Örebro University
School of Medical Sciences Degree project, 15 ECTS January 2019
Lithium’s impact on proinflammatory cytokines in patients with
bipolar disorder, schizophrenia and major depressive disorder:
a systematic literature review.
Author: Johanna Helgesen, MS Supervisors:
Nikolaos Venizelos, Professor Senior. Mussie Msghina, MD, Assoc. Professor School of Medical Sciences, Campus USÖ, Örebro University, Sweden
Abstract
Background: Psychiatric diseases such as bipolar disorder, major depressive disorder (MDD) and schizophrenia are chronic ailments that severely affect daily function and quality of life. A relationship between elevated levels of proinflammatory cytokines and these disorders has been suggested in several studies. Lithium is used as a treatment in bipolar disorder, and as an adjunctive treatment in MDD and schizophrenia. Despite the extensive use of lithium, it’s mechanism of action is not fully understood. One of the proposed hypotheses of lithium’s mechanism of action is reduction in the levels of proinflammatory cytokines.
Aim: The aim of this systematic literature review is to describe the effects of lithium on proinflammatory cytokines (IL-1, IL-6, INF- γ, TNF) in bipolar disorder, MDD and schizophrenia.
Methods: The study was conducted through a search in the electronic database PubMed. Using the PICOS format, inclusion and exclusion criteria were specified. Search words and filters were combined using both Medical Subject Headings (MeSH) terms and free text search words. The search initially resulted in 105 articles and through inclusion and exclusion criteria full-text articles were estimated for eligibility. Risk of bias was estimated using the Cochrane Handbook.
Results: A total of 10 eligible articlets were included in this study. Nine out of 10 articles investigated bipolar patients, and one article investigated patients with MDD. No article investigated patients with schizophrenia. Four out of 9 articles regarding bipolar patients reported a significant decrease in some proinflammatory cytokines after lithium
treatment, 4 articles reported a significant increase, and one article reported no change in the proinflammatory cytokines. In two well conducted studies where bipolar disorder was
investigated, lithium had differential effects, namely decreasing proinflammatory cytokines in lithium responders compared to lithium non-responders. No significant change in
proinflammatory cytokine levels after lithium treatment were found in the article studying MDD patients.
Conclusions: The results indicated that lithium may have different effects depending on which specific cytokine was studied and on the specific characteristics of the studied population. Therefore, the outcomes of our review cannot unequivocally answer whether lithium acts by increasing or decreasing proinflammatory cytokines, or both depending on the experimental conditions. Further research is needed to fully elucidate the relationship between lithium and proinflammatory cytokines in bipolar disorder, MDD and schizophrenia.
Abbreviations
LAT1 – L-type amino acid transporter 1
MDD – Major depressive disorder
IL-1 – Interleukin 1
IL-6 – Interleukin 6
TNF-α – Tumour necrosis factor alfa
CRP – C-reactive protein
NSAID - Non-Steroidal Anti-Inflammatory Drugs
IL-1β – Interleukin 1 beta
INF- γ – Interferon gamma
1 INTRODUCTION ... 5
1.1 B
IPOLAR DISORDER... 5
1.2 M
AJOR DEPRESSIVE DISORDER... 5
1.3 S
CHIZOPHRENIA... 5
1.4 P
ROINFLAMMATORY CYTOKINES... 6
1.5 L
ITHIUM... 7
2 AIM ... 7
3 METHOD ... 7
3.1 D
ESIGN... 7
3.2 I
NCLUSION CRITERIA... 8
3.3 E
XCLUSION CRITERIA... 8
3.4 S
EARCH PLAN... 8
3.5 F
ILTER STAGES... 9
3.6 R
ISK OF BIAS... 9
3.7 E
THICAL CONSIDERATIONS... 9
4 RESULT ... 10
4.1 S
EARCH... 10
4.2 S
IGNIFICANT INCREASE IN PROINFLAMMATORY CYTOKINE LEVELS... 13
4.3 S
IGNIFICANT DECREASE IN PROINFLAMMATORY CYTOKINE LEVELS... 13
4.4 N
O SIGNIFICANT OR DETECTABLE CHANGE IN PROINFLAMMATORY CYTOKINE LEVELS... 14
4.1 R
ISK OF BIAS... 16
5 DISCUSSION ... 17
6 CONCLUSION ... 21
1 Introduction
1.1 Bipolar disorder
Bipolar disorder is a psychiatric disease involving periodic mood swings and shifts in energy levels[1]. It is divided into bipolar I and bipolar II, depending on the type of mood elevation. The prevalence in the world is 1-2% and 60 million people worldwide are estimated to be affected by this disease. It is a lifelong disorder that can severely affect function and quality of life. One aspect of the disorder is the episodic state of mania or hypomania, which are characterized by elevated mood and increased motor activity. During a manic episode a large number of affected individuals meet the criteria for psychotic symptoms. Bipolar depression, on the other hand, is characterized by low mood, anhedonia, dysphoria as well as lethargy and low levels of energy. Bipolar depression is generally evaluated using the same criteria used for unipolar depression. The risk for suicide in bipolar disorder is roughly 20 times higher compared to the rest of the population. The rate is so high that during a lifetime up to half of all bipolar patients attempt suicide at least once. The pathogenesis of bipolar disorder is not yet completely understood, and no biomarkers have been identified that could be used as diagnostic markers. To this day, bipolar disorder is diagnosed through interview and rating scales [1].
1.2 Major depressive disorder
Major depressive disorder is an episodic disorder with a high lifetime prevalence, and a higher prevalence in women compared to men. The symptoms are low mood, anhedonia and dysphoria [2]. A neuroinflammatory hypothesis of the pathogenesis of MDD has been postulated, and many studies have reported increased proinflammatory cytokine levels in MDD, including IL-1, IL-6, TNF-α and CRP [3]. There have also been studies suggesting a beneficial effect of adjunctive NSAID therapy together with anti-depressive medication in some cases of MDD, but there are also studies reporting the opposite [3], which might reflect on the heterogeneous nature of the pathogenesis of MDD.
1.3 Schizophrenia
Schizophrenia is a disorder characterized by psychotic episodes and motivational and cognitive impairment, displaying a vast spectrum of symptoms [4]. Diagnosis often occurs during the onset of the first psychotic episode, but there have been arguments that signs of schizophrenia might already be present many months prior to diagnosis. Studies have shown
evidence of a decreased level of intracranial volume in patients with schizophrenia, which argues for the disorder being active before 13 years of age when the size of the brain reaches its maximum [4]. Also, evidence points to a decrease in brain volume implicating both the white matter and grey matter. The pathophysiology of schizophrenia is not clearly understood, and there are several different hypotheses, as well as arguments for a heterogeneity of its pathophysiology. One of the hypotheses is increased proinflammatory status that could affect neurodevelopment and neuronal damage through neurotoxic substances (e.g. free radicals and proinflammatory cytokines) [4].
1.4 Proinflammatory cytokines
Proinflammatory cytokines are soluble mediators produced by immune cells. These cells are activated at the location of infection or lesions, but also through different mechanisms such as in response to pathogen-associated molecular patterns (PAMPS). Proinflammatory cytokines trigger changes in the body such as endocrine, autonomic and behavioral [5,6].
The relationship between elevated levels of proinflammatory cytokines and a number of neuropsychiatric disorders such as schizophrenia, bipolar disorder, MDD and autism has been suggested from different studies [7–11]. Studies have also proposed favorable symptomatic effects using immunomodulatory drugs on patients with schizophrenia [12]. Also, some antidepressants medication show reduction in levels of proinflammatory cytokines in patients with MDD. There is a measurable elevation of these cytokines in MDD patients, even in the absence of somatic illness. As reduction in proinflammatory cytokines is not seen for all antidepressant treatments, although pretreatment levels of cytokines are high in patients, thus the possibility that some antidepressants inhibit the effect of proinflammatory cytokines on the brain instead of reducing the levels has been suggested [13].
Disturbed transport of tyrosine across the cell membrane of fibroblasts cultured from of patients with bipolar disorder and schizophrenia has been described [14,15]. Transport of the amino acids tryptophan (precursor of serotonin) and tyrosine (precursor of dopamine and noradrenaline) is largely facilitated through the LAT1 isomer of system-L, an amino acid transporter in fibroblasts [16,17]. The same is assumed to be true for micro vascular endothelial cells in the human brain that form a part of the blood brain barrier [18]. In fibroblasts, 90% of the transport of tyrosine was shown to be facilitated through LAT1 [17]. The effect of proinflammatory cytokines (IL-1β, IL-6 and TNF-α) on tyrosine transport in
fibroblast was demonstrated and indicated a significant decrease in transport. The effect of IL-1β measured the most impact and decreased the transport with 47%. The same study also showed a decreased transport of tyrosine in fibroblasts when exposed to reactive oxygen species [19]. There has been evidence that suggest a relationship between proinflammatory cytokines and the induction of oxidative stress, as well as the conversely [20].
1.5 Lithium
Lithium has been used as medication for over 150 years. Around 1870 it was used for “general nervousness” and shortly after that it began to be prescribed as treatment for mania [21]. In the present day lithium is used as a treatment for bipolar disorder, schizophrenia and major depressive disorder [22–24]. There has also been strong evidence of an anti-suicidal effect of lithium maintenance therapy [25]. Although the pharmacological use of lithium is extensive, and in some cases a first-line treatment for disorders such as bipolar disorder, its mechanism of action is not clearly understood [26,27].
Many hypothesis have been proposed to explain lithium’s pharmacodynamics, some of them include neurotransmitters, apoptosis and neuronal structures [27]. There have been studies suggesting a positive effect on the cell’s ability to resist oxidative stress when treated with lithium [28,29]. Also, a reduction in levels of proinflammatory cytokines after treatment with LiCl (lithium chloride) has been demonstrated [30,31]. In this review, we will overview the presence of proinflammatory cytokines and lithium therapy to investigate the relationship between these two.
2 Aim
The aim of this systematic literature review is to describe the effects of lithium on
proinflammatory cytokines in bipolar disorder, schizophrenia and major depressive disorder.
3 Method
3.1 Design3.2 Inclusion criteria
The survey was transacted using the PRISMA statements guidelines [32] (an evidence-based minimum set of items for reporting in systematic reviews and meta-analyze) and population, interventions, comparator, outcome and study design (PICOS) format specified below:
Population: Patients with Bipolar disorder, Schizophrenia, Major depressive disorder, Mania/Manic disorder
Intervention: Lithium Comparison:
Outcome: Proinflammatory cytokines (IL-1, IL-6, INF- γ, TNF) Study design: English
3.3 Exclusion criteria
In this study articles were excluded by the following criteria:
1) Articles investigating other subjects than human or human cell populations 2) Scientific articles describing lithium in other usages than therapeutic purposes 3) Articles not compliant with the criteria for the population
4) Articles not presenting data for the proinflammatory cytokines mentioned earlier or lithium, or the relationship between these two
5) Secondary sources (e. g. reviews). 6) Articles not available in English. 7) Articles not available in full text
3.4 Search plan
Using the databases PubMed a thorough literature search was performed. Based on PICO format search words and filters were combined using both Medical Subject Headings (MeSH) terms and free text search words (see Table 1). The search was accomplished through
combination of the search words for Population, Intervention and Outcome parted with the word “AND”. The words within each section was parted with the word “OR”. The search resulted in 105 articles. Latest searched December 4, 2018.
Table 1. Search plan and criteria
Inclusion criteria Search words Filters
Population Bipolar disorder, Schizophrenia, Major depression disorder, Manic disorder, Human subjects
“Bipolar” “Schizo*” depressi*” “Mania” “Manic”
“Psychiatry and Psychology Category"[Mesh]
“Human”
Intervention Lithium “lithium” “LiCl” “lithium chloride” Comparison Outcome Proinflammatory cytokines “interleukin-1” “IL-1” “interleukine-6” “IL-6”
“interferon gamma” “INF-gamma” “INF-γ”
“tumor necrosis factor” “tumour necrosis factor” “TNF”
“proinflammatory cytokine” “inflammatory cytokine”
Study design Written in English “English”
3.5 Filter stages
Search words were combined with the filter word “human”. The titles were screened, and exclusion criteria were applied. At the second stage the selected abstracts were screened, and inclusion and exclusion criteria were applied. At the third and last stage the full text of each article was screened and once again the inclusion and exclusion criteria were applied, which resulted in the included articles of the systematic literature review.
3.6 Risk of bias
The Cochrane Handbook for Systematic Reviews of Intervention was used to assess the risk of bias in the reviewed articles [33]. Selection bias, performance bias, detection bias was evaluated and graded as low risk, unclear risk and high-risk using instructions from chapter 8 in the handbook.
3.7 Ethical considerations
No original data was used when composing this systematic literature study. All data extracted from included articles was processed and published. Despite this, it could be of value to
emphasize that seven out of ten of the included articles stated that they had ethical approval. One stated that they had approval for the study protocol and written consent from the
participants but did not specify an ethical approval. One study stated that they had a written approval from the participating patients. One study did not present either an ethical approval or a written consent from the participants.
4 Result
4.1 SearchFlow chart of stages in search, exclusion and result are presented in Figure 1. Initially, the search gave a total of 105 articles when combining the different search words as described above. When filters were applied this gave a result of 54 articles. After screening the titles for eligibility this resulted in 44 articles. The remaining articles were then screened by abstract for eligibility which resulted in 32 articles. The 32 articles were then screened by full text and resulted in 10 eligible articles that were included in the systematic literature study (Table 2). The articles excluded were dismissed due to the intervention not being eligible (3), the population not being eligible (15), the measurements not being eligible (2), the study design not being eligible (1) and full-text article not being available (1).
Table 2. Included articles presented with study size, study design and follow-up period relevant to the association between bipolar disorder, schizophrenia or
MDD and proinflammatory cytokines and lithium treatment.
Authors n Study size PICO Follow-up period Adjusted factors
Ricken et al. 2018 [34] Patients: 95 P: MDD
I: Lithium treatment C: Before treatment O: IL-6, TNF-α, IFN-γ 4 weeks Li et al. 2015 [35] Patients: 41 Controls: 77 P: Bipolar I disorder
I: Combination treatment lithium and quetiapine C: HC
O: TNF-α
8 weeks
Guloksuz et al. 2012 [36] Patients: 60 P: Euthymic state bipolar disorder I: Lithium
C: Lithium response scale O: TNF-α
Guloksuz et al. 2010 [37] Patients: 31 Controls:16
P: Euthymic state bipolar disorder I: Lithium
C: Medication free O: IFN-γ, TNF-α
>8 weeks
Padmos et al. 2008 [38] Patients: 42 P: Bipolar disorder I: Lithium
C: BD without treatment
O: Gene expression IL-6, IL1B, TNF, TNFAIP3 Knijff et al. 2007 [39] Patients: 80
Healthy controls: 59
P: Bipolar disorder I: Lithium
C: Non-lithium users O: IL‐1β, IL-6
Himmerich et al. 2005 [40] Patients:10 P: Bipolar or monopolar disorder I: Lithium
C: Before lithium treatment O: TNF-α
4 weeks
Boufidou et al. 2004 [41] Patients: 40 Controls: 20
P: Euthymic state bipolar disorder I: Lithium
C: HC
O: IFN-γ-, IL-6-producing cells Rapaport et al. 1999 [42] Patients: 17
Controls: 18
P: Rapid cycling bipolar disorder I: Lithium
C: HC O: IFN-γ, IL-6
30 days
Munkholm et al. 2018 [43] Patients: 92 Controls: 71
P: Rapid cycling bipolar disorder or hospitalization with manic or mixed episodes
I: Treatment as usual C: HC
O: Leukocytes, Neutrophils,
6-12 months age, gender, BMI, smoking status, alcohol,
hypertension, antipsychotic, lithium,
4.2 Significant increase in proinflammatory cytokine levels
In 4 out of 10 eligible articles (37, 39,40, 43) a significant increase in proinflammatory cytokine levels after lithium treatment was reported (Table 3). An increased level of TNF-α during lithium treatment was found by two different studies; one comparing cytokine levels in bipolar patients during lithium treatment [40] and one comparing euthymic bipolar patients with lithium
treatment to those with no treatment [37]. One study [43] investigating the association of
leukocytes, neutrophils, lithium and proinflammatory cytokines, found a significant relationship between lithium and 20% higher leukocyte count and 34% higher neutrophil count compared to no lithium treatment. The same study presented a statistically significant positive relationship between IL-6 and leukocyte count as well as neutrophil count, and the same was to be true for TNF-α [43]. In one study [39] stimulating monocytes of bipolar patients with LPS in vitro, the ratio IL‐1β/ IL-6 production was significantly higher in bipolar patients with lithium treatment compared to those without, with or without LPS stimulation. In bipolar patients treated with lithium the ratio was instead comparable with the ratio of healthy controls. There was a significant increase of IL‐1β production as well as a reduction of IL-6 production in patients treated with lithium compared to those without, when stimulated with LPS. The same change in cytokines was not statistically significant when not stimulated with LPS. The study could not find a significant correlation between duration of lithium treatment and production of IL‐1β or IL-6, nor did it suggest a change in cytokine production during different mood symptoms under lithium treatment [39].
4.3 Significant decrease in proinflammatory cytokine levels
In 5 out of 10 eligible articles (35, 36, 38, 41) a significant decrease in proinflammatory cytokine levels were reported (Table 3). One study [35] showed a decreased level of TNF-α in bipolar patients who attained remission after combination treatment with lithium and quetiapine, compared to the patients who were not in remission (p < 0.05) [35]. In agreeing with this, one study [36] found a decrease in TNF-α levels in bipolar patients who responded well to lithium treatment compared to patients who were partial or poor responders (p = 0.002) [36]. One study [38] investigating the gene expression of some inflammatory cytokines in monocytes suggest a decrease in TNF andTNFAIP3 gene expression in bipolar patients with lithium treatment compared to those without lithium treatment. There was a lower but not statistically significant expression of IL-6 or IL-1B genes in patients with lithium treatment [38]. In one of the studies [41] the number of IFN-γ-, IL-6-cytokine secreting cells in peripheral blood was measured and
suggested a statistically significant lower number of cytokine secreting cells in lithium treated patients compared to healthy controls in response to stress (stimulated with phytohemagglutinin). The same study also investigated cytokine producing cell count in never before medicated
bipolar patients, before and after 3 month of lithium treatment. The patients showed a statistically significant lower count in cytokine secreting cells after treatment, compared to before treatment, when stimulated with phytohemagglutinin [41]. As mentioned in the previous section, one article [39] found a change in proinflammatory cytokine production in monocytes of bipolar patients, including significant reduction of IL-6 production in patients treated with lithium compared to those without when stimulated with LPS.
4.4 No significant or detectable change in proinflammatory cytokine levels
In 3 out of 10 eligible articles (34, 37, 42) no significant change in certain proinflammatory cytokine levels was found (Table 3). One out of 10 articles [34] did not find any change in the concerning pro-inflammatory cytokines (IL-6, TNF-α, IFN-γ) after lithium augmentation in patients with MDD. Article [37] found no indication of significant change in INF-γ levels in euthymic bipolar patients with lithium treatment. Study [42] investigated IL-6 and IFN-γ serum levels before and after 30 days of lithium treatment in rapid cycling bipolar patients and healthy controls, and no detectable levels of cytokine were found before or after lithium treatment in neither patients nor healthy controls [42] (Table 3).
Table 3. Lithium’s effects on proinflammatory cytokine levels in bipolar disorder, schizophrenia and MDD,compared to healthy controls or within patient group. The articles are presented in type of disorder examined.
Authors
Bipolar disorder MDD Schizophrenia
Lithium’s effect comparing HC and
patient group*
Lithium’s effect within patient group* Lithium’s ffect comparing HC and patient
group*
Lithium’s effect within patient group* Lithium’s effect comparing HC and patient group* Lithium’s effect within patient group* Ricken et al. 2018 [34] No significant change in IL-6, TNF-α, IFN-γ serum
levels before LT compared to after Li et al. 2015
[35] remitted patients compared to non-remitted after Significant decrease in TNF-α plasma level for treatment
Guloksuz et al. 2012 [36]
Significant decrease in TNF-α plasma levels in good response patients compared to partial or poor response
patients Guloksuz et al.
2010 [37]
Significant increase in TNF-α serum level in patients with LT compared to medication free, no significant
change in INF-γ serum level Padmos et al.
2008 [38]
Significant decrease in gene expression of TNF (and TNFAIP3) in patients with LT compared to patients
without LT Knijff et al.
2007 [39] Significant higher IL‐1β/IL-6 ratio in patients with LT compared to patients without LT, including increased IL‐1β production and decreased IL-6 production when
stimulating with LPS Himmerich et
al. 2005 [40]
Significant increase in TNF-α plasma level after LT Boufidou et al.
2004 [41] of cytokine secreting cells Significant lower number in LT patients compared
to HC
Significant decrease in IFN-γ-, IL-6-cytokine producing cells after LT
Rapaport et al.
1999 [42] No detectable IFN-γ, IL-6 serum levels in patients nor HC before or after LT Munkholm et
al. 2018 [43] Significant increased lymphocyte and neutrophil count in LT patients compared to patients without LT
Risk of bias was estimated through the Cochrane Handbook chapter 8 [33]. The estimation resulted in an overall “unclear risk of bias” (Figure 2).
Two of the included articles were estimated as “high risk of selection bias”. In one of the articles, bipolar patients without medication who were investigated had chosen themselves to stop medication before the study was performed. In the other article, the control group was comprised of medical staff members and medical students. The remaining 8 articles did not specify a randomized sequence generation of participants and were therefore estimated as “unclear risk of bias”.
One article was described as an open-labeled study and was estimated as “high risk of performance bias”. The remaining 9 blinding of participants and personnel was not clearly specified and were therefore estimated as “unclear risk of bias”.
Two articles stated a blinding of outcome assessment and were estimated as “low risk of bias”. The remaining articles did not specify how outcome was assessed and were estimated as “unclear risk of bias”.
Figure 2. Estimated risk of bias in included articles.
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Blinding of outcome assessment (detection bias)
Blinding of participants and personnel (performance bias)
Random sequence generation (selection bias)
Risk of bias
5 Discussion
The aim of this review is to describe the effects of lithium on proinflammatory cytokines in bipolar disorder, schizophrenia and major depressive disorder. A relationship between
proinflammatory cytokines and the previous mentioned psychiatric disorders has been described in several articles. The pharmacodynamics of lithium have not fully been described, wherefore the interaction between lithium and proinflammatory cytokines were explored in this review. This was done through a systematic search plan using PubMed, the most relevant database for medical sciences. The systematic literature data search gave a total result of 105 articles from start. By application of exclusion criteria, the articles were reduced to 10 eligible articles that fulfilled both inclusion and exclusion criteria, which were included in this systematic literature study (Table 2).
Of the 10 eligible articles included in this study, only one article investigated lithium treatment in MDD. 9 of the included articles investigated lithium treatment in bipolar disorder, and none of the included articles investigated lithium treatment in schizophrenia (Table 3). The achieved results apply only to bipolar disorder, which will be discussed subsequent below.
Considering the results of the 9 articles describing lithium treatment in bipolar patients in general, fiveof these articles suggested a decrease in proinflammatory cytokine levels, and four suggested an increase in proinflammatory cytokine levels during and after lithium treatment of patients with bipolar disorders. One article [42] demonstrated no detectable serum levels of cytokines in patients nor HC before or after lithium treatment, which implies that the patients were not affected by any inflammation at this time. One article [37] found no indication of significant change in INF-γ levels in bipolar patients. One out of 10 articles [34] did not find any change in the concerning proinflammatory cytokines (IL-6, TNF-α, IFN-γ) in serum after lithium augmentation in patients with MDD. The above outcomes show that we have a group of five studies, which are predicting that lithium treatment has muted effects on certain proinflammatory cytokine levels thus also on inflammation, and one group of four studies, which are predicting that lithium treatment has the opposite effects on certain proinflammatory cytokine levels that could indicate a proinflammatory effect. Also, three of the included articles indicates no significant change in proinflammatory cytokine levels which means no effect at all on inflammation.
Regarding the outcomes of the particular proinflammatory cytokines IL-1, IL-6, INF- γ, and TNF- α, two of the articles [35, 36] suggested a decrease of the TNF-α level in patients with good response or remission. Agreeing with the previous [35, 36], one study [38] suggested a decrease in TNF gene expression in lithium treated bipolar patients. Contradictory, two articles [40, 37] described an increase in TNF-α in bipolar patients when treated with lithium. Two studies [34, 37] measured INF-γ but could not find any change in cytokine level after lithium treatment in bipolar patients. One of these [34] also measured IL-6 but could not detect any level of the cytokines measured before or after lithium treatment. One study [39] suggested a change in IL-1β/IL-6 ratio in bipolar patients after lithium treatment when stimulated by a stressor and noted an increase of IL-1β production and a decrease of IL-6. Concerning cytokine secreting cells, one study [41] suggested a decreased number of cells producing IFN-γ and IL-6 in bipolar patients when treated with lithium, and one study [43] reported increased level of leukocytes and neutrophils associated with IL-6 and TNF- α, when treating bipolar patients with lithium. Two studies [39, 41] reported a different response in cytokine secreting cells of bipolar patients on lithium treatment when stimulated with lithium and stressor in vitro compared to only stimulate with stressor.
The results of this review are shown to be inconsistent, possibly because of the diverse methods, populations (no homogeny of patient subgroups) and interventions of the different articles. In this context, needs to be mentioned also that not many similar studies of this subject have been made, which makes it difficult to comparethe results but also to draw a relevant conclusion of lithium’s effects. There are several suggestions to why this result was attained. Some aspects of bipolar disorder that should be taken into consideration are (i) the differences between lithium responders and non-responders, (ii) if the patients had depressive, manic or hypomanic relapse when they entered the study, (iii) if the study is investigating bipolar I, II or mixed state, (iv) if lithium was introduced as acute treatment in destabilized patients or if it was a preventive treatment in stable but symptomatic patients, or completely euthymic patients, (v) if the diagnosis was correct considering that it sometimes is difficult to separate schizoaffective syndrome from bipolar disorder. These aspects could possibly affect the outcome of lithium’s effect on proinflammatory cytokine levels in bipolar patients and therefore inflict on the result of this review. Moreover, it could be of interest in scientific research to take these aspects into consideration.
One finding that could be of great interest, as well as an explanation to lithium’s diverse effect on proinflammatory cytokine levels, is the result from two of the included articles [35, 36]. The articles investigated bipolar disorder and suggested that lithium had differential effects, namely decreasing proinflammatory cytokines in lithium responders compared to lithium
non-responders. If this were to be correct, not specifying the attained response to lithium treatment in the patient group of the included articles would be a source of error affecting the result. There may be other explanations to the obtained result. An elevated level of inflammatory cytokine receptors in bipolar patients have been observed in different studies [8], and also suggestions of decreased levels of proinflammatory cytokine receptors when treating bipolar patients with lithium [42].The effects of cytokines are transduced through receptors, and alternated levels of proinflammatory cytokine receptors would most certainly affect the inflammatory state. As mentioned in the introduction an inflammatory state has been suggested to affect different neuropsychiatric disorders. Also of interest, two of the studies [39,41] included reported different effect of lithium when stimulating cells in vitro compared to lithium treatment. This suggests that result from studies only investigating lithium’s effect in vitro might not reflect the lithium’s effect as a medication. Furthermore, one article [42] reported altered effects of lithium treatment in healthy controls compared to bipolar patients. These results accentuate the
importance of using bipolar patients as participants in studies investigating the
pharmacodynamics of lithium as a bipolar treatment, and not translate the effect in healthy controls to bipolar patients. In the included articles, only serum levels or cytokine producing blood cells were investigated. As the concerning disorders are neuropsychiatric disorders it would be relevant to investigate lithium’s effect on cytokine levels or proinflammatory cytokine producing cells in cerebrospinal fluid instead of serum. Lithium is one of the primary treatments of bipolar disorder and is not used as often in MDD or schizophrenia, explaining why not as many articles on lithium and the two later disorders are to be expected.
A limitation of this review is the narrow search. It is desirable to accomplish a wide search to verify that relevant articles are not left out. Likewise, although the database PubMed is extensive and comprehensive regarding medical research, it is appropriate to include at least two databases when performing a systematic literature review. Another limitation is the possibility that only statistically significant data has been published. This would restrict the information included in the study. Also, only English published articles are included in this study, which might affect the result. One study was not available in full text and therefore excluded, with a risk of affecting the result.
Limitations regarding the included articles are many. Primary, none of the included studies were randomized. Only two articles stated blinding of outcome, and no article stated blinding of participants and personnel. On the contrary, one article stated an open-labelled study. When excluding participants, a majority of the included articles checked for recent illness. No article mentioned participants being examined during the trail for illness and being omitted. Some of the studies required participants to self-report if their bipolar symptoms worsened or changed, so that they could be examined by the physicians. This trusts that participants will have a correct subjective judgement of their symptoms and be willing to contact the study. Likewise, in a number of the included studies the participants self-reported the amount of tobacco and alcohol used. This subjective self-estimation was used as an exclusion criterion. Once again, no article mentioned participants being asked about tobacco and alcohol during the study. A number of the included studies were already aware of who of the participants belonged to the intervention group or not when beginning a trail. Also, when including patients in a trail, physicians use a rating scale to evaluate psychiatric disorders. This could allow for subjective judgement when including participants. It varies in how the studies have gone about when evaluating disorders, regarding if they used one researcher or a team. Some studies only mentioned the ordination of lithium to the patients and did not measure serum lithium level. This would potentiate for pooling of the drug and might affect the outcome. The included articles had a variation of follow-up periods which is problematic for comparison. Also, the effect of lithium might vary depending of the treatment duration.
In summary, both results of increased as well as decreased proinflammatory cytokine production during and after lithium treatment has been documented. A possible explanation to the diverse result could be the dissimilar study designs of included articles. An elevated level of
inflammatory cytokine receptors in bipolar patients have been observed in different studies [8], and there have been suggestions of a decreased levels of proinflammatory cytokine receptors when treating bipolar patients with lithium [42]. Since no given conclusion of the impact of lithium can be drawn from this review, taken into consideration also the small study sizes of the included articles in this review, more extensive studies on this subject would be required to understand the effects and mechanism of lithium on proinflammatory cytokine. It might also be proposed to investigate other aspects of lithium.
6 Conclusion
The obtained results of this study indicate that lithium may have different effects depending on which specific cytokine studied and on the specific characteristics of the studied population. Therefore, the outcomes of our review cannot unequivocally answer whether lithium acts by increasing or decreasing proinflammatory cytokines, or both depending on the experimental conditions. Further research is needed to fully elucidate the relationship between lithium and proinflammatory cytokines in bipolar disorder, MDD and schizophrenia.
Considering the selective passage of proinflammatory cytokines through the blood brain barrier would be desirable.
7 References
1. Bipolar disorder. The Lancet 2016;387:1561–72.
2. Schmidt HD, Shelton RC, Duman RS. Functional Biomarkers of Depression: Diagnosis, Treatment, and Pathophysiology. Neuropsychopharmacology 2011;36:2375–94.
3. Young JJ, Bruno D, Pomara N. A review of the relationship between
proinflammatory cytokines and major depressive disorder. Journal of Affective Disorders 2014;169:15–20.
4. Kahn RS, Sommer IE. The neurobiology and treatment of first-episode schizophrenia. Mol Psychiatry 2015;20:84–97.
5. Dantzer R, O’Connor JC, Freund GG, Johnson RW, Kelley KW. From
inflammation to sickness and depression: when the immune system subjugates the brain. Nat Rev Neurosci 2008;9:46–56.
6. Quan N, Whiteside M, Herkenham M. Time course and localization patterns of interleukin-1β messenger rna expression in brain and pituitary after peripheral administration of lipopolysaccharide. Neuroscience 1998;83:281–93.
7. Meta-Analysis of Cytokine Alterations in Schizophrenia: Clinical Status and Antipsychotic Effects - ScienceDirect [Internet]. [cited 2018 Nov 22];Available from:
https://www-sciencedirect-com.db.ub.oru.se/science/article/pii/S0006322311004045
8. Munkholm K, Braüner JV, Kessing LV, Vinberg M. Cytokines in bipolar disorder vs. healthy control subjects: A systematic review and meta-analysis. Journal of Psychiatric Research 2013;47:1119–33.
9. Söderlund J, Olsson SK, Samuelsson M, Walther-Jallow L, Johansson C, Erhardt S, et al. Elevation of cerebrospinal fluid interleukin-1β in bipolar disorder. J Psychiatry Neurosci 2011;36:114–8.
10. Simon NM, McNamara K, Chow CW, Maser RS, Papakostas GI, Pollack MH, et al. A detailed examination of cytokine abnormalities in Major Depressive Disorder. European Neuropsychopharmacology 2008;18:230–3.
11. Jyonouchi H, Sun S, Le H. Proinflammatory and regulatory cytokine production associated with innate and adaptive immune responses in children with autism spectrum disorders and developmental regression. Journal of Neuroimmunology 2001;120:170–9.
12. Müller N, Ulmschneider M, Scheppach C, Schwarz MJ, Ackenheil M, Möller H-J, et al. COX-2 inhibition as a treatmentapproach in schizophrenia: Immunological considerations andclinical effects of celecoxib add-on therapy. European Archives of Psychiatry and Clinical Neurosciences 2004;254:14–22.
13. Hannestad J, DellaGioia N, Bloch M. The Effect of Antidepressant Medication Treatment on Serum Levels of Inflammatory Cytokines: A Meta-Analysis. Neuropsychopharmacology 2011;36:2452–9.
14. Aberrant Tyrosine Transport Across the Cell Membrane in Patients With
Schizophrenia | Psychiatry | JAMA Psychiatry | JAMA Network [Internet]. [cited 2018 Nov 22];Available from:
https://jamanetwork-com.db.ub.oru.se/journals/jamapsychiatry/fullarticle/481833
15. Persson ML, Johansson J, Vumma R, Raita J, Bjerkenstedt L, Wiesel F-A, et al. Aberrant amino acid transport in fibroblasts from patients with bipolar disorder. Neuroscience Letters 2009;457:49–52.
16. Vumma R, Johansson J, Lewander T, Venizelos N. Tryptophan Transport in Human Fibroblast Cells—A Functional Characterization. Int J Tryptophan Res
2011;4:19–27.
17. Vumma R, Wiesel F-A, Flyckt L, Bjerkenstedt L, Venizelos N. Functional characterization of tyrosine transport in fibroblast cells from healthy controls. Neurosci. Lett. 2008;434:56–60.
18. Umeki N, Fukasawa Y, Ohtsuki S, Hori S, Watanabe Y, Kohno Y, et al. mRNA Expression and Amino Acid Transport Characteristics of Cultured Human Brain Microvascular Endothelial Cells (hBME). Drug Metabolism and Pharmacokinetics 2002;17:367–73.
19. Vumma R, Johansson J, Venizelos N. Proinflammatory Cytokines and Oxidative Stress Decrease the Transport of Dopamine Precursor Tyrosine in Human
Fibroblasts. NPS 2017;75:178–84.
20. Hwang YS, Jeong M, Park JS, Kim MH, Lee DB, Shin BA, et al. Interleukin-1β stimulates IL-8 expression through MAP kinase and ROS signaling in human gastric carcinoma cells. Oncogene 2004;23:6603–11.
21. Full Text PDF [Internet]. [cited 2018 Nov 26];Available from:
https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1399-5618.2009.00706.x 22. Licht RW. Lithium: Still a Major Option in the Management of Bipolar Disorder.
CNS Neuroscience & Therapeutics 2012;18:219–26.
23. Leucht S, Helfer B, Dold M, Kissling W, McGrath JJ. Lithium for schizophrenia. Cochrane Database of Systematic Reviews [Internet] 2015 [cited 2018 Nov 26];Available from:
https://www-cochranelibrary-com.db.ub.oru.se/cdsr/doi/10.1002/14651858.CD003834.pub3/abstract 24. Bschor T. Lithium in the Treatment of Major Depressive Disorder. Drugs
2014;74:855–62.
25. Baldessarini RJ, Tondo L, Hennen J. Lithium Treatment and Suicide Risk in Major Affective Disorders: Update and New Findings. J Clin Psychiatry 2003;64:44–52. 26. Goodwin G. Evidence-based guidelines for treating bipolar disorder: revised
second edition—recommendations from the British Association for Psychopharmacology. J Psychopharmacol 2009;23:346–88.
27. Won E, Kim Y-K. An Oldie but Goodie: Lithium in the Treatment of Bipolar Disorder through Neuroprotective and Neurotrophic Mechanisms. Int J Mol Sci [Internet] 2017 [cited 2018 Nov 27];18. Available from:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5751281/
28. Nciri R, Desmoulin F, Allagui MS, Murat J-C, Feki AE, Vincent C, et al. Neuroprotective effects of chronic exposure of SH-SY5Y to low lithium
concentration involve glycolysis stimulation, extracellular pyruvate accumulation and resistance to oxidative stress. Int J Neuropsychopharmacol 2013;16:365–76. 29. Macêdo DS, de Lucena DF, Queiroz AIG, Cordeiro RC, Araújo MM, Sousa FC, et
al. Effects of lithium on oxidative stress and behavioral alterations induced by lisdexamfetamine dimesylate: Relevance as an animal model of mania. Progress in Neuro-Psychopharmacology and Biological Psychiatry 2013;43:230–7.
30. Wang H-M, Zhang T, Li Q, Huang J-K, Chen R-F, Sun X-J. Inhibition of glycogen synthase kinase-3β by lithium chloride suppresses 6-hydroxydopamine-induced inflammatory response in primary cultured astrocytes. Neurochemistry
International 2013;63:345–53.
31. Watanabe S, Iga J, Nishi A, Numata S, Kinoshita M, Kikuchi K, et al. Microarray analysis of global gene expression in leukocytes following lithium treatment. Human Psychopharmacology: Clinical and Experimental 2014;29:190–8.
32. Moher D, Liberati A, Tetzlaff J, Altman DG, Group TP. Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLOS Medicine 2009;6:e1000097.
33. 8.7 Summary assessments of risk of bias [Internet]. [cited 2018 Dec 17];Available from:
https://handbook-5-1.cochrane.org/chapter_8/8_7_summary_assessments_of_risk_of_bias.htm 34. Ricken R, Busche M, Schlattmann P, Himmerich H, Bopp S, Bschor T, et al.
Cytokine serum levels remain unchanged during lithium augmentation of
antidepressants in major depression. Journal of Psychiatric Research 2018;96:203– 8.
35. Li H, Hong W, Zhang C, Wu Z, Wang Z, Yuan C, et al. IL-23 and TGF-β1 levels as potential predictive biomarkers in treatment of bipolar I disorder with acute manic episode. Journal of Affective Disorders 2015;174:361–6.
36. Guloksuz S, Altinbas K, Aktas Cetin E, Kenis G, Bilgic Gazioglu S, Deniz G, et al. Evidence for an association between tumor necrosis factor-alpha levels and lithium response. Journal of Affective Disorders 2012;143:148–52.
37. Guloksuz S, Aktas Cetin E, Cetin T, Deniz G, Oral ET, Nutt DJ. Cytokine levels in euthymic bipolar patients. Journal of Affective Disorders 2010;126:458–62.
38. Padmos RC, Hillegers MHJ, Knijff EM, Vonk R, Bouvy A, Staal FJT, et al. A Discriminating Messenger RNA Signature for Bipolar Disorder Formed by an Aberrant Expression of Inflammatory Genes in Monocytes. Arch Gen Psychiatry 2008;65:395–407.
39. Knijff EM, Breunis MN, Kupka RW, Wit HJD, Ruwhof C, Akkerhuis GW, et al. An imbalance in the production of IL-1β and IL-6 by monocytes of bipolar patients: restoration by lithium treatment. Bipolar Disorders 2007;9:743–53. 40. Himmerich H, Koethe D, Schuld A, Yassouridis A, Pollmächer T. Plasma levels of
leptin and endogenous immune modulators during treatment with carbamazepine or lithium. Psychopharmacology 2005;179:447–51.
41. Boufidou F, Nikolaou C, Alevizos B, Liappas IA, Christodoulou GN. Cytokine production in bipolar affective disorder patients under lithium treatment. Journal of Affective Disorders 2004;82:309–13.
42. Rapaport MH, Guylai L, Whybrow P. Immune parameters in rapid cycling bipolar patients before and after lithium treatment. Journal of Psychiatric Research
1999;33:335–40.
43. Munkholm K, Jacoby AS, Lenskjold T, Bruunsgaard H, Vinberg M, Kessing LV. Leukocytes in peripheral blood in patients with bipolar disorder – Trait and state alterations and association with levels of cytokines and C-reactive protein. Psychiatry Research 2018;261:383–90.
