If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
1 Present address: School of Health and Care Management, Arden University Berlin, 10963 Berlin, Germany and Charité-Universitätsmedizin Berlin, Corporate member of Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, 10117, Berlin, Germany.
1 Present address: School of Health and Care Management, Arden University Berlin, 10963 Berlin, Germany and Charité-Universitätsmedizin Berlin, Corporate member of Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, 10117, Berlin, Germany. 2 Co-first authors.
Affiliations
Departments of Psychiatry and Clinical Neuroscience, Hotchkiss Brain Institute and Mathison Centre, University of Calgary, Canada
2 Co-first authors. 1 Present address: School of Health and Care Management, Arden University Berlin, 10963 Berlin, Germany and Charité-Universitätsmedizin Berlin, Corporate member of Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, 10117, Berlin, Germany.
Deep brain stimulation (DBS) has shown promising therapeutic results for treatment resistant depression (TRD). The subcallosal cingulate gyrus (SCG), having extensive connections to limbic and cortical areas, is a common target [
]. Reasons for this variable effectiveness may involve underlying biomarkers related to mechanisms of action. For example, work using animal models suggests that SCG-DBS in TRD may have similar mechanisms of action as selective serotonin reuptake inhibitor (SSRI) antidepressants [
]. Stimulation of SCG homologous regions in rodents significantly increased release of serotonin (5-hydroxytryptamine, 5-HT) in anatomically connected areas and 5-HT depletion eliminated benefits of DBS in the model [
]. If this is similar in humans, then baseline 5-HT level in TRD patients and changes in 5-HT level during stimulation may be a biomarker of treatment outcome for SCG-DBS.
We investigated the predictive and therapeutic role of the serotonergic system in SCG-DBS outcomes for TRD patients using the loudness dependence of auditory evoked potentials (LDAEP). The LDAEP is an index of global serotonergic activity and a non-invasive measure of in vivo serotonin in animals and humans [
]. It is a component of the evoked response to increases in stimulus intensity as measured by change in amplitude of N1/P2, and is inversely related to central serotonergic activity, with a high LDAEP reflecting weak serotonergic neurotransmission and vice versa [
Comparison between the analysis of the loudness dependency of the auditory N1/P2 component with LORETA and dipole source analysis in the prediction of treatment response to the selective serotonin reuptake inhibitor citalopram in major depression.
]. In animal studies, microinjection into the dorsal raphe nucleus or systemic administration of a serotonin agonist or antagonist led to a decrease or increase, respectively, in the intensity dependence of the auditory evoked potential. Previous studies showed that this can be used as a predictor of response to selective serotonin reuptake inhibitors (SSRIs) for patients with major depressive disorder [
The loudness dependence of the auditory evoked potential (LDAEP) in schizophrenia, bipolar disorder, major depressive disorder, anxiety disorder, and healthy controls.
Twelve individuals who received DBS for TRD participated in baseline preoperative EEG recording, with six completing a 6-month EEG session, and seven completing a 12-month EEG session (Supplementary Table 1). Detailed description of participants is reported in our clinical trial2. EEG was recorded using a Brain Vision antiCHamp device (Brain Products GmbH, Gilching, Germany) and a 64-channel (Cz reference) EasyCap with 10/20 electrode montage. Participants were sitting on a comfortable chair and electrode impedances were kept under 17 kΩ. For the LDAEP paradigm, auditory stimuli were presented binaurally through headphones during EEG recording. Pure tones (1000 Hz, 30 ms duration, 10 ms rise/fall time and ISI jittered from 1200 to 1800 ms) of five different intensities (60, 70, 80, 90, and 100 dB SPL; 100 stimuli per intensity) were presented pseudo-randomly [
The loudness dependence of the auditory evoked potential (LDAEP) in schizophrenia, bipolar disorder, major depressive disorder, anxiety disorder, and healthy controls.
Data were analyzed offline using ERPLab (https://erpinfo.org/erplab). Signals were bandpass filtered at 0.1–30Hz. Independent component analysis (fastICA, https://erpinfo.org/erplab) was used to remove eyeblink and eye movement artifacts. Data were segmented into 100–500 ms epochs. Epochs were excluded if they exceeded a peak-to-peak threshold±75μV in a moving window of 200 ms. FCz was selected for analyses as the central midline point shows maximal amplitude for N1/P2 components of the LDAEP [
The loudness dependence of the auditory evoked potential (LDAEP) in schizophrenia, bipolar disorder, major depressive disorder, anxiety disorder, and healthy controls.
]. For the N1, we extracted the amplitude of the most negative peak between 50 and 150 ms, and for P2, the most positive peak between 150 and 250 ms. Then, we calculated the peak-to-peak N1/P2 amplitudes for five different intensities (i.e., 60-, 70-, 80-, 90-, and 100-dB SPL) and LDAEP slopes were achieved as the slope of linear regression fit among them [
The loudness dependence of the auditory evoked potential (LDAEP) in schizophrenia, bipolar disorder, major depressive disorder, anxiety disorder, and healthy controls.
To test our hypothesis that the LDAEP slope at baseline could predict treatment outcome, we performed a correlation analysis between LDAEP slopes at baseline and 6- and 12-month percentage change in Hamilton Rating Scale for Depression 17 (HRSD) scores. Then, we compared the N1/P2 LDAEP component in remitters (defined as those achieving HRSD<7) and non-remitters using a non-parametric permutation t-test [
] with 10000 random shuffles, an appropriate approach to evaluate differences between groups with small sample size. We also performed a correlation analysis between the percentage change in HRSD and LDAEP slopes at 6- and 12-months.
Correlation analysis for N1/P2 LDAEP (Fig. 1-A) showed significant negative relationships between the LDAEP slope at baseline, and baseline to 6-month percent change in HRSD (r = −0.625, p = 0.030) (Fig. 1-A1), and between baseline LDAEP slope, and baseline to 12-month percent change in HRSD (r = −0.704, p = 0.011) (Fig. 1-A2). A shallower LDAEP slope was associated with greater clinical response to DBS. The permutation t-test showed that non-remitters at 12-months exhibited a significantly steeper LDAEP at baseline compared to remitters (t(2,10) = -2.1055, p = 0.048) (Fig. 1-B). At 12-months, significant correlation between changes in HRDS and LDAEP slope was observed (r = −0.927, p = 0.003) (Fig. 1-C2) but no correlation was seen at 6-months (Fig. 1-C1). Moreover, one participant (7024) with the highest LDAEP slope at baseline together with another showing greatest change in slope over time (7027) exhibited the worst result with DBS (Fig. 1-C).
Fig. 1Relationship between LDAEP and outcomes. Each participant is identified by a specific color. The black line in each plot shows the linear fit across points. A) Scatter plot of individual results based on HRSD at different assessments and LDAEP slope at baseline. Both A1 and A2 show significant correlations between percentage change in HRSD (6 months-baseline, A1) and baseline LDAEP slope, as well as HRSD (12 months-baseline, A2) and baseline LDAEP slope. B) Slope of LDAEP in remitters and non-remitters at 12 months after surgery. Non-parametric permutation t-test showed significant differences between two groups. C) There was no correlation between percent change in HRSD from baseline to 6 months and LDAEP slope (C1); however, there was a significant relationship between change in HRSD from baseline to 12 months and LDAEP slope. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
These preliminary results suggest that LDAEP can be used as a biomarker to predict SCG-DBS outcome in TRD patients and confirms our hypothesis that individual differences in central 5-HT activity is correlated with DBS outcomes. Participants who exhibited shallower LDAEP at baseline showed better response to DBS and continuing shallow LDAEP slope over time correlated with reduction in depression scores. Since central 5-HT activity is correlated with shallowness of the LDAEP slope, it suggests that TRD patients with relatively higher baseline 5-HT activity were likely to improve with SCG-DBS. Additionally, our findings are consistent with animal studies, confirming a role for serotonergic neurotransmission in the therapeutic mechanism of SCG-DBS clinical outcome [
Interestingly, our results on prediction are contrary to studies where responders to SSRIs had steep LDAEP at baseline, meaning lower central 5-HT activity [
Comparison between the analysis of the loudness dependency of the auditory N1/P2 component with LORETA and dipole source analysis in the prediction of treatment response to the selective serotonin reuptake inhibitor citalopram in major depression.
]. These discrepant findings may relate to different patient populations (TRD vs. major depression) and neurochemical mechanisms of DBS and SSRIs on 5-HT release or reuptake. DBS patients were exposed and concurrently treated with multiple antidepressants. Although both DBS and SSRIs significantly increase 5-HT release, they may induce different alterations in 5-HT1A and 5-HT1B receptor expression [
] expression. Beyond these opposite effects of DBS and SSRIs on SERT, DBS may change ionic currents and presynaptic physiology to cause more complex synaptic alterations compared to SSRIs [
]. There are no comparable studies of long-term DBS effects on 5-HT in humans; therefore, we can only speculate that the considerable increase of LDAEP slope (decrease of 5-HT activity) over time in two participants with worst outcomes may be related to complex synaptic alterations of SCG-DBS.
Overall, despite small number of patients, our results suggest that EEG (an accessible, non-invasive and inexpensive technology) could predict SCG-DBS outcomes. This approach can easily be tested in future studies with larger sample sizes.
Author contributions
ECB designed the study with input from DLC, ABP, ZHTK and RR. ECB and DLC performed the data collection. AG and ECB analyzed the data and wrote the manuscript. RR did the clinical assessments. AG made the figures and tables. All authors interpreted the data, reviewed the final manuscript, and provided comments.
Declaration of competing interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Fundings provided by Alberta Innovates Health Solutions (AIHS) and Natural Sciences and Engineering Research Council of Canada (NSERC) have been received by Dr. Kiss (04126-2017) and Dr. Protzner (05299-2020). Dr. Kiss was an AIHS clinical scholar; Drs. Brown and Clark were both post-doctoral fellows with AIHS and received additional funding from NSERC-CREATE and the Mathison Centre. Dr. Ghaderi was funded by an Eyes High Postdoctoral Award from the University of Calgary. Dr. Ramasubbu has received honorarium for serving in the advisory committee of Astra Zeneca, Lundbeck, Janssen, and Otsuka. He also received an investigator-initiated grant from Astra Zeneca and Pfizer.
These data were presented as a poster at the 74th Society of Biological Psychiatry annual meeting in May 2019. All authors report no potential conflicts of interest.
Appendix A. Supplementary data
The following is the Supplementary data to this article:
Comparison between the analysis of the loudness dependency of the auditory N1/P2 component with LORETA and dipole source analysis in the prediction of treatment response to the selective serotonin reuptake inhibitor citalopram in major depression.
The loudness dependence of the auditory evoked potential (LDAEP) in schizophrenia, bipolar disorder, major depressive disorder, anxiety disorder, and healthy controls.
00844-5&id=gr1.jpg){kind=link}