Lack of clinical therapeutic benefit of antidepressants is associated overall activation of the inflammatory system

Journal of Affective Disorders 148 (2013) 136–140

Contents lists available at SciVerse ScienceDirect

Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad

Brief report

Lack of clinical therapeutic benefit of antidepressants is associated overall activation of the inflammatory system L.A. Carvalho a,b,n, J.P. Torre b, A.S. Papadopoulos c, L. Poon c, M.F. Juruena c, K. Markopoulou c, A.J. Cleare c, C.M. Pariante a a

Section of Perinatal Psychiatry & Stress, Psychiatry and Immunology Laboratory King’s College London, Institute of Psychiatry, London, UK Department of Epidemiology and Public Health, University College London, London, WC1E 7HB, UK c Affective Disorders Unit, Bethlem Royal Hospital, London, UK b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 3 August 2012 Received in revised form 26 October 2012 Accepted 27 October 2012 Available online 27 November 2012

Despite the evidence of an association between depression and increased inflammatory markers, still little is known in relation to the most severe cases of the disorder i.e., those who fail to respond to antidepressants. We have assessed the cytokine profile and cortisol levels in 21 healthy controls (HC) and 19 medicated patients with depression with treatment-resistance (TRD) moderately ill. As an initial exploratory analysis, we have also related cytokine profile to the patient’s clinical treatment outcome after an inpatient admission. Cytokine profile was measured in the serum by the Cytokine Array I kit (Randoxs). Plasma cortisol was carried out using a commercially available for the IMMULITEs system. When compared to healthy controls, depressed patients had higher levels of cortisol, IL-6, IL-10, but lower levels of IL-4 and VEGF. Our exploratory analysis showed subjects who did not go on to respond to the inpatient admission treatment package had lower levels of MCP-1, and a trend toward lower levels of VEGF. Taking together, these data suggest that lack of clinical therapeutic benefit of antidepressants is associated with overall activation of the inflammatory system. & 2012 Elsevier B.V. All rights reserved.

Keywords: Antidepressant agents Biological markers Cytokines Neuroinflammation Psychoneuroimmunology Endophenotype

1. Introduction There is now evidence that some patients with major depression (MDD) present inflammatory activation even in the absence of physical illnesses. Meta-analysis studies confirmed the involvement of inflammatory cytokines in MDD patients (Dowlati et al., 2010; Hiles et al., 2012; Howren et al., 2009). The evidence for inflammatory changes in the brain in depression suggests that an increase in inflammation-induced apoptosis, together with a reduction in the synthesis of neurotrophic factors caused by a rise in brain glucocorticoids, may play a role in the pathology of these disorders. If this is the case, it is expected that the more severe cases of depression – those who are resistant to antidepressant treatment – present further inflammatory disturbance. Fewer studies have attempted understand whether a pattern of cytokines could present a biomarkers of treatment response at baseline. To our knowledge, only two studies so far have investigated cytokine levels in patients who were refractory to antidepressants. O’Brien et al. (2007) showed that antidepressants reduce cytokine

n Corresponding author at: Department of Epidemiology and Public Health, University College London, 1–19 Torrington Place, London WC1E 7HB, Rm356a, UK. Tel.: þ44 20 7679 5973; fax: þ44 20 7813 0242. E-mail address: [email protected] (L.A. Carvalho).

0165-0327/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jad.2012.10.036

levels only in those who respond to antidepressant treatment. Yoshimura et al. (2009) also showed that refractoriness to antidepressant treatment is associated with higher levels of IL-6 levels. Nevertheless, two recent meta-analyses have confirmed the antiinflammatory role of some classes of antidepressants (Hannestad et al., 2011). They did not, however, characterize treatment-resistant or-responsive patients under a biological perspective. In the present study, we have investigated the cytokine profile of patients with historically defined treatment-resistant major depression. We have also conducted exploratory analysis to understand whether cytokine profile at baseline could vary in relation to treatment outcome.

2. Materials and methods The study protocol was approved by the Research Ethics Committee of the Institute of Psychiatry, King’s College London and Maudsley Hospital (London). All subjects gave their written and informed consent. Participants: Twenty-one healthy controls were recruited through members of the local community, and free of any self-reported psychiatric illness. Concomitantly, nineteen treatment-resistant depressed (MDD) inpatients were examined one-two weeks after their admission at the National Affective Disorders Unit, Bethlem Royal Hospital, London, UK. Inclusion criteria were: (1) receipt of a

L.A. Carvalho et al. / Journal of Affective Disorders 148 (2013) 136–140

period of intensive inpatient treatment (rather than assessment); (2) diagnosis of a primary affective disorder; (3) failure to respond to at least one prior adequate medication trial; and (4) HRSD score of Z16 on admission. At admission, all patients underwent intensive psychopharmacology in-patient treatment using combinations of medications as indicated by the Maudsley prescribing guidelines (Taylor et al., 2007). Detailed evaluation of this specialist inpatient treatment for refractory affective disorder unit is extensively described (Wooderson et al., 2011). Some of these patients have been used on a previous study in treatment-resistance patients (Carvalho et al., 2008). Clinical assessment: Diagnoses were made according to the 10th revision of the International Classification of Diseases (ICD-10) (World Health Organisation, 1992) and use of the SCID-II (Structured Clinical Interview for DSM-IV, version II). Exclusion criteria for participants were: history of hypersensitivity to corticosteroids or steroid use; heavy smokers (i.e., more than 20 cigarettes/day); drugs known to modify immune and endocrine functions for at least one month before blood sampling, including oral contraceptives, pregnant/lactating women; alcohol dependence; and significant physical illnesses. Healthy controls were further excluded if they used any psychotropic medications. For ethical reasons, it was not possible to withdraw antidepressants and assess the patients in a drug free state; however, a switch in medication was avoided for at least 14 days before experimental procedures. Treatment-resistance: Further measures of treatment-resistance involved historical assessment by Sackeim’s definition, whereby resistance to a given treatment is concluded if, despite continued adherence to the same medication and dosage that produced an initial response, a patient experienced relapse or recurrence of a depressive episode (Sackeim, 2001). We also used Thase and Rush (1997) staging criteria, which recognises five stages of treatment-resistance according to the number of treatment trials adequately delivered.

137

Clinical severity: Depression severity was examined using the 21-item Hamilton rating scale (HAM-D, Hamilton, 1960) and Beck depression inventory (BDI) at admission and discharge. To check for other symptoms we used: anxiety, Beck anxiety inventory (BAI) (Beck et al., 1988), suicide ideation, BECK Suicide Ideation (BSI), Beck et al., 1979), hopelessness, Beck hopelessness scale (BHS), Beck et al., 1974), and recent life events, recent life events questionnaire (RLCQ), Casey et al., 1967) at admission. Pharmacological response to treatment was defined using the a priori definition of a reduction in HAM-D score of 50% or greater. Sample measurements were conducted with commercially available kits according to manufacturer’s protocol: Serum multiplex cytokine was analysed by (Randox Laboratories Ltd) and plasma cortisol by IMMULITEs (Diagnostics Products Corporation, Los Angeles, CA). For cortisol: CV values—7.1% within run, 7.8% between runs, detection limit 5.5 nmol/L.

2.1. Statistical analysis Two groups were compared for continuous variables by Mann– Whitney, and for discrete variables by the Chi-Square test. Since the majority of the cytokines, checked by the Kolmogorov–Smirnov test, were not normally distributed and transformations did not improve linearity non-parametric tests were used. We conducted two subsequent analyses. First, we compared controls versus all depressed patients. Second, as exploratory analysis we compared responsive versus refractory patients to antidepressant treatment. Results are expressed as median q25–q75 of raw cytokine levels. Raw values of cytokines were correlated with scale scores using the Spearman’s coefficient. Multiple testing corrections were applied via the method of Simes’ for correlated parameters (Rødland, 2006). The significance level was set to pr0.05 (two-tailed) and computer statistical

Table 1 Socio-demographic and clinical characteristics of the treatment-resistant depressed who responded (MDD R) or not (MDD NR) to pharmacological treatment, and healthy controls. Percentages are given in brackets. Parameters

MDD R

MDD NR

Healthy controls

p-value

N Sex (M/F) Age (AVG7 SEM) BMI (AVG7 SEM) BDI HAM-D admission HAM-D discharge BHS BAI RLCQ BSI Duration of current episode (years) Number previous hospital admissions Current medications

6 3/3 47.2 7 3.0 29.1 7 2.6 32.7 7 3.7 21.8 7 1.9 6.0 72.5 13.2 7 2.2 14.3 7 4.7 413.0 7 150.1 14.33 7 3.0 3.3 71.2 5.2 7 1.3 1 Drug free 4 Mood stabilizers 5 SSRI/SNRI 1 Benzodiazepines 2 Atypical antipsychotics 1 Tri/tetracyclic 2 MAOI Other antipsychotics Non psychotropics 4/6

13 03/11 50.9 73.6 28.80 7 1.1 37.6 7 3.6 21.7 7 2.1 13.4 7 4.3 15.7 7 1.2 25.1 7 3.7 334.6 7 48.9 20.07 2.8 6.3 71.8 4.4 71.0 6 Drug free 5 Mood stabilizers 8 SSRI/SNRI 3 Benzodiazepines 1 Atypical antipsychotics 4 Tri/tetracyclic 1 MAOI 1 Other antipsychotics 2 Non psychotropics 10/13

70% 30% 00%

70% 20% 10%

21 06/15 45.9 72.4 28.90 70.99 2.1 70.6 – – – – – – – Drug free – – – – – – – – – – – – – –

– 0.275 F ¼1.03, 0.314 F ¼0.02, 0.889 F ¼214.4, 0.001 0.767 0.002 0.239 0.080 0.535 0.183 0.259 0.689 – – – – – – – – – 0.517 0.665 – – –

ECT in the past Treatment-resistant stage: n (%) Stage 4þ failure to respond to ECT: Stage 3þ no response to 2nd augmentation Stage 2þ no response to 1st augmentation

AVG—average, F—female, BMI—body mass index, ECT—eletroconvulsotherapy, Treatment Resistance Stage—Thase and Rush treatment resistance criteria, MAOI—monoamine oxidase inhibitor, SSRI—selective serotonin reuptake inhibitor, SNRI—selective noradrenaline reuptake inhibitor, M—males, BHS—Beck hopelessness scale, BAI—Beck Anxiety Inventory, RLCQ—recent life changes questionnaire, BSI—Beck Suicide Scale, HAM-D—Hamilton depression scale, MDD R—treatment responsive, MDD NR—treatment non responsive, SEM—standard error of the mean.

138

L.A. Carvalho et al. / Journal of Affective Disorders 148 (2013) 136–140

differ at admission in relation to recent stressful life events (RLCQ, p¼0.535), hopelessness (BHS, p ¼0.239), suicide ideation (BSI, p¼0.183), duration of current episode (p ¼0.259), number of previous hospital admissions (p ¼0.689), number of patients who had had previous ECT in the past (p ¼0.517), or stage of treatment-resistance (p ¼0.665). There was a trend for increased anxiety symptoms in refractory patients at baseline as measured by the BAI (p ¼0.08).

packages (Graph-Pad 4.1 San Diego, California, USA and SPSS 20.0, Chicago, Illinois, USA) used.

3. Results 3.1. Demographic and clinical assessment 3.1.1. Healthy controls versus depressed patients Socio-demographic and clinical characteristics of healthy controls are shown in Table 1. There were no significant differences in age, gender distribution and body mass index between controls and MDD patients. At baseline, MDD patients had higher HAM-D and BDI scores than controls (po0.001).

3.2. Cortisol levels and cytokine profile We conducted two subsequent analyses. First, in order to analyse whether cortisol and cytokine levels were different between depressed patients and healthy controls, we compared the entire group of depressed patients with healthy controls (Fig. 1A). Second, in order to understand whether depressed patients were biologically different at the start of treatment, we also compared responsive and refractory at baseline (Fig. 1B).

Cytokine levels (pg/mL)

3.1.2. MDD responsive and refractory to antidepressant treatment All patients were taking medication at the time of testing. Using the a priori definition of treatment outcome, 6 of 19 patients showed a response to treatment (‘responsive’) and 13 did not show a response (‘refractory’). Socio-demographic and clinical characteristics separately for MDD responsive and refractory to antidepressants are shown in Table 1. The mean average of HAM-D scores at admission in responsive did not differ from refractory MDD patients (p ¼0.767). Between admission and discharge, depressive symptoms improved in responsive patients (p o0.001), but did not in refractory (p ¼0.239). Responsive or refractory patients did not

3.3. Plasma cortisol levels MDD patients had higher levels of cortisol when compared to healthy controls, Fig. 1A (p ¼0.0014). There was no difference of cortisol levels between responsive and refractory at baseline, (p ¼0.600) Fig. 1B.

400 300 200 100 30

*

20

control patients

*

10 6 4 2 0

*

* IL-10

*

IL-6

TNF

IL-4

MCP-1

VEGF

p = 0.008

500 250 0 6 5 4 3 2 1 0

p = 0.012 p = 0.028

8 7 6 5 4 3 2 1 0

p = 0.05

400 MCP-1 (pg/mL)

p<0.001

35 30 25 20 15 10 5 0

p = 0.03 HC MDD_R MDD_NR

300 200 100 0 3

p = 0.012 IL-10 (pg/mL)

750

VEGF (pg/mL)

1000

IL-4 (pg/mL)

IL-6 (pg/mL)

Cortisol (pg/mL)

p = 0.010 p = 0.006

Cortisol

p = 0.035

2 1 0

Fig. 1. (A) Comparison between cytokine levels in healthy controls and depressed patients. Multiple comparisons were then conducted using Simes’ correction procedure, (B) Comparison between cytokine levels in healthy controls and in depressed patients in relation to treatment outcome. Treatment outcome had a significant major effect on cytokines serum levels in treatment-resistant depressed (MDD_R) responsive and refractory (MDD_NR) when compared to controls. First, Kruskal–Wallis nonparametric test was used to analyse whether cytokine levels differed in relation to group (MDD responsive, MDD refractory, Healthy controls). Multiple comparisons were then conducted using Simes’ correction procedure. Statistical significance was considered when p r 0.05.

L.A. Carvalho et al. / Journal of Affective Disorders 148 (2013) 136–140

3.3.1. IL-4 MDD patients had lower levels of IL-4 when compared to healthy controls (p ¼0.035), Fig. 1A. There was no difference between responsive and refractory in relation to IL-4 levels at baseline (p ¼0.087), Fig. 1B. 3.3.2. IL-6 MDD patients had higher levels of IL-6 when compared to healthy controls (p ¼0.035), Fig. 1A. There was no difference between responsive and refractory in relation to IL-6 levels at baseline (p ¼0.467), Fig. 1B. 3.3.3. IL-10 MDD patients had higher levels of IL-10 when compared to healthy controls (p ¼0.035), Fig. 1A. There was no difference between responsive and refractory in relation to IL-10 levels at baseline (p ¼0.898), Fig. 1B. 3.3.4. MCP-1 MDD patients did not differ in the levels of MCP-1 when compared to healthy controls (p¼0.298), Fig. 1A. However, refractory had significantly lower levels of MCP-1 levels than responsive at baseline (p¼0.036), Fig. 1B. 3.3.5. VEGF MDD patients had lower levels of VEGF when compared to healthy controls (p ¼0.047), Fig. 1A. There was a trend difference for lower levels of VEGF in refractory when compared to responsive at baseline (p ¼0.058), Fig. 1B. 3.4. Clinical correlations We examined correlations between clinical variables and the levels of the VEGF and MCP-1. Specifically, both VEGF and MCP-1 negatively correlated with depressive symptoms at baseline as measured by the HAM-D scores (VEGF, r ¼  0.548, p ¼0.043, and MCP-1, r ¼  0.636, p ¼0.014). MCP-1 and VEGF correlated with each other in MDD patients (VEGF, r¼0.556, p ¼0.011), but not in healthy controls (VEGF, r ¼0.319, p ¼0.170).

4. Discussion We found that all depressed patients, irrespective of treatment outcome, had clear evidence of immune activation as shown by higher levels of IL-6 and IL-10, but lower levels of IL-4 in the context of HPA axis disturbance as shown by higher cortisol levels. Moreover, we have explored whether patients with different levels of treatment-resistance present different degrees of inflammatory disturbance. The data suggests lower VEGF and MCP-1 levels in refractory than in responsive patients at baseline, and a negatively correlation with depressive symptoms in refractory patients. We found increased levels of IL-6, but not TNF, in patients with a moderate degree of treatment-resistance. This is in line with meta-analysis studies that show increased IL-6 levels in depression (Dowlati et al., 2010; Hiles et al., 2012; Howren et al., 2009). We also found, IL-4 and IL-10, respectively, lower and higher in depressed patients. Similar to our work, others had previously demonstrated that depressed patients had lower IL-4 levels compared to normal controls (Myint et al., 2005), but this was not confirmed by some who reported higher (Hernandez et al., 2008) or even unchanged (Schlatter et al., 2004) IL-4 levels in depression. There was no support for the involvement IL-2, IL-8, IFN-a, IL-1a or IL-1b in our sample.

139

To our knowledge, only a couple of studies have investigated inflammatory cytokines specifically looking at treatment-resistant depression. Similar to our work, O’Brien et al. (2007) found higher levels of IL-6 in depressed patients who failed to respond to SSRI. As they did not include a SSRI-responsive group, direct comparisons with our study is not straightforward. Yoshimura also found higher levels of IL-6 in depressed patients when compared to controls (Yoshimura et al., 2009). They showed, however, increased levels of IL-6 in refractory when compared to responsive patients (Yoshimura et al., 2009). Their group of depressed patients were younger than ours (mean age Yoshimura 4079 years; ours 49.672.6 years), and not medicated. The lack of information on other clinical characteristics halts further comparison. Neither authors analysed MCP-1 or VEGF levels. We have shown for the first time an association between MCP-1, VEGF and treatment-outcome. The role of MCP-1 in depression is not yet clear. Research showed depressed patients without traditional risk factors for heart disease present elevated levels of circulating markers of atherosclerosis vulnerability, like MCP-1 (Piletz et al., 2009; Rajagopalan et al., 2001). Others who used similar technology have shown down-regulation of MCP-1 in patients with longer duration of illness (Lehto et al., 2010). Noradrenaline induces production of MCP-1 in astrocytes and mediates neuroprotective actions (Hinojosa et al., 2011). Chronic stress decreases noradrenaline induced MCP-1 (Madrigal et al., 2010). MCP-1 has previously shown to be involved in cardiac repair and to have protective effect under certain conditions (Morimoto et al., 2008). Our data also suggests a role played by VEGF in the most severe cases of depression. In accordance, others have also found decreased levels of protective factors, like VEGF (Fournier and Duman, 2012) in depression. Lower VEGF levels has been associated with completed suicide (Isung et al., 2011), and resistance to antidepressant treatment (Viikki et al., 2010). VEGF is an angiogenic protein with both neuroprotection and neurotrophic actions (Segi-Nishida et al., 2008). We ourselves have previously shown an involvement of other neurotrophic factors in mood disorders (Barbosa et al., 2011, 2012). It is possible that inflammation in the presence of glucocorticoid resistance contributes to disturbed angiogenic and neurogenic disturbances leading to depression (Carvalho and Pariante, 2008; Zunszain et al., 2010). Neurogenesis in the hippocampus region is a known effect of antidepressants (Anacker et al., 2011) and might have therapeutic value. Limitations include the cross-sectional design that does not allow causal conclusions, limited sample size, and medicated patients. Nevertheless, generalized activation of the inflammatory system was observed despite the anti-inflammatory properties of some antidepressants (Hannestad et al., 2011). Thus, we believe that our data is still relevant for understanding pathophysiology of depression. In conclusion, our data support the hypothesis that lack of clinical therapeutic benefit of antidepressants is associated with further immunological impairment in the presence of a concomitant HPA axis disturbance. Biomarkers may be able to identify patients who may benefit from early access to adjuvant therapies.

Role of funding source This research has been funded by the UK Medical Research Council, the NARSAD, the South London and Maudsley NHS Foundation Trust & Institute of Psychiatry NIHR Biomedical Research Centre for Mental Health, and the Commission of European Communities 7th Framework Programme Collaborative Project Grant Agreement no. 22963 (Mood Inflame) to Carmine Pariante. Livia A Carvalho is funded by the NARSAD Young Investigator Award 2009 and the ECNP Young Investigator Award.

Conflict of interest The authors have no conflict of interest to disclose.

140

L.A. Carvalho et al. / Journal of Affective Disorders 148 (2013) 136–140

Acknowledgments The authors would like to thank you all participants that took part in the study.

References Anacker, C., Zunszain, P.A., Cattaneo, A., Carvalho, L.A., Garabedian, M.J., Thuret, S., Price, J., Pariante, C.M., 2011. Antidepressants increase human hippocampal neurogenesis by activating the glucocorticoid receptor. Molecular Psychiatry 16, 738–750. Barbosa, I.G., Huguet, R.B., Sousa, L.P., Abreu, M.N., Rocha, N.P., Bauer, M.E., Carvalho, L.A., Teixeira, A.L., 2011. Circulating levels of GDNF in bipolar disorder. Neuroscience Letters 502, 103–106. Barbosa, I.G., Rocha, N.P., Huguet, R.B., Ferreira, R.A., Salgado, J.V., Carvalho, L.A., Pariante, C.M., Teixeira, A.L., 2012. Executive dysfunction in euthymic bipolar disorder patients and its association with plasma biomarkers. Journal of Affective Disorders 137, 151–155. Beck, A.T., Epstein, N., Brown, G., Steer, R.A., 1988. An inventory for measuring clinical anxiety: psychometric properties. Journal of Consulting and Clinical Psychology 56, 893–897. Beck, A.T., Kovacs, M., Weissman, A., 1979. Assessment of suicidal intention: the scale for suicide ideation. Journal of Consulting and Clinical Psychology 47, 343–352. Beck, A.T., Weissman, A., Lester, D., Trexler, L., 1974. The measurement of pessimism: the hopelessness scale. Journal of Consulting and Clinical Psychology 42, 861–865. Carvalho, L.A., Pariante, C.M., 2008. In vitro modulation of the glucocorticoid receptor by antidepressants. Stress 11, 411–424. Carvalho, L.A., Juruena, M.F., Papadopoulos, A.S., Poon, L., Kerwin, R., Cleare, A.J., Pariante, C.M., 2008. Clomipramine in vitro reduces glucocorticoid receptor function in healthy subjects but not in patients with major depression. Neuropsychopharmacology 33 (13), 3182–3189. Casey, R.L., Masuda, M., Holmes, T.H., 1967. Quantitative study of recall of life events. Journal of Psychosomatic Research 11, 239–247. David Taylor, Carol Paton, Robert Kerwin, 2007. The Maudsley Prescribing Guidelines. Informa Healthcare. Dowlati, Y., Herrmann, N., Swardfager, W., Liu, H., Sham, L., Reim, E.K., Lanctot, K.L., 2010. A meta-analysis of cytokines in major depression. Biological Psychiatry 67, 446–457. Fournier, N.M., Duman, R.S., 2012. Role of vascular endothelial growth factor in adult hippocampal neurogenesis: implications for the pathophysiology and treatment of depression. Behavioural Brain Research 227, 440–449. Hamilton, M., 1960. A rating scale for depression. Journal of Neurology, Neurosurgery, Psychiatry 23, 56–62. Hannestad, J., DellaGioia, N., Bloch, M., 2011. The effect of antidepressant medication treatment on serum levels of inflammatory cytokines: a metaanalysis. Neuropsychopharmacology 36, 2452–2459. Hernandez, M.E., Mendieta, D., Martinez-Fong, D., Loria, F., Moreno, J., Estrada, I., Bojalil, R., Pavon, L., 2008. Variations in circulating cytokine levels during 52 week course of treatment with SSRI for major depressive disorder. European Neuropsychopharmacology 18, 917–924. Hiles, S.A., Baker, A.L., de, M.T., Attia, J., 2012. A meta-analysis of differences in IL-6 and IL-10 between people with and without depression: exploring the causes of heterogeneity. Brain, Behavior, and Immunity 26, 1180–1188. Hinojosa, A.E., Garcia-Bueno, B., Leza, J.C., Madrigal, J.L., 2011. CCL2/MCP-1 modulation of microglial activation and proliferation. Journal of Neuroinflammation 8, 77. Howren, M.B., Lamkin, D.M., Suls, J., 2009. Associations of depression with C-reactive protein, IL-1, and IL-6: a meta-analysis. Psychosomatic Medicine 71, 171–186.

Isung, J., Mobarrez, F., Nordstrom, P., Asberg, M., Jokinen, J., 2011. Low plasma vascular endothelial growth factor (VEGF) associated with completed suicide. World Journal of Biological Psychiatry. Lehto, S.M., Niskanen, L., Miettola, J., Tolmunen, T., Viinamaki, H., Mantyselka, P., 2010. Serum anti-inflammatory markers in general population subjects with elevated depressive symptoms. Neuroscience Letters 484, 201–205. Madrigal, J.L., Garcia-Bueno, B., Hinojosa, A.E., Polak, P., Feinstein, D.L., Leza, J.C., 2010. Regulation of MCP-1 production in brain by stress and noradrenalinemodulating drugs. Journal of Neurochemistry 113, 543–551. Morimoto, H., Hirose, M., Takahashi, M., Kawaguchi, M., Ise, H., Kolattukudy, P.E., Yamada, M., Ikeda, U., 2008. MCP-1 induces cardioprotection against ischaemia/reperfusion injury: role of reactive oxygen species. Cardiovascular Research 78, 554–562. Myint, A.M., Leonard, B.E., Steinbusch, H.W., Kim, Y.K., 2005. Th1, Th2, and Th3 cytokine alterations in major depression. Journal of Affective Disorder 88, 167–173. O’Brien, S.M., Scully, P., Fitzgerald, P., Scott, L.V., Dinan, T.G., 2007. Plasma cytokine profiles in depressed patients who fail to respond to selective serotonin reuptake inhibitor therapy. Journal of Psychiatric Research 41, 326–331. Piletz, J.E., Halaris, A., Iqbal, O., Hoppensteadt, D., Fareed, J., Zhu, H., Sinacore, J., Lindsay, D.C., 2009. Pro-inflammatory biomakers in depression: treatment with venlafaxine. World Journal of Biological Psychiatry, 1–11. Rødland, Einar Andreas, 2006. Simes procedure is ‘valid on average’. Biometrika 93, 742–746. Rajagopalan, S., Brook, R., Rubenfire, M., Pitt, E., Young, E., Pitt, B., 2001. Abnormal brachial artery flow-mediated vasodilation in young adults with major depression. American Journal of Cardiology 88 (196-8), A7. Sackeim, H.A., 2001. The definition and meaning of treatment-resistant depression. Journal of Clinical Psychiatry 62 (16), 10–17. Schlatter, J., Ortuno, F., Cervera-Enguix, S., 2004. Lymphocyte subsets and lymphokine production in patients with melancholic versus nonmelancholic depression. Psychiatry Research 128, 259–265. Segi-Nishida, E., Warner-Schmidt, J.L., Duman, R.S., 2008. Electroconvulsive seizure and VEGF increase the proliferation of neural stem-like cells in rat hippocampus. Proceedings of the National academy of Sciences of the United States of America 105, 11352–11357. Thase, M.E., Rush, A.J., 1997. When at first you don’t succeed: sequential strategies for antidepressant nonresponders. Journal of Clinical Psychiatry 58 (13), 23–29. Viikki, M., Anttila, S., Kampman, O., Illi, A., Huuhka, M., Setala-Soikkeli, E., Mononen, N., Lehtimaki, T., Leinonen, E., 2010. Vascular endothelial growth factor (VEGF) polymorphism is associated with treatment-resistant depression. Neuroscience Letters 477, 105–108. World Health Organisation, 1992. The ICD-10 Classification of Mental and Behavioural Disorders: Clinical Descriptions and Diagnostic Guidelines. Geneva. Wooderson, S.C., Juruena, M.F., Fekadu, A., Commane, C., Donaldson, C., Cowan, M., Tomlinson, M., Poon, L., Markopoulou, K., Rane, L., Donocik, J., Tunnard, C., Masterson, B., Cleare, A.J., 2011. Prospective evaluation of specialist inpatient treatment for refractory affective disorders. Journal of Affective Disorders 131 (1–3), 92–103. Yoshimura, R., Hori, H., Ikenouchi-Sugita, A., Umene-Nakano, W., Ueda, N., Nakamura, J., 2009. Higher plasma interleukin-6 (IL-6) level is associated with SSRI- and SNRI-refractory depression. Progress in Neuropsychopharmacological and Biological Psychiatry 33, 722–726. Zunszain, P.A., Anacker, C., Cattaneo, A., Carvalho, L.A., Pariante, C.M., 2010. Glucocorticoids, cytokines and brain abnormalities in depression. Progress in Neuropsychopharmacological and Biological Psychiatry.