- Email: [email protected]
Brain functional alterations observed 4-weekly in major depressive disorder following antidepressant treatment
Journal of Affective Disorders 252 (2019) 25–31
Contents lists available at ScienceDirect
Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad
Research paper
Brain functional alterations observed 4-weekly in major depressive disorder following antidepressant treatment
T
⁎
Bun Yamagataa, , Kaori Yamanakaa, Yuichi Takeib, Shogo Hottaa, Jinichi Hiranoa, Hajime Tabuchia, Masaru Mimuraa a b
Department of Neuropsychiatry, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo 160-8582, Japan Department of Psychiatry and Neuroscience, Gunma University Graduate School of Medicine, Maebashi, Gunma, Japan
A R T I C LE I N FO
A B S T R A C T
Keywords: Depression Near-infrared spectroscopy Antidepressant treatment Longitudinal study Neuroimaging Sertraline
Background: Major depressive disorder (MDD) is a heterogeneous condition. Identifying the brain responses to antidepressant treatment is of particular interest as these may represent potential neural networks related to treatment response, forming one aspect of the biological markers of MDD. Near-infrared spectroscopy (NIRS) is suitable for repeated measurements with short intervals because of its noninvasiveness, and can provide detailed time courses of functional alterations in prefrontal regions. Methods: We conducted a 12-week longitudinal study to explore prefrontal hemodynamic changes at 4-week intervals following sertraline treatment in 11 medication-naïve participants with MDD using 52-channel NIRS. Results: While all participants achieved remission after treatment, intra-class correlation coefficient of oxygenated hemoglobin [oxy-Hb] values throughout the 12-week observation was moderate at the spatially and temporally contiguous cluster located in the left inferior frontal and temporal gyri. There was a significant negative correlation between mean [oxy-Hb] values in the significant cluster at 4 weeks and changes in Hamilton Rating Scale for Depression total score from 4 to 8 weeks (r = −0.73, P = 0.011) and from 4 to 12 weeks (r = −0.63, P = 0.039). Limitations: Without healthy controls for comparison, we were unable to fully evaluate whether improvement of [oxy-Hb] activations after treatment in MDD reached normal levels or not. Conclusion: Our NIRS findings of detailed prefrontal hemodynamic alterations over short interval observations such as 4 weeks may have revealed potential trait marker for MDD and biological maker for predicting clinical response to sertraline treatment in MDD.
1. Introduction Major depressive disorder (MDD), a leading cause of disability worldwide, is characterized by affective, cognitive, and somatic symptoms (Kessler et al., 2005; Zhang et al., 2011). A number of neuroimaging studies have demonstrated alterations in widespread neural networks, including the medial prefrontal cortex and limbic, striatal, thalamic, and basal forebrain areas, in individuals with MDD (Diener et al., 2012; Pizzagalli, 2011; Price and Drevets, 2012). Despite the importance of understanding the etiology of MDD, there is still no established biological marker for its diagnosis and treatment. Given the
heterogeneity of this disorder, identifying brain alterations in response to antidepressant treatment is particularly important as this may indicate potential neural networks related to treatment response, as one aspect of the biological markers of MDD. When comparing pre- and post-treatment time-points, recent resting-state functional magnetic resonance imaging (rs-fMRI) studies examining responses to medication have consistently reported that functional connectivity (FC) between the prefrontal and limbic brain regions increases after treatment (Dichter et al., 2015). This indicates greater inhibitory control over neural circuits related to emotion processing and regulation (Dichter et al., 2015). One study found that
Abbreviations: MDD, major depressive disorder; rs-fMRI, resting-state functional magnetic resonance imaging; FC, functional connectivity; DMN, default mode network; NIRS, near-infrared spectroscopy; oxy-Hb, oxygenated hemoglobin; deoxy-Hb, deoxygenated hemoglobin; VFT, verbal fluency task; HC, healthy controls; HRS-D, Hamilton Rating Scale for Depression; total-Hb, total-hemoglobin; BA, Brodmann's area; BOLD, blood-oxygenation level-dependent; ANOVA, analysis of variance; ECT, electroconvulsive therapy ⁎ Corresponding author. E-mail address: [email protected] (B. Yamagata). https://doi.org/10.1016/j.jad.2019.04.001 Received 17 December 2018; Received in revised form 29 March 2019; Accepted 2 April 2019 Available online 03 April 2019 0165-0327/ © 2019 Elsevier B.V. All rights reserved.
Journal of Affective Disorders 252 (2019) 25–31
B. Yamagata, et al.
Table 1 Demographic data.
Age (years) Education (years) Sex (male/female) HRS-D total score VFT performance Dose of sertraline (mg/day)
Baseline
Week 4
Week 8
Week 12
F-value
P-value
14.5 ± 5.6 14.1 ± 6.0 0
8.5 ± 5.8 12.4 ± 3.6 29.5 ± 10.1
4.6 ± 4.1 13.6 ± 3.2 61.4 ± 20.5
4.6 ± 2.7 13.2 ± 3.6 65.9 ± 23.1
11.7 1.72 43.4
0.003 0.247 <0.0001
36.3 ± 11.2 15.4 ± 2.3 5/6
Abbreviations: HRS-D, Hamilton rating scale for depression; VFT, Verbal fluency task.
Hb] values alter after initiation of antidepressant treatment until achieving clinical improvement. We therefore performed a longitudinal study to explore frontal hemodynamic changes at 4-week intervals using 52-channel NIRS following treatment with the antidepressant sertraline in medication-naïve MDD participants, to reveal whether the hypofrontality observed in MDD alters between starting antidepressant treatment and as depressive symptoms improve.
connectivity of the posterior cingulate cortex, part of the default mode network (DMN), to the right lateral parietal cortex and right inferior temporal gyrus normalized after duloxetine therapy (Posner et al., 2013). Additionally, a recent study reported that baseline FC between the amygdala and insula was associated with a greater reduction in depressive symptoms 3 months later, suggesting that FC between these neural circuit potentially serves as a candidate biomarker for treatment response (Connolly et al., 2017). Although these normalized brain networks after treatment have been identified as biological markers of treatment response in MDD, prior studies have had a potential drawback: In most prior studies, individuals with MDD have been scanned at relativity long intervals such as 8 or 12 weeks, and detailed alterations in neural activities at shorter intervals have not been sufficiently investigated. For example, examining whether brain response at the beginning of therapy contributes to treatment reactivity is of particular interest. Additionally, MRI scanning has disadvantages, such as high cost, loud noise, and being in an enclosed space for a long period of time, so may not be suitable for repeated measurements at short intervals such as 2 or 4 weeks. Near-infrared spectroscopy (NIRS) is an optical neuroimaging technique that allows the noninvasive measurement of changes in the concentrations of oxygenated and deoxygenated hemoglobin ([oxy-Hb] and [deoxy-Hb], respectively), which reflect changes in regional cerebral blood volume (Hoshi and Tamura, 1993; Villringer et al., 1993). NIRS has the following advantages; (1) complete noninvasiveness of the measurement, facilitating repeated measurements, (2) higher temporal resolution (0.1 s time resolution) than fMRI, and (3) portability and compactness of the apparatus, enabling measurements under natural conditions with subjects sitting on a comfortable chair. We therefore hypothesized that NIRS would be more suitable for frequent measurements, such as 2- or 4-week intervals, to observe treatment responses in psychiatric conditions. Prior NIRS studies have consistently reported that [oxy-Hb] significantly decreases in the fronto-temporal brain regions during a verbal fluency task (VFT) in participants with MDD compared with healthy controls (HC) (Kameyama et al., 2006; Matsuo et al., 2002; Pu et al., 2008; Suto et al., 2004). A meta-analysis of NIRS studies further supports the previous findings of hypofrontality in MDD (Zhang et al., 2015). Such observations suggest that NIRS may be a useful tool for investigating neural activity with fine time resolution. However, it is still unclear whether NIRS signal changes during VFT are state- or trait-dependent since previous longitudinal studies have not obtained a consistent finding. For instance, a prior 12-week longitudinal study demonstrated that there were no [oxy-Hb] changes between pre- and post-antidepressant treatments in the MDD despite significant improvement in depressive symptoms, suggesting that hypofrontality response to a VFT may represent a potential trait marker for depression rather than a state marker (Tomioka et al., 2015). On the other hand, a recent study has reported that during 1.5-year follow-up, [oxy-Hb] changes in the right inferior frontal gyrus synchronized with changes of depressive symptoms, whereas [oxy-Hb] activation in the bilateral middle frontal gyrus was rather stable over time. This study suggested that NIRS would separate state- or trait-related brain regions in MDD, respectively (Satomura et al., 2019). Furthermore, there has been no study to measure repeatedly with a short interval how [oxy-
2. Methods 2.1. Participants This study included eleven drug-naïve adult individuals with MDD (Table 1). All participants were recruited from Keio University Hospital (Tokyo, Japan). After an extensive description of the study, written informed consent was obtained from all participants. The study protocol was prepared in accordance to the ethical standards of the Declaration of Helsinki and approved by the Ethics Committee of Keio University School of Medicine in February 27th, 2012 (registration number: UMIN000017269). We then started recruiting participants from March 1st, 2012 and completed collecting data in December 31st, 2016. The clinical diagnoses of depression were based on the DSM-IV-TR (American Psychiatric Association, 2000). Participants were confirmed to have had no lifetime history of bipolar disorder, psychosis, obsessivecompulsive disorder, or drug or alcohol misuse, or any neurological disorder. Furthermore, none of the participants reported any unstable medical condition or history of significant head trauma. Depression severity was evaluated using the 17-item Hamilton Rating Scale for Depression (HRS-D). We performed NIRS scans and clinical assessments at four time-points: before starting antidepressant treatment (baseline), and at 4, 8, and 12 weeks. The participants started treatment with 25 mg/day sertraline after the baseline scan. Partial remission was defined as ≤ 10 on the 17-item HRS-D based on DSM-IV criteria (American Psychiatric Association, 2000; Frank et al., 1991).
2.2. Verbal fluency task The task procedure in the current study was similar to that used by (Takizawa et al., 2014). Briefly, [oxy-Hb] changes were measured during the VFT (letter version), which is composed of a 30 s pre-task baseline, 60 s task, and a 60 s post-task baseline. For the pre- and posttask baseline periods, the participants were instructed to consecutively repeat the five Japanese vowels (“a”, “i”, “u”, “e”, “o”) aloud. During the task period, participants were requested to produce as many nouns as possible beginning with a designated syllable without the use of repetitions and proper nouns. The three sets of initial syllables (A;/to/, /se/, /o/, B; /a/, /ki/, /ha/, C; /na/, /i/, /ta/) were presented in counterbalanced order among the participants and each syllable changed every 20 s during the 60 s task.
26
Journal of Affective Disorders 252 (2019) 25–31
B. Yamagata, et al.
described in (Maris and Oostenveld, 2007). This method effectively controls the Type I-error rate with respect to multiple comparisons over channels and time-points because NIRS has 52 channels over the prefrontal and temporal regions and there are 1251 time-points during the whole experimental period (total duration is 125.1 s with 0.1 s time resolution). This is achieved by clustering neighboring channels and time-points. For the first-level statistics, channels and time-points over a threshold (F > 3.72) were identified from the repeated-measures ANOVA and spatially and temporally contiguous points in terms of channels and time-points over this threshold were subsequently defined as a cluster. Then the sum of the F values for a given cluster was used for the cluster-level statistics. By repeatedly resampling the data across all subjects, cluster-level F values were created from 10,000 randomization routines. For the second-level statistics, the P value was estimated according to the proportion of the randomization distribution exceeding the observed maximum cluster-level statistics (see details of methods in our previous study (Yamagata et al., 2017)). We used a “spatio_temporal_cluster_test” in MNE-python (Gramfort et al., 2013) for the cluster-based nonparametric randomization test of the NIRS data. We then extracted mean [oxy-Hb] data from the cluster showing the significant main effect of time-point and performed the Friedman test followed by the Wilcoxon Signed-rank Test with Bonferroni correction for a post-hoc analysis to examine the differences in [oxy-Hb] at the four time-points (baseline, 4, 8, and 12 weeks). In order to examine relationship between [oxy-Hb] values and clinical response to sertraline treatment, we first performed a correlation analysis between mean [oxy-Hb] values at baseline extracted from the spatially and temporally contiguous cluster mentioned above and changes of HRS-D total scores at any time-points. Second, in order to investigate whether changes of NIRS signals would represent state-dependent, we conducted the Spearman's rank correlation analysis between intra-individual longitudinal changes in [oxy-Hb] values at the spatially and temporally contiguous cluster ([oxy-Hb] at 12 weeks – [oxy-Hb] at baseline) and HRS-D total scores (HRS-D at 12 weeks – HRS-D at baseline). Additionally, to identify whether brain activation remains rather constant throughout the 12-week observation, we calculated intra-class correlations (ICCs, one-way random, single measure) between [oxy-Hb] values at baseline, 4, 8, and 12 weeks at the spatially and temporally contiguous cluster. Statistical significance was defined as P < 0.05 and all analyses were performed using SPSS (IBM SPSS Statistics, Version 25.0. Armonk, NY) except for the cluster-based nonparametric randomization test.
2.3. NIRS measurements The current study utilized a 52-channel NIRS machine (ETG-4000 Optical Topography System; Hitachi, Ltd., Japan) using two different wavelengths (695 nm and 830 nm) with 0.1 s time resolution to measure relative changes of absorbed near-infrared light. These changes are transformed into [oxy-Hb], [deoxy-Hb] and [total-hemoglobin (totalHb); sum of oxy-Hb and deoxy-Hb] concentration changes, as indicators of brain activity, by means of a modified Beer–Lambert law (Obrig and Villringer, 2003). The units were changes in chromophore concentration (mM) multiplied by the path length (mm) of the near-infrared light. We utilized 33 probes composed of 17 light emitters and 16 detectors with an inter-optrode distance of 30 mm. A measuring point of activation (channel) was defined as the region between one emitter and one detector. Thus, the probe set covered the brain regions including the bilateral prefrontal (approximately dorsolateral [Brodmann's area (BA) 9, 46], ventrolateral [BA 44, 45, 47], frontopolar [BA 10]) and superior and middle temporal cortical regions (BA 22, 41, 42). The correspondence between the probe positions and the measurement areas on the cerebral cortex was confirmed based on a previous multisubject study of anatomical craniocerebral correction via the international 10–20 system (Okamoto et al., 2004; Tsuzuki et al., 2007) (Fig. 1). The obtained data were analyzed using integral mode; the pre-task baseline was the mean of a 10 s period just prior to the task period, and the post-task baseline was the mean of the last 5 s of the post-task period; linear fitting was applied to the data between these two baselines. To remove short-term motion artifacts, we applied a moving average method using a window width of 5 s. We then calculated the median and standard deviation at each channel, rejecting data above or below 3 standard deviations of the median value. [oxy-Hb] represents regional neural activation and its change is assumed to more directly reflect cognitive activation than [deoxy-Hb] change, as shown by a stronger correlation with blood-oxygenation level-dependent (BOLD) signal measured by fMRI study (Strangman et al., 2002). 2.4. Data analysis Regarding the clinical and behavioral data, we compared total HRSD scores and total VFT numbers between four time-points including baseline, 4, 8, and 12 weeks using a repeated-measures analysis of variance (ANOVA). To examine differences in [oxy-Hb] between the four time-points, we first compared the time-course [oxy-Hb] changes during the whole baseline and VFT periods using a cluster-based nonparametric randomization test (P < 0.05 corrected) based on Monte-Carlo estimates, as
Fig. 1. Location of the 52 channels in the near-infrared spectroscopy. Corresponding anatomical site of each channel. Yellow circle represents “superior frontal gyrus”, green circle represents “Middle frontal gyrus”, light blue circle represents “inferior frontal gyrus”, blue circle represents “superior and middle temporal gyri”, red circle represents “precentral gyrus”, orange circle represents “postcentral gyrus”, and brown circle represents “supramarginal gyrus”. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) 27
Journal of Affective Disorders 252 (2019) 25–31
B. Yamagata, et al.
Fig. 2. Grand average waveforms of [oxy-Hb] changes during VFT in the fronto-temporal brain regions. A) The 52 measuring positions are labeled Ch1 to Ch52, from the right temporal to the left temporal regions. Orange colored regions represent a spatially and temporally contiguous cluster showing a significant main effect of time-point in [oxy-Hb] values during VFT in drug-naïve major depressive disorder (P < 0.05 corrected; a cluster-based nonparametric randomization test). B) Example of detailed time-course [oxy-Hb] changes with a 0.1 s time resolution between four time-points at Ch50 and Ch42. Purple line represents “baseline”, green line represents “4 weeks”, blue line represents “8 weeks”, and orange line represents “12 weeks”. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
3. Results
(P = 0.002). [oxy-Hb] at baseline significantly declined after starting antidepressant treatment until 4 weeks and then decreased [oxy-Hb] compared with baseline persisted until 12 weeks [baseline > 4 weeks (P < 0.001); baseline > 8 weeks (P = 0.003); baseline > 12 weeks (P = 0.021)]. Furthermore, [oxy-Hb] values at 4 weeks gradually increased, corresponding with clinical improvements but there were no significant differences between 4 and 8 weeks (P = 0.509), between 4 and 12 weeks (P = 0.186), and between 8 and 12 weeks (P = 0.509) (Fig. 3).
3.1. Demographic characteristics There were no significant differences between four time-points in VFT performance (F(3,8) = 1.73, P = 0.247), but depressive symptoms based on HRS-D scores significantly improved after starting sertraline treatment (F(3,8) = 11.67, P < 0.003) (Table 1). At 12 weeks, all participants showed at least partial remission based on DSM-IV criteria (American Psychiatric Association, 2000; Frank et al., 1991).
3.3. Relationship between [oxy-Hb] values and clinical response to sertraline treatment
3.2. Longitudinal comparison of [oxy-Hb] time-courses between baseline, 4, 8, and 12 weeks
There were no significant correlations between mean [oxy-Hb] values at baseline extracted from the spatially and temporally contiguous cluster and changes of HRS-D total scores at any time point. Furthermore, changes of mean [oxy-Hb] values at the cluster ([oxy-Hb] at 12 weeks – [oxy-Hb] at baseline) were not significantly correlated with changes of HRS-D total scores (HRS-D at 12 weeks –HRS-D at baseline). Given that our results showed that [oxy-Hb] values increased from 4 to 12 weeks, after an initial reduction from baseline to 4 weeks, we further investigated whether [oxy-Hb] values at 4 weeks would
Cluster-based nonparametric randomization testing identified a significant main effect of time-point in [oxy-Hb] activation during VFT at one spatially and temporally contiguous cluster located in the left inferior frontal gyrus, postcentral gyrus, and superior and middle temporal gyri (this cluster consists of Ch 31, 39, 41, 42, 50, 51, and 52; duration time = 17.7 s–62.5 s; P = 0.046 corrected) (Fig. 2). A post-hoc analysis of mean [oxy-Hb] values extracted from the contiguous cluster revealed a significant main effect of time-point 28
Journal of Affective Disorders 252 (2019) 25–31
B. Yamagata, et al.
Fig. 3. Comparison of [oxy-Hb] during VFT at baseline, week 4, 8, and 12. Each bar represents the mean [oxy-Hb] values extracted a spatially and temporally contiguous cluster showing a significant main effect of time-point in [oxy-Hb] values during VFT in drug-naïve major depressive disorder. Friedman test revealed significant differences in [oxy-Hb] at the four time-points. * P < 0.05, error bar represents standard error of the mean.
emotional networks compared with healthy controls, indicating that these brain dysfunctions may potentially be predominant biological makers for depression (Kanske et al., 2012; Okada et al., 2009; Smoski et al., 2013; Usami et al., 2014). These discrepancies among studies may be because of differences in brain imaging methodologies, activation paradigms including cognitive or emotional tasks, and methods of treatment (e.g. antidepressant or electroconvulsive therapy [ECT]). Additionally, it has been controversial as to whether NIRS signals represent state- or trait-dependent since only a few longitudinal studies have been conducted so far. A previous longitudinal NIRS study identified that hypoactivation during VFT in the bilateral prefrontal and temporal cortices persisted from baseline to 12 weeks despite improvement of depressive symptoms in MDD, suggesting that [oxy-Hb] changes are trait-dependent (Tomioka et al., 2015). On the other hand, Satomura et al. performed a long-term longitudinal study of 1.5 years and revealed that NIRS signals separately exhibit state- or trait-related brain sub-regions in MDD (Satomura et al., 2019). In the current study, [oxy-Hb] changes in the left inferior frontal and temporal regions did not synchronize with changes of depressive symptoms, whereas [oxyHb] activation in these regions remained consistent throughout the 12week observation. Therefore, we surmise that our findings of persistent functional abnormality in the left prefrontal and temporal cortices before and after antidepressant treatment may be a trait maker representing neurobiological vulnerability for depression rather than a marker specific to the patient's current state. Intriguingly, although ICC analysis demonstrated consistency of [oxy-Hb] activations in the left inferior frontal and temporal regions from baseline to 12 weeks, our frequent 4-weekly measurements identified atypical [oxy-Hb] time-course changes throughout the 12week observation. In particular, [oxy-Hb] significantly decreased from baseline to 4 weeks, and then increased from 4 to 12 weeks, correlating with clinical improvement. A previous fMRI study showed that activation in the left prefrontal cortex and right premotor cortex decreased during a linguistic task with paroxetine treatment (Peran et al., 2008). Other studies have also reported sedative effects of antidepressants or
predict treatment response. There was a significant negative correlation between mean [oxy-Hb] values in the significant cluster at 4 weeks and changes in HRS-D total score from 4 to 8 weeks (r = −0.73, P = 0.011) and from 4 to 12 weeks (r = −0.63, P = 0.039). These results may indicate that less [oxy-Hb] in the left inferior frontal and temporal regions at 4 weeks may relate to greater clinical improvements in depressive symptoms after sertraline treatment. 3.4. Consistency of [oxy-Hb] values from baseline to 12 weeks The significant ICC of [oxy-Hb] among baseline, 4, 8, and 12 weeks was demonstrated at the spatially and temporally contiguous cluster located in the left inferior frontal gyrus and temporal regions [ICC (1,1) = 0.468; P < 0.001]. 4. Discussion This study revealed that [oxy-Hb] during VFT in medication-naïve individuals with MDD at 12 weeks after starting sertraline treatment are significantly lower than at baseline in the left inferior frontal gyrus, postcentral gyrus, and superior and middle temporal gyri. This is despite significant improvement in depressive symptoms. In particular, during the 12-week treatment, [oxy-Hb] values initially reduced from baseline to 4 weeks and then gradually increased from 4 to 12 weeks. Furthermore, the intra-class correlation coefficient of [oxy−Hb] values between 4 measurement points was moderate in the left inferior frontal gyrus and temporal gyri (ICC interpretation scale; moderate [0.41 to 0.60]). Correlation analysis revealed that [oxy-Hb] values at 4 weeks may be a potential predictor of treatment response in MDD. Prior neuroimaging studies demonstrated reversal or normalization of abnormal brain function following antidepressant treatment in MDD (Delaveau et al., 2011; Fitzgerald et al., 2008). However, previous fMRI studies using VFT or emotional tasks have reported that even in the absence of current symptoms, individuals with remitted MDD showed reduced activity in brain networks related to the verbal fluency or 29
Journal of Affective Disorders 252 (2019) 25–31
B. Yamagata, et al.
Acknowledgment
antihistamines on NIRS signals (Kohmura et al., 2013; Tsujii et al., 2009, 2007). Additionally, our recent study revealed that high-doses of antidepressants are associated with attenuated [oxy-Hb] during a cognitive task (Takamiya et al., 2017). Based on these observations, we assumed that our findings of attenuated [oxy-Hb] values at 4 weeks after commencing sertraline therapy may be the consequence of introducing antidepressant treatment, rather than clinical responses. Moreover, given the consistency of [oxy-Hb] values over 12 weeks, it is speculated that the effects of artifacts due to repeated measurements on atypical [oxy-Hb] time-course changes are minimal. Taken together, our frequent 4-weekly measurements identified atypical [oxy-Hb] timecourse alterations throughout the 12-week observation, revealing not only the effect of antidepressant-induced reduction of [oxy-Hb] values but also the effect of consistency of [oxy-Hb] activations as a trait marker of MDD. The current study provides additional evidence that the impact of antidepressants on NIRS signals should be taken into account when NIRS data are analyzed. Several limitations of this study should be noted. First, as we did not enroll healthy controls, we were unable to fully evaluate whether improvement of [oxy-Hb] activation after treatment in MDD returned to normal levels or not. Second, the number of participants was relatively small. Future research including both participants with MDD and healthy controls, and with a larger sample size, would overcome these limitations. However, the advantage of this study was the ability to perform NIRS scans with 4-week intervals, giving a more detailed time course of [oxy-Hb] alterations in the prefrontal and temporal regions in relation to antidepressant treatment. To our knowledge, this is the first longitudinal NIRS study to investigate the effects of sertraline treatment on neural activation with such a frequent observation interval. Additionally, we only enrolled participants who at baseline had never taken any antidepressant drugs and we only used sertraline for the treatment. Thus, this study excludes confounding effects of continuous antidepressant use and various types of drugs on prefrontal cortex function at baseline and allowed us to observe a specific effect of drug treatment response in MDD. In summary, the current study provides further evidence to support the potential usefulness of NIRS in capturing the details of brain functional alterations along with improvements of clinical symptoms in major psychiatric conditions. Even though depressive symptoms significantly improved after sertraline treatment, remitted depression showed persistent hypofrontality during VFT. Furthermore, our frequent measurements, every 4 weeks, identified atypical [oxy-Hb] timecourse alterations throughout the 12-week observation, revealing not only the effect of antidepressant-induced reduction of [oxy-Hb] values but also the effect of consistency of [oxy-Hb] activations as a trait marker of MDD. The correlation analyses also suggest that NIRS signals 4 weeks after the initiation of treatment hold potential as a biological marker to predict patient's clinical response to antidepressant treatment.
The authors are grateful to Mr. Shingo Kawasaki at Hitachi Co. for technical support. Declaration of interest None. Ethical statement This study was approved by the Institutional Review Board at the Keio University School of Medicine. Written informed consent was obtained from all participants prior to the study. Supplementary materials Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.jad.2019.04.001. References American Psychiatric Association, 2000. Diagnostic and statistical manual. Text Revision (DSM-IV-TR), 4th edn. American Psychiatric Association, Washington. Connolly, C.G., Ho, T.C., Blom, E.H., LeWinn, K.Z., Sacchet, M.D., Tymofiyeva, O., Simmons, A.N., Yang, T.T., 2017. Resting-state functional connectivity of the amygdala and longitudinal changes in depression severity in adolescent depression. J. Affect. Disord. 207, 86–94. Delaveau, P., Jabourian, M., Lemogne, C., Guionnet, S., Bergouignan, L., Fossati, P., 2011. Brain effects of antidepressants in major depression: a meta-analysis of emotional processing studies. J. Affect. Disord. 130, 66–74. Dichter, G.S., Gibbs, D., Smoski, M.J., 2015. A systematic review of relations between resting-state functional-MRI and treatment response in major depressive disorder. J. Affect. Disord. 172, 8–17. Diener, C., Kuehner, C., Brusniak, W., Ubl, B., Wessa, M., Flor, H., 2012. A meta-analysis of neurofunctional imaging studies of emotion and cognition in major depression. NeuroImage 61, 677–685. Fitzgerald, P.B., Laird, A.R., Maller, J., Daskalakis, Z.J., 2008. A meta-analytic study of changes in brain activation in depression. Hum. Brain Mapp. 29, 683–695. Frank, E., Prien, R.F., Jarrett, R.B., Keller, M.B., Kupfer, D.J., Lavori, P.W., Rush, A.J., Weissman, M.M., 1991. Conceptualization and rationale for consensus definitions of terms in major depressive disorder. Remission, recovery, relapse, and recurrence. Arch. Gen. Psychiatry 48, 851–855. Gramfort, A., Luessi, M., Larson, E., Engemann, D.A., Strohmeier, D., Brodbeck, C., Goj, R., Jas, M., Brooks, T., Parkkonen, L., Hamalainen, M., 2013. MEG and EEG data analysis with MNE-Python. Front. Neurosci. 7, 267. Hoshi, Y., Tamura, M., 1993. Dynamic changes in cerebral oxygenation in chemically induced seizures in rats: study by near-infrared spectrophotometry. Brain Res. 603, 215–221. Kameyama, M., Fukuda, M., Yamagishi, Y., Sato, T., Uehara, T., Ito, M., Suto, T., Mikuni, M., 2006. Frontal lobe function in bipolar disorder: a multichannel near-infrared spectroscopy study. NeuroImage 29, 172–184. Kanske, P., Heissler, J., Schonfelder, S., Wessa, M., 2012. Neural correlates of emotion regulation deficits in remitted depression: the influence of regulation strategy, habitual regulation use, and emotional valence. NeuroImage 61, 686–693. Kessler, R.C., Chiu, W.T., Demler, O., Merikangas, K.R., Walters, E.E., 2005. Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication. Arch. Gen. Psychiatry 62, 617–627. Kohmura, K., Iwamoto, K., Aleksic, B., Sasada, K., Kawano, N., Katayama, H., Noda, Y., Noda, A., Iidaka, T., Ozaki, N., 2013. Effects of sedative antidepressants on prefrontal cortex activity during verbal fluency task in healthy subjects: a near-infrared spectroscopy study. Psychopharmacology (Berl.) 226, 75–81. Maris, E., Oostenveld, R., 2007. Nonparametric statistical testing of EEG- and MEG-data. J. Neurosci. Methods 164, 177–190. Matsuo, K., Kato, N., Kato, T., 2002. Decreased cerebral haemodynamic response to cognitive and physiological tasks in mood disorders as shown by near-infrared spectroscopy. Psychol. Med. 32, 1029–1037. Obrig, H., Villringer, A., 2003. Beyond the visible–imaging the human brain with light. J. Cereb. Blood Flow Metab. 23, 1–18. Okada, G., Okamoto, Y., Yamashita, H., Ueda, K., Takami, H., Yamawaki, S., 2009. Attenuated prefrontal activation during a verbal fluency task in remitted major depression. Psychiatry Clin. Neurosci. 63, 423–425. Okamoto, M., Dan, H., Sakamoto, K., Takeo, K., Shimizu, K., Kohno, S., Oda, I., Isobe, S., Suzuki, T., Kohyama, K., Dan, I., 2004. Three-dimensional probabilistic anatomical cranio-cerebral correlation via the international 10–20 system oriented for transcranial functional brain mapping. NeuroImage 21, 99–111. Peran, P., Demonet, J.F., Cardebat, D., 2008. Paroxetine-induced modulation of cortical activity supporting language representations of action. Psychopharmacology (Berl.) 195, 487–496.
Disclosures The authors report no conflicts of interest in this work. Authors’ contributions BY, KY, SH, and MM designed and conducted this study. BY and YT analyzed the data. All authors interpreted the data and wrote the manuscript. All authors have approved the final article. Role of the funding source This work was partially supported by Pfizer (to M.M.) and JSPS KAKENHI (grant number JP 16K10224 to B.Y.). The funder of this study had no role in the study design, data collection and analysis, publication decision, or manuscript preparation. 30
Journal of Affective Disorders 252 (2019) 25–31
B. Yamagata, et al.
G., 2014. Neuroimaging-aided differential diagnosis of the depressive state. NeuroImage 85, 498–507 Pt 1. Tomioka, H., Yamagata, B., Kawasaki, S., Pu, S., Iwanami, A., Hirano, J., Nakagome, K., Mimura, M., 2015. A longitudinal functional neuroimaging study in medication-naive depression after antidepressant treatment. PLoS One 10, e0120828. Tsujii, T., Masuda, S., Yamamoto, E., Ohira, T., Akiyama, T., Takahashi, T., Watanabe, S., 2009. Effects of sedative and nonsedative antihistamines on prefrontal activity during verbal fluency task in young children: a near-infrared spectroscopy (NIRS) study. Psychopharmacology (Berl.) 207, 127–132. Tsujii, T., Yamamoto, E., Ohira, T., Saito, N., Watanabe, S., 2007. Effects of sedative and non-sedative H1 antagonists on cognitive tasks: behavioral and near-infrared spectroscopy (NIRS) examinations. Psychopharmacology (Berl.) 194, 83–91. Tsuzuki, D., Jurcak, V., Singh, A.K., Okamoto, M., Watanabe, E., Dan, I., 2007. Virtual spatial registration of stand-alone fNIRS data to MNI space. NeuroImage 34, 1506–1518. Usami, M., Iwadare, Y., Kodaira, M., Watanabe, K., Saito, K., 2014. Near infrared spectroscopy study of the frontopolar hemodynamic response and depressive mood in children with major depressive disorder: a pilot study. PLoS One 9, e86290. Villringer, A., Planck, J., Hock, C., Schleinkofer, L., Dirnagl, U., 1993. Near infrared spectroscopy (NIRS): a new tool to study hemodynamic changes during activation of brain function in human adults. Neurosci. Lett. 154, 101–104. Yamagata, B., Takei, Y., Itahashi, T., Pu, S., Hirano, J., Mimura, M., Iwanami, A., 2017. Aberrant spatial and temporal prefrontal activation patterns in medication-naive adults with ADHD. Front. Psychiatry 8, 274. Zhang, J., Wang, J., Wu, Q., Kuang, W., Huang, X., He, Y., Gong, Q., 2011. Disrupted brain connectivity networks in drug-naive, first-episode major depressive disorder. Biol. Psychiatry 70, 334–342. Zhang, W., Qu, Z., Wang, Y., Yao, C., Bai, X., Bian, S., Zhao, B., 2015. Near-infrared reflectance spectroscopy (NIRS) for rapid determination of ginsenoside Rg1 and Re in Chinese patent medicine Naosaitong pill. Spectrochim. Acta A Mol. Biomol. Spectrosc. 139, 184–188.
Pizzagalli, D.A., 2011. Frontocingulate dysfunction in depression: toward biomarkers of treatment response. Neuropsychopharmacology 36, 183–206. Posner, J., Hellerstein, D.J., Gat, I., Mechling, A., Klahr, K., Wang, Z., McGrath, P.J., Stewart, J.W., Peterson, B.S., 2013. Antidepressants normalize the default mode network in patients with dysthymia. JAMA Psychiatry 70, 373–382. Price, J.L., Drevets, W.C., 2012. Neural circuits underlying the pathophysiology of mood disorders. Trends Cogn. Sci. 16, 61–71. Pu, S., Matsumura, H., Yamada, T., Ikezawa, S., Mitani, H., Adachi, A., Nakagome, K., 2008. Reduced frontopolar activation during verbal fluency task associated with poor social functioning in late-onset major depression: multi-channel near-infrared spectroscopy study. Psychiatry Clin. Neurosci. 62, 728–737. Satomura, Y., Sakakibara, E., Takizawa, R., Koike, S., Nishimura, Y., Sakurada, H., Yamagishi, M., Shimojo, C., Kawasaki, S., Okada, N., Matsuoka, J., Kinoshita, A., Jinde, S., Kondo, S., Kasai, K., 2019. Severity-dependent and -independent brain regions of major depressive disorder: a long-term longitudinal near-infrared spectroscopy study. J. Affect. Disord. 243, 249–254. Smoski, M.J., Keng, S.L., Schiller, C.E., Minkel, J., Dichter, G.S., 2013. Neural mechanisms of cognitive reappraisal in remitted major depressive disorder. J. Affect. Disord. 151, 171–177. Strangman, G., Culver, J.P., Thompson, J.H., Boas, D.A., 2002. A quantitative comparison of simultaneous BOLD fMRI and NIRS recordings during functional brain activation. NeuroImage 17, 719–731. Suto, T., Fukuda, M., Ito, M., Uehara, T., Mikuni, M., 2004. Multichannel near-infrared spectroscopy in depression and schizophrenia: cognitive brain activation study. Biol. Psychiatry 55, 501–511. Takamiya, A., Hirano, J., Ebuchi, Y., Ogino, S., Shimegi, K., Emura, H., Yonemori, K., Shimazawa, A., Miura, G., Hyodo, A., Hyodo, S., Nagai, T., Funaki, M., Sugihara, M., Kita, M., Yamagata, B., Mimura, M., 2017. High-dose antidepressants affect nearinfrared spectroscopy signals: a retrospective study. NeuroImage 14, 648–655. Takizawa, R., Fukuda, M., Kawasaki, S., Kasai, K., Mimura, M., Pu, S., Noda, T., Niwa, S., Okazaki, Y., Joint Project for Psychiatric Application of Near-Infrared Spectroscopy,
31
Contents lists available at ScienceDirect
Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad
Research paper
Brain functional alterations observed 4-weekly in major depressive disorder following antidepressant treatment
T
⁎
Bun Yamagataa, , Kaori Yamanakaa, Yuichi Takeib, Shogo Hottaa, Jinichi Hiranoa, Hajime Tabuchia, Masaru Mimuraa a b
Department of Neuropsychiatry, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-Ku, Tokyo 160-8582, Japan Department of Psychiatry and Neuroscience, Gunma University Graduate School of Medicine, Maebashi, Gunma, Japan
A R T I C LE I N FO
A B S T R A C T
Keywords: Depression Near-infrared spectroscopy Antidepressant treatment Longitudinal study Neuroimaging Sertraline
Background: Major depressive disorder (MDD) is a heterogeneous condition. Identifying the brain responses to antidepressant treatment is of particular interest as these may represent potential neural networks related to treatment response, forming one aspect of the biological markers of MDD. Near-infrared spectroscopy (NIRS) is suitable for repeated measurements with short intervals because of its noninvasiveness, and can provide detailed time courses of functional alterations in prefrontal regions. Methods: We conducted a 12-week longitudinal study to explore prefrontal hemodynamic changes at 4-week intervals following sertraline treatment in 11 medication-naïve participants with MDD using 52-channel NIRS. Results: While all participants achieved remission after treatment, intra-class correlation coefficient of oxygenated hemoglobin [oxy-Hb] values throughout the 12-week observation was moderate at the spatially and temporally contiguous cluster located in the left inferior frontal and temporal gyri. There was a significant negative correlation between mean [oxy-Hb] values in the significant cluster at 4 weeks and changes in Hamilton Rating Scale for Depression total score from 4 to 8 weeks (r = −0.73, P = 0.011) and from 4 to 12 weeks (r = −0.63, P = 0.039). Limitations: Without healthy controls for comparison, we were unable to fully evaluate whether improvement of [oxy-Hb] activations after treatment in MDD reached normal levels or not. Conclusion: Our NIRS findings of detailed prefrontal hemodynamic alterations over short interval observations such as 4 weeks may have revealed potential trait marker for MDD and biological maker for predicting clinical response to sertraline treatment in MDD.
1. Introduction Major depressive disorder (MDD), a leading cause of disability worldwide, is characterized by affective, cognitive, and somatic symptoms (Kessler et al., 2005; Zhang et al., 2011). A number of neuroimaging studies have demonstrated alterations in widespread neural networks, including the medial prefrontal cortex and limbic, striatal, thalamic, and basal forebrain areas, in individuals with MDD (Diener et al., 2012; Pizzagalli, 2011; Price and Drevets, 2012). Despite the importance of understanding the etiology of MDD, there is still no established biological marker for its diagnosis and treatment. Given the
heterogeneity of this disorder, identifying brain alterations in response to antidepressant treatment is particularly important as this may indicate potential neural networks related to treatment response, as one aspect of the biological markers of MDD. When comparing pre- and post-treatment time-points, recent resting-state functional magnetic resonance imaging (rs-fMRI) studies examining responses to medication have consistently reported that functional connectivity (FC) between the prefrontal and limbic brain regions increases after treatment (Dichter et al., 2015). This indicates greater inhibitory control over neural circuits related to emotion processing and regulation (Dichter et al., 2015). One study found that
Abbreviations: MDD, major depressive disorder; rs-fMRI, resting-state functional magnetic resonance imaging; FC, functional connectivity; DMN, default mode network; NIRS, near-infrared spectroscopy; oxy-Hb, oxygenated hemoglobin; deoxy-Hb, deoxygenated hemoglobin; VFT, verbal fluency task; HC, healthy controls; HRS-D, Hamilton Rating Scale for Depression; total-Hb, total-hemoglobin; BA, Brodmann's area; BOLD, blood-oxygenation level-dependent; ANOVA, analysis of variance; ECT, electroconvulsive therapy ⁎ Corresponding author. E-mail address: [email protected] (B. Yamagata). https://doi.org/10.1016/j.jad.2019.04.001 Received 17 December 2018; Received in revised form 29 March 2019; Accepted 2 April 2019 Available online 03 April 2019 0165-0327/ © 2019 Elsevier B.V. All rights reserved.
Journal of Affective Disorders 252 (2019) 25–31
B. Yamagata, et al.
Table 1 Demographic data.
Age (years) Education (years) Sex (male/female) HRS-D total score VFT performance Dose of sertraline (mg/day)
Baseline
Week 4
Week 8
Week 12
F-value
P-value
14.5 ± 5.6 14.1 ± 6.0 0
8.5 ± 5.8 12.4 ± 3.6 29.5 ± 10.1
4.6 ± 4.1 13.6 ± 3.2 61.4 ± 20.5
4.6 ± 2.7 13.2 ± 3.6 65.9 ± 23.1
11.7 1.72 43.4
0.003 0.247 <0.0001
36.3 ± 11.2 15.4 ± 2.3 5/6
Abbreviations: HRS-D, Hamilton rating scale for depression; VFT, Verbal fluency task.
Hb] values alter after initiation of antidepressant treatment until achieving clinical improvement. We therefore performed a longitudinal study to explore frontal hemodynamic changes at 4-week intervals using 52-channel NIRS following treatment with the antidepressant sertraline in medication-naïve MDD participants, to reveal whether the hypofrontality observed in MDD alters between starting antidepressant treatment and as depressive symptoms improve.
connectivity of the posterior cingulate cortex, part of the default mode network (DMN), to the right lateral parietal cortex and right inferior temporal gyrus normalized after duloxetine therapy (Posner et al., 2013). Additionally, a recent study reported that baseline FC between the amygdala and insula was associated with a greater reduction in depressive symptoms 3 months later, suggesting that FC between these neural circuit potentially serves as a candidate biomarker for treatment response (Connolly et al., 2017). Although these normalized brain networks after treatment have been identified as biological markers of treatment response in MDD, prior studies have had a potential drawback: In most prior studies, individuals with MDD have been scanned at relativity long intervals such as 8 or 12 weeks, and detailed alterations in neural activities at shorter intervals have not been sufficiently investigated. For example, examining whether brain response at the beginning of therapy contributes to treatment reactivity is of particular interest. Additionally, MRI scanning has disadvantages, such as high cost, loud noise, and being in an enclosed space for a long period of time, so may not be suitable for repeated measurements at short intervals such as 2 or 4 weeks. Near-infrared spectroscopy (NIRS) is an optical neuroimaging technique that allows the noninvasive measurement of changes in the concentrations of oxygenated and deoxygenated hemoglobin ([oxy-Hb] and [deoxy-Hb], respectively), which reflect changes in regional cerebral blood volume (Hoshi and Tamura, 1993; Villringer et al., 1993). NIRS has the following advantages; (1) complete noninvasiveness of the measurement, facilitating repeated measurements, (2) higher temporal resolution (0.1 s time resolution) than fMRI, and (3) portability and compactness of the apparatus, enabling measurements under natural conditions with subjects sitting on a comfortable chair. We therefore hypothesized that NIRS would be more suitable for frequent measurements, such as 2- or 4-week intervals, to observe treatment responses in psychiatric conditions. Prior NIRS studies have consistently reported that [oxy-Hb] significantly decreases in the fronto-temporal brain regions during a verbal fluency task (VFT) in participants with MDD compared with healthy controls (HC) (Kameyama et al., 2006; Matsuo et al., 2002; Pu et al., 2008; Suto et al., 2004). A meta-analysis of NIRS studies further supports the previous findings of hypofrontality in MDD (Zhang et al., 2015). Such observations suggest that NIRS may be a useful tool for investigating neural activity with fine time resolution. However, it is still unclear whether NIRS signal changes during VFT are state- or trait-dependent since previous longitudinal studies have not obtained a consistent finding. For instance, a prior 12-week longitudinal study demonstrated that there were no [oxy-Hb] changes between pre- and post-antidepressant treatments in the MDD despite significant improvement in depressive symptoms, suggesting that hypofrontality response to a VFT may represent a potential trait marker for depression rather than a state marker (Tomioka et al., 2015). On the other hand, a recent study has reported that during 1.5-year follow-up, [oxy-Hb] changes in the right inferior frontal gyrus synchronized with changes of depressive symptoms, whereas [oxy-Hb] activation in the bilateral middle frontal gyrus was rather stable over time. This study suggested that NIRS would separate state- or trait-related brain regions in MDD, respectively (Satomura et al., 2019). Furthermore, there has been no study to measure repeatedly with a short interval how [oxy-
2. Methods 2.1. Participants This study included eleven drug-naïve adult individuals with MDD (Table 1). All participants were recruited from Keio University Hospital (Tokyo, Japan). After an extensive description of the study, written informed consent was obtained from all participants. The study protocol was prepared in accordance to the ethical standards of the Declaration of Helsinki and approved by the Ethics Committee of Keio University School of Medicine in February 27th, 2012 (registration number: UMIN000017269). We then started recruiting participants from March 1st, 2012 and completed collecting data in December 31st, 2016. The clinical diagnoses of depression were based on the DSM-IV-TR (American Psychiatric Association, 2000). Participants were confirmed to have had no lifetime history of bipolar disorder, psychosis, obsessivecompulsive disorder, or drug or alcohol misuse, or any neurological disorder. Furthermore, none of the participants reported any unstable medical condition or history of significant head trauma. Depression severity was evaluated using the 17-item Hamilton Rating Scale for Depression (HRS-D). We performed NIRS scans and clinical assessments at four time-points: before starting antidepressant treatment (baseline), and at 4, 8, and 12 weeks. The participants started treatment with 25 mg/day sertraline after the baseline scan. Partial remission was defined as ≤ 10 on the 17-item HRS-D based on DSM-IV criteria (American Psychiatric Association, 2000; Frank et al., 1991).
2.2. Verbal fluency task The task procedure in the current study was similar to that used by (Takizawa et al., 2014). Briefly, [oxy-Hb] changes were measured during the VFT (letter version), which is composed of a 30 s pre-task baseline, 60 s task, and a 60 s post-task baseline. For the pre- and posttask baseline periods, the participants were instructed to consecutively repeat the five Japanese vowels (“a”, “i”, “u”, “e”, “o”) aloud. During the task period, participants were requested to produce as many nouns as possible beginning with a designated syllable without the use of repetitions and proper nouns. The three sets of initial syllables (A;/to/, /se/, /o/, B; /a/, /ki/, /ha/, C; /na/, /i/, /ta/) were presented in counterbalanced order among the participants and each syllable changed every 20 s during the 60 s task.
26
Journal of Affective Disorders 252 (2019) 25–31
B. Yamagata, et al.
described in (Maris and Oostenveld, 2007). This method effectively controls the Type I-error rate with respect to multiple comparisons over channels and time-points because NIRS has 52 channels over the prefrontal and temporal regions and there are 1251 time-points during the whole experimental period (total duration is 125.1 s with 0.1 s time resolution). This is achieved by clustering neighboring channels and time-points. For the first-level statistics, channels and time-points over a threshold (F > 3.72) were identified from the repeated-measures ANOVA and spatially and temporally contiguous points in terms of channels and time-points over this threshold were subsequently defined as a cluster. Then the sum of the F values for a given cluster was used for the cluster-level statistics. By repeatedly resampling the data across all subjects, cluster-level F values were created from 10,000 randomization routines. For the second-level statistics, the P value was estimated according to the proportion of the randomization distribution exceeding the observed maximum cluster-level statistics (see details of methods in our previous study (Yamagata et al., 2017)). We used a “spatio_temporal_cluster_test” in MNE-python (Gramfort et al., 2013) for the cluster-based nonparametric randomization test of the NIRS data. We then extracted mean [oxy-Hb] data from the cluster showing the significant main effect of time-point and performed the Friedman test followed by the Wilcoxon Signed-rank Test with Bonferroni correction for a post-hoc analysis to examine the differences in [oxy-Hb] at the four time-points (baseline, 4, 8, and 12 weeks). In order to examine relationship between [oxy-Hb] values and clinical response to sertraline treatment, we first performed a correlation analysis between mean [oxy-Hb] values at baseline extracted from the spatially and temporally contiguous cluster mentioned above and changes of HRS-D total scores at any time-points. Second, in order to investigate whether changes of NIRS signals would represent state-dependent, we conducted the Spearman's rank correlation analysis between intra-individual longitudinal changes in [oxy-Hb] values at the spatially and temporally contiguous cluster ([oxy-Hb] at 12 weeks – [oxy-Hb] at baseline) and HRS-D total scores (HRS-D at 12 weeks – HRS-D at baseline). Additionally, to identify whether brain activation remains rather constant throughout the 12-week observation, we calculated intra-class correlations (ICCs, one-way random, single measure) between [oxy-Hb] values at baseline, 4, 8, and 12 weeks at the spatially and temporally contiguous cluster. Statistical significance was defined as P < 0.05 and all analyses were performed using SPSS (IBM SPSS Statistics, Version 25.0. Armonk, NY) except for the cluster-based nonparametric randomization test.
2.3. NIRS measurements The current study utilized a 52-channel NIRS machine (ETG-4000 Optical Topography System; Hitachi, Ltd., Japan) using two different wavelengths (695 nm and 830 nm) with 0.1 s time resolution to measure relative changes of absorbed near-infrared light. These changes are transformed into [oxy-Hb], [deoxy-Hb] and [total-hemoglobin (totalHb); sum of oxy-Hb and deoxy-Hb] concentration changes, as indicators of brain activity, by means of a modified Beer–Lambert law (Obrig and Villringer, 2003). The units were changes in chromophore concentration (mM) multiplied by the path length (mm) of the near-infrared light. We utilized 33 probes composed of 17 light emitters and 16 detectors with an inter-optrode distance of 30 mm. A measuring point of activation (channel) was defined as the region between one emitter and one detector. Thus, the probe set covered the brain regions including the bilateral prefrontal (approximately dorsolateral [Brodmann's area (BA) 9, 46], ventrolateral [BA 44, 45, 47], frontopolar [BA 10]) and superior and middle temporal cortical regions (BA 22, 41, 42). The correspondence between the probe positions and the measurement areas on the cerebral cortex was confirmed based on a previous multisubject study of anatomical craniocerebral correction via the international 10–20 system (Okamoto et al., 2004; Tsuzuki et al., 2007) (Fig. 1). The obtained data were analyzed using integral mode; the pre-task baseline was the mean of a 10 s period just prior to the task period, and the post-task baseline was the mean of the last 5 s of the post-task period; linear fitting was applied to the data between these two baselines. To remove short-term motion artifacts, we applied a moving average method using a window width of 5 s. We then calculated the median and standard deviation at each channel, rejecting data above or below 3 standard deviations of the median value. [oxy-Hb] represents regional neural activation and its change is assumed to more directly reflect cognitive activation than [deoxy-Hb] change, as shown by a stronger correlation with blood-oxygenation level-dependent (BOLD) signal measured by fMRI study (Strangman et al., 2002). 2.4. Data analysis Regarding the clinical and behavioral data, we compared total HRSD scores and total VFT numbers between four time-points including baseline, 4, 8, and 12 weeks using a repeated-measures analysis of variance (ANOVA). To examine differences in [oxy-Hb] between the four time-points, we first compared the time-course [oxy-Hb] changes during the whole baseline and VFT periods using a cluster-based nonparametric randomization test (P < 0.05 corrected) based on Monte-Carlo estimates, as
Fig. 1. Location of the 52 channels in the near-infrared spectroscopy. Corresponding anatomical site of each channel. Yellow circle represents “superior frontal gyrus”, green circle represents “Middle frontal gyrus”, light blue circle represents “inferior frontal gyrus”, blue circle represents “superior and middle temporal gyri”, red circle represents “precentral gyrus”, orange circle represents “postcentral gyrus”, and brown circle represents “supramarginal gyrus”. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) 27
Journal of Affective Disorders 252 (2019) 25–31
B. Yamagata, et al.
Fig. 2. Grand average waveforms of [oxy-Hb] changes during VFT in the fronto-temporal brain regions. A) The 52 measuring positions are labeled Ch1 to Ch52, from the right temporal to the left temporal regions. Orange colored regions represent a spatially and temporally contiguous cluster showing a significant main effect of time-point in [oxy-Hb] values during VFT in drug-naïve major depressive disorder (P < 0.05 corrected; a cluster-based nonparametric randomization test). B) Example of detailed time-course [oxy-Hb] changes with a 0.1 s time resolution between four time-points at Ch50 and Ch42. Purple line represents “baseline”, green line represents “4 weeks”, blue line represents “8 weeks”, and orange line represents “12 weeks”. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
3. Results
(P = 0.002). [oxy-Hb] at baseline significantly declined after starting antidepressant treatment until 4 weeks and then decreased [oxy-Hb] compared with baseline persisted until 12 weeks [baseline > 4 weeks (P < 0.001); baseline > 8 weeks (P = 0.003); baseline > 12 weeks (P = 0.021)]. Furthermore, [oxy-Hb] values at 4 weeks gradually increased, corresponding with clinical improvements but there were no significant differences between 4 and 8 weeks (P = 0.509), between 4 and 12 weeks (P = 0.186), and between 8 and 12 weeks (P = 0.509) (Fig. 3).
3.1. Demographic characteristics There were no significant differences between four time-points in VFT performance (F(3,8) = 1.73, P = 0.247), but depressive symptoms based on HRS-D scores significantly improved after starting sertraline treatment (F(3,8) = 11.67, P < 0.003) (Table 1). At 12 weeks, all participants showed at least partial remission based on DSM-IV criteria (American Psychiatric Association, 2000; Frank et al., 1991).
3.3. Relationship between [oxy-Hb] values and clinical response to sertraline treatment
3.2. Longitudinal comparison of [oxy-Hb] time-courses between baseline, 4, 8, and 12 weeks
There were no significant correlations between mean [oxy-Hb] values at baseline extracted from the spatially and temporally contiguous cluster and changes of HRS-D total scores at any time point. Furthermore, changes of mean [oxy-Hb] values at the cluster ([oxy-Hb] at 12 weeks – [oxy-Hb] at baseline) were not significantly correlated with changes of HRS-D total scores (HRS-D at 12 weeks –HRS-D at baseline). Given that our results showed that [oxy-Hb] values increased from 4 to 12 weeks, after an initial reduction from baseline to 4 weeks, we further investigated whether [oxy-Hb] values at 4 weeks would
Cluster-based nonparametric randomization testing identified a significant main effect of time-point in [oxy-Hb] activation during VFT at one spatially and temporally contiguous cluster located in the left inferior frontal gyrus, postcentral gyrus, and superior and middle temporal gyri (this cluster consists of Ch 31, 39, 41, 42, 50, 51, and 52; duration time = 17.7 s–62.5 s; P = 0.046 corrected) (Fig. 2). A post-hoc analysis of mean [oxy-Hb] values extracted from the contiguous cluster revealed a significant main effect of time-point 28
Journal of Affective Disorders 252 (2019) 25–31
B. Yamagata, et al.
Fig. 3. Comparison of [oxy-Hb] during VFT at baseline, week 4, 8, and 12. Each bar represents the mean [oxy-Hb] values extracted a spatially and temporally contiguous cluster showing a significant main effect of time-point in [oxy-Hb] values during VFT in drug-naïve major depressive disorder. Friedman test revealed significant differences in [oxy-Hb] at the four time-points. * P < 0.05, error bar represents standard error of the mean.
emotional networks compared with healthy controls, indicating that these brain dysfunctions may potentially be predominant biological makers for depression (Kanske et al., 2012; Okada et al., 2009; Smoski et al., 2013; Usami et al., 2014). These discrepancies among studies may be because of differences in brain imaging methodologies, activation paradigms including cognitive or emotional tasks, and methods of treatment (e.g. antidepressant or electroconvulsive therapy [ECT]). Additionally, it has been controversial as to whether NIRS signals represent state- or trait-dependent since only a few longitudinal studies have been conducted so far. A previous longitudinal NIRS study identified that hypoactivation during VFT in the bilateral prefrontal and temporal cortices persisted from baseline to 12 weeks despite improvement of depressive symptoms in MDD, suggesting that [oxy-Hb] changes are trait-dependent (Tomioka et al., 2015). On the other hand, Satomura et al. performed a long-term longitudinal study of 1.5 years and revealed that NIRS signals separately exhibit state- or trait-related brain sub-regions in MDD (Satomura et al., 2019). In the current study, [oxy-Hb] changes in the left inferior frontal and temporal regions did not synchronize with changes of depressive symptoms, whereas [oxyHb] activation in these regions remained consistent throughout the 12week observation. Therefore, we surmise that our findings of persistent functional abnormality in the left prefrontal and temporal cortices before and after antidepressant treatment may be a trait maker representing neurobiological vulnerability for depression rather than a marker specific to the patient's current state. Intriguingly, although ICC analysis demonstrated consistency of [oxy-Hb] activations in the left inferior frontal and temporal regions from baseline to 12 weeks, our frequent 4-weekly measurements identified atypical [oxy-Hb] time-course changes throughout the 12week observation. In particular, [oxy-Hb] significantly decreased from baseline to 4 weeks, and then increased from 4 to 12 weeks, correlating with clinical improvement. A previous fMRI study showed that activation in the left prefrontal cortex and right premotor cortex decreased during a linguistic task with paroxetine treatment (Peran et al., 2008). Other studies have also reported sedative effects of antidepressants or
predict treatment response. There was a significant negative correlation between mean [oxy-Hb] values in the significant cluster at 4 weeks and changes in HRS-D total score from 4 to 8 weeks (r = −0.73, P = 0.011) and from 4 to 12 weeks (r = −0.63, P = 0.039). These results may indicate that less [oxy-Hb] in the left inferior frontal and temporal regions at 4 weeks may relate to greater clinical improvements in depressive symptoms after sertraline treatment. 3.4. Consistency of [oxy-Hb] values from baseline to 12 weeks The significant ICC of [oxy-Hb] among baseline, 4, 8, and 12 weeks was demonstrated at the spatially and temporally contiguous cluster located in the left inferior frontal gyrus and temporal regions [ICC (1,1) = 0.468; P < 0.001]. 4. Discussion This study revealed that [oxy-Hb] during VFT in medication-naïve individuals with MDD at 12 weeks after starting sertraline treatment are significantly lower than at baseline in the left inferior frontal gyrus, postcentral gyrus, and superior and middle temporal gyri. This is despite significant improvement in depressive symptoms. In particular, during the 12-week treatment, [oxy-Hb] values initially reduced from baseline to 4 weeks and then gradually increased from 4 to 12 weeks. Furthermore, the intra-class correlation coefficient of [oxy−Hb] values between 4 measurement points was moderate in the left inferior frontal gyrus and temporal gyri (ICC interpretation scale; moderate [0.41 to 0.60]). Correlation analysis revealed that [oxy-Hb] values at 4 weeks may be a potential predictor of treatment response in MDD. Prior neuroimaging studies demonstrated reversal or normalization of abnormal brain function following antidepressant treatment in MDD (Delaveau et al., 2011; Fitzgerald et al., 2008). However, previous fMRI studies using VFT or emotional tasks have reported that even in the absence of current symptoms, individuals with remitted MDD showed reduced activity in brain networks related to the verbal fluency or 29
Journal of Affective Disorders 252 (2019) 25–31
B. Yamagata, et al.
Acknowledgment
antihistamines on NIRS signals (Kohmura et al., 2013; Tsujii et al., 2009, 2007). Additionally, our recent study revealed that high-doses of antidepressants are associated with attenuated [oxy-Hb] during a cognitive task (Takamiya et al., 2017). Based on these observations, we assumed that our findings of attenuated [oxy-Hb] values at 4 weeks after commencing sertraline therapy may be the consequence of introducing antidepressant treatment, rather than clinical responses. Moreover, given the consistency of [oxy-Hb] values over 12 weeks, it is speculated that the effects of artifacts due to repeated measurements on atypical [oxy-Hb] time-course changes are minimal. Taken together, our frequent 4-weekly measurements identified atypical [oxy-Hb] timecourse alterations throughout the 12-week observation, revealing not only the effect of antidepressant-induced reduction of [oxy-Hb] values but also the effect of consistency of [oxy-Hb] activations as a trait marker of MDD. The current study provides additional evidence that the impact of antidepressants on NIRS signals should be taken into account when NIRS data are analyzed. Several limitations of this study should be noted. First, as we did not enroll healthy controls, we were unable to fully evaluate whether improvement of [oxy-Hb] activation after treatment in MDD returned to normal levels or not. Second, the number of participants was relatively small. Future research including both participants with MDD and healthy controls, and with a larger sample size, would overcome these limitations. However, the advantage of this study was the ability to perform NIRS scans with 4-week intervals, giving a more detailed time course of [oxy-Hb] alterations in the prefrontal and temporal regions in relation to antidepressant treatment. To our knowledge, this is the first longitudinal NIRS study to investigate the effects of sertraline treatment on neural activation with such a frequent observation interval. Additionally, we only enrolled participants who at baseline had never taken any antidepressant drugs and we only used sertraline for the treatment. Thus, this study excludes confounding effects of continuous antidepressant use and various types of drugs on prefrontal cortex function at baseline and allowed us to observe a specific effect of drug treatment response in MDD. In summary, the current study provides further evidence to support the potential usefulness of NIRS in capturing the details of brain functional alterations along with improvements of clinical symptoms in major psychiatric conditions. Even though depressive symptoms significantly improved after sertraline treatment, remitted depression showed persistent hypofrontality during VFT. Furthermore, our frequent measurements, every 4 weeks, identified atypical [oxy-Hb] timecourse alterations throughout the 12-week observation, revealing not only the effect of antidepressant-induced reduction of [oxy-Hb] values but also the effect of consistency of [oxy-Hb] activations as a trait marker of MDD. The correlation analyses also suggest that NIRS signals 4 weeks after the initiation of treatment hold potential as a biological marker to predict patient's clinical response to antidepressant treatment.
The authors are grateful to Mr. Shingo Kawasaki at Hitachi Co. for technical support. Declaration of interest None. Ethical statement This study was approved by the Institutional Review Board at the Keio University School of Medicine. Written informed consent was obtained from all participants prior to the study. Supplementary materials Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.jad.2019.04.001. References American Psychiatric Association, 2000. Diagnostic and statistical manual. Text Revision (DSM-IV-TR), 4th edn. American Psychiatric Association, Washington. Connolly, C.G., Ho, T.C., Blom, E.H., LeWinn, K.Z., Sacchet, M.D., Tymofiyeva, O., Simmons, A.N., Yang, T.T., 2017. Resting-state functional connectivity of the amygdala and longitudinal changes in depression severity in adolescent depression. J. Affect. Disord. 207, 86–94. Delaveau, P., Jabourian, M., Lemogne, C., Guionnet, S., Bergouignan, L., Fossati, P., 2011. Brain effects of antidepressants in major depression: a meta-analysis of emotional processing studies. J. Affect. Disord. 130, 66–74. Dichter, G.S., Gibbs, D., Smoski, M.J., 2015. A systematic review of relations between resting-state functional-MRI and treatment response in major depressive disorder. J. Affect. Disord. 172, 8–17. Diener, C., Kuehner, C., Brusniak, W., Ubl, B., Wessa, M., Flor, H., 2012. A meta-analysis of neurofunctional imaging studies of emotion and cognition in major depression. NeuroImage 61, 677–685. Fitzgerald, P.B., Laird, A.R., Maller, J., Daskalakis, Z.J., 2008. A meta-analytic study of changes in brain activation in depression. Hum. Brain Mapp. 29, 683–695. Frank, E., Prien, R.F., Jarrett, R.B., Keller, M.B., Kupfer, D.J., Lavori, P.W., Rush, A.J., Weissman, M.M., 1991. Conceptualization and rationale for consensus definitions of terms in major depressive disorder. Remission, recovery, relapse, and recurrence. Arch. Gen. Psychiatry 48, 851–855. Gramfort, A., Luessi, M., Larson, E., Engemann, D.A., Strohmeier, D., Brodbeck, C., Goj, R., Jas, M., Brooks, T., Parkkonen, L., Hamalainen, M., 2013. MEG and EEG data analysis with MNE-Python. Front. Neurosci. 7, 267. Hoshi, Y., Tamura, M., 1993. Dynamic changes in cerebral oxygenation in chemically induced seizures in rats: study by near-infrared spectrophotometry. Brain Res. 603, 215–221. Kameyama, M., Fukuda, M., Yamagishi, Y., Sato, T., Uehara, T., Ito, M., Suto, T., Mikuni, M., 2006. Frontal lobe function in bipolar disorder: a multichannel near-infrared spectroscopy study. NeuroImage 29, 172–184. Kanske, P., Heissler, J., Schonfelder, S., Wessa, M., 2012. Neural correlates of emotion regulation deficits in remitted depression: the influence of regulation strategy, habitual regulation use, and emotional valence. NeuroImage 61, 686–693. Kessler, R.C., Chiu, W.T., Demler, O., Merikangas, K.R., Walters, E.E., 2005. Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication. Arch. Gen. Psychiatry 62, 617–627. Kohmura, K., Iwamoto, K., Aleksic, B., Sasada, K., Kawano, N., Katayama, H., Noda, Y., Noda, A., Iidaka, T., Ozaki, N., 2013. Effects of sedative antidepressants on prefrontal cortex activity during verbal fluency task in healthy subjects: a near-infrared spectroscopy study. Psychopharmacology (Berl.) 226, 75–81. Maris, E., Oostenveld, R., 2007. Nonparametric statistical testing of EEG- and MEG-data. J. Neurosci. Methods 164, 177–190. Matsuo, K., Kato, N., Kato, T., 2002. Decreased cerebral haemodynamic response to cognitive and physiological tasks in mood disorders as shown by near-infrared spectroscopy. Psychol. Med. 32, 1029–1037. Obrig, H., Villringer, A., 2003. Beyond the visible–imaging the human brain with light. J. Cereb. Blood Flow Metab. 23, 1–18. Okada, G., Okamoto, Y., Yamashita, H., Ueda, K., Takami, H., Yamawaki, S., 2009. Attenuated prefrontal activation during a verbal fluency task in remitted major depression. Psychiatry Clin. Neurosci. 63, 423–425. Okamoto, M., Dan, H., Sakamoto, K., Takeo, K., Shimizu, K., Kohno, S., Oda, I., Isobe, S., Suzuki, T., Kohyama, K., Dan, I., 2004. Three-dimensional probabilistic anatomical cranio-cerebral correlation via the international 10–20 system oriented for transcranial functional brain mapping. NeuroImage 21, 99–111. Peran, P., Demonet, J.F., Cardebat, D., 2008. Paroxetine-induced modulation of cortical activity supporting language representations of action. Psychopharmacology (Berl.) 195, 487–496.
Disclosures The authors report no conflicts of interest in this work. Authors’ contributions BY, KY, SH, and MM designed and conducted this study. BY and YT analyzed the data. All authors interpreted the data and wrote the manuscript. All authors have approved the final article. Role of the funding source This work was partially supported by Pfizer (to M.M.) and JSPS KAKENHI (grant number JP 16K10224 to B.Y.). The funder of this study had no role in the study design, data collection and analysis, publication decision, or manuscript preparation. 30
Journal of Affective Disorders 252 (2019) 25–31
B. Yamagata, et al.
G., 2014. Neuroimaging-aided differential diagnosis of the depressive state. NeuroImage 85, 498–507 Pt 1. Tomioka, H., Yamagata, B., Kawasaki, S., Pu, S., Iwanami, A., Hirano, J., Nakagome, K., Mimura, M., 2015. A longitudinal functional neuroimaging study in medication-naive depression after antidepressant treatment. PLoS One 10, e0120828. Tsujii, T., Masuda, S., Yamamoto, E., Ohira, T., Akiyama, T., Takahashi, T., Watanabe, S., 2009. Effects of sedative and nonsedative antihistamines on prefrontal activity during verbal fluency task in young children: a near-infrared spectroscopy (NIRS) study. Psychopharmacology (Berl.) 207, 127–132. Tsujii, T., Yamamoto, E., Ohira, T., Saito, N., Watanabe, S., 2007. Effects of sedative and non-sedative H1 antagonists on cognitive tasks: behavioral and near-infrared spectroscopy (NIRS) examinations. Psychopharmacology (Berl.) 194, 83–91. Tsuzuki, D., Jurcak, V., Singh, A.K., Okamoto, M., Watanabe, E., Dan, I., 2007. Virtual spatial registration of stand-alone fNIRS data to MNI space. NeuroImage 34, 1506–1518. Usami, M., Iwadare, Y., Kodaira, M., Watanabe, K., Saito, K., 2014. Near infrared spectroscopy study of the frontopolar hemodynamic response and depressive mood in children with major depressive disorder: a pilot study. PLoS One 9, e86290. Villringer, A., Planck, J., Hock, C., Schleinkofer, L., Dirnagl, U., 1993. Near infrared spectroscopy (NIRS): a new tool to study hemodynamic changes during activation of brain function in human adults. Neurosci. Lett. 154, 101–104. Yamagata, B., Takei, Y., Itahashi, T., Pu, S., Hirano, J., Mimura, M., Iwanami, A., 2017. Aberrant spatial and temporal prefrontal activation patterns in medication-naive adults with ADHD. Front. Psychiatry 8, 274. Zhang, J., Wang, J., Wu, Q., Kuang, W., Huang, X., He, Y., Gong, Q., 2011. Disrupted brain connectivity networks in drug-naive, first-episode major depressive disorder. Biol. Psychiatry 70, 334–342. Zhang, W., Qu, Z., Wang, Y., Yao, C., Bai, X., Bian, S., Zhao, B., 2015. Near-infrared reflectance spectroscopy (NIRS) for rapid determination of ginsenoside Rg1 and Re in Chinese patent medicine Naosaitong pill. Spectrochim. Acta A Mol. Biomol. Spectrosc. 139, 184–188.
Pizzagalli, D.A., 2011. Frontocingulate dysfunction in depression: toward biomarkers of treatment response. Neuropsychopharmacology 36, 183–206. Posner, J., Hellerstein, D.J., Gat, I., Mechling, A., Klahr, K., Wang, Z., McGrath, P.J., Stewart, J.W., Peterson, B.S., 2013. Antidepressants normalize the default mode network in patients with dysthymia. JAMA Psychiatry 70, 373–382. Price, J.L., Drevets, W.C., 2012. Neural circuits underlying the pathophysiology of mood disorders. Trends Cogn. Sci. 16, 61–71. Pu, S., Matsumura, H., Yamada, T., Ikezawa, S., Mitani, H., Adachi, A., Nakagome, K., 2008. Reduced frontopolar activation during verbal fluency task associated with poor social functioning in late-onset major depression: multi-channel near-infrared spectroscopy study. Psychiatry Clin. Neurosci. 62, 728–737. Satomura, Y., Sakakibara, E., Takizawa, R., Koike, S., Nishimura, Y., Sakurada, H., Yamagishi, M., Shimojo, C., Kawasaki, S., Okada, N., Matsuoka, J., Kinoshita, A., Jinde, S., Kondo, S., Kasai, K., 2019. Severity-dependent and -independent brain regions of major depressive disorder: a long-term longitudinal near-infrared spectroscopy study. J. Affect. Disord. 243, 249–254. Smoski, M.J., Keng, S.L., Schiller, C.E., Minkel, J., Dichter, G.S., 2013. Neural mechanisms of cognitive reappraisal in remitted major depressive disorder. J. Affect. Disord. 151, 171–177. Strangman, G., Culver, J.P., Thompson, J.H., Boas, D.A., 2002. A quantitative comparison of simultaneous BOLD fMRI and NIRS recordings during functional brain activation. NeuroImage 17, 719–731. Suto, T., Fukuda, M., Ito, M., Uehara, T., Mikuni, M., 2004. Multichannel near-infrared spectroscopy in depression and schizophrenia: cognitive brain activation study. Biol. Psychiatry 55, 501–511. Takamiya, A., Hirano, J., Ebuchi, Y., Ogino, S., Shimegi, K., Emura, H., Yonemori, K., Shimazawa, A., Miura, G., Hyodo, A., Hyodo, S., Nagai, T., Funaki, M., Sugihara, M., Kita, M., Yamagata, B., Mimura, M., 2017. High-dose antidepressants affect nearinfrared spectroscopy signals: a retrospective study. NeuroImage 14, 648–655. Takizawa, R., Fukuda, M., Kawasaki, S., Kasai, K., Mimura, M., Pu, S., Noda, T., Niwa, S., Okazaki, Y., Joint Project for Psychiatric Application of Near-Infrared Spectroscopy,
31