University of Groningen
An inflamed mood Yang, Chenghao DOI:
10.33612/diss.98153713
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Yang, C. (2019). An inflamed mood: studies on the role of inflammation in the pathophysiology and treatment outcome of major depressive disorder. University of Groningen.
https://doi.org/10.33612/diss.98153713
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Chapter 3
Inflammatory markers and treatment outcome in
treatment resistant depression: A systematic review
Chenghao Yang, Klaas J. Wardenaar, Fokko J. Bosker, Jie Li, and Robert A. Schoevers
J Affect Disord. 2019 Jul 5;257:640-649
Abstract
Background: A substantial percentage of depressed patients do not respond satisfactorily to conventional antidepressant treatment. This treatment resistant depression (TRD) may be partly related to inflammatory processes in the central nervous system. Accordingly, peripheral inflammatory markers might serve to predict treatment response with novel but still experimental forms of antidepressant treatment.
Methods: A literature search on treatment of TRD and inflammatory markers was performed
using the PubMed/Medline database on November 8th 2018, and 95 articles were retrieved
initially, which were subsequently screened and selected only when the inclusion and exclusion criteria were met.
Results: Ten studies were recruited. In five studies higher baseline interleukin-6 (IL-6) or C-reactive protein (CRP)/high-sensitivity-CRP (hsCRP) in blood predicted better response to medication with anti-inflammatory characteristics, such as ketamine and infliximab. One study found that higher IL-6 predicted worse response to antidepressant treatment in patients with TRD. No evidence was found for the predictive value of other inflammatory markers (e.g., Tumor Necrosis Factor-α, Interferon-γ).
Limitations: The number of available studies was limited; included studies showed considerable methodological variation and used different definitions for TRD.
Conclusion: The inflammatory markers IL-6 and CRP/hsCRP could hold promise as markers for the prediction of treatment response in TRD. Clearly, this field of research is still far from mature but it could pave the way for novel and efficacious treatments for at least the inflammatory type of TRD with more well-designed studies and more convincing results.
1. Introduction
ccording to the World Health Organization (WHO), depression will become the second leading cause of disability worldwide in 2020, thus forming an increasing burden for both individuals and society [1]. The etiology of depression is still far from clear which limits the effectiveness of current treatment strategies. It has been estimated that up to 60% of depressed patients show an insufficient response to their first antidepressant treatment. Moreover, approximately 30% of the patients respond poorly to various forms of antidepressant treatment, eventually falling into the category of treatment resistant depression (TRD) [2-4]. In order to develop more efficacious antidepressant treatments, it is thus necessary to better understand the underlying pathophysiology in TRD and to identify peripheral markers of the biological processes involved.
Since early studies in the 1980s reported higher concentrations of leukocytes among depressed subjects compared to the general population [5, 6], evidence has been accumulating showing that inflammatory dysregulation may play an important role in the pathophysiology of depression, although studies have also suggested that this may be restricted to specific symptoms and possibly subtypes of depression [7]. Moreover, preclinical studies have shown that both peripheral and central inflammatory cytokines could be causal factors in developing depression-like behaviors in animals [8-10]. Inflammatory responses can be triggered by many peripheral and central processes through activation of macrophages and lymphocytes as well as microglia and astrocytes (see Figure 1) [11]. Increased production of pro-inflammatory cytokines, activation of the kynurenine pathway and the hypothalamic-pituitary-adrenal axis dysfunction have been reported to contribute to the pathophysiology of depression, likely through a decrease of neurotrophic factors and increased oxidative stress [12, 13]. Indeed, several studies have shown dysregulations of inflammatory markers in both major depressive disorder (MDD) and bipolar disorder (BD) [14-17], including dysregulations in Interleukin-6 (IL-6), Tumor Necrosis Factor alpha (TNF-α), Interferon gamma (IFN-γ) and C-reactive protein (CRP) [18-20], which are vital contributors to the immune system (see table 1). In addition, excessive central activation of the inflammatory system can interfere with various processes that are increasingly being implicated in the pathophysiology of depression such as neurotrophic support, oxidative stress, neurogenesis, and apoptosis [21-23].
Some clinical studies also indicate that increased inflammatory activity plays a role in antidepressant-treatment resistance. For instance, a study by Carvalho et al. showed that depressive patients benefit less from antidepressants in the presence of a hyperactivity of inflammatory system [24]. An animal study also indicated that overexpression of brain IL-6 in mice contributed to resistance to fluoxetine treatment [25]. Furthermore, a number of studies have demonstrated the antidepressant efficacy of anti-inflammatory drugs in TRD [26, 27]. For example, Raison and co-workers reported that the functional TNF-α antagonist infliximab significantly relieved depressive symptoms in TRD patients with higher baseline levels of high-sensitivity CRP (hsCRP) and TNF-α, while significantly reducing hsCRP compared to baseline levels [26]. Other studies, however, failed to relate antidepressant efficacy to changes of inflammatory cytokine levels [28, 29]. In addition, treatment strategies recommended for TRD are also known to possess anti-inflammatory properties, such treatments include vagal nerve stimulation, electroconvulsive therapy (ECT), lithium, and ketamine, which occupy
advantage in treatment of TRD [30, 31]. To sum up, higher inflammatory cytokines seem to contribute to treatment resistance in at least part of TRD patients. In patients with hyperactive inflammation, anti-inflammatory treatment could be a promising augmentation strategy to improve overall treatment response. However, in order for such a strategy to plausibly work, it needs to be established that (1) inflammatory biomarker levels are predictive of treatment response and that (2) anti-inflammatory treatment improves anti-depressive treatment response in patients with inflammatory hyperactivity.
However, it is noteworthy that results, in term of the role of inflammatory markers in pathophysiology of MDD/TRD, are inconsistent because of multiple confounding factors, such as sample characteristics, heterogeneity of depression, and diverse definition of TRD, which evidently limits the generalization of these findings [32, 33]. Given the inconclusive results so far it is hard to determine if and how inflammation levels are related to antidepressant treatment response in TRD and, by extension, if and how targeting of inflammation in treatment could improve treatment response in certain TRD patients. To determine which patients could potentially benefit from such treatments, it is first important to find out which inflammatory markers are actually associated with treatment response in TRD and could be screened for further clinical research. Therefore, this study aimed to systematically review the relevant available literature on this topic and to critically evaluate the role of baseline inflammatory marker levels as predictors of treatment response in TRD.
Figure 1. Potential mechanisms of inflammatory cytokine effects on brain monoamine, glutamate, and BDNF neurotransmitter systems (taken from Felger and Lotrich with permission). Peripheral cytokines can access the central nervous system and increases production of local inflammatory mediators such as cyclooxegenase-2 (COX-2), prostaglandin E2 (PGE2), nitric oxide (NO), cytokines, and chemokines by endothelial cells, perivascular macrophages, and microglia. Production of monocyte chemotactic protein-1 (MCP-1) recruits peripheral immune cells into the brain that produce even more cytokines and inflammatory mediators. Inflammatory cytokines are associated with increased oxidative stress and generation of reactive oxygen and reactive nitrogen species (ROS and RNS). Increased ROS and RNS contribute to oxidation of tetrahydrobiopterin (BH4), a cofactor required for the synthesis of monoamines. Furthermore, evidence exists indicating that inflammatory cytokines and their signal transduction pathways, such as p38 mitogen-activated protein kinases (MAPK),
may decrease expression or function of the vesicular monoamine transporter 2 (VMAT2) and/or increase expression or function of serotonin and dopamine transporters (5-HTT/DAT). Cytokines can also decrease brain-derived neurotrophic factor (BDNF) and interfere with TrkB receptor signaling, which may adversely influence neurogenesis and neuroplasticity. Finally, inflammatory cytokines can affect the glutamate (Glu) system by activation of the enzyme, indoleamine-2,3-dioxygenase (IDO), that catabolizes tryptophan, the primary amino-acid precursor of 5-HT, into kynurenine. Kynurenine is further broken down in the CNS into the neuroactive metabolites kynurenic acid and quinolinic acid (QUIN). Although not pictured, kynurenic acid can antagonize Glu receptors and decrease Glu release leading to decreased Glu neurotransmission. QUIN can directly activate the n-methyl-d-aspartate receptor (NMDAR), increase Glu release, and inhibit Glu uptake by astrocytes via the excitatory amino acid transporter (EAAT), thus allowing increased access of Glu to extrasynaptic NMDARs and contributing to excitotoxicity.
Note: 3-HAO, 3-hydroxyanthranilic acid oxygenase; 5-HT, serotonin; 5-HTT, serotonin transporter; AMPAR, 2-amino-3-(5-methyl-3-oxo-1,2- oxazol-4-yl) propanoic acid receptor; BH4, tetrahydrobiopterin; BDNF, brain-derived neurotrophic factor; COX-2, cyclooxygenase-2; DAT, dopamine transporter; glu, glutamate; EAAT, excitatory amino acid transporter; IDO, indoleamine 2,3 dioxygenase; IFN, interferon; iNOS, inducible nitric oxide synthase; IL, Interleukin; KAT II, kynurenine aminotransferase II; KMO, kynurenine 3-monooxygenase; MAPK, mitogen-activated protein kinases; MCP-1, monocyte chemotactic protein-1; NMDAR, N-Methyl-D-aspartic acid receptor; NF-kB, nuclear factor-kappa B; PGE2, prostaglandin E2; QUIN, quinolinic acid; RNS, reactive nitrogen species; ROS, reactive oxygen species; STAT, signal transducer and activator of transcription; TH, tyrosine hydroxylase; TNF, tumor necrosis factor; TrkB, tyrosine kinase receptor B; VMAT2, vesicular monoamine transporter 2.
Ta b le 1 . O v e rv iew o f TNF -α, I FN -γ, CR P an d in te rl eu ki ns: so u rc e, ta rg ets, re ce pt or s an d fu n ct io ns. C y to k in e C y to k in e S o u rc e C y to k in e T a rg et s C y to k in e R ec ep to r C y to k in e M a in Fun ct io n C y to k in e D isea se A ss o ci a ti o n T N F -α M ac ro p h ag es, C D 4 + l y mp h o cy te s, N K c el ls, n eu tr o p h il s, m ast c el ls, eo si n o p h il s, an d n eu ro n s M o st t iss u es i n t h e b o d y (T N F R 1 ) an d c el ls o f th e imm u n e sy st em ( T N F R 2 ). T N F R 1 ( C D 1 2 0 a) a n d T N F R 2 (C D 1 2 0 b ) R eg u la ti o n o f imm u n e ce ll s; c el l ap o p to si s, in d u ce s fe v er , ca ch ex ia , an d i n fl amm at io n ; in h ib it io n o f o n co g en esi s a n d v ir al r ep li ca ti o n ; re sp o n d s to se p si s v ia I L 1 & I L 6 p ro d u ci n g c el ls. ↑= p ro mo te s t he in fl amm at or y re sp on se ; c au se s man y c li n ic al p ro b le ms a sso ci at ed w it h au to imm u n e d iso rd er s; i n d u ce s fe v er , ce ll d ea th , an d s h o ck -l ik e sy mp to ms. ↓= in du ce s ca ch ex ia . IF N -γ T h c el ls, cy to to x ic T c el ls, N K ce ll s , mac ro p h ag es, an d mu co sal ep it h el ia l ce ll s h et er o d ime ri c re ce p to r co n si st in g o f IF N G R 1 /2 , g ly co sam in o g ly ca n h ep ar an su lf at e at t h e ce ll su rf ac e IF N G R 1 , IF N G R 2 A n ti v ir al , imm u n o re g u la to ry , an d a n ti -t u mo r p ro p er ti es A sso ci at ed w it h a n u mb er o f au to -i n fl amm at o ry an d a u to imm u n e d ise ase s C R P /h sC R P S y n th esi ze d b y t h e li v er i n r esp o n se to b io mark er s re le ase d b y mac ro p h ag es a n d a d ip o cy te s D amag ed c el ls, d ea d c el ls, co mp le me n t sy st em, b ac te ri a P h o sp h o ch o li n e o n t h e su rf ac e o f d ea d o r d y in g c el ls a n d so me b ac te ri a; F cg amm aR I , F cg amm aR II a, a n d F cg amm aR II b A ct iv at es t h e co mp le me n t sy st em, p ro mo te s p h ag o cy to si s o f d ea d o r d y in g c el ls (o r b ac te ri a) b y mac ro p h ag es. ↑ = in cr ea se s p ha go cy to si s a nd r el ea se o f cy to k in es; b in d s to d amag ed me m b ra n es; in cr ea se s c le ar an ce o f ap o p to ti c ce ll s; mas k s au to an ti g en s fr o m th e imm u n e sy st em o r en h an ce s t h ei r cl ea ra n ce . IL -1α M ac ro p h ag es, ep it h el ia l ce ll s, man y o th er s M ac ro p h ag es, th y mo cy te s, C N S , o th er s H as t w o si te s o f b in d in g t o IL 1 R I an d I L 1 R -A cP In fl amm at o ry ; p ro mo te s a ct iv at io n , co st imu la ti o n , an d se cr et io n o f cy to k in es a n d o th er a cu te -p h ase p ro te in s; p y ro g en ic ; k il ls a l im it ed n u mb er o f tu mo r ce ll s t y p es ↑ = i n fl amm at o ry b o n e re so rp ti o n ; g o u t; p ro mo te s T h 1 7 r esp o n se ; in te ra ct io n w it h T N F -α in v o lv in g i n i n su li n r esi st an ce IL -1β M ac ro p h ag es, ma n y o th er s M ac ro p h ag es, th y mo c y te s, C N S , o th er s IL 1 R I an d I L 1 R -A cP In fl amm at o ry ; co n tr ib u te s p ai n se n si ti v it y ; p ro m o te s a ct iv at io n , co st im u la ti o n , an d se cr et io n o f cy to k in es a n d o th er a cu te -p h ase p ro te in s; p y ro g en ic ↑ = i n fl amm at o ry b o n e re so rp ti o n ; g o u t; p ro mo te s T h 1 7 r esp o n se IL -2 T c el ls T , B , N K c el ls, an d ma cr o p h ag es IL 2R α, IL 2R β, an d IL 2R γ P ro li fe ra ti o n ; en h an ce m en t o f cy to to xi ci ty , I FN γ se cr et io n , an d a n ti b o d y p ro d u ct io n ↓ = ly mp ho pr ol if er at iv e di se ase a nd s usce pt ib ili ty to a u to imm u n e d ise ase ; re d u ce d T re g de ve lo pme nt . ↑ = r ed uc ed T h1 7 de ve lo pme nt . IL -6 M ac ro p h ag es, T c el ls, fi b ro b la st s, an d o th er s W id e v ar ie ty o f ce ll s: B c el ls, T ce ll s, th y mo c y te s, my el o id c el ls, o st eo cl ast s IL 6R α an d g p 1 3 0 In fl amm at o ry a n d c o st im u la to ry a ct io n ; in d u ce s p ro li fe ra ti o n a n d d if fe re n ti at io n ; sy n er g iz es w it h TG Fβ t o dr iv e T h1 7 ↓ = d ef ic ie nt in na te imm un it y an d ac ut p h ase re sp o n se s, ly mp h o p en ia IL -10 D if fe re n ti at ed T h el p er c el ls, Tr eg s, B c el ls, d en d ri ti c ce ll s, o th er s M ac ro p h ag es, T c el ls, d en d ri ti c ce ll s, B c el ls I L 1 0 R 1 a n d I L 1 0 R 2 Imm u n e su p p re ssi o n ; d ec re ase s a n ti g en p re se n ta ti o n a n d M H C c la ss I I ex p re ssi o n o f d en d ri ti c ce ll s; d o w n - re g u la te s p at h o g en ic T h 1 , T h 2 , an d T h 1 7 r esp o n se s ↓ = imm un e pa th ol og y du e to u nc on tr ol le d in fl amm at io n . ↑ = in hi bi ts s te ri le imm un it y to so me p at ho ge ns. N o te : I L -1 α/1 β/ 2 /6 /1 0 , In te rl eu k in -1 α/1 β /2 /6 /1 0 ; IL 1 R I, i n te rl eu k in -1 r ec ep to r, t y p e I; I L 1 R -A cP , I L -1 R a cc ess o ry p ro te in ; C N S , ce n tr al n er v o u s sy st e m; T h 1 /2 /1 7 , T h el p er 1 /2 /1 7 c el l; I L 2 R α/R β/ 2 R γ, i n te rl eu k in 2 re ce p to r α/ β/ γ; N K c el l, n at u ra l k il le r ce ll ; IF N -γ , in te rf er on γ ; T G F -β, tr an sf or mi ng g ro w th f ac to r β; M H C : maj or h ist oc om pa tib ili ty c omp le x; TN F -α, T umo r ne cr osi s fa ct or α; T N FR 1/ 2, T umo r n ec ro si s fa ct o r re ce p to r 1 /2 ; C D 1 2 0 a/ b , cl u st er o f d if fe re n ti at io n 1 2 0 a/ b ; C R P /h sC R P , C r ea ct iv e p ro te in /h ig h se n si ti v it y C R P ; IF N G R 1 /2 , in te rf er o n g amm a re ce p to r 1 .
2. Methods
The PubMed (Medline) database was used to search relevant literature using the following search strategy:
((((((((Biomarkers[MeSH Terms]) OR C-Reactive Protein[MeSH Terms]) OR Interferon-gamma[MeSH Terms]) OR Interleukins[MeSH Terms]) OR Tumor Necrosis Factor-alpha[MeSH Terms])) OR (biological marker*[Text Word] OR inflammatory marker*[Text Word] OR c-reactive protein[Text Word] OR interferon-gamma[Text Word] OR interleukins[Text Word] OR tumor necrosis factor-alpha[Text Word] OR CRP[Text Word] OR high-sensitive CRP[Text Word])))) AND ((((Treatment Resistant Depression[MeSH Terms]) OR Refractory Depression[MeSH Terms]) OR treatment resistant depression[Text Word]) OR refractory depression[Text Word])
The literature search was performed on November 8th 2018, which yielded a total number of
95 articles (see Figure 2 below for the screening process). Next, the titles and abstracts of all articles were scanned to check if they met the inclusion criteria. Inclusion criteria were that the articles should report on (1) clinical trials of treatments in TRD patients, (2) human subjects and (3) the predictive effects of baseline inflammatory markers on treatment response. The clinical trials could be single/double-blind randomized placebo-controlled trials and open-label studies. Both studies that dichotomized between responders and non-responders and those measuring severity changes over time without dichotomizing were included. Although the definitions of treatment resistance varied considerably, studies were included irrespective of the used definition. It is important to note that the intervention strategies were not restricted to monotherapy or adjunctive treatment with anti-inflammatory agents, but also included agents with unrecognized anti-inflammatory properties. If there was any doubt whether an article should be included or not, the whole text was read. Previous review studies and meta-analyses were excluded, but their reference lists were scanned for potentially interesting articles that might have been missed by the PubMed search. Reference lists of the recruited articles were also scanned for the same purpose. Studies that recruited TRD patients with severe comorbid somatic diseases, such as autoimmune diseases, unstable hypertension, cardiovascular diseases or cancer were excluded.
Figure 2. Flowchart of methodology 3. Results
3.1 Overview of included articles
Eventually 10 articles were included in the present review. Brief information of these studies is presented in table 2. In summary, a total number of 540 participants were included in the described studies, consisting of 440 patients and 100 healthy controls (sample sizes per study are shown in table 2). The most widely studied biomarkers were IL-6, TNF-alpha and CRP/hsCRP. Seven studies were carried out in patients with unipolar TRD while the other 3 studies included both unipolar and bipolar TRD. All studies used DSM-IV-(TR) criteria to diagnose the depressive episode, but the definition of TRD varied considerably between the studies (see table 3). We used the gathered information to evaluate the evidence on predictive validity of the various inflammatory markers for TRD treatment response. Finally, when reported in articles, we also reviewed the results on correlations between the change in marker levels and change in depression severity over the treatment period. Although not our primary research interest, these results were included here as well as they can provide secondary clues about the link between cytokine markers and TRD treatment response.
3.2 IL-1β
Yang et al. [27] found that baseline levels of IL-1β were significantly higher in responders compared to non-responders (p<0.05) and controls (p<0.01) in an open-label trial of ketamine infusion. They reported a significant decrease of IL-1β levels at 230 minutes and 1 day following ketamine infusion in the responder group (p=0.013) but not in the non-responder group (p>0.05).
Kiraly et al. [35] did not find a correlation between the baseline level of IL-1β and change in severity of depressive symptoms following ketamine treatment, indicating no predictive role of baseline IL-1β level for treatment response (all p>0.05). Also, Kruse et al. [36] found that baseline IL-β level was not related to changes in severity of depressive
symptoms following ECT (p>0.05), providing no evidence for a predictive role of baseline level of IL-1β for response to ECT.
3.3 IL-6
Chen et al. [37] showed that higher baseline IL-6 levels were significantly related to a better treatment response in a 0.5 mg/kg ketamine infusion group (OR: 8.93, 95% CI: 1.16–68.86), but not in a 0.2 mg/kg ketamine and placebo group. Kruse et al. [36] reported that higher baseline IL-6 levels were predictive of lower Montgomery-Åsberg Depression Rating Scale (MADRS) scores following the last ECT (p=0.01), especially in the female subgroup (p=0.02) but not the male subgroup after stratification for sex. Yang et al. [27] found significantly higher baseline IL-6 levels in responders to ketamine infusion compared to the non-responder group (p<0.01).
Kagawa et al. [29] reported that IL-6 levels did not significantly differ between responders and non-responders after lamotrigine treatment, with no predictive validity of baseline IL-6 level for treatment response (data not shown). Moreover, they found no significant correlation between the change in IL-6 level and the improvement in MADRS scores (p>0.05). Allen et al. [38] found no predictive association of IL-6 with ketamine/ECT treatment response: the correlation between the reduction in depression severity and higher baseline IL-6 was only found at 24 hours following the first infusion (p=0.007). Yoshimura et al. [39] found that baseline IL-6 levels had no predictive association with response to treatment with Selective Serotonin Reuptake Inhibitors (SSRIs)/Serotonin and Noradrenaline Reuptake Inhibitors (SNRIs). Kiraly et al. [35] found no predictive effects of baseline IL-6 levels for ketamine treatment response. Kranaster et al. [40] showed that IL-6 had no predictive value for the clinical outcomes and was not correlated with the reduction of depressive symptoms during ECT, neither in cerebrospinal fluid (CSF) nor in serum (both p>0.05).
3.4 IL-8
Allen et al. [38] found that higher baseline IL-8 levels were associated with a reduction of depression severity at 2 hours following a third ketamine infusion (p=0.048), but not at other follow-up time points. Two other studies [35, 36] found no predictive associations of baseline IL-8 levels with treatment response.
3.5 IL-10
Two studies investigated the predictive value of IL-10, both of which [35, 38] found no predictive associations between IL-10 levels and treatment response.
3.6 IFN-γ
Two studies that investigated IFN-γ [35, 38] found no associations between IFN-γ levels and treatment response.
3.7 TNF-α
Raison et al. [26] found that higher baseline TNF-α levels predicted a better treatment response to infliximab (p<0.05).
Other studies did not find any predictive associations. Chen et al. [37] found that TNF-α levels were not predictive for ketamine treatment outcome and Yoshimura et al. [39] found that baseline TNF-α levels were not predictive of SSRIs/SNRIs treatment response. Yang et al. [27] found that baseline levels of TNF-α did not differ significantly between responders and non-responders, and Kruse et al. [36] and Kiraly et al. [35] found no association between baseline TNF-α levels and treatment response.
3.8 CRP/hsCRP
Raison et al. [26] reported that baseline values of hsCRP>5mg/L predicted a greater decrease in Hamilton rating scale of depression (HAMD)-17 scores in infliximab-treated patients than placebo-treated subjects. Papakostas et al. [41] observed that changes on the HAMD-28 scoring list were significantly greater than baseline scores in a subpopulation with hsCRP levels above the study median value when comparing L-methylfolate with placebo treatment (p<0.05). Kruse et al. [36] showed that baseline CRP was not associated with changes of MADRS scores but it was significantly correlated with the end-of-treatment scores for women (p=0.04). Finally, Chen et al. [37] reported that log-transformed CRP levels were not predictive for ketamine treatment response.
3.9 Biomarker panel consisting of IL-1α, IL-2, IL-4, IL-5, IL-7, IL-9, IL-12p40, IL-12 p70, IL-13, IL-15, and TNF-α and β
Kiraly et al. [35] measured a panel of cytokines and assessed their predictive value for ketamine treatment response and found that none of the inflammatory markers possessed predictive value for ketamine treatment response.
Ta b le 2 . S u m m a ry o f re su lts Au th o r s (r ef. n r) Dia g n o sti c cr iter ion Cy to k in e m e a su re s (so u rc e) Un ip o la r/B ip o la r TRD T re a tm e n t str a te g y (a n ti -i n fla m m a to ry a ctio n s: y es/n o ) Ra tin g sc a le (N: stu d y g ro u p v s co n tr o l g ro u p ) Ad ju ste d c o n fo u n d er s (d iffer e n ce i n r esu lts a fte r a d ju st m e n t) Ou tc o m e s K iral y. DD [ 3 5 ] D SM -I V (S CID) IL -1 α, IL -1 β, IL -2 , IL -3 , IL -4 , IL -5 , IL -6 , IL -7 , IL -8 , IL -9 , IL -1 0 , IF N -α, IL -1 2 p 4 0 , IL -1 2 p 7 0 , IL -1 3 , IL -1 5 a n d T NF -α /β (se ru m ) Un ip o lar T RD Ke ta m in e/p lac eb o (y es) M A DRS (N: 3 3 v s 2 6 ) A d ju stm en t fo r co v ariate s d id in a n y c a se , n o co n fo u n d ers sh o w n (n o ) No p re d ictiv e ass o ciatio n s f o r trea tm en t re sp o n se f o r all c y to k in es; l ev els o f IL -6 , IL -1 α, IL -1 3 , IL -7 , an d IL -8 c h an g ed af ter k eta m in e in fu sio n a t v aried ti m e p o in ts b u t th ere w ere n o co rre latio n s w it h t h e ch an g es in se v erit y o f d ep re ss iv e sy m p to m s. K ra n aste r. L [4 0 ] DSM -IV IL -6 (se ru m , CS F ) Bo th ECT + p rio r m ed ica ti o n (y es) H A M D -21 (N: 1 2 , n o co n tro l g ro u p ) Be n jam in i-Ho ch b er g p ro ce d u re f o r m u lt ip le co m p ariso n s f o r CS F IL -6 (y es) No p re d ictiv e v alu e w as fo u n d f o r b o th IL -6 i n se ru m a n d CS F ; o n ly a d if fe re n ce o f CS F IL -6 c h an g e w a s f o u n d b etw ee n re m it ters a n d non -re m it ters , b u t it w as n o t co n sid ere d sig n if ica n t af ter t h e Be n jam in i-Ho ch b er g p ro ce d u re fo r m u lt ip le co m p ariso n s. Ka g a wa . S [2 9 ] D SM -IV IL -6 (se ru m ) B o th L a m o trig in e + p rio r p sy ch o ti cs (n o ) M A DRS (N: 5 6 , n o co n tro l g ro u p ) N o (N/A ) No p re d ictiv e ef fe cts we re f o u n d fo r trea tm en t re sp o n se ; n o co rre latio n w as f o u n d b etw ee n ch an g es in se v erit y o f d ep re ss io n . A ll en . A P [3 8 ] D SM -IV IL -6 , IL -8 , IL -1 0 , IF N -γ (p las m a) Un ip o lar T RD K eta m in e/E CT (y es) H A M D -17 (N: 3 7 v s 2 0 ) No (N/A ) Hig h er b ase li n e IL -6 o r IL -8 p re d icte d b ett er re sp o n se s a t sp ec if ic b u t n o t all ti m e p o in ts; n o co rre latio n s w ere f o u n d b etw ee n ch an g es in m ar k er lev el s a n d d ep re ss iv e sy m p to m s; K ru se . JL [ 3 6 ] D SM -IV -TR CR P , IL -1 β, IL -6 , IL -8 , T NF -α (p las m a) Un ip o lar T RD E CT (y es) M A DRS (N: 2 9 , n o co n tro l g ro u p ) Du ra ti o n o f cu rre n t ep iso d e, a g e, sy m p to m se v erit y a t b ase li n e (No ) Hig h er b ase li n e IL -6 a n d CR P p re d icte d g re ater i m p ro v e m en ts i n w o m en ; lev els o f CRP a n d IL -6 v aried a t d if fe re n t ti m e p o in ts b u t n o c o rre lati o n s w ere f o u n d b etw ee n c h an g es in th e m ar k ers
an d i n d ep re ss iv e sy m p to m s. C h en . M H [ 3 7 ] D SM -IV -TR CR P , IL -6 , T NF -α (se ru m ) Un ip o lar T RD Ke ta m in e/p lac eb o (y es) M A DRS (N: 4 7 v s 2 4 ) De m o g ra p h ic d ata, n o d etail sh o w n (No ) Hig h er b ase li n e IL -6 p re d icte d b ett er trea tm en t re sp o n se ; T NF -alp h a co rre late d t o c h an g es in d ep re ss iv e sy m p to m s in a sp ec if ic ti m e fra m e. Y o sh im u ra . R [3 9 ] D SM -IV IL -6 , T NF -α (p las m a) Un ip o lar T RD P ar o x eti n e, se rtralin e, flu v o x a m in e, m il n ac ip ra n (n o ) H A M D (N: 5 1 v s 3 0 ) Bo n fe rro n i ad ju stm en t, n o co n fu n d ers sh o w n in d etail (No ) Hig h er b ase li n e IL -6 p re d icte d a w o rse o u tco m e in S S RIs/S NRI trea tm en t; c h an g e in IL -6 w as re late d to t h e ch an g es o f HA M D sc o re s in S S RIs/ S NRI s-re sp o n siv e p ati en ts b u t n o t th o se th at w ere re fr ac to ry . R aiso n . CL [2 6 ] D SM -I V (S CID) h sCR P , T NF -α (p las m a) Bo th In fli x im ab /p lac eb o (y es) HA M D -1 7 a n d CG I-S (N: 3 0 v s 3 0 ) N o (N/A ) Hig h er b ase li n e h sCR P a n d T NF -alp h a p re d icte d b ett er re sp o n se s to in fli x im ab trea tm en t; g re ater d ec re ase in h sCR P c o rre late d w it h b ett er re sp o n se to i n fli x ima b ; Y an g . JJ [ 2 7 ] D SM -IV IL -1 β, IL -6 , T NF -α (se ru m ) Un ip o lar T RD Ke ta m in e/p lac eb o (y es) M A DRS (N: 1 6 v s 2 4 ) N o (N/A ) T h e h ig h er b ase li n e IL -1 β o r IL -6 p re d icte d re sp o n se t o k etam in e trea tm en t; th e ch an g es in IL -1 β or IL -6 re late d to c h an g es in d ep re ss iv e sy m p to m s. P ap ak o sta s. G I [4 1 ] D SM -IV h sCR P (se ru m ) Un ip o lar T RD L -M eth y lf o late /p lac eb o (y es) H A M D -7 /2 8 (N: 4 7 v s 2 8 ) Ag e, se x , ra ce , BM I, an d b ase li n e lev el o f HA M D -28 (No ) Hig h er b ase li n e h sCR P p re d icte d sig n if ica n tl y g re ater i m p ro v e m en ts o n HA M D -2 8 . No d ata w as p ro v id ed f o r co rre latio n s. No te: HA M D -1 7 /2 1 /2 8 : Ha m il to n ra ti n g sc ale o f d ep re ss io n -1 7 /2 1 /2 8 it em s; E CT : ele ctro co n v u lsiv e th era p y ; CS F : ce re b ro sp in al flu id ; M A DRS: M o n tg o m er y -A sb er g De p re ss io n Ra ti n g S ca le; CG I-S : Cli n ica l G lo b al Im p re ss io n S ev erit y ; IL : In terle u k in ; T NF -α /β : T u m o r Ne cro sis F ac to r alp h a/b eta; IF N -γ: In terf ero n ga m m a; CRP : C -re ac ti v e p ro tein ; h sCR P : h ig h -se n siti v it y CRP ; T RD: trea tm en t re sista n t d ep re ss io n ; S S RIs : S elec ti v e S ero to n in R eu p tak e In h ib it o rs; S NRIs : S er o to n in a n d N o ra d re n al in e Re u p tak e In h ib it o rs .
T a b le 3 . S u p p le m e n t o f re su lts Au th o r s (r ef. n r) De fin iti o n o f TRD Cr iter io n o f fa il u re t o r es p o n d De g re e o f re sista n ce a n d se v e rity o f d e p re ssio n K iral y . DD [ 3 5 ] n o t re sp o n d ed to a t lea st t w o th era p eu ti c tri als o f an a n ti d ep re ss an t ac co rd in g to th e crit eria o f th e A T HF re sista n ce ra ti n g o f 3 o r h ig h er ac co rd in g to A THF ≥2 ty p es o f an ti d ep re ss an ts as se ss ed w it h A T HF K ra n aste r. L [ 4 0 ] fa il ed to a ch iev e re sp o n se o r re m issio n t o a t lea st t w o p ro v en an ti d ep re ss an t tr ials w it h a d eq u at e d o sin g a n d d u ra ti o n fa il ed to a ch iev e re sp o n se o r re m issio n ( n o d etail p ro v id ed ) ≥2 ty p es o f an ti d ep re ss an ts Ka g a wa . S [2 9 ] in su ff icie n t re sp o n se to a t lea st 3 p sy ch o tro p ics in cl u d in g an ti d ep re ss an ts, m o o d sta b il ize rs, o r aty p ica l an ti p sy ch o ti cs d esp it e su ff icie n t d o se a n d d u ra ti o n In su ff icie n t re sp o n se (n o d etail p ro v id ed ) ≥3 p sy ch o tro p ics ; M A DRS ≥2 1 A ll en . A P [3 8 ] fa il ed to re sp o n d t o a t lea st t w o a d eq u ate t rials o f an ti d ep re ss an t m ed ica ti o n , as a ss ess ed w it h a m o d if ied v ersio n o f th e A n ti d ep re ss an t T re at m en t Histo ry F o rm re sista n ce ra ti n g o f 3 o r h ig h er ac co rd in g to A THF ≥2 ty p es o f an ti d ep re ss an ts as se ss ed w it h AT HF K ru se . JL [ 3 6 ] tw o p rio r m ajo r d ep re ss iv e ep iso d es a t lea st, an d f ail ed to re sp o n d t o a t lea st t w o p ri o r an ti d ep re ss an t m ed ica ti o n s fa il ed to a ch iev e re sp o n se (n o d et ail p ro v id ed ) ≥2 ty p es o f an ti d ep re ss an ts an d c o n firm ed w it h M INI C h en . M H [ 3 7 ] fa il ed to re sp o n d t o m o re th an tw o a d eq u ate an ti d ep re ss an t tri als fa il ed to a ch iev e re sp o n se (n o d et ail p ro v id ed ) ≥2 ty p es o f an ti d ep re ss an ts; M A DRS ≥1 8 Yo sh im u ra . R [3 9 ] HA M D sc o re s d ec re ase d les s th an 5 0 % f ro m b ase li n e HA M D sc o re s d ec re ase d les s th an 5 0 % a fter trea ti n g w it h S S RIs/S NRIs S S RIs/S NRIs re fra cto ry d ep re ss io n if HA M D sc o re s d ec re ase d les s th an 5 0 % af ter t re ati n g w it h S S RIs/S NRIs R aiso n . CL [ 2 6 ] M G H -S ≥2 a n d QID S -SR -16 ≥1 4 i n th e cu rre n t ep iso d e M G H -S ≥2 in th e cu rre n t ep iso d e MG H -S ≥2 a n d QID S -SR -16 ≥14 Y an g . JJ [ 2 7 ] Im p ro v ed les s th an 5 0 % a fter k eta m in e trea tm en t h ad a les s th an 5 0 % im p ro v e m en t af ter k eta m in e trea tm en t No re sp o n se to k etam in e trea tm e n t P ap ak o sta s. G I [4 1 ] F ail ed to m o re th an 2 a d eq u ate an ti d ep re ss an t tri als d u ri n g th e cu rre n t ep is o d e n o i n fo rm ati o n w as p ro v id ed f o r fa il u re to re sp o n d No re sp o n se s to a d eq u ate S S RIs tre at m en ts ass ess ed b y AT RQ; an d QID S -SR ≥1 2 No te: A T HF: A n ti d ep re ss an t T re at m en t Histo ry F o rm ; H A M D: Ha m il to n ra ti n g sc ale o f d ep re ss io n ; M A DRS: M o n tg o m er y -As b erg De p re ss io n Ra ti n g S ca le; M INI: M in i-In tern ati o n al Ne u ro p sy ch iatric In terv ie w ; QID S -SR -1 6 : Qu ick In v en to ry o f De p re ss iv e S y m p to m ato lo g y -se lf re p o rt -1 6 ; A TRQ: M ass ac h u se tt s G en era l Ho sp it al A n ti d ep re ss an t T re at m en t Re sp o n se Qu estio n n aire; M G H -S : M ass ac h u se tt s G en era l Ho sp it al S tag in g .
4. Discussion
The present review is aimed to evaluate whether baseline inflammatory marker levels are associated with treatment response in TRD. Regarding the predictive validity of IL-1β, IL-8, IL-10, INF-γ and TNF-α baseline levels results were mostly negative. However, several studies on IL-6 and CRP/hsCRP did find predictive associations between marker levels and treatment response. Here, studies found that baseline levels were associated with better response to compounds with known anti-inflammatory characteristics, such as infliximab and ketamine. However, one study found that higher baseline levels of IL-6 were associated with worse response to SSRIs/SNRIs treatment.
The concept of TRD was already introduced in 1974 [42, 43], but there is still no consensus on the definition of TRD [44, 45]. In five of the ten included studies patients were considered to be treatment resistant when not sufficiently responding to at least 2 types of antidepressants, which is in line with the most widely applied definition of TRD in clinical research. The other included studies used different criteria such as resistance to at least 3 psychotropics [29], no response to ketamine treatment [27] or scores on the Massachusetts General Hospital Staging (MGH-S) and Quick Inventory of Depressive Symptomatology-selfreport-16 (QIDS-SR-16) lists [26]. The divergent definitions may have partly contributed to the sometimes conflicting outcomes in the TRD studies, but the variety in study methods and designs may also have played a role.
Increased inflammation has been shown to contribute to the treatment resistance in depression [24, 39, 46]. Accordingly, agents with anti-inflammatory properties would be promising options for treating TRD. This is supported by most of the studies with the rapid acting antidepressant ketamine, which also has anti-inflammatory properties [47, 48]. In these studies, the IL-6 baseline concentration showed predictive validity for ketamine treatment response. Interestingly, these studies also found that the marker levels changed during the course of treatment [27, 36, 37]. However, in contrast with these studies, Allen et al. found no change of IL-6 levels with ketamine infusion or ECT, notwithstanding significant decreases of the HAMD-17 scores [38]. A parsimonious explanation would be that in the latter study a different patient population was investigated with relatively more patients with non-inflammation-driven depression, because inflammatory hypothesis contributes partly to the heterogeneous etiology of depression [33]. The latter idea is supported by the fact that in the negative studies, IL-6 levels were not significantly different between healthy controls and patients or between responders and non-responders [28, 29, 38]. In contrast, studies with positive results had significant differences in IL-6 levels when comparing remitters and non-remitters or responders and non-responders [27, 36, 37]. This finding is also in line with the study by Raison et al. showing that TRD patients with higher inflammatory activity respond better to the functional TNF-α antagonist infliximab than those with lower levels of inflammation [26]. Moreover, when patients were stratified by their baseline levels of hsCRP/TNF-α, higher levels of inflammation predicted lower HAMD-17 scores with infliximab intervention.
It is important to note that in intervention studies with agents that are assumed to possess anti-inflammatory properties, including ketamine, infliximab, L-methylfolate, and ECT, higher levels of inflammatory markers were indeed often found to predict outcomes of
TRD treatment. However, in those studies with agents that are assumed to have no anti-inflammatory properties, such as lamotrigine (Kagawa et al., 2017), erythropoietin [28], and SSRIs/SNRIs used in recruited studies, no relationship was found between levels of inflammatory markers and outcomes of TRD treatment, safe for one study [39], where IL-6 was found to be associated with worse outcome in SSRIs/SNRIs treatment. The differential outcomes presented above indicate that agents with an anti-inflammatory property could affect the inflammatory hyperactivity and also alleviate depressive symptoms. At the same time, the findings described above could be taken as evidence for the involvement of inflammatory process in the maintenance of depressive symptoms, and thus, a possible role of these markers as predictors of symptom persistence and TRD treatment outcome.
A factor that could have played a role in the observed inconsistency of research results is the measurement of depression severity. The enrolled studies measured severity with the MADRS or different versions of HAMD, the latter being widely used as the standard for assessing treatment responses in clinical trials of depression. However, the HAMD has been criticized for its lack of sensitivity to detect clinical changes and its limited power to discriminate remission, although it performs better in detecting physical symptoms than the MADRS [49]. Furthermore, the longer HAMD-28 version is more sensitive to detect changes in atypical or melancholic symptoms than the HAMD-17 [50, 51]. The shorter HAMD-7 version, on the other hand, has less somatic and anxiety related items thus focusing on the core symptoms of depression, which is believed to be more sensitive for changes in clinical trials of depression [52]. In view of the higher incidence of somatic problems in patients with bipolar depression, the HAMD may also be a better choice than the MADRS when both unipolar and bipolar depressed patients are included in TRD studies [53, 54]. Although it is not possible to determine the exact role of the rating scales in the treatment outcomes, it is not inconceivable that the different sensitivities of the scales have contributed to the divergent results from the included TRD studies. Other confounding factors could be sex and the type of depression (unipolar vs. bipolar), which both have been reported to influence the state of the immune system [55-58]. Three studies have included patients with treatment resistant bipolar depression, but only the ECT study by Kranaster et al. hinted at an effect of bipolarity in treatment response. The other two groups used alternative forms of treatment and could not demonstrate significant differences in the percent of improvement or in response rate among patients with major depressive disorder, bipolar I disorder and bipolar II disorder [29] or did not provide data on this topic [26]. Although previous studies indicated a difference in the level of inflammatory cytokines and change of concentrations after treatment in unipolar and bipolar depression, the limited available data did not allow us to check if specific cytokines were predictive of or changed in specific types of depression. Given the potential clinical relevance of such associations, it would be worthwhile to investigate this by measuring and comparing a broad range of cytokines in patients with different types of TRD. A role of sex was reported in two studies. Kruse et al. found that higher IL-6 and CRP levels were significant predictors of a lower end-point MADRS following ECT for women (p=0.02) but not men (p=0.10). This result is broadly compatible with the outcome of the study by Kranaster et al. showing that the remitters with ECT consisted of a significantly higher percentage of females (p=0.0028) than the non-remitters. Arguably, female patients with increased inflammation respond more positively to ECT, but it is also possible that depressed
patients with elevated inflammatory activity respond less robustly to antidepressant medications [59, 60].
There were still multiple variables that contributed to the inconsistency of findings such as sample size, bio-sample source, and study arm. First, two out of three studies with ECT concluded that higher baseline IL-6 level predicted better treatment response while one study with a small sample size (n=12) and a single arm did not show such an association [36, 38, 40]. The latter study was also unique in that it took both serum and CSF samples. Here, CSF IL-6 was found to be predictive of treatment response, but no longer after adjustment for multiple comparisons [40]. Second, three out of four studies with ketamine showed predictive value of higher IL-6 level at baseline for treatment outcome [27, 37, 38]. The remaining, fourth study by Kiraly et al. did not find a predictive effect, which could be due to the fact that this was the only study to include patients with somatic diseases like hypertension, hyperlipidemia and type-II diabetes [35], which are all known to have an effect on the inflammatory system [61-63]. The latter could explain the discrepant findings across ketamine studies, although it is difficult to discriminate exactly between the influences of different variables on the total observed predictive effect. Finally, differences across studies in treatment intensity may also have contributed to varied outcomes. For instance, in the ECT studies, patients in the study by Kranaster et al. had an average of 10.6 sessions of ECT while patients in the study by Kruse. et al. had an average of 11.5 sessions. It is possible that this also explains some of the differences between study outcomes.
Our PubMed search was restricted to cytokines and hsCRP as putative markers for neuro-inflammation. Yet several studies have used alternative approaches to assess inflammatory activity such as diffusion tensor imaging [64], glucose and lipid-metabolism related biomarkers [65], biomarkers of macrophage/microglia activation such as neopterin, sCD14, sCD163 [40], receptors of pro-inflammatory cytokines and the kynurenine pathways [38]. Interestingly most of these studies yielded positive results with respect to predictive validity, indicating that measuring blood borne cytokines and hsCRP may not be the final answer to assess the inflammation driven type of TRD. Still, the current focus was motivated by our interest in the use of markers that are easily measured in a routine clinical practice.
5. Limitations
There are several limitations in this review. First of all, there is a lack of consensus about the definition of TRD. Divergent definitions of TRD could implicate that the severity and resistance spectrum varies among the studies, which complicates comparisons of the included studies considerably. Other limitations are the relatively small number of studies and the varying research conditions, such as type of intervention, number of inflammatory markers investigated, type of depression and the rating scale being used. Furthermore, the studies showing predictive validity of cytokine markers all were quite small (six studies containing 355 participants in total), which suggests that larger studies are needed. The studies varied in their analytical approaches and not all studies documented full quantitative results, including statistics, descriptive information (e.g., means, standard deviations) of marker levels in responders and non-responders. Also, effect sizes were not reported by all authors, making
comparison across studies harder. Finally, not all studies adjusted for variations in patient characteristics (e.g. metabolic syndrome) in the data analyses.
6. Conclusion and clinical implications
Notwithstanding the limitations of the present review the tentative conclusion can be drawn that the inflammatory markers IL-6 and CRP/hsCRP hold some promise as markers for the prediction of treatment response in TRD. Furthermore, according to our previous review study, different subtypes of MDD show differences in their inflammatory profiles, like 6 and IL-1β for melancholic depression and CRP for non-melancholic depression [33]. Combined with the finding of the current study, we argue that predictive ability of specific cytokines for treatment response could differ across subtypes of depression. This field of research is still far from mature and several hurdles have to be taken before adjuvant or monotherapy with anti-inflammatory agents will become a serious option in the treatment of TRD. Firstly, it is important to reach worldwide consensus on the definition of TRD. The next step should be well-designed intervention studies using a broad panel of inflammatory markers enabling to investigate the complex relation between inflammation and depression. This could pave the way for novel and efficacious treatments for at least the inflammatory type of TRD with more well-designed studies and more convincing results.
Acknowledgements
The work was supported by funds from the Tianjin Finance Bureau and Tianjin Key Programs for Science and Technology Development in Health Industry (No.13KG118) for the conduct of the preparation of the article, but the funding did not play the role in the decision to submit the article for publication or other issues.
3. Scott, J. and B. Dickey, Global burden of depression: the intersection of culture and medicine. Br J Psychiatry, 2003. 183: p. 92-4.
4. Gibson, R.C., J.S. Martin, and S.M. Neita, Mental illness and public health: exploring the role of general hospital physicians at a teaching hospital in Jamaica. West Indian Med J, 2010. 59(6): p. 662-7.
5. Kronfol, Z., Multiple sclerosis and depression. Arch Neurol, 1985. 42(4): p. 310.
6. Kronfol, Z., et al., Depression, cortisol metabolism and lymphocytopenia. J Affect Disord, 1985. 9(2): p. 169-73. 7. White, J., et al., Association of inflammation with specific symptoms of depression in a general population of older
people: The English Longitudinal Study of Ageing. Brain Behav Immun, 2017. 61: p. 27-30.
8. Yirmiya, R., Endotoxin produces a depressive-like episode in rats. Brain Res, 1996. 711(1-2): p. 163-74.
9. Plata-Salaman, C.R. and J.P. Borkoski, Centrally administered bacterial lipopolysaccharide depresses feeding in rats. Pharmacol Biochem Behav, 1993. 46(4): p. 787-91.
10. Kent, S., et al., Different receptor mechanisms mediate the pyrogenic and behavioral effects of interleukin 1. Proc Natl Acad Sci U S A, 1992. 89(19): p. 9117-20.
11. Felger, J.C. and F.E. Lotrich, Inflammatory cytokines in depression: neurobiological mechanisms and therapeutic implications. Neuroscience, 2013. 246: p. 199-229.
12. Quan, N. and W.A. Banks, Brain-immune communication pathways. Brain Behav Immun, 2007. 21(6): p. 727-35. 13. Sas, K., et al., Mitochondria, metabolic disturbances, oxidative stress and the kynurenine system, with focus on
neurodegenerative disorders. J Neurol Sci, 2007. 257(1-2): p. 221-39.
14. Soczynska, J.K., et al., The effect of tumor necrosis factor antagonists on mood and mental health-associated quality of life: novel hypothesis-driven treatments for bipolar depression? Neurotoxicology, 2009. 30(4): p. 497-521.
15. Rosenblat, J.D., et al., Inflamed moods: a review of the interactions between inflammation and mood disorders. Prog Neuropsychopharmacol Biol Psychiatry, 2014. 53: p. 23-34.
16. Hiles, S.A., et al., A meta-analysis of differences in IL-6 and IL-10 between people with and without depression: exploring the causes of heterogeneity. Brain Behav Immun, 2012. 26(7): p. 1180-8.
17. Howren, M.B., D.M. Lamkin, and J. Suls, Associations of depression with C-reactive protein, IL-1, and IL-6: a meta-analysis. Psychosom Med, 2009. 71(2): p. 171-86.
18. Blume, J., S.D. Douglas, and D.L. Evans, Immune suppression and immune activation in depression. Brain Behav Immun, 2011. 25(2): p. 221-9.
19. Miller, A.H., V. Maletic, and C.L. Raison, Inflammation and its discontents: the role of cytokines in the pathophysiology of major depression. Biol Psychiatry, 2009. 65(9): p. 732-41.
20. Simon, N.M., et al., A detailed examination of cytokine abnormalities in Major Depressive Disorder. Eur Neuropsychopharmacol, 2008. 18(3): p. 230-3.
21. Goshen, I., et al., A dual role for interleukin-1 in hippocampal-dependent memory processes. Psychoneuroendocrinology, 2007. 32(8-10): p. 1106-15.
22. Ben Menachem-Zidon, O., et al., Intrahippocampal transplantation of transgenic neural precursor cells overexpressing interleukin-1 receptor antagonist blocks chronic isolation-induced impairment in memory and neurogenesis. Neuropsychopharmacology, 2008. 33(9): p. 2251-62.
23. Wu, C.W., et al., Treadmill exercise counteracts the suppressive effects of peripheral lipopolysaccharide on hippocampal neurogenesis and learning and memory. J Neurochem, 2007. 103(6): p. 2471-81.
24. Carvalho, L.A., et al., Lack of clinical therapeutic benefit of antidepressants is associated overall activation of the inflammatory system. J Affect Disord, 2013. 148(1): p. 136-40.
25. Sukoff Rizzo, S.J., et al., Evidence for sustained elevation of IL-6 in the CNS as a key contributor of depressive-like phenotypes. Transl Psychiatry, 2012. 2: p. e199.
26. Raison, C.L., et al., A randomized controlled trial of the tumor necrosis factor antagonist infliximab for treatment-resistant depression: the role of baseline inflammatory biomarkers. JAMA Psychiatry, 2013. 70(1): p. 31-41. 27. Yang, J.J., et al., Serum interleukin-6 is a predictive biomarker for ketamine's antidepressant effect in
treatment-resistant patients with major depression. Biol Psychiatry, 2015. 77(3): p. e19-e20.
28. Vinberg, M., et al., Effect of recombinant erythropoietin on inflammatory markers in patients with affective disorders: A randomised controlled study. Brain Behav Immun, 2016. 57: p. 53-57.
29. Kagawa, S., et al., Both Serum Brain-Derived Neurotrophic Factor and Interleukin-6 Levels Are Not Associated with Therapeutic Response to Lamotrigine Augmentation Therapy in Treatment-Resistant Depressive Disorder. Neuropsychobiology, 2017. 75(3): p. 145-150.
14(3): p. 716-727.
31. Beurel, E. and R.S. Jope, Inflammation and lithium: clues to mechanisms contributing to suicide-linked traits. Transl Psychiatry, 2014. 4: p. e488.
32. Allison, D.J., B. Sharma, and B.W. Timmons, The efficacy of anti-inflammatory treatment interventions on depression in individuals with major depressive disorder and high levels of inflammation: A systematic review of randomized clinical trials. Physiol Behav, 2019. 207: p. 104-112.
33. Yang, C., et al., Interleukin, tumor necrosis factor-alpha and C-reactive protein profiles in melancholic and non-melancholic depression: A systematic review. J Psychosom Res, 2018. 111: p. 58-68.
34. Marnell, L., C. Mold, and T.W. Du Clos, C-reactive protein: ligands, receptors and role in inflammation. Clin Immunol, 2005. 117(2): p. 104-11.
35. Kiraly, D.D., et al., Altered peripheral immune profiles in treatment-resistant depression: response to ketamine and prediction of treatment outcome. Transl Psychiatry, 2017. 7(3): p. e1065.
36. Kruse, J.L., et al., Inflammation and Improvement of Depression Following Electroconvulsive Therapy in Treatment-Resistant Depression. J Clin Psychiatry, 2018. 79(2).
37. Chen, M.H., et al., Rapid inflammation modulation and antidepressant efficacy of a low-dose ketamine infusion in treatment-resistant depression: A randomized, double-blind control study. Acta Psychiatr Scand, 2018. 269: p. 207-211.
38. Allen, A.P., et al., Kynurenine pathway metabolism and the neurobiology of treatment-resistant depression: Comparison of multiple ketamine infusions and electroconvulsive therapy. J Psychiatr Res, 2018. 100: p. 24-32. 39. Yoshimura, R., et al., Higher plasma interleukin-6 (IL-6) level is associated with SSRI- or SNRI-refractory
depression. Prog Neuropsychopharmacol Biol Psychiatry, 2009. 33(4): p. 722-6.
40. Kranaster, L., et al., Antidepressant efficacy of electroconvulsive therapy is associated with a reduction of the innate cellular immune activity in the cerebrospinal fluid in patients with depression. World J Biol Psychiatry, 2018. 19(5): p. 379-389.
41. Papakostas, G.I., et al., Effect of adjunctive L-methylfolate 15 mg among inadequate responders to SSRIs in depressed patients who were stratified by biomarker levels and genotype: results from a randomized clinical trial. J Clin Psychiatry, 2014. 75(8): p. 855-63.
42. Heimann, H., Therapy-resistant depressions: symptoms and syndromes. Contributions to symptomatology and syndromes. Pharmakopsychiatr Neuropsychopharmakol, 1974. 7(3): p. 139-44.
43. Lehmann, H.E., Therapy-resistant depressions - a clinical classification. Pharmakopsychiatr Neuropsychopharmakol, 1974. 7(3): p. 156-63.
44. Fava, M. and K.G. Davidson, Definition and epidemiology of treatment-resistant depression. Psychiatr Clin North Am, 1996. 19(2): p. 179-200.
45. Berman, R.M., M. Narasimhan, and D.S. Charney, Treatment-refractory depression: definitions and characteristics. Depress Anxiety, 1997. 5(4): p. 154-64.
46. Strawbridge, R., et al., Inflammation and clinical response to treatment in depression: A meta-analysis. Eur Neuropsychopharmacol, 2015. 25(10): p. 1532-43.
47. Clarke, M., et al., Ketamine modulates hippocampal neurogenesis and pro-inflammatory cytokines but not stressor induced neurochemical changes. Neuropharmacology, 2017. 112(Pt A): p. 210-220.
48. Li, Y., et al., Effects of Ketamine on Levels of Inflammatory Cytokines IL-6, IL-1beta, and TNF-alpha in the Hippocampus of Mice Following Acute or Chronic Administration. Front Pharmacol, 2017. 8: p. 139.
49. Ballesteros, J., et al., Sensitivity to change, discriminative performance, and cutoff criteria to define remission for embedded short scales of the Hamilton depression rating scale (HAMD). J Affect Disord, 2007. 102(1-3): p. 93-9. 50. Nemeroff, C.B., The burden of severe depression: a review of diagnostic challenges and treatment alternatives. J
Psychiatr Res, 2007. 41(3-4): p. 189-206.
51. Cusin, C., H. Yang, and A. Yeung, Rating Scale for Depression. Handbook of Clinical Rating Scale and Assessment in Psychiatry and Mental Health. 2010, New York: NY: Humana Press.
52. McIntyre, R.S., et al., Measuring the severity of depression and remission in primary care: validation of the HAMD-7 scale. Cmaj, 2005. 173(11): p. 1327-34.
53. Chauvet-Gelinier, J.C., et al., [Bipolar disorders and somatic comorbidities: a focus on metabollic syndrome, diabetes and cardiovascular disease]. Encephale, 2012. 38 Suppl 4: p. S167-72.
54. Vancampfort, D., et al., Sedentary behavior and physical activity levels in people with schizophrenia, bipolar disorder and major depressive disorder: a global systematic review and meta-analysis. World Psychiatry, 2017. 16(3): p. 308-315.
55. Goldsmith, D.R., M.H. Rapaport, and B.J. Miller, A meta-analysis of blood cytokine network alterations in psychiatric patients: comparisons between schizophrenia, bipolar disorder and depression. Mol Psychiatry, 2016. 21(12): p. 1696-1709.
58. Mao, R., et al., Different levels of pro- and anti-inflammatory cytokines in patients with unipolar and bipolar depression. J Affect Disord, 2018. 237: p. 65-72.
59. Eller, T., et al., Pro-inflammatory cytokines and treatment response to escitalopram in major depressive disorder. Prog Neuropsychopharmacol Biol Psychiatry, 2008. 32(2): p. 445-50.
60. Lanquillon, S., et al., Cytokine production and treatment response in major depressive disorder. Neuropsychopharmacology, 2000. 22(4): p. 370-9.
61. Prado, N.J., et al., Anti-Inflammatory Effects of Melatonin in Obesity and Hypertension. Curr Hypertens Rep, 2018. 20(5): p. 45.
62. Paudel, Y.N., et al., "2-(4-Fluorobenzamido)-4-methylthiazole-5-carboxylic acid" a novel thiazole compound, ameliorates insulin sensitivity and hyperlipidaemia in streptozotocin-induced diabetic rats: Plausible role of inflammatory and oxidative stress markers. Biomed Pharmacother, 2017. 95: p. 1232-1241.
63. Klisic, A., et al., Relationship between Oxidative Stress, Inflammation and Dyslipidemia with Fatty Liver Index in Patients with Type 2 Diabetes Mellitus. Exp Clin Endocrinol Diabetes, 2018. 126(6): p. 371-378.
64. Grieve, S.M., et al., Prediction of nonremission to antidepressant therapy using diffusion tensor imaging. J Clin Psychiatry, 2016. 77(4): p. e436-43.
65. Bekhbat, M., et al., Glucose and lipid-related biomarkers and the antidepressant response to infliximab in patients with treatment-resistant depression. Psychoneuroendocrinology, 2018. 98: p. 222-229.
