder, NMDAR subunit(s), disease states, and included study are summarized in Fig. 4, Tables S3, and S4. 3.3. Secondary outcome: N-methyl-D-aspartate receptor antibody titers Two studies provided longitudinal data on NMDAR antibody serum titers in psychiatric disorders. For the study by Dickerson

et al. (2013), we estimated values of serum NMDAR antibody titers from a graph in the original report because of missing outcome data on the absolute titer values and associated test statistics (we were unable to obtain these data from the authors). Serum titers of IgG class NMDAR antibodies against NR2A/NR2B subunits were significantly higher among participants with acute mania, as compared with healthy controls (P b .01). Among those with acute mania, titers decreased by about 13% after an average of 4.0 ± 2.3 days (~ 1.6 vs ~ 1.4 ng/mL) and by about 25% after an average of 191.0 ± 12.9 days (~ 1.6 vs ~ 1.2 ng/mL): test of trend for antibody titers vs. time (P b .007). There were no significant differences in average serum titers of IgG class NMDAR antibodies against NR2A/NR2B subunits for comparisons between each of the remaining psychiatric groups and the healthy control group. Govorin and Vasil'eva (2011) collected serum samples from 23 participants with first episode schizophrenia and 10 healthy controls. Mean ± standard deviation serum NMDAR antibody titers of unspecified immunoglobulin class against NR2A/NR2B subunits were significantly higher among those with acute schizophrenia, as compared with healthy controls (1.94 ± 0.76 vs 0.75 ± 0.32 ng/mL; mean difference, 1.19; 95% CI 0.68–1.70; P = .0001). At eight-week follow-up, titers remained higher among the schizophrenia group (1.23 ± 0.38 vs 0.75 ± 0.32 ng/mL; mean difference, 0.48; 95% CI, 0.20–0.76; P b .0001). Levels were about 58% lower than at baseline (1.94 ± 0.76 vs 1.23 ± 0.38 ng/mL; mean difference, 0.71; 95% CI, 0.35–1.07; P = .002). The test of trend for the effect of time on antibody titers was statistically significant (P b .0002). In addition, among those in the schizophrenia group the authors found significant positive correlations between

Please cite this article as: Pearlman, D.M., Najjar, S., Meta-analysis of the association between N-methyl-D-aspartate receptor antibodies and schizophrenia, schizoaffective disorder, bipolar disorder, and ..., Schizophr. Res. (2014), http://dx.doi.org/10.1016/j.schres.2014.05.001

D.M. Pearlman, S. Najjar / Schizophrenia Research xxx (2014) xxx–xxx

7

Fig. 3. Odds of N-methyl-D-aspartate receptor antibody seropositivity for comparisons between (A) first-episode versus chronic schizophrenia or schizoaffective disorder, and (B) psychiatric plus healthy controls at high-specificity versus low-specificity seropositivity thresholds. Summary estimates and heterogeneity were computed using DerSimonian and Laird's random effects model. Square data markers and horizontal error bars indicate individual study point estimates of the odds ratio and the corresponding 95% confidence intervals, respectively. The size of each square is directly proportional to the weight the point estimate contributes to the pooled summary estimate (diamond). The solid vertical line reflects the null hypothesis (odds ratio = 1.00); point estimates with error bars that overlap with this line, are statistically insignificant at α = 0.05. CI = confidence interval; NR = not reported.

A

57

60 50 38

40

Seropositive 30 Participants

11

20

17

10

1

2

5

0

0

1

IgG

0 0

IgA

IgM

IgG, IgA

NR1

0

IgG, IgM

0

IgA, IgM

0

NR1/NR2B

IgG, IgA, IgM

B 60

51

50 40

31

Seropositive 30 Participants

11

20

15

10

1 5

0 IgG

0

7

IgA

2

7

IgM

0

IgG, IgA

0

IgG, IgM

2

IgA, IgM

Chronic/Unspecified 0

First episode/Acute

IgG, IgA, IgM

Fig. 4. Distribution of immunoglobulin class antibodies against N-methyl-D-aspartate receptors among seropositive participants with psychiatric disorders, by (A) N-methyl-D-aspartate receptor subunits and (B) disease states. The distribution of data illustrated here, by included study, is shown in Tables S4 and S5 in the supplement. Ig = immunoglobulin. For each row, events shown are mutually exclusive (e.g., an individual with IgA and IgM antibodies against NR1 is only counted in the IgA + IgM bar, and not in either the IgA [alone] or IgM [alone] bars).

Please cite this article as: Pearlman, D.M., Najjar, S., Meta-analysis of the association between N-methyl-D-aspartate receptor antibodies and schizophrenia, schizoaffective disorder, bipolar disorder, and ..., Schizophr. Res. (2014), http://dx.doi.org/10.1016/j.schres.2014.05.001

8

D.M. Pearlman, S. Najjar / Schizophrenia Research xxx (2014) xxx–xxx

NR2A/NR2B antibody titers and levels of brain-derived neurotropic factor (r = 0.42, P b .05), glial acid fibrillary protein (r = 0.46, P b .05), and particularly with neuron-specific enolase (r = 0.77, P b .005). 4. Discussion To our knowledge, this is the first systematic review of studies providing evidence on the relationships between NMDAR antibody seropositivity or titers and schizophrenia, schizoaffective disorder, bipolar disorder, and major depressive disorder. There are several findings of note. First, we found that individuals with these psychiatric disorders are collectively about three times more likely to have elevated NMDAR antibody titers compared with healthy controls based on high-specificity, but not low-specificity, seropositivity thresholds. In contrast to prior reports indicating that psychotropic medication can induce serum autoantibody production (Schwartz et al., 2009; Shen et al., 2009), that our results indicate the odds of NMDAR antibody seropositivity and/or average titers are both higher among participants with first-episode presentations of psychiatric disorders relative to healthy controls in several (Govorin and Vasil'eva, 2011; Zandi et al., 2011; Steiner et al., 2013), though not all included studies (Masdeu et al., 2012; Hammer et al., 2013), supports the possibility anticipated by Ezeoke et al. (2013) that the association between NMDAR antibodies and selected psychiatric disorders can be observed independent of medication effects. Second, we found that NMDAR antibody seropositive participants with first-episode or acute schizophrenia or schizoaffective disorder, bipolar or major depressive disorders had various combinations of IgG, IgM, and/or IgA class NMDAR antibodies against NR1, NR1/NR2B, and/or NR2A/NR2B subunits. With the exception of 3% (4 of 132) of participants with bipolar or major depressive disorders from one study ( Hammer et al., 2013), none of these participants had IgG class NMDAR antibodies against the NR1 subunit alone. This suggests that our finding of increased odds of NMDAR antibody seropositivity cannot be attributed to the inadvertent inclusion of undiagnosed anti-NMDAR encephalitis cases in psychiatric groups. Third, both studies reporting longitudinal data on serum titers of NMDAR antibodies based on ELISAs documented significant reductions in antibody titers over time (Govorin and Vasil'eva, 2011; Dickerson et al., 2013). These findings underscore the possibility that a subset of seronegative participants may have previously been seropositive during first-episode or acute disease states, as previously suggested (Lennox et al., 2009; Zandi et al., 2011; Haussleiter et al., 2012). Fourth, although there is no optimal dilution cut-off point with established sensitivity and specificity for distinguishing individuals with psychiatric disorders from healthy controls based on NMDAR antibody seropositivity, results of the two studies contributing data to the dilution-stratified subgroup analysis suggest that 1:10 is a sub-optimal standard in this regard (Hammer et al., 2013; Steiner et al., 2013). This underscores the need for future investigations aimed at determining the optimal dilution cut-off point in these populations, parallel to what has been done in studies of acute cortical ischemia and systemic lupus erythematosus, which used a receiver–operator characteristic curve (Dambinova et al., 2003) and standardized the cut-off based on a given number of standard deviations above the average controls' titers (Arinuma et al., 2008), respectively. Several factors might contribute to the association between serum NMDAR antibodies and schizophrenia, schizoaffective disorder, bipolar disorder and major depressive disorder. These include (a) genetic predisposition to autoimmunity, (b) molecular mimicry secondary to infection, and (c) hyperglutamatergic-dependent, NMDAR-mediated, neuronal excitotoxicity with secondary intrathecal NMDAR antibody synthesis. Regarding genetic predisposition to autoimmunity, in 2009, three independent genome-wide association studies identified a common risk variant for schizophrenia and bipolar disorder on chromosome 6.22p1, a segment containing

genes that encode the major histocompatibility complex (MHC) class II proteins involved in human leukocyte antigen (HLA) presentation (International Schizophrenia et al., 2009; Shi et al., 2009; Stefansson et al., 2009). In 2013, a fourth independent genomewide association study, which addressed several potential limitations concerning the earlier three studies, replicated these findings (Walters et al., 2013). Epidemiological data demonstrate robust personal and familial associations between psychotic as well as mood disorders and autoimmune diseases (Gilvarry et al., 1996; Eaton et al., 2006; Eaton et al., 2010; Benros et al., 2011; Chen et al., 2012; Benros et al., 2013; Benros et al., 2014a; Benros et al., 2014b). Hammer et al. (2013) specifically assessed the relationship between MHC class II mutations and NMDAR antibody seropositivity, finding a nominally significant association with a single nucleotide polymorphism in the HLA-A03 gene (P = 0.01). However, the authors found a mutation reaching genome-wide significance in rs524991 (OR, 2.22; 95% CI, 1.65–3.00; P b 6.1 ×10− 8). This gene resides 218.59 kb away from the transcription factor nuclear factor I subtype A, which is putatively implicated in neuroprotection and neuronal plasticity following NMDAR activation (Zheng et al., 2010), but not in autoimmunity per se. Regarding molecular mimicry secondary to infection, epidemiological studies provide consistent evidence of an association between psychotic as well as mood disorders and prior history of infection (Yolken and Torrey, 1995; Torrey et al., 1997). Along these lines, studies have also implicated several specific pathogens, including herpes simplex virus type II (Brown et al., 2005; Mortensen et al., 2007), influenza (Brown and Derkits, 2010) and toxoplasma gondii (Buka et al., 2001; Torrey et al., 2007; Buka et al., 2008; Dickerson et al., 2014). Herpes simplex virus has also recently been linked to anti-NMDAR encephalitis, though the direction of this association is unclear (Prüss et al., 2012; Armangue et al., 2014). Hammer et al. (2013) reported a significant association between NMDAR antibody seropositivity and prior exposure to influenza among participants with schizophrenia or schizoaffective disorder. Hyperglutamatergic-dependent NMDAR-mediated excitotoxicity with secondary intrathecal NMDAR antibody synthesis may represent another contributing factor to NMDAR antibody mediated neuropsychiatric sequelae. In acute cortical ischemia, NMDAR-mediated excitotoxicity is thought to induce serine-protease cleavage and release of NMDAR subunit peptide fragments into the blood, where they are recognized as foreign antigens, prompting autoantibody synthesis and opsonization (Dambinova et al., 2003; Bokesch et al., 2006; Dambinova et al., 2012). Although mood disorders are also known to be associated with hyperglutamatergia (Najjar et al., 2013a), the relevance of this mechanism to NMDAR antibody seropositivity in mood disorders is unknown. The functional significance of serum NMDAR antibodies found in psychiatric disorders, from directly pathogenic to biomarkers of concomitant autoimmunity to mere epiphenomena, remains to be clarified. Emerging evidence links blood–brain barrier dysfunction and hyperpermeability to primary psychotic and mood disorders (Shalev et al., 2009; Bechter et al., 2010; Najjar et al., 2013a; Najjar et al., 2013b). An independent line of evidence suggests that increased blood–brain barrier permeability is necessary for NMDAR antibodyinduced neuropsychiatric symptomatology (Arinuma et al., 2008; Lauvsnes and Omdal, 2012; Hammer et al., 2013; Yoshio et al., 2013). IgG NMDAR antibodies against NR2A/NR2B subunits in cerebrospinal fluid are substantially more predictive of neuropsychiatric systemic lupus erythematosus than are those in serum (Lauvsnes and Omdal, 2012). A recent study showed that IgG class antibodies against NR2A/ NR2B subunits obtained from cerebrospinal fluid of individuals with neuropsychiatric systemic lupus erythematosus can promote inflammatory and vascular abnormalities that increase blood–brain barrier permeability by cross-reacting with endothelial double-stranded DNA in a dose-dependent manner (Yoshio et al., 2013). These inflammatory and vascular abnormalities included increased endothelial expression of

Please cite this article as: Pearlman, D.M., Najjar, S., Meta-analysis of the association between N-methyl-D-aspartate receptor antibodies and schizophrenia, schizoaffective disorder, bipolar disorder, and ..., Schizophr. Res. (2014), http://dx.doi.org/10.1016/j.schres.2014.05.001

D.M. Pearlman, S. Najjar / Schizophrenia Research xxx (2014) xxx–xxx

endothelial leukocyte adhesion molecule 1, vascular cell adhesion molecule 1 and intercellular adhesion molecule 1, and increased endothelial production of interleukin-6 (IL-6) and IL-8 (but not tumor necrosis factor α or IL-1β), in association with NF-κB activation. Although many of these same abnormalities have been described in primary psychotic and mood disorders (Najjar et al., 2013a), the effects of NMDAR antibodies on blood–brain barrier integrity in these disorders remain unknown. Hammer et al. (2013) found a significant positive association between risk factors for blood–brain barrier damage (history of neurotrauma and birth complications) and severity of neuropsychiatric features in NMDAR antibody seropositive participants with schizophrenia or schizoaffective disorder. Whereas the majority of evidence for the functional relevance of NMDAR antibodies has focused on the now well-established pathogenic role of IgG class antibodies in anti-NMDAR encephalitis, the pathogenicity of IgA (Pruss et al., 2012a; Hammer et al., 2013) and IgM (Choe et al., 2013) class NMDAR antibodies has also been demonstrated in vitro and in vivo. Prüss and colleagues showed that in a cohort of seven patients with cognitive impairment and reduced metabolic brain activity, elevated serum IgA class (cerebrospinal fluid in three of seven), but not IgG class, NMDAR antibodies substantially decreased the expression of NMDARs and synaptophysin, and inhibited NMDARmediated currents in a dose-dependent manner in vitro (via NR1HEK cell-based assay; Pruss et al., 2012b). Further, these immunologic and clinical abnormalities were reversed after removing patients' serum from the culture media and initiating immunotherapy plus plasma exchange (Pruss et al., 2012b). Several authors have also hypothesized that NMDAR antibodies might exert differential functional effects according to the underlying neurobiology of psychiatric disorders (Lapteva et al., 2006; Gono et al., 2011). That is, given that NMDAR hypofunction is implicated in schizophrenia and that NMDAR hyperfunction is implicated in major depressive disorder (Najjar et al., 2013a), it is conceivable that NMDAR antibodies might exacerbate the neuropsychiatric symptoms of schizophrenia by augmenting NMDAR hypofunction, while ameliorating those of mood disorders. This is analogous to either pro-psychotic or antidepressant effects of NMDAR antagonists, such as ketamine, MK-801, or phencyclidine. Our study has several important limitations. First, included studies largely provided data on clinical and demographic characteristics at the group rather than the individual level of analysis. As such, we were unable to perform regression analyses aimed at assessing the potential for confounding associated with these characteristics. Second, only four of the included studies explicitly reported blind outcome assessment of case/control status. Third, heterogeneous methods in the included studies might have influenced the pretest odds of NMDAR antibody seropositivity. Fourth, variability in the combinations of immunoglobulin classes and NMDAR subunits assessed across studies increases the risk of detection bias. One study did not detect any NMDAR antibody seropositivity among participants with first-episode schizophrenia, specifically assayed for IgG class NMDAR antibodies against the NR1 subunit, but not for other immunoglobulin classes or NMDAR subunits (Masdeu et al., 2012). Fifth, none of the studies assayed antibody seropositivity against NR2A or NR2B subunits alone. ELISA methods that can distinguish between these subunits were recently described (Gono et al., 2011). Whereas the more commonly assayed peptide DWEYSVWLSN does not disambiguate between these subunits, the N-terminal domain sequences DWDYS and DEWDY are specific to NR2A and NR2B, respectively (Gono et al., 2011). Future studies should do so, not only to minimize the risk of detection bias, but also because NR2A and NR2B subunits have distinctive physiologic (von Engelhardt et al., 2009; Paoletti et al., 2013) and pathologic (Li et al., 2010; Gono et al., 2011) roles. Sixth, the estimates of NMDAR antibody seropositivity we report are based on a combination of different immunoassay techniques that are characterized by differences in their inherent reliability (e.g., results from cell-based assays are generally

9

considered more valid than those from ELISAs). Last, the possibility that other methodological factors not reported might have significantly influenced the sensitivity and specificity of immunostaining assays used across studies cannot be excluded. In conclusion, we found that individuals with schizophrenia or schizoaffective, bipolar, or major depressive disorders are collectively about three times more likely to have elevated NMDAR antibody titers compared with healthy controls based on high-specificity, but not low-specificity, seropositivity thresholds, though considerable methodological and statistical heterogeneity exists. Evidence concerning the effect of disease state and time of serum acquisition is varied and consistent, repectively. Adequately powered longitudinal studies employing standardized assay methods and seropositivity threshold definitions, and quantifying NMDAR antibodies in both sera and cerebrospinal fluid are needed to further elucidate the clinical and pathophysiological implications of this association. Role of the funding source None. Contributions Both authors contributed substantially to the conception and design of the study, data collection, and analyses and interpretation of data. Both authors revised the paper critically for important intellectual content and gave final approval of the version to be published. Conflicts of interest Both authors declare that we have nothing to disclose. Acknowledgments None.

Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.schres.2014.05.001. References Arinuma, Y., Yanagida, T., Hirohata, S., 2008. Association of cerebrospinal fluid anti-NR2 glutamate receptor antibodies with diffuse neuropsychiatric systemic lupus erythematosus. Arthritis Rheum. 58 (4), 1130–1135. Armangue, T., Leypoldt, F., Malaga, I., Raspall-Chaure, M., Marti, I., Nichter, C., et al., 2014. Herpes simplex virus encephalitis is a trigger of brain autoimmunity. Ann. Neurol. 75 (2), 317–323. Bechter, K., Reiber, H., Herzog, S., Fuchs, D., Tumani, H., Maxeiner, H.G., 2010. Cerebrospinal fluid analysis in affective and schizophrenic spectrum disorders: identification of subgroups with immune responses and blood-CSF barrier dysfunction. J. Psychiatr. Res. 44 (5), 321–330. Benros, M.E., Nielsen, P.R., Nordentoft, M., Eaton, W.W., Dalton, S.O., Mortensen, P.B., 2011. Autoimmune diseases and severe infections as risk factors for schizophrenia: a 30-year population-based register study. Am. J. Psychiatry 168 (12), 1303–1310. Benros, M.E., Waltoft, B.L., Nordentoft, M., Ostergaard, S.D., Eaton, W.W., Krogh, J., et al., 2013. Autoimmune diseases and severe infections as risk factors for mood disorders: a nationwide study. JAMA Psychiatry 70 (8), 812–820. Benros, M.E., Eaton, W.W., Mortensen, P.B., 2014a. The epidemiologic evidence linking autoimmune diseases and psychosis. Biol. Psychiatry 75 (4), 300–306. Benros, M.E., Pedersen, M.G., Rasmussen, H., Eaton, W.W., Nordentoft, M., Mortensen, P.B., 2014b. A nationwide study on the risk of autoimmune diseases in individuals with a personal or a family history of schizophrenia and related psychosis. Am. J. Psychiatry 171 (2), 218–226. Bokesch, P.M., Izykenova, G.A., Justice, J.B., Easley, K.A., Dambinova, S.A., 2006. NMDA receptor antibodies predict adverse neurological outcome after cardiac surgery in highrisk patients. Stroke 37 (6), 1432–1436. Brown, A.S., Derkits, E.J., 2010. Prenatal infection and schizophrenia: a review of epidemiologic and translational studies. Am. J. Psychiatry 167 (3), 261–280. Brown, A.S., Schaefer, C.A., Quesenberry Jr., C.P., Liu, L., Babulas, V.P., Susser, E.S., 2005. Maternal exposure to toxoplasmosis and risk of schizophrenia in adult offspring. Am. J. Psychiatry 162 (4), 767–773. Buka, S.L., Tsuang, M.T., Torrey, E.F., Klebanoff, M.A., Bernstein, D., Yolken, R.H., 2001. Maternal infections and subsequent psychosis among offspring. Arch. Gen. Psychiatry 58 (11), 1032–1037. Buka, S.L., Cannon, T.D., Torrey, E.F., Yolken, R.H., Collaborative Study Group on the Perinatal Origins of Severe Psychiatric, D, 2008. Maternal exposure to herpes simplex virus and risk of psychosis among adult offspring. Biol. Psychiatry 63 (8), 809–815. Chen, S.J., Chao, Y.L., Chen, C.Y., Chang, C.M., Wu, E.C., Wu, C.S., et al., 2012. Prevalence of autoimmune diseases in in-patients with schizophrenia: nationwide populationbased study. Br. J. Psychiatry 200 (5), 374–380.

Please cite this article as: Pearlman, D.M., Najjar, S., Meta-analysis of the association between N-methyl-D-aspartate receptor antibodies and schizophrenia, schizoaffective disorder, bipolar disorder, and ..., Schizophr. Res. (2014), http://dx.doi.org/10.1016/j.schres.2014.05.001

10

D.M. Pearlman, S. Najjar / Schizophrenia Research xxx (2014) xxx–xxx

Choe, C.U., Karamatskos, E., Schattling, B., Leypoldt, F., Liuzzi, G., Gerloff, C., et al., 2013. A clinical and neurobiological case of IgM NMDA receptor antibody associated encephalitis mimicking bipolar disorder. Psychiatry Res. 208 (2), 194–196. Dalmau, J., Gleichman, A.J., Hughes, E.G., Rossi, J.E., Peng, X., Lai, M., et al., 2008. AntiNMDA-receptor encephalitis: case series and analysis of the effects of antibodies. Lancet Neurol. 7 (12), 1091–1098. Dambinova, S.A., Khounteev, G.A., Izykenova, G.A., Zavolokov, I.G., Ilyukhina, A.Y., Skoromets, A.A., 2003. Blood test detecting autoantibodies to N-methyl-D-aspartate neuroreceptors for evaluation of patients with transient ischemic attack and stroke. Clin. Chem. 49 (10), 1752–1762. Dambinova, S.A., Bettermann, K., Glynn, T., Tews, M., Olson, D., Weissman, J.D., et al., 2012. Diagnostic potential of the NMDA receptor peptide assay for acute ischemic stroke. PLoS One 7 (7), e42362. DerSimonian, R., Laird, N., 1986. Meta-analysis in clinical trials. Control. Clin. Trials 7 (3), 177–188. Dickerson, F., Stallings, C., Vaughan, C., Origoni, A., Khushalani, S., Yolken, R., 2012. Antibodies to the glutamate receptor in mania. Bipolar Disord. 14 (5), 547–553. Dickerson, F., Stallings, C., Origoni, A., Vaughan, C., Katsafanas, E., Khushalani, S., et al., 2013. A combined marker of inflammation in individuals with mania. PLoS One 8 (9), e73520. Dickerson, F., Stallings, C., Origoni, A., Vaughan, C., Katsafanas, E., Khushalani, S., et al., 2014. Antibodies to Toxoplasma gondii in individuals with mania. Bipolar Disord. 16 (2), 129–136. Eaton, W.W., Byrne, M., Ewald, H., Mors, O., Chen, C.Y., Agerbo, E., et al., 2006. Association of schizophrenia and autoimmune diseases: linkage of Danish national registers. Am. J. Psychiatry 163 (3), 521–528. Eaton, W.W., Pedersen, M.G., Nielsen, P.R., Mortensen, P.B., 2010. Autoimmune diseases, bipolar disorder, and non-affective psychosis. Bipolar Disord. 12 (6), 638–646. Ezeoke, A., Mellor, A., Buckley, P., Miller, B., 2013. A systematic, quantitative review of blood autoantibodies in schizophrenia. Schizophr. Res. 150 (1), 245–251. Faust, T.W., Chang, E.H., Kowal, C., Berlin, R., Gazaryan, I.G., Bertini, E., et al., 2010. Neurotoxic lupus autoantibodies alter brain function through two distinct mechanisms. Proc. Natl. Acad. Sci. U. S. A. 107 (43), 18569–18574. Gilvarry, C.M., Sham, P.C., Jones, P.B., Cannon, M., Wright, P., Lewis, S.W., et al., 1996. Family history of autoimmune diseases in psychosis. Schizophr. Res. 19 (1), 33–40. Gleichman, A.J., Spruce, L.A., Dalmau, J., Seeholzer, S.H., Lynch, D.R., 2012. Anti-NMDA receptor encephalitis antibody binding is dependent on amino acid identity of a small region within the GluN1 amino terminal domain. J. Neurosci. 32 (32), 11082–11094. Gono, T., Kawaguchi, Y., Kaneko, H., Nishimura, K., Hanaoka, M., Kataoka, S., et al., 2011. Anti-NR2A antibody as a predictor for neuropsychiatric systemic lupus erythematosus. Rheumatology (Oxford) 50 (9), 1578–1585. Govorin, N.V., Vasil'eva, A.I., 2011. Effect of haloperidol and risperidone on neuromarkers and indices of endothelial dysfunction in patients with acute schizophrenia. Zh. Nevropatol. Psikhiatr. Im. S. S. Korsakova 111 (3), 54–57. Hammer, C., Stepniak, B., Schneider, A., Papiol, S., Tantra, M., Begemann, M., et al., 2013 Sep 3. Neuropsychiatric disease relevance of circulating anti-NMDA receptor autoantibodies depends on blood-brain barrier integrity. Mol. Psychiatry. http://dx.doi.org/ 10.1038/mp.2013.110 [Epub ahead of print] PubMed PMID: 23999527. Haussleiter, I.S., Emons, B., Schaub, M., Borowski, K., Brune, M., Wandinger, K.P., et al., 2012. Investigation of antibodies against synaptic proteins in a cross-sectional cohort of psychotic patients. Schizophr. Res. 140 (1–3), 258–259. Higgins, J.P., Thompson, S.G., Deeks, J.J., Altman, D.G., 2003. Measuring inconsistency in meta-analyses. BMJ 327 (7414), 557–560. Hughes, E.G., Peng, X., Gleichman, A.J., Lai, M., Zhou, L., Tsou, R., et al., 2010. Cellular and synaptic mechanisms of anti-NMDA receptor encephalitis. J. Neurosci. 30 (17), 5866–5875. International Schizophrenia, C., Purcell, S.M., Wray, N.R., Stone, J.L., Visscher, P.M., O'Donovan, M.C., et al., 2009. Common polygenic variation contributes to risk of schizophrenia and bipolar disorder. Nature 460 (7256), 748–752. Irani, S.R., Bera, K., Waters, P., Zuliani, L., Maxwell, S., Zandi, M.S., et al., 2010. N-methyl-Daspartate antibody encephalitis: temporal progression of clinical and paraclinical observations in a predominantly non-paraneoplastic disorder of both sexes. Brain 133 (Pt 6), 1655–1667. Jentsch, J.D., Roth, R.H., 1999. The neuropsychopharmacology of phencyclidine: from NMDA receptor hypofunction to the dopamine hypothesis of schizophrenia. Neuropsychopharmacology 20 (3), 201–225. Kayser, M.S., Titulaer, M.J., Gresa-Arribas, N., Dalmau, J., 2013. Frequency and characteristics of isolated psychiatric episodes in anti-N-methyl-D-aspartate receptor encephalitis. JAMA Neurol. 70 (9), 1133–1139. Lapteva, L., Nowak, M., Yarboro, C.H., Takada, K., Roebuck-Spencer, T., Weickert, T., et al., 2006. Anti-N-methyl-D-aspartate receptor antibodies, cognitive dysfunction, and depression in systemic lupus erythematosus. Arthritis Rheum. 54 (8), 2505–2514. Lauvsnes, M.B., Omdal, R., 2012. Systemic lupus erythematosus, the brain, and anti-NR2 antibodies. J. Neurol. 259 (4), 622–629. Lauvsnes, M.B., Maroni, S.S., Appenzeller, S., Beyer, M.K., Greve, O.J., Kvaloy, J.T., et al., 2013. Memory dysfunction in primary Sjogren's syndrome is associated with antiNR2 antibodies. Arthritis Rheum. 65 (12), 3209–3217. Lennox, B., Zandi, M., Irani, S., Coles, A.J., Vincent, A., 2009. Anti-GAD65, NMDAR and VGKC antibodies detected in first episode psychosis. Schizophr. Bull. 35 (suppl 1), 3. Li, N., Lee, B., Liu, R.J., Banasr, M., Dwyer, J.M., Iwata, M., et al., 2010. mTOR-dependent synapse formation underlies the rapid antidepressant effects of NMDA antagonists. Science 329 (5994), 959–964.

Masdeu, J.C., Gonzalez-Pinto, A., Matute, C., Ruiz De Azua, S., Palomino, A., De Leon, J., et al. , 2012. Serum IgG antibodies against the NR1 subunit of the NMDA receptor not detected in schizophrenia. Am. J. Psychiatry 169 (10), 1120–1121. Mortensen, P.B., Norgaard-Pedersen, B., Waltoft, B.L., Sorensen, T.L., Hougaard, D., Torrey, E.F., et al., 2007. Toxoplasma gondii as a risk factor for early-onset schizophrenia: analysis of filter paper blood samples obtained at birth. Biol. Psychiatry 61 (5), 688–693. Najjar, S., Pearlman, D.M., Alper, K., Najjar, A., Devinsky, O., 2013a. Neuroinflammation and psychiatric illness. J. Neuroinflammation 10, 43. Najjar, S., Pearlman, D.M., Hirsch, S., Friedman, K., Strange, J., Reidy, J., et al., 2013 Sep 13b. Brain biopsy findings link major depressive disorder to neuroinflammation, oxidative stress, and neurovascular dysfunction: A case report. Biol. Psychiatry pii: S00063223(13)00741-5. http://dx.doi.org/10.1016j.biopsych.2013.07.041 [Epub ahead of print] PubMed PMID: 24075735. Olney, J.W., Farber, N.B., 1995. Glutamate receptor dysfunction and schizophrenia. Arch. Gen. Psychiatry 52 (12), 998–1007. Omdal, R., Brokstad, K., Waterloo, K., Koldingsnes, W., Jonsson, R., Mellgren, S.I., 2005. Neuropsychiatric disturbances in SLE are associated with antibodies against NMDA receptors. Eur. J. Neurol. 12 (5), 392–398. Paoletti, P., Bellone, C., Zhou, Q., 2013. NMDA receptor subunit diversity: impact on receptor properties, synaptic plasticity and disease. Nat. Rev. Neurosci. 14 (6), 383–400. Pollak, T.A., McCormack, R., Peakman, M., Nicholson, T.R., David, A.S., 2013. Prevalence of anti-N-methyl-D-aspartate (NMDA) antibodies in patients with schizophrenia and related psychoses: a systematic review and meta-analysis. Psychol. Med. 1–13. Prüss, H., Finke, C., Holtje, M., Hofmann, J., Klingbeil, C., Probst, C., et al., 2012. N-methyl-Daspartate receptor antibodies in herpes simplex encephalitis. Ann. Neurol. 72 (6), 902–911. Pruss, H., Holtje, M., Maier, N., Gomez, A., Buchert, R., Harms, L., et al., 2012a. IgA NMDA receptor antibodies are markers of synaptic immunity in slow cognitive impairment. Neurology 78 (22), 1743–1753. Pruss, H., Holtje, M., Maier, N., Gomez, A., Buchert, R., Harms, L., et al., 2012b. IgA NMDA receptor antibodies are markers of synaptic immunity in slow cognitive impairment. Neurology 1743–1753 (United States). Rhoads, J., Guirgis, H., McKnight, C., Duchemin, A.M., 2011. Lack of anti-NMDA receptor autoantibodies in the serum of subjects with schizophrenia. Schizophr. Res. 129 (2–3), 213–214. Schwartz, M., Kormilachev, M., Kushnir, M., Weller, B., Rochas, M., Toubi, E., 2009. Lupus anticoagulant and anticardiolipin antibodies in serum of patients treated with risperidone. J. Clin. Psychiatry 70 (5), 769–771. Shalev, H., Serlin, Y., Friedman, A., 2009. Breaching the blood–brain barrier as a gate to psychiatric disorder. Cardiovasc. Psychiatry Neurol. 2009, 278531. Shen, H., Li, R., Xiao, H., Zhou, Q., Cui, Q., Chen, J., 2009. Higher serum clozapine level is associated with increased antiphospholipid antibodies in schizophrenia patients. J. Psychiatr. Res. 43 (6), 615–619. Shi, J., Levinson, D.F., Duan, J., Sanders, A.R., Zheng, Y., Pe'er, I., et al., 2009. Common variants on chromosome 6p22.1 are associated with schizophrenia. Nature 460 (7256), 753–757. Stefansson, H., Ophoff, R.A., Steinberg, S., Andreassen, O.A., Cichon, S., Rujescu, D., et al., 2009. Common variants conferring risk of schizophrenia. Nature 460 (7256), 744–747. Steiner, J., Walter, M., Glanz, W., Sarnyai, Z., Bernstein, H.G., Vielhaber, S., et al., 2013. Increased prevalence of diverse N-methyl-D-aspartate glutamate receptor antibodies in patients with an initial diagnosis of schizophrenia: specific relevance of IgG NR1a antibodies for distinction from N-methyl-D-aspartate glutamate receptor encephalitis. JAMA Psychiatry 70 (3), 271–278. Stroup, D.F., Berlin, J.A., Morton, S.C., Olkin, I., Williamson, G.D., Rennie, D., et al., 2000. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA 283 (15), 2008–2012. Torrey, E.F., Miller, J., Rawlings, R., Yolken, R.H., 1997. Seasonality of births in schizophrenia and bipolar disorder: a review of the literature. Schizophr. Res. 28 (1), 1–38. Torrey, E.F., Bartko, J.J., Lun, Z.R., Yolken, R.H., 2007. Antibodies to Toxoplasma gondii in patients with schizophrenia: a meta-analysis. Schizophr. Bull. 33 (3), 729–736. Tsutsui, K., Kanbayashi, T., Tanaka, K., Boku, S., Ito, W., Tokunaga, J., et al., 2012. AntiNMDA-receptor antibody detected in encephalitis, schizophrenia, and narcolepsy with psychotic features. BMC Psychiatry 12, 37. von Engelhardt, J., Doganci, B., Seeburg, P.H., Monyer, H., 2009. Synaptic NR2A — but not NR2B-Containing NMDA receptors increase with blockade of ionotropic glutamate receptors. Front. Mol. Neurosci. 2, 19. Walters, J.T., Rujescu, D., Franke, B., Giegling, I., Vasquez, A.A., Hargreaves, A., et al., 2013. The role of the major histocompatibility complex region in cognition and brain structure: a schizophrenia GWAS follow-up. Am. J. Psychiatry 170 (8), 877–885. Yolken, R.H., Torrey, E.F., 1995. Viruses, schizophrenia, and bipolar disorder. Clin. Microbiol. Rev. 8 (1), 131–145. Yoshio, T., Okamoto, H., Hirohata, S., Minota, S., 2013. IgG anti-NR2 glutamate receptor autoantibodies from patients with systemic lupus erythematosus activate endothelial cells. Arthritis Rheum. 65 (2), 457–463. Zandi, M.S., Irani, S.R., Lang, B., Waters, P., Jones, P.B., McKenna, P., et al., 2011. Diseaserelevant autoantibodies in first episode schizophrenia. J. Neurol. 258 (4), 686–688. Zheng, S., Eacker, S.M., Hong, S.J., Gronostajski, R.M., Dawson, T.M., Dawson, V.L., 2010. NMDA-induced neuronal survival is mediated through nuclear factor I-A in mice. J. Clin. Invest. 120 (7), 2446–2456.

Please cite this article as: Pearlman, D.M., Najjar, S., Meta-analysis of the association between N-methyl-D-aspartate receptor antibodies and schizophrenia, schizoaffective disorder, bipolar disorder, and ..., Schizophr. Res. (2014), http://dx.doi.org/10.1016/j.schres.2014.05.001