Introduction
Schizophrenia is a heterogeneous and complex psychiatric disorder affecting up to
1% of the population worldwide (1). Although the precise development of schizophrenia
is not yet fully understood, it is now admitted to be underpinned by the entanglement
of genetic, environmental, and immuno-inflammatory factors. Among schizophrenic patients,
it is assumed that at least 30% will not respond to conventional antipsychotics (2).
These data underlie the importance of precision medicine in psychiatry, in other words,
the need to identify subgroups of patients with specific signatures who will benefit
from treatment targeting these specific biological pathways. Reviving an area of exploration
older than a century, recent and abundant literature emphasized the importance of
the immune system in the pathophysiology of schizophrenia [for review, see Ref. (3)].
In psychiatry, the link between psychotic disorders, particularly schizophrenia, and
immune system deregulations, including autoimmunity, is an old concept that regained
strong support; thanks to the better characterization of brain inflammation-induced
psychotic symptoms and autoimmune encephalitis (3). Moreover, recent epidemiological
studies evidenced a high prevalence of multiple autoimmune diseases in schizophrenic
patients (4). In a recent meta-analysis, autoantibodies against neuronal receptors
have indeed been identified in the circulation of patients with neuropsychiatric disorders,
constituting, today, one of the hottest topics in psychiatry (5–10). This new era
fosters debate on (i) how to explain the increased burden of autoimmunity in schizophrenia,
(ii) what could be the precise target(s) and the pathogenic implication(s) of the
autoantibodies on the disease onset and development, (iii) how to define patient subgroups
carrying such autoantibodies to facilitate their diagnosis, and (iv) how should we
treat these patients using appropriate protocols such as immunotherapy (i.e., corticotherapy
or plasmapheresis). Several neurological autoimmune diseases are, for instance, efficiently
treated once autoantibodies against neurotransmitter receptors and ion channels have
been identified (11, 12). The discovery of the autoimmune encephalitis due to anti-N-methyl-d-aspartic
acid receptor (NMDAR) has greatly revived the relationship between autoimmunity and
psychosis. Indeed, directed against the NMDAR N-methyl-d-aspartate receptors antibodies
(NMDAR-Ab), the autoantibodies are directly responsible for the psychotic symptoms
and catatonia, followed by profound neurologic deterioration (13, 14). In patients
with schizophrenia, the prevalence and clinical significance of circulating NMDAR-Ab
remains controversial with detection prevalence rates varying considerably between
studies (15–27). Inspite of such imprecisions, defining and isolating seropositive
patients, suffering from “autoimmune psychosis,” is a major challenge for appropriate
treatments. In this review, we focus our attention on the potential elements possibly
helping to define an “autoimmune psychosis” subgroup of schizophrenic patients. Furthermore,
we outline some of the specific clinical presentation of these patients that will
be of great importance to optimize the diagnostic and subsequent therapies.
Autoimmunity and Psychosis: Roots
Autoimmune disorders occur after the failure of self-recognition processes with consequent
production of pathogenic autoantibodies directed against specific or multiple organs.
They are heterogeneous disorders, representing more than 80 different diseases. Several
risk factors contribute to the high prevalence of autoimmunity including genetic and
environmental ones and their interplay. Within this context, the “immunogenetic” contribution
is largely dominated by the major histocompatibility complex (MHC) genetic diversity
and, at a lesser extent, by mutational events affecting cytokines encoding genes (28).
On the other side, environmental stressors are also of major importance for the onset
of autoimmunity. For example, infections during pregnancy or in childhood are associated
with an increased risk of type 1 diabetes (29). Moreover, and within the GxE context,
it has been postulated that the risk of autoimmunity is enhanced through the perturbation
of gut microbiota or dysbiosis (30, 31). This dysbiosis seems to be the origin of
the emergence of different autoantibodies, even if the exact mechanisms involved are
still under debate (32).
In psychosis settings, similar epidemiological associations with early infections,
autoimmune disorders, and dysbiosis have pinpointed the possible existence of an autoimmune
psychosis subgroup in schizophrenia. Maternal exposure to influenza or toxoplasmosis
during pregnancy has been associated with schizophrenia. Childhood autoimmune diseases
as well as inflammatory diseases, such as asthma, are known to be associated with
an increased number of psychotic experiences in adolescence but also with an increased
incidence of schizophrenia in the adulthood. Moreover, in patients with autoimmune
conditions, the risk to develop schizophrenia increases linearly with the number of
severe infectious episodes (4). The other way around, patients with schizophrenia
and their first degree relatives, also exhibit a higher prevalence for autoimmune
disorders (33). Last, associations between autoimmunity, gastrointestinal symptoms,
and dysbiosis are starting to emerge (34). These data, along with the strong association
between the interindividual immunogenetic background and the whole array of brain
and peripheral autoantibodies, in at least a subgroup of schizophrenic patients, led
us to propose the concept of “autoimmune psychosis.” Accordingly, our goal is to review
the evocative characteristics that should prompt the search of autoantibodies in front
of a patient, in particular, in cases of first episode, resistant ones, or schizophrenia
with neurological comorbidity.
Biological and Clinical Features of Patients with Autoimmune Psychosis
Autoimmune Psychosis: Genetic and Environmental Risk Factors
Several genome-wide association studies (GWAS) confirmed an association between the
MHC region (chromosome 6) and psychosis (35, 36). Moreover, a recent landmark GWAS
analysis produced by the largest consortium on genetics of schizophrenia has shown
that, like in autoimmune disorders, the MHC region was the most strongly associated
(best p-value: MHC-region: p = 3.86e−32; 36,989 cases and 113,075 controls) (37, 38).
The consequences of these mutations are still to be fully understood because some
of them are found in non-coding region. However, in a matter of interest, the MHC
region include the human leukocyte antigen (HLA) cluster, which is the most polymorphic
and gene-dense genomic part of the human genome (39) encompassing more than 250 genes
(4 Mb) and 14,000 alleles as reported to date (IMGT/HLA database; http://www.ebi.ac.uk/imgt/hla).
Governing the specific adaptive immune responses, the HLA molecules were widely explored
in disease-association studies (40) especially concerning those classified as autoimmune
disorders (40–42). Even if more studies are needed to understand the link between
immunogenetic and psychosis, disentangling such diversity might help to delineate
the concept of autoimmune psychosis, at least on a genetic point of view.
On the other side, although data on gene–environment interactions are scarce, several
environmental risk factors have been associated with schizophrenia and would be worth
testing with MHC/HLA haplotypes. In particular, the occurrence of infections by pathogens
such as, influenza, herpes simplex type 2, cytomegalovirus, and Toxoplasma gondii
and/or increased C-reactive protein plasma levels during pregnancy are known to be
associated with an increased risk of developing schizophrenia in adulthood (43–45).
In the same context, hospitalization for infection increased the risk of schizophrenia
by 60%, and there is a dose–response relationship between the number of hospital contact
with infection and psychosis (46). Altogether, the reported deep intricacies between
infection and autoimmunity, either under an additive or a more complex framework,
with a consequent risk of psychosis, reinforce the concept of autoimmune psychosis
concept (46).
More than separated risk factors, actual studies argue for a complex interaction between
genetic risk factors conferring susceptibility to environmental injuries. For example,
polymorphisms of the innate system genes, like IL1B, IL6, TNF alpha, or interferon,
will lead to a bigger release of pro-inflammatory cytokines in response to environmental
stressors (47).
In summary, immunogenetic dissection especially of the MCH/HLA region according to
the natural history of deleterious immune processes including early infection and/or
autoimmune features might be a promising route to better understand the interactions
between gene and environment.
Autoimmune Psychosis: Peripheral Biomarkers from Dysbiosis to Autoantibodies?
Similar to autoimmunity, dysbiosis is found in patients with psychosis (48). The intestinal
microbiota seems essential for the development and functioning of the nervous central
system, shedding light on the concept of a gut–brain axis (49). Dysbiosis is a well-known
cause of increased intestinal permeability (so-called “leaky gut”) in schizophrenia
(50). This increased intestinal permeability is demonstrated by the high circulating
levels of CD14, a biomarker of bacterial translocation (49). The release of such pro-inflammatory
innate sensor in a repetitive manner could allow, under the framework of particular
genetic framework (HLA), to the breakdown of immune tolerance with consequent emergence
of autoantibodies. Along this line, various autoantibodies have been found in subgroups
of schizophrenic patients. For example, increased anti-bovine casein antibodies have
been found in psychosis (51). Meta-analysis found threefold to fourfold times more
anti-transglutaminase and anti-gliadine autoantibodies in patients with schizophrenia
than in general population. Autoantibodies, specifically against the central nervous
system, have also been found in schizophrenic patients. These patients have a higher
prevalence of circulating antibodies against hippocampus and hypothalamus as compared
to healthy control (52). A recent meta-analysis has confirmed and specified these
results, showing that schizophrenic patients are three times more likely to have high
levels of anti-glutamate receptor antibodies, N-methyl-d-aspartic acid receptor (NMDAR),
compared to controls (22). The latter being of major importance. For the first time,
they might make the bridge between autoimmune psychosis and the glutamate theory of
psychosis and, doing so, sheds light on the pathophysiology of autoimmune psychosis.
In summary, there is a whole array of peripheral and central autoantibodies in schizophrenia,
which deserve further exploration to explore their pathogenic role and to describe
possible associated clinico-biological signatures, helping to more precise the concept
of autoimmune psychosis.
Autoimmune Psychosis: Clinical Picture?
We have seen that, among the heterogeneous group of schizophrenic patients, it is
possible to hypothesize the existence of an autoimmune psychosis subgroup. The question
is now in front of which clinical history or symptoms should we search for autoimmunity
(53). The literature is still heterogeneous in the field and some have found no differences
between patients (17, 20, 23). However, based on a French cohort of patients with
psychiatric symptoms and autoantibodies against NMDA-R, we described clinical characteristics
of patients that should lead to search of biological markers of an autoimmune psychosis
[for details, see Ref. (54)]. While the mean age of onset in schizophrenia is 25–35 years
old, we observed the first episode of autoimmune psychosis to occur around the 24th
years of life (55). It is well known that schizophrenia is associated with the presence
of neurological soft signs (56). More than that, we have been able to put forward
that 50% of the patients with autoantibodies against NMDA-R had neurological symptoms
including headaches, disorientation, paresthesia, anterograde amnesia, or abnormal
movements. These results are in agreement with others who have also found neurological
comorbidities in autoimmune psychosis cases (26, 57). Catatonia is a complex neuropsychiatric
syndrome related to schizophrenia in 20% of cases (58). Its exact physiopathology
is still unknown but seems underpinned by a deregulation between glutamatergic and
GABAergic signaling (59). Catatonia, schizophrenia, and NMDAR-Ab have been extensively
associated in the literature, which might indicate catatonia as a sign of autoimmune
psychosis (60–62).
In summary, in front of an early age at onset of psychosis, discrete neurological
symptoms, and catatonia, search for autoantibodies should be performed.
Treatment Response
More than 30% of schizophrenic patients are resistant to conventional antipsychotics
(63). Among them, 41% exhibited biological signs of immune activation (64). For example,
treatment-resistant patient has been strongly associated with increased cytokines
level (65–69). It has also been reported the specific presence of NMDAR-Ab in treatment-resistant
patients (70). Finally, we have also been able to underlie the tight link between
presence of NMDAR Ab and neuroleptic intolerance (54). These data seem to indicate
a different pathophysiology, not related to the classical dopaminergic hypothesis,
in patients with an autoimmune psychosis.
All these arguments have led to propose that all treatment-resistant/intolerant patients
should have an autoantibodies screening, and particularly NMDAR-Ab, as a part of the
diagnostic process.
In summary, based on epidemiological studies, genetic and biological biomarkers but
also environmental risk factors, there are many arguments to suggest, that among schizophrenia,
it is useful to ensure the identification of a subgroup of autoimmune psychosis. It
is possibly characterized by (i) history of early infections or severe stress, (ii)
autoimmune or infections during childhood or early adulthood, (iii) clinical presentation
with the presence of gastrointestinal/neurological symptoms, catatonia, and (iv) presence
of one or several autoantibodies associated with schizophrenia leading to a resistant
form of schizophrenia (Figure 1).
Figure 1
Diagnostic elements possibly supporting the definition of autoimmune psychosis. The
diagnosis is suspected if the patient presents psychiatric symptoms (e.g., early age
of onset, catatonia), neurological signs, resistant or intolerance to antipsychotic
treatment, history of autoimmune disorder, and severe infections. The diagnosis relies
on the detection in the circulation of autoantibodies and, particularly, directed
against neurotransmitter receptors, such as the glutamate NMDA receptor.
Perspectives for Appropriate Treatments
Now that we are able to isolate autoimmune psychosis, future clinical trials should
evaluate if different types of immunotherapy may be helpful, in particular, involving
those routinely used in immune/autoimmune-related common disorders such as cortisone
pulse therapy, intravenous immunoglobulins, plasmapheresis, humanized monoclonal antibodies
(e.g., rituximab), or immunosupressor (e.g., cyclophosphamide).
Among the new therapeutic approaches already used with success in autoimmune psychosis,
three can be considered (10). The first one is based on the use of immunotherapies
from non-selective immunosuppressive ones like minocycline, steroids, plasma exchange,
or cyclophosphamide to a more selective one like the anti CD-20 monoclonal antibody
Rituximab (71). Schematically, CD-20 is a potent marker of B-lymphocytes. By targeting
CD-20, Rituximab will be able to inhibit B lymphocyte and, doing so, to prevent antibodies
circulation (72). The second approach also focuses on antibodies and has been proposed
by Diamond and colleagues. They propose to use d-peptide in order to prevent pathogenic
antibodies to reach their target, theoretically, without affecting receptor function
(73). It has been tested in mice model and seems to indeed prevent autoantibodies
effect (73). The last one is more specific and consists in the use of a co-agonist
of the NMDAR, the d-serine. It has been used by Heresco-Levy in an open label case
study and has shown a dramatic improvement in the psychosis symptomatology (74). The
potential mechanism of action behind d-serine is that it will enhance NMDAR activity
by increasing the frequency of channel opening to counter act the action of the antibodies.
Conclusion
Today, the main problem of the so-called autoimmune psychosis is that patients are
not diagnosed. In order to help the physician to evocate it and to consider an autoantibody
screening, we propose to gather elements enabling to build a risk score for autoimmune
psychosis. This score should take into account the personal and/or familial history
of early infections, autoimmune disorders, the demographic and clinical characteristics,
and the presence of blood biomarkers such as CD14 and a panel of autoantibodies (anti-bovine
casein, anti-transglutaminase, anti-folate receptor, anti-central nervous system,
etc.). Of course, such risk score will need to be built and validated, as it should
enable to allow early detection of autoimmune psychosis to prevent misdiagnosis with
long-term deleterious consequences. Clinical trials targeting specific mechanisms
and performed in homogeneous subgroups of autoimmune psychosis will allow to test
and to select the most efficient treatment. More than that, the discovery of N-methyl-d-aspartate
receptors antibodies (NMDAR-Ab) is also of major importance for a better comprehension
of the neurobiological basis not only of autoimmune psychosis but also psychosis in
general. We hope that, in a few years, personalized psychiatry will become the rule
and not the exception anymore.
Author Contributions
PE made the bibliography and wrote the article. ML, LG, and RT have corrected the
manuscript.
Conflict of Interest Statement
The authors declare that the research was conducted in the absence of any commercial
or financial relationships that could be construed as a potential conflict of interest.