Impact on cortisol and antidepressant efficacy of quetiapine and escitalopram in depression

Impact on cortisol and antidepressantefficacy of quetiapine and escitalopram indepression§

Nina Sarubin a,c,*, Caroline Nothdurfter a,b, Christian Schmotz c,Anna-Maria Wimmer c, Julia Trummer c, Martin Lieb a,Manfred Uhr b, Thomas C. Baghai a, Thomas C. Wetter a,Markus Buhner d, Rainer Rupprecht a,b, Cornelius Schule c

aDepartment of Psychiatry and Psychotherapy, University Regensburg, Regensburg, GermanybMax-Planck-Institute of Psychiatry, Munich, GermanycDepartment of Psychiatry and Psychotherapy, Ludwig-Maximilian-University, Munich, GermanydDepartment of Psychology/Statistics and Evaluation, Ludwig-Maximilian-University, Munich, Germany

Received 23 July 2013; received in revised form 15 October 2013; accepted 15 October 2013

Psychoneuroendocrinology (2014) 39, 141—151

KEYWORDSMajor depression;Escitalopram;Quetiapine;HPA system

Summary

Background: In this study, the impact of quetiapine fumarate extended release (QXR) andescitalopram (ESC) on HPA axis activity was investigated in depressed patients in relationshipto antidepressant efficacy.Methods: In a randomized, open-label 5-week trial 60 inpatients suffering from major depression(DSM-IV criteria) were treated for 5 weeks with either QXR (300 mg/day) or ESC (10 mg/day). Thedexamethasone/CRH (DEX/CRH) test was performed before treatment, after 1, and after 5 weeksof treatment. Cortisol (COR) AUC values were used to assess HPA axis function. The HamiltonDepression Rating Scale was used weekly to estimate antidepressant efficacy.Results: QXR and ESC showed comparable antidepressant effects but strongly differed in theirimpact on HPA axis activity. In the QXR group, a marked inhibition of COR AUC levels was observedwhich was most pronounced after one week of treatment but showed a partial re-increase after 5weeks of treatment. In contrast, ESC transiently stimulated COR AUC values (week 1) whereasCOR AUC levels at week 0 and week 5 were comparable. COR improvement at week 1 (defined asCOR peak value reduction between DEX/CRH test 1 and 2) was significantly associated with betterclinical outcome.

§ Clinicaltrials.gov identifier: NCT00953108.* Corresponding author at: Department of Psychiatry and Psychotherapy, Ludwig-Maximilian-University 80336 Munich, Germany.

Tel.: +49 89-51605382; fax: +49 89-51603930.E-mail address: [email protected] (N. Sarubin).

Available online at www.sciencedirect.com

ScienceDirect

j our na l h omepa g e: www.e l se v ie r.c om/l oca te/ psyne ue n

0306-4530/$ — see front matter # 2013 Elsevier Ltd. All rights reserved.

Conclusion: Apparently, different effects on HPA axis activity reflect distinct pharmacoendocri-nological properties of psychotropic drugs.# 2013 Elsevier Ltd. All rights reserved.

142 N. Sarubin et al.

1. Introduction

One of the most widely discussed moderating or mediatingfactors regarding the etiology of major depression is adysregulation of hypothalamic-pituitary-adrenocortical(HPA) axis activity (Holsboer, 2000, 2001). Proponents ofthe corticosteroid receptor hypothesis (neuroendocrinolo-gical hypothesis) of depression argue that a gradual normal-ization of HPA system dysregulation, as measured by serialdexamethasone/CRH (DEX/CRH) tests, precedes or coin-cides with the response to antidepressant treatment (Hols-boer, 2000, 2001; Ising et al., 2007). The reduction of HPAaxis activity is a necessary prerequisite for clinical remis-sion to become manifest and also assert that all antide-pressants developed so far may have a dampening impacton HPA axis function, irrespective of their type of actionwithin monoaminergic systems (Ising et al., 2007; Schuleet al., 2009).

However, the rate of non-suppression in the DEX/CRH testin acutely depressed patients is inconsistent in more recentstudies and lies within the range between 20 and 30 percent(Ising et al., 2005; Schule et al., 2009). A considerable part ofacutely depressed patients shows normally regulated HPAaxis activity in the DEX/CRH test already before antidepres-sant treatment and may nevertheless benefit from thesedrugs. Some researchers even found an enhanced HPA systemactivity at discharge in a certain proportion of depressedpatients in spite of clinical recovery (Zobel et al., 2001;Schule et al., 2009). On the other hand, down-regulationof HPA axis activity in depressed patients is not necessarilyfollowed by a favorable response (Schule et al., 2009). Theseapparently contradictory findings suggest that down-regula-tion of an increased HPA axis activity may be partly involvedin the pathophysiology of depression and its treatment but isa neither necessary nor sufficient condition for a beneficialtreatment outcome (Schule et al., 2009).

Moreover, the assumption of a uniform normalizing influ-ence of different types of antidepressants on HPA axis hyper-activity — irrespective of their mode of action — is challengedby the finding that various antidepressants show differenteffects on cortisol (COR) and acetylcholine (ACTH) releaseafter acute administration in healthy volunteers (Laakmann,1988). Whereas inhibitors of noradrenaline and/or serotoninreuptake acutely stimulate COR and ACTH secretion (Laak-mann, 1988), other antidepressants with different pharma-codynamic properties, such as mirtazapine, do not (Schule,2007). Mirtazapine is not a reuptake inhibitor but acts as anantagonist at presynaptic a2-receptors and at postsynaptic 5-HT2, 5-HT3, and histamine H1 receptors (De Boer, 1995). Inseveral investigations, our research group could demonstratethat single administration of 15 mg mirtazapine acutely inhi-bits ACTH and COR release in healthy male controls, pre-sumably due to central antagonism at H1 and 5-HT2 receptorsthereby reducing the hypothalamic CRH output (Schuleet al., 2002). Whereas reuptake inhibitors of serotonin(Bschor et al., 2012) or noradrenaline (Schule et al., 2006)

show a delayed and gradual down-regulation of HPA axisactivity in depressed patients, mirtazapine markedly inhibitsHPA system function already after 1 week of treatment,followed by a partial re-increase of COR levels in the DEX/CRH test after 5 weeks of therapy (Schule et al., 2006).

In the present study, the influence of 5-week treatmentwith the atypical antipsychotic drug quetiapine fumarateextended release (QXR) and of the selective serotonin-reup-take inhibitor escitalopram (ESC) on the time course of HPAaxis activity was investigated in depressed inpatients. Que-tiapine is an atypical antipsychotic which acts as an antago-nist with high affinity at histaminergic H1 receptors; withmoderate affinity at serotonergic 5-HT1A and 5-HT2, adre-nergic a1, and dopaminergic D1—3 receptors; and almost noaffinity for D4, a2, and muscarinic acetylcholine (ACh) recep-tors or benzodiazepine binding sites (Green, 1999). Thepharmacological effects of quetiapine have some similaritiescompared to mirtazapine: both drugs are antagonists at 5-HT2 and H1 receptors which may cause inhibition of hypotha-lamic CRH release and consecutively inhibit ACTH and CORsecretion (Schule, 2007). In addition, quetiapine is anantagonist at a1-receptors; this pharmacological propertyis also expected to reduce hypothalamic CRH secretion(Schule, 2007). In fact, both mirtazapine (Laakmann,1988; Schule et al., 2002) and quetiapine (de Borja et al.,2005) have been shown to acutely inhibit ACTH and/or CORsecretion after single administration in healthy male volun-teers.

Quetiapine, which has primarily been developed as anatypical antipsychotic, has also been demonstrated to haveantidepressant properties as monotherapy in the acute treat-ment of bipolar depression at dosages of 300 and 600 mg/day(Calabrese et al., 2005; McElroy et al., 2010). Moreover,extended release quetiapine fumarate (QXR) may exert sig-nificant antidepressant efficacy both in acute treatment(Cutler et al., 2009; Katila et al., 2012) and maintenancetherapy (Liebowitz et al., 2010) of unipolar depression, whengiven as monotherapy at dosages of 150 or 300 mg/day. Inaddition, the onset of antidepressant action during treat-ment with QXR appears to be earlier than that of conven-tional antidepressant drugs such as duloxetine (Cutler et al.,2009). The efficacy of quetiapine or QXR as an augmentationstrategy in the treatment of major depressive episodes hasalso been shown in several controlled (McIntyre et al., 2007;Bauer et al., 2010; El-Khalili et al., 2010) or open-label(Devarajan et al., 2006; Sagud et al., 2006; Andersonet al., 2009) trials.

The aim of the study was to investigate the influence of 5-week treatment with QXR monotherapy on HPA axis activityin depressed inpatients and to answer the question whetherputative changes in HPA system function are related to theantidepressant effects of QXR. Moreover, ESC which is the S-enantiomer of citalopram and is one of the most potent SSRIsdue to its special affinity for the allosteric and orthostericbinding sites of the serotonin-transporter (Plenge et al.,2007) served as an active comparator drug.

Impact on cortisol and antidepressant efficacy of quetiapine and escitalopram in depression 143

2. Materials and methods

2.1. Study overview

In this randomized open-label 5-week trial, 60 depressedinpatients, treated in the Department of Psychiatry andPsychotherapy, Ludwig-Maximilian-University of Munich,Germany, were included. The recruitment period lasted fromAugust 2009 to February 2012. The patients were treated for5 weeks with either QXR (300 mg/day; group 1) or ESC(10 mg/day; group 2). Randomization was performed by anonline computer randomization program (http://www.ran-domization.com). The allocation of the patients to bothtreatment groups occurred in a randomized order withoutany restrictions (such as blocking or block size). To implementthe random allocation sequence patients were includedconsecutively in the study after eligibility was proven andwritten informed consent was obtained and then sequentiallynumbered. The allocation to the treatment group was doneby the principal investigator (C. Schule) according to the pre-defined randomization plan.

We abstained from blinding the medication since the sideeffect profiles of QXR and ESC differ markedly (e.g., sedativeeffects of QXR but not of ESC) and in clinical practice, ESC isgiven preferably in the morning whereas the main dosage ofQXR is usually administered in the evening.

The administration of QXR was titrated at the beginning ofthe study in that the patients of group 1 received at 18:00 h50 mg QXR from day 1 to day 2; 150 mg QXR from day 3 to day 4;and 300 mg QXR from day 5 to day 35. With respect to group 2,the patients received 2.5 mg ESC at 08:00 h on day 1 and 2;5 mg ESC on day 3 and 4; and 10 mg ESC from day 5 to day 35.Zopiclon (up to 7.5 mg/day) and lorazepam (up to 2 mg/day)were allowed as concomitant medication in case of sleepdifficulties or inner tension/anxiety. The time of administra-tion of QXR and ESC was stable throughout the study.

The dexamethasone/CRH test (DEX/CRH test) was per-formed before treatment, after 1 and after 5 weeks oftreatment calculating COR AUC (area under the curve) valuesto assess HPA axis function. The 21-item version of theHamilton Depression Rating Scale (21-HAMD) was used atdays 0 and 4 and weekly thereafter to estimate the severityof depression.

The study was carried out according to the Declaration ofHelsinki (http://www.wma.net) and had been approved bothby a local ethics committee (intramural review panel of theLudwig-Maximilian-University of Munich, Faculty of Medi-cine) and the Federal Institute for Drugs and Medical Devicesin Germany.

2.2. Eligibility

60 depressed male and female unrelated inpatients agedbetween 18 and 65 years and suffering from a major depres-sive episode according to DSM-IV criteria (American Psychia-tric Association, 1994) entered the study after theprocedures had been fully explained and written informedconsent had been obtained. The recruitment period lastedfrom August 2009 up to February 2012. The patients werediagnosed by experienced and trained psychiatrists using theStructured Clinical Interview for DSM-IV, German version

(Wittchen et al., 1997). The sample size was determinedby a statistical power analysis based on a former study of ourresearch group investigating reboxetine and mirtazapine in asimilar study design (Schule et al., 2006). The patients werefree of comorbid psychiatric and somatic disorders, formernon-response or intolerance to ESC or QXR led to exclusionfrom the study.

53 patients (25 patients treated with QXR; 28 patientstreated with ESC) completed the full study period up to week5 including the performance of 3 DEX/CRH tests (CONSORT2010 Flow Diagram). As predefined in the study protocol,statistical analysis was restricted to the completers becauseperformance of all 3 DEX/CRH test (week 0, week 1, week 5)was mandatory to investigate the relationship between HPAaxis function and clinical response (completer analysis).

Further clinical and demographic characteristics of theparticipants who completed the study are given in Table 1.

2.3. Outcome measures

Since the study was designed to investigate the relationshipbetween clinical response and impact on HPA axis function,21-HAMD sum scores and COR AUC values within the DEX/CRHtests were pre-defined as primary outcome measures.

Severity of depression was estimated weekly, during thefirst week also after 4 days of treatment (days �1, 4, 7, 14,21, 28, 35) using the 21-HAMD in the morning between 09:00and 11:00 h. All raters were experienced psychiatrists andblind to hormonal measurements. Raters were not changed insingle patients throughout the whole study period. Clinicalresponse was defined by a reduction of at least 50% in the 21-HAMD sum score after five weeks of treatment with QXR orESC.

The DEX/CRH test was performed before treatment (test1), after 1 week (test 2) and again after 5 weeks (test 3) ofpharmacotherapy with QXR (300 mg/die) or ESC (10 mg/die).Participants pre-treated with an oral dose of 1.5 mg dexa-methasone (DEX) at 23.00 h followed by an intravenousinjection of 100 mg hCRH on the next day at 15:00 h. Bloodsamples were collected at 15:00, 15:30, 15:45, 16:00, and16:15 h. The total COR AUC values (total area under theconcentration time curve), determined by the trapezoid ruleaccording to Simpson (Forsythe et al., 1969), were used fordetermination of the COR response to the hCRH challenge inthe DEX pretreated patients representing the combinedeffects of altered glucocorticoid receptor (GR) functionand hyperdrive of endogenous CRH and vasopressin.

A COR peak improver was defined by a lower COR peakconcentration during DEX/CRH test 2 (after 1 week of treat-ment) as compared to test 1 (week 0 before treatment);otherwise, a COR peak non-improver was assumed. The peakCOR level was used for the categorization into HPA systemimprovers and non-improvers instead of the COR AUC value tobe in line with previous definitions of HPA system improve-ment during acute treatment of depression (Ising et al.,2005, 2007; Schule et al., 2009).

2.4. Statistical analysis

Normal distribution of the data was established using theKolmogorov—Smirnov test. Demographic and clinical para-

Table 1 Clinical and demographic data in 53 out of 60 depressive patients treated with either QXR (n = 25; 300 mg/day) or ESC(n = 28; 10 mg/day) for 5 weeks (completer analysis). Data represent mean � SD (standard deviation). Suppression status atbaseline (week 0): S = Suppressors; NS = Non-Suppressors. Medical pre-treatment: Pre = Pre-Treatment; NPre = No Pre-Treatment.Response: R = Responders; NR = Non-Responders. Statistical parameters of Chi square tests and Fisher’ exact test for qualitativevariables as well as Mann—Whitney U-tests (non-parametric quantitative variables) and T-tests (parametric quantitative variables)are provided. Significant results are provided in bold letters.

All patients (n = 53) QXR (n = 25) ESC (n = 28) Statistical evaluationChi-square test(qualitative variables)

x 2 df p-Value

Chi-square test (qualitative variables)Sex [M/F] 38/15 21/4 17/11 0.074a

Medical pre-treatment (Pre/NPre) 16/37 9/16 7/21 0.758 1 0.384Suppression status (S/NS) 27/26 15/10 12/16 1.553 1 0.213

All patients (n = 53) QXR (n = 25) ESC (n = 28) T df p-Value

T-test (quantitative variables)21-HAMD, week 0 22.04 � 5.23 21.24 � 4.14 22.75 � 6.03 �1.050 51 0.299Age [years] 40.25 � 12.57 41.24 � 11.04 39.36 � 13.97 0.540 51 0.592Age of onset [years] 34.91 � 10.76 34.16 � 11.14 35.57 � 10.56 �0.473 51 0.638Number of depressive episodes 2.23 � 1.85 2.44 � 1.92 2.04 � 1.79 0.793 51 0.432Duration of total illness [months] 75.43 � 116.57 104.38 � 143.17 49.68 � 80.56 1.684 36.869 0.101Duration of index episode [weeks] 11.17 � 10.76 11.24 � 8.30 11.11 � 12.71 0.044 51 0.965Duration of wash-out period [days] 25.34 � 14.83 23.48 � 15.67 27.00 � 14.11 �0.860 51 0.394COR AUC at baseline (week 0)[nmol/l x min]

8784.53 � 8016.34 7388.33 � 8314.48 10,031.13 � 7674.45 0.084b

Weight [kg] 80.92 � 14.99 84.44 � 16.19 77.79 � 13.36 1637 51 0.108Height [cm] 174.87 � 12.59 173.96 � 16.67 175.68 � 7.56 �0.492 51 0.625a Fisher’s exact test (two-sided).b : Mann—Whitney U test (two-sided).

144 N. Sarubin et al.

meters were compared between the QXR group and the ESCgroup by the Pearson Chi-Square test for contingency tablesor by Fisher’ exact test with respect to qualitative variables(gender distribution, medical pre-treatment) and by a one-way ANOVA for independent samples with regard to quanti-tative variables (current age, age of onset, number of depres-sive episodes, duration of total illness, duration of indexepisode, duration of wash-out period, COR AUC value atbaseline [week 0], weight, and height).

For statistical comparison of 21-HAMD sum scores betweenboth treatment groups, an analysis of variance with arepeated measures design (rmANOVA) was performed (WilksLambda). Thereby ‘‘time’’ (days 0, 4, 7, 14, 21, 28, 35) and‘‘group’’ (QXR versus ESC) were considered as within-sub-jects and between-subjects factors with seven (‘‘time’’) andtwo (‘‘group’’) levels, respectively. RmANOVAs were alsocarried out to compare the influence of QXR and ESC onendocrinological parameters (COR AUC values) using ‘‘time’’(weeks 0, 1, 5) and ‘‘group’’ (QXR versus ESC) as within-subjects and between-subjects factors with 3 (‘‘time’’) and 2(‘‘group’’) levels, respectively. Moreover, to compare theclinical amelioration of depressive symptoms between CORsuppressors and COR non-suppressors, rmANOVAs were per-formed in the total sample but also separately for bothtreatment groups, using ‘‘time’’ (days 0, 4, 7, 14, 21, 28,35) and ‘‘group’’ (COR suppressors versus COR non-suppres-sors) as within-subjects and between-subjects factors withseven (‘‘time’’) and two (‘‘group’’) levels, respectively.

For all rmANOVA procedures, post hoc tests were addi-tionally performed whenever a significant ‘‘group’’ factorwas found, in order to compare the single time points of themeasurements individually. Moreover, a correction wasapplied to the F-value by means of adjusting the degreesof freedom by a factor Epsilon, if the sphericity test(Mauchly-W test) was significant indicating heterogeneityof the covariances (Huynh-Feldt correction). In addition, alog-transformation was conducted (natural logarithm) toachieve normal distribution if variables did not show Gaussiandistribution.

Statistical analyses were performed using SPSS for Win-dows (Release 20, SPSS Inc., Chicago, IL 60606, USA). As anominal level of significance, alpha = 0.05 was accepted. Alltests were conducted two-tailed unless stated differently.

3. Results

3.1. Clinical characteristics

At baseline, group 1 (QXR treatment) and group 2 (ESCtreatment) did not differ significantly in gender distribution(Fisher’ exact test: p = 0.074) nor were there any significantdifferences between group 1 and group 2 with regard to otherclinical or demographic variables ( p > 0.05 for all compar-isons; Table 1). Also, there was no significant differencebetween response/non-response and COR level in the first

Impact on cortisol and antidepressant efficacy of quetiapine and escitalopram in depression 145

dex/crh-test (Mann—Whitney U-test, p = 0.673). The dura-tion of the wash-out period and the kind of psychopharma-cological pre-treatment did not have any significant impacton the results (Table 1).

3.2. Assessment of HPA axis function(performance of DEX/CRH tests, completeranalysis)

3.2.1. Comparison of patients treated with QXR andESCPatients who received QXR treatment showed a marked andsignificant reduction in COR secretion during the DEX/CRHtest already after one week of therapy. However, after fiveweeks of QXR treatment the CRH-induced COR secretionincreased again to some extent. In patients receiving ESCan increase of COR concentrations during the DEX/CRH testcould be observed after 1 week of treatment which cameback to baseline levels after 5 weeks of treatment (Fig. 1).

The performance of rmANOVA using ‘‘time’’ (DEX/CRHtest 1—3) and ‘‘group’’ (QXR versus ESC) as within and

Figure 1 Mean COR concentrations during 3 DEX/CRH tests(week 0, 1, 5) in depressed patients treated with QXR (300 mg/day) or ESC (10 mg/day). SEM (standard error of mean) indicat-ed. *:p < 0.05. **:p < 0.01.

between subjects factors revealed a significant ‘‘time’’effect (F = 3.511; df = 2, 50; p = 0.037), a highly significant‘‘group’’ effect (F = 11.113; df = 1, 51; p = 0.002), and ahighly significant ‘‘time x group’’ interaction effect(F = 5.358; df = 2, 50; p = 0.008), indicating a significantimpact of both treatment options on HPA axis function overtime which differed strongly and significantly between theQXR group and the ESC group (Fig. 1).

3.2.2. Separate analysis of HPA axis function insuppressors and non-suppressorsSince a generally accepted cut-off criterion defining suppres-sion status has not been established yet for the DEX/CRH test,we used the median of the COR AUC values at baseline (DEX/CRH test 1 at week 0) to differentiate between baselinesuppressors and non-suppressors. The median COR AUC valueat baseline was 5624.175 nmol/l � min in our sample. There-fore, patients with COR AUC values above 5624.175 nmol/l � min (test 1) were defined as DEX/CRH test non-suppres-sors at baseline; all other patients were regarded as baselinesuppressors. A separate analysis of non-suppressors and sup-pressors at baseline (week 0) demonstrated that this rapidinhibition of HPA axis activity mainly occurred in baselinenon-suppressors whereas in suppressors who already showeda strong suppression of CRH-induced COR secretion by pre-treatment with dexamethasone. Almost no further reductionof HPA axis function could be observed in this group (Fig. 2).

When analyzing non-suppressors and suppressors sepa-rately, it became obvious that ESC increases HPA axis functionin particular in suppressors after 1 week of treatment andcauses elevated COR concentrations even after 5 weeks oftreatment. However, in non-suppressors there was only aweak and transient increase in COR levels after 1-weektreatment with ESC whereas the COR concentrations weredown-regulated and lower as compared to baseline (week 0)after 5 weeks of ESC treatment (Fig. 2 and Table 2).

Since the ESC-induced increase in HPA axis function wasmost pronounced after 1 week of treatment and only tran-sient insofar as the COR levels fell again after 5 weeks of ESCtherapy at least to some extent, no significant ‘‘time’’ effectfor COR AUC levels (tests 1—3) could be demonstrated,neither in non-suppressors (F = 1.927; df = 2, 14; p = 0.182)nor in suppressors (F = 0.277; df = 2, 10; p = 0.763). A RmA-NOVA demonstrated a significant ‘‘time’’ effect for COR AUCvalues (tests 1—3) in the QXR group for baseline non-sup-pressors (F = 11.998; df = 2, 8; p = 0.004) but not for baselinesuppressors (F = 1.098; df = 2, 13; p = 0.362).

It is also important to note that in our trial, the fastinhibiting effects of QXR on HPA axis activity in depressedpatients could predominantly be observed in baseline non-suppressors who showed an approximately 4-fold decrease ofCOR levels in the DEX/CRH test after 1 week of QXR treat-ment. However, suppressors were already characterized byvery low COR levels before treatment and therefore werescarcely influenced in their DEX/CRH test results by the CORinhibitor quetiapine.

3.2.3. Separate analysis of HPA axis function inresponders and non-respondersIn the QXR group, 15 out of 25 completers (60.0% of patients)were responders, whereas 18 out of 28 completers (64.3% of

Figure 2 Mean COR concentrations during 3 DEX/CRH tests (week 0, 1, 5) in depressed patients treated with QXR (300 mg/day) or ESC(10 mg/day) subdivided into non-suppressors and suppressors. SEM (standard error of mean) indicated.

146 N. Sarubin et al.

patients) in the ESC group showed a clinical response (morethan 50% reduction in the 21-HAMD sum score after 5 weeks oftreatment). A RmANOVA did not show any significant ‘‘group’’effects or interaction effects when comparing 21-HAMD sum

Table 2 COR responses to 3 DEX/CRH tests (week 0, 1, 5) in 53 ouanalysis). Data are given for all patients (n = 53) and separately for

deviation) is indicated.

CORAUC[nmol/l � min]

Week 0

All patients (n = 53)Complete sample (n = 53) 8784.53 � 8016.34

Non-suppressors (n = 26) 14,849.58 � 7541.94

Suppressors (n = 27) 2944.11 � 1273.88

Non-responders (n = 20) 8001.12 � 7278.95

Responders (n = 33) 9259.32 � 8505.90

QXR (n = 25)Complete sample (n = 25) 7388.33 � 8314.48

Non-suppressors (n = 10) 14,564.02 � 9354.83

Suppressors (n = 15) 2604.53 � 976.23

Non-responders (n = 10) 6498.97 � 5194.95

Responders (n = 15) 7981.23 � 10,010.94

ESC (n = 28)Complete sample (n = 28) 10,031.13 � 7674.45

Non-suppressors (n = 16) 15,028.05 � 6496.57

Suppressors (n = 12) 3368.58 � 1506.29

Non-responders (n = 10) 9503.37 � 8935.96

Responders (n = 18) 10,324.40 � 7142.36

scores during the 5-week treatment period between patientstreated with QXR and patients treated with ESC; however, asignificant ‘‘time’’ effect could be demonstrated indicatingthat both treatment options reduced 21-HAMD sum scores

t of 60 depressed patients treated with QXR or ESC (completerthe QXR group (n = 25) and the ESC group (n = 28). SD (standard

Week 1 Week 5

8493.83 � 11,089.55 7765.70 � 8266.4712,411.07 � 11,222.03 11,040.46 � 8830.874721.68 � 9733.69 4612.22 � 6374.39

12,546.81 � 15,056.14 7799.51 � 7535.736037.49 � 6982.47 7745.20 � 8793.46

3410.99 � 3430.39 5387.41 � 7105.914701.54 � 4407.31 8271.82 � 8800.832550.62 � 2388.12 3464.47 � 5183.084994.72 � 4868.60 3834.92 � 3746.332355.16 � 1396.14 6422.40 � 8641.13

13,032.09 � 13,457.57 9889.17 � 8764.1417,229.53 � 11,569.15 1,277,086 � 8671.647435.52 � 14,220.46 6046.92 � 7603.19

20,098.89 � 5728.15 11,764.11 � 8421.419106.09 � 8260.40 8847.54 � 9012.77

Table 3 21-HAMD sum scores in depressive patients treated with QXR (300 mg/day) or ESC (10 mg/day) for 5 weeks, respectively.SD (standard deviation) is indicated.

21-HAMD

Day 0 Day 4 Day 7 Day 14 Day 21 Day 28 Day 35

QXR (n = 25) 21.24 � 4.13 15.16 � 4.18 13.24 � 4.40 11.36 � 5.02 11.32 � 6.09 10.56 � 6.48 9.64 � 6.42ESC (n = 28) 22.75 � 6.03 16.67 � 5.14 12.60 � 5.33 10.82 � 4.73 10.17 � 4.96 9.50 � 5.78 8.82 � 6.10

Impact on cortisol and antidepressant efficacy of quetiapine and escitalopram in depression 147

over time (‘‘time’’: F = 74.645; df = 3.020, 154.029;p < 0.001; ‘‘group’’: F = 0.021; df = 1, 51; p = 0.884;‘‘time � group’’: F = 1.261; df = 3.020, 154.029; p = 0.290)(Table 3).

When comparing responders (n = 33) and non-responders(n = 20) in the total patient sample using a rmANOVA(‘‘time’’: test 1—3; ‘‘group’’: responders versus non-respon-ders), a significant ‘‘time � group’’ interaction effect(F = 4.002; df = 2, 50; p = 0.024) but no significant ‘‘group’’effect (F = 0.465; df = 1, 51; p = 0.498) could be demon-strated. The significant ‘‘time � group’’ interaction effectgave reason to the assumption that the time course of theHPA axis function during 5-week treatment with QXR or ESC isassociated with response/non-response probability. A sepa-rate analysis of both treatment groups revealed that thesignificant ‘‘time � group’’ interaction effect could mainlybe demonstrated in the ESC group reaching highly statisticalsignificance (F = 6.276; df = 2, 25; p = 0.006) whereas in theQXR group only a statistical trend for a ‘‘time � group’’interaction effect was observed (F = 2.651; df = 2, 22;p = 0.093) (Fig. 3).

Figure 3 Mean COR concentrations during 3 DEX/CRH tests (week 0,(10 mg/day) subdivided into non-responders and responders. SEM (s

As can be seen in Fig. 3, the interaction between responseand time course of HPA axis activity in the QXR group could benotably attributed to the fact that a pronounced and rapiddown-regulation of HPA axis function already after 1 week oftreatment followed by a marked ‘‘rebound’’ effect (partialre-increase of COR concentrations) after 5 weeks of therapywas mainly demonstrated in responders to QXR (Fig. 3).Moreover, responders and non-responders to ESC differedin as much as a marked increase of COR levels within theDEX/CRH test after 1 week of treatment followed by a partialbut nonetheless clear re-decrease of COR concentrationsafter 5 weeks of ESC therapy. This effect could be demon-strated in particular in ESC non-responders whereas inresponders a weak but step-by-step down-regulation ofHPA axis function over the 5-week treatment period wasseen (Fig. 3). In line with this, a significant correlationbetween COR AUC level and 21-HAMD sum score both onday 7 was only observed in the ESC (r = 0.457, p = 0.015), butnot in the QXR group (r = 0.354, p = 0.193). This result indi-cates that the mental state after one week of treatment withESC is associated with the cortisol response on day 7: high

1, 5) in depressed patients treated with QXR (300 mg/day) or ESCtandard error of mean) indicated.

148 N. Sarubin et al.

HAMD scores and high COR AUC levels seem to be related,especially within the non-response group (Fig. 3).

4. Discussion

The goal of the present randomized, active-controlled, open-label study was to investigate the impact of 5-week pharma-cotherapy with QXR or ESC on HPA axis activity as measuredby serial DEX/CRH tests in depressed inpatients and theantidepressant efficacy.

At baseline (week 0) patients treated with QXR or ESC didnot differ significantly in clinical or demographic variables inthe present investigation. Although the DEX/CRH test hasbeen reported not to be largely influenced by disease-unre-lated factors, such as the presence or absence of antidepres-sant pharmacotherapy at admission (Kunzel et al., 2003),effects of drug withdrawal on the DEX/CRH test cannot becompletely ruled out. However, the mean duration of thewash-out period and the type of antidepressant pre-treat-ment were comparable between the QXR group and the ESCgroup. Moreover, patients with and without pre-treatmentdid not differ significantly with regard to DEX/CRH testresults.

According to the 21-HAMD sum scores, the antidepressantefficacy of ESC and QXR was comparable. The response ratesin our study were in line with former clinical studies inves-tigating the antidepressant effects of QXR (Komossa et al.,2010) or ESC (Cipriani et al., 2009).

In spite of similar clinical effects, QXR and ESC stronglydiffered in their impact on the time course of HPA axisactivity in depressed patients, thereby challenging theassumption of a uniform down-regulating effect of all psy-chotropic drugs with antidepressant properties, irrespectiveof their mode of action. In contrast to QXR, the selectiveserotonin reuptake inhibitor ESC transiently increased HPAaxis activity in depressed patients after 1 week of treatmentin our study, but returned to primary COR levels in the DEX/CRH test after 5 weeks of treatment. This is the first studydemonstrating transient enhancement of HPA axis activity inacutely depressed patients treated with an antidepressantdrug for several weeks. Notably, the transient enhancementof HPA axis activity during ESC treatment observed in ourstudy was more pronounced in baseline suppressors which areapparently more sensitive to the stimulatory effects of ESC.The effect was apparent even after 5 weeks of pharma-cotherapy with ESC in these patients. However, in baselinenon-suppressors, a down-regulation of HPA system functionto some extent was seen after 5-week treatment with ESC,which may be explained by a restoration of the disturbednegative feedback control.

In healthy controls, the different acute effects of singlechallenge with QXR (de Borja et al., 2005) and ESC (Nadeemet al., 2004) on COR secretion are reflected by a differentialimpact of these drugs on the time course of the HPA systemfunction in major depression. QXR acutely inhibits CORsecretion in healthy subjects due to a reduction of hypotha-lamic CRH release via antagonism at central 5-HT2, H1, and a1

receptors. The present study shows that, in depressedpatients, QXR also rapidly down-regulates HPA axis activitywithin 1 week of therapy, followed by a partial re-increase ofCOR concentrations in the DEX/CRH test after 5 weeks of

treatment which may be explained by an compensatory up-regulation and increased sensitivity of CRH receptors at thepituitary level. These results are exactly in line withthe findings of a former study of our research group, reportingthe same endocrinological effects on the HPA system indepressed patients after 5-week treatment with the 5-HT2and H1-blocker mirtazapine (Schule et al., 2006). Besideseffects of quetiapine and its metabolite norquetiapine onnoradrenergic (McIntyre et al., 2007), serotonergic (Gold-stein, 1999), and dopaminergic (Ichikawa et al., 2002) neu-rotransmission, the strong inhibitory impact of quetiapine onHPA axis activity shown in our study may also contribute to itsantidepressant properties.

A general and gradual down-regulation of HPA axis hyper-activity in depressed patients, as measured by serial DEX/CRH tests, has been demonstrated for tricyclic antidepres-sants such as amitriptyline (Heuser et al., 1996), doxepin(Deuschle et al., 1997), or trimipramine (Holsboer-Trachsleret al., 1991; Hatzinger et al., 2002; Frieboes et al., 2003), forthe noradrenaline-reuptake inhibitor reboxetine (Schuleet al., 2006), for the SSRIs paroxetine (Nickel et al., 2003)and citalopram (Bschor et al., 2012), and for tianeptinewhich enhances the presynaptic reuptake of serotonin(Nickel et al., 2003). This study supports the general assump-tion, that a long-term normalization of the HPA axis hyper-activity is a correlate of clinical improvement (response),regardless of the mechanism of action of the antidepressantdrug and regardless of the initial changes of the HPA axisduring week 1.

Regarding the underlying mechanisms, it can by hypothe-sized that reuptake inhibiting antidepressants such as rebox-etine, SSRIs or tricyclic antidepressants acutely stimulateCOR secretion both in healthy subjects (Laakmann, 1988;Schule, 2007) and in depressed patients (Asnis et al., 1992)after single administration and may gradually normalize HPAaxis hyperactivity in depressed patients when administereddaily for several weeks via up-regulation of mineralocorticoidreceptor and glucocorticoid receptor mRNA levels (Bradyet al., 1991; Seckl and Fink, 1992), down-regulation ofpro-opiomelanocortin mRNA expression in the pituitary gland(Jensen et al., 2001), and decrease of CRH gene expressionand CRH mRNA synthesis in the paraventricular nucleus (Moriet al., 1998; Stout et al., 2002), thereby enhancing miner-alocorticoid receptor and glucocorticoid receptor functionand restoring the disturbed feedback control (‘‘reset’’mechanism). It can also be assumed that these effects ofantidepressants on gene expression represent physiologicaladaptive mechanisms which are triggered by the primarilyacute stimulatory effects of reuptake inhibitors on the ACTHand COR release and take several weeks to become effective.This hypothesis is in line with clinical studies showing thatshort-term treatment with mineralocorticoid or glucocorti-coid agonists such as dexamethasone (Dinan et al., 1997) orhydrocortisone (DeBattista et al., 2000) may enhance thefeedback control of the HPA system and may have antide-pressant effects in humans. Recent studies have highlightedadditional mechanisms by which antidepressants regulate GRand the HPA axis, including direct stimulation of GR phos-phorylation and increased neurogenesis (Anacker et al.,2011; Anacker and Pariante, 2012), which might also playa role in the reduction of HPA axis activity after 5 weeks andthe associated clinical improvement.

Impact on cortisol and antidepressant efficacy of quetiapine and escitalopram in depression 149

Limitations of this study include the lack of a placebogroup as well as the missing assessment of estrogen levels(concerning neuroendocrinological factors). However, thedirection of the estrogen effect seems to be inconsistentand the precise mechanism is not known yet (Unschuld et al.,2010; Bangasser, 2013; Fernandez-Guasti et al., 2012).Another limitation of the present investigation lies withinthe lack of individualized antidepressant treatment, sinceESC and QXR were administered using fixed doses. The usualdose of ESC in unipolar depressed patients varies between 10and 15 mg/day and results in significant improvement on allclinical measures with both doses (Undurraga and Baldessar-ini, 2012). Quetiapine fumarate has been demonstrated tohave significant antidepressant efficacy in unipolar depres-sion when given as a monotherapy at doses of 150 or 300 mg/day (Cutler et al., 2009; Katila et al., 2012; Liebowitz et al.,2010). In this study, ESC was given as a fixed dose of 10 mg/day and QXR was given as a fixed dose of 300 mg/day.

The ‘‘minimum’’ dosage of 10 mg of ESC does not lead toworse clinical outcomes than 15 mg (Undurraga and Baldes-sarini, 2012), whereas the ‘‘medium’’ dosage of 300 mg QXRmay be more effective in patients with severe depressionthan 150 mg (Weisler et al., 2012). Thus, both treatmentregimens are within the therapeutic range. However, itcannot be excluded that the relatively higher dose of QXRcontributes to its clinical efficacy (Han et al., 2013). Never-theless, clinical efficacy of both treatment regimens werecomparable in our study (Table 2: 10 QXR non-responders and10 ESC non-responders). The goal of our study was to inves-tigate the effects of both treatment regimens on HPA axisactivity and not to compare clinical efficacy of these treat-ments. We chose the respective dosage regimens in order tocompare the maximum acute stimulating effect of ESC (Asniset al., 1992) and the assumed maximum acute inhibitoryimpact of QXR on HPA axis activity in order to clearly separateboth treatment regimens at an endocrinological level. Also,the results are limited by the circumstance that participantswere not prohibited from continuing with or seeking psycho-logical treatment for the duration of the study.

In conclusion, our data provide intriguing insight in theregulation of HPA system of depressed patients treated withantidepressants or atypical antipsychotics with antidepres-sant properties. Apparently, different effects on HPA axisactivity reflect distinct pharmacoendocrinological propertiesof psychotropic drugs being used in the treatment of depres-sion.

Role of funding source

This study was supported by an unrestricted grant fromAstraZeneca Germany.

Conflicts of interest

This study was supported by an unrestricted grant fromAstraZeneca Germany. Rainer Rupprecht has been on Astra-Zeneca advisory boards. Thomas C. Baghai accepted paidspeaking engagements and acted as a consultant for Astra-Zeneca, Glaxo-Smith-Kline, Janssen-Cilag, Pfizer and Ser-vier. Nina Sarubin, Caroline Nothdurfter, Manfred Uhr, Mar-tin Lieb, Christian Schmotz, Anna-Maria Wimmer, Julia

Trummer, Thomas Wetter, Markus Buhner and CorneliusSchule reported no direct or indirect financial or personalrelationships, interests, and affiliations relevant to thesubject matter of the manuscript that have occurred overthe last three years, or that are expected in the foreseeablefuture.

Acknowledgements

Cornelius Schule, M.D., who is the principal investigator ofthe study and is independent of any commercial funder hadfull access to all of the data in the study and takes respon-sibility for the integrity of the data and the accuracy of thedata analysis.

Appendix A. Supplementary data

Supplementary data associated with this article can befound, in the online version, at http://dx.doi.org/10.1016/j.psyneuen.2013.10.008.

References

American Psychiatric Association, 1994. Diagnostic and StatisticalManual of Mental Disorders, 4th ed. American Psychiatric Press,Washington, DC.

Anacker, C., Pariante, C.M., 2012. Can adult neurogenesis bufferstress responses and depressive behaviour? Mol. Psychiatry 17,9—10.

Anacker, C., Zunszain, P.A., Cattaneo, A., Carvalho, L.A., Garabe-dian, M.J., Thuret, S., Price, J., Pariante, C.M., 2011. Antide-pressants increase human hippocampal neurogenesis byactivating the glucocorticoid receptor. Mol. Psychiatry 16,738—750.

Anderson, I.M., Sarsfield, A., Haddad, P.M., 2009. Efficacy, safety andtolerability of quetiapine augmentation in treatment resistantdepression: an open-label, pilot study. J. Affect. Disord. 117,116—119.

Asnis, G.M., Sanderson, W.C., van Praag, H.M., 1992. Cortisol re-sponse to intramuscular desipramine in patients with majordepression and normal control subjects: a replication study.Psychiatry Res. 44, 237—250.

Bangasser, D.A., 2013. Sex differences in stress-related receptors:‘‘micro’’ differences with ‘‘macro’’ implications for mood andanxiety disorders. Biol. Sex Differ. 4, 2—13.

Bauer, M., El-Khalili, N., Datto, C., Szamosi, J., Eriksson, H., 2010. Apooled analysis of two randomised, placebo-controlled studies ofextended release quetiapine fumarate adjunctive to antidepres-sant therapy in patients with major depressive disorder. J. Affect.Disord. 127, 19—30.

Brady, L.S., Whitfield Jr., H.J., Fox, R.J., Gold, P.W., Herkenham, M.,1991. Long-term antidepressant administration alters corticotro-pin-releasing hormone, tyrosine hydroxylase, and mineralocorti-coid receptor gene expression in rat brain. Therapeuticimplications. J. Clin. Invest. 87, 831—837.

Bschor, T., Ising, M., Erbe, S., Winkelmann, P., Ritter, D., Uhr, M.,Lewitzka, U., 2012. Impact of citalopram on the HPA system. Astudy of the combined DEX/CRH test in 30 unipolar depressedpatients. J. Psychiatr. Res. 46, 111—117.

Calabrese, J.R., Keck Jr., P.E., Macfadden, W., Minkwitz, M., Ketter,T.A., Weisler, R.H., Cutler, A.J., McCoy, R., Wilson, E., Mullen, J.,2005. A randomized, double-blind, placebo-controlled trial ofquetiapine in the treatment of bipolar I or II depression. Am. J.Psychiatry 162, 1351—1360.

150 N. Sarubin et al.

Cipriani, A., Santilli, C., Furukawa, T.A., Signoretti, A., Nakagawa,A., McGuire, H., et al., 2009. Escitalopram versus other anti-depressive agents for depression. CDS Rev. CD006532.

Cutler, A.J., Montgomery, S.A., Feifel, D., Lazarus, A., Astrom, M.,Brecher, M., 2009. Extended release quetiapine fumarate mono-therapy in major depressive disorder: a placebo- and duloxetine-controlled study. J. Clin. Psychiatry 70, 526—539.

De Boer, T., 1995. The effects of mirtazapine on central noradrener-gic and serotonergic neurotransmission. Int. Clin. Psychopharma-col. 10 (Suppl. 4) 19—23.

De Borja Goncalves Guerra, A., Castel, S., Benedito-Silva, A.A., Calil,H.M., 2005. Neuroendocrine effects of quetiapine in healthyvolunteers. Int. J. Neuropsychopharmacol. 8, 49—57.

DeBattista, C., Posener, J.A., Kalehzan, B.M., Schatzberg, A.F., 2000.Acute antidepressant effects of intravenous hydrocortisone andCRH in depressed patients: a double-blind, placebo-controlledstudy. Am. J. Psychiatry 157, 1334—1337.

Deuschle, M., Schmider, J., Weber, B., Standhardt, H., Korner, A.,Lammers, C.H., Schweiger, U., Hartmann, A., Heuser, I., 1997.Pulse-dosing and conventional application of doxepin: effects onpsychopathology and hypothalamus-pituitary-adrenal (HPA) sys-tem. J. Clin. Psychopharmacol. 17, 156—160.

Devarajan, S., Ali, J., Dursun, S.M., 2006. Quetiapine plus SSRI intreatment-resistant depression: possible mechanisms. Psycho-pharmacology (Berl). 185 (3) 402—403.

Dinan, T.G., Lavelle, E., Cooney, J., Burnett, F., Scott, L., Dash, A.,Thakore, J., Berti, C., 1997. Dexamethasone augmentation intreatment-resistant depression. Acta Psychiatr. Scand. 95, 58—61.

El-Khalili, N., Joyce, M., Atkinson, S., Buynak, R.J., Datto, C.,Lindgren, P., et al., 2010. Extended-release quetiapine fumarate(quetiapine XR) as adjunctive therapy in major depressive disor-der (MDD) in patients with an inadequate response to ongoingantidepressant treatment: a multicentre, randomized, double-blind, placebo-controlled study. Int. J. Neuropsychopharmacol.13, 917—932.

Fernandez-Guasti, A., Fiedler, J.L., Herrera, L., Handa, R.J., 2012.Sex, stress, and mood disorders: at the intersection of adrenaland gonadal hormones. Horm. Metab. Res. 44 (8) 607—618.

Forsythe, A.I., Keenan, T.A., Organick, E.I., Stenberg, W., 1969.Computer Science: A First Course. John Wiley and Sons Inc.,New York.

Frieboes, R.M., Sonntag, A., Yassouridis, A., Eap, C.B., Baumann, P.,Steiger, A., 2003. Clinical outcome after trimipramine in patientswith delusional depression — a pilot study. Pharmacopsychiatry36, 12—17.

Goldstein, J.M., 1999. Quetiapine fumarate (Seroquel): a new atypi-cal antipsychotic. Drugs Today (Barc). 35, 193—210.

Green, B., 1999. Focus on quetiapine. Curr. Med. Res. Opin. 15, 145—151.

Han, C., Wang, S.M., Kato, M., Lee, S.J., Patkar, A.A., Masand, P.S.,Pae, C.U., 2013. Second-generation antipsychotics in the treat-ment of major depressive disorder: current evidence. Expert Rev.Neurother. 13 (7) 851—870.

Hatzinger, M., Hemmeter, U.M., Baumann, K., Brand, S., Holsboer-Trachsler, E., 2002. The combined DEX-CRH test in treatmentcourse and long-term outcome of major depression. J. Psychiatr.Res. 36, 287—297.

Heuser, I.J., Schweiger, U., Gotthardt, U., Schmider, J., Lammers,C.H., Dettling, M., et al., 1996. Pituitary-adrenal-system regu-lation and psychopathology during amitriptyline treatment inelderly depressed patients and normal comparison subjects.Am. J. Psychiatry 153, 93—99.

Holsboer, F., 2000. The corticosteroid receptor hypothesis of depres-sion. Neuropsychopharmacology 23, 477—501.

Holsboer, F., 2001. Stress, hypercortisolism and corticosteroid recep-tors in depression: implications for therapy. J. Affect. Disord. 62,77—91.

Holsboer-Trachsler, E., Stohler, R., Hatzinger, M., 1991. Repeatedadministration of the combined dexamethasone-human cortico-tropin releasing hormone stimulation test during treatment ofdepression. Psychiatry Res. 38, 163—171.

Ichikawa, J., Li, Z., Dai, J., Meltzer, H.Y., 2002. Atypical antipsy-chotic drugs, quetiapine, iloperidone, and melperone, preferen-tially increase dopamine and acetylcholine release in rat medialprefrontal cortex: role of 5-HT1A receptor agonism. Brain Res.956, 349—357.

Ising, M., Kunzel, H.E., Binder, E.B., Nickel, T., Modell, S., Holsboer,F., 2005. The combined dexamethasone/CRH test as a potentialsurrogate marker in depression. Prog. Neuropsychopharmacol.Biol. Psychiatry 29, 1085—1093.

Ising, M., Horstmann, S., Kloiber, S., Lucae, S., Binder, E.B., Kern, N.,et al., 2007. Combined dexamethasone/corticotropin releasinghormone test predicts treatment response in major depression —a potential biomarker? Biol. Psychiatry 62, 47—54.

Jensen, J.B., Mork, A., Mikkelsen, J.D., 2001. Chronic antidepressanttreatments decrease pro-opiomelanocortin mRNA expression inthe pituitary gland: effects of acute stress and 5-HT(1A) receptoractivation. J. Neuroendocrinol. 13, 887—893.

Katila, H., Mezhebovsky, I., Mulroy, A., Berggren, L., Eriksson, H.,Earley, W., et al., 2012. Randomized, double-blind study of theefficacy and tolerability of extended release quetiapine fumarate(Quetiapine XR) monotherapy in elderly patients with majordepressive disorder. Am. J. Geriatr. Psychiatry, http://dx.doi.org/10.1097/JGP.0b013e3182545e50.

Komossa, K., Depping, A.M., Gaudchau, A., Kissling, W., Leucht, S.,2010. Second-generation antipsychotics for major depressivedisorder and dysthymia. CDS Rev. CD008121.

Kunzel, H.E., Binder, E.B., Nickel, T., Ising, M., Fuchs, B., Majer, M.,et al., 2003. Pharmacological and nonpharmacological factorsinfluencing hypothalamic-pituitary-adrenocortical axis reactiv-ity in acutely depressed psychiatric in-patients, measuredby the Dex-CRH test. Neuropsychopharmacology 28, 2169—2178.

Laakmann, G., 1988. Psychopharmaco-endocrinology and depressionresearch. Monogr. Gesamtgeb. Psychiatr. Psychiatry Ser. 46, 1—220.

Liebowitz, M., Lam, R.W., Lepola, U., Datto, C., Sweitzer, D.,Eriksson, H., 2010. Efficacy and tolerability of extended releasequetiapine fumarate monotherapy as maintenance treatment ofmajor depressive disorder: a randomized, placebo-controlledtrial. Depress. Anxiety 27, 964—976.

McElroy, S.L., Weisler, R.H., Chang, W., Olausson, B., Paulsson, B.,Brecher, M., et al., 2010. A double-blind, placebo-controlledstudy of quetiapine and paroxetine as monotherapy in adults withbipolar depression (EMBOLDEN II). J. Clin. Psychiatry 71, 163—174.

McIntyre, A., Gendron, A., McIntyre, A., 2007. Quetiapine adjunct toselective serotonin reuptake inhibitors or venlafaxine in patientswith major depression, comorbid anxiety, and residual depressivesymptoms: a randomized, placebo-controlled pilot study. De-press. Anxiety 24, 487—494.

Mori, S., Zanardi, R., Popoli, M., Garbini, S., Brunello, N., Smeraldi,E., et al., 1998. cAMP-dependent phosphorylation system aftershort and long-term administration of moclobemide. J. Psychiatr.Res. 32, 111—115.

Nadeem, H.S., Attenburrow, M.J., Cowen, P.J., 2004. Comparison ofthe effects of citalopram and escitalopram on 5-Ht-mediatedneuroendocrine responses. Neuropsychopharmacology 29,1699—1703.

Nickel, T., Sonntag, A., Schill, J., Zobel, A.W., Ackl, N., Brunnauer,A., et al., 2003. Clinical and neurobiological effects of tianeptineand paroxetine in major depression. J. Clin. Psychopharmacol.23, 155—168.

Plenge, P., Gether, U., Rasmussen, S.G., 2007. Allosteric effects of R-and S-citalopram on the human 5-HT transporter: evidence for

Impact on cortisol and antidepressant efficacy of quetiapine and escitalopram in depression 151

distinct high- and low-affinity binding sites. Eur. J. Pharmacol.567, 1—9.

Sagud, M., Mihaljevic-Peles, A., Muck-Seler, D., Jakovljevic, M.,Pivac, N., 2006. Quetiapine augmentation in treatment-resistantdepression: a naturalistic study. Psychopharmacology (Berl) 187,511—514.

Schule, C., 2007. Neuroendocrinological mechanisms of actions ofantidepressant drugs. J. Neuroendocrinol. 19, 213—226.

Schule, C., Baghai, T., Goy, J., Bidlingmaier, M., Strasburger, C.,Laakmann, G., 2002. The influence of mirtazapine on anteriorpituitary hormone secretion in healthy male subjects. Psycho-pharmacology (Berl) 163, 95—101.

Schule, C., Baghai, T.C., Eser, D., Zwanzger, P., Jordan, M., Buechs,R., et al., 2006. Time course of hypothalamic-pituitary-adreno-cortical axis activity during treatment with reboxetine and mir-tazapine in depressed patients. Psychopharmacology (Berl) 186,601—611.

Schule, C., Baghai, T.C., Eser, D., Hafner, S., Born, C., Herrmann, S.,et al., 2009. The combined dexamethasone/CRH Test (DEX/CRHtest) and prediction of acute treatment response in major de-pression. PLoS ONE 4, e4324, http://dx.doi.org/10.1371/jour-nal.pone.0004324.

Seckl, J.R., Fink, G., 1992. Antidepressants increase glucocorticoidand mineralocorticoid receptor mRNA expression in rat hippo-campus in vivo. Neuroendocrinology 55, 621—626.

Stout, S.C., Owens, M.J., Nemeroff, C.B., 2002. Regulation of corti-cotropin-releasing factor neuronal systems and hypothalamic-pituitary-adrenal axis activity by stress and chronic antidepres-sant treatment. J. Pharmacol. Exp. Ther. 300, 1085—1092.

Undurraga, J., Baldessarini, R.J., 2012. Randomized, placebo-con-trolled trials of antidepressants for acute major depression:thirty-year meta-analytic review. Neuropsychopharmacology 37(4) 851—864.

Unschuld, P.G., Ising, M., Roeske, D., Erhardt, A., Specht, M., Kloiber,S., Uhr, M., Mueller-Myhsok, B., Holsboer, F., Binder, E.B., 2010.Gender-specific association of galanin polymorphisms with hpa-axis dysregulation, symptom severity, and antidepressant treat-ment response. Neuropsychopharmacology 35, 1583—1592.

Weisler, R.H., Montgomery, S.A., Earley, W.R., Szamosi, J., Lazarus,A., 2012. Efficacy of extended release quetiapine fumaratemonotherapy in patients with major depressive disorder: a pooledanalysis of two 6-week, double-blind, placebo-controlled studies.Int. Clin. Psychopharmacol. 27 (1) 27—39.

Wittchen, H.U., Wunderlich, U., Zaudig, M., Fydrich, T., 1997.Strukturiertes Klinisches Interview fur DSM-IV, Achse I: PsychischeStorungen. Interviewheft. Hogrefe, Gottingen.

Zobel, A.W., Nickel, T., Sonntag, A., Uhr, M., Holsboer, F., Ising, M.,2001. Cortisol response in the combined dexamethasone/CRHtest as predictor of relapse in patients with remitted depression.a prospective study. J. Psychiatr. Res. 35, 83—94.