Facing the fear – clinical and neural effects of cognitive behavioural and pharmacotherapy in panic disorder with agoraphobia

https://doi.org/10.1016/j.euroneuro.2016.01.004Get rights and content

Abstract

Introduction: Cognitive behavioural therapy (CBT) and pharmacological treatment with selective serotonin or serotonin-noradrenalin reuptake inhibitors (SSRI/SSNRI) are regarded as efficacious treatments for panic disorder with agoraphobia (PD/AG). However, little is known about treatment-specific effects on symptoms and neurofunctional correlates. Experimental procedures: We used a comparative design with PD/AG patients receiving either two types of CBT (therapist-guided (n=29) or non-guided exposure (n=22)) or pharmacological treatment (SSRI/SSNRI; n=28) as well as a wait-list control group (WL; n=15) to investigate differential treatment effects in general aspects of fear and depression (Hamilton Anxiety Rating Scale HAM-A and Beck Depression Inventory BDI), disorder-specific symptoms (Mobility Inventory MI, Panic and Agoraphobia Scale subscale panic attacks PAS-panic, Anxiety Sensitivity Index ASI, rating of agoraphobic stimuli) and neurofunctional substrates during symptom provocation (Westphal-Paradigm) using functional magnetic resonance imaging (fMRI). Comparisons of neural activation patterns also included healthy controls (n=29). Results: Both treatments led to a significantly greater reduction in panic attacks, depression and general anxiety than the WL group. The CBT groups, in particular, the therapist-guided arm, had a significantly greater decrease in avoidance, fear of phobic situations and anxiety symptoms and reduction in bilateral amygdala activation while the processing of agoraphobia-related pictures compared to the SSRI/SSNRI and WL groups. Discussion: This study demonstrates that therapist-guided CBT leads to a more pronounced short-term impact on agoraphobic psychopathology and supports the assumption of the amygdala as a central structure in a complex fear processing system as well as the amygdala׳s involvement in the fear system׳s sensitivity to treatment.

Introduction

Panic disorder with agoraphobia (PD/AG) is one of the most prevalent anxiety disorders and is accompanied by severe impairment of functionality and quality of life (Wittchen et al., 2011). The main characteristics are panic attacks, fear of panic symptoms and phobic anxiety or avoidance of situations from which it is difficult or embarrassing to escape (American Psychiatric Association, 2000). Cognitive behavioural therapy (CBT) and pharmacotherapy with selective serotonin or serotonin-noradrenalin reuptake inhibitors (SSRI/SSNRI) have been proven to be effective treatments (Bandelow et al., 2014, McHugh et al., 2009). While a key goal of CBT is the fear extinction through in situ behavioural exposure, SSRI/SSNRI treatment is thought to have a rather unspecific effect.

The aim of this study was to achieve a better understanding of the underlying mechanisms behind PD/AG treatment by directly comparing treatment-specific effects. In recent studies that have contrasted the clinical effects of CBT and pharmacotherapy, samples were often heterogeneous (PD and/or AG) and the assessment batteries used encompassed a range of symptoms that are common to several anxiety disorders, such as anxiety-related physiological symptoms, psychological restlessness or tension (Pull and Damsa, 2008, Roshanaei-Moghaddam et al., 2011, van Apeldoorn et al., 2010). By additionally assessing panic attacks, avoidance and the fear of physical sensations, we aimed to obtain more precise conclusions.

Moreover, we aimed to investigate how these potential changes in psychopathology are represented on a neural network level. Many studies investigating the neural correlates of anxiety disorders have suggested that the amygdala is the central structure for anxiety related information processing. For example, in analysing the evidence for functional neurobiological abnormalities in patients with phobic disorders and fear conditioned healthy controls (HC) who were confronted with aversive stimuli, Etkin and Wager (2007) concluded that amygdalar hyperactivation was the basis of the participant׳s shared anxiety symptoms such as fear and avoidance of critical cues. Regarding PD, Gorman et al. (2000) also described the amygdala as one of the central regions of fear processing. In their review, pharmacotherapy was believed to desensitise the pathologically hypersensitive anxiety network at the level of the amygdala by increasing the transmission of serotonin and noradrenalin. A direct effect of monoaminergic neurotransmission on functional amygdala activation has indeed been shown. (Kienast et al., 2008). Psychotherapy, on the other hand, is assumed to lead to a strengthened prefrontal top-down regulation of the amygdala (Yang et al., 2014). Nevertheless, findings about patterns of neuroplastic changes in the fear network following treatment have been inconsistent. Dresler et al. (2013) critically reflected on this neuroanatomical hypothesis of PD by examining the latest results of neuroimaging studies. Although the integral role of the amygdala in fear-related information processing was supported, a general hypersensitivity could not be reported. Sakai et al. (2005) found an increased bilateral glucose uptake in the amygdala, hippocampus and thalamus in patients with PD compared to HC, while Hayano et al. (2009) reported a negative correlation between amygdala volume and state anxiety in PD patients. On the other hand, evidence from a combined CBT and pharmacological study that investigated the effects of treatment on brain metabolism in a small PD sample (n=6) with or without AG during rest found that there were changes in the frontal and temporal cortical regions, but no effects in the hippocampus, parahippocampal gyrus or amygdala (Prasko et al., 2004). Furthermore, decreased limbic and paralimbic activation has been found in phobic disorders following CBT (Linden, 2006), whereas increased glucose metabolism in cortical and limbic areas such as the amygdala has been reported after paroxetine treatment in PD patients (Sim et al., 2010). In most of these studies, patients were assessed in a resting state. Another study used a social fear-inducing task to identify fear perception-related alterations and found an equal reduction in cerebral blood flow in the bilateral amygdala, hippocampus and surrounding areas for social phobia patients receiving CBT or citalopram (Furmark et al., 2002). However, because the task lacked a control condition, the specificity of the effects remains unclear. On the contrary, changes in amygdalar reactivity in PD patients and HC were not detected when performing an affective go/no go task before and after a short-term psychodynamic treatment, despite a significant symptom reduction (Beutel et al. (2010). However, the cognitive demanding task could have led to cognitive avoidance and may have prevented the activation of the fear circuit. In summary, the results of these studies are limited by small sample sizes, the lack of a control condition, the resting state measures or task demands.

The research network PANIC-NET conducted the multicentre study “Mechanisms of Action in CBT” (MAC) (Gloster et al., 2009, Gloster et al., 2011), in which a large sample of patients with PD/AG was investigated using different paradigms and functional magnetic resonance imaging (fMRI) to examine the structural and functional characteristics of the disorder and its sensitivity to CBT. Here, studies on fear conditioning detected an association with the fear circuitry network. Although a comparison between patients and HC in simple and differential conditioning revealed no stronger amygdala activation but instead enhanced activation of the midbrain and the bilateral dorsal inferior frontal gyrus, respectively (Lueken et al., 2014a), an altered connection between limbic and anterior cingulate cortex activation had already been found at baseline in CBT non-responders (Lueken et al., 2013). These patients exhibited enhanced amygdala activation when confronted with safety signals, indicating that their ability to distinguish between safe and unsafe stimuli was disturbed. Kircher et al. (2013) also found a reduction in amygdala activity after treatment in direct pre-post single group analysis. While these previous investigations have focused on the effects of CBT only, and predominantly on the neural correlates of fear conditioning, in this study we investigate the differential effects of CBT and SSRI/SSNRI on disorder-specific symptoms and the neural correlates of agoraphobia-specific stimuli processing. Consequently, we compared the effects of CBT (further discriminating between different variants of exposure therapy) with a SSRI/SSNRI-treated group, a wait-list and a HC group. We assumed that both treatments would lead to a rapid improvement in general aspects of anxiety and associated depression but that CBT would have a greater impact on disorder-specific symptoms. By addressing the supposed limitations of the previously mentioned imaging studies, we expected to find increased amygdala activation before therapy in patients compared to HC and hypothesised that the reduction in disorder-specific symptoms, particularly in the CBT groups, should be reflected in a stronger decrease in amygdalar activation compared to the SSRI/SSNRI and the control groups.

Section snippets

Participants

This study was based on the MAC study, which was a randomised controlled clinical trial of CBT for PD/AG (Gloster et al., 2009, Gloster et al., 2011) in eight German centres (Aachen, Berlin-Adlershof, Berlin-Charité, Bremen, Dresden, Greifswald, Muenster, and Wuerzburg). Three hundred sixty-nine patients who met the DSM-IV-TR criteria for PD/AG were randomly assigned to three different study conditions, namely therapist-guided exposure therapy (CBTT+), non-guided exposure therapy (CBTT) and

Sample characteristics

At baseline, all of the groups were comparable with regard to sample characteristics such as age, sex, education, handedness (Oldfield, 1971), smoking status and the results of the neuropsychological testing. The patient groups did not differ significantly in disorder severity, as measured by the clinical inventories and the amount of comorbid DIA-X-diagnoses (see Table 1). The most frequent comorbid mental disorders were depression (40.4%), other anxiety disorders (38.3%) and somatoform

Discussion

Despite evidence that CBT and SSRI/SSNRI are effective treatments of PD/AG, there is a lack of knowledge about their specific effects on the psychopathology, neural substrates and neuroplastic changes following treatment. This is the first comparison study with a homogenous disorder, a sufficient sample size, a WL and a HC group, a wide range of disorder-specific and disorder-unspecific symptom scales and a symptom provocation approach. The subjective, clinical and neurofunctional data showed

Role of funding source

The MAC study was funded by the German Federal Ministry of Education and Research (Project No. 01GV0615) as part of the Psychotherapy Research Funding Initiative. (Panic-Net, to ALG, TK, AS, HUW, VA, AH and JD; for details see web page http://www.paniknetz.de/netzwerk.html or acknowledgements of other published studies of the Panic-net). Doctoral funding for CL was provided by the “FAZIT-Stiftung” and the “Elsa-Neumann-Stipendium”. The foundations had no influence on the study design, the

Contributors

Individual contributions to the study: principal investigators (PIs) with respective areas of responsibility in the MAC study are V. Arolt (Münster: Overall MAC Program Coordination), H.U. Wittchen (Dresden: PI for the Randomised Clinical Trial (RCT) and Manual Development), A. Hamm (Greifswald: PI for Psychophysiology), A.L. Gerlach (Münster: PI for Psychophysiology and Panic subtypes), A. Ströhle (Berlin: PI for Experimental Pharmacology), T. Kircher (Marburg: PI for Functional Neuroimaging),

Conflict of interest

The following authors report no conflict of interest concerning the content of this paper: C. Liebscher, A. Wittmann, J. Gechter, U. Lueken, B. Straube, B. Pfleiderer, L. Fehm, A. L. Gerlach, T. Kircher, T. Fydrich, J. Deckert, H.-U. Wittchen, A. Heinz, and V. Arolt. F. Schlagenhauf received funding from the German Research Foundation (SCHL 1969/1-1 & 2-1) and was supported by the Max Planck Society. Jens Plag received speaker honoraria from Pfizer and Grünenthal. Andreas Ströhle received

Acknowledgments

This work is part of the German multicentre trial “Mechanisms of Action in CBT” (MAC) that was conducted by the research group PANIC-NET. The multicentre study was registered with the trial name "Improving cognitive behavioural therapy for panic by identifying the active ingredients and understanding the mechanisms of action: a multicentre study; http://www.controlled-trials.com/ISRCTN80046034/ with the registration number ISRCTN80046034. The RCT project was approved by the Ethics Committee of

References (43)

  • H.U. Wittchen et al.

    The size and burden of mental disorders and other disorders of the brain in Europe 2010

    Eur. Neuropsychopharmacol.: J. Eur. Coll. Neuropsychopharmacol.

    (2011)
  • Y. Yang et al.

    The neural correlates of cognitive behavioral therapy: recent progress in the investigation of patients with panic disorder

    Behav. Res. Ther.

    (2014)
  • American Psychiatric Association

    Diagnostic and Statistical Manual of Mental Disorders

    (2000)
  • B. Bandelow

    Panic and Agoraphobia Scale (PAS)

    (1997)
  • B. Bandelow et al.

    The diagnosis of and treatment recommendations for anxiety disorders

    Dtsch. Arzteblatt Int.

    (2014)
  • B. Bandelow et al.

    Efficacy of treatments for anxiety disorders: a meta-analysis

    Int. Clin. Psychopharmacol.

    (2015)
  • A.T. Beck et al.

    An inventory for measuring depression

    Arch. Gen. Psychiatry

    (1961)
  • T. Dresler et al.

    Revise the revised? New dimensions of the neuroanatomical hypothesis of panic disorder

    J. Neural Trans.

    (2013)
  • A. Etkin et al.

    Functional neuroimaging of anxiety: a meta-analysis of emotional processing in PTSD, social anxiety disorder, and specific phobia

    Am. J. Psychiatry

    (2007)
  • T. Furmark et al.

    Common changes in cerebral blood flow in patients with social phobia treated with citalopram or cognitive-behavioral therapy

    Arch. Gen. Psychiatry

    (2002)
  • A.T. Gloster et al.

    Mechanism of action in CBT (MAC): methods of a multi-center randomized controlled trial in 369 patients with panic disorder and agoraphobia

    Eur. Arch. Psychiatry Clin. Neurosci.

    (2009)
  • Cited by (15)

    • Neurocircuitry of treatment in anxiety disorders

      2022, Biomarkers in Neuropsychiatry
      Citation Excerpt :

      However, psychodynamic psychotherapy decreased hippocampal activation and increased vlPFC activity (Beutel et al., 2010). Additionally, CBT decreased emotional image-related amygdala activation in two studies (Liebscher et al., 2016; Reinecke et al., 2018), although this effect was not observed in a third CBT study (Wittmann et al., 2018). In addition, CBT decreased insular activation to emotional images (Wittmann et al., 2018) and increased middle frontal gyrus and superior parietal lobule activation during a Flanker task (Neufang et al., 2019).

    • Neuroimaging findings as predictors of treatment outcome of psychotherapy in anxiety disorders

      2019, Progress in Neuro-Psychopharmacology and Biological Psychiatry
      Citation Excerpt :

      CBT produced several brain function changes in SAD patients (Mansson et al., 2013; Klumpp et al., 2013; Klumpp et al., 2014; Klumpp et al., 2014a; Klumpp et al., 2016; Klumpp et al., 2017a; Yuan et al., 2016; Yuan et al., 2017; Yuan et al., 2018; Furmark et al., 2002; Young et al., 2017; Goldin et al., 2014; Mansson et al., 2016), including modifications in: left and right amygdala, caudate, cerebellum, dlPFC, putamen, mOFG, ACC, supramarginal gyrus, dmPFC, right dlPFC, precuneus, hippocampus, insula, right sFG, IPL, MOG and left pSTG. In the PDAG and GAD studies (Lueken et al., 2013; Liebscher et al., 2016; Kircher et al., 2013; Straube et al., 2014; Ball et al., 2014; Fonzo et al., 2014), CBT produced changes in: ACC, hippocampus, pre and post central gyri, amygdala, left middle temporal gyrus, fusiform gyrus, inferior occipital gyrus, inferior frontal gyrus, left uncus, left transversal temporal gyrus and right substantia nigra. Although the effectiveness of psychotherapy has been proved and described in several studies (Bandelow et al., 2017) the underlying mechanism of psychotherapy are still not well clarified and to determine that a considerable numbers of studies applied fMRI to identify the brain circuits that change during psychotherapy. (

    • Survival circuits in affective disorders

      2018, Current Opinion in Behavioral Sciences
      Citation Excerpt :

      In contrast to psychological treatments that primarily impact cortical brain activity, evidence to date suggests that pharmacological treatments target subcortical regions of survival circuitry. One recent study directly compared changes in neural functioning in response to CBT and SSRI treatment for panic disorder with agoraphobia [28•]. Focusing on amygdala reactivity, both treatment types were found to reduce amygdala reactivity to agoraphobic stimuli, but this reduction was greater in the CBT group than the SSRI group.

    View all citing articles on Scopus
    View full text