Psychiatry Research: Neuroimaging Section 98 Ž2000. 133]143
Increased left posterior parietal]temporal cortex activation after D-fenfluramine in women with panic disorder U
Jeffrey H. Meyera, , Richard Swinsonb, Sidney H. Kennedya , Sylvain Houlea , Gregory M. Browna a
The Centre for Addiction and Mental Health, Department of Psychiatry, Uni¨ ersity of Toronto, Clarke Di¨ ision, 250 College Street, Toronto, Ontario, Canada M5T 1R8 b Department of Psychiatry, McMaster Uni¨ ersity, Hamilton, Ontario, Canada Received 10 August 1999; received in revised form 31 January 2000; accepted 1 February 2000
Abstract It is unclear whether the functional changes found in panic disorder reflect disturbed physiology of particular neurotransmitters. One method of investigating altered neurotransmission is to assess regional brain activations in response to agonist challenges. D-Fenfluramine is a medication that induces neuronal release of serotonin. Using w15OxH2 O and positron emission tomography ŽPET., measurements of regional cerebral blood flow ŽrCBF. were done at t s y20, y5, q20 and q35 relative to the IV D-fenfluramine injection Ž t s 0. in nine panic-disordered and 18 healthy subjects. Subjects were otherwise healthy, right-handed, non-smoking and not taking psychotropic medication. w15OxH2 O PET scans were assessed with Statistical Parametric Mapping using individual global cerebral blood flow as a covariate. Comparisons of the Žbaseline . first two scans between healthy and panic-disordered subjects showed a decreased rCBF in the left posterior parietal]superior temporal cortex in the patient group. Fenfluramine-induced increases as defined by the last two scans minus the first two scans were compared between groups and a significantly greater increase in the same left posterior parietal]superior temporal region was found in panic-disordered subjects. Consistent with this finding, differences between the last two scans Žpost-fenfluramine. of the healthy and panic-disordered subjects showed an increased rCBF in the left superior temporal cortex in panic-disordered subjects. Functional pathology in the left parietal]superior temporal cortex in panic disorder may be related to abnormal modulation by serotonin. Q 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Anxiety disorder; Serotonin; Parietal cortex; Temporal cortex; Positron emission tomography; Regional cerebral blood flow
U
Corresponding author. Tel.: q1-416-535-8501; fax: q1-416-979-6821. E-mail address:
[email protected] ŽJ.H. Meyer. 0925-4927r00r$ - see front matter Q 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 9 2 5 - 4 9 2 7 Ž 0 0 . 0 0 0 4 8 - 2
134
J.H. Meyer et al. r Psychiatry Research: Neuroimaging 98 (2000) 133]143
1. Introduction In panic disorder, investigations using anatomical and functional measures report abnormalities reflecting both brainstem and cortex dysfunction. Brainstem dysfunction is implicated by susceptibility to panicogens such as lactate and CO2 ŽGorman et al., 1984; Liebowitz et al., 1984.. Cortex pathology is detectable by positron emission tomography ŽPET. and structural magnetic resonance ŽMRI. techniques. With w18 Fxfluorodeoxyglucose PET, decreased left inferior parietal cortex metabolism ŽNordahl et al., 1990., increased left hippocampus]parahippocampus metabolism ŽBisaga et al., 1998., and decreased right inferior parietal]superior temporal cortex metabolism were found in panic disorder ŽBisaga et al., 1998.. Abnormalities in the inferior parietal]superior temporal cortex have also been detected by MRI ŽOntiveros et al., 1989; Fontaine et al., 1990.. In lactate-sensitive patients, a decreased left-to-right ratio in parahippocampal cortex blood flow was found with w15 OxH2 O PET ŽReiman et al., 1986.. It is unclear whether the functional changes found in panic disorder are associated with neurotransmitter pathology in the same regions. One method of investigating altered neurotransmission is to assess regional brain activations in response to agonist challenges in patients and healthy subjects ŽGrasby et al., 1992; Kapur et al., 1994; Mann et al., 1996; Meyer et al., 1996, 1998; Siever et al., 1999.. In this study, we investigate the effects of an intravenous D-fenfluramine challenge upon regional cerebral blood flow ŽrCBF. in panicdisordered and healthy subjects. Subjects received w15 OxH2 O PET scans before and after a Žonce off. dose of intravenous D-fenfluramine. D-Fenfluramine is a medication that induces the neuronal release of serotonin ŽInvernizzi et al., 1986; Bonanno et al., 1994.. We chose to investigate serotonin agonist effects because several serotonin-related abnormalities have been found in patients with panic disorder: elevated anxiety and hormone changes in response to serotonin agonists; decreased platelet serotonin transporter sites; and increased anxiety in response to sero-
tonin reuptake inhibitors ŽNorman et al., 1986; Gorman et al., 1987; Butler et al., 1992; Targum 1992; George et al., 1995..
2. Materials and methods This study was approved by the review committee on the use of human subjects at the University of Toronto. Nine panic-disordered and 18 healthy women aged 18]38 years were recruited by advertisement. All were physically healthy, right-handed, non-smoking, and not pregnant. All subjects had been drug-free for the previous 6 weeks except for two panic-disordered subjects who were taking oral contraceptive pills. For each subject, written informed consent was obtained after the procedures had been fully explained. The healthy subjects have been previously described ŽMeyer et al., 1996, 1998.. Healthy subjects were screened using the Structured Clinical Interview for DSM-IV ŽSCID. ŽFirst et al., 1995.. A diagnosis of panic disorder was confirmed by the SCID, which was administered by a trained research assistant ŽFirst et al., 1995.. Six patients with panic disorder also had a diagnosis of agoraphobia. A score of less than 8 on the Hamilton Depression Rating Scale was required for entry of both healthy and panic-disordered subjects into the study ŽHamilton, 1960.. Each panicdisordered subject had routine tests to rule out common medical causes of depression Žthyroid function, electrolytes, complete blood cell count.. On the scan day, 0.3 mgrkg of intravenous ŽIV. D-fenfluramine was administered over 3 min. Relative to the IV D-fenfluramine injection Ž t s 0., w15 OxH2 O PET measurements of regional cerebral blood flow ŽrCBF. were done at t s y20, y5, q20 and q35 min. In vivo microdialysis studies demonstrate that detectable serotonin release occurs 20 min after locally applied D-fenfluramine ŽPuig de Parada et al., 1995; Rothman et al., 1998.. A trained research assistant was present throughout the scanning time and a psychiatrist ŽJ.M.. was present during IV D-fenfluramine administration and the subsequent 5 min. PET scanning was carried out with a GEMS-
J.H. Meyer et al. r Psychiatry Research: Neuroimaging 98 (2000) 133]143
Scanditronix PC-2048 15B head scanner. Subjects received a bolus intravenously of 60 mCi of w15 OxH2 O. Data acquisition began when the activity threshold exceeded 8000 countsrs, which is approximately 30 s after the bolus. Data were taken in one 60-s frame for each scan. A 68 Ge transmission scan was used to correct attenuation. To standardize cognitive activity, subjects did a visual continuous performance task ŽCPT. during scanning ŽMeyer et al., 1996, 1998.. To maximize the reliability of the CPT-related activation, the CPT began 20 s before the bolus ŽHurtig et al., 1994.. w15 OxH2 O PET scans were assessed with Statistical Parametric Mapping ŽSPM 95., version 1995 ŽFriston et al., 1994, 1995a,b.. First, on an individual subject basis, SPM realigns images to the same orientation. Then SPM has a normalization step in which realigned images are transformed and deformed into a common space. Then the images are spatially smoothed with a Gaussian filter. In this study the Gaussian filter chosen was 12 mm Žfull width at half maximum.. Then with these transformedrdeformed images, SPM generates an analysis at each voxel, which assigns a P-value to each voxel. This can be referred to as a probability map. In this study the threshold for display within the probability map was set at PF 0.001. The voxels displayed do not account for multiple comparisons. Given the known distributions of random probability values in threedimensional space, SPM can correct for multiple comparisons by assessing the number of adjacent suprathreshold voxels Žcluster significance. or by setting extremely low individual voxel value thresholds ŽFriston et al., 1994.. To avoid false positives, a conservative approach was used: regions displayed were considered only if the change was significant for cluster significance Ž P- 0.05.. The 1995 version of SPM was used because the spatial normalisation transformationrdeformation algorithms in this version were designed for PET scanners with a limited field of view in the z-axis that does not include the full brain. All analyses used global cerebral blood flow as a covariate. Differences in regional activations examined were relative to global brain activations.
135
Each scan was weighted equally for all comparisons. To detect panic attacks, the diagnostic symptom questionnaire ŽDSQ. ŽSanderson et al., 1989. was used before and after fenfluramine was given. Also, patients were asked to report if they were experiencing panic symptoms. To monitor ongoing mood and anxiety changes, the activation-euphoria-dysphoria-physical symptoms scale ŽAEDP. was done at t s y30, y15, q10 and q25 min relative to the IV D-fenfluramine injection Ž t s 0.. The AEDP scale has subscales for anxiety and physical symptoms, and it has been previously demonstrated to be sensitive to both intravenous D-fenfluramine and the effects of stimulants ŽVan Kammen and Murphy, 1975; Meyer et al., 1996..
3. Results The mean age of panic-disordered and healthy subjects was 27.5 ŽS.D.s 4.6., and 25.9 ŽS.D.s 5.3. years, respectively Žno significant differences, t26 s 0.8, Ps 0.43.. Two panic-disordered, and three healthy subjects were in the luteal phase of the menstrual cycle, and the remaining subjects were in the follicular phase Žno significant differences, Fisher’s Exact Test, Ps 0.46.. First, differences in baseline rCBF between panic-disordered and healthy subjects were analyzed using SPM ŽFriston et al., 1994, 1995a,b.. Baseline differences were defined as the differences between the first two scans of the healthy and the first two scans of the panic-disordered subjects. In panic-disordered subjects, lower rCBF was found in the posterior, left parietal]superior temporal cortex Žsee Fig. 1.. Peak Talairach coordinates ŽTalairach and Tournoux, 1988., Z score, number of 2 = 2 = 4 mm voxels, P value corrected for number of voxels and P value for corrected peak voxel were xs y44, y s y62, z s 28; Z s 5.57; n s 1607 voxels; P Žcluster size corrected. - 0.001; P Žpeak voxel corrected. 0.001. There was also lower rCBF in the cingulate gyrus Žsee Fig. 1 baseline .. Peak Talairach coordinates ŽTalairach and Tournoux, 1988., Z score, number of 2 = 2 = 4 mm voxels, P value cor-
136
J.H. Meyer et al. r Psychiatry Research: Neuroimaging 98 (2000) 133]143
rected for number of voxels and P value for corrected peak voxel were xs 8, y s y4, z s 32; Z s 4.91; n s 1607; P Žsize corrected. - 0.001, P
Žpeak voxel corrected. s 0.005. No regions showed significantly greater rCBF in the panic-disordered patients relative to the healthy subjects.
Fig. 1. Significant differences in rCBF between panic-disordered ŽPD. and healthy subjects: at baseline, with fenfluramine-induced changes and at post-fenfluramine.
J.H. Meyer et al. r Psychiatry Research: Neuroimaging 98 (2000) 133]143
Next, changes in rCBF with respect to intravenous D-fenfluramine were examined. Increases in rCBF were defined as the last two scans minus the first two scans in the paradigm. On the basis of analysis with SPM 95 ŽFriston et al., 1994, 1995a,b., increases in panic-disordered and healthy subjects were found. As previously reported ŽMeyer et al., 1996, 1998., the healthy group had bilateral increases in medial frontal cortex, including the anterior cingulate gyrus Žsee Table 1.. In the panic-disordered group, increases were found bilaterally in medial frontal cortex and cingulate cortex as well as the left dorsolateral prefrontal cortex Žsee Table 1.. Decreases in rCBF were defined as the first two scans minus the last two scans in the paradigm. In healthy subjects, there was decreased rCBF bilaterally in the thalamus and posterior temporal cortex Žsee
137
Table 1.. In panic-disordered subjects, there was decreased rCBF in the right posterior temporal cortex only Žsee Table 1.. Although the data in Table 1 would suggest that there was an increased rCBF after D-fenfluramine in the left posterior temporal cortex in panic-disordered subjects located proximal to a decrease in rCBF in healthy subjects, the true test for a difference between groups is to contrast the change in rCBF with respect to intavenous Dfenfluramine between panic-disordered and healthy subjects. With this contrast approach, a relatively greater fenfluramine-induced increase in rCBF in panic-disordered as compared to healthy subjects was found in the left posterior inferior parietal]superior temporal cortex. Peak Talairach coordinates ŽTalairach and Tournoux, 1988., Z score, number of 2 = 2 = 4 mm voxels, P
Table 1 Changes in rCBF after intravenous D-fenfluramine within healthy and panic disorder groups Change
Group
Region
Brodmann area
Side
x Increase
Healthy
Panic disorder
Anterior cingulate
Healthy
Cluster P-value
z
16 16 y22 y20
34 34 38 24
20 0 4 28
4.06 3.95 4.94 4.15
594
0.016
777
0.004
12 12
y16 y2
32 32
4.71 4.57
1428
- 0.001
24r32r23
Bilateral Right ) left
Dorsolateral prefrontal cortex
9
Left
y30
y2
40
4.98
1788
- 0.001
37r19r posterior21
Left
y24 y40
y64 y26
32 y4
4.02 3.89
816
0.003
37r20r21
Bilateral
52 62 y58 y56 y4
y58 I28 y52 y48 y14
8 y4 y4 y16 4
4.07 3.84 4.93 4.91 5.77
577
0.018
483
0.037
1020
0.001
60
y30
0
4.49
2409
- 0.001
Posterior temporal cortex Thalamus
Panic disorder
Bilateral
y
Extent Žvoxels.
Cingulate
Occipital and posterior temporal cortex Decrease
24r32
Z score
Talairach coordinates
Posterior temporal cortex
Bilateral 20r anterior21
Right
138
J.H. Meyer et al. r Psychiatry Research: Neuroimaging 98 (2000) 133]143
value for number of voxels by chance and P value for corrected significance of the peak voxel were xs y42, y s y62, zs 28; Z s 4.34; n s 626; P Žvoxel cluster . s 0.02; P Žpeak voxel corrected. s 0.043 Žsee Fig. 1.. Decreased rCBF with respect to fenfluramine was compared and no significant decreases in rCBF in panic-disordered subjects as compared to healthy subjects were found. Then, differences in post-fenfluramine rCBF between panic-disordered and healthy subjects were investigated with SPM 95 ŽFriston et al., 1994, 1995a,b. . These differences were defined as being between the last two scans of the healthy and the last two scans of the panic-disordered subjects. There was a significantly greater rCBF in the left superior temporal cortex in panic disordered subjects as compared to healthy subjects. Peak Talairach coordinates ŽTalairach and Tournoux, 1988., Z score, number of 2 = 2 = 4 mm voxels, P value corrected for number of voxels and P value for corrected peak voxel were x s y48, y s y54, zs 12; Z s 5.23; n s 183; P Žsize corrected. s 0.02; P Žpeak voxel corrected. s 0.001 Žsee Fig. 1.. There was also significantly greater rCBF in the posterior cingulate gyrus and cuneus. Peak Talairach coordinates ŽTalairach and Tournoux, 1988., Z score, number of 2 = 2 = 4 mm voxels, P value corrected for number of voxels and P value for corrected peak voxel were x s y2, y s y62, z s 8; Z s 5.24; n s 222; P Žsize corrected. s 0.01; P Žpeak voxel corrected. s 0.001 Žsee Fig. 1.. The two patients taking oral contraceptives ranked 7th and 8th among patients with respect to the relative increase at the peak voxel in left superior temporal cortex. Hence, their data do not account for the finding present in panic disorder patients. Whether phase of menstrual cycle was related to the peak voxel change was also examined. Among the combined set of healthy and panic-disordered patients, repeated measures analysis of variance showed no effect of luteal phase Ž F3,21 s 0.33, Ps 0.81. or diagnosis by luteal phase Ž F3,21 s 0.56, Ps 0.65.. Given that in panic-disordered subjects we found the pre-fenfluramine rCBF lower in the left posterior parietal superior temporal cortex, a
post-fenfluramine rCBF greater in left superior temporal cortex, and a relative increase in the posterior parietal]superior temporal cortex, it was of interest to understand whether there was overlap amongst all three regions. Use of a high supravoxel threshold for the statistical analysis in SPM avoids false positives but also limits the size of detectable regional differences. To better assess the extent of these regional findings, we decided to reexamine rCBF for these contrasts using an individual voxel threshold of P- 0.05. Under less stringent conditions, these analyses Žbaseline healthy vs. panic, rCBF less in panic Ž n s 1949 voxels.; post-fenfluramine healthy vs. panic, rCBF greater in panic after fenfluramine Ž n s 388 voxels.; and pre- vs. post-fenfluramine, greater increase in panic Ž n s 1474 voxels.. had a common overlap over the left superior temporal cortex region within Brodmann’s areas 37, 39, 22 and 21. Whether rCBF changes differently over time between scan 1 and 2 depending upon whether there is a diagnosis of panic disorder was also investigated using SPM 95 ŽFriston et al., 1994, 1995a,b. . No differences in rCBF change between healthy and panic-disordered groups were found. SPM 95 was also used to assess whether rCBF changes differently between scans 3 and 4 depending upon whether there is a diagnosis of panic disorder. No such differences were found. Repeated measures analysis of variance for the response rate of the cognitive performance task found that the effects of diagnosis, as well as the interaction of diagnosis and repetition, were all non-significant with respect to the peak voxel value Ždiagnosis: F1,52 s 1.3, Ps 0.26; repeated performance = diagnosis: F3,50 s 1.6, Ps 0.20.. Next the relationship between the peak voxel value representing the differential change Žincrease. in panic-disordered patients after Dfenfluramine, and subscale scores on the activation-euphoria-dysphoria-physical symptom scores was examined. The peak voxel value corresponding to xs y42, y s y62 and z s 28, divided by the mean voxel value, was found for each smoothed normalized image from each panicdisordered patient Ž n s 36 images.. The peak voxel was chosen on smoothed normalized images
J.H. Meyer et al. r Psychiatry Research: Neuroimaging 98 (2000) 133]143
because it is representative of a region of 18 = 18 = 20 mm surrounding the peak Ž r s 0.95, P0.001, Pearson; r s 0.93, P- 0.001, Spearman.. A multiple stepwise linear regression was done with peak voxel values as the dependent variable and subscale scores prior to each scan representing the predictor variables. It was found that the euphoria subscale best predicted peak voxel value Ž F1,34 s 11.49, Ps 0.002, r2 s 0.25.. The anxiety subscale was highly negatively correlated with the euphoria subscale Ž t34 s y3.5, Ps 0.001.. The Spearman rank correlation coefficient between the peak voxel and the euphoria subscale was also significant and showed a negative correlation Ž rs s y0.502, Ps 0.002.. As defined by the presence of four or more symptoms rated with a severity of 4 out of 8 or greater on the diagnostic symptom questionnaire ŽDSQ., five of nine panic-disordered patients and two of 18 healthy patients experienced a panic attack after intravenous D-fenfluramine. All subjects experiencing a panic attack as defined by the DSQ reported such during or immediately after administration of D-fenfluramine and all panic attacks subsided within the 5 min that the psychiatrist ŽJ.M.. was present. Significantly more people with panic disorder had a panic attack ŽFisher’s Exact Test, Ps 0.023.. Within panic-disordered subjects, the correlation of the presence of a panic attack and the change in peak voxel activity Žscans 3 and 4 minus scans 1 and 2. was not significant Ž r s 0.52, Ps 0.15, Spearman; r s 0.54, Ps 0.13, Pearson.. We also examined how the activationreuphoriardysphoriarphysical symptoms scale differed between patients and subjects over time. A repeated measures analysis of variance was used to assess the interaction of diagnosis with each of the six symptom subscales. Without correcting for multiple comparisons, the sleepiness subscale was significantly different between groups over time Ž F3,23 s 3.7, Ps 0.025.. No significant change on the sleepiness subscale was found within patients Ž F3,15 s 2.6, Ps 0.09. or within control subjects Ž F3,6 s 1.8, Ps 0.24. alone. With rigorous application of SPM criteria for cluster significance Žsee Section 2., our results
139
appeared consistent with a previous report of regional metabolic changes in panic disorder ŽNordahl et al., 1990.. There are reports of abnormal rCBF and brain metabolism in other brain regions in patients with panic disorder such as the right inferior parietal]superior temporal cortex ŽBisaga et al., 1998. and the hippocampus]parahippocampus region ŽReiman et al., 1986; Bisaga et al., 1998.. We also examined our data with respect to these other reports using less rigorous criteria. Thus, we decided to compare rCBF in these other brain regions between groups using an individual voxel threshold of P- 0.05 without requiring cluster significance. We found bilateral decreases in parahippocampal rCBF that were spatially larger on the left: peak Talairach coordinates, Z score, number of 2 = 2 = 4 mm voxels were x s y26, y s y24, z s y16, 84 and xs 20, y s y24, z s y16, 24. No decreases were found in the right inferior temporal cortex of panicdisordered patients.
4. Discussion The major finding is that the posterior left parietal]superior temporal cortex has a greater increase in activity after D-fenfluramine in panicdisordered subjects than in healthy subjects. This region initially has lower relative rCBF in panicdisordered subjects as compared to healthy subjects. After D-fenfluramine there is greater activity in the left superior temporal cortex relative to the rest of the brain in panic-disordered patients. Such findings were not present in our previous study, which found no difference between depressed and healthy subjects with respect to the effects of intravenous D-fenfluramine upon rCBF ŽMeyer et al., 1998.. When agonist challenge regional cerebral blood flowrbrain metabolism studies are done, the differences between groups can be interpreted in two major ways. One is that the difference between groups is attributable to a neurochemical difference in sensitivity to the agonist in the region found. The second is that the difference between groups is attributable to differences in behavior during scanning as a result of the ago-
140
J.H. Meyer et al. r Psychiatry Research: Neuroimaging 98 (2000) 133]143
nist. To minimize the second, one attempts to obtain the same behavior during both scanning periods. This is done by choosing an easy task that both groups perform at ceiling levels during repeated scanning. Typical tasks are either the resting state condition or the continuous performance test. With a continuous performance task ŽCPT., the behavior of the subject may be monitored during scanning; however, for long scanning periods a CPT may be impractical due to fatigue. Scanning times for w15 OxH2 O PET are short, being 1 min; hence a CPT is feasible. We chose a continuous performance task that was performable at ceiling levels and had an outcome performance that could be measured ŽMeyer et al., 1996, 1998.. In the current study the performance of the CPT had no relationship to diagnosis. Hence, the best interpretation of our results is that they are secondary to a neurochemical difference in the brain of panic-disordered patients. The left posterior parietal]superior temporal cortex could have a role in panic disorder. Electrical stimulation of the posterior parietal superior temporal cortex is associated with some symptoms of panic attacks such as vertigo, nausea, and suffocation ŽPenfield and Jasper, 1954.. Seizures within this region have also been associated with the generation of panic attacks ŽPenfield and Jasper, 1954; Swartz, 1995; Weilburg et al., 1995.. A previous w18 Fxfluorodeoxyglucose PET study of patients with panic disorder found decreased metabolism in the left inferior parietal cortex ŽNordahl et al., 1990.. Further study of rCBF changes in this region during experiences of nausea, vertigo and suffocation could clarify the relationship between specific symptoms and the activity of this region. It it unlikely that the present finding reflects the state of being in a full panic attack because panic attacks, when present, were over within 5 min of fenfluramine administration and the post-fenfluramine PET scans were done at 25 and 40 min after fenfluramine administration. Also, this region is not known to have altered rCBF during the presence of panic attacks ŽStewart et al., 1988; Javanmard et al., 1999.. The relationship between serotonin and activation in the left posterior parietal superior tem-
poral cortex requires further investigation. The rCBF change found may be a direct consequence of the release of serotonin in this region: A tonically decreased release of serotonin in this region could create a low activation at baseline, and increased sensitivity to fenfluramine-induced serotonin release. It is also possible that abnormal coupling of serotonin receptors to second messenger systems could create functional abnormalities. While chronic manipulations of serotonin do not necessarily relate to acute change, it is interesting that decreased left inferior parietal cortex metabolism has been reported to normalize in imipramine-treated panic-disordered subjects ŽNordahl et al., 1998.. Another explanation is that the activation found is secondary to serotonin release in other brain structures. Even chronically decreased release of serotonin in other Žsmall. brain structures could have secondary downstream effects upon baseline activation in the posterior parietal superior temporal cortex. PET studies of serotonin receptors ŽMeyer et al., 1999a. using parametric binding potential maps ŽGunn et al., 1997; Meyer et al., 1999b. could resolve whether serotonin receptor binding or affinity abnormalities occur in the posterior parietal superior temporal cortex or in other structures. This study did not use a placebo; hence it is possible that the posterior left parietal]superior temporal cortex activation could be induced by any anxiety-provoking agonist challenge. This is unlikely because previous activation studies using lactate, and yohimbine did not report the present finding ŽStewart et al., 1988; Woods et al., 1988.. Our finding is not attributable to an order effect because no such difference between panicdisordered and healthy subjects was found when comparing changes between the first two scans Ždone before D-fenfluramine administration. or changes between the last two scans Ždone after D-fenfluramine admistration.. The design of the current study has some limitations often present in PET work using challenges and rCBF or metabolic tracers. The measure of difference in rCBF is relative; thus, in theory, the absolute change within the left posterior parietal]superior temporal cortex in panic
J.H. Meyer et al. r Psychiatry Research: Neuroimaging 98 (2000) 133]143
disorder could be a greater increase in rCBF compared to the rest of the brain or a milder decrease in rCBF compared to a greater decrease in the rest of the brain. However, it is known that absolute global rCBF increases after intravenous D-fenfluramine ŽSmith et al., 1998.; hence this finding is most likely a greater absolute increase relative to the rest of the brain. Changes in regional cerebral blood flow are highly associated with changes in neuronal activity. There is also the question as to whether D-fenfluramine directly influences cerebral blood flow within localized blood vessels and, more importantly, whether such an effect would alter the interpretation of our findings. The locations of the activations found after D-fenfluramine indicate that the activations are not direct vascular effects of fenfluramine because these locations do not reflect a vascular pattern of supply and they are similar in location to our previous oral D,L-fenfluramine challenge ŽKapur et al., 1994. which used w18 FxFDG, a measure of glucose metabolism, as a radiotracer. There is no evidence that there are localized differences in vasoconstriction by serotonin agonists. Of cerebral arteries assessed in humans: basilar ŽParsons et al., 1989., middle cerebral ŽHamel and Bouchard, 1991., and smaller arteries within the temporal]parietal cortex ŽJansen et al., 1993., the rank potency of vasoconstriction by specific agonists is similar. D-Fenfluramine may have a global effect upon cerebral vasculature; however, statistical parametric mapping ŽFriston et al., 1994, 1995a,b. accounts for this by using global rCBF as a covariate. In summary the posterior left parietal]superior temporal cortex has a greater increase in activity after D-fenfluramine in panic-disordered subjects as compared to healthy subjects. This region has been previously implicated in panic disorder, and pathology in this region is associated with panic symptoms. Whether serotonin hypersensitivity andror a decreased tonic release of serotonin occurs within the posterior left parietal]superior temporal cortex will require further investigation.
Acknowledgements We acknowledge the support of the Medical
141
Research Council of Canada ŽMT-13137.. Dr Jeffrey Meyer received funding from the Medical Research Council of Canada and the National Alliance for Research on Schizophrenia and Depression. We acknowledge and thank research assistants Fiona Downie, Shari Rafi-Tari, Beata Eisfeld and Corey Jones, technicians Doug Hussey, Kevin Cheung, Armando Garcia and Li Jin and radiochemists Dr Alan Wilson and Dr Jean DaSilva. References Bisaga, A., Katz, J.L., Antonini, A., Wright, C.E., Margouleff, C., Gorman, J.M., Eidelberg, D., 1998. Cerebral glucose metabolism in women with panic disorder. American Journal of Psychiatry 155, 1178]1183. Bonanno, G., Fassio, A., Severi, P., Ruelle, A., Raiteri, M., 1994. Fenfluramine releases serotonin from human brain nerve endings by a dual mechanism. Journal of Neurochemistry 63, 1163]1166. Butler, J., Halloran, A., Leonard, B.E., 1992. The Galway Study of Panic Disorder. II: Changes in some peripheral markers of noradrenergic and serotonergic function in DSM III-R panic disorder. Journal of Affective Disorders 26, 89]99. First, M., Spitzer, R., Williams, J., Gibbon, M., 1995. Structured Clinical Interview for DSM-IV Axis I Disorders. Patient Edition SCID-P version 2. Biometrics Research, New York. Fontaine, R., Breton, G., Dery, R., Fontaine, S., Elie, R., 1990. Temporal lobe abnormalities in panic disorder an MRI study. Biological Psychiatry 27, 304]310. Friston, K., Frith, C., Liddle, P., Frackowiak, R., 1994. Assessing the significance of focal activations using their spatial extent. Human Brain Mapping 2, 214]220. Friston, K., Ashburner, J., Poline, J., Frith, C., Heather, J., Frackowiak, R., 1995a. Spatial realignment and normalization of images. Human Brain Mapping 2, 165]169. Friston, K., Holmes, A., Worsley, K., Poline, J., Frith, C., Frackowiak, R., 1995b. Statistical parametric maps in functional imaging: a general linear approach. Human Brain Mapping 2, 189]210. George, D.T., Nutt, D.J., Rawlings, R.R., Phillips, M.J., Eckardt, M.J., Potter, W.Z., Linnoila, M., 1995. Behavioral and endocrine responses to clomipramine in panic disorder patients with or without alcoholism. Biological Psychiatry 37, 112]119. Gorman, J.M., Askanazi, J., Liebowitz, M.R., Fyer, A.J., Stein, J., Kinney, J.M., Klein, D.F., 1984. Response to hyperventilation in a group of patients with panic disorder. American Journal of Psychiatry 141, 857]861. Gorman, J.M., Liebowitz, M.R., Fyer, A.J., Goetz, D., Campeas, R.B., Fyer, M.R., Davies, S.O., Klein, D.F., 1987. An open trial of fluoxetine in the treatment of panic attacks. wPublished erratum appears in Journal of Clinical Psy-
142
J.H. Meyer et al. r Psychiatry Research: Neuroimaging 98 (2000) 133]143
chopharmacology 1988 8 Ž1. 13x. Journal of Clinical Psychopharmacology 7, 329]332. Grasby, P.M., Friston, K.J., Bench, C., Cowen, P.J., Frith, C.D., Liddle, P.F., Frackowiak, R.S., Dolan, R.J., 1992. Effect of the 5-HT1A partial agonist buspirone on regional cerebral blood flow in man. Psychopharmacology 108, 380]386. Gunn, R.N., Lammertsma, A.A., Hume, S.P., Cunningham, V.J., 1997. Parametric imaging of ligand]receptor binding in PET using a simplified reference region model. Neuroimage 6, 279]287. Hamel, E., Bouchard, D., 1991. Contractile 5-HT1 receptors in human isolated pial arterioles: correlation with 5-HT1D binding sites. British Journal of Pharmacology 102, 227]233. Hamilton, M., 1960. A rating scale for depression. Journal of Neurology, Neurosurgery and Psychiatry 23, 56]62. Hurtig, R.R., Hichwa, R.D., O’Leary, D.S., Boles Ponto, L.L., Narayana, S., Watkins, G.L., Andreasen, N.C., 1994. Effects of timing and duration of cognitive activation in w15 Oxwater PET studies. Journal of Cerebral Blood Flow and Metabolism 14, 423]430. Invernizzi, R., Berettera, C., Garattini, S., Samanin, R., 1986. D- and L-isomers of fenfluramine differ markedly in their interaction with brain serotonin and catecholamines in the rat. European Journal of Pharmacology 120, 9]15. Jansen, I., Olesen, J., Edvinsson, L., 1993. 5-Hydroxytryptamine receptor characterization of human cerebral middle meningeal and temporal arteries: regional differences. Acta Physiologica Scandinavica 147, 141]150. Javanmard, M., Shlik, J., Kennedy, S.H., Vaccarino, F.J., Houle, S., Bradwejn, J., 1999. Neuroanatomic correlates of CCK-4-induced panic attacks in healthy humans: a comparison of two time points. Biological Psychiatry 45, 872]882. Kapur, S., Meyer, J., Wilson, A.A., Houle, S., Brown, G.M., 1994. Modulation of cortical neuronal activity by a serotonergic agent: a PET study in humans. Brain Research 646, 292]294. Liebowitz, M.R., Fyer, A.J., Gorman, J.M., Dillon, D., Appleby, I.L., Levy, G., Anderson, S., Levitt, M., Palij, M., Davies, S.O., 1984. Lactate provocation of panic attacks. I. Clinical and behavioral findings. Archives of General Psychiatry 41, 764]770. Mann, J.J., Malone, K.M., Diehl, D.J., Perel, J., Cooper, T.B., Mintun, M.A., 1996. Demonstration in vivo of reduced serotonin responsivity in the brain of untreated depressed patients. wPublished erratum appears in American Journal of Psychiatry 1996 Apr 153 Ž4. 588 see commentsx. American Journal of Psychiatry 153, 174]182. Meyer, J., Kapur, S., Houle, S., DaSilva, J., Owczarek, B., Brown, G., Wilson, A., Kennedy, S., 1999a. Prefrontal cortex 5-HT2 receptors in depression: a w18 Fx setoperone PET imaging study. American Journal of Psychiatry 156, 1029]1034. Meyer, J.H., Gunn, R.N., Myers, R., Grasby, P.M., 1999b. Assessment of spatial normalization of PET ligand images using ligand-specific templates. Neuroimage 9, 545]553.
Meyer, J.H., Kapur, S., Wilson, A.A., DaSilva, J.N., Houle, S., Brown, G.M., 1996. Neuromodulation of frontal and temporal cortex by intravenous D-fenfluramine: an w15 OxH2 O PET study in humans. Neuroscience Letters 207, 25]28. Meyer, J.H., Kennedy, S., Brown, G.M., 1998. No effect of depression on w15 OxH2 O PET response to intravenous Dfenfluramine. American Journal of Psychiatry 155, 1241]1246. Nordahl, T.E., Semple, W.E., Gross, M., Mellman, T.A., Stein, M.B., Goyer, P., King, A.C., Uhde, T.W., Cohen, R.M., 1990. Cerebral glucose metabolic differences in patients with panic disorder. Neuropsychopharmacology 3, 261]272. Nordahl, T.E., Stein, M.B., Benkelfat, C., Semple, W.E., Andreason, P., Zametkin, A., Uhde, T.W., Cohen, R.M., 1998. Regional cerebral metabolic asymmetries replicated in an independent group of patients with panic disorders. Biological Psychiatry 44, 998]1006. Norman, T.R., Judd, F.K., Gregory, M., James, R.H., Kimber, N.M., McIntyre, I.M., Burrows, G.D., 1986. Platelet serotonin uptake in panic disorder. Journal of Affective Disorders 11, 69]72. Ontiveros, A., Fontaine, R., Breton, G., Elie, R., Fontaine, S., Dery, R., 1989. Correlation of severity of panic disorder and neuroanatomical changes on magnetic resonance imaging. Journal of Neuropsychiatry and Clinical Neuroscience 1, 404]408. Parsons, A.A., Whalley, E.T., Feniuk, W., Connor, H.E., Humphrey, P.P., 1989. 5-HT1-like receptors mediate 5-hydroxytryptamine-induced contraction of human isolated basilar artery. British Journal of Pharmacology 96, 434]440. Penfield, W., Jasper, H., 1954. Epilepsy and the Functional Anatomy of the Human Brain. Little Brown, Boston. Puig de Parada, M., Parada, M.A., Pothos, E., Hoebel, B.G., 1995. D-Fenfluramine but not D-norfenfluramine uses calcium to increase extracellular serotonin. Life Sciences 56, L415]L420. Reiman, E.M., Raichle, M.E., Robins, E., Butler, F.K., Herscovitch, P., Fox, P., Perlmutter, J., 1986. The application of positron emission tomography to the study of panic disorder. American Journal of Psychiatry 143, 469]477. Rothman, R.B., Elmer, G.I., Shippenberg, T.S., Rea, W., Baumann, M.H., 1998. Phentermine and fenfluramine. Preclinical studies in animal models of cocaine addiction. Annals of the New York Academy of Sciences 844, 59]74. Sanderson, W.C., Rapee, R.M., Barlow, D.H., 1989. The influence of an illusion of control on panic attacks induced via inhalation of 5.5% carbon dioxide-enriched air. Archives of General Psychiatry 46, 157]162. Siever, L.J., Buchsbaum, M.S., New, A.S., Spiegel-Cohen, J., Wei, T., Hazlett, E.A., Sevin, E., Nunn, M., Mitropoulou, V., 1999. D,L-Fenfluramine response in impulsive personality disorder assessed with w18 Fxfluorodeoxyglucose positron emission tomography. Neuropsychopharmacology 20, 413]423. Smith, D.F., Poulsen, P.H., Ishizu, K., Sakoh, M., Hansen, S.B., Gee, A.D., Bender, D., Gjedde, A., 1998. Quantitative PET analysis of regional cerebral blood flow and glucose
J.H. Meyer et al. r Psychiatry Research: Neuroimaging 98 (2000) 133]143 and oxygen metabolism in response to fenfluramine in living porcine brain. Journal of Neuroscience Methods 86, 17]23. Stewart, R.S., Devous, M.D., Rush, A.J., Lane, L., Bonte, F.J., 1988. Cerebral blood flow changes during sodium]lactateinduced panic attacks. American Journal of Psychiatry 145, 442]449. Swartz, B., 1995. Unusual seizure types. Seminars in Neurology 15, 151]157. Talairach J., Tournoux P., 1988. Co-Planar Stereotaxic Atlas of the Human Brain. Thieme Medical, New York. Targum, S.D., 1992. Cortisol response during different anxiogenic challenges in panic disorder patients. Psychoneuroendocrinology 17, 453]458.
143
Van Kammen, D.P., Murphy, D.L., 1975. Attenuation of the euphoriant and activating effects of D- and L-amphetamine by lithium carbonate treatment. Psychopharmacologia 44, 215]224. Weilburg, J.B., Schachter, S., Worth, J., Pollack, M.H., Sachs, G.S., Ives, J.R., Schomer, D.L., 1995. EEG abnormalities in patients with atypical panic attacks. Journal of Clinical Psychiatry 56, 358]362. Woods, S.W., Koster, K., Krystal, J.K., Smith, E.O., Zubal, I.G., Hoffer, P.B., Charney, D.S., 1988. Yohimbine alters regional cerebral blood flow in panic disorder wletterx. Lancet 2, 678.
