Lack of evidence for predictive utility from resting state fMRI data for individual exposure-based cognitive behavioral therapy outcomes: A machine learning study in two large multi-site samples in anxiety disorders.

Data-based predictions of individual Cognitive Behavioral Therapy (CBT) treatment response are a fundamental step towards precision medicine. Past studies demonstrated only moderate prediction accuracy (i.e. ability to discriminate between responders and non-responders of a given treatment) when using clinical routine data such as demographic and questionnaire data, while neuroimaging data achieved superior prediction accuracy. However, these studies may be considerably biased due to very limited sample sizes and bias-prone methodology. Adequately powered and cross-validated samples are a prerequisite to evaluate predictive performance and to identify the most promising predictors. We therefore analyzed resting state functional magnet resonance imaging (rs-fMRI) data from two large clinical trials to test whether functional neuroimaging data continues to provide good prediction accuracy in much larger samples. Data came from two distinct German multicenter studies on exposure-based CBT for anxiety disorders, the Protect-AD and SpiderVR studies. We separately and independently preprocessed baseline rs-fMRI data from n = 220 patients (Protect-AD) and n = 190 patients (SpiderVR) and extracted a variety of features, including ROI-to-ROI and edge-functional connectivity, sliding-windows, and graph measures. Including these features in sophisticated machine learning pipelines, we found that predictions of individual outcomes never significantly differed from chance level, even when conducting a range of exploratory post-hoc analyses. Moreover, resting state data never provided prediction accuracy beyond the sociodemographic and clinical data. The analyses were independent of each other in terms of selecting methods to process resting state data for prediction input as well as in the used parameters of the machine learning pipelines, corroborating the external validity of the results. These similar findings in two independent studies, analyzed separately, urge caution regarding the interpretation of promising prediction results based on neuroimaging data from small samples and emphasizes that some of the prediction accuracies from previous studies may result from overestimation due to homogeneous data and weak cross-validation schemes. The promise of resting-state neuroimaging data to play an important role in the prediction of CBT treatment outcomes in patients with anxiety disorders remains yet to be delivered.

Association between resting-state connectivity patterns in the defensive system network and treatment response in spider phobia-a replication approach.

Although highly effective on average, exposure-based treatments do not work equally well for all patients with anxiety disorders. The identification of pre-treatment response-predicting patient characteristics may enable patient stratification. Preliminary research highlights the relevance of inhibitory fronto-limbic networks as such. We aimed to identify pre-treatment neural signatures differing between exposure treatment responders and non-responders in spider phobia and to validate results through rigorous replication. Data of a bi-centric intervention study comprised clinical phenotyping and pre-treatment resting-state functional connectivity (rsFC) data of n = 79 patients with spider phobia (discovery sample) and n = 69 patients (replication sample). RsFC data analyses were accomplished using the Matlab-based CONN-toolbox with harmonized analyses protocols at both sites. Treatment response was defined by a reduction of >30% symptom severity from pre- to post-treatment (Spider Phobia Questionnaire Score, primary outcome). Secondary outcome was defined by a reduction of >50% in a Behavioral Avoidance Test (BAT). Mean within-session fear reduction functioned as a process measure for exposure. Compared to non-responders and pre-treatment, results in the discovery sample seemed to indicate that responders exhibited stronger negative connectivity between frontal and limbic structures and were characterized by heightened connectivity between the amygdala and ventral visual pathway regions. Patients exhibiting high within-session fear reduction showed stronger excitatory connectivity within the prefrontal cortex than patients with low within-session fear reduction. Whereas these results could be replicated by another team using the same data (cross-team replication), cross-site replication of the discovery sample findings in the independent replication sample was unsuccessful. Results seem to support negative fronto-limbic connectivity as promising ingredient to enhance response rates in specific phobia but lack sufficient replication. Further research is needed to obtain a valid basis for clinical decision-making and the development of individually tailored treatment options. Notably, future studies should regularly include replication approaches in their protocols.

Mechanisms of action underlying virtual reality exposure treatment in spider phobia: Pivotal role of within-session fear reduction.

Although virtual-reality exposure treatment (VRET) for anxiety disorders is an efficient treatment option for specific phobia, mechanisms of action for immediate and sustained treatment response need to be elucidated. Towards this aim, core therapy process variables were assessed as predictors for short- and long-term VR treatment outcomes. In a bi-centric study, n = 186 patients with spider phobia completed a baseline-assessment, a one-session VRET, a post-therapy assessment, and a 6-month-follow-up assessment (ClinicalTrials.gov, ID: NCT03208400). Short- and long-term outcomes regarding self-reported symptoms in the spider phobia questionnaire (SPQ) and final patient-spider distance in the behavioral avoidance test (BAT) were predicted via logistic regression models with the corresponding baseline score, age, initial fear activation, within-session fear reduction and fear expectancy violation as predictors. To predict long-term remission status at 6-month-follow-up, dimensional short-term changes in the SPQ and BAT were additionally included. Higher within-session fear reductions predicted better treatment outcomes (long-term SPQ; short- and long-term BAT). Lower initial fear activation tended to be associated with better long-term outcomes (SPQ), while fear expectancy violation was not associated with any outcome measure. Short-term change in the SPQ predicted remission status. Findings highlight that in VRET for spider phobia, the experience of fear reduction is central for short- and long-term treatment success and should be focused by therapists.

Behavioral and Magnetoencephalographic Correlates of Fear Generalization Are Associated With Responses to Later Virtual Reality Exposure Therapy in Spider Phobia.

BACKGROUND: Because overgeneralization of fear is a pathogenic marker of anxiety disorders, we investigated whether pretreatment levels of fear generalization in spider-phobic patients are related to their response to exposure-based treatment to identify pretreatment moderators of treatment success. METHODS: A total of 90 patients with spider phobia completed pretreatment clinical and magnetoencephalography assessments, one session of virtual reality exposure therapy, and a posttreatment clinical assessment. Based on the primary outcome (30% symptom reduction in self-reported symptoms), they were categorized as responders or nonresponders. In a pretreatment magnetoencephalography fear generalization paradigm involving fear conditioning with 2 unconditioned stimuli (UCS), we obtained fear ratings, UCS expectancy ratings, and event-related fields to conditioned stimuli (CS: CS-, CS+) and 7 different generalization stimuli on a perceptual continuum from CS- to CS+. RESULTS: Before treatment, nonresponders showed behavioral overgeneralization indicated by more linear generalization gradients in fear ratings. Analyses of magnetoencephalography source estimations revealed that nonresponders showed a decline of their (inhibitory) frontal activations to safety-signaling CS- and generalization stimuli compared with CS+ over time, while responders maintained these activations at early (<300 ms) and late processing stages. CONCLUSIONS: Results provide initial evidence that pretreatment differences of behavioral and neural markers of fear generalization may act as moderators of later responses to behavioral exposure. Stimulating further research on fear generalization as a potential predictive marker, our findings are an important first step in the attempt to identify patients who may not benefit from exposure therapy and to personalize and optimize treatment strategies for this vulnerable patient group.

Clinical predictors of treatment response towards exposure therapy in virtuo in spider phobia: A machine learning and external cross-validation approach.

While being highly effective on average, exposure-based treatments are not equally effective in all patients. The a priori identification of patients with a poor prognosis may enable the application of more personalized psychotherapeutic interventions. We aimed at identifying sociodemographic and clinical pre-treatment predictors for treatment response in spider phobia (SP). N = 174 patients with SP underwent a highly standardized virtual reality exposure therapy (VRET) at two independent sites. Analyses on group-level were used to test the efficacy. We applied a state-of-the-art machine learning protocol (Random Forests) to evaluate the predictive utility of clinical and sociodemographic predictors for a priori identification of individual treatment response assessed directly after treatment and at 6-month follow-up. The reliability and generalizability of predictive models was tested via external cross-validation. Our study shows that one session of VRET is highly effective on a group-level and is among the first to reveal long-term stability of this treatment effect. Individual short-term symptom reductions could be predicted above chance, but accuracies dropped to non-significance in our between-site prediction and for predictions of long-term outcomes. With performance metrics hardly exceeding chance level and the lack of generalizability in the employed between-site replication approach, our study suggests limited clinical utility of clinical and sociodemographic predictors. Predictive models including multimodal predictors may be more promising.

Effects of low-level visual information and perceptual load on P1 and N170 responses to emotional expressions.

Emotional facial expressions lead to modulations of early event-related potentials (ERPs). However, it has so far remained unclear how far these modulations represent face-specific effects rather than differences in low-level visual features, and to which extent they depend on available processing resources. To examine these questions, we conducted two preregistered independent experiments (N = 40 in each experiment) using different variants of a novel task that manipulates peripheral perceptual load across levels but keeps overall visual stimulation constant. At the display center, we presented task-irrelevant angry, neutral, and happy faces and their Fourier phase-scrambled versions, which preserved low-level visual features. The results of both studies showed load-independent P1 and N170 emotional expression effects. Importantly, by using Bayesian analyses we could confirm that these facial expression effects were face-independent for the P1 but not for the N170 component. We conclude that firstly, ERP modulations during the P1 interval strongly depend on low-level visual information, while the N170 modulation requires the processing of figural facial expression features. Secondly, both P1 and N170 modulations appear to be immune to a large range of variations in perceptual load.

Stimulus arousal drives amygdalar responses to emotional expressions across sensory modalities.

The factors that drive amygdalar responses to emotionally significant stimuli are still a matter of debate - particularly the proneness of the amygdala to respond to negatively-valenced stimuli has been discussed controversially. Furthermore, it is uncertain whether the amygdala responds in a modality-general fashion or whether modality-specific idiosyncrasies exist. Therefore, the present functional magnetic resonance imaging (fMRI) study systematically investigated amygdalar responding to stimulus valence and arousal of emotional expressions across visual and auditory modalities. During scanning, participants performed a gender judgment task while prosodic and facial emotional expressions were presented. The stimuli varied in stimulus valence and arousal by including neutral, happy and angry expressions of high and low emotional intensity. Results demonstrate amygdalar activation as a function of stimulus arousal and accordingly associated emotional intensity regardless of stimulus valence. Furthermore, arousal-driven amygdalar responding did not depend on the visual and auditory modalities of emotional expressions. Thus, the current results are consistent with the notion that the amygdala codes general stimulus relevance across visual and auditory modalities irrespective of valence. In addition, whole brain analyses revealed that effects in visual and auditory areas were driven mainly by high intense emotional facial and vocal stimuli, respectively, suggesting modality-specific representations of emotional expressions in auditory and visual cortices.

Theranostic markers for personalized therapy of spider phobia: Methods of a bicentric external cross-validation machine learning approach.

OBJECTIVES: Embedded in the Collaborative Research Center "Fear, Anxiety, Anxiety Disorders" (CRC-TRR58), this bicentric clinical study aims at identifying biobehavioral markers of treatment (non-)response by applying machine learning methodology with an external cross-validation protocol. We hypothesize that a priori prediction of treatment (non-)response is possible in a second, independent sample based on multimodal markers. METHODS: One-session virtual reality exposure treatment (VRET) with patients with spider phobia was conducted on two sites. Clinical, neuroimaging, and genetic data were assessed at baseline, post-treatment and after 6 months. The primary and secondary outcomes defining treatment response are as follows: 30% reduction regarding the individual score in the Spider Phobia Questionnaire and 50% reduction regarding the individual distance in the behavioral avoidance test. RESULTS: N = 204 patients have been included (n = 100 in Würzburg, n = 104 in Münster). Sample characteristics for both sites are comparable. DISCUSSION: This study will offer cross-validated theranostic markers for predicting the individual success of exposure-based therapy. Findings will support clinical decision-making on personalized therapy, bridge the gap between basic and clinical research, and bring stratified therapy into reach. The study is registered at ClinicalTrials.gov (ID: NCT03208400).

Brain activation during disorder-related script-driven imagery in panic disorder: a pilot study.

Despite considerable effort, the neural correlates of altered threat-related processing in panic disorder (PD) remain inconclusive. Mental imagery of disorder-specific situations proved to be a powerful tool to investigate dysfunctional threat processing in anxiety disorders. The current functional magnetic resonance imaging (fMRI) study aimed at investigating brain activation in PD patients during disorder-related script-driven imagery. Seventeen PD patients and seventeen healthy controls (HC) were exposed to newly developed disorder-related and neutral narrative scripts while brain activation was measured with fMRI. Participants were encouraged to imagine the narrative scripts as vividly as possible and they rated their script-induced emotional states after the scanning session. PD patients rated disorder-related scripts as more arousing, unpleasant and anxiety-inducing as compared to HC. Patients relative to HC showed elevated activity in the right amygdala and the brainstem as well as decreased activity in the rostral anterior cingulate cortex, and the medial and lateral prefrontal cortex to disorder-related vs. neutral scripts. The results suggest altered amygdala/ brainstem and prefrontal cortex engagement and point towards the recruitment of brain networks with opposed activation patterns in PD patients during script-driven imagery.

One executive function never comes alone: monitoring and its relation to working memory, reasoning, and different executive functions.

Monitoring is involved in many daily tasks and is described in several theoretical approaches of executive functioning. This study investigated the relative relationship of cognitive processes that are theoretically relevant to monitoring, such as concept formation, reasoning, working memory, and general cognitive control functions. Data from 699 participants who performed the Balanced Switching Task, aiming at capturing monitoring, were used. Subsamples also performed standard tasks assessing the processes assumed to be related to monitoring. Structural equation modeling revealed that general cognitive control processes are particularly relevant. They mediate the relationship between working memory, reasoning, and monitoring. Updating and maintaining information, as well as concluding from information which strategies can guide behavior toward predefined goals, is required for the ability to exert general cognitive control, which again may be relevant for implementing strategies in a goal-directed way. Together, these processes seem to be necessary to adequately monitor behavior in complex tasks.

Interpersonal violence in posttraumatic women: brain networks triggered by trauma-related pictures.

Interpersonal violence (IPV) is one of the most frequent causes for the development of posttraumatic stress disorder (PTSD) in women. Trauma-related triggers have been proposed to evoke automatic emotional responses in PTSD. The present functional magnetic resonance study investigated the neural basis of trauma-related picture processing in women with IPV-PTSD (n = 18) relative to healthy controls (n = 18) using a newly standardized trauma-related picture set and a non-emotional vigilance task. We aimed to identify brain activation and connectivity evoked by trauma-related pictures, and associations with PTSD symptom severity. We found hyperactivation during trauma-related vs neutral picture processing in both subcortical [basolateral amygdala (BLA), thalamus, brainstem] and cortical [anterior cingulate cortex (ACC), medial prefrontal cortex (mPFC), insula, occipital cortex] regions in IPV-PTSD. In patients, brain activation in amygdala, ACC, insula, occipital cortex and brainstem correlated positively with symptom severity. Furthermore, connectivity analyses revealed hyperconnectivity between BLA and dorsal ACC/mPFC. Results show symptom severity-dependent brain activation and hyperconnectivity in response to trauma-related pictures in brain regions related to fear and visual processing in women suffering from IPV-PTSD. These brain mechanisms appear to be associated with immediate responses to trauma-related triggers presented in a non-emotional context in this PTSD subgroup.

Specifically altered brain responses to threat in generalized anxiety disorder relative to social anxiety disorder and panic disorder.

BACKGROUND: Despite considerable effort, the neurobiological underpinnings of hyper-responsive threat processing specific to patients suffering from generalized anxiety disorder (GAD) remain poorly understood. The current functional magnetic resonance imaging (fMRI) study aims to delineate GAD-specific brain activity during immediate threat processing by comparing GAD patients to healthy controls (HC), to social anxiety disorder (SAD) and to panic disorder (PD) patients. METHOD: Brain activation and functional connectivity patterns to threat vs. neutral pictures were investigated using event-related fMRI. The sample consisted of 21 GAD, 21 PD, 21 SAD and 21 HC. RESULTS: GAD-specific elevated activity to threat vs. neutral pictures was found in cingulate cortex, dorsal anterior insula/frontal operculum (daI/FO) and posterior dorsolateral prefrontal cortex (dlPFC). Defining these effects as seed regions, we detected GAD-specific increased functional connectivity to threat vs. neutral pictures between posterior dlPFC and ventrolateral prefrontal cortex, between cingulate cortex and amygdala, between cingulate cortex and anterior insula, as well as decreased functional connectivity between daI/FO and mid-dlPFC. CONCLUSION: The findings present the first evidence for GAD-specific neural correlates of hyper-responsive threat processing, possibly reflecting exaggerated threat sensitivity, maladaptive appraisal and attention-allocation processes.

Monitoring supports performance in a dual-task paradigm involving a risky decision-making task and a working memory task.

Performing two cognitively demanding tasks at the same time is known to decrease performance. The current study investigates the underlying executive functions of a dual-tasking situation involving the simultaneous performance of decision making under explicit risk and a working memory task. It is suggested that making a decision and performing a working memory task at the same time should particularly require monitoring-an executive control process supervising behavior and the state of processing on two tasks. To test the role of a supervisory/monitoring function in such a dual-tasking situation we investigated 122 participants with the Game of Dice Task plus 2-back task (GDT plus 2-back task). This dual task requires participants to make decisions under risk and to perform a 2-back working memory task at the same time. Furthermore, a task measuring a set of several executive functions gathered in the term concept formation (Modified Card Sorting Test, MCST) and the newly developed Balanced Switching Task (BST), measuring monitoring in particular, were used. The results demonstrate that concept formation and monitoring are involved in the simultaneous performance of decision making under risk and a working memory task. In particular, the mediation analysis revealed that BST performance partially mediates the influence of MCST performance on the GDT plus 2-back task. These findings suggest that monitoring is one important subfunction for superior performance in a dual-tasking situation including decision making under risk and a working memory task.

Among three different executive functions, general executive control ability is a key predictor of decision making under objective risk.

Executive functioning is supposed to have an important role in decision making under risk. Several studies reported that more advantageous decision-making behavior was accompanied by better performance in tests of executive functioning and that the decision-making process was accompanied by activations in prefrontal and subcortical brain regions associated with executive functioning. However, to what extent different components of executive functions contribute to decision making is still unclear. We tested direct and indirect effects of three executive functions on decision-making performance in a laboratory gambling task, the Game of Dice Task (GDT). Using Brand's model of decisions under risk (2006) we tested seven structural equation models with three latent variables that represent executive functions supposed to be involved in decision making. The latent variables were general control (represented by the general ability to exert attentional and behavioral self-control that is in accordance with task goals despite interfering information), concept formation (represented by categorization, rule detection, and set maintenance), and monitoring (represented by supervision of cognition and behavior). The seven models indicated that only the latent dimension general control had a direct effect on decision making under risk. Concept formation and monitoring only contributed in terms of indirect effects, when mediated by general control. Thus, several components of executive functioning seem to be involved in decision making under risk. However, general control functions seem to have a key role. They may be important for implementing the calculative and cognitively controlled processes involved in advantageous decision making under risk.

Stress and decision making: neural correlates of the interaction between stress, executive functions, and decision making under risk.

Stress and additional load on the executive system, produced by a parallel working memory task, impair decision making under risk. However, the combination of stress and a parallel task seems to preserve the decision-making performance [e.g., operationalized by the Game of Dice Task (GDT)] from decreasing, probably by a switch from serial to parallel processing. The question remains how the brain manages such demanding decision-making situations. The current study used a 7-tesla magnetic resonance imaging (MRI) system in order to investigate the underlying neural correlates of the interaction between stress (induced by the Trier Social Stress Test), risky decision making (GDT), and a parallel executive task (2-back task) to get a better understanding of those behavioral findings. The results show that on a behavioral level, stressed participants did not show significant differences in task performance. Interestingly, when comparing the stress group (SG) with the control group, the SG showed a greater increase in neural activation in the anterior prefrontal cortex when performing the 2-back task simultaneously with the GDT than when performing each task alone. This brain area is associated with parallel processing. Thus, the results may suggest that in stressful dual-tasking situations, where a decision has to be made when in parallel working memory is demanded, a stronger activation of a brain area associated with parallel processing takes place. The findings are in line with the idea that stress seems to trigger a switch from serial to parallel processing in demanding dual-tasking situations.

Performing a secondary executive task with affective stimuli interferes with decision making under risk conditions.

Previous studies demonstrated that executive functions are crucial for advantageous decision making under risk and that therefore decision making is disrupted when working memory capacity is demanded while working on a decision task. While some studies also showed that emotions can affect decision making under risk, it is unclear how affective processing and executive functions predict decision-making performance in interaction. The current experimental study used a between-subjects design to examine whether affective pictures (positive and negative pictures compared to neutral pictures), included in a parallel executive task (working memory 2-back task), have an impact on decision making under risk as assessed by the Game of Dice Task (GDT). Moreover, the performance GDT plus 2-back task was compared to the performance in the GDT without any additional task (GDT solely). The results show that the performance in the GDT differed between groups (positive, negative, neutral, and GDT solely). The groups with affective pictures, especially those with positive pictures in the 2-back task, showed more disadvantageous decisions in the GDT than the groups with neutral pictures and the group performing the GDT without any additional task. However, executive functions moderated the effect of the affective pictures. Regardless of affective influence, subjects with good executive functions performed advantageously in the GDT. These findings support the assumption that executive functions and emotional processing interact in predicting decision making under risk.