Psychiatry Research: Neuroimaging 202 (2012) 12–19

Contents lists available at SciVerse ScienceDirect

Psychiatry Research: Neuroimaging

j ourna l homepage: www.e lsev ie r .com/ locate /psychresns

Cognitive and volumetric predictors of response to repetitive transcranial magneticstimulation (rTMS) — A prospective follow-up study

Christina P. Furtado a,⁎, Kate E. Hoy a, Jerome J. Maller a, Greg Savage b,Zafiris J. Daskalakis c, Paul B. Fitzgerald a

a Monash Alfred Psychiatry Research Centre, The Alfred and Monash University School of Psychology and Psychiatry, Melbourne, Australiab Department of Psychology and ARC Centre of Excellence in Cognition and Its Disorders, Macquarie University, Sydney, NSW, Australiac Centre for Addiction and Mental Health, Toronto, Canada

⁎ Corresponding author at: Monash Alfred PsychiatrOld Baker Building, The Alfred Hospital, Commercial RoAustralia. Tel.: +61 3 9076 6564; fax: +61 3 9076 658

E-mail address: Christina.Furtado@monash.edu (C.P

0925-4927/$ – see front matter © 2012 Elsevier Irelanddoi:10.1016/j.pscychresns.2012.02.004

a b s t r a c t

a r t i c l e i n f o

Article history:Received 29 September 2011Received in revised form 30 January 2012Accepted 8 February 2012

Keywords:Volumetric analysisImagingTreatment resistant depressionCognitive functionAutomated segmentation

As the prevalence of treatment resistant depression (TRD) continues to rise, it remains a clinically importantissue to identify neurobiological-, patient- and treatment-related factors that could potentially predict responseto treatment. Medial temporal lobe (MTL) structures, in particular the hippocampus and amygdala have beenimplicated in inferior treatment response. The role of related structures such as the entorhinal cortex and the im-pact of MTL abnormalities on neurocognitive function, however, have not been systematically examined. Thecurrent study investigated MTL abnormalities and neurocognitive characteristics of eventual treatment re-sponders and non-responders to a course of repetitive transcranialmagnetic stimulation (rTMS) in order to iden-tify potential predictors of treatment outcome. Prior to rTMS treatment all patients underwent magneticresonance imaging (MRI) and neuropsychological assessment. MRI analysis was conducted using FreeSurfer5.0. Therewas a 50% response rate following up to a 6-week course of daily rTMS treatments. Treatment responsewas defined as 50% reduction in Hamilton Depression Rating Scale and BDI-II scores from baseline. There was nodifference in pre-treatment neurocognitive profiles and MTL volumes between eventual treatment respondersand non-responders. Smaller pre-treatment left hippocampus volume showed a trend towards predicting even-tual subjective improvement in depressive symptomatology. Although preliminary, our findings suggest thatstructural abnormalities may have some potential for predicting outcome to rTMS.

© 2012 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Major depressive disorder (MDD) is a common, disabling anddifficult-to-treat psychiatric disorder. Although, there is a range ofestablished treatments for MDD, including antidepressant medications,psychotherapy and electroconvulsive therapy (ECT) (Fitzgerald andDaskalakis, 2011), MDD is often resistant to treatment with standardapproaches, with approximately 30% of patients meeting standard def-initions for treatment resistant depression (TRD) (Fava, 2003). Themain treatment option for patients with TRD is ECT (Brakemeier et al.,2007). Despite proven high efficacy of ECT in MDD, its use is often lim-ited due to patient refusal from concerns about marked cognitive sideeffects, risks from repeated general anaesthetics and stigma(Fitzgerald and Daskalakis, 2011). Repetitive transcranial magneticstimulation (rTMS) is a relatively new brain stimulation techniquethat offers a potential alternative for patients with TRD. rTMS involvesthe production of a magnetic field via an alternating electric current

y Research Centre, First Floor,ad, Melbourne, Victoria 3004,8.. Furtado).

Ltd. All rights reserved.

(Barker, 1991). This magnetic field passes into the brain and stimulateselectrical activity in neurons; high-frequency rTMS increases brain ac-tivity, whereas low-frequency stimulation decreases brain activity(Hoy and Fitzgerald, 2010). Imaging studies have demonstrated thatMDDmay involve dysregulation of cortical activity and rTMS is thoughtto act by normalising hypoexcitability over the left prefrontal cortexand normalising hyperexcitability over the right hemisphere(Daskalakis et al., 2008).

Several meta-analyses have provided support for rTMS treatmenthaving significantly more antidepressant efficacy than sham treatment(Burt et al., 2002; Kozel and George, 2002; Martin et al., 2003;Couturier, 2005; Loo and Mitchell, 2005; Fregni et al., 2006; Lam et al.,2008; Schutter, 2009). In particular, both high-frequency left (HFL; usu-ally between 5 and 20 Hz) and low-frequency right (LFR; ≤1Hz) rTMSapplied over the dorsolateral prefrontal cortex (DLPFC) have beenshown to be effective in the treatment of TRD (Daskalakis et al., 2008;Fitzgerald et al., 2010; Rossini et al., 2010; Fitzgerald and Daskalakis,2011). Overall, there is overwhelming support for a significant reduc-tion in depressive symptomatology following rTMS; however, the per-centage of patients responding to a course of rTMS treatment is lessthan 50% (Burt et al., 2002; Loo and Mitchell, 2005; Brakemeier et al.,2007; Fitzgerald et al., 2010; Rossini et al., 2010). Therefore, it remains

13C.P. Furtado et al. / Psychiatry Research: Neuroimaging 202 (2012) 12–19

a clinically important issue to identify biological-, patient- andtreatment-related factors that could potentially predict response torTMS at baseline and aid in patient selection (Loo and Mitchell, 2005;Brakemeier et al., 2008; Lisanby et al., 2009).

Although a number of studies have explored this question to date, nostrong predictors have been identified that would be able to predict re-sponse to rTMS with sufficient sensitivity and specificity to be of thera-peutic use (Fitzgerald and Daskalakis, 2011). Of the potential predictorsthat have emerged, degree of treatment resistance and demographicfactors have shown significant results. That is, younger age (Fregni etal., 2006) and a lower degree of medication resistance in the current ep-isode (Fregni et al., 2006; Brakemeier et al., 2007; Brakemeier et al.,2008; Lisanby et al., 2009) were found to predict better antidepressantresponse to rTMS. Furthermore, in one study the likelihood of responseto rTMS was four times higher if patients had only received one unsuc-cessfulmedication trial before rTMS in comparisonwith patients havingreceived two or more unsuccessful trials (O'Reardon et al., 2007). Otherpositive predictors of improvementwith rTMS include short duration ofthe current episode (Brakemeier et al., 2007; Lisanby et al., 2009), a highlevel of sleep disturbances (Brakemeier et al., 2007), female gender, ab-sence of a co-morbid anxiety disorder and a higher baseline depressionseverity (Lisanby et al., 2009; Rossini et al., 2010). Conversely, presenceof personality disorder (Fitzgerald and Daskalakis, 2011) and psychoticsymptoms (Grunhaus et al., 2000) have been linked to inferior rTMSoutcome.

Neuropsychological assessments are often included in clinical trials ofrTMS efficacy to monitor the safety of the technique. These assessmentshave frequently demonstrated that rTMS has no detrimental cognitiveside-effects after several weeks of daily treatments in clinical samples(Triggs et al., 1999; Little et al., 2000; Loo and Mitchell, 2005; Fitzgeraldet al., 2006; Januel et al., 2006; Vanderhasselt, 2009). In fact, some studieshave reported improvement in cognitive function following a course ofrTMS, namely in the domains of attention, concentration, workingmem-ory and processing speed, with probable flow-on effects leading to im-provements in learning, memory and aspects of executive functioning(Loo et al., 2008; Vanderhasselt, 2009). These beneficial effects are pri-marily reported amongst treatment responders and therefore neuropsy-chological gains are likely secondary to symptomatic remission (Loo etal., 2008). There is some preliminary evidence, however, that suggeststhat cognitive changes may occur earlier than mood improvement(Vanderhasselt, 2009). These early changes in cognition may provide auseful clinical tool to predict eventual response to rTMS. Further researchinvestigating cognitive characteristics of rTMS responders and non-responders is highly warranted.

Recent studies have also investigated the influence of neurobiologicalmarkers on treatment outcome and identified lower concentrations ofglutamate in the DLPFC (Luborzewski et al., 2007) and reduced cerebralblood flow in the amygdala at baseline (Nadeau et al., 2002) as positivepredictors of rTMS outcome. It has also been demonstrated that rTMS af-fects underlying regional brain activation and perfusion (Langguth et al.,2007; Lisanby et al., 2009). Left DLPFC stimulation has been shown to in-duce changes in deeper regions such as the anterior cingulate cortex(ACC), basal ganglia, thalamus and limbic system, areas known to be in-volved in the pathophysiology of MDD (Albus et al., 1996; Soares andMann, 1997; Sheline, 2006; Langguth et al., 2007; Price and Drevets,2010). Accordingly, onemight speculate that rTMS exerts its antidepres-sant effects by modulating this cortico-limbic connectivity (Langguth etal., 2007; Price and Drevets, 2010) with fMRI studies confirming abnor-mal functioning in frontal (e.g., ACC and DLPFC) and limbic connections(e.g., hippocampus, amygdala) in depression (Lorenzetti et al., 2009;McClintock et al., 2010).

The limbic structures are of particular interest as volumetric reduc-tions especially in the hippocampus have been linked to decreased re-sponse to antidepressant medications, poor clinical outcomes, andincreased rates of relapse (Frodl et al., 2004, 2008; Kronmüller et al.,2008; MacQueen, 2009; Li et al., 2010). Several studies have shown

that patients who achieve remission have larger pre-treatment hippo-campal volumes bilaterally than patients who remain depressed(Vakili et al., 2000; Frodl et al., 2004; MacQueen et al., 2008;MacQueen, 2009). Furthermore, in prospective studies, smaller baselinehippocampal volumes were found to be predictive of poorer clinicaloutcome at 1 and3 years follow-up (Frodl et al., 2004, 2008). The sourceof hippocampal volume abnormalities in MDD have been postulated toresult from the ‘neurotoxic effects of stress’ through repeated episodesof hypercortisolemia, glutamate neurotoxicity, decreased neurogenesis,glial cell loss, and decreased expression of brain derived neurotrophicfactor (Caetano et al., 2004; Duman and Monteggia, 2006; Czéh andLucassen, 2007). Given shared glutamatergic transmission betweenthe hippocampus and nearby structures of the medial temporal lobe(MTL), it is not surprising that volumetric reductions have also beenfound in the amygdala (Drevets, 2003; Hamilton et al., 2008; vanEijndhoven et al., 2009; Lorenzetti et al., 2010) and entorhinal cortex(EC) (Bell-McGinty et al., 2002; Furtado et al., 2008; Gerritsen et al.,2011). Reductions in volumes of the amygdala have also been foundto correlate with poor clinical outcomes such as greater number of re-current MDD episodes and longer illness duration (Caetano et al.,2004; Frodl et al., 2004; Hastings et al., 2004; Lorenzetti et al., 2009).The EC is of additional interest because of its intimate connectionswith the hippocampus. Early findings from animal models haveshown that lesioning the EC could prevent up to 70% of hippocampaldamage caused by chronic stress (Sunanda and Raju, 1997) and there-fore it is of interest to investigate the role of this structure in chronictreatment resistant MDD illness.

Gender differences in MTL volumes in MDD have not been wellexplored. Some studies have specifically focussed on female patientsand consistently found hippocampal volume reductions comparedto female controls (Sheline et al., 1999; McKinnon et al., 2009). Thepicture of volumetric abnormalities in males is less clear. MacMasterand Kusumakar (2004) found smaller left hippocampal volumes inpaediatric MDD males, while Frodl et al. (2002) similarly found smal-ler left hippocampal volumes in adult male patients compared tohealthy controls. Maller et al. (2007) also found smaller hippocampalvolumes in males, whilst Kronmüller et al. (2008) found smaller hip-pocampal volumes in males were predictive of recurrence in MDD.When both sexes are examined together, however, conflicting find-ings have been reported (Eker and Gonul, 2010), with a meta-analysis failing to find any significant contribution of sex on hippo-campal volume (Campbell et al., 2004). It is well known that genderdifferences exist in brain size and volume. Male brains are on average,about 10% heavier and larger than female brains, however, femaleshave higher percentages of grey matter volume (GMV), even aftercorrecting for brain size (Clarke, 2003; Yan et al., 2011). Therefore,an examination of neuroanatomic abnormalities in the sexes sepa-rately would allow for genuine differences to be found.

To date no study has investigated gender differences in the volumesof the EC or hippocampus and amygdala or the role of these volumes inpredicting response to rTMS treatment. Therefore, the current studyaimed to investigate neuroanatomical (split by sex) and neuropsycho-logical characteristics of rTMS treatment responders and non-responders. To investigate these relationships, we defined response torTMS as the a priori primary endpoint measure and divided baselinemoderators into four domains: 1) demographic features (e.g., age andgender), 2) clinical features (e.g., age at onset of MDD, length of illness,length of current episode, number of previous major depressive epi-sodes, concurrent medical and Axis I psychiatric disorder [e.g., anxiety],symptom severity, resistance to treatment [e.g., number of failed trialsof antidepressant medication]), 3) structural biomarkers (volumes ofthe hippocampus, amygdala and entorhinal cortex) and 4) cognitivemarkers (e.g., attention, processing speed, memory and executive func-tion). We hypothesised that larger pre-treatment volumes of the hippo-campus, amygdala and entorhinal cortex in both sexes would beassociated with superior antidepressant efficacy of rTMS.

Table 1Means and standard deviations for ICV, TBV in litres and TBV/ICV∗100 (%).

Males (N=18) Females (N=24)

Mean S.D. Mean S.D.

ICV 1.70 0.12 1.56 0.16TBV 1.17 0.18 1.12 0.10TBV/ICV 68.81 9.78 72.02 3.25

14 C.P. Furtado et al. / Psychiatry Research: Neuroimaging 202 (2012) 12–19

2. Method

2.1. Subjects

The experimental group consisted of 46 patients aged between 22 and 63 years(M=42.67, S.D.=11.06); comprising 25 females (M=41.92, S.D.=10.94) amongwhom 21 were right-handed and four were left-handed, and 21 males (M=43.57,S.D.=11.41), among whom 17 were right-handed, two were left-handed and twowere ambidextrous. All patients were recruited either from the outpatient depart-ment of a public mental health service (Alfred Psychiatry, Melbourne) or the inpa-tient department of a private psychiatric hospital (Victoria Clinic, Melbourne), or byreferral from a variety of private psychiatrists between June 2008 and April 2011.The inclusion criteria for the study included: age 18–65 years, a diagnosis of MDD(not bipolar disorder), and a score on the Hamilton Depression Rating Scale [HDRS;(Hamilton, 1960)] of at least 20 indicating the presence of moderate to severe depres-sion. All patients fulfilled DSM-IV-TR (American Psychiatric Association, 2000) cri-teria for major depressive disorder. Diagnosis was made using the MiniInternational Interview for Neuropsychiatric Disorders (Sheehan et al., 1998) by astudy psychiatrist. All patients had to be in a current major depressive episode. Exclu-sion criteria included a concurrent or previous DSM-IV-TR Axis I disorder, DSM-IV-TRdiagnosis of alcohol or substance dependence, significant current active medicalproblem and known neurological disease or a contraindication to rTMS (e.g., historyof seizure disorder, presence of a pacemaker or metal somewhere in the head otherthan in the teeth) or to MRI scanning (e.g., aneurysm clips, stents or metal anywherein the body).

Treatment resistance was defined as non-responsiveness to at least two courses ofantidepressant medications for at least 6 weeks [Stage II, Thase and Rush (1997) defi-nition; mean number of courses across episodes±S.D.=5.85±3.7], as determined bytheir primary treating clinician and patient judgement of medication effectiveness.Medications were not allowed to have changed in the 4 weeks prior to commencementof the trial or during the trial itself. Thirty-five patients were taking antidepressantmedication during the trial: 19 a serotonin-norepinephrine reuptake inhibitor(SNRI), 10 were taking a selective serotonin reuptake inhibitor (SSRI), three a tricyclicantidepressant (TCA), one a reversible inhibitor of monoamine oxidase A antidepres-sant (RIMA), one a norepinephrine (noradrenaline) reuptake inhibitor (NRI) and onea melatonergic antidepressant/5-HT2C receptor antagonist. Eleven patients were anti-depressant medication-free. Concomitant medication also remained unchangedthroughout the study and included: benzodiazepines (N=17), mood stabilisers(N=12; lithium [n=5], sodium valproate [n=4], lamotrigine [n=1]) and augment-ing antipsychotic medications (N=14; quetiapine [n=10], amisulpride [n=2], aripi-prazole [n=1] and pericyazine [n=1]).

After complete description of the study to the subjects, written informed consentwas obtained from all patients in accordance with approval of the protocol by the Al-fred Human Subjects Research and Ethics Committee and the Melbourne ClinicHuman Research Ethics Committee.

2.2. TMS treatment

rTMS treatment was provided using a Medtronic Magpro30 magnetic stimula-tors (Medtronic Inc., USA) using 70-mm figure-of-eight coils held in custom-madestands. The coil was held tangential to the scalp with the handle pointing back andaway from the midline at 45°. rTMS treatment was provided for five consecutivedays per week for up to 6 weeks (M=26.07 daily sessions, S.D.=6.11) administeredas either 10 Hz left DLPFC stimulation (N=21) or sequential bilateral DLPFC stimu-lation (N=23). Two patients withdrew from the trial after three treatments each asthey were unable to tolerate rTMS treatment due to side-effects, and one patientcompleted suicide during the trial; thus three sets of data were removed from thesample.

2.3. Clinical assessment

The primary outcome for the study was scores on the 17-item version of the HDRS.In addition, all patients completed the 21-item Beck Depression Inventory [BDI-II;(Beck et al., 1996)] as a subjective and secondary measure of treatment response.These measures were repeated at end of active treatment (endpoint). Treatment re-sponse was defined as 50% reduction in HDRS and BDI-II scores from baseline. 23 pa-tients were treatment responders with percentage improvement of HDRS frombaseline ranging from 50% to 96.7% (M=68.37, S.D.=15.81); 7 patients attained re-mission (HDRSb7). Subjectively, on BDI-II 19 patients were treatment responderswith percentage improvement on BDI-II from baseline ranging from 50% to 100%(M=71.75, S.D.=16.28).

A neuropsychological assessment was performed at baseline using the followingtests: Wechsler Test of Adult Reading [WTAR;(Wechsler, 2001)]; Brief VisuospatialMemory Test-Revised [BVMT-R; (Benedict, 1997)] Rey Auditory Verbal Learning Test[RAVLT; (Schmidt, 1996)] Trail Making Test (Reitan and Wolfson, 1985); Digit Spansubtest from the Wechsler Adult Intelligence Scale—Third Edition [WAIS-III;(Wechsler, 1997)]; Controlled Oral Word Association Test (Benton et al., 1994) andthe Stroop Test (Golden, 1978). All assessments were conducted by clinically-trainedassessors.

2.4. Magnetic resonance image acquisition

Thirty-nine patients (16 males, 23 females) underwent magnetic resonance image(MRI) scanning. Two patients declined MRI due to severe claustrophobia and one pa-tient was unable to be scanned due to weight limitations of the scanner. Imageswere acquired on a 1.5 T Signa scanner (General Electric Medical Systems, Milwaukee,WI) using an eightchannel head coil. After a scout scan an anterior-commissureposterior-commissure aligned sagittal SPGR T1-weighted sequence was acquired forvolumetric estimations (TR=100, TE=450, matrix size=224×224, voxel dimen-sions=0.94×0.94×1.4 mm3, NEX=1, flip angle=15°). Slices were then convertedto be isotropic (0.94 mm3) so that volumetric analysis could be conducted in the coro-nal plane (see Maller et al., 2007).

2.4.1. Volumetric analysisAll scans were de-identified and entered into a database so that volumetric analy-

sis occurred blind to responder group. Image analysis was performed on a PC-compatible computer using ANALYZE® software (Version 9.0; Biomedical Imaging Re-source, Mayo Foundation Rochester, MN, USA) and SPM5 software (Wellcome Depart-ment of Cognitive Neurology, UK). All outputs were manually inspected to ensureaccurate and valid data. Cerebrospinal fluid (CSF) volumes, intracranial volumes(ICV; grey matter plus white matter plus CSF) and total brain volumes (TBV; grey mat-ter plus white matter) were computed for each subject using the SPM5 software fornormalisation purposes and as measures of brain atrophy; measures are presented inlitres (Table 1). In view of the fact, however, that TBV has been directly correlated inthe peer-reviewed literature with disease-related atrophy, the ICV was selected asthe normalisation variable to minimise interpersonal effects as past research hasshown no significant association of this variable with MDD pathology and it has beenfound to exert no significant effects on volumes of regions of interest (ROIs) (Furtadoet al., 2008).

2.4.2. Automated MRI data analysisAutomated segmentation was used in this study because it holds twomajor advan-

tages over manual segmentation procedures: significant improvement in reliabilitysince automated algorithms are usually subject and user-independent allowing forreplication of study results, and consume less operator time (Bergouignan et al.,2009; Pruessner et al., 2010). Moreover, manual, VBM and automated segmentationhave been shown to result in comparable degrees of MTL volumes in acute depressionand normal populations (Bergouignan, et al., 2009; Pruessner, et al., 2010). The onlydisadvantage of automated procedures is the limited validity of the results; however,all automated volumes were individually examined by an expert in medial temporallobe anatomy (JM) to overcome this limitation.

FreeSurfer 5.0 (http://surfer.nmr.mgh.harvard.edu) was used to estimate regionalvolumetrics. FreeSurfer is a set of tools developed to parcellate the brain's cortical andsubcortical anatomy based on T1-weighted MRI scan. Although from this, data summa-ries for each cortical region are computed (including regional volume, surface area,curvature, and thickness), only the volumetrics of the MTL structures are of interestin the current study (see Fig. 1). Inspection of automated volumes resulted in oneTRD patient's (male) volumetric data being excluded due to incorrect processingthrough FreeSurfer.

2.5. Data analysis

All datawere statistically analysed usingSPSS forWindows version 19.0 (SPSS, Chicago,Ill). In order to investigate potential predictors of outcome, percentage change scores weregenerated for the primary clinical outcome variable (HDRS) and the secondary outcomemeasure (BDI-II). Raw scores for all cognitivemeasureswere analysed expect for the Stroopwhere an interference score for performance was calculated according to the method de-scribed in the testmanual (Golden, 1978). Independent samples t-tests were used to assessdifferences between treatment responders and non-responders as determined by theHDRSand BDI-II in age, depression rating scale scores, estimated premorbid intellect (i.e., WTAR-predicted IQ) and performance on individual tests within the cognitive battery. In order toadjust for the effects of multiple comparisons pwas set at 0.01.

To adjust for differences in head size, hippocampal, amygdala and entorhinal cor-tex volumes were divided by total ICV for normalisation purposes. A series of correla-tions were then undertaken between percentage change scores on depression ratingscales and baseline volumetrics separately for males and females. Analyses weretwo-tailed and evaluated for significance at the 0.01 alpha level. For the sake of parsi-mony only significant and trend level results are reported in text and the specifics of allcomparisons are summarised in tables. Finally, a linear regression was undertaken

Fig. 1. FreeSurfer 5.0 output of single scan from coronal perspective at the level wherethe medial temporal lobe structures of the amygdala (AMYG), hippocampus (HC), andentorhinal cortex (EC) can be seen.

15C.P. Furtado et al. / Psychiatry Research: Neuroimaging 202 (2012) 12–19

which investigated the role of baseline volumetrics and gender as potential predictorsof antidepressant response to rTMS as measured by the dependent variable of eventualimprovement in depression rating scales.

3. Results

3.1. Demographics, clinical variables and psychological tests

There was no significant difference at baseline in age, demographicsor clinical variables between groups on both the HDRS and BDI-II re-sponder groups. Depression rating outcome scores significantly differedat treatment end (see Table 2). BDI-II treatment non responders scoredsignificantly higher than BDI-II treatment responders on endpointHDRS (t(40)=3.68, pb0.001) and BDI-II (t(40)=6.57, pb0.001). Per-centage improvement in clinical outcome scores also significantly dif-fered between responder groups (pb0.001). BDI-II treatment

Table 2Means and standard deviations of demographic and clinical variables in eventual treatmen

Whole grouppatients

Eventual rTMS treatm

HDRS responders

Number of subjects 42 23Age (years) 42.76 (11.53) 43.04 (10.93)Gender (males/females) 18/24 9/14Treatment group (unilateral/bilateral) 20/22 13/10Co-morbid GAD (% of sample) 19 (45.2%) 10 (43.5%)Age of onset 29.85 (13.24) 30.63 (13.20)Est. disease duration (years) 14.98 (10.51) 13.58 (11.00)Est. current episode duration (years) 6.36 (10.19) 4.77 (11.04)No. of past depressive episodes 7.11 (12.12) 9.68 (14.50)No. of failed AD medications 5.79 (3.67) 5.43 (3.72)Depression rating scale scores at baselineHDRS 24.83 (4.11) 25.13 (4.34)BDI-II 38.21 (10.10) 37.22 (9.43)Depression rating scale scores at treatment endHDRS 13.72 (8.04) 7.75 (3.87)⁎⁎

BDI-II 22.26 (14.45) 15.52 (11.50)⁎⁎

% Improvement in depression rating scale scores between baseline and endpointHDRS 43.65 (31.76) 68.37 (15.81)⁎⁎

BDI-II 42.83 (32.77) 57.56 (28.63)⁎⁎

Note: HDRS = Hamilton Depression Rating Scale; BDI-II = Beck Depression Inventory-II; G⁎⁎ Significant at pb0.01.

responders showed greater improvements in HDRS (t(40)=−3.53,pb0.001) and BDI-II (t(40)=−9.47, pb0.001) than BDI-II treatmentnon-responders.

3.2. Cognitive and behavioural data

No significant differences were found between HDRS and BDI-IIresponder groups at baseline in estimated premorbid intellect, atten-tion, processing speed, verbal and visual memory, and executivefunction (Table 3.)

3.3. Relationship between depression symptom outcomes and medialtemporal lobe volumes

There were no significant relationships across both genders be-tween the primary outcome measures of HDRS response in all MTLvolumes (Table 4). Similarly, no significant relationships were foundon the secondary outcome measures of BDI-II response. Trend levelnegative relationships were observed, however, in males betweenBDI-II percentage change and left hippocampus (r=-0.474,p=0.07) and right hippocampus (r=−0.491, p=0.06), althoughthey failed to reach significance.

3.4. Predictors of rTMS treatment outcome

Due to left and right hippocampus volumes themselves beingstrongly correlated (males: r=.917, pb0.001; females: r=0.836,pb0.001; combined: r=0.854, pb0.001), left hippocampus volumewas chosen as the predictor as it showed a modest correlation withBDI-II percentage change treatment non-responders (Fig. 2).

Gender and left hippocampal volume were found to correlate witheventual BDI-II improvement (Table 5). Therefore, amultiple regressionmodel comprising the predictors of left hippocampus volume and gen-der and the dependent variable of eventual BDI-II improvement wasundertaken which approached significance (F(2,35)=3.424,r=0.405, pb0.05) (Table 6).

4. Discussion

The objective of this study was to investigate baseline neuroana-tomical (in the sexes separately) and neurocognitive characteristicsof eventual treatment responders and non-responders to a course of

t responders and treatment non-responders.

ent outcome

HDRS non-responders BDI-II responders BDI-II non-responders

19 19 2342.42 (12.51) 44.00 (11.66) 41.74 (11.56)9/10 6/13 12/117/12 10/9 10/139 (47.4%) 10 (52.6%) 9 (39.1%)28.87 (13.68) 31.14 (15.12) 28.98 (12.08)16.68 (9.91) 15.28 (10.89) 14.74 (10.43)8.28 (8.96) 6.30 (12.48) 6.41 (8.12)3.07 (5.23) 10.71 (16.25) 3.89 (5.28)6.21 (3.65) 4.95 (3.72) 6.48 (3.54)

24.47 (3.89) 24.53 (3.37) 25.09 (4.69)39.42 (10.98) 38.63 (9.40) 37.87 (10.84)

20.95 (5.32)⁎⁎ 9.32 (6.34)⁎⁎ 17.35 (7.57)⁎⁎

30.42 (13.63)⁎⁎ 10.95 (6.86)⁎⁎ 31.61 (12.19)⁎⁎

14.33 (17.57)⁎⁎ 60.48 (27.47)⁎⁎ 29.74 (28.56)⁎⁎

23.40 (27.78)⁎⁎ 71.74 (16.28)⁎⁎ 17.62 (20.02)⁎⁎

AD = Generalised Anxiety Disorder; AD = Antidepressant.

Table 3Means and standard deviations of cognitive data in eventual rTMS treatment responders and treatment non-responders.

Whole grouppatients

Eventual rTMS treatment outcome

HDRS responders HDRS non-responders BDI-II responders BDI-II non-responders

Number of subjects 42 23 19 19 23WTAR Pred IQ 108.80 (10.83) 107.00 (10.13) 111.00 (11.54) 106.72 (11.67) 110.50 (10.05)BVMT-R t 21.57 (6.56) 22.30 (6.00) 20.68(7.25) 22.21 (6.73) 21.04 (6.52)BVMT-R delay 8.76 (2.51) 9.39 (2.21) 8.00 (2.69) 9.11 (2.42) 8.48 (2.59)RAVLT t 54.36 (7.88) 55.43 (7.30) 53.05 (8.55) 54.37 (7.93) 54.35 (8.02)RAVLT delay 10.71 (3.04) 10.73 (2.90) 10.68 (3.28) 10.74 (3.23) 10.68 (2.95)TMT A (seconds) 27.05 (8.72) 26.70 (8.17) 27.47 (9.56) 27.16 (8.71) 26.96 (8.92)TMT B (seconds) 75.48 (37.01) 77.91 (40.26) 72.53 (33.51) 81.00 (44.67) 70.91 (29.53)COWAT 36.88 (11.53) 36.91 (11.17) 36.84 (12.27) 37.74 (12.14) 36.17 (11.24)DSF 10.26 (2.50) 10.09 (2.86) 10.47 (2.04) 9.95 (2.67) 10.52 (2.37)DSB 6.69 (2.03) 6.26 (2.09) 7.21 (1.87) 6.23 (2.05) 7.04 (1.99)Stroop 0.20 (8.02) 0.04 (8.17) 0.36 (8.08) −0.05 (8.74) 0.41 (7.57)

Note: WTAR Pred IQ=Wechsler Test of Adult Reading Predicted Intelligence Quotient; BVMT-R t=Benton Visual Memory Test total learning score; BVMT-R delay = Benton VisualMemory Test delayed recall; RAVLT t = Rey Auditory Verbal Learning Test total learning score; RAVLT delay = Rey Auditory Verbal Learning Test delayed recall; TMT A = TrailMaking Test A; TMT B = Trail Making Test B; COWAT = Controlled Oral Word Association Test; DSF = Digit Span (forwards); DSB = Digit Span (backwards); Stroop =Stroopinterference score.

16 C.P. Furtado et al. / Psychiatry Research: Neuroimaging 202 (2012) 12–19

rTMS in a clinical sample of severe TRD patients in order to identifypotential predictors of treatment outcome. Although no statisticallysignificant differences in baseline neurocognitive profiles and MTLstructural volumes were found between treatment responders andnon-responders, several findings of this study are worthy of discus-sion. First, approximately 50% of patients achieved clinical responseto any single course or rTMS and 30% of all responders attained clin-ical remission (HDRSb7). Second, a relationship between pre-treatment left hippocampus volume and antidepressant response torTMS treatment was found when controlled for gender, but this wasof insufficient strength to have statistical value as a predictor offavourable outcome.

To the best of our knowledge, this is the first study that has pro-spectively explored volumetric and cognitive differences betweentreatment responders and non-responders to rTMS. In line with ourfindings, non-significant results were also reported from a largersample of 61 drug-naive patients suffering from depression, whichdid not find any objective neurological markers that could predictclinical outcome to a course of single antidepressant medication(Gong et al., 2011). Although we did not find significant differences,the results provide an initial first step in the utility of biological

Table 4Correlations for HDRS and BDI-II percentage improvement and MTL volumes bygender.

HDRS percentagechange (baselineto end rTMStreatment)

BDI-II percentagechange (baseline toend rTMS treatment)

Males (n=15) R Sig. R Sig.

HDRS% change 0.472 0.048⁎

Left hippocampus 0.055 0.846 −0.474 0.074Right hippocampus −0.084 0.767 −0.491 0.063Left amygdala −0.374 0.169 −0.202 0.471Right amygdala −0.128 0.648 −0.254 0.361Left entorhinal cortex −0.066 0.814 −0.441 0.100Right entorhinal cortex 0.198 0.479 0.091 0.748

Females (n=23) R Sig. R Sig.HDRS% change 0.723 0.001⁎⁎

Left hippocampus −0.131 0.552 −0.225 0.301Right hippocampus 0.060 0.785 −0.143 0.516Left amygdala 0.353 0.098 0.270 0.212Right amygdala 0.224 0.304 −0.027 0.901Left entorhinal cortex 0.151 0.492 0.287 0.184Right entorhinal cortex 0.030 0.891 0.166 0.449

⁎ Significant at pb0.05.⁎⁎ Significant at pb0.001.

markers in the prediction of clinical outcome. In particular, a relation-ship between smaller left hippocampal volumes at baseline whencontrolled for the effects of gender was found to be associated withpositive subjective treatment outcome (BDI-II percentage improve-ment from baseline to end treatment). This finding is in keepingwith previous studies on the directionality and asymmetry of hippo-campus volume in MDD (Mervaala et al., 2000; Frodl et al., 2004;Koolschijn, 2009; Gerritsen et al., 2011). Interestingly, several studieshave implicated left hippocampal abnormalities in MDD specificallywith male gender (Frodl et al., 2002; Maller et al., 2007; Kronmülleret al., 2008; MacMaster et al., 2008; Eker and Gonul, 2010) consistentwith our results. Furthermore, hippocampal volume of healthy men isknown to be greater than that of healthy women but decreases overthe lifespan only in men (Kronmüller et al., 2008). Thus, it may bespeculated that hippocampal volume reduction is a gender specificfactor in the pathophysiology of depressive illness and may hold po-tential predictive value. Given our limited sample size, however, itis unlikely that small hippocampal volume in males predicts re-sponse. A strong impetus is therefore provided for future large scalestudies to investigate gender and age-related differences in hippo-campal volume and rTMS treatment response outcomes in TRD.

Fig. 2. Scatter plot of percentage improvement in BDI-II scores and pre-treatment lefthippocampus volume split by eventual BDI-II responder (r=−0.151, p=.58) andBDI-II non-responder (r=−0.416, p=.05).

Table 5Correlations between gender, left hippocampus volume and BDI-II change at end rTMStreatment.

BDI-II percentage change(baseline to end rTMS treatment)

R Sig.

Gender 0.263 0.056Left hippocampus volume −0.283 0.042⁎

⁎ Significant at pb0.05.

17C.P. Furtado et al. / Psychiatry Research: Neuroimaging 202 (2012) 12–19

An extensive preclinical literature has implicated the hippocampusas a key target of antidepressant medication, supporting the hypothesisthat the integrity of the hippocampus may have predictive value as amarker of pharmacotherapy treatment responsiveness (MacQueen,2009). Several prospective studies have shown smaller left and righthippocampal volumes at baseline were associated with worse clinicaloutcomes at one (Frodl et al., 2004) and 3 years follow-up (Frodl etal., 2008). Although our findings are contradictory of these studies,with smaller pre-treatment left hippocampus volumes being associatedwith higher subjective improvement, it is important to note that oursample included patients with chronic illness and lengthy durations ofunremitted current episodes (mean length of 6.36 years), in line withpast research of smaller hippocampus volumes being associated withprotracted illness (Sheline, 2003; McKinnon et al., 2009). The “neuro-toxic hypothesis” of depression postulates that recurrence and lengthof unremitted depressive episodes may enhance the deleterious effectsof depression on the structural and functional plasticity of cortico-limbic circuits including the hippocampus and amygdala (Fossati etal., 2004; Eker and Gonul, 2010). Thus, smaller baseline volumes maybe reflective of our patient's severe unremitting, treatment-resistantcourse of depressive illness.

The discrepancy in objective and subjective rating of rTMS treatmentresponse in detecting differences in pre-treatment volumes is worthy ofcomment. Although the most widely applied criteria of ‘responders’ and‘non-responders’ are based on objective clinical rating scales such as theHDRS, the BDI-II is often used as a comparativemeasure andprovides pa-tient opinion of degree of symptomatic burden. In clinical TRD samplesresponse to treatment is the primary outcome and subjective ratings ofresponse are of comparable importance to objective ratings if not ofgreater importance as perceived improvement has flow-on effects on so-cial, occupational and interpersonal functioning (Souery et al., 1999). Inour sample there was a strong positive correlation between HDRS per-centage improvement scores and BDI-II percentage improvement scores,supporting the fact that these two rating scales essentially measured thesame outcomes but to varying degrees. This was evidenced by our nearsignificant regression model of smaller pre-treatment left hippocampusvolume predicting eventual subjective improvement in depressivesymptomatology as measured by the BDI-II but not on objective im-provement as determined by the HDRS. It may well be that rTMS exertsa more favourable effect in individuals with smaller left hippocampus

Table 6Unstandardised and standardised β coefficients, confidence intervals and p values forpredictors of eventual degree of percentage improvement in BDI-II scores.

Unstandardisedcoefficients

Standardisedcoefficients

95% CI pvalue

β coefficient Std. error β coefficient

Gender 19.479 10.427 0.290 −1.69 to 40.65 0.07Lefthippocampus

−0.047 0.023 −0.309 −0.09 to 0.001 0.05⁎

R2=0.164a

Adjusted R2=0.116R=0.405

a Unique variability=0.011, shared variability=0.153.⁎ Significant at pb0.05.

volumes resulting in perceived improvements in depressive symptoms,however, as there was no MRI follow-up post treatment it is unknownwhat effect rTMS had on all MTL volumes.

Our lack of significant findings is likely to be related to our samplesize, which subsequently may not have provided sufficient power todetect subtle differences between treatment responders and non-responders. Even with considerably larger sample sizes, however,no strong predictors of response to rTMS have yet been elucidated.Several potential predictors have emerged in literature and are pre-dominately related to the degree of treatment resistance and demo-graphic factors. Increasing age and treatment refractoriness havebeen identified as significant negative predictors of depression im-provement in 150 MDD patients (Fregni et al., 2006), whilst lesstreatment resistance, shorter duration of the current episode(O'Reardon et al., 2007) and higher levels of sleep disturbance(Brakemeier et al., 2007) have been identified as positive predictorsof treatment response to rTMS in large MDD samples. The emergingpredictors, however, have not been consistently replicated with nullfindings of degree of treatment resistance (Fitzgerald et al., 2006;Brakemeier et al., 2008) and age (O'Reardon et al., 2007; Lisanby etal., 2009) being reported, consistent with our findings.

A second limitation is that the study did not have a control group forcomparison. Our main focus, however, was on prediction of treatmentoutcome, not establishing structural abnormalities and cognitive prob-lems in acute TRDpatients compared to healthy subjects, and so a controlgroup was not included as they would not be receiving any treatment.Additionally, although patients received different types of rTMS stimula-tion (unilateral or sequential bilateral), our aim was not to investigatethe efficacy of the two approaches. Recent treatment studies haveshown that applying sequential bilateral rTMS is just as effective as uni-lateral HFL treatment approaches (Fitzgerald et al., 2006; Fitzgerald andDaskalakis, 2011), thereby emphasising the comparability of the two ap-proaches on prediction of outcome. In regard to laterality differences inMTL volumes, this finding may be the result of inadequate study poweras well as true gender differences. For example, several studies havereported significant differences between patients and controls for righthippocampal volume only (Vakili et al., 2000; Frodl et al., 2004; Langeand Irle, 2004), while others have found significant differences for lefthippocampal volume only (Mervaala et al., 2000; Frodl et al., 2002). In-terestingly, Frodl et al. (2002) reported smaller left hippocampus vol-umes in male patients compared to controls, while their femalepatients had a tendency towards larger hippocampal volumes thantheir female controls, supporting the notion that laterality differencesare dependent to some extent on gender variability in brainmorphology.

A strength of our study is that it has ecological validity, in that pa-tients recruited were actively seeking health care or were hospita-lised for acute depressive illness. TRD is associated with a particulartreatment response (non-response) and severity (occurred for ex-tended periods of time), that allows for the investigation from amore homogenous population and subsequently more consistent re-sults, free from the influence of treatment factors such as type anddosage of current medications and time on current medications(Kozel and George, 2005). The fact that there was a 50% responserate to rTMS with just over 25% attaining clinical remission is encour-aging and in direct contrast to prior studies of higher treatment resis-tance being a negative predictor of rTMS outcome. Whether subtletrends towards structural abnormalities between responders andnon-responders to rTMS arose as a consequence of protracted illnessor degree of treatment refractoriness, or instead predate illness onsetas an underlying susceptibility to a more severe MDD illness course, isunknown. Furthermore, the relationship between these volumetricabnormalities and eventual treatment response to rTMS needs to beexplored further. Thus, future large scale, sham controlled, prospec-tive studies are needed and should include healthy matched controlsat both pre-treatment and follow-up to investigate the potential ofMTL volumetrics in predicting response to rTMS treatment.

18 C.P. Furtado et al. / Psychiatry Research: Neuroimaging 202 (2012) 12–19

In conclusion, this study did not find structural or cognitive predic-tors of response to rTMS. Although the data are preliminary, gender spe-cific structural abnormalities and eventual outcome may hold thepotential for predicting outcome to rTMS and therefore replication ofthis study in a larger sample is highly warranted. Recent findings fromanimal models have shown that rTMS has similar therapeutic potentialas antidepressant pharmacotherapy in reversing the neurotoxic effectsof stress and promoting neurogenesis (Gersner et al., 2011). Prospectivelongitudinal studies exploring the same patients while depressed andafter remission to examine the neural and cognitive correlates of re-sponse and non-response to rTMS are also needed to provide furtherclarification as to whether changes are illness traits or state-related.There continues to be a great need to identify clinical tools such as cog-nitive measures to predict response to rTMS with sufficient sensitivityand specificity to be of therapeutic use especially in guiding patient se-lection and reducing the overwhelming burden of treatment resistanceon patients and their families.

References

Albus, M., Hubmann,W., Wahlheim, C., Sobizack, N., Franz, U., Mohr, F., 1996. Contrasts inneuropsychological test profile between patients with first-episode schizophreniaand first-episode affective disorders. Acta Psychiatrica Scandinavica 94, 87–93.

American Psychiatric Association, 2000. Diagnostic and Statistical Manual ofMental Disor-ders, Fourth Edition, Text Revision. American Psychiatric Association,Washington, DC.

Barker, A.T., 1991. An introduction to the basic principles of magnetic nerve stimula-tion. Journal of Clinical Neurophysiology 8, 26–37.

Beck, A.T., Steer, R.A., Brown, G.K., 1996. Manual for Beck Depression Inventory-II. Psy-chological Corporation, San Antonio, TX.

Bell-McGinty, S., Butters, M.A., Meltzer, C.C., Greer, P.J., Reynolds, C.F., Becker, J.T., 2002.Brainmorphometric abnormalities in geriatric depression: long-term neurobiologicaleffects of illness duration. The American Journal of Psychiatry 159, 1424–1427.

Benedict, R.H.B., 1997. Brief Visuospatial Memory Test-Revised: Professional Manual.Psychological Assessment Resources, Inc., Odessa, FL.

Benton, A.L., Hamsher, K.d., Sivan, A.B., 1994. Multilingual Aphasia Examination, 3rd ed.The Psychological Corporation, San Antonio, TX.

Bergouignan, L., Chupin, M., Czechowska, Y., Kinkingnéhun, S., Lemogne, C., Le Bastard,G., Lepage, M., Garnero, L., Colliot, O., Fossati, P., 2009. Can voxel based morphom-etry, manual segmentation and automated segmentation equally detect hippo-campal volume differences in acute depression? NeuroImage 45, 29–37.

Brakemeier, E.-L., Luborzewski, A., Danker-Hopfe, H., Kathman, N., Bajbouj, M., 2007.Positive predictors for antidepressive response to prefrontal repetitive transcranialmagnetic stimulation (rTMS). Journal of Psychiatric Research 41, 395–403.

Brakemeier, E.L., Wilbertz, G., Rodax, S., Danker-Hopfe, H., Zinka, B., Zwanzger, P.,Grossheinrich, N., Várkuti, B., Rupprecht, R., Bajbouj, M., Padberg, F., 2008. Patternsof response to repetitive transcranial magnetic stimulation (rTMS) in major de-pression: replication study in drug-free patients. Journal of Affective Disorders108, 59–70.

Burt, T., Lisanby, S.H., Sackeim, H.A., 2002. Neuropsychiatric applications of transcranialmagnetic stimulation: a meta analysis. The International Journal of Neuropsycho-pharmacology 5, 73–103.

Caetano, S.C., Hatch, J.P., Brambilla, P., Sassi, R.B., Nicoletti, M., Mallinger, A.G., Frank, E.,Kupfer, D.J., Keshavan, M.S., Soares, J.C., 2004. Anatomical MRI study of hippocampusand amygdala in patients with current and remittedmajor depression. Psychiatry Re-search: Neuroimaging 132, 141–147.

Campbell, S.,Marriott,M., Nahmias, C.,MacQueen, G.M., 2004. Lower hippocampal volume inpatients suffering from depression: a meta-analysis. The American Journal of Psychiatry161, 598–607.

Clarke, S., 2003. Complexity of human interhemispheric connections. In: Zaidel, E.,Iacoboni, M. (Eds.), The Parallel Brain: The Cognitive Neuroscience of the CorpusCallosum. Massachusetts Institute of Technology Press, Cambridge, MA.

Couturier, J.L., 2005. Efficacy of rapid-rate repetitive transcranial magnetic stimulationin the treatment of depression: a systematic review and meta-analysis. Journal ofPsychiatry & Neuroscience 30, 83–90.

Czéh, B., Lucassen, P.J., 2007. What causes the hippocampal volume decrease in depres-sion? Are neurogenesis, glial changes and apoptosis implicated? European Archivesof Psychiatry and Clinical Neuroscience 257, 250–260.

Daskalakis, Z.J., Levinson, A.J., Fitzgerald, P.B., 2008. Repetitive transcranial magneticstimulation for major depressive disorder: a review. Canadian Journal of Psychiatry53, 555–566.

Drevets, W.C., 2003. Neuroimaging abnormalities in the amygdala in mood disorders,Vol. 985, pp. 420–444.

Duman, R.S., Monteggia, L.M., 2006. A neurotrophic model for stress-related mood dis-orders. Biological Psychiatry 59, 1116–1127.

Eker, C., Gonul, A.S., 2010. VolumetricMRI studies of the hippocampus inmajor depressivedisorder: meanings of inconsistency and directions for future research. The WorldJournal of Biological Psychiatry 11, 19–35.

Fava, M., 2003. Diagnosis and definition of treatment-resistant depression. BiologicalPsychiatry 53, 649–659.

Fitzgerald, P.B., Daskalakis, Z.J., 2011. The effects of repetitive transcranial magneticstimulation in the treatment of depression. Expert Review of Medical Devices 8,85–95.

Fitzgerald, P.B., Benitez, J., de Castella, A., Daskalakis, Z.J., Brown, T.L., Kulkarni, J., 2006.A randomised, controlled trial of sequential bilateral repetitive transcranial mag-netic stimulation for treatment resistant depression. The American Journal of Psy-chiatry 163, 88–94.

Fitzgerald, P.B., Hoy, K.E., Gunewardene, R., Slack, C., Ibrahim, S., Bailey, M., Daskalakis, Z.J.,2010. A randomized trial of unilateral and bilateral prefrontal cortex transcranial mag-netic stimulation in treatment-resistant major depression. Psychological Medicine 41,1187–1196.

Fossati, P., Radtchenko, A., Boyer, P., 2004. Neuroplasticity: from MRI to depressivesymptoms. European Neuropsychopharmacology 14, S503–S510.

Fregni, F., Marcolin, M.A., Myczkowski, M., Amiaz, R., Hasey, G., Rumi, D.O., Rosa, M.,Rigonatti, S.P., Camprodon, J., Walpoth, M., Heaslip, J., Grunhaus, L., Hausmann,A., Pascual-Leone, A., 2006. Predictors of antidepressant response in clinical trialsof transcranial magnetic stimulation. The International Journal of Neuropsycho-pharmacology 9, 641–654.

Frodl, T., Meisenzahl, E.M., Zetzsche, T., Born, C., Groll, C., Jäger, M., Leinsinger, G.,Bottlender, R., Hahn, K., Möller, H.J., 2002. Hippocampal changes in patients witha first episode of major depression. The American Journal of Psychiatry 159,1112–1118.

Frodl, T., Meisenzhal, E.M., Zetzsche, T., Hohne, T., Banac, S., Schorr, C., Jager, M.,Leinsinger, G., Bottlender, R., Reiser, M., Moller, H.-J., 2004. Hippocampal andamygdala changes in patients with major depressive disorder and healthy controlsduring a 1-year follow-up. The Journal of Clinical Psychiatry 65, 492–499.

Frodl, T., Jäger, M., Smajistrlova, I., Born, C., Bottlender, R., Palladino, T., Reiser, M., Möller,H.J., Meisenzahl, E.M., 2008. Effect of hippocampal and amygdala volumes on clinicaloutcomes in major depression: a 3-year prospective magnetic resonance imagingstudy. Journal of Psychiatry & Neuroscience 33, 423–430.

Furtado, C.P., Maller, J.J., Fitzgerald, P.B., 2008. A magnetic resonance imaging study ofthe entorhinal cortex in treatment- resistant depression. Psychiatry Research:Neuroimaging 163, 133–142.

Gerritsen, L., Comijs, H.C., Van Der Graaf, Y., Knoops, A.J.G., Penninx, B.W.J.H., Geerlings,M.I., 2011. Depression, hypothalamic pituitary adrenal axis, and hippocampal andentorhinal cortex volumes—the SMART Medea study. Biological Psychiatry 70,373–380.

Gersner, R., Kravetz, E., Feil, J., Pell, G., Zangen, A., 2011. Long-term effects of repetitivetranscranial magnetic stimulation on markers for neuroplasticity: differentialoutcomes in anesthetized and awake animals. Journal of Neuroscience 31,7521–7526.

Golden, C.J., 1978. Stroop Color and Word Test: A Manual for Clinical and ExperimentalUses. Skoelting, Chicago, Illinois.

Gong, Q., Wu, Q., Scarpazza, C., Lui, S., Jia, Z., Marquand, A., Huang, X., McGuire, P.,Mechelli, A., 2011. Prognostic prediction of therapeutic response in depressionusing high-field MR imaging. NeuroImage 55, 1497–1503.

Grunhaus, L., Dannon, P.N., Schreiber, S., Dolberg, O.H., Amiaz, R., Ziv, R., Lefkifker, E., 2000.Repetitive transcranialmagnetic stimulation is as effective as electroconvulsive therapyin the treatment of nondelusional major depressive disorder: an open study. BiologicalPsychiatry 47, 314–324.

Hamilton, M., 1960. A rating scale for depression. Journal of Neurology, Neurosurgery,and Psychiatry 23, 56–62.

Hamilton, J.P., Siemer,M., Gotlib, I.H., 2008. Amygdala volume inmajor depressivedisorder:a meta-analysis of magnetic resonance imaging studies. Molecular Psychiatry 13,993–1000.

Hastings, R.S., Parsey, R.V., Oquendo,M.A., Arango, V., Mann, J.J., 2004. Volumetric analysisof the prefrontal cortex, amygdala, and hippocampus inmajor depression. Neuropsy-chopharmacology 29, 952–959.

Hoy, K.E., Fitzgerald, P.B., 2010. Brain stimulation in psychiatry and its effects on cognition.Nature Reviews. Neurology 6, 267–275.

Januel, D., Dumortier, G., Verdon, C.M., Stamatiadis, L., Saba, G., Cabaret, W., Benadhira,R., Rocamora, J.F., Braha, S., Kalalou, K., Vicaut, P.E., Fermanian, J., 2006. A double-blind sham controlled study of right prefrontal repetitive transcranial magneticstimulation (rTMS): therapeutic and cognitive effect in medication free unipolardepression during 4 weeks. Progress in Neuro-Psychopharmacology & BiologicalPsychiatry 30, 126–130.

Koolschijn, P.C.M.P., 2009. Brain volume abnormalities in major depressive disorder: ameta-analysis of magnetic resonance imaging studies. Human Brain Mapping 30,3719–3735.

Kozel, F., George, M.S., 2002. Meta-analysis of left prefrontal repetitive transcranialmagnetic stimulation (rTMS) to treat depression. Journal of Psychiatric Practice8, 270–275.

Kozel, F.A., George, M.S., 2005. Neuroimaging and depression with inadequate treatmentresponse. Primary Psychiatry 12, 30–35.

Kronmüller, K.T., Pantel, J., Köhler, S., Victor, D., Giesel, F., Magnotta, V.A., Mundt, C.,Essig, M., Schröder, J., 2008. Hippocampal volume and 2-year outcome in depres-sion. The British Journal of Psychiatry 192, 472–473.

Lam, R.W., Chan, P., Wilkins-Ho, M., Yatham, L.N., 2008. Repetitive transcranial magneticstimulation for treatment-resistant depression: a systematic review and meta-analysis. Canadian Journal of Psychiatry 53, 621–631.

Lange, C., Irle, E., 2004. Enlarged amygdala volume and reduced hippocampal volumein young women with major depression. Psychological Medicine 34, 1059–1064.

Langguth, B., Wiegand, R., Kharraz, A., Landgrebe, M., Marienhagen, J., Frick, U., Hajak,G., Eichhammer, P., 2007. Pre-treatment anterior cingulate activity as a predictorof antidepressant response to repetitive transcranial magnetic stimulation(rTMS). Neuroendocrinology Letters 28, 633–638.

19C.P. Furtado et al. / Psychiatry Research: Neuroimaging 202 (2012) 12–19

Li, C.T., Lin, C.P., Chou, K.H., Chen, I.Y., Hsieh, J.C., Wu, C.L., Lin, W.C., Su, T.P., 2010. Struc-tural and cognitive deficits in remitting and non-remitting recurrent depression: avoxel-based morphometric study. NeuroImage 50, 347–356.

Lisanby, S.H., Husain, M.M., Rosenquist, P.B., Maixner, D., Gutierrez, R., Krystal, A., Gilmer,W.,Marangell, L.B., Aaronson, S., Daskalakis, Z.J., Canterbury, R., Richelson, E., Sackeim,H.A., George,M.S., 2009. Daily left prefrontal repetitive transcranialmagnetic stimula-tion in the acute treatment of major depression: clinical predictors of outcome in amultisite, randomized controlled clinical trial. Neuropsychopharmacology 34,522–534.

Little, J.T., Kimbrell, T.A., Wassermann, E.M., Grafman, J., Figueras, S., Dunn, R.T., Danielson,A., Repella, J., Huggins, T., George, M.S., Post, R.M., 2000. Cognitive effects of 1- and 20-hertz repetitive transcranial magnetic stimulation in depression: preliminary report.Neuropsychiatry, Neuropsychology, and Behavioral Neurology 13, 119–124.

Loo, C.K., Mitchell, P.B., 2005. A review of the efficacy of transcranial magnetic stimulation(TMS) treatment for depression, and current and future strategies to optimize efficacy.Journal of Affective Disorders 88, 255–267.

Loo, C.K., McFarquhar, T.F., Mitchell, P.B., 2008. A review of the safety of repetitive tran-scranial magnetic stimulation as a clinical treatment for depression. The Interna-tional Journal of Neuropsychopharmacology 11, 131–147.

Lorenzetti, V., Allen, N.B., Fornito, A., Yücel, M., 2009. Structural brain abnormalities inmajor depressive disorder: a selective reviewof recentMRI studies. Journal of AffectiveDisorders 117, 1–17.

Lorenzetti, V., Allen, N.B., Whittle, S., Yücel, M., 2010. Amygdala volumes in a sample ofcurrent depressed and remitted depressed patients and healthy controls. Journal ofAffective Disorders 120, 112–119.

Luborzewski, A., Schubert, F., Seifert, F., Danker-Hopfe, H., Brakemeier, E.L., Schlattmann, P.,Anghelescu, I., Colla, M., Bajbouj, M., 2007. Metabolic alterations in the dorsolateralprefrontal cortex after treatmentwith high-frequency repetitive transcranialmagneticstimulation in patients with unipolarmajor depression. Journal of Psychiatric Research41, 606–615.

MacMaster, F.P., Kusumakar, V., 2004. Hippocampal volume in early onset depression.BioMed Central Medicine 2 (2), 6.

MacMaster, F.P., Mirza, Y., Szeszko, P.R., Kmiecik, L.E., Easter, P.C., Taormina, S.P., Lynch,M.,Rose, M., Moore, G.J., Rosenberg, D.R., 2008. Amygdala and hippocampal volumes infamilial early onset major depressive disorder. Biological Psychiatry 63, 385–390.

MacQueen, G., 2009. Magnetic resonance imaging and prediction of outcome in pa-tients with major depressive disorder. Journal of Psychiatry & Neuroscience 34,343–349.

MacQueen, G.M., Yucel, K., Taylor, V.H., Macdonald, K., Joffe, R., 2008. Posterior hippocampalvolumes are associatedwith remission rates in patientswithmajor depressive disorder.Biological Psychiatry 64, 880–883.

Maller, J.J., Daskalakis, Z.J., Fitzgerald, P.B., 2007. Hippocampal volumetrics in depres-sion: the importance of the posterior tail. Hippocampus 17, 1023–1027.

Martin, J.L.R., Barbanoj,M.J., Schlaepfer, T.E., Thompson, E., Perez, V., Kulisevsky, J., 2003. Re-petitive transcranial magnetic stimulation for the treatment of depression: systematicreview and meta-analysis. The British Journal of Psychiatry 182, 480–491.

McClintock, S.M., Husain, M.M., Greer, T.L., Cullum, C.M., 2010. Association between de-pression severity and neurocognitive function inmajor depressive disorder: a reviewand synthesis. Neuropsychology 24, 9–34.

McKinnon, M.C., Yucel, K., Nazarov, A., MacQueen, G.M., 2009. A meta-analysis examiningclinical predictors of hippocampal volume in patientswithmajor depressive disorder.Journal of Psychiatry & Neuroscience 34, 41–54.

Mervaala, E., Fohr, J., Kononen, M., Valkonen-Korhonen, M., Vainio, P., Partanen, K.,Partanen, J., Tiihonen, J., Viinamaki, H., Karjalainen, A.-K., Lehtonen, J., 2000. Quan-titative MRI of the hippocampus and amygdala in severe depression. PsychologicalMedicine 30, 117–125.

Nadeau, S.E., McCoy, K.J.M., Crucian, G.P., Greer, R.A., Rossi, F., Bowers, D., Goodman,W.K., Heilman, K.M., Triggs, W.J., 2002. Cerebral blood flow changes in depressedpatients after treatment with repetitive transcranial magnetic stimulation: evi-dence of individual variability. Neuropsychiatry, Neuropsychology, and BehavioralNeurology 15, 159–175.

O'Reardon, J.P., Solvason, H.B., Janicak, P.G., Sampson, S., Isenberg, K.E., Nahas, Z.,McDonald, W.M., Avery, D., Fitzgerald, P.B., Loo, C., Demitrack, M.A., George, M.S.,Sackeim, H.A., 2007. Efficacy and safety of transcranial magnetic stimulation in

the acute treatment of major depression: a multisite randomized controlled trial.Biological Psychiatry 62, 1208–1216.

Price, J.L., Drevets, W.C., 2010. Neurocircuitry of mood disorders. Neuropsychopharma-cology 35, 192–216.

Pruessner, J.C., Dedovic, K., Pruessner, M., Lord, C., Buss, C., Collins, L., Dagher, A.,Lupien, S.J., 2010. Stress regulation in the central nervous system: evidence fromstructural and functional neuroimaging studies in human populations - 2008Curt Richter Award Winner. Psychoneuroendocrinology 35, 179–191.

Reitan, R.M., Wolfson, D., 1985. The Halstead-Reitan Neuropsychological Test Battery.Neuropsychology Press, Tucson.

Rossini, D., Lucca, A., Magri, L., Malaguti, A., Smeraldi, E., Colombo, C., Zanardi, R., 2010.A symptom-specific analysis of the effect of high-frequency left or low-frequencyright transcranial magnetic stimulation over the dorsolateral prefrontal cortex inmajor depression. Neuropsychobiology 62, 91–97.

Schmidt, M., 1996. Rey Auditory Verbal Learning Test. Western Psychological Services,Los Angeles.

Schutter, D.J.L.G., 2009. Antidepressant efficacy of high-frequency transcranial magneticstimulation over the left dorsolateral prefrontal cortex in double-blind sham-controlled designs: a meta-analysis. Psychological Medicine 39, 65–75.

Sheehan, D.V., Lecrubier, Y., Sheehan, K.H., Amorim, P., Janavs, J., Weiller, E., Hergueta,T., Baker, R., Dunbar, G.C., 1998. The Mini-International Neuropsychiatric Interview(M.I.N.I.): the development and validation of a structured diagnostic psychiatricinterview for DSM-IV and ICD-10. The Journal of Clinical Psychiatry 59, 22–33.

Sheline, Y., 2003. Neuroimaging studies of mood disorder effects on the brain. BiologicalPsychiatry 54, 338–352.

Sheline, Y.I., 2006. Brain structural changes associated with depression. In: Gilliam, F.G.,Kanner, A.M., Sheline, Y.I. (Eds.), Depression and Brain Dysfunction. Taylor & Francis,London.

Sheline, Y., Sanghavi, M., Mintun, M.A., Gado, M.H., 1999. Depression duration but notage predicts hippocampal volume loss in medically healthy women with recurrentmajor depression. The Journal of Neuroscience 19, 5034–5043.

Soares, J.C., Mann, J.J., 1997. The anatomy of mood disorders — review of strucuturalneuroimaging studies. Biological Psychiatry 41, 86–106.

Souery, D., Amsterdam, J., Montigny, C., Lecrubier, Y., Montgomery, S., Lipp, O., Racagni,G., Zohar, J., Mendlewicz, J., 1999. Treatment resistant depression: methodologicaloverview and operational criteria. European Neuropsychopharmacology 9, 83–91.

Sunanda, M.B.L., Raju, T.R., 1997. Entorhinal cortex lesioning protects hippocampal CA3neurons from stress-induced damage. Brain Research 770, 302–306.

Thase, M.E., Rush, A.J., 1997. When at first you don't succeed: sequential strategies forantidepressant non-responders. Journal of Clinical Psychiatry 58, 23–29.

Triggs, W.J., McCoy, K.J.M., Greer, R., Rossi, F., Bowers, D., Kortenkamp, S., Nadeau, S.E.,Heilman, K.M., Goodman, W.K., 1999. Effects of left frontal transcranial magneticstimulation on depressed mood, cognition, and corticomotor threshold. BiologicalPsychiatry 45, 1440–1446.

Vakili, K., Pillay, S.S., Lafer, B., Fava, M., Renshaw, P.F., Bonello-Cintron, C.M., Yurgelun-Tood, D.A., 2000. Hippocampal volume in primary unipolar major depression: amagnetic resonance imaging study. Biological Psychiatry 47, 1087–1090.

van Eijndhoven, P., van Wingen, G., van Oijen, K., Rijpkema, M., Goraj, B., Jan Verkes, R.,Oude Voshaar, R., Fernández, G., Buitelaar, J., Tendolkar, I., 2009. Amygdala volumemarks the acute state in the early course of depression. Biological Psychiatry 65,812–818.

Vanderhasselt, M.A., 2009. Acute effects of repetitive transcranial magnetic stimulation onattentional control are related to antidepressant outcomes. Journal of Psychiatry &Neuroscience 34, 119–126.

Wechsler, D., 1997. Wechsler Adult Intelligence Scale — Third Edition. The PsychologicalCorporation, San Antonio, TX.

Wechsler, D., 2001. Wechsler Test of Adult Reading. The Psychological Corporation, SanAntonio, TX.

Yan, C., Gong, G., Wang, J., Wang, D., Liu, D., Zhu, C., Chen, Z.J., Evans, A., Zang, Y., He, Y.,2011. Sex- and brain size-related small-world structural cortical networks inyoung adults: a DTI tractography study. Cerebral Cortex 21, 449–458.