Biomarkers of Response to Treatment With Frontal Cortex Stimulation for Anxious Depression

Overview

This is an add-on study to an existing multimodal neuroimaging study in MDD by investigating the acute effects of DLPFC tDCS on threat vigilance in 24-44 patients with MDD, as part of an open-label treatment intervention study. Behavioral and neural measures of threat vigilance will be taken acutely and investigated as predictors of subsequent treatment response to a four-week, fourteen-session DLPFC tDCS intervention, using a novel home-tDCS protocol. The design takes advantage of an existing rich set of candidate baseline behavioral, neural and molecular measures from the existing neuroimaging study, which could be used to predict treatment response to tDCS and thereby, aid future patient selection for clinical trials.

BACKGROUND Importance: Over 50% of patients with major depressive disorder (MDD) do not respond to initial treatment and relapse is common. Poor treatment response is likely due to limited treatment options, the heterogeneous nature of MDD, its high comorbidity with other psychiatric conditions (particularly anxiety disorders) and the lack of data to support a targeted treatment approach. In addition, existing treatments can have negative side effects. Thus, there is a great need for novel, more targeted treatments. Transcranial direct current stimulation (tDCS) is a novel intervention that targets neural excitability and plasticity in regions implicated in the cognitive features of MDD and anxiety disorders. This proposal applies cognitive neuroscience approaches in a tDCS treatment study to establish biomarkers of response to inform treatment selection and improve patient outcomes. Cognitive neuropsychological features of mood and anxiety disorders: Depressed and anxious patients typically show negative biases in emotional perception and memory, and such biases are believed to play a fundamental role in the maintenance of emotional disorders. In terms of neural correlates, functional magnetic resonance imaging (fMRI) studies have confirmed hyperactive amygdala and/or hypoactive prefrontal activity in patients with MDD and anxiety disorders, indicating an imbalance of activity within this cortico-limbic circuit. There is evidence that treatment with antidepressant drugs (Paulus et al. 2005) can reduce amygdala hyperactivity and cognitive behavioral therapy can increase frontal activation (Ritchey et al. 2011). In addition, emerging fMRI findings indicate that pre-treatment neural markers can be used to predict whether patients will respond to behavioral or drug treatments (McGrath et al. 2014). Following the administration of a single dose of anxiolytic or antidepressant treatment, early changes in emotional processing have been observed in healthy people and clinical groups, in the absence of acute mood improvements. Critically, among patients, acute cognitive effects - such as a reduction in vigilance to threat (e.g. fearful faces) - have been shown to predict response to drug and behavioral treatments (Tranter et al. 2009; Reinecke et al. 2013). Thus, baseline and acute behavioral and neural markers could be leveraged to identify characteristics of likely responders to different treatments. Specifically, treatments aiming to remediate prefrontal and amygdala dysfunction (underlying negative biases) could be a critical target in patients with MDD and anxiety disorders exhibiting these deficits. Transcranial direct current stimulation: Mounting evidence from clinical trials indicates that repeated administration (10-15 sessions over 2-4 weeks) of tDCS to the dorsolateral prefrontal cortex (DLPFC) is a potentially effective treatment for MDD (Brunoni et al. 2015). However, underlying mechanisms of action are unclear, although spectroscopy imaging suggests that tDCS causes alterations in inhibitory neurotransmitter GABA (Stagg \& Nitsche 2011), and a recent review (Ironside \& Perlo 2018) implicated protective effects on cognition as a potential mechanism of action for DLPFC tDCS. An initial investigation in healthy volunteers by the applicant revealed an anxiolytic-like effect (reduced threat vigilance) from a single session of DLPFC tDCS vs sham tDCS on a behavioral test of proven clinical relevance (Ironside et al. 2016). To investigate the neural correlates of this effect, the applicant followed up with an fMRI study which found that, in a sample of trait anxious females, a single session of DLPFC vs. sham tDCS reduced amygdala response to fearful faces (p \< 0.05) whilst simultaneously increasing frontal attentional control signals (p \< 0.001) (Ironside et al. 2019). This provides causal evidence that modulating activity directly in the DLPFC inhibits amygdala response to threat in humans, highlighting a potential neural mechanism for the prior behavioral reduction in vigilance. In addition, this offers initial mechanistic insights into the efficacy tDCS in the treatment of MDD and anxiety disorders. Candidate mechanisms of action for tDCS: These acute cognitive neuropsychological effects of tDCS can mirror acute effects of antidepressant and anxiolytic treatment. This provides a rationale for patient investigations of tDCS to characterize these as potential markers of treatment response. The current proposal seeks to apply - the investigators believe for the first time - these findings to patients receiving tDCS treatment for MDD, using behavioral, neural and molecular measures to test the predictive validity of acute effects of tDCS on behavioral threat vigilance (Aim 1), amygdala response to fear (Aim 2) and resting brain activation (Aim 3) as markers of future treatment response. Additionally, evidence from magnetic resonance spectroscopy (MRS) indicates that a change in the inhibitory neurotransmitter GABA is implicated in the effects of tDCS (Stagg et al. 2009) and thus exploratory analyses will probe the role of baseline GABA in treatment response. SPECIFIC AIMS AIM 1: Evaluate behavioral vigilance to threat as a biomarker of treatment response to tDCS in MDD: Hypothesis 1: The significant acute reduction in fear vigilance from a single session of tDCS is similar to that seen with anxiolytic treatments in the same cognitive paradigm, which, for anxiolytic treatment, was predictive of treatment response (Reinecke et al. 2013). This is hypothesized as a potential mechanism of action for the clinical effects of tDCS in MDD. Specifically, it is expected that, compared to non-responders, subsequent responders to tDCS treatment will show reduced threat vigilance following acute tDCS administration. AIM 2: Evaluate amygdala response to fear as a biomarker of treatment response to tDCS in MDD: Hypothesis 2: Previous findings show that frontal tDCS can reduce amygdalar threat reactivity in high trait anxious females. The investigators hypothesize that this reduction in amygdala threat reactivity is predictive of treatment response. Specifically, it is expected that, compared to non-responders, subsequent responders to tDCS treatment will show decreased amygdala response to fearful faces, following acute tDCS administration. AIM 3: Evaluate functional connectivity changes as a biomarker of treatment response to tDCS in MDD: Hypothesis 3: The reduction in amygdala threat reactivity by tDCS is caused by increased connectivity between frontal attentional control networks and the amygdala. Specifically, it is expected that tDCS will increase resting state functional connectivity between frontal attentional control networks and the amygdala in responders. EXPLORATORY AIM: To evaluate baseline measures of DLPFC GABA as a predictor of treatment response. STUDY DESIGN Participants: Forty-four unmedicated adults with MDD will be offered an optional add-on study to an existing large NIH-funded patient study. In current funding applications, the investigators are requesting funding for 20 patients (existing fellowship funding is used for 24 pilot participants). Interventions: After the multimodal imaging protocol (MRS and fMRI, see Day 2; Table 1) from the existing study is complete, participants will receive 14 sessions of bilateral prefrontal bipolar tDCS (2mA for 20 mins/session) over four weeks (Day 3 - Day 33). TDCS is a non-invasive neuromodulatory technique that uses weak electrical current to modify cortical excitability and neural plasticity. Sessions 3-14 will be self-administered, in the participants' homes, using a validated remote supervised protocol and specialized home use device (Soterix Medical, New York, 2018). Tasks: Participants will carry out computerized tasks, including the validated dot-probe measurement of behavioral vigilance to threat (Reinecke et al. 2013; Ironside et al. 2016) and validated fMRI scanner-based attentional control task with fearful distractor faces (Bishop et al. 2007; Ironside et al. 2019). Crucially, these tasks will measure acute effects of tDCS after only 1-2 sessions, before any treatment effects emerge. Measurements: MRS and fMRI data will be acquired on a 3 Tesla (3T) Prisma scanner using a 64-channel head coil. Imaging measurements are taken at baseline (existing study) and after the first tDCS session. The Montgomery-Asberg Depression Rating Scale (MADRS), Quick Inventory of Depressive Symptomology (QIDS), State-Trait Anxiety Inventory-State (STAI-S), and Smith-Hamilton Pleasure Scale (SHAPS) will be administered weekly to assess depression, anxiety, and anhedonia. Data analysis: Treatment response will be established as \>50% reduction in MADRS scores following all treatment and baseline/ acute measures will be tested for their power to predict this response using regression analyses in R. FMRI data (task-based and resting) will be pre-processed and analyzed using SPM12 and CONN. RELEVANCE OF THE PROPOSED RESEARCH: MDD is one of the leading causes of disease burden worldwide, characterized by treatment resistance and high relapse. Clinical investigations of tDCS as a treatment for MDD show promise but the mechanisms of action remain unclear. Therefore, an experimental medicine model is needed to establish the behavioral, neural, and molecular characteristics of responders to tDCS. Preliminary evidence suggests that tDCS reduces behavioral/neural vigilance to threat but this has not yet been linked to treatment response in MDD. Through an innovative approach taking advantage of an existing clinical sample and baseline behavioral, neural and molecular measurements from an ongoing patient investigation, the proposed study is expected to provide a better understanding of the mechanisms of action of tDCS in the treatment of MDD. In addition, any potential biomarkers identified could be used for patient selection and stratification in larger clinical trials and ultimately - after replications - in treatment selection in the clinic. Finally, the employment of an innovative, home use tDCS protocol including remote supervision and patient training/ assessment progresses feasibility of this novel treatment, furthering the case for translation.