In 39 patients with major depressive disorder treated with six subanesthetic ketamine infusions, increased resting-state functional connectivity between the left amygdala and left medial superior frontal gyrus after treatment correlated with reduction in depressive symptoms, and baseline connectivity between the bilateral amygdalae and the right putamen predicted antidepressant and antisuicidal response (AUCs 0.739 and 0.827).
Papers cited by this study that are also in Blossom
Background
Dysfunction of the amygdala is the core pathogenesis of major depressive disorder (MDD). However, it remains unclear whether ketamine treatment could modulate characteristics of amygdala-related networks. We aimed to explore the relationship between changes in the resting-state functional connectivity (RSFC) of the amygdala and the treatment of ketamine in MDD patients and to identify important neuroimaging predictors of treatment outcome.
Methods
Thirty-nine MDD patients received six subanesthetic dose infusions of ketamine. Depressive and suicidal symptoms were assessed and magnetic resonance imaging (MRI) scans were performed before and after six ketamine infusions. Forty-five healthy controls also underwent once MRI scans. Seed-based RSFC analyses were performed, focusing on the bilateral amygdala.
Results
After ketamine treatment, the RSFC between the left amygdala (LA) and the left medial superior frontal gyrus (mSFG) of MDD patients enhanced significantly, and this change was positively correlated with the reduction in depressive symptoms (r = 0.40, p = 0.012). The combination baseline RSFC of LA – right putamen and right amygdala (RA) – right putamen was related to the antidepressant and antisuicidal effects of ketamine. The combination baseline RSFC of LA – right putamen and RA – right putamen could predict the ineffective antidepressant (AUC = 0.739, p = 0.011) and antisuicidal effects of ketamine (AUC = 0.827, p = 0.001).
Conclusion
Ketamine can regulate the relevant circuits of amygdala and mSFG, and the baseline RSFC between bilateral amygdala and right putamen may be a predictor of the response of ketamine’s antidepressant and antisuicidal treatment.Clinical trial registrationhttps://www.chictr.org.cn/showproj.aspx?proj=20875, identifier ChiCTR-OOC-17012239.
The data came from a clinical trial that explored the antidepressant effect of repeated ketamine infusions on MDD patients (clinical trial No. ChiCTR-OOC-17012239). Patients who met the following criteria (regardless of sex) were included in the study: (1) 18-65 years old; (2) diagnosed with MDD without psychotic features according to the structured clinical interview for Diagnostic and Statistical Manual of Mental Diseases-5 (DSM-5); (3) failed to respond to at least two suitable antidepressant agents with adequate dosage or exhibited suicidal ideation [confirmed by the Beck Scale for Suicidal Ideation (SSI) -Part I, with a score ≥2 at screening]; (4) total score of 17 items of Hamilton Depression Rating Scale (HAMD-17) ≥17; (5) no current pregnancy or lactation; (6) without serious medical or nervous system diseases and substance dependence; (7) no metal implants or other magnetic resonance imaging (MRI) contraindications; and (8) underwent resting-state functional MRI (rs-fMRI) scans at baseline and follow-up. Forty-five healthy controls (HCs) were recruited from the community. All HCs were in a healthy state without any previous or current mental illness or drug abuse or dependence. At the same time, there were no contraindications to MRI. All research procedures were conducted in accordance with the Declaration of Helsinki ethical principles and approved by the Review Committee of the Brain Hospital Affiliated to Guangzhou Medical University. All subjects had provided written informed consent before participation.
All MDD patients received a 40-min open-label infusion of ketamine (0.5 mg/kg) three times a week for 2 weeks. The details of our methods have been described in our previous studies. Hemodynamics and clinical status were monitored during this period. The MDD patients were not restricted from using psychiatric drugs throughout the study, but were only eligible to use if they took a stable dosage of antidepressants for at least 4 weeks before entering the study and continuing to receive the same regimen and dose throughout the study.MDD patients were assessed for clinical symptoms and underwent rs-fMRI scans 24 h prior to ketamine infusion (T0) and 24 h after the sixth ketamine infusion (T1). HCs underwent only one rs-fMRI scan.
Create a free account to open full-text PDFs.
Lally, N., Nugent, A. C., Luckenbaugh, D. A. et al. · Translational Psychiatry (2014)
Murrough, J. W., Perez, A. M., Pillemer, S. et al. · Biological Psychiatry (2012)
Wilkinson, S. T., Ballard, E. D., Bloch, M. H. et al. · American Journal of Psychiatry (2017)
Zheng, W., Zhou, Y-L., Liu, W. et al. · Journal of Psychiatric Research (2018)
Zheng, W., Zhou, Y-L., Wang, C-Y. · Journal of Psychopharmacology (2019)
Liu, W., Wang, C., Zhan, Y. et al. · Journal of Psychopharmacology (2018)
The severity of depressive symptoms was assessed with the HAMD-17; the higher their score on this scale, the more severe the depressive symptoms were. The SSI scale was used to assess suicidal ideation and which was consisted of 19 items administered by a clinician. We utilized the first five items of the SSI (SSI-5), which included hope to live, hope to die, reasons for living or dying, desire to actively attempt suicide, and passive suicidal thoughts. The higher score of SSI-5 indicated a higher risk of suicide. The interrater reliability of the clinicians who administered the HAMD-17 was assessed; the intraclass correlation coefficients were >0.9. The reduction rate ( HAMD-17% and SSI-5%) was used to index the antidepressant and antisuicide effect of ketamine. The reduction rate was calculated with the following equation: pretreatment score minus post-treatment score, then divided by the pre-treatment score, and finally multiplied by 100%. Responders were defined as having a reduction rate in HAMD-17 score or SSI-5 score ≥50%.
All imaging data were acquired by the Philips Achieva X-series 3T scanner with an eight-channel phased-array head coils. The gradientecho echo-planar imaging (GRE-EPI) sequence was used to acquire blood oxygen level dependent (BOLD) images with an echo time of 30 ms, repetition time of 2,000 ms, flip angle of 90 • , filed of view of 220 mm × 220 mm, matrix of 64 × 64, slice thickness of 4 mm, slice gap of 0.6 mm. The functional run lasted 8 min and 240 volumes acquired. Total of 33 transverse interleaved slices covered the whole brain. During the scan, the subjects were instructed to remain still with their eyes closed, avoid systematic thinking and remain awake.
All rsfMRI data was preprocessed using the Data Processing and Analysis of Brain Imaging (DPABI, version 5.0) 1 . The main steps were as follows: (1) removing the first 10 volumes; (2) slice timing; (3)realigning (subjects who had excessive head motion >2.0 mm translation and/or >2.0 • rotation during the scan were excluded); (3) spatially normalizing by using EPI templates; (4) smoothing using a 4-mm full width at half maxima (FWHM) isotropic Gaussian kernel; (5) removing linear and quadratic trends; (6) regressing out head motion effects using the Friston 24-parameter model, the white matter and cerebrospinal fluid. Moreover, the global signal was regressed in the whole brain analysis at the same time; (7) temporal band-pass filtering (0.01-0.1 Hz); and (8) "scrubbed" one time points before and one time points after bad images, whose frame displacement (FD) >0.5. 1
The bilateral amygdala was defined as seed-based two ROI according to the automated anatomical labels. Then, whole-brain voxel-level RSFC of each amygdala was mapped for all subjects. The main steps were as follows: (1) calculating the mean time series of each amygdala; (2) calculating the Pearson correlation coefficients between the mean time series of each amygdala and that of each voxel in the rest of the brain; and (3) converting each correlation coefficient z values using Fisher's z transformation to improve normality.
Independent-sample t-tests and χ 2 tests were conducted to determine whether the MDD and HC groups differed in age and sex with the Statistical Package for Social Sciences (SPSS 25.0). All significance levels were set at p < 0.05.
The DPABI (version 5.0) toolbox was used to obtain the RSFC from the bilateral amygdala to the whole-brain of HCs, pre-treatment and post-treatment MDD patients. The difference in the mean time course of the rs-fMRI scans before and after ketamine treatment was defined as the change in the RSFC of the amygdala. First, the statistical parametric mapping version 12 (SPM12), 2 was used to apply two sample t-test comparing the RSFC images between the HCs and pre-treatment MDD groups, controlling for age, sex, and head motion. If aberrant RSFC was found in MDD patients, the RSFC of the amygdala in this region was extracted for correlation analysis with the HAMD-17 score, SSI-5 score. Second, to investigate the relationship between the changes in abnormal RSFC of amygdala and the improvements in clinical symptoms after ketamine treatment, paired t-test was applied with SPM12 to compare the RSFC of pre-treatment and pro-treatment MDD patients, with head motion as a covariate. If the change in the RSFC of the amygdala was found in post-treatment patients, we would extract the above RSFC for correlation analysis with the reduction rate of HAMD-17 and SSI-5 scores. Third, the correlation analysis between abnormal amygdala RSFC at baseline and the reduction rate of depressive symptoms and suicidal symptoms after ketamine treatment were conducted, and ROC curves were performed to determine whether the abnormal RSFC of the amygdala at baseline could predict the treatment effect of ketamine. All resulting group-level analyses had a threshold of p < 0.05 [cluster-corrected using the familywise error rate (FWE) with a height threshold of p < 0.001]. All Images were displayed through the DPABI Viewer.4. Results
A total of 45 MDD patients were enrolled in the study. Of these, one exhibited serious artifacts in the rs-fMRI scan, and another five were excluded due to head movement during the scans. Therefore, 39 MDD patients were included in the final analysis. The two groups (HCs and MDD) exhibited a significant difference in age; specifically, the mean age of the MDD group was higher than that of the HC group (36.5 ± 12.1 vs. 31.4 ± 8.0, p = 0.031). The baseline characteristics were shown in Table.
For each of the two a priori-defined amygdala seeds, we assessed and compared its whole-brain RSFCs in MDD patients and HCs. Compared to HCs, MDD patients displayed hypoconnectivity between LA and bilateral putamen, right MCC and left insula. In addition, MDD patients showed hyperconnectivity between the LA and the bilateral posterior central gyrus (PCG) (Tableand Figure). Compared to HCs, the hypoconnectivity was showed between the RA and left cerebellum_6 as well as the RA and the right putamen in MDD patients. Moreover, MDD patients displayed hyperconnectivity between RA and left calcarine gyrus, left lingual, the left middle occipital gyrus (MOG) and left paracentral lobule (Tableand Figure).
At baseline, showed as Figure, HAMD-17 score was negatively correlated with RSFC in the LA and left insula (r = -0.44, p = 0.005). SSI-5 score was positively correlated with the abnormal RSFCs of LA and left PCG (r = 0.47, p = 0.002) as well as LA and right PCG (r = 0.35, p = 0.030). SSI-5 score was also negatively correlated with abnormal RSFC in the LA and right MCC (r = -0.41, p = 0.010), the LA and the right putamen (r = -0.35, p = 0.030). No abnormal RSFC was found in the RA associated with HAMD-17 and SSI-5 score.
After ketamine treatment, the HAMD-17 score significantly decreased by 11.0 points (23.2 ± 4.6 vs. 12.2 ± 7.4, p < 0.001), and the SSI-5 score significantly decreased by 3.3 points (9.0 ± 3.5 vs. 6.0 ± 2.3, p < 0.001). After ketamine treatment, the RSFC between the LA and the left medial superior frontal gyrus (mSFG) was strengthened (Figure). The altered RSFC between the LA and the left mSFG [cluster size = 43; Montreal Neurological Institute (MNI) coordinates: x = -9, y = 60, z = 15; pFWE-corr = 0.020), a brain region that exhibited significant changes after ketamine treatment, was positively correlated with the HAMD-17 reduction rate (r = 0.40, p = 0.012, Figure). But the altered RSFC between the LA and the left mSFG was not correlated with the SSI-5 reduction rate. There was no significant difference in RSFC of left LA -left mSFG between the pre-treatment MDD patients and HCs and between post-treatment MDD patients and HCs (Figure). The RSFC of the RA did not exhibit significant changes after treatment.
At baseline, the RSFC between the LA and the right putamen (r = -0.37, p = 0.021, Figure) and the RSFC between the RA and the right putamen (r = -0.40, p = 0.011, Figure) were negatively correlated with the reduction rate of HAMD-17 score, respectively. The RSFC between the LA and the right putamen (r = -0.53, p = 0.001, Figure), the RSFC between the RA and the right putamen (r = -0.38, p = 0.017, Figure) and the LA and right MCC (r = -0.38, p = 0.017, Figure), were negatively correlated with the reduction rate of SSI-5, respectively. For antidepressant response, ROC curves analysis was showed that the baseline amygdala RSFC was significant predictor, with the AUC of the LA -right putamen connectivity was 0.689 [95% confidence interval (CI), 0.504-0.875; p = 0.043], the AUC of the RA -right putamen connectivity was 0.726 (95% CI, 0.561-0.892;for details. For antisuicidal response, the AUC of the LA -right putamen connectivity was 0.799 (95% CI, 0.641 to 0.958; p = 0.003), and the AUC of the RA-right putamen connectivity was 0.787 (95% CI, 0.628 to 0.946; p = 0.005) and the AUC of combination of LA -right putamen connectivity and RA-right putamen connectivity was 0.827 (95% CI, 0.676 to 0.978; p = 0.001). See Figureand Tablefor details.
In this study, we found that after ketamine treatment, the RSFC between the LA and the left mSFG of MDD patients enhanced Previous studies reported that the RSFC of the amygdala in MDD patients differ from that of HCs, and our findings further validate this conclusion. Our study found that the strength of the RSFC between the amygdala and insula was associated with the severity of depression; the weaker the RSFC was, the more severe the depressive symptoms, consistent with previous study. In addition, we found that the strength of the RSFC between amygdala and putamen, MCC and PCG were related to suicide, and these regions associated with mood and the (A,B,D-F) Correlation between abnormal amygdala RSFC at baseline and the remission rate of depressive symptoms and suicidal symptoms after ketamine treatment and ROC curve analysis. (C) Neuroimaging predictors of treatment non-response, with the non-responders defined with <50% improvements in HAMD-17 obtained after the ketamine infusions. The area under the ROC curve was 0.689 (p = 0.043) for the RSFC of the LA -right putamen, with a sensitivity of 91.4%, and specificity of 59.1%; the area under the ROC curve was 0.726 (p = 0.016) for RA -right putamen connectivity, with a sensitivity of 82.4%, and specificity of 68.2%; and the area under the ROC curve was 0.739 (p = 0.011) for LA -right putamen connectivity and RA-right putamen connectivity, with a sensitivity of 75.0%, and specificity of 73.7%. (G) Non-responders defined with <50% improvements SSI-5 score obtained after the ketamine infusions. The area under the ROC curve was 0.799 (p = 0.003) for the RSFC of the LA -right putamen, with a sensitivity of 77.8%, and specificity of 83.3%; the area under the ROC curve was 0.787 (p = 0.005) for the RA -right putamen connectivity, with a sensitivity of 59.3%, and specificity of 91.7%; and the area under the ROC curve was 0.827 (p = 0.001) for LA -right putamen connectivity and RA -right putamen connectivity, with a sensitivity of 74.1%, and specificity of 83.3%. AUC, area under curve; LA, left amygdala; RA, right amygdala. risk of suicide. However, few studies have explored the relationship between the RSFC of the amygdala and suicidal ideation in MDD patients. One study reported that suicide was associated with the RSFC between the amygdala and the insula, middle temporal gyrus, and supraorbital frontal regions. Possible explanations for the lack of consistent findings were differences in the methods used to analyze the RSFC and the heterogeneity of subjects. In addition, the sample sizes of our study and the above studies was small; further investigations should increase the sample size. In the present study, we found strengthened RSFC between the LA and left mSFG after ketamine treatment; these changes were positively correlated with the improvements in depressive symptoms. The mSFG is part of the medial PFC (mPFC), which is normally extensively associated with the amygdala and exerts top-down inhibitory control over its activity, thereby regulating emotional expression. Hyperactivity of the amygdala in MDD patients has been suggested to result from reduced inhibitory input. Although our study did not find a significant difference in the RSFC between the amygdala and mPFC between MDD patients and HCs, previous studies of MDD patients have found weakened RSFC in the amygdala and PFC. A previous study found that the RSFC between the amygdala and PFC was strengthened after treatment with conventional antidepressants and that the RSFC of the LA was more significantly strengthened than that of the RA. The effect of ketamine may be related to mPFC activity. One study reported that the concentration of glutamate and glutamine complex in mPFC of ketamine responder began to rise rapidly when ketamine was injected. Another study showed that subanesthetic doses of ketamine cause a glutamatergic burst in the mPFC and enhanced mPFC activity, whereas silencing the mPFC blocks the effects of ketamine. Additionally, MDD patients exhibited reduced PFC volume, decreased neurotrophic factor release, and neuronal atrophy. Ketamine could promote the production and release of neurotrophic factors within the mPFC and restores its function. Therefore, the strengthened RSFC between the LA and left mSFG after ketamine treatment further indicated that the antidepressant effect of ketamine may be related to enhanced PFC activity to regulate amygdala function. Our study found that the RSFC between the amygdala and the right putamen may be a predictor of clinical response to ketamine treatment. MDD patients with greater decoupling of amygdala and right putamen may be the target population for ketamine treatment. The putamen is an important part of the striatum, a system that increases responses to negative emotions and is a core region of the reward network. Previous studies have demonstrated that loss of pleasure in MDD patients was associated with abnormal reward circuits. One study showed that glycolysis in the putamen of bipolar depressive patients with treatment-resistant was associated with reduced anhedonia and the anti-anhedonic effects of ketamine was due to ketamine-induced increases in glucose metabolism in the putamen and MCC. Moreover,found that female MDD patients had reduced amygdala and putamen volume and decreased RSFC between amygdala and putamen, which was similar to our results. In our study, we also found that the lower RSFC between amygdala and putamen was associated with stronger suicidal ideation, which may suggest that functional decoupling between amygdala and putamen may be related to the occurrence of suicidal ideation. Previous studies suggest that the putamen was also involved in suicidal ideation. Putamen volume was decreased in MDD patients with suicidal ideation, and that lower putamen activation during motor tasks was associated with higher suicidal ideation in adult BD patients. After ketamine treatment, putamen gray matter volume increased, and activity in brain regions such as the putamen and caudate nucleus increased. Our findings may further confirm that the antidepressant and antisuicide effects of ketamine may be related to the functional changes of reward related brain regions. In addition, our study found that RSFC between the amygdala and MCC was associated with improvement in suicidal ideation. MCC is also thought to be part of the reward circuit, involved in action reinforcement association and selection based on the rewarding or aversive nature of the underlying movement. Animal studies have found that this amygdala -MCC pathway is associated with the influence of the amygdala's facial processing subsystem on the perception of emotional facial expressions, such as fearful, sad, and happy expressions. Previous study has suggested that the sustained antidepressant effect of a lowdose ketamine infusion was mediated by increased activation of the supplementary motor area (SMA) and MCC; a single optimal dose of ketamine facilitated glutamatergic neurotransmission in the SMA and MCC, activating both regions, whereas a greater increase in the activation of MCC was associated with a reduction in depressive symptoms. Except for ketamine, other drug also seem to target the RSFC between the amygdala and MCC. One study showed that after 8-weeks of quetiapine treatment, decreased HAMD scores in anxiety depressive patients were associated with increased RSFC in the amygdala-MCC and amygdala-precuneus. Our results reveal a connection between the amygdala and MCC that may also be a target for ketamine treatment. This study had several limitations. First, the sample size was small. Since MDD patients commonly experience symptoms of anxiety, they are prone to irritability during scans, causing excessive head movement, which affects the acquisition of high-quality data. Second, due to the open-label design and the lack of placebo-treated controls, the influence of temporal effects on functional changes in the brain cannot be excluded. Third, RSFC describes the temporal correlation of blood signals between brain regions, that is, the synchronization of functional activity in related brain regions; it cannot determine anatomical directionality or causal relationships. Fourth, as all MDD patients were not restricted in terms of the use of psychiatric drugs throughout the study, we cannot rule out a group effect due to medication. Fifth, we focused on the amygdala and explored its RSFC with the whole brain, but the results indicated that the amygdala is not a single homogeneous structure and exhibits structural and functional subdivisions; thus, further exploration of the subregions of the amygdala is needed.
Short-term repeated use of ketamine may modulate the RSFC of the amygdala in MDD patients. The mechanism of improving depressive symptoms by ketamine may be related to it is regulation of the RSFC of amygdala. The combination baseline RSFC of bilateral amygdala and right putamen may be a predictor of the response of ketamine's antidepressant and antisuicidal treatment.