Plasma Brain-Derived Neurotrophic Factor in treatment-resistant depressed patients receiving electroconvulsive therapy

Page 1

www.e l sev i e r. com/ loca te /eu roneu ro

European Neuropsychopharmacology (2009) 19, 349–355

Plasma Brain-Derived Neurotrophic Factor intreatment-resistant depressed patientsreceiving electroconvulsive therapyArmando Piccinni a, Alessandro Del Debbio a, Pierpaolo Medda a,Carolina Bianchi a, Isabella Roncaglia a, Antonello Veltri a, Sara Zanello a,Enrico Massimetti a, Nicola Origlia b, Luciano Domenicib,c,Donatella Marazziti a,⁎, Liliana Dell'Osso a

a Dipartimento di Psichiatria, Neurobiologia, Farmacologia e Biotecnologie, University of Pisa, Italyb Institute of Neuroscience, National Research Council, Pisa, Italyc Dipartimento di Scienze e Tecnologie Biomediche, University of L'Aquila, Italy

Received 7 October 2008; received in revised form 23 December 2008; accepted 8 January 2009

⁎ Corresponding author. Dipartimentgia, Farmacologia e Biotecnologie, Un56100 Pisa, Italy. Tel.: +39 050835412;

E-mail address: [email protected]

0924-977X/$ - see front matter © 200doi:10.1016/j.euroneuro.2009.01.002

KEYWORDSBDNF;ECT;Plasma;Treatment-resistantdepression;HRSD;Remission

Abstract

There is an increasing evidence that the Brain-Derived Neurotrophic Factor (BDNF) could beinvolved in the mode of action of antidepressants and, perhaps, of ECT. This study aimed toinvestigate whether the clinical course of medication-resistant depressed patients following acourse of ECT might be associated with changes of plasma BDNF concentrations. Our findingsshowed that at T0 (baseline) plasma BDNF levels of patients were significantly lower than thoseof control subjects, and that at T2 (after ECT) were significantly increased in parallel with thedecrease of the Hamilton Rating Scale for Depression (HRSD) total score. However, only remitter

patients who showed higher baseline BDNF levels than non-remitters reached normalized BDNFlevels after ECT. These findings would suggest the potential usefulness of baseline plasma BDNFlevels as predictors of response to ECT in treatment-resistant depressed patients.© 2009 Elsevier B.V. and ECNP. All rights reserved.

o di Psichiatria, Neurobiolo-iversity of Pisa, Via Roma 67,fax: +39 05021581d.unipi.it (D. Marazziti).

9 Elsevier B.V. and ECNP. All ri

ght

1. Introduction

Electroconvulsive therapy (ECT) represented for over70 years, and still represents, one of the most effectivetreatments for both major depression (MD) and a few othersevere psychiatric conditions (Zomberg and Pope, 1993;Pagnin et al., 2004; Taylor, 2008) with an effectiveness rategreater than 60%, although it is not without side effects(MacQueen et al., 2007). The morbidity/mortality of ECT

s reserved.

Page 2

350 A. Piccinni et al.

ranges between 2 and 4 per 100,000 sessions and 1 per 10,000patients, similar to that of anaesthetic induction in minorsurgery (Fink, 1979; Philbert et al., 1995). In spite of thisevidence, its mechanism of action is still elusive (Altar et al.,2004; Frazer et al., 2005). Different parameters, such ascortisol, adrenocorticotropic hormone, corticotrophin-re-leasing factor, thyroid-releasing hormone, thyroid-stimulatinghormone, prolactin, oxytocin, vasopressin, dehydroepiandros-terone sulfateand tumor necrosis factorα, havebeenproposedas possible substrates of the effect of ECT, but most of the dataderived from animal and human research are generallyinconsistent (Wahlund and von Rosen, 2003).

Over the past decades, different studies have suggestedthat Brain-Derived Neurotrophic Factor (BDNF), a neurotro-phin recognized to mediate the survival, differentiation andoutgrowth of selected neurons during development andadulthood, as well as to modulate the synaptic functionsand the neuronal plasticity in several brain areas, might beinvolved in the pathophysiology of mood disorders (Dumanet al., 1997; Duman, 2004). BDNF is also present in peripheraltissues, in particular it is mainly stored in human plateletsand circulates in plasma at levels 100-fold lower than thoseof serum (Yamamoto and Gurney, 1990). It has been sug-gested that the difference between serum and plasma BDNFmay correspond to the amount of BDNF stored in circulatingplatelets (Fujimura et al., 2002). Although the regulation ofBDNF in plasma is still poorly understood and there are otherpotential cellular sources of plasma BDNF including vascularendothelial and smooth muscle cells, activated macrophagesand lymphocytes, the amount of plasma BDNF has beenconsidered to partly reflect BDNF secretion in the centralnervous system (Lommatzsch et al., 2005); interestingly, ithas been observed that central and peripheral BDNF changesare positively correlated in rodents (Karege et al., 2002).Recently, decreased plasma and serum BDNF levels havebeen observed in drug-free depressed patients, as comparedwith those detected in healthy subjects (Shimizu et al.,2003; Gonul et al., 2005; Aydemir et al., 2005; Piccinni et al.,2008a,b).

Electroconvulsive seizures (ECS) have been shown toincrease the levels of BDNF mRNA, proteins and the tyrosinekinase receptor B (TrkB) mRNA in the rat hippocampus, whilechronic ECS administration blocked the down-regulation ofBDNF mRNA in the same area in response to restraint-inducedstress (Lindefors et al., 1995; Nibuya et al., 1995; Angelucciet al., 2002). Preclinical observations in animals on the ECS-inducedmossy fiber sprouting of hippocampal neurons (Dumanand Vaidya, 1998; Vaidya et al., 1999; Lamont et al., 2001) andneurogenesis (Parent et al., 1997; Madsen et al., 2000) haveanticipated more recent studies indicating a possible relation-ship between the neurotrophic effect of ECTand the increaseof the brain levels of N-acetylaspartate (Michael et al., 2003;Lang et al., 2007), an index of neuron functionality (Tsai et al.,1995; Sager et al., 2001). A few studies in depressed patientshave shown that ECT may increase the amount of serum andplasma BDNF (Taylor, 2008). The first evidence of serum BDNFlevels increase in treatment-resistant depressed patientsreceiving ECT was recently published (Bocchio-Chiavettoet al., 2006) and subsequently confirmed only in responders(Okamoto et al., 2008). A similar finding was reported also forBDNF plasma levels (Marano et al., 2007). These results suggestthat one of the putativemechanisms of ECTmight bemediated

by BDNF and related substances (Taylor, 2008). Two of theabovementioned studies (Marano et al., 2007; Okamoto et al.,2008) included both unipolar and bipolar depressed patientswith a 2:1 ratio (Maranoet al., 2007). Although the relationshipbetween the polarity of a patient's illness and the ECToutcomeis still controversial, most observations up-to present suggestthat the unipolar/bipolar distinction may have no predictivevalue in determining ECToutcome (Daly et al., 2001).

Many factors including sex, age, diagnosis, presence ofpsychosis, duration of index episode, medication treatmentfailure prior to ECT and medication during ECT course, havebeen proposed as predictors of response to ECT in patientssuffering from depression (Bloch et al., 2005; Kho et al.,2005; Pluijms et al., 2006), however no agreement exists onthe predictive values of such variables (Dombrovski et al.,2005). The aim of the present study was, therefore, to assessplasma BDNF levels in depressed patients who failed torespond to medications and received ECT, as well as toexplore the possible correlation between the biologicalparameter and the clinical changes along the ECT course.

2. Experimental procedures

2.1. Subjects

Eighteen inpatients (9 men and 9 women, mean age±SD 44.9±17 years) who met the DSM-IV-TR (APA, 2000) criteria for currentmajor depressive episode (MDE) with or without psychotic features(16 bipolar disorder and 2 major depressive disorder) were selectedat the Dipartimento di Psichiatria, Neurobiologia, Farmacologia eBiotecnologie, University of Pisa, between September 2006 andDecember 2007. Subjects were eligible if they were 18 years or older,not suffering from amajor neurological ormedical illness that limitedthe use of ECT, and had a chronic (N2 years) or recurrent MDE.Participants were required to have at least three MD treatment fails,including an adequate trial with a tricyclic antidepressant (TCA)(Thase and Rush, 1997; Fava, 2003). More specifically, treatment non-response in patients with bipolar depression was defined as persistingdepressive symptoms despite 2 trials of at least 8 weeks, consisting of1 trial withmood stabilizer(s) plus a (TCA) (200mg/day of imipramineor the equivalent, or the maximum tolerable dose) and one trial witha selective serotonin reuptake inhibitor (SSRI) (40 mg/day offluoxetine or the equivalent) combined with mood stabilizer(s) anda TCA. Diagnosis was confirmed by means of the Mini InternationalNeuropsychiatric Interview (MINI) (Sheehan et al., 1998). For eachpatient, an attending-level ECT psychiatrist recommended ECT ac-cording to clinical judgment based on the patient's failure to me-dication trials and severity or urgency of illness.

Fifteen healthy subjects (3 men and 12 women, mean age±SD:36.9±9.2 years), with no history of past or current chronic physicalor mental disorders and not taking regular medications, wererecruited as the control group.

A written informed consent was obtained from all subjects toparticipate in the study, whichwas approved by the Ethics Committee ofthe University of Pisa, after procedure and effects were fully explained.

2.2. ECT treatment procedures

Before undergoing ECT, each patient was screened for generalmedical conditions through an accurate clinical evaluation includingthe collection of a detailed medical history, a physical and neu-rological examination, blood and urine tests, electrocardiogram,chest film, and a cerebral computed tomography scan. Anesthesiawas induced with intravenous thiopental (2–4 mg/kg i.v.) and

Page 3

Table 1 Characteristics at baseline of patients and controls.

Treatment-resistantdepressed subjects(n=18)

Controlsubjects(n=15)

p value

Mean age±S.D.(years)

44.9±17.0 36.9±9.2

.212 a

Gender, female 9 (50%) 12 (80%) .145 b

DiagnosisUnipolardepression

2 (11%) N/A –

Bipolardepression

16 (89%)

Psychosis presence 8 (44%) N/A –Mean age of onset±

S.D. (years)23.2±11.3 N/A –

Mean duration oflast episode±S.D. (m)

7.4±6.7 N/A –

Mean number ofECT treatments±S.D.

8.3±1.2 N/A –

Mean HRSD21

score±S.D.26.4±6.0 N/A –

Mean CGI-Sscore±S.D.

5.6±.7 N/A –

Mean plasmaBDNF level±S.D. (ng/ml)

2.1±1.2 5.4±2.3 b.001

ECT = electroconvulsive therapy, HRSD21 = 21-item HamiltonRating Scale for Depression, BDNF = Brain-Derived NeurotrophicFactor, NA = not applicablea p Value from Mann Whitney U-test.b p Value from χ2 test.

351Plasma BDNF in treatment-resistant depressed patients receiving electroconvulsive therapy

succinylcholine (0.5–1 mg/kg i.v.). Bilateral brief-pulse, square-wavestimuli were delivered using a Mecta Spectrum 5000 (Mecta Corpora-tion, Lake Oswego, USA) two times a week between 7:00 and 9:00 a.m. Stimulus setting was initially based on age, and the length of theseizures measured by an electroencephalogram (EEG) was kept above25 s. If the duration of the seizure fell below 25 s, the stimulus wasraised (1.5 times) at the next session. For all patients the electrodepositioning and stimulus dosing were similar. Other physiologicalparameters were monitored by means of a pulse oximetry and anelectrocardiogram. Patients were ventilated with 100% oxygen untilresumption of spontaneous respiration. ECT treatment was completedon the basis of the clinical judgment of the treating clinicians. Themean number of treatments received was 8.3 (range 6–10).

Patients were maintained on the same drug treatment for at leastfour weeks before and throughout the entire study period (4paroxetine, 2 citalopram, 2 escitalopram, 2 fluoxetine, 1 fluvoxamine,3 trimipramine, 2 nortriptyline, 1 imipramine, and 1 amitriptyline). Inbipolar subjects mood stabilizers were discontinued for three daysbefore starting ECTand were withheld until the end of ECT treatment.

2.3. BDNF assay and clinical assessments

All venous blood samples were taken in themorning (between 8:00 and9:00 a.m., following an overnight fast) before starting ECT treatment(baseline,T0), aweekafter the third ECTsession (T1), andaweekafterthe last ECTsession (T2). Bloodwas drawn into EDTA-coated tubes thatwere kept on ice, centrifugated at 2000 ×g for 10 min at 4 °C andrefrigerated at −20 °C. The choice of EDTA as anticoagulant for plasmaBDNFassay is consistentwith the results of a recent study, showing thatheparin, but not EDTA, may interfere in some way with BDNF assay(Begliuomini et al., 2007). To measure the amount of total BDNF,acidification and subsequent neutralization of the samples werefollowed before proceeding with the enzyme-linked immunosorbentassay (ELISA) protocol, according to manufacturer's instruction(Promega, Wallisellen, Switzerland). 96-well plates were coated withanti-BDNF monoclonal antibody and incubated at 4 °C for 18 h. Theplates were then incubated in a blocking buffer for 1 h at roomtemperature, then samples were added. The samples and BDNFstandardsweremaintained at room temperature under shaking for 2 h,followed by washing with the appropriate buffer. The plates weresuccessively incubated with anti-human BDNF polyclonal antibody atroom temperature for 2 h, washed, and incubated with anti-IgGantibody conjugated to horseradish peroxidase for 1 h at room tem-perature. The plates were incubated in peroxidase substrate andtetramethylbenzidine solution to produce a colour reaction. The reac-tion was stopped with 1 M HCl. The absorbance at 450 nm was mea-sured with a microplate reader (Model 550, Bio Rad Laboratories) todetermine BDNF values that are expressed as pg/ml.

The severity of depression was assessed by means of the 21-itemHamilton Rating Scale for Depression (HRSD21) (Hamilton, 1960) andthe Clinical Global Impressions — Severity of Illness scale (CGI-S)(Guy, 1976) that were administered at baseline and after each ECTsession by trained raters two-times a week. Remission was con-sidered when HRSD total scores were b10 at the last 2 consecutiveassessments (Lecrubier, 2002).

2.4. Statistical analyses

Since BDNF levels and HRSD scores were not normally distributed, asshown by the Kolmogorov–Smirnov test, non-parametric tests wereused. To compare these variables amongst independent samples, theKruskal–Wallis and Mann–Whitney U-tests were used. Comparisonswithin each group were analyzed by means of the Friedman andDunnet test. A p-value of b.05 was judged as statistically significant.All analyses were carried-out using the SPSS, version 12.1, by meansof personal computers.

3. Results

As shown in Table 1, patients and control subjects were notdifferent in terms of age and sex. The plasma BDNF levels ofpatients were significantly lower than those of control subjectsat all assessment times (T0: Z=−3.941, pb .001; T1: Z=−3.580,pb .001; T2: Z=−2.169, p=.030).

As depicted in Fig. 1, plasma BDNF concentrations showedsignificant variations over time in all patients (mean±SD, ng/ml; T0: 2.1±1.2; T1: 2.6±.8; T2: 3.7±2.0; χ2=8.725, df=2,p=.013), particularly between T0 and T2 (q1=4.03, pb0.05).

The HRSD total scores decreased progressively from T0 to T2(mean values±SD; T0: 26.4±6.0; T1: 14.3±7.0; T2: 10.4±4.6;χ2=30.789, df=2, pb .001): the difference was significant be-tween t0 and t1 (q1=5.05, pb0.05) and between T0 and T2(q1=7.64, pb0.05). Also the CGI-S scores (mean values±SD)showed a progressive and significant reduction during the ECTcourse (T0: 5.6±.7; T1: 4.2±1.3; T2: 3.3±.8; χ2=29.129,pb .001): the difference was significant between T0 and T1(q1=3.69, pb0.05) and between T0 and T2 (q1=6.62, pb0.05).

At T0, no significant correlation was observed betweenbaseline plasma BDNF levels and age, gender, diagnosis, presenceof psychosis, age of onset or duration of the last episode (Table 2);on the contrary, a significant and negative correlation was foundat T2 between percent change in plasma BDNF levels and percentchange in HRSD score (Rho=−0.543, p=0.02).

Page 4

Figure 1 Longitudinal changes in plasma BDNF mean values inremitters and non-remitters treatment-resistant depressedpatients during a course of ECT, as compared with those ofcontrol subjects. Histograms show the means of plasma BDNFlevels at three different sampling times. T0=baseline; T1=afterthe third ECT session; T3=after the last ECT session. PlasmaBDNF levels of all patients (a) and non-remitters (b) differsignificantly from those of control subjects at all assessmenttimes. BDNF concentrations of remitters differ significantly fromcontrol subjects' values at T0 and T1 (c) BDNF levels of remittersdiffer significantly from nonremitters' values at T0 and T2 (d)(⁎pb05; ⁎⁎pb0.1; ⁎⁎⁎pb001).

Table 2 Characteristics at baseline of patients achieving ornot remission following ECT.

Remitters(n=8)

Non-remitters(n=10)

p value

Mean age±S.D. (years) 47.4±16.7 42.9±17.9 .563 a

Gender, female 5 (62%) 4 (40%) .637 b

DiagnosisUnipolar depression 1 (13%) 1 (10%) 1.000 b

Bipolar depression 7 (87%) 9 (90%)Psychosis presence 2 (25%) 6 (60%) .188 b

Mean age of onset±S.D. (years)

21.9±11.1

24.2±11.9 .263 a

Mean duration of lastepisode±S.D. (m)

7.4±7.2 7.4±6.6 .929 a

Mean number of ECTtreatments±S.D.

8.6±.9 8.0±1.3 .263 a

Mean HRSD21 score±S.D. 24.1±5.3 28.2±6.3 .212 a

Mean CGI-S score±S.D. 5.6±.9 5.5±.5 .623 a

Mean plasma BDNFlevel±S.D. (ng/ml)

2.9±1.3 1.5±.5 .023 a

ECT = electroconvulsive therapy, HRDS21 = 21-item HamiltonRating Scale for Depression, BDNF = Brain-Derived NeurotrophicFactor, NA = not applicable.a p Value from Mann Whitney U-test.b p Value from χ2 test.

352 A. Piccinni et al.

At T2, eight (44.4%) patients showed clinical remission, asdocumented by the HRSD total score which was 6.8±3.1, whilethat of non-remitter patients was 13.2±2.4. When comparingremitters and non-remitters, no difference was detected interms of age, gender, presence of psychotic symptoms, age atonset, duration of the last episode, HRSD or CGI-S scores at T0.However, the BDNF levels of remitters resulted to be signifi-cantly higher at T0 and T2 (2.9±1.3 vs. 1.5±.5; Z=−2.267,p=0.02; 5.0±1.8 vs. 2.7±1.4; Z=−2.667, p=0.008, respec-tively) than those of non-remitters. In addition, the BDNF levelsof remitters at T2 were similar to those of healthy subjects,while those of non-remitters remained significantly lower (T0,Z=−3.827, pb .001; t1, Z=−3.273, p=.001 and T2, Z=−2.940,p=.003) (Fig. 1). Within-group analyses showed that BDNFincreased significantly from baseline to T2 only in remitters(χ2=8.194, df=2, p=0.017; q1=2.96, pb0.05). At the end ofthe course of ECT, remitters scored significantly lower than non-remitters at the CGI-S (2.7±.7 vs. 3.7±.6, Z=−2.663, p=.021).

Given that the sample was mostly composed by depressedpatients with a lifetime diagnosis of bipolar disorder (16 of 18patients), we further replicated our analyses in this group.Plasma BDNF levels (T0: 2.1±1.2; T1: 2.6±.9; T2: 3.9±2.1;χ2=9.000, df=2, p=.011), HRSD total scores (T0: 26.0±6.3; T1:14.2±7.4; T2: 10.6±4.7; χ2=27.125, df=2, pb .001) and CGI-Sscores (T0: 5.6±.7; T1: 4.2±1.3; T2: 3.4±1.3; χ2=21.855,pb .001) showed the same significant variations over time.Similarly, remitters (n=7) had significantly higher BDNF levels atT0 and T2 than non-remitters (3.2±1.3 vs. 1.4±.4; Z=−2.488,p=0.01; 5.3±1.8 vs. 2.7±1.4; Z=−2.595, p=0.008, respec-tively); only in remitters BDNF concentrations increased sig-nificantly from baseline to T2 (χ2=6.889, df=2, p=0.032; q1=2.85, pb0.05) and equaled those of controls.

4. Discussion

The results of the present study showed that at baseline theplasma BDNF levels of depressed patients who failed to

respond to medication treatment and underwent an ECTcourse were significantly lower than those of healthy controlsubjects. In addition, after the ECT, the BDNF concentrationsincreased significantly, in parallel with the improvement ofdepressive symptoms, as documented by the decrease of theHRSD total scores. However, this was true only for patientsachieving clinical remission, while in the others, BDNF levels,although increased as compared with baseline values,remained lower than those of healthy subjects. Groupanalyses revealed a difference in plasma BDNF levels atbaseline, with remitters showing higher mean values thannon-remitters. While no difference in plasma BDNF concen-trations between the two groups was measured at T1 (afterthe third ECT session), a significant change occurred at T2 inremitters who showed higher BDNF levels than non-remitters.

The results of the present study are in agreement withthose of previous reports (Marano et al., 2007), showing anincrease of plasma BDNF in 12 out of 13 unipolar or bipolardepressed subjects who responded to ECT, or enhancedserum BDNF levels in refractory depressed patients whoresponded to ECT, with no difference before ECT betweenresponders and non-responders (Okamoto et al., 2008). Onthe contrary, our data seem to suggest that higher plasmaBDNF concentration at baseline may be predictive ofremission following ECT. Previously, different factors, suchas age, duration of index episode, medication resistance,chronicity or the presence of psychotic symptoms, have beenproposed as predictors of remission following ECT (Dom-brovski et al., 2005; Kho et al., 2005; O'Connor et al., 2001;Petrides et al., 2001); however, we could not confirm thesefindings. In this study, the percentage of patients achievingremission following an ECT course was lower than those re-ported by others in treatment-resistant patients with major

Page 5

353Plasma BDNF in treatment-resistant depressed patients receiving electroconvulsive therapy

depression adopting the same remission criteria (44.5% vs.53.3% and 65%, respectively) (Husain et al., 2004; Khalidet al., 2008); this might be due to our small sample size andthe preponderance of bipolar patients (16 out of the total of18). Interestingly, another study reported a lower remissionrate (around 42%) for patients with major depression thatdid not differ from that of bipolar depressed patients (Dalyet al., 2001).

It has been supposed that changes of BDNF of depressedpatients treated with ECT may be time-related: a significantincrease in serum BDNF levels has been observed to occuronly one month after the end of ECT (Bocchio-Chiavettoet al., 2006), and plasma BDNF level has been reported toenhance following the fourth ECT session (Marano et al.,2007). We could not find any significant changes in plasmaBDNF after the third ECT treatment either in remitter or innon-remitter patients, but in both the groups only at the endof the ECT course. However, in non-remitters plasma BDNFlevels remained significantly lower than those of controlsubjects. It is possible that lower plasma BDNF levels mightrepresent a peculiar trait of a subgroup of treatment-resistantdepressed patients or of patients less responsive to ECT.Alternatively, a longer duration of ECT may be needed toobtain clinical remission and a normalization of BDNF values.The increase of plasma BDNF levels seemed indeed to occur atthe same time of the clinical improvement, as result of thesignificant and negative correlation between BDNF levels anddepressive symptoms, as assessed by the HRSD total score,from baseline to endpoint. Other longitudinal studies reportedsignificant relationship between increase in serum or plasmaBDNF and response to antidepressants (Gervasoni et al., 2005;Yoshimura et al., 2007). The increase and normalization ofserum or plasma BDNF levels may, therefore, represent abiological, state-dependent marker of remission followingtreatmentswith antidepressants or ECT. It is noteworthy that aresolution of the hypothalamic-pituitary-adrenocortical (HPA)dysregulation has been observed in depressed patients whohad failed trials of antidepressants and achieved remissionwith ECT (Yuuki et al., 2005); since a down-regulation of BDNFmRNA and protein by corticosteroid hormones has beenreported (Schaaf et al., 1998) in the rat and a relationshipbetween cortisol and BDNF plasma levels in humans has beendescribed also in physiological conditions (Begliuomini et al.,2008), it is possible that a normalization of cortisol and BDNFexpressions may play a role in treatment outcome ofdepression. As observed in recent meta-analyses, the growingevidence concerning both the reduction of BDNF levels indepression and the normalization of such levels followingantidepressant treatment, suggests that blood BDNFmeasure-ment may represent a useful biological marker for depressionor a predictor of antidepressant efficacy (Sen et al., 2008).

In this report, the same pattern of longitudinal BDNFvariations in remitters/non-remitters patients has been ob-served when the analyses were narrowed to bipolar de-pressed patients only. Overall, these results seem to suggestthat a dysregulation in BDNF expression may be a commonfeature of both unipolar and bipolar treatment-resistantdepressed patients and perhaps linked to ECT outcome.

This study suffers several limitations. The main bias isrepresented by the small sample size, so that the compar-isons of clinical characteristics and outcome between re-mitters and non-remitters are based on a few cases (8 vs. 10

subjects). Moreover, a common limitation of studies ad-dressing the relationship between BDNF, depression andpharmacological treatments is linked to the extent to whichblood BDNF levels may reflect brain BDNF concentrations.We chose to investigate the longitudinal changes of BDNF inplasma, because poor-platelet plasma BDNF is minimallyaffected by the amount of BDNF stored in platelets and,therefore, may represent a more reliable and sensitivemarker of BDNF variations occurring in the brain and pe-riphery (Lommatzsch et al., 2005). Nevertheless, plasmaBDNF has shown high inter-individual variability. It shouldalso be mentioned that our absolute plasma BDNF valueswere higher than those observed in recent publications(Lommatzsch et al., 2005; Marano et al., 2007; Begliuominiet al., 2008), however, as previously reported (Piccinniet al., 2008a,b), we assayed total BDNF in plasma by theacidification and neutralization procedures, while othersmight have measured the amount of the free mature form.Therefore, according to us, different methodological pro-cedure might contribute to explain the controversial datapresent in literature (Karege et al., 2005; Palomino et al.,2006). With this limitation in mind, the present findings(Bocchio-Chiavetto et al., 2006; Marano et al., 2007; Oka-moto et al., 2008), suggest a possible role of BDNF in themechanism of action of ECT, that, as observed in preclinicalstudies (Balu et al., 2008), would be shared by that of anti-depressants (Mannari et al., 2008).

Another limitation of our study is represented by theconcomitant use of antidepressant medications in depressedpatients, which might have contributed to the plasma BDNFincrease following ECT, as already reported (Piccinni et al.,2008a,b). In the present study, treatment-resistant de-pressed patients were maintained on the same drug for atleast four weeks before ECT and during the entire studyperiod, and, at baseline, they all showed significant lowerplasma BDNF levels than control subjects.

In conclusion, our results provide further evidence thatECT treatment is associated with changes in plasma BDNF intreatment-resistant depressed patients. The correlation ofplasma BDNF with depressive symptoms and the normal-ization of such levels up to the values found in controlsubjects that occurred in remitters would suggest a potentialusefulness of plasma BDNF levels in the clinical practice aspredictors of response of depression to ECT.

Role of the funding source

Funding for this study was provided by the University of Pisa; theUniversity of Pisa had no further role in the study design; in thecollection, analysis and interpretation of data; in the writing of thereport; and in the decision to submit the paper for publication.

Contributors

Drs. Piccinni, Medda, Marazziti and Dell'Osso designed the study andwrote the protocol. Drs. Del Debbio, Bianchi, Roncaglia managed theliterature searches and analyses. Drs. Veltri and Zanello contributedto the data analyses. Dr. Massimetti undertook the statisticalanalysis. Drs. Origlia and Domenici managed the biochemical partof the study. Drs. Del Debbio and Bianchi wrote the first draft andDrs. Piccinni, Medda, Marazziti and Dell'Osso wrote the final version

Page 6

354 A. Piccinni et al.

of the manuscript. All authors contributed to the study and approvedthe final manuscript.

Conflict of interest

All authors declare that they have no conflicts of interest.

Acknowledgements

This study was carried out at the Institute of Psychiatry, University ofPisa. The authors want to thank Drs. Mariotti, Schiavi and Palla forsubjects' referral.

References

Altar, C.A., Laeng, P., Jurata, L.W., Brockman, J.A., Lemire, A.,Bullard, J., Bukhman, Y.V., Young, T.A., Charles, V., Palfreyman,M.G., 2004. Electroconvulsive seizures regulate gene expressionof distinct neurotrophic signaling pathways. J. Neurosci. 24,2667–2677.

American Psychiatric Association, 2000. Diagnostic and StatisticalManual of Mental Disorders, Fourth Edition. American PsychiatricPress, Washington DC. Text Revision.

Angelucci, F., Aloe, L., Jiménez-Vasquez, P., Mathé, A.A., 2002.Electroconvulsive stimuli alter the regional concentrations ofnerve growth factor, brain-derived neurotrophic factor, and glialcell line-derived neurotrophic factor in adult rat brain. J. ECT 18,138–143.

Aydemir, O., Deveci, A., Taneli, F., 2005. The effect of chronicantidepressant treatment on serum brain-derived neurotrophicfactor levels in depressed patients: a preliminary study. Prog.Neuropsychopharmacol. Biol. Psychiatry 29, 261–265.

Balu, D.T., Hoshaw, B.A., Malberg, J.E., Rosenzweig-Lipson, S.,Schechterl, L.E., Lucki, I., 2008. Differential regulation ofcentral BDNF protein levels by antidepressant and non-anti-depressant drug treatments. Brain Res. 1211, 37–43.

Begliuomini, S., Casarosa, E., Pluchino, N., Lenzi, E., Centofanti,M., Freschi, L., Pieri, M., Genazzani, A.D., Luisi, S., Genazzani,A.R., 2007. Influence of endogenous and exogenous sexhormones on plasma brain-derived neurotrophic factor. Hum.Reprod. 22, 995–1002.

Begliuomini, S., Lenzi, E., Ninni, F., Casarosa, E., Merlini, S.,Pluchino, N., Valentino, V., Luisi, S., Luisi, M., Genazzani, A.R.,2008. Plasma brain-derived neurotrophic factor daily variationsin men: correlation with cortisol circadian rhythm. J. Endocrinol.197, 429–435.

Bloch, Y., Ratzoni, G., Sobol, D., Mendlovic, S., Gal, G., Levkovitz,Y., 2005. Gender differences in electroconvulsive therapy: aretrospective chart review. J. Affect. Disord. 84, 99–102.

Bocchio-Chiavetto, L., Zanardini, R., Bortolomasi, M., Abate, M.,Segala, M., Giacopuzzi, M., Riva, M.A., Marchina, E., Pasqualetti,P., Perez, J., Gennarelli, M., 2006. Electroconvulsive Therapy(ECT) increases serum Brain Derived Neurotrophic Factor (BDNF)in drug resistant depressed patients. Eur. Neuropsychopharma-col. 16, 620–624.

Daly, J.J., Prudic, J., Devanand, D.P., Nobler, M.S., Lisanby, S.H.,Peyser, S., Roose, S.P., Sackeim, H.A., 2001. ECT in bipolar andunipolar depression: differences in speed of response. BipolarDisord. 3, 95–104.

Dombrovski, A.Y., Mulsant, B.H., Haskett, R.F., Prudic, J., Begley, A.E., Sackeim, H.A., 2005. Predictors of remission after electro-convulsive therapy in unipolar major depression. J. Clin.Psychiatry 66, 1043–1049.

Duman, R.S., 2004. Role of neurotrophic factors in the etiology andtreatment of mood disorders. Neuromol. Med. 5, 11–25.

Duman, R.S., Vaidya, V.A., 1998. Molecular and cellular actions ofchronic electroconvulsive seizures. J. ECT 14, 181–193.

Duman, R.S., Heninger, G.R., Nestler, E.J., 1997. A molecular andcellular theory of depression. Arch. Gen. Psychiatry 54, 597–605.

Fava, M., 2003. Diagnosis and definition of treatment-resistantdepression. Biol. Psychiatry 53, 649–659.

Fink, M., 1979. Convulsive Therapy: Theory and Practice. RavenPress, New York, NY.

Frazer, C.J., Christensen, H., Griffiths, K.M., 2005. Effectiveness oftreatments for depression in older people. Med. J. Aust. 182,627–632.

Fujimura, H., Altar, C.A., Chen, R., Nakamura, T., Nakahashi, T.,Kambayashi, J., Sun, B., Tandon, N.N., 2002. Brain-derivedneurotrophic factor is stored in human platelets and released byagonist stimulation. Thromb. Haemost. 87, 728–734.

Gervasoni, N., Aubry, J.M., Bondolfi, G., Osiek, C., Schwald, M.,Bertschy, G., Karege, F., 2005. Partial normalization of serumbrain-derived neurotrophic factor in remitted patients after amajor depressive episode. Neuropsychobiology 51, 234–238.

Gonul, A.S., Akdenz, F., Taneli, A., Donat, O., Eker, C., Vahip, S.,2005. Effect of treatment on serum brain-derived neurotrophicfactor in depressed patients. Eur. Arch. Psychiatry Clin. Neurosci.255, 381–386.

Guy, W., 1976. ECDEU Assessment Manual for Psychopharmacology.US Dept Health, Education, and Welfare Publication (ADM) 76–338. National Institute of Mental Health, Rockville, Md.

Hamilton, M., 1960. A rating scale for depression. J. Neurol. Neu-rosurg. Psychiatry 23, 56–62.

Husain, M.M., Rush, A.J., Fink, M., Knapp, R., Petrides, G.,Rummans, T., Biggs, M.M., O'Connor, K., Rasmussen, K., Litle,M., Zhao, W., Bernstein, H.J., Smith, G., Mueller, M., McClintock,S.M., Bailine, S.H., Kellner, C.H., 2004. Speed of response andremission in major depressive disorder with acute electrocon-vulsive therapy (ECT): a Consortium for Research in ECT (CORE)report. J. Clin. Psychiatry. 65, 485–491.

Karege, F., Bondolfi, G., Gervasoni, N., Schwald, M., Aubry, J.M.,Bertschy, G., 2005. Low brain-derived neurotrophic factor(BDNF) levels in serum of depressed patients probably resultsfrom lowered platelet BDNF release unrelated to platelet re-activity. Biol. Psychiatry 57, 1068–1072.

Karege, F., Schwald, M., Cisse, M., 2002. Postnatal developmentalprofile of brain-derived neurotrophic factor in rat brain andplatelets. Neurosci. Lett. 328, 261–264.

Khalid, N., Atkins, M., Tredget, J., Giles, M., Champney-Smith, K.,Kirov, G., 2008. The effectiveness of electroconvulsive therapy intreatment-resistant depression: a naturalistic study. J. ECT 24,141–145.

Kho, K.H., Zwinderman, A.H., Blansjaar, B.A., 2005. Predictors forthe efficacy of electroconvulsive therapy: chart review of anaturalistic study. J. Clin. Psychiatry 66, 894–899.

Lamont, S.R., Pauls, A., Stewart, C.A., 2001. Repeated electro-convulsive stimulation, but not antidepressant drugs, inducesmossy fibre sprouting in the rat hippocampus. Brain Res. 893,53–58.

Lang, U.E., Hellweg, R., Seifert, F., Schubert, F., Gallinat, J., 2007.Correlation between serum brain-derived neurotrophic factorlevel and an in vivo marker of cortical integrity. Biol. Psychiatry62, 530–535.

Lecrubier, Y., 2002. How do you define remission? Acta Psychiatr.Scand. 415, 7–11.

Lindefors, E., Brodin, E., Metsis, M., 1995. Spatiotemporal selectiveeffects on brain-derived neurotrophic factor and trk B messengerRNA in rat hippocampus by electroconvulsive shock. Neu-roscience 65, 661–670.

Lommatzsch, M., Zingler, D., Schuhbaeck, K., Schloetcke, K.,Zingler, C., Schuff-Werner, P., 2005. The impact of age, weightand gender on BDNF levels in human platelets and plasma.Neurobiol. Aging 26, 115–123.

Page 7

355Plasma BDNF in treatment-resistant depressed patients receiving electroconvulsive therapy

MacQueen, G., Parkin, C., Marriott, M., Bégin, H., Hasey, G., 2007.The long-term impact of treatment with electroconvulsivetherapy on discrete memory systems in patients with bipolardisorder. J. Psychiatry Neurosci. 32, 241–249.

Madsen, T.M., Treschow, A., Bengzon, J., Bolwig, T.G., Lindvall, O.,Tingström, A., 2000. Increased neurogenesis in a model ofelectroconvulsive therapy. Biol. Psychiatry 47, 1043–1049.

Mannari, C., Origlia, N., Scatena, A., Del Debbio, A., Catena, M.,Dell'agnello, G., Barraco, A., Giovannini, L., Dell'Osso, L.,Domenici, L., Piccinni, A., 2008. BDNF level in the rat prefrontalcortex increases following chronic but not acute treatment withduloxetine, a dual acting inhibitor of noradrenaline and serotoninre-uptake. Cell. Mol. Neurobiol. 28, 457–468.

Marano, C.M., Phatak, P., Vemulapalli, U.R., Sasan, A., Nalbandyan,M.R., Ramanujam, S., Soekadar, S., Demosthenous, M., Rege-nold, W.T., 2007. Increased plasma concentration of brain-derived neurotrophic factor with electroconvulsive therapy: apilot study in patients with major depression. J. Clin. Psychiatry68, 512–517.

Michael, N., Erfurth, A., Ohrmann, P., Arolt, V., Heindel, W.,Pfleiderer, B., 2003. Neurotrophic effects of electroconvulsivetherapy: a proton magnetic resonance study of the leftamygdalar region in patients with treatment-resistant depres-sion. Neuropsychopharmacology 28, 720–725.

Nibuya, M., Morinobu, S., Duman, R.S., 1995. Regulation of BDNF andTrkB mRNA in rat brain by chronic electroconvulsive seizure andantidepressant drug treatments. J. Neurosci. 15, 7539–7547.

O'Connor, M.K., Knapp, R., Husain, M., Rummans, T.A., Petrides, G.,Smith, G., Mueller, M., Snyder, K., Bernstein, H., Rush, A.J., Fink,M., Kellner, C., 2001. The influence of age on the response ofmajor depression to electroconvulsive therapy: a C.O.R.E.Report. Am. J. Geriatr. Psychiatry 9, 382–390.

Okamoto, T., Yoshimura, R., Ikenouchi-Sugita, A., Hori, H., Umene-Nakano, W., Inoue, Y., Ueda, N., Nakamura, J., 2008. Efficacy ofelectroconvulsive therapy is associated with changing bloodlevels of homovanillic acid and brain-derived neurotrophic factor(BDNF) in refractory depressed patients: a pilot study. Prog.Neuropsychopharmacol. Biol. Psychiatry 32, 1185–1190.

Pagnin, D., de Queiroz, V., Pini, S., Cassano, G.B., 2004. Efficacy ofECT in depression: a meta-analytic review. J. ECT 20, 13–20.

Palomino, A., Vallejo-Illarramendi, A., González-Pinto, A., Aldama,A., González Gómez, C., Mosquera, F., González-García, G.,Matute, C., 2006. Decreased levels of plasma BDNF in first-episode schizophrenia and bipolar disorder patients. Schizophr.Res. 86, 321–322.

Parent, J.M., Yu, T.W., Leibowitz, R.T., Geschwind, D.H., Sloviter, R.S.,Lowenstein, D.H., 1997. Dentate granule cell neurogenesisis increased by seizures and contributes to aberrant networkreorganization in the adult hippocampus. J. Neurosci. 17,3727–3738.

Petrides, G., Fink, M., Husain, M.M., Knapp, R.G., Rush, A.J.,Mueller, M., Rummans, T.A., O'Connor, K.M., Rasmussen Jr., K.G.,Bernstein, H.J., Biggs, M., Bailine, S.H., Kellner, C.H., 2001. ECTremission rates in psychotic versus nonpsychotic depressedpatients: a report from CORE. J. ECT 17, 244–253.

Philbert, R.A., Richards, L., Lynch, C.F., Winokur, G., 1995. Effect ofECT on mortality and clinical outcome in geriatric unipolardepression. J. Clin. Psychiatry 56, 390–394.

Piccinni, A., Marazziti, D., Catena, M., Domenici, L., Del Debbio, A.,Bianchi, C., Mannari, C., Martini, C., Da Pozzo, E., Schiavi, E.,Mariotti, A., Roncaglia, I., Palla, A., Consoli, G., Giovannini, L.,Massimetti, G., Dell'Osso, L., 2008a. Plasma and serum brain-derived neurotrophic factor (BDNF) in depressed patients during1 year of antidepressant treatments. J. Affect. Disord. 105,279–283.

Piccinni, A., Marazziti, D., Del Debbio, A., Bianchi, C., Roncaglia, I.,Mannari, C., Origlia, N., Catena Dell'Osso, M., Massimetti, G.,Domenici, L., Dell'Osso, L., 2008b. Diurnal variations of plasmabrain-derived neurotrophic factor (BDNF) in humans. A genderanalysis. Chronobiol. Int. 25, 819–826.

Pluijms, E.M., Birkenhäger, T.K., Mulder, P.G., van den Broek, W.W.,2006. Influence of episode duration of major depressive disorderon response to electroconvulsive therapy. J. Affect. Disord. 90,233–237.

Sager, T.N., Topp, S., Torup, L., Hanson, L.G., Egestod, B., Moller, A.,2001. Evaluation of CA1 damage using single-voxel 1H-MRS andun-biased stereology: can non-invasive measures of N-acetyl-aspartate following global ischemia be used as a reliable measureof neuronal damage? Brain Res. 892, 166–175.

Schaaf, M.J.M., Durland, R., de Kloet, E.R., Vreugdenhil, E., 1998.Downregulation of BDNF mRNA and protein in rat hippocampusbycorticosterone. Brain Res. 813, 112–120.

Sen, S., Duman, R., Sanacora, G., 2008. Serum brain-derivedneurotrophic factor, depression, and antidepressant medica-tions: meta-analyses and implications. Biol. Psychiatry 64,527–532.

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 develop-ment and validation of a structured diagnostic psychiatricinterview for DSM-IV and ICD-10. J. Clin. Psychiatry 59 (Suppl.20), 22–33.

Shimizu, E., Hashimoto, K., Okamura, N., Koike, K., Komatsu, N.,Kumakiri, C., Nakazato, M., Watanabe, H., Shinoda, N., Okada,S., Iyo, M., 2003. Alterations of serum levels of brain-derivedneurotrophic factor (BDNF) in depressed patients with or withoutantidepressants. Biol. Psychiatry 54, 70–75.

Taylor, S.M., 2008. Electroconvulsive therapy, brain-derived neuro-trophic factor, and possible neurorestorative benefit of theclinical application of electroconvulsive therapy. J. ECT 24,160–165.

Thase, M.E., Rush, A.J., 1997. When at first you don't succeed:sequential strategies for antidepressant nonresponders. J. Clin.Psychiatry 58 (Suppl. 13), 23–29.

Tsai, G., Passani, L.A., Slusher, B.S., Carter, R., Kleinman, J.E.,Coyle, J.T., 1995. Abnormal excitatory neurotransmitter meta-bolism in schizophrenic brains. Arch. Gen. Psychiatry 52,829–836.

Vaidya, V.A., Siuciak, J.A., Du, F., Duman, R.S., 1999. Hippocampalmossy fiber sprouting induced by chronic electroconvulsiveseizures. Neuroscience 89, 157–166.

Wahlund, B., von Rosen, D., 2003. ECTof major depressed patients inrelation to biological and clinical variables: a brief overview.Neuropsychopharmacology 28 (Suppl. 1), 21–26.

Yamamoto, H., Gurney, M.E., 1990. Human platelets contain brain-derived neurotrophic factor. J. Neurosci. 10, 3469–3478.

Yoshimura, R., Mitoma, M., Sugita, A., Hori, H., Okamoto, T.,Umene, W., Ueda, N., Nakamura, J., 2007. Effects of paroxetineor milnacipran on serum brain-derived neurotrophic factor indepressed patients. Prog. Neuropsychopharmacol. Biol. Psychia-try 31, 1034–1037.

Yuuki, N., Ida, I., Oshima, A., Kumano, H., Takahashi, K., Fukuda,M., Oriuchi, N., Endo, K., Matsuda, H., Mikuni, M., 2005. HPA axisnormalization, estimated by DEX/CRH test, but less alteration oncerebral glucose metabolism in depressed patients receiving ECTafter medication treatment failures. Acta Psychiatr. Scand. 112,257–265.

Zomberg, G.L., Pope, H.G., 1993. Treatment of depression in bipolardisorder: new directions for research. J. Clin. Psychopharmacol.13, 397–408.