Department of Psychiatry, University of Milan, Milan, Italy
Correspondence: Massimiliano Buoli, MD, Department of Psychiatry, University of Milan; Fondazione IRCCS Ca' Granda, Ospedale Maggiore Policlinico, Via F. Sforza 35, 20122, Milan, Italy. Email: email@example.com
Several lines of evidence point to the key role of neurobiological mechanisms and shared genetic background in schizophrenia and bipolar disorder. For both disorders, neurodevelopmental and neurodegenerative processes have been postulated to be relevant for the pathogenesis as well as dysregulation of immuno-inflammatory pathways. Inflammation is a complex biological response to harmful stimuli and it is mediated by cytokines cascades, cellular immune responses, oxidative factors and hormone regulation. Cytokines, in particular, are supposed to play a critical role in infectious and inflammatory processes, mediating the cross-talk between the brain and the immune system; they also possibly contribute to the development of the central nervous system. From this perspective, even though mixed results have been reported, it seems that both schizophrenia and bipolar disorder are associated with an imbalance in inflammatory cytokines; in fact, some of these could represent biological markers of illness and could be possible targets for pharmacological treatments. In light of these considerations, the purpose of the present paper was to provide a comprehensive and critical review of the existing literature about immunological abnormalities in bipolar disorder with particular attention to the similarities and differences with schizophrenia.
BIPOLAR DISORDER (BD) and schizophrenia are severe disabling disorders, which for several decades have been considered as distinct clinical entities. However, they share common biological features, such as brain changes in cases of long illness duration. Moreover, alterations in inflammatory pathways could lead to a common pathogenetic hypothesis involving immune dysregulation. The sharing of common inflammatory mechanisms does not match with the ‘classical’ binary classification of psychiatric disorders, but instead seems to support the hypothesis of a ‘continuum’ in pathogenetic patterns between BD and schizophrenia.[3-5] Of note, several data indicate that cytokines can be the mediator of metabolic/brain changes associated with clinical symptoms both of schizophrenia and BD.
Cytokines are a family of polypeptides that are essential to the immune system and are crucial systemic mediators of host response to infection, representing a robust marker of infectious and inflammatory conditions. Cytokines are synthesized and secreted by both immune and non-immune cells and their effects are mediated by specific receptors expressed on the surface of target cells. Cytokine receptors are available also in soluble forms and their activity can either enhance or inhibit the activity of cytokines. The action of cytokines is exerted by activation and recruitment of immune cells, increased vascular permeability and blood supply to inflamed tissues. Cytokines are frequently regulated in cascades, where induction of early molecules stimulate the production of later ones. They interact with hypothalamic–pituitary axis (HPA), monoaminergic system (dopamine, serotonin and glutamate) and autonomic system. Examples of cytokines include interleukins (IL), interferons (INF), tumor necrosis factors (TNF), transforming growth factors (TGF) and chemokines. These molecules and their receptors are expressed physiologically in central nervous system (CNS) cells. However, in normal conditions, many cytokines are virtually undetectable in the CNS and, because of their large size, their ability to cross the brain–blood barrier is reduced. Therefore, it has been argued that cytokines could access the CNS through several pathways, including: (i) binding to specific transporters on brain endothelium; (ii) activation of peripheral vagal fibres, which could transmit the signal to deep brain nuclei, such as nucleus of the solitary tract, which has the role of relay station to other brain nuclei; (iii) through leaky areas, such as the circumventricular organs; and (iv) increased permeability of brain–blood barrier.[10, 11] As well as entering from the periphery, cytokines can be secreted directly in the CNS, mainly by astrocytes and microglia.[12-15] Centrally secreted cytokines play a role in modulating neuronal development and plasticity, tissue repair and synaptogenesis[16, 17] or even promoting neural damage after brain insult. Rapid induction and sustained elevation of cytokines in immune and glial cells under inflammatory response play a role in the dysregulation of neural cell homeostasis. This contributes to abnormal neural cell development, cytotoxicity and loss of oligodendrocytes.
Actually, abnormal levels of proinflammatory cytokines and their receptors have been found in peripheral blood and cerebrospinal fluid of schizophrenic patients.[20-22] Recently, studies have focused on the role of cytokines in BD; however, less evidence is present with respect to schizophrenia. In addition, recent studies have clarified the reasons for cytokine dysfunction, and the role of immune polymorphisms and antioxidants, providing alternative views about the pathophysiology of major psychoses.
On the basis of the mentioned data, the present article aims to provide a comprehensive updated overview of the main acquisitions in the field of immuno-inflammatory abnormalities in BD with particular attention to the similarities and differences with schizophrenia.
Literature for this overview was identified by searching database sources (Medline, PsycINFO, Isi Web of Knowledge, Medscape) and Cochrane Libraries in two steps. First, a search was carried out identifying articles published related to inflammatory processes in schizophrenia and BD. Specifically, the keywords ‘schizophrenia’ and/or ‘bipolar disorder’ were variably combined with the terms ‘immune system’, ‘inflammation’ and ‘cytokines’. Furthermore, we performed a manual search for relevant articles, examining the reference lists of publications retrieved in the primary search. No restriction criteria were established in relation to study design so that the present review includes researches with different methodologies (e.g. open-label and controlled studies). Studies with mixed samples or without a DSM-IV-TR diagnosis were excluded.
Results of the retrieved publications are presented in five main sections. The first one describes the current hypotheses about the cause of immune abnormalities in major psychoses. In the second, neurodevelopment and progress theories are discussed. The third deals with immunological factors that could be thought of as possible candidates of markers of illness. The fourth describes the contribution of immunity to a dimensional diagnostic approach and the fifth describes studies comparing immuno-inflammatory features in schizophrenia and BD.
Hypotheses about immunological alterations in major psychoses
Historically two arms of research have studied the possible relation between immunological abnormalities and major psychoses. The first was focused on the role of infectious agents (‘infective hypothesis’), while the other took into consideration the dysregulation of immuno-inflammatory mechanisms (‘inflammatory hypothesis’).
Schizophrenia has been historically associated with cytomegalo and herpes' viral infections that would compromise the development of the CNS. In addition, the serum levels of toxoplasma antibodies were found to be higher in schizophrenic patients than in the general population. Toxoplasma could have a role in the cause of schizophrenia, as this parasite increases the CNS levels of kynurenic acid, an antagonist of glutamatergic N-methyl-D-aspartate receptor (NMDA) receptors. Finally, a flu infection during pregnancy increases the risk of future development of schizophrenia probably due to a dysregulation of GABAergic transmission.
With regard to BD, there are more scanty evidences about an association of infections and onset of symptoms; however, the infections of similar pathogens have been associated with both schizophrenia and BD. For example, with the exception of one paper, the seroprevalence of Toxoplasma gondii infection was found to be higher in individuals with BD with respect to unaffected controls.[29-31] In addition the exposure to influenza epidemics was found to increase the risk for adult BD. Finally, Pavuluri and Smith reported the case of a patient affected by mononucleosis that subsequently developed BD and showed remission of symptoms and normalization of neuroimmune dysfunction after treatment with lithium.
The ‘inflammation’ hypothesis of schizophrenia was firstly formulated after the observation of the positive effects of anti-inflammatory treatment and of lower rates of rheumatoid arthritis in patients suffering from schizophrenia compared to the general population.
Similarly, autoimmune thyroiditis was found to be frequently associated with BD, as well as a good response to the anti-inflammatory agent minocycline.
Praecox ‘stressors’, such as partum complications, are thought to be associated with overexpression of specific polymorphisms (e.g. inflammatory cytokines) that would be responsible for immune dysregulation associated with schizophrenia and BD.
Role of the immune system in the pathophysiology of schizophrenia and BD
Biological changes of schizophrenia, including immunological abnormalities, were traditionally explained by these two views:
A neurodevelopmental hypothesis suggests that a disruption of brain development during early life is responsible for later emergence of symptoms during adulthood
A progressive hypothesis suggests that symptoms and related neurotransmitter dysregulation is responsible for immunological changes that can be observed during the course of illness.
BD was traditionally thought of as an early onset disorder that is clinically and biologically stable during the course of illness with the exception of acute relapses. The finding of a different biological profile in recent-onset versus chronic euthymic bipolars has changed this traditional vision in recent years in favor of the role of illness in determining biological changes, including immunological dysfunction.[39, 40] In addition, similarly to schizophrenia, the role of praecox stressors has been postulated in explaining prefrontal–subcortical dysfunction observed in BD.
Contribution of immune system research to the neurodevelopmental model
The neurodevelopmental model of schizophrenia/BD states that immune alterations due to several factors, including decreased vitamin D, hypoferremia and zinc deficiency, are responsible for brain development abnormalities.23 Cytokines would be the mediators between immune abnormalities and CNS development. In fact, cytokines may play an important role during neurodevelopment at all stages from the differentiation of ectoderm into neuroepithelium, to the renewal of neuroepithelial cells, which act as precursors for all neurons, microglia and adult progenitors as well as a framework for radial neuron migration. During neurogenesis, there is an overproduction of neurons and glia and cytokines act crucially to promote survival of cells, which are properly connected in the neural network, as well as to induce apoptosis of cells with impaired connections.
In other words, cytokines represent a link between altered immune system, neurotransmission dysfunction and impaired neurodevelopment, all of which contribute to the onset of schizophrenia/BD in childhood or later life. In recent years, several studies have focused on infection and subsequent activation of inflammatory pathways during neurodevelopment.[43-45]
Two major hypotheses have been postulated: one argues that individual infections increase the risk of developing schizophrenia/BD; the second suggests that different infections act through common pathways, which may alter fetal brain development and increase risk of the disorder. On the basis of clinical and preclinical findings, cytokines have been proposed as key mediators of this common mechanism.[43, 45, 46]
As mentioned above, an increase of cytokines following a maternal infection may alter the immune status of the fetal brain, causing abnormal cellular development with subsequent brain damage.[44, 45] It has been well described that maternal immune activation induces increase of cytokines in the placenta (IL-1β, IL-6, TNF-α) and amniotic fluid (IL-6, TNF- α).[47, 48] Cytokines could act on the placenta, altering transfer of nutrients, oxygen, growth factors and maternal antibodies, all of which have a crucial effect on fetal development.
However, it is not clear whether cytokines are altered in the fetal brain, even though animal models suggest that there could be a significant increase in IL-1β, IL-6, and INF-γ.[49, 50] Cytokines seem to act directly on developing neurons and cause a disruption on maturation of oligodendroglia as well as white matter abnormalities. In addition they could activate different potential mechanisms, such as stimulation of astroglia and microglia, to produce cytokines potentiating excitatory amino acids and nitric oxide, which lead to disrupted maturation of oligodendrocytes.[45, 51, 52]
Specifically, overall increase of maternal cytokines, including IL-8 and TNF-α, is correlated with several infections. Of note, IL-8 appears to be a key factor for neutrophil attraction as well as for release of lysosomal enzymes from neutrophils, leading to discharge of oxygen free radicals. TNF-α has been associated with chorioamnionitis[45, 53] and fetal infections.[45, 54]
Along with maternal infections, other factors have been considered to be involved in maternal immune activation alteration. For instance, nutritional deficits and stress have been linked with increase of inflammatory cytokines. In fact, chronic and acute stress has been associated with increased production of proinflammatory cytokines[55, 56] and decreased levels of anti-inflammatory ones.[57-59] On the other hand, life in an urban area and birth in winter–spring can enhance the risk of maternal infections with subsequent inflammatory response.[48, 60-62] However, it has not been clarified if cytokine fetal imbalance is able to induce itself an immune dysregulation in the adult brain.
Contribution of immune system research to the progressive model
As an alternative to the ‘neurodevelopmental’ hypothesis of schizophrenia/BD, the ‘neurodegenerative’ hypothesis states that immunological alterations are secondary to the onset of illness and they can be observed during the progression of illness. It has been hypothesized that immunological alterations of schizophrenia and BD may be an adaptive response to microglial activation in mature CNS. Of note, IL-6 seems to play a crucial role in both the disorders in determining biological changes observed during the course of illness. IL-6 potentiates B lymphocyte proliferation: the hyperactivation of humoral immunity stimulates the tryptophan 2,3-dioxygenase enzyme with an increase of the transformation of the amino acid tryptophan in kynurenic acid that acts as an NMDA antagonist.[65, 66] Another metabolite of tryptophan catabolism is 3-hydroxykynurenine, which is considered an endogenous oxidative stress generator, probably involved in cognitive symptoms.[67, 68] In support of this hypothesis, some studies show that a long duration of illness in schizophrenia/BD is associated with higher serum levels of IL-6.[69-71]
Immunological factors as biological markers of illness
Immune dysregulation appears to play some role in the pathophysiology and outcome of major psychoses along with genetic factors and neurodevelopmental disturbances.[51, 72-74] For this reason, in recent years, researchers have tried to identify neuroimmunological markers, defined as any genetically or environmentally biological parameter, with the aim to identify individuals at risk and monitor the progression of illness.
Immunological markers of schizophrenia
As mentioned above, a number of studies indicate the presence of immuno-inflammatory abnormalities in schizophrenia.[73, 76-81]
In recent years, several different models have been proposed for explaining the cytokine dysregulation observed in schizophrenia. These models have been later used to explain also immunological abnormalities in BD.
The ‘macrophage T-lymphocyte hypothesis’ argues that chronically activated macrophages produce cytokines, such as IL-1, IL-2, TNF, IFN-α and IFN-γ, which could be the key mediators for schizophrenia.[83, 84] The illness would be, therefore, the result of a hyper-activation of cell-mediated immunity. In line with this hypothesis, some authors found higher levels of Th-1 cytokines in schizophrenic patients with respect to controls.
The ‘Th-2 hypothesis’ postulates that in schizophrenia, there is a predominance of Th-2 lymphocytes.[22, 47, 72] In contrast with the previous hypothesis, the illness would be the result of a hyperactivation of humoral immunity. This is actually the most corroborated hypothesis in light of:
Decreased in vitro production of IL-2 and IFN-γ pointing to a blunted production of Th1 cytokines;
decreased levels of neopterin, a product of activated macrophages;
increased CSF IL-4 levels: the key cytokine for the Th2 immune response.
It can be hypothesized that patients with different symptoms show a predominance of humoral or cell-mediated immunity. In this sense, IL-6 can be thought of as the unifier cytokine, being increased by activated monocytes and Th2 immune response.
More recently, a variant of the macrophage T-lymphocyte hypothesis known as the ‘microglial hypothesis’ has been formulated, which suggests that proinflammatory cytokines and free radicals are released by activated CNS microglia, thus causing neural degradation, abnormal neurogenesis and white matter abnormalities, which could contribute to the pathogenesis of the disorder.
Finally, a recent meta-analysis showed that cytokine alterations may vary with clinical status of schizophrenic patients, as in other inflammatory non-psychotic diseases, such as systemic lupus erythematosus and celiac disease. Some cytokines (namely, IL-1β, IL-6, and TGF-β) might be state markers for acute relapse, while some others (IL-12, IFN-γ, TNF-α and sIL2R) could represent trait markers of schizophrenia.
Chemokines act as chemoattractants to guide the migration of leukocytes. Most of the studies show that monocyte chemoattractant protein (MCP)-1 chemokine serum levels are increased in schizophrenic patients.[90, 91] Some authors hypothesize a role of this chemokine in the development of metabolic syndrome. Of note, in schizophrenics, there is frequently a polymorphism of the gene of the chemokine MCP-1, which seems responsible for resistance to antipsychotics and could be present in patients with more severe subtypes. In addition, one study found higher serum levels of chemokine CCL11 in schizophrenic patients with respect to healthy controls: this finding would reinforce the view that schizophrenic disorder may be associated with a Th1/Th2 imbalance with a shift toward a Th2 immune response, being CCL11 ligand preferentially expressed in Th2 lymphocytes.
Immune cells counting
The results are inconsistent as the majority of studies measured the effects of antipsychotics and not lymphocyte distribution in drug-naïve patients. However, study results are generally concordant in showing reduced T cells and increased B cells in drug-naïve patients and an inverse pattern when patients are medicated.
Polyunsaturated fatty acids
A number of data showed decreased concentrations of polyunsaturated fatty acids (PUFA) in both brain and peripheral membranes of schizophrenics. It has been hypothesized that this decrease can be due to the increased degradation of arachidonic acid as a result of altered immune function and overproduction of prostaglandins, particularly prostaglandin E. In line with this view, recent studies showed that adjuvant aspirin and celecoxib reduce symptoms of schizophrenia spectrum disorders.[97-99] The rationale for using cyclooxygenase antagonists in the treatment of major psychoses is, in fact, the upregulation of arachidonic acid metabolism associated with inflammation and excess proinflammatory cytokines.[99, 100] Actually, a very recent study found an overexpression of genes related to the arachidonic acid signaling pathway in schizophrenic patients compared to healthy controls.
Antioxidant defense system
In recent years, some authors have speculated that an impairment of the antioxidant defense system (AODS) could be responsible for neuron damage in schizophrenic patients. AODS, in fact, prevents cellular damage of free radicals that are overproduced during inflammatory processes. AODS includes enzymatic antioxidants (superoxide dismutase, catalase and glutathione peroxidase) and non-enzymatic ones (albumin, bilirubin, uric acid, ascorbic acid, tocopherol, glutathione). Schizophrenic patients are expected to have cellular/plasma decreased levels of antioxidants. With regard to non-enzymatic antioxidants, albumin, bilirubin, uric acid, ascorbic acid, glutathione and tocopherol were all found to be decreased in schizophrenics' plasma. Contrasting results were found regarding superoxide-dismutase serum/cellular concentrations: it has been speculated that this enzyme might be raised in chronic patients as a compensatory response to oxidative stress. Nitric oxide, which plays an important regulatory/modulatory role in a variety of inflammatory conditions, is a source of free radicals and is supposed to be increased in schizophrenic patients. However, the inconsistent findings[109, 110] exclude NO as a diagnostic marker to distinguish schizophrenics from healthy controls.
One of the effects of immune dysregulation of schizophrenics is the HPA dysfunction indicated by high cortisol and adrenocorticotropic hormone (ACTH). In addition, prolactin serum levels were found to be increased in drug-naïve schizophrenic patients in comparison with controls, but the correlation with symptoms severity and outcome is uncertain. Hyperinsulinemia has been found in schizophrenic patients and it has been associated with abnormal inflammation responses. Finally, high plasma levels of T3 and T4 were found in schizophrenics and thyroid hormogenesis was thought to be enhanced by higher peroxides plasma levels in schizophrenics compared to healthy controls.
Immunological markers of BD
The central role of cytokines in mood disorders is predominately supported by studies in unipolar depression. The macrophage hypothesis was sustained after the recognition of induction of a depressive syndrome in up to 50% of healthy volunteers by INF-α administration and of a high comorbidity between depression and inflammatory illnesses. More recently, there has been increasing evidence suggesting that chronic, mild inflammatory processes both in the periphery and the brain are involved in the pathophysiology of BD. Some complement factors (C3, C4 and C6) and C-reactive protein (CRP) levels have been found to be higher in BD than controls and acute phase protein response has been shown not only in major depression, but also in schizophrenia and mania, suggesting that all these conditions may result in an activation of inflammatory response. Significantly increased levels of cytokines, such as IL-1 and its receptors, have been found in the post-mortem frontal cortex of bipolar patients, compared to healthy controls.
Some studies demonstrated altered production of proinflammatory cytokines in bipolar patients compared to healthy controls, with differences in cytokine serum levels between different mood episodes. A key feature of BD is shifts in mood states, and it has been suggested that immune system activation may vary across mood episodes of different polarity. During depressive episodes, an increase in proinflammatory cytokines, such as IL-8, CRP and TNF-α, was significant among patients with BD versus controls;[20, 119, 120] whereas in mania, levels of IL-6, IL-8 and TNF-α were found to be increased, with inconsistent results for IL-4.[9, 20, 121] In another study, cytokine levels were compared in depressed, manic and euthymic patients with BD, finding an increase in IL-2, IL-4 and IL-6 levels in mania and only increase of IL-6 during depression, whereas euthymic patients showed an increase of IL-4. These data would suggest that changes in proinflammatory cytokines could be related to mood state, as some differences were more pronounced during acute episodes than euthymia. This hypothesis is consistent with another report showing that elevation of IL-6 in manic patients resolves with clinical remission after treatment with mood stabilizers, while TNF-α levels do not appear to change. Therefore, some components of inflammatory cascade could be state markers, but others, such as TNF-α, could be trait markers of BD. However, other results do not support these findings and the discrepancy could be explained by use of smaller sample size, their heterogeneity regarding mood symptoms, length of illness and effect of medications.[116, 119, 120]
There are also data suggesting that BD is associated with cytokine genetic polymorphisms.[122-125] Interestingly, a study found that IL-6 alleles had different distribution between adults with BD and controls, and in offspring (with and without mood disorder) versus controls as well as in offspring with versus without mood disorders. Contradictory findings were reported for TNF-α polymorphism: two studies described an association between TNF-α2 allele and BD,[122, 123] whereas some others studies were negative.[126, 127]
In summary, aberrant expression of inflammatory genes may be considered a biological marker or an endophenotype for BD, consistent with the above findings about state and trait markers of the disorder. In contrast to this consideration, a recent study with twins affected by BD demonstrated that proinflammatory gene expression observed in bipolar patients is mostly explained by environmental factors, although in a small sub-cluster of genes, genetic influences could dominate.
In general, hypotheses on the mechanisms by which alterations in inflammatory pathways could induce episodes and symptoms in BD are lacking or doubtful. Recent studies, given the established link between muscarinic M2 receptor and cognition, focused their attention on reduction of muscarinic M2 receptors levels in peripheral tissue inflammation. They highlighted the role of TNF-α in reducing the expression of this receptor and suggested that increased levels of TNF-α (as shown in depressive disorder) could be implicated in decreased levels of M2 receptors in BD and depression. According to this hypothesis, the link between M2 receptor, inflammation and cognition could provide a mechanism by which changes in inflammatory pathways could be one of the causes of cognitive deficits associated with BD. This is consistent with the fact that cognitive impairment is one of the core features of the disorder and at least some of it is related to lifetime number of episodes.
An alternative pathogenetic hypothesis states, similarly to schizophrenia, that bipolar mania is the result of an increase of tryptophan breakdown pathway. This would be consistent with the observation of an increase of IL-6 serum levels during manic episodes, associated with the activation of tryptophan 2,3-dioxygenase. Kynurenine pathway overactivation can be due also to lower serum levels of the immunosuppressive TGF-β1 observed in manic patients compared to healthy controls.
In manic patients, one study found increased serum levels of CCL11 similarly to schizophrenics, while in euthymic bipolars, no differences were observed with respect to controls. In addition, a study on a polymorphism of the chemokine MCP-1 (a cytokine playing a key role in innate immunity, named alternatively CCL2) showed a significantly higher frequency of the allele A in BD compared to unipolar depression, with a more frequent number of lifetime suicide attempts independent of diagnosis. Finally, increased serum levels of CCL2 in BD were found supporting the hypothesis of a Th1 hyperactivation.
Immune cells counting
In BD, a monocyte proinflammatory state and cell-mediate immunity activation has been reported as a result of indirect evidences like the plasma increases of monocyte-related cytokines. However, a study reported leukocytosis and increased monocytes in bipolar patients.
Preliminary data would indicate a disturbance in brain lipid concentrations of bipolar patients with the majority of results derived form post-mortem studies. In a post-mortem study, arachidonic acid concentrations in orbitofrontal cortex were found to be decreased compared to healthy controls. Interestingly, a recent paper reported elevations in cyclooxygenase and membrane prostaglandin E synthase in the BD cortex, which indicate a hyperactivation of arachidonic acid cascade in bipolars.
Preliminary data would indicate oxidative imbalance in BD as well as in schizophrenia. Total oxidant status would be high in the first stages of illness, while in chronic patients, increased AODS was found. In support of this speculation, a paper reported increased levels of the anti-oxidant enzyme glutathione S-transferase in the late stage of illness.
In depression, cytokines seem to activate the HPA axis, to increase CRH in both hypothalamus and amygdala, and to affect the monoaminergic system. In contrast, during manic phases, an inverse hormonal pattern would be observed with a decrease of plasma cortisol levels.[144, 145]
Clinical dimensions and immunological patterns
Some authors propose to substitute the categorical approach to psychiatric nosography in favor of a dimensional one in light of overlapping symptoms and biological features between diagnoses. In this section the contribution of immunology to the dimensional approach in schizophrenia and BD is summarized.
Dimensionality and schizophrenia
Positive symptoms have been hypothesized to be the result of changes in the enzymatic activity of tyrosine hydroxylase, which is supposed to be enhanced by proinflammatory activity. In addition, increased autoimmunity was found in patients with positive symptoms: a study showed that the progression of positive symptoms was coupled with gradual increase of autoantibodies to nerve growth factor.
Indoleamine-pyrrole 2,3-dioxygenase activation secondary to proinflammatory state would be involved in the pathogenesis of negative symptoms of schizophrenia by two mechanisms: the depletion of serotonin and the production of kynurenic acid. In support of this view, several studies report an association between peripheral or central inflammation and severity of negative symptoms. Of note, concentrations of plasma IL-6 and IL-8 were reported to positively correlated with the severity of negative symptoms. Furthermore, increased CRP levels were found in patients with predominant negative symptoms.
Similarly, cognitive schizophrenia symptoms would be the result of the effects of kynurenic acid on NMDA receptor complex as well as on cholinergic neurotransmission (blockade of nicotinic acetylcholine receptors, especially α7 types). In support of the correlation between inflammation and cognitive symptoms, a study found an association between CRP and severity of cognitive symptoms.
Finally, a study found an association between microglial density and suicide in schizophrenic patients, but these results have not been replicated.
Dimensionality and BD
Studies associating immunity and clinical dimension in BD are preliminary; as are those about HPA axis dysregulation. Higher cortisol plasma levels were found in dysphoric bipolars than in euphoric ones. In addition, haplotype variation at the CRHR2 locus was associated with suicidal behaviour.
Studies comparing abnormalities in schizophrenia and BD
In general, few studies have investigated similarities and differences in immuno-inflammatory dysregulation between BD and schizophrenia. From one side, genetic studies have shown an association of polymorphism in cytokines with both schizophrenia and mood disorders.
A significant excess of a polymorphism in the promoter region of the IL-1B gene was found among subjects with BD and schizophrenia compared to controls, especially in those who had a family history of these disorders. In other words, IL-1 cluster genetic variability could represent a common genetic susceptibility for BD and schizophrenia.
On the other side, there are conflicting results concerning the association between polymorphisms of TNF-α and both schizophrenia and BD. In some studies, a specific TNF-α polymorphism was present in both schizophrenic and bipolar patients; but in another study, this was only present in schizophrenic patients, and not in controls or BD patients.
Plasma levels of inflammatory markers in BD and schizophrenic patients versus healthy controls have been investigated. The patients, independently from the diagnosis, shared the same inflammatory profile and they had highly statistically significantly elevated levels of two inflammatory markers, named sTNF-R1 (a stable and reliable marker of the activity of TNF-α system) and von Wildebrand factor (vWf), which is an endothelial-related inflammatory marker.[7, 135] The combination of elevation of TNF-α and vWf would also indicate the presence of endothelial-related inflammation in both disorders.
The comparison of cytokine levels in bipolar euthymic and schizophrenic patients versus healthy controls showed an increase in IL-6 in schizophrenia compared to BD and healthy controls, and in the same study, IL-10 was increased in both schizophrenic and bipolar patients versus controls. On the other side, manic patients showed an immunological profile similar to schizophrenic patients versus controls, where IL-6 was more specific for schizophrenia and IL-6 and IL-2R for mania (Table 1).
Table 1. Similarities and differences of BD versus schizophrenia with regard to immunology
Shared features: Association with Toxoplasma gondii infection and flu epidemics[25-32]
Shared features: A role of praecox stressors (neurodevelopment model) and further biological changes after onset (neuroprogression)[38-41]
Cytokines level can vary according to clinical symptoms (state markers)
Differences: Increased IL-6 can be the result of monocyte activation (BD)[115, 136] or Th2 activation (schizophrenia)[22, 47, 72]
Increased CCL11 in manic patients and schizophrenia
Increased levels of MCP-1 alternatively named CCL2
Immune cells counting
Differences: Increased B cells and decreased T cells for schizophrenia, while increased monocytes for BD[94, 137]
Shared features: Decreased PUFA in brain membranes as a result of arachidonic acid cascade hyperactivation[95, 96, 101, 138-140]
Antioxidant defense system
Shared features: Increased oxidant status in first years of illness and potentiated antioxidant system in chronic patients[102-108, 141, 142]
Shared features: High cortisol levels in schizophrenia and bipolar depression[111, 143]
Differences: Decreased plasma cortisol levels in mania[144, 145]
Shared features: Polymorphism in the promoter region of IL-1B
The role of medications has to be taken into account when comparing immunological abnormalities between schizophrenia and BD. It is actually debated if the different abnormalities observed in the two disorders are the result of a distinct pathogenesis or of the effects of medications, such as lithium or clozapine.
Immunological abnormalities are present in major psychoses with more evidences regarding schizophrenia. The similarities (e.g. serum increase of IL-6 concentrations) between schizophrenia and BD support a ‘continuum’ between the two disorders. In particular, cytokine dysregulation could be responsible for the neurodegenerative aspects observed both in schizophrenia and in BD, particularly in patients with long duration of illness. On the other hand, schizophrenic illness would appear to be more related to a Th2 immune overactivation, while bipolar patients would show more frequently a monocyte-related immune response. It can be speculated that the different symptoms of schizophrenia versus BD would be the result of a distinct immunological profile with a shared genetic predisposition. This model explains the different course of first-episode psychotic patients. In some of these patients, a prevalent Th2 immune response will be associated with negative/cognitive symptoms and a diagnosis of schizophrenia, while a monocyte-related response with affective symptoms will be associated with a diagnosis of BD. This hypothesis is concordant with the current view that intermediate phenotypes between schizophrenia and BD are existing and there is not a clear distinction between the two disorders in terms of biological markers (Fig. 1).
Moreover, schizophrenia and BD frequently share the same pharmacological treatment with atypical antipsychotics and there are some available data concerning the effects of these drugs on patients' immunity and their effect, specifically on the immuno-inflammatory system. In this light, preliminary findings indicate that atypical antipsychotics may reverse immuno-inflammatory abnormalities at the same time with the improvement of clinical symptoms.[85, 163, 164] Interestingly, a recent paper showed that quetiapine treatment was associated with a parallel improvement of positive symptoms and a normalization of IL-2 receptor serum concentrations.
All these data open the way to new treatment options, including anti-inflammatory medications. Accordingly, a recent study on augmentative treatment with minocycline to antipsychotics found this combination to be effective in treating negative symptoms in early schizophrenia. Similarly, low doses of aspirin resulted in reduced risk of clinical deterioration in BD patients treated with lithium. Taken as a whole, these data indicate that the use of anti-inflammatory agents in the treatment of schizophrenia and BD are promising, but further research is needed to assess to which extent anti-inflammatory agents improve symptoms, particularly cognitive impairment, which is associated with neurodegenerative features in major psychoses.
Some limitations of the present paper have to be briefly described. A number of important conditions (e.g. age, lifestyle, obesity, smoking) influence the levels of circulating cytokines in psychiatric patients. These factors might have influenced the results of the cited studies. In addition, there is increasing evidence of immunological abnormalities in major psychoses, but the importance of these findings is still debated. This paper tried to give a comprehensive vision of the topic as much as possible, but only future research will be able to indicate the importance of these data in terms of diagnosis and pharmacological management of psychiatric disorders.
Professor Altamura is a Merck and Astra Zeneca consultant, and a member of the speaker's bureaus for Sanofi, Lilly and Pfizer. Dr Buoli is a Roche consultant. Dr Pozzoli does not have any affiliation with or financial interest in any organization that might pose a conflict of interest with the present article.