The change of prefrontal qeeg theta cordance as a predictor of response to bupropion treatment in patients who had failed to respond to previous antidepressant treatments

European Neuropsychopharmacology (2010) 20, 459–466 w w w . e l s e v i e r . c o m / l o c a t e / e u r o n e u r o The change of prefrontal QEEG theta cordance as apredictor of response to bupropion treatment inpatients who had failed to respond to previousantidepressant treatments Martin Bares , Martin Brunovsky , Tomas Novak ,Miloslav Kopecek Pavla Stopkova , Peter Sos Vladimir Krajca Cyril Höschl a Prague Psychiatric Center, Ustavni 91, Prague 8-Bohnice, 181 03, Czech Republicb The Department of Psychiatry and Medical Psychology, 3rd Faculty of Medicine, Charles University, Ruska 87, Prague 10,100 00, Czech Republicc The Department of Neurology, Faculty Hospital Na Bulovce, Prague 8, 180 81, Czech Republicd Faculty of Biomedical Engineering, Czech Technical University in Prague, nam. Sitna 3105, Kladno, 272 01, Czech Republic Received 21 October 2009; received in revised form 10 February 2010; accepted 14 March 2010 The aim of the study was to examine whether the reduction of theta prefrontal quantitative EEG (QEEG) cordance after one week of bupropion administration is a predictor of response to a 4- week treatment in patients that had failed to respond to previous antidepressant treatments.
Method: EEG data of 18 inpatients were monitored at baseline and after one week. QEEGcordance was computed at 3 frontal electrodes (Fp1, Fp2, Fz). Response to treatment wasdefined as a ≥50% reduction of MADRS score. Results: Nine of the eleven responders and one ofthe seven non-responders showed decreased prefrontal cordance value after the first week oftreatment (p = 0.01). Positive and negative predictive values of cordance reduction for theprediction of response to the treatment were 0.9 and 0.75, respectively. Conclusion: Similar toother antidepressants, the reduction of prefrontal QEEG cordance might be helpful in theprediction of the acute outcome of bupropion treatment.
2010 Elsevier B.V. and ECNP. All rights reserved.
⁎ Corresponding author. Prague Psychiatric Center, Ustavni 91, Prague 8-Bohnice, 181 03, Czech Republic. Tel.: +420 266003330;fax: +420 266003337.
Major depressive disorder (MDD) is considered to be a chronic, relapsing and remitting illness. A large percentage 0924-977X/$ - see front matter 2010 Elsevier B.V. and ECNP. All rights reserved.
doi: of patients (30–50%) fail to respond to an initial course of consent to participate in the research was obtained from all subjects. Study was carried out in accordance with the latest Since a large number of patients fail to respond to version of the Declaration of Helsinki.
antidepressants (AD), there is a clear need for methods thatselect the right treatment for the right patient. A consider-able body of research supports the assertion that antide- pressant medication effects are physiologically detectable inthe EEG (for review see QEEG cordance This single-centre, open-label study was performed as a part of a is one of the promising tools for the prediction of response grant project addressing the evaluation of the relationship betweenQEEG cordance and response to the treatment with various which has generated research interest. Cordance is a QEEG method which combines complementary information fromabsolute (the amount of power in a frequency band at a givenelectrode) and relative power (the percentage of power contained in a frequency band relative to the total spectrum)of EEG spectra ). Since cordance Our sample comprised 18 inpatients (10 men, 8 women, mean age values are correlated with regional cerebral blood flow, 46.1 ± 10.1 years) with major depressive disorder (recurrent or single findings with this measure could be interpreted within the episode) diagnosed according to DSM IV criteria ), confirmed using The Mini-International same conceptual framework as other functional neuroima- Neuropsychiatric Interview—M.I.N.I., Czech version 5.0.0 ( ). We included subjects who reached at least the total demonstrating an abnormal pattern of metab- score of 20 in Montgomery–Åsberg Depression Rating Scale (MADRS, olism or perfusion in the prefrontal cortex and the anterior Clinical Global Impression (CGI, ). All patients were hospitalized at the Open Department of Prague Psychiatric Center electrical activity in theta frequency band has been between May 2006 and December 2008. They fulfilled at least Stage I associated with the function of these structures and previous criteria for resistant depression (≥ 1 adequate antidepressant research has linked higher pretreatment theta activity of the anterior cingulate with clinical response to nortriptyline and adequacy of previous medication in the index episode was based onthe Antidepressant Treatment History Form () with a score of at least 3 (more than 4 weeks of treatment at an adequate dose). The most recent medications before enrollment to the study Several studies have demonstrated that a reduction of were SSRI (n = 6), noradrenergic and specific serotonergic AD (NaSSA, prefrontal QEEG theta cordance value after 1 or 2 weeks of n = 2), SNRI (n = 1), various combinations of AD (n = 4) and augmen- treatment with selective serotonin reuptake inhibitors (SSRI) tation of AD with atypical antipsychotics (n = 5) — for more details and selective serotonin–norepinephrine reuptake inhibitors see . We excluded subjects with suicidal risk assessed by (SNRI) can predict clinical response to 8-week treatment in clinical examination, current psychiatric comorbidity on Axes I and non-resistant patients or non-responders to SSRI II, serious unstable medical illness or neurologic disorder (e.g., epilepsy, head trauma with loss of consciousness) and patients using were different from those observed in placebo responders any treatment (including electroconvulsive therapy within 3 monthsbefore start of study) which can strongly affect EEG, as well as patients who were resistant to bupropion in the past.
relationship between an early change in prefrontal cordanceand clinical outcome for resistant patients treated with variousAD (n = 17) and venlafaxine monotherapy (n = 25) in two open- 2.2. Treatment trial and clinical assessments label studies . The positive predictivevalues (PPV) and negative predictive values (NPV) for the All patients were antidepressants and antipsychotics free at least reduction of theta cordance as a predictor of response were one day before initializing of bupropion treatment — for more details 0.7 and 1.0 in the first study, and 0.7 and 0.9 in the second one.
see The continuation of benzodiazepines was allowed in As far as we know, no study examining predictive value of unchanged dosage in patients who used them before the study toavoid withdrawal effect and possible EEG changes. The last dose of prefrontal QEEG cordance changes for AD other than SSRI or benzodiazepines before EEG recording was given at 9 p.m. of SNRI in resistant subjects has been published. Bupropion, an antidepressant which does not act primary via serotonergic The length of bupropion treatment was four weeks. We used mechanism was selected for our study because it is generally sustained-release form of bupropion. The bupropion was used in a well tolerated (including low rate of sexual side effects) and minimal dose of 150 mg/p.d. with possibility of titration of dose switching to bupropion is a popular strategy for the after 5 days of treatment according to clinical status, tolerability and judgement of attending psychiatrist, with average daily doses of 183.3 ± 64.2 mg at week 1 and 287.5 ± 38.6 mg at week 4. Zolpidem about its efficacy in the treatment of resistant depression and hydroxyzine were permitted as a concomitant (emergency) treatment in case of severe insomnia or anxiety.
The primary outcome measure for the study was the score change We hypothesized that the reduction of theta prefrontal in the MADRS. Clinical response was defined as equal to or more than QEEG cordance value after 1 week of bupropion administra- 50% reduction of the MADRS score. The patients were assessed with tion would be associated with response to 4-week treatment MADRS, Beck Depression Inventory—Short Form (BDI-S, in patients resistant to previous antidepressive treatments.
) and CGI before a wash-out period of 1 to 5 days, at baseline The Prague Psychiatric Center Institutional Review Board and after 1 and 4 weeks of treatment. Ratings were made by reviewed and approved the study and written informed experienced clinical psychiatrists (M.B., T.N., M.K., P.S.) who were The change of prefrontal QEEG theta cordance as a predictor of response to bupropion treatment Baseline characteristics of subjects and clinical features of depression.
risperidone-1, mirtazapine +olanzapine-1, dibenzepine +quetiapine-1 IQR—interquartile range, NA—not applicable, NS—nonsignificant, p.d.—per day.
trained to the criterion of intraclass correlation N0.80 for each detection was performed visually to exclude all epochs containing clinician prior to conducting ratings.
eye blink, eye rolling artifact, head movements, muscle artifacts,decrease in alertness or epoch in which any channel had a voltagedeflection greater than ± 75 μV. EEG reviewer was blind to the 2.3. Apparatus and physiological recording outcome of treatment. The number of artifact-free, 2-secondsepochs averaged 18.4 ± 2.5 (range 15–23) across subjects, indicating EEG data were collected at baseline and after 1st week of treat- that on average, at least 30 s of the available recordings were ment. The EEG examination was regularly carried out between 8 a.
submitted to a spectral analysis. The number of epochs we processed m. and 9 a.m. We used a standard 32-channel digital EEG amplifier did not differ between responders and non-responders. Preceding BrainScope (unimedis, Prague) with 21 Ag/AgCl surface electrodes the Fast Fourier Transform (FFT), a linear interpolation between placed according to the international 10/20 system and referenced adjacent raw values (sampled at a frequency of 250 Hz) was carried to the electrode situated between electrodes Fz and Cz in the out, followed by a second sampling procedure at a frequency of midline (FCz). All scalp electrode impedances were below 5 kΩ 256 Hz, yielding 512 values in the 2.0 s spectral window. With this (within 1 kΩ of homologous sites). The EEG recording system algorithm, the frequency resolution of the power spectra is 0.5 Hz.
acquires the data with a 16 bit depth and 7.63 nV/bit resolution FFT was used to calculate absolute and relative power in each of (i.e. ∼130 bit/µV) with the dynamic range of ±250 µV. The data four non-overlapping frequency bands ): delta sampling rate was 250 Hz and the acquired signals were filtered with (0.5–4 Hz), theta (4–8 Hz), alpha (8–12 Hz), and beta (12–20 Hz).
digital high- and low-pass filtering at 0.15 and 70 Hz, respectively.
The EEG was recorded with the patients in a semi- recumbentposition, with eyes closed in a maximally alert state in a sound- attenuated room with subdued lighting. During the recording thealertness was controlled. If the patterns of drowsiness appeared in QEEG cordance was calculated by our EEG software (WaveFinder the EEG, the subjects were aroused by acoustic stimuli.
v.1.70, unimedis, Prague) using the algorithm for the cordancecalculation which has been described elsewhere in greater detail (This algorithm normalizes power acrossboth electrode sites and frequency bands in three consecutive steps: The first 50 sequential, non-overlapping, 2-second epochs collected first, absolute power values are reattributed to each individual during resting periods with eyes closed were selected to be electrode by averaging power from all bipolar electrode pairs sharing processed for each recording. Before analysis of the data, artifact that electrode (for example, the reattributed absolute power for Fp1 electrode is calculated as an average of the bipolar absolute power of level of 0.05 was adopted. The baseline characteristics, scores in pairs Fp1–F7, Fp1–F3, and Fp1–Fp2) (). This rating scales as well as values of cordance were expressed as a electrode referencing method is similar to the single source method median and interquartile ratio (IQR). The primary analysis was of ), in which voltage signals are recombined, except that conducted to detect difference between the number of responders the current method averages power from neighboring electrode pairs and non-responders who decreased cordance (Fisher Exact Test).
whereas the Hjorth transformation averages voltage amplitudes. It The difference in cordance value changes between responders and has been previously shown that electrode referencing on the basis of non-responders after one week of treatment was assessed using power averaging provides a stronger association between QEEG Mann–Whitney U Test. PPV, NPV, number need to diagnose (NND) measures and perfusion of underlying brain than either the linked- with exact binomial 95% confidence intervals (95 CI%) as well as post- ears reference or the conventional Hjorth method hoc effect size were also calculated. Based on our previous results Then the relative power values (percentage of power in each (), we planned our sample size to detect a frequency band) are calculated on the basis of dividing reattributed large effect size (difference between responders and non-respon- absolute power values by total power values for each electrode site in ders who reduced prefrontal cordance). Total sample size of 18 each frequency band. In the second step, for each individual EEG patients would be sufficient to detect an effect size (w) of 0.66 with recording the maximum absolute and relative power values (AMAXf, 81% power at a 5% level of statistical significance.
RMAXf) in each frequency band (f) are determined to obtain normal-ized absolute (ANORM (s,f)) and normalized relative (RNORM (s,f)) power values (each absolute and relative power values are divided by AMAXfand RMAXf respectively). This normalization process places absoluteand relative power values into a common unit (yielding values 3.1. Baseline and treatment characteristics and between 0 and 1) which allows them to be combined. In the third step, the cordance values at each electrode site (s) for each frequencyband (f) are calculated by summing the ANORM and RNORM values, after We analyzed 18 patients who finished 4 weeks of treatment.
a half-maximal value (0.5 on the normalized scale) are subtracted: Eleven (61%) out of 18 subjects responded to the treatment.
With the exception of gender, no baseline differences were ðs;fÞ = ANORM ðs;fÞ–0:5 + RNORM ðs;fÞ−0:5 : found between responders and non-responders in demographic For each individual EEG record, the cordance values from 3 fron- and clinical characteristics, duration of wash-out period or in tal electrodes (Fp1, Fp2 and Fz) in theta frequency band (4–8 Hz) average daily doses of bupropion at week 1 and week 4 (see were averaged and subjected to statistical analysis similar to We also did not find any significant difference in the number of patients taking zolpidem and hydroxyzine in both groups at week 1 (a day before the 2nd EEG session) and week 4.
2.6. Statistical methods and data analyses The scores of the clinical rating scales in patients over time aresummarized in and there were no differences between Analyses were performed using SPPS version 13. Due to the small responders and non-responders at baseline. Both groups sample size and non-normal data distribution we used nonparamet- differed in MADRS and CGI after first week of treatment but ric statistical tests (Fisher Exact Test, Mann–Whitney U Test, were not significantly different in the reductions of MADRS and Spearman's Rho). All tests were 2-sided and an exact significance Results of the clinical rating scales.
BDI-S—Beck Depression Inventory—Short Form, CGI—Clinical Global Impression, IQR—interquartile range, MADRS—Montgomery and ÅsbergDepression Rating Scale, NS—nonsignificant.
The change of prefrontal QEEG theta cordance as a predictor of response to bupropion treatment 3.2. Predictive value of prefrontal cordance ≤12 points. PPV and NPV were 0.6 (95%CI, 0.26–0.88) and 1.0 Nine of eleven responders and only one of seven non-respondersshowed a decrease in prefrontal QEEG cordance after the first week of drug administration (Fisher Exact Test, p = 0.01). Usingthe decrease of prefrontal cordance value after one week of The primary finding of this study was that the reduction treatment as an indicator of response to bupropion, PPV and of prefrontal QEEG cordance value in theta frequency band NPV of this test were 0.9 (95% CI, 0.56–1.0) and 0.75 (95% CI, after one week of bupropion treatment predicted clinical 0.35–0.97), respectively. NND for response was 1.48 (95% CI, response to 4-week treatment in patients who had failed 1.16–4.17) with the effect size (w) for response of 0.7. When to previous antidepressant treatments. We also found inter- cordance values were analyzed as continuous variables, we group difference (responders vs. non-responders) in cordance detected significant difference in prefrontal cordance value value changes at this time point. As far as we know, this is the changes between responders and non-responders (Mann–Whit- first study using the frontal theta band QEEG cordance as an ney U Test, U = 15, p = 0.03) after first week of bupropion early predictor of response to an antidepressant whose treatment. The higher baseline cordance value was found in mechanism of action does not involve inhibition of serotonin responders (Mann–Whitney U Test, U = 6, p = 0.002). For numer- reuptake. Previous study demonstrated predictive effect of a reduction of prefrontal cordance for SSRI or SNRI ( We found significant relationships between percentage reduc- tion of MADRS score from baseline to final visit and both the The decrease of theta prefrontal cordance we observed baseline cordance (Spearmen's Rho, rs = 0.64, p = 0.004) and the might be a potential correlate of early activity changes in change of cordance value after week 1 (Spearmen's Rho, rs = anterior cingulate and prefrontal cortex coupling with anti- −0.55, p=0.02). There were no correlations between baseline depressant response. The changes of metabolic activity in cordance value and severity of depression (baseline MADRS score).
anterior cingulate and adjacent orbital and prefrontal We also did not detect any relationship between benzodia- cortices were associated with response to treatment with zepine equivalent dose ) and baseline cordance value for all patients (rs = 0.02, p = 0.94) as well as between benzodiazepine equivalent dose and the change of cordance However, the link between decrease of theta pre- value after week 1 (rs = 0.21, p = 0.4). The baseline cordance frontal cordance and early activity changes in anterior values were not different between patients with benzodiaze- cingulate and prefrontal cortex is currently supported by pines (0.64, IQR 0.60–0.69) and without benzodiazepines (0.50, IQR 0.31–0.70, Mann–Whitney U Test, U = 18, p = 0.33) as well as observed significantly higher baseline cordance value in the change of cordance value after week 1 (−0.02, IQR −0.06– responders as well as the relationship between baseline 0.07 and 0.02, IQR −0.12–0.13, resp.; Mann–Whitney U Test, prefrontal cordance value and a reduction of MADRS in the whole sample that were not seen in previous studies Although we observed significant gender difference in final response rate, there was no significant difference in the change activity in theta frequency band reflects mainly the func- of cordance value after week 1 between males and females (Mann–Whitney U Test, U = 32, p = 0.51). The same result was achieved by comparing the cordance change between males and females who responded to the treatment (Mann–Whitney U our finding could be hypothetically consistent with test, U = 6, p = 0.28). We found significantly higher baseline the results of previous studies linking higher baseline metab- cordance value in females (0.68, IQR 0.64–0.72) comparing to olism as well as higher theta activity of anterior cingulate males (0.57, IQR 0.46–0.62) in whole sample (Mann–Whitney U with response to antidepressant treatment ( test, U = 13, p = 0.02) and no gender difference of baseline cordance value in responders group (Mann–Whitney U test, Reviewing previous “cordance” studies we identified the In addition, we calculated predictive parameters of cor- same pattern of results (higher cordance value in responders) dance reduction for remission (n = 6) defined as a MADRS score Prefrontal cordance values during study.
CF1—cordance value at baseline, CF2—cordance value after week 1, IQR—interquartile range.
), but none reached statistical significance.
(Antidepressant Treatment History Form — Based on our data we hypothesize that both parameters (baseline cordance value and early change of cordance) duration to response in responders to bupropion in Level 2 of closely interact and may predict changes in depressive STAR*D study that involved patients with similar degree of failure to previous AD treatment as in our project We suppose that reduction of cordance value outlined as a ). Moreover, at least four previous studies found dichotomous variable is more suitable for prediction of that the change after the first 2 or 4 weeks of treatment treatment outcome in clinical practice than baseline predicted the outcome at 6, 8 and 12 weeks ( cordance because there is no cut-off of cordance value to The increase of cordance value in non-responders Second, we used only a short wash-out period to prevent observed in our study might be also a promising predictor potential side effects of rapid switching as in our venlafa- but it has not yet been supported by sufficient body of xine study ). Since a previous study with evidence contrary to cordance decrease in responders randomized clinical trial design employed a wash-out and a placebo lead-in period prior to enrollment processes are coupled with increase of cordance in non- ) and detected similiar sensitivity, specificity and pre- responders. It might be due to non-response to treatment or dictive values as a study without a wash-out period ( a consequence of the ongoing changes related to pathophys- ), we supposed that the wash-out period might not be essential for the correct detection of prefrontal Accidentally, we detected significant gender difference cordance change in patients with a new antidepressant in the response rate. Several analyses have found that gender, menopausal status, and age can affect response to AD, Third, we did not include placebo control arm because whereas others have failed to show such differences Institutional Review Board of Prague Psychiatric Centre ). The recent pooled analysis did not find would not have approved a placebo-controlled study in the gender-related difference in the antidepressant efficacy of Fourth, the raters were not blind to medication; however, found a gender difference in baseline cordance value in whole they were blind to EEG results during the study. Next, the sample but not in responders and no differences were observed relatively small sample size could be a further limitation.
between males and females in changes of cordance values Nevertheless, our sample size calculation was based on after week 1 in whole sample or responders. We are not able to effect sizes observed in previous studies ( say if baseline cordance gender difference is a true gender ) and post-hoc effect size (w) estimated from this difference or a consequence of baseline cordance difference sample for response was in the large range ).
between responders and non-responders combined with Final limitation of our study is that we did not record EEG in unequal response gender ratio in our study. The study is too the end of study in all patients as it was not a part of our a small to elucidate this question. Since a previous study did not priori hypothesis. We collected EEG records after finishing detect gender baseline difference in cordance value the study only in responders in the framework of another in depressive patients we do not suppose any study (not yet published) evaluating stability of various EEG robust influence of gender on cordance prediction, however parameters in responders to acute treatment. Calculating we cannot exclude some smaller effect in patients treated cordance value we found four responders who did not reduce cordance after finishing the study. Two responders with We evaluated confounding influence of benzodiazepines increase of cordance value after week 1 continued as administration (stable dose during the study) on prefrontal cordance non-reducers. Since a previous study has demon- cordance change. We examined the relationship between strated a different pattern of cordance changes in placebo benzodiazepine equivalent dose and baseline cordance value responders (increase of cordance value) in comparison with for all patients as well as between benzodiazepine equiva- medication responders after 4 weeks of treatment, cordance lent dose and the change of cordance value after week 1 and changes might possibly differentiate true medication respon- found no significant correlation. Moreover, there was no ders and false (placebo) medication responders ( difference between patients with and without benzodiaze- However, there is no clear evidence supporting pines in baseline cordance values and in the change of such approach in individual patients.
cordance value after week 1. The relation between cordance Despite the limitations of this and other cordance studies and benzodiazepines, if present, did not appear to influence the early change of cordance value remains a promising tool in the prediction of antidepressant response. The data of our It is important to note several limitations of the current study together with the results of previous clinical trials study. First, the duration of 4 weeks might be too short to assess clinical response to bupropion and we cannot exclude the possibility of further clinical change decision whether to stop or continue with a given AD and thus emerging during longer treatment. In an outpatients' to reduce the period of ineffective treatment.
fluoxetine study, non-responders after 4 weeks of treatment There is a clear need of cordance studies combined with achieved better remission rate at week 12 but the response some neuroimaging or other neurophysiological methods to rate after week 4 of treatment was still substantial — 50% clearly determine physiological or pathophysiological mean- ings of cordance. This approach and combination or used as a cut-off point of antidepressant treatment adequacy comparison of cordance with other potential predictors of The change of prefrontal QEEG theta cordance as a predictor of response to bupropion treatment response will define the role and significance of cordance in Bares, M., Brunovsky, M., Kopecek, M., Novak, T., Stopkova, P., Kozeny, J., Sos, P., Krajca, V., Höschl, C., 2008. Early reductionin prefrontal theta QEEG cordance value predicts response tovenlafaxine treatment in patients with resistant depressive disorder. Eur. Psychiatry 23, 350–355.
Bazire, S., 2003. Psychotropic Drug Directory 2003/04. The Profes- This study was supported by a grant from Internal Grant Agency of sionals' Pocket Handbook & Aide Memoire. Five Pin Publishing Ministry of Health of Czech Republic (IGA MZ CR) Nr.9330-3. The IGA MZ CR had no further role in study design, in the collection and Beck, A.T., Rial, W.Y., Rickels, K., 1974. Short form of depression interpretation of the data, in the preparation of this report, and in inventory: cross-validation. Psychol. Rep. 34, 1184–1186.
the decision to publish this manuscript.
Cohen, J., 1988. Statistical Power Analyses for the Behavioral Sciences, 2nd ed. Lawrence Earlbaum Association, Hillsdale.
Cook, I.A., 2008. Biomarkers in psychiatry: potentials, pitfalls, and pragmatics. Primary Psychiatry 15, 54–59.
Cook, I.A., Leuchter, A.F., 2001. Prefrontal changes and treatment Dr. Bares designed the study, wrote protocol, coordinated all response prediction in depression. Semin. Clin. Neuropsychiatry project activities, participated in clinical part of project and wrote Cook, I.A., O'Hara, R., Uijtdehaage, S.H., Mandelkern, M., Leuchter, Dr. Brunovsky designed the study, wrote protocol and provided A.F., 1998. Assessing the accuracy of topographic EEG mapping assistance with EEG hardware and contributed to the analysis of the for determining local brain function. Electroencephalogr. Clin.
Dr. Novak assessed the patients throughout the study and Cook, I.A., Leuchter, A.F., Morgan, M., Witte, E., Stubbeman, W.F., Abrams, M., Rosenberg, S., Uijtdehaage, S.H., 2002. Early Dr. Kopecek designed study, wrote the protocol, assessed changes in prefrontal activity characterize clinical responders patients and managed literature searches.
to antidepressants. Neuropsychopharmacology 27, 120–131.
Dr. Stopkova participated in clinical part of project, discussed Cook, I.A., Leuchter, A.F., Morgan, M.L., Stubbeman, W., Siegman, B., Abrams, M., 2005. Changes in prefrontal activity characterize clinical response in SSRI nonresponders: a pilot study. J.
Mr. Krajca provided assistance with EEG hardware and software.
Dr. Höschl designed study and discussed results.
Drevets, W.C., 1998. Functional neuroimaging studies of depression: All authors contributed, revised and approved the final the anatomy of melancholia. Annu. Rev. Med. 49, 341–361.
Drevets, W.C., Price, J.L., Furey, M.L., 2008. Brain structural and functional abnormalities in mood disorders: implications forneurocircuitry models of depression. Brain Struct. Funct. 213, Guy, W., 1976. ECDEU assessment manual for psychopharmacology— revised. US Dept. Health, Education and Welfare Publication In the last three years, Dr. Höschl has received travel grant from Eli (ADM) 76-338. National Institute of Mental Health, Rockville, MD, Lilly and comp. He has been a consultant for Servier, and he has been a speaker for Eli Lilly & Co. and Bristol-Myers Squibb Company. He is also Fredman, S.J., Fava, M., Kienke, A.S., White, C.N., Nierenberg, A.A., a faculty member, Lundbeck International Neuroscience Foundation.
Rosenbaum, J.F., 2000. Partial response, nonresponse, and He has received consulting honoraria from United Biosource Corpora- relapse with selective serotonin reuptake inhibitors in major tion. He has no shares, no other conflicts of interests.
depression: a survey of current “next-step” practices. J. Clin.
Other authors declare that they have no conflicts of interests.
Hjorth, B., 1975. An on-line transformation of EEG scalp potentials into orthogonal source derivations. Electroencephalogr. Clin.
Neurophysiol. 39, 526–530.
Hunter, A.M., Cook, I.A., Leuchter, A.F., 2007. The promise of the The authors thank Jan Volavka, M.D. for valuable advice and help quantitative electroencephalogram as a predictor of antidepres- with the final revision of the manuscript, Ms. Kveta Vonaskova and sant treatment outcomes in major depressive disorder. Psy- Ms. Jolana Sediva for administrative and technical support, and Ms.
chiatr. Clin North. Am. 30, 105–124.
Prajzlerova, Ms.Vrubelova and Ms. Rihakova for collecting the EEG Keller, M.B., 2005. Issues in treatment-resistant depression. J. Clin.
Kennedy, S.H., Fulton, K.A., Bagby, R.M., Greene, A.L., Cohen, N.L., Rafi-Tar, S., 2006. Sexual function during bupropion or paroxetinetreatment of major depressive disorder. Can. J. Psychiatry 51, American Psychiatric Association, 1994. Diagnostic and Statistical Manual of Mental Disorders, 4th ed. American Psychiatric Press, Kornstein, S.G., Sloan, D.M.E., 2006. Depression and gender. In: Stein, D.J., Kupfer, D.J., Schatzberg, A.F. (Eds.), Textbook of Asada, H., Fukuda, Y., Tsunoda, S., Yamaguchi, M., Tonoike, M., Mood Disorders. American Psychiatric Publishing, Inc., Arlington, 1999. Frontal midline theta rhythms reflect alternative activa- tion of prefrontal cortex and anterior cingulate cortex in humans.
Leuchter, A.F., Cook, I.A., Lufkin, R.B., Dunkin, J., Newton, T.F., Cummings, J.L., Mackey, J.K., Walter, D.O., 1994a. Cordance: a Bares, M., Brunovsky, M., Kopecek, M., Stopkova, P., Novak, T., new method for assessment of cerebral perfusion and metabo- Kozeny, J., Höschl, C., 2007. Changes in QEEG prefrontal lism using quantitative electroencephalography. Neuroimage 1, cordance as a predictor of response to antidepressants in patients with treatment resistant depressive disorder: a pilot Leuchter, A.F., Cook, I.A., Mena, I., Dunkin, J.J., Cummings, J.L., study. J. Psychiatr. Res. 41, 319–325.
Newton, T.F., Migneco, O., Lufkin, R.B., Walter, D.O., Lachenbruch, P.A., 1994b. Assessment of cerebral perfusion Benca, R.M., Davidson, R.J., 2001. Anterior cingulate activity as using quantitative EEG cordance. Psychiatry Res. 55, 141–152.
a predictor of degree of treatment response in major depression: Leuchter, A.F., Cook, I.A., Witte, E.A., Morgan, M., Abrams, M., evidence from brain electrical tomography analysis. Am. J.
2002. Changes in brain function of depressed subjects during treatment with placebo. Am. J. Psychiatry 159, 122–129.
Pizzagalli, D.A., Oakes, T.R., Davidson, R.J., 2003. Coupling of theta Leuchter, A.F., Uijtdehaage, S.H., Cook, I.A., O'Hara, R., Mandelk- activity and glucose metabolism in the human rostral anterior ern, M., 1999. Relationship between brain electrical activity and cingulate cortex: an EEG/PET study of normal and depressed cortical perfusion in normal subjects. Psychiatry Res. 90, subjects. Psychophysiology 40, 939–949.
Quitkin, F.M., Petkova, E., McGrath, P.J., Tailor, B., Beasley, C., Lozano, A.M., Mayberg, H.S., Giacobbe, P., Hamani, C., Craddock, Stewart, J., Amsterdam, J., Fava, M., Rosenbaum, J., Reimherr, R.C., Kennedy, S.H., 2008. Subcallosal cingulate gyrus deep brain F., Fawcett, J., Chen, Y., Klein, D., 2003. When should a trial of stimulation for treatment-resistant depression. Biol. Psychiatry fluoxetine for major depression be declared failed? Am. J.
Mayberg, H.S., Brannan, S.K., Mahurin, R.K., Jerabek, P.A., Brick- Rush, A.J., Kraemer, H.C., Sackeim, H.A., Fava, M., Trivedi, M.H., man, J.S., Tekell, J.L., Silva, J.A., McGinnis, S., Glass, T.G., Frank, E., Ninan, P.T., Thase, M.E., Gelenberg, A.J., Kupfer, D.
Martin, C.C., Fox, P.T., 1997. Cingulate function in depression: a J., Regier, D.A., Rosenbaum, J.F., Ray, O., Schatzberg, A.F., potential predictor of treatment response. Neuroreport 8, Task Force, A.C.N.P., 2006a. Report by the ACNP Task Force on response and remission in major depressive disorder. Neuropsy- Mayberg, H.S., Brannan, S.K., Tekell, J.L., Silva, J.A., Mahurin, R.
K., McGinnis, S., Jerabek, P.A., 2000. Regional metabolic effects Rush, A.J., Trivedi, M.H., Wisniewski, S.R., Stewart, J.W., Nieren- of fluoxetine in major depression: serial changes and relationship berg, A.A., Thase, M.E., Ritz, L., Biggs, M.M., Warden, D., to clinical response. Biol. Psychiatry 48, 830–843.
Luther, J.F., Shores-Wilson, K., Niederehe, G., Fava, M., STAR*D Mayberg, H.S., Lozano, A.M., Voon, V., McNeely, H.E., Seminowicz, Study Team, 2006b. Bupropion-SR, sertraline, or venlafaxine-XR D., Hamani, C., Schwalb, J.M., Kennedy, S.H., 2005. Deep brain after failure of SSRIs for depression. N. Engl. J. Med. 354, stimulation for treatment-resistant depression. Neuron 45, Sackeim, H.A., 2001. The definition and meaning of treatment- Montgomery, S.A., Asberg, M., 1979. A new depression scale resistant depression. J. Clin. Psychiatry 62 (Suppl 16), 10–17.
designed to be sensitive to change. Br. J. Psychiatry 134, Sheehan, D.V., Lecrubier, Y., Sheehan, K.H., Amorim, P., Janavs, J., Weiller, E., Hergueta, T., Baker, R., Dunbar, G.C., 1998. The Morgan, M.L., Witte, E.A., Cook, I.A., Leuchter, A.F., Abrams, M., Mini-International Neuropsychiatric Interview (M.I.N.I.): the Siegman, B., 2005. Influence of age, gender, health status, and development and validation of a structured diagnostic psychiat- depression on quantitative EEG. Neuropsychobiology 52, 71–76.
ric interview for DSM-IV and ICD-10. J. Clin. Psychiatry 59 (Suppl Mulert, C., Juckel, G., Brunnmeier, M., Karch, S., Leicht, G., Mergl, R., Möller, H.J., Hegerl, U., Pogarell, O., 2007. Rostral anterior Souery, D., Oswald, P., Massat, I., Bailer, U., Bollen, J., Demytte- cingulate cortex activity in the theta band predicts response to naere, K., Kasper, S., Lecrubier, Y., Montgomery, S., Serretti, antidepressive medication. Clin. EEG Neurosci. 38, 78–81.
A., Zohar, J., Mendlewicz, J., Group for the Study of Resistant Nierenberg, A.A., Farabaugh, A.H., Alpert, J.E., Gordon, J., Depression, 2007. Clinical factors associated with treatment Worthington, J.J., Rosenbaum, J.F., Fava, M., 2000. Timing of resistance in major depressive disorder: results from a European onset of antidepressant response with fluoxetine treatment. Am.
multicenter study. J. Clin. Psychiatry 68, 1062–1070.
Szegedi, A., Jansen, W.T., van Willigenburg, A.P., van der Meulen, Nuwer, M.R., Lehmann, D., da Silva, F.L., Matsuoka, S., Sutherling, E., Stassen, H.H., Thase, M.E., 2009. Early improvement in the W., Vibert, J.F., 1999. IFCN guidelines for topographic and first 2 weeks as a predictor of treatment outcome in patients frequency analysis of EEGs and EPs. Electroencephalogr. Clin.
with major depressive disorder: a meta-analysis including 6562 patients. J. Clin. Psychiatry 70, 344–353.
Papakostas, G.I., Kornstein, S.G., Clayton, A.H., Soares, C.N., Thase, M.E., Rush, A.J., 1997. When at first you don't succeed: Hallett, L.A., Krishen, A., Tucker, V.L., 2007a. Relative sequential strategies for antidepressant nonresponders. J. Clin.
antidepressant efficacy of bupropion and the selective sero- tonin reuptake inhibitors in major depressive disorder: Trivedi, M.H., Morris, D.W., Grannemann, B.D., Mahadi, S., 2005.
gender–age interactions. Int. Clin. Psychopharmacol. 22, Symptom clusters as predictors of late response to antidepres- sant treatment. J. Clin. Psychiatry 66, 1064–1070.
Papakostas, G.I., Petersen, T., Sklarsky, K.G., Nierenberg, A.A., Trivedi, M.H., Rush, A.J., Wisniewski, S.R., Nierenberg, A.A., Alpert, J.E., Fava, M., 2007b. Timing of clinical improvement and Warden, D., Ritz, L., Norquist, G., Howland, R.H., Lebowitz, symptom resolution in the treatment of major depressive B., McGrath, P.J., Shores-Wilson, K., Biggs, M.M., Balasubra- disorder. Psychiatry Res. 149, 195–200.
mani, G.K., Fava, M., STAR*D Study Team, 2006. Evaluation of Papakostas, G.I., Fava, M., Thase, M.E., 2008. Treatment of SSRI- outcomes with citalopram for depression using measurement- resistant depression: a meta-analysis comparing within-versus based care in STAR*D: implications for clinical practice. Am. J.
across-class switches. Biol. Psychiatry 63, 699–704.
Pizzagalli, D., Pascual-Marqui, R.D., Nitschke, J.B., Oakes, T.R., Larson, C.L., Abercrombie, H.C., Schaefer, S.M., Koger, J.V.,

Source: http://www.pcp.lf3.cuni.cz/pcpout/dokumenty/sos_publikace/04_The_change_of_prefrontal_QEEG.pdf