University of Groningen The anxious brain Boshuisen, Marjolein Larissa IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record Publication date: 2003 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): Boshuisen, M. L. (2003). The anxious brain: neuroimaging of panic and anxiety. s.n. Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons). The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license. More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne- amendment. Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum. Download date: 17-03-2023 17355 Bosh B 22-10-2003 14:20 Pagina 55 4 Chapter Partial normalization of rCBF patterns after succesful pharmacological treatment of panic disorder patients: an H15O PET scan study 2 M. L. Boshuisen MD, G. J. Ter Horst PhD, A. M. J. Paans PhD, J. A. den Boer MD PhD Submitted 17355 Bosh B 22-10-2003 14:20 Pagina 56 The Anxious Brain Abstract Background: Selective serotonin reuptake inhibitors (SSRIs) are considered the first line treatment of panic disorder (PD). In this study we combined neuroimaging and pentagastrin (a challenge agent) to evaluate the effect of SSRI treatment in PD patients. Methods: Six patients were studied twice with H215O PET scan, once before and once after 12 weeks of successful treatment with sertraline. Four H215O PET scans were made, all immediately after injection of pentagastrin (once) and saline (three times). The data were compared with other pre- and posttreatment data of patients who only participated in the neuroimaging study once, either before or after treat- ment. Results: Before treatment all patients experienced a panic attack after pentagastrin injection, while after succesful treatment only two of the six patients panicked. After treatment we observed increases in the rCBF in the parahippocampal gyrus, the midbrain, the thalamus and the hypothalamus as compared to the situation before treatment, in rest as well as after pentagastrin injection. In the prefrontal cortex and the orbitofrontal cortex posttreatment, a decrease was seen during rest, while an increase was noticed during pentagastrin challenge as compared to the pretreat- ment situation. Conclusions: The results of this study suggest that after succesful sertraline treat- ment of PD patients there are still changes in rCBF during pentagastrin challenge, but these changes are comparable to the changes seen in healthy control subjects, and thus more related to a normal reaction on pentagastrin. This confirms our hypothesis of partial normalization of rCBF patterns in PD patients after succesful pharmacological treatment. Introduction Panic disorder (PD) is a rather common disorder with a lifetime prevalence of approximately 2% (American Psychiatric Association 1994). The pharmacological treatment of first choice for PD at this moment is a selective serotonin reuptake inhibitor (SSRI) (American Psychiatric Association Committee on Nomenclature and Statistics 1998). SSRIs were shown in a number of studies to be more effective than placebo and as effective as earlier treatments, such as benzodiazepines and tricyclic antidepressants (Bakker et al 2000; Pollack and Marzol 2000; Sheehan 1999). Because of the success of these pharmacological interventions, several biological theories of PD have been proposed, including dysfunction of serotonergic and nora- drenergic neurotransmitter systems (Westenberg et al 1996). At this moment neu- roimaging, sometimes combined with treatment studies and/or symptom provocati- on, is the most widely used technique to further clinically explore the neurobiology of psychiatric diseases such as PD. A decreased binding to the benzodiazepine receptor was observed in PD patients compared to control subjects (Bremner et al 2000; Kaschka et al 1995; Kuikka et al 1995; Malizia et al 1998; Schlegel et al 1994). Several neuroimaging studies evalua- 56 17355 Bosh B 22-10-2003 14:20 Pagina 57 Partial normalization of rCBF pattern after succesful SSRI treatment of PD patients ted the effect of SSRIs in obsessive compulsive disorder (OCD) (Baxter et al 1992; Benkelfat et al 1990; Saxena et al 1998; Saxena et al 2002; Swedo et al 1992), social phobia (Furmark et al 2002; Van der Linden et al 2000), and PD (Nordahl et al 1998). The majority of these studies showed decreases of regional cerebral blood flow (rCBF) in the caudate nucleus and/or the orbitofrontal cortex after succesful treatment. Furthermore, the effect of behavioral treatment (cognitive behavioral the- rapy or CBT) in OCD and in social phobia was evaluated by neuroimaging and com- pared with the effect of pharmacological treatment (SSRI) (Baxter et al 1992; Brody et al 1998; Furmark et al 2002). In all these studies both treatment strategies show- ed similar effects, in symptom improvement as well as in decreases in brain activati- on in different anxiety related brain regions. In addition, in social phobia patients this design was combined with a symptom provocation test (Furmark et al 2002). Significant posttreatment decreases in rCBF during symptom provocation were observed in responders to both treatment strategies, SSRI and CBT, in the amygda- la, hippocampus and parahippocampal cortex. Two case studies using symptom provocation in posttraumatic stress disorder (PTSD) patients before and after SSRI treatment were published (Fernandez et al 2001; Levin et al 1999). The subjects in both studies felt significantly less anxious during the posttreatment challenge, and in both studies changes in brain activation were noted. In the study of Levin et al (1999) increases in rCBF were observed after treatment during the challenge in the anterior cingulate cortex and the left frontal lobe as compared to the pretreatment condition. In the study of Fernandez et al (2001) it was noted that in the untreated condition trauma reminders resulted in rCBF decreases in several anxiety related brain regions like the insula, the inferior frontal and the prefrontal cortex, while after treatment this effect was not seen. In PD patients only one study has been performed studying treatment effects with neu- roimaging, and no provocation test was used in this case (Nordahl et al 1998). Furthermore, the patients were not studied before and after treatment. The study group consisted of only succesfully treated patients and was compared to a previ- ous study group of unmedicated PD patients (Nordahl et al 1990). The aim of the present study was to examine rCBF changes associated with a phar- macological treatment response in PD patients. Therefore we used H215O PET to exa- mine rCBF in PD patients before and after 12 weeks of sertraline treatment, both during a pentagastrin challenge and at rest. As stated above, SSRIs are the first in line of treatment for PD. Sertraline is one of the SSRIs proven to be effective in the treatment of PD (Hirschfeld 2000; Londborg et al 1998; Pohl et al 1998; Pollack et al 1998; Rapaport et al 2001). Pentagastrin is a five-amino-acid-long peptide very much like cholecystokinin-4 (CCK ). Both agents have been shown to reliably induce panic attacks and anxiety, 4 especially in PD patients, but dependent on the dose also in healthy volunteers (Abelson and Nesse 1990; de Montigny 1989; Van Megen et al 1994). The effect of succesful fluvoxamine treatment in PD patients on CCK -induced panic and anxiety 4 was studied before (van Megen et al 1997a). In this study 26 PD patients received, before and after a double blind 8-week treatment period with fluvoxamine (n = 17) or placebo (n = 9), a single blind bolus injection with 50 micrograms CCK . Treatment 4 with fluvoxamine (150 mg daily) significantly decreased the sensitivity of PD patients 57 17355 Bosh B 22-10-2003 14:20 Pagina 58 The Anxious Brain for CCK while placebo was without effect. Of the patients who responded to treat- 4 ment, 83% no longer experienced a panic attack when rechallenged with CCK , whe- 4 reas in the non-responders group this was only 28%. This study, among others, also shows that CCK is a reliable panic provoking agent, which provokes panic attacks in 4 PD patients even when challenged for the second time (Bradwejn et al 1992b; van Megen et al 1994; van Megen et al 1996a; van Megen et al 1997a). Based on the outcome of previous neuroimaging studies in anxiety disorders, we expected to observe rCBF changes during pentagastrin challenge in anxiety related brain regions such as the amygdala, the parahippocampal gyrus, the orbitofrontal cortex, the prefrontal cortex, the thalamus and the hypothalamus, in the pretreat- ment condition. In the posttreatment condition we hypothesized a partial normali- zation of the rCBF in the aforementioned regions during pentagastrin challenge. Methods and Materials Subjects Six patients with PD (2 men/4 women) participated in the study twice (for demogra- phic data see table 1). Another group of 10 PD patients participated only once, befo- re treatment, while a small group of 3 patients was studied once after a succesful 12- week treatment period with sertraline. All patients met the DSM-IV criteria of PD with or without agoraphobia, as diagnosed by a psychiatrist (American Psychiatric Association 1994). The subjects were physically healthy, as determined by history, physical and neurologic examination, and had no comorbid psychiatric or neurolo- gic diseases. All patients were free from medication for at least three weeks prior to the first PET scan (in the case of fluoxetine for at least 6 weeks). The patients were allowed to use oxazepam in the week before the PET scans, but not during the last 24 hours before the scan. The present levels of anxiety and depression in the patients were obtained by the Hamilton rating scales for anxiety (HAS) (Hamilton 1959) and for depression (HDS) (Hamilton 1967) on the day of the PET scan (+/- two days). None of the subjects was taking medication (other than oral contracepti- ves: three women) or had a history of substance abuse. In the week following the first PET scan the patients started with sertraline 50 mg once daily. The dose could be increased after one and after three weeks untill the maximum dose of 150 mg daily was reached. After twelve weeks the effect of the tre- atment was evaluated and the subjects were asked to participate again in the PET scan and challenge procedure. Treatment response was defined as a decrease of more than 50% on the HAS and/or a decrease in the number of panic attacks in the two weeks preceding the assessment of at least 50%. None of the treatment non- responders was willing to participate in the second scanning procedure. Finally nine treatment responders decided to participate in the procedure, six of whom also par- ticipated in the first scanning procedure. Because of the nature of the challenge test, in this study we choose to define a panic attack with stricter criteria than stated in the DSM-IV: severe anxiety had to be present during the panic attack, in addition to the criterion of at least 4 out of the 13 58 17355 Bosh B 22-10-2003 14:20 Pagina 59 Partial normalization of rCBF pattern after succesful SSRI treatment of PD patients symptoms. Pentagastrin in a dose of 40 micrograms induces panic-like symptoms in 100% of the subjects, and most of the subjects also report feelings of anxiety. All subjects gave oral and written informed consent after complete explanation of the nature and possible consequences of the study. The study was approved by the medical ethical committee of the Groningen University Hospital. Procedure The scanning procedure was performed twice, once before and again after a treat- ment period of 12 weeks of sertraline treatment. During the scanning procedure, the subjects were positioned in the scanner using a molded head rest with a strap over the head to minimize head movement. Two marks were placed on the forehead and nose of the subject to check for movement between the scans. The subjects were instructed to lie quietly with their eyes open in a dimly lit room, without conversati- on or noise, other than the noise from the scanner. First a 20-minute transmission scan, used for attenuation correction, was made. After this scan all subjects were scanned twice, before and after a pentagastrin challenge. Immediately after each scan blood pressure (BP) and heart rate (HR) were measured, and the subjects were asked about the panic-related symptoms experienced during the scan, with a semi-structured interview based on the DSM-IV criteria for panic attacks (Acute panic inventory, API). The API score was derived by counting the experienced anxiety symptoms multiplied by the severity rating each subject gave per experienced symp- tom. Furthermore subjects had to point out on a separate five-point scale how anxi- ous they felt immediately after the injection. Scanning was done between 11.30 am and 2.30 p.m., with a Siemens ECAT EXACT HR+ whole-body PET camera operating in high sensitivity 3D mode. Subjects recei- ved a total of 500 MBq of H215O per scan over 20 seconds through a forearm canula. The first scan was started at the moment of injection and lasted 120 seconds. Since the time required for the radioactive water to reach the brain is between 20 and 25 seconds in the injection system used, the effective scanning time is 95-100 seconds. Data are reconstructed by filtered backprojection with a zoom factor of 2.25. A Hamming filter with a cut-off frequency of 0.5 cycles/pixel is used. The resulting images have a nearly isotropical spatial resolution of roughly 5 mm FWHM (full width at half maximum). Data-analysis Statistical analyses for differences in the demographic, behavioral and physiological data were performed using SPSS for Windows version 10. Within and between group analyses were performed by t-tests and analysis of variance (ANOVA), using a level of significance of 5% in two-tailed testing. The PET scan data were processed and analyzed with statistical parametric mapping software (SPM99, Wellcome department of Cognitive Neurology, London, UK) (Friston et al 1995), implemented in Matlab (Mathworks, Sherborn, MA, USA). The scans of all subjects were realigned and normalized to the MNI (Montreal Neurological Institute) template, using heavy regularization. Images were smoothed 59 17355 Bosh B 22-10-2003 14:20 Pagina 60 The Anxious Brain with an 8 mm FWHM Gaussian kernel. Activation differences in terms of hyper- and hypoactivity in rCBF pre- and posttreat- ment were tested voxel-by-voxel by ANOVA. The analyses concerning pentagastrin- induced panic were performed on the data of the patients who experienced penta- gastrin-induced panic pretreatment and the patients who did not panic posttreat- ment. Differences over the scans in the main group of six subjects who participated twice in the procedure were analyzed by a so-called multisubject analysis, an ANOVA within-group analysis. Overall group and state differences were also tested, in all PD subjects scanned before treatment versus all PD subjects scanned after succesful treatment, by a multigroup analysis, wherein both groups were compared with ANOVA, with one between-group factor (before versus after treatment) and one within-group factor (pentagastrin-induced panic attack versus no panic attack). Clusters with voxel differences achieving a threshold of Z = 3.30 (corresponding to a p-value < 0.001 uncorrected) were displayed in three orthogonal projections on SPM glass brain projections. The statistical parametric maps (SPMs) were inspec- ted for the presence of findings in unpredicted regions; these were reported signifi- cant only if resisting correction for multiple comparisons based on the Gaussian random field theory (Friston et al 1996) and using the statistical procedures for con- trolling the false discovery rate (Genovese et al 2002) (p<0.05). Secondly the SPMs were inspected for the presence of voxel clusters of statistical significance in regions where abnormalities were predicted a priori (regions of interest were the different regions of the fear network in the brain); these regions were reported only if the uncorrected p-value was less than 0.001. SPM99 calculates corrected statistics across the whole brain, by working out the shape and size of the whole brain volu- me in the analysis and calculating the correction accordingly (Worsley et al 1996). The stereotactic coordinates of the peak differences were determined using the Talairach and Tournoux atlas (Talairach and Tournoux 1988), after translating the MNI coordinates to coordinates according to the Talairach template, by using the matlab function mni2tal.m as described by M. Brett on the website of the Cambridge Imagers Group (http//www.mrc-cbu.cam.ac.uk/Imaging/mnispace.html). Results The effects of twelve-week treatment with sertraline The scores on the HAS and the HDS, and the number of experienced panic attacks per week decreased significantly over the twelve-week treatment period. The mean score on the HAS decreased more than 50%, from a mean score of 25 (4) to a mean score of 11 (5) (mean (SD); see table 1; t = 5.4; df = 10; p < 0.000), and the score on the HDS decreased from 12 (2) to 5 (3) (mean (SD); t = 4.2; df = 10; p = 0.002). Only two of the patients experienced a panic attack in the week preceding the second PET scanning procedure, while they were free of panic attacks in the two weeks before. The other patients were free of panic attacks for at least four weeks. 60 17355 Bosh B 22-10-2003 14:20 Pagina 61 Partial normalization of rCBF pattern after succesful SSRI treatment of PD patients Table 1 Pre- Post- P treatment treatment Demographic Criteria Mean age (SD) 40 (16.7) NA Mean no. panic attacks/2 wks (SD) 4.5 (1.0) 0.3 (0.5) < 0.01 No. males / female 2/4 NA Mean HAS (SD) 25.3 (4.0) 10.7 (5.3) < 0.01 Mean HDS (SD) 11.8 (2.0) 4.8 (3.5) < 0.01 Behavioral and physiological results MABP prechallenge (SD) 115.2 (15.3) 109.2 (6.2) 0.393 MABP challenge (SD) 116.3 (18.0) 116 (7.3) 0.967 MABP postchallenge (SD) 109.8 (17.5) 116 (20.6) 0.588 Mean HR prechallenge (SD) 80.7 (21.6) 76.5 (12.3) 0.690 Mean HR challenge (SD) 81.5 (19.6) 78.7 (12.1) 0.769 Mean HR postchallenge (SD) 78.8 (18.2) 70.5 (10.1) 0.351 API prechallenge (SD)3.2 (3.2) 0.8 (1.0) 0.118 API challenge (SD) 22.3 (6.3 13.3 (5.6)) 0.026 API postchallenge (SD) 1.3 (1.8) 0.3 (0.5) 0.209 Legend:SD = standard deviation, HAS = Hamilton rating scale for anxiety, HDS = Hamilton rating scale for depression, MABP = mean arterial blood pressure, HR = heart rate, API = Acute Panic Inventory. The mean number of panic attacks in the two weeks preceding the scanning proce- dure decreased from 4.5 (1) to 0.3 (0.5) (mean (SD); see table 1; t = 8.7; df = 10; p < 0.000). Differences in blood pressure and heart rate with treatment No significant differences were seen in mean arterial blood pressure (MABP) and heart rate (HR) between the pretreatment and the posttreatment state (table 1). During the pentagastrin challenge no significant increase of MABP and HR was seen, although there was a trend visible, especially in the posttreatment condition (p-values between 0.06 and 0.10). Differences in symptomatology after pentagastrin injection following sertraline treatment In the pretreatment condition all PD patients experienced a panic attack according to DSM-IV criteria during pentagastrin challenge. After treatment only two of these 61 17355 Bosh B 22-10-2003 14:20 Pagina 62 The Anxious Brain Table 2 Coordinates and z-scores of peak voxels of brain regions where differences were observed in PD patients during rest and after pentagastrin challenge after 12-week succesful treatment Decreases posttreatment x-coord y-coord z-coord Z-score in rest parahippoc gyrus right 12 -46 -8 4,03 parahippoc gyrus right 32 -34 -10 2,66 parahippoc gyrus left -34 -32 -10 3,12 parahippoc gyrus left -22 -38 -8 2,69 thalamus 14 -6 12 4,13 hypothalamus 6 -4 -2 3,73 midbrain 8 -38 -16 5,23 ant cing gyrus left -4 40 0 2,51 ant cing gyrus right 6 34 2 2,36 Increases posttreatment x-coord y-coord z-coord Z-score in rest insula left -36 20 18 4,54 insula right 40 0 8 3,24 med frontal gyrus left -42 40 16 3,47 med frontal gyrus right 42 38 8 3,37 Decreases posttreatment x-coord y-coord z-coord Z-score during pentagastrin challenge parahippoc gyrus right 8 -30 -20 5,23 parahippoc gyrus right 12 -38 -10 3,94 parahippoc gyrus left -14 -22 -24 3,27 parahippoc gyrus left -16 -30 -20 3,12 basal ganglia/thalamus 16 -4 14 3,37 hypothalamus -4 -8 -4 3,39 temporal lobe right 28 -6 -34 3,36 temporal lobe left -18 4 -30 2,49 amygdala right 20 -4 -20 2,93 Decreases posttreatment x-coord y-coord z-coord Z-score during pentagastrin challenge orbitofrontal cortex right 16 32 -18 4,12 orbitofrontal cortex left -20 52 -18 3,51 ant cing gyrus ri 2 20 50 4 precentr gyrus ri 38 0 20 3,17 precentr gyrus le -40 -2 -20 2,52 Legend:The coordinates of maximal change of the main voxels according to the Talairach and Tournoux atlas (1988), after translation of the coordinates via the function mni2tal (Brett, 1999); Z-scores presented are signifi- cant with a P < 0.05 corrected with false discovery rate (Genovese et al, 2002). 62 17355 Bosh B 22-10-2003 14:20 Pagina 63 Partial normalization of rCBF pattern after succesful SSRI treatment of PD patients Fig. 4-1A Fig. 4-1B SPM glass brain showing the regions SPM glass brain showing the where in rest decreases were observed regions where in rest increases after 12-week succesful sertraline were observed after 12-week treatment succesful sertraline treatment patients experienced a panic attack according to DSM-IV criteria, although the num- ber of symptoms experienced during this panic attack were less than before, and they felt less anxious. Both before (t = 6.6; df = 5; p < 0.001) and after (t = 5.6; df = 5; p < 0.003) treatment the API score increased significantly during the pentagastrin challenge, and decreased again during the resting condition. As can be seen in table 1, after treatment with sertraline there was a diminished increase in the API score during the challenge procedure. Changes in rCBF in rest pretreatment versus posttreatment Figure 1 shows the differences in brain activation between the pretreatment and the posttreatment resting state (the last scan). Posttreatment rCBF decreased (fig. 4-1A) in the right parahippocampal gyrus, the right orbitofrontal cortex, the right anterior cingulate cortex, the right thalamus, the hypothalamus, the right basal ganglia, the midbrain and the pons, and increased (fig. 4-1B) in the precuneus bilaterally, and in the left medial and inferior frontal cortex. In table 2 the coordinates and z-scores of the peak voxels are shown. When we compared these data with the data of all sub- jects scanned pretreatment (N=16) with all subjects scanned posttreatment (N=9) a comparable rCBF pattern was seen. We used the final scan as the resting condition because anticipatory anxiety was pre- sent during the first scan (the prechallenge scan) in the PD patients, as described in an earlier paper (Boshuisen et al 2002). Changes in rCBF during pentagastrin challenge: pretreatment During pentagastrin challenge in the pretreatment condition we observed increases in rCBF of the inferior frontal gyrus, the right postcentral gyrus, the right thalamus, the left cerebellum and the left orbitofrontal gyrus (for coordinates of the peak voxels see table 2). Decreases in rCBF pretreatment during pentagastrin challenge 63