Short communication

99m

Tc-ECD brain perfusion SPECT in hyperalgesic fibromyalgia

Eric Guedj1, David Taieb1, Serge Cammilleri1, David Lussato1, Catherine de Laforte1, Jean Niboyet2, Olivier Mundler1
1

Service Central de Biophysique et de Mdecine Nuclaire, Assistance Publique des Hpitaux de Marseille, Centre Hospitalo-Universitaire de la Timone, 264 rue Saint Pierre, 13385 Marseille Cedex 05, France 2 Unit dEtude et de Traitement de la Douleur, Clinique La Phocanne, 143 route des 3 Lucs, 13012 Marseille, France Received: 9 March 2006 / Accepted: 11 May 2006 Springer-Verlag 2006

Abstract. Purpose: Neuro-imaging studies with 99mTcHMPAO SPECT in fibromyalgia (FM) patients have reported only limited subcortical hypoperfusion. 99mTcECD SPECT is known to provide better evaluation of areas of high cerebral blood flow and regional metabolic rate. We evaluated a homogeneous group of hyperalgesic patients with FM using 99mTc-ECD SPECT. The aim of this study was to investigate brain processing associated with spontaneous pain in FM patients. Methods: Eighteen hyperalgesic FM women (mean age 49 years, range 2563 years; American College of Rheumatology criteria) and ten healthy women matched for age were enrolled in the study. A voxel-by-voxel group analysis was performed using SPM2 (p<0.05, corrected for multiple comparisons). Visual Analogue Scale score for pain was 824 at the time of the SPECT study. Results: Compared with control subjects, we observed individual brain SPECT abnormalities in FM patients, confirmed by SPM2 analysis, with hyperperfusion of the somatosensory cortex and hypoperfusion of the frontal, cingulate, medial temporal and cerebellar cortices. Conclusion: In the present study, performed without noxious stimuli in hyperalgesic FM patients, we found significant hyperperfusion in regions of the brain known to be involved in the sensory dimension of pain processing and significant hypoperfusion in areas assumed to be associated with the affective-attentional dimension. As current pharmacological and non-pharmacological therapies act differently on the two components of pain, we hypothesise that SPECT could be a valuable and readily

available tool to guide individual therapeutic strategy and provide objective follow-up of pain processing recovery under treatment.
99m

Keywords: Pain Fibromyalgia Brain SPECT Tc-ECD Voxel-based analysis

Eur J Nucl Med Mol Imaging DOI 10.1007/s00259-006-0174-7

Introduction Fibromyalgia (FM) syndrome is a chronic pain condition characterised by widespread musculoskeletal aches and pain and stiffness, soft tissue tenderness, general fatigue and sleep disturbances, without a clinically demonstrable peripheral nociceptive cause [1]. Although a psychogenic cause was initially postulated, recent fMRI activation studies have clearly demonstrated global dysfunction of central pain processing, consolidating the hypothesis of central sensitisation. Similar painful pressure applied in patients and in controls did not result in activation of any common cerebral areas and showed greater effects in patients [2]. In contrast to these studies on painful stimuli [3], in resting subjects 99mTc-hexamethylpropylene amine oxime (HMPAO) single-photon emission computed tomography (SPECT) revealed only limited subcortical hypoperfusion, involving the thalamus, the caudate nucleus and the inferior pontine tegmentum [4, 5]. In this prospective study, we evaluated brain perfusion SPECT in a homogeneous group of hyperalgesic FM patients using 99mTc-ethyl cysteinate dimer (ECD). We performed a voxel-based analysis in comparison to a control group, matched for age and gender. To date, use of this radiotracer, known to provide better evaluation of high-blood flow regions and regional metabolic rate than 99mTc-HMPAO [6], has never been reported in FM patients. Under such conditions, we made the assumption that significant cerebral perfusion abnormalities could be demonstrated, evidencing altered cerebral processing associated with spontaneous pain in FM patients.

Eric Guedj ()) Service Central de Biophysique et de Mdecine Nuclaire, Assistance Publique des Hpitaux de Marseille, Centre Hospitalo-Universitaire de la Timone, 264 rue Saint Pierre, 13385 Marseille Cedex 05, France e-mail: eric.guedj@ap-hm.fr Tel.: +33-4-91385558, Fax: +33-4-91384769

European Journal of Nuclear Medicine and Molecular Imaging Vol. 00, No. 0, Month 20xx

Materials and methods

Patients and control subjects Eighteen consecutive hyperalgesic FM women (49 11 years, range 2563) who fulfilled the American College of Rheumatology (ACR) criteria [7] were enrolled in the study. All patients underwent a general medical assessment by one investigator to confirm the diagnosis. All patients failed to respond optimally to non-opioid analgesics or weak opioids and had to be managed in a pain management unit. Patients with psychiatric disease were excluded. No patient had any other significant medical illness. No change was made in treatment during the month preceding inclusion. No patient was receiving treatment with a strong opioid, tricyclic antidepressant, selective serotonin reuptake inhibitor, benzodiazepine or anticonvulsant agent. For comparison of imaging findings, a control group of ten women matched for age was also included (527 years, p=0.21, using the Mann-Whitney U test). All subjects provided informed consent according to institutional guidelines.

Fig. 1. Brain 99mTc-ECD SPECT abnormalities (anatomical con- " vention: right side = right hemisphere; left side = left hemisphere). a and b Examples of individual results. a Right centro-parietal hyperperfusion (frontal slice, patient no. 5). b Left to right: normal perfusion (axial slice, patient no. 4), diffuse hypoperfusion (axial slice, patient no. 9), frontal and left medial temporal hypoperfusion (frontal slice, patient no. 2; anatomical convention: right side = right hemisphere; left side = left hemisphere), inferior frontal hypoperfusion (axial slice, patient no. 8), medial frontal and cingulate hypoperfusion (axial slice, patient no. 8).c Voxel-by-voxel SPM2 group analysis (18 FM patients vs 10 healthy women). Anatomical localisation of regions of significant hyperperfusion (yellow-orange) and significant hypoperfusion (blue), projected onto the axial section of a normal MRI set spatially normalised into the standard SPM2 template (P voxel level <0.001; P cluster level <0.05, corrected for multiple comparisons)

SPECT protocol and statistical analysis All brain SPECT studies were performed before any change was made in therapy in the pain care unit. Patients were injected with 740 MBq of 99mTc-ECD (Neurolite, BMS) and placed at rest for 1 h, in quiet surroundings with their eyes closed. AVisual Analogue Scale (VAS) score for pain, evaluated in the immediate pre-injection period, was 824 (range 7590). SPECT image acquisitions were performed using a double-headed rotating gamma camera (ECAM, Siemens) equipped with a fan-beam collimator. Thirty-two projections of 40 s were collected per head, in a 128128 format. Tomographic 3D reconstruction was performed using a filtered backprojection algorithm (Butterworth filter of order 4 with a cut-off frequency of 0.4 cm1) and Changs attenuation correction. Each

brain SPECT study was first visually interpreted by two nuclear medicine physicians (E.G. and C.D.L.). A voxel-by-voxel group study was then performed using SPM2 (Welcome Department of Cognitive Neurology, University College, London, UK, running on Matlab 6.0 Mathworks Inc, Sherborn, MA, USA). Images were initially converted from the DICOM to the Analyze format using MRIcro (http://www.mricro.com), and transferred to SPM2. The data were then standardised with the Montreal Neurological Institute (MNI) atlas by using a 12-parameter affine transformation, followed by non-linear transformations and a trilinear interpolation. Dimensions of the resulting voxels were 222 mm. Standardised data were then smoothed by a Gaussian filter (FWHM 12 mm). FM and control groups were compared using the compare-populations one scan/ subject routine, which carries out a fixed-effects simple t test for each voxel. Global normalisation was performed using proportional scaling. The SPM{T} maps were initially obtained at a height threshold of p<0.001, then an extent threshold of 50 voxels was applied to obtain a statistical threshold corrected for multiple comparisons for the cluster (p<0.05). A complementary analysis was performed using small volume correction with a VOI sphere of

Table 1. Characteristics of FM hyperalgesic patients (n=18). SPECT images were assessed qualitatively by two independent observers (E.G. and C.D.L.) FM patient Age (yrs) Pain, VAS Visual interpretation Hypoperfusion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 60 56 56 52 34 63 55 49 31 50 49 52 44 62 25 30 58 47 85 85 90 80 80 90 75 85 75 80 80 85 80 80 80 80 85 80 Frontal and medial temporal (weak and prevailing on the left side) Frontal, cingulate and medial temporal (severe and bilateral) Frontal, cingulate and medial temporal (severe and bilateral) None None None Cingulate (weak and bilateral) Frontal, medial temporal and cingulate (severe and bilateral) Frontal, temporal and occipital (severe and bilateral) Frontal (weak and bilateral) Cingulate (weak and bilateral) None Frontal (weak and bilateral) Frontal (weak and bilateral) Frontal and medial temporal (weak and bilateral) Frontal and medial temporal (severe and bilateral) Frontal and medial temporal (severe and prevailing on the right side) Frontal (weak and bilateral) Hyperperfusion None None None None Right centro-parietal None None None None None None None Right centro-parietal None None None None None

European Journal of Nuclear Medicine and Molecular Imaging Vol. 00, No. 0, Month 20xx

European Journal of Nuclear Medicine and Molecular Imaging Vol. 00, No. 0, Month 20xx

10 mm radius, and a height threshold of p<0.05. MNI coordinates were finally converted into Talairach coordinates using the Talairach Daemon database (http://ric.uthscsa.edu/projects/talairachdaemon. html).

Results There was no disagreement between observers in the qualitative visual interpretation of brain SPECT. Areas of hypoperfusion were found in 14/18 patients. This hypoperfusion was bilateral and involved the frontal, cingulate and/or the medial temporal lobes. In three cases, SPECT images were interpreted as normal. In two cases, we found hyperperfusion in the right centroparietal lobe (SI and SII) corresponding to a pain syndrome dominating on the opposite side. Results are detailed in Fig. 1a and b and in Table 1. Statistical group analysis found a significant difference between SPECT findings in FM patients and control subjects (Fig. 1c, Table 2). The FM patient group had bilateral hyperperfusion of the centro-parietal lobe (BA1, BA2, BA3 and BA5), the superior parietal lobe dominating on the right side (BA7), and the left inferior parietal lobe (BA39 and BA40). The FM patient group also had bilateral hypoperfusion of the medial frontal lobe (BA10), the anterior cingulate (BA32), the posterior cingulate (BA29 and BA30), and cerebellar cortices and the medial temporal structures (amygdala and parahippocampal cortex).
Table 2. Regions of significant hyperperfusion and hypoperfusion in the FM group. Results are listed by clusters. k value, Z score, Talairach coordinates of peak voxel, Brodmann area (BA) and anatomic localisation are provided for each cluster. k value k Hyperperfusion 7,840 Z score 6.14 5.71 5.47 5.18 5.08 6.25 5.50 5.15 5.88 5.53 4.71 5.67 5.40 5.04 4.99 4.83 4.82 4.63 x (mm) 18 57 63 22 30 32 22 0 20 24 36 28 22 30 18 16 20 14

Small volume correction was performed in order to explore perfusion of the thalamus, the caudate nucleus, the putamen and the brain stem, showing hypoperfusion of the right and left pontine tegmentum (p<0.001; Z score of 5.88 and 4.56; Talairach coordinates: 20, 35, 27 and 20, 36, 30, respectively), the right and left thalamus (p=0.012 and 0.012; Z score of 2.25 and 2.24; Talairach coordinates: 12, 11, 13 and 12, 13, 14, respectively) and the right putamen (p=0.020; Z score of 2.04; Talairach coordinates: 28, 28, 5). Discussion Previous studies based on SPECT, PET and fMRI have demonstrated that pain stimuli increase synaptic activity in many cerebral areas involved in the sensory dimension (somatosensory and inferior parietal cortices) and the affective-attentional dimension of pain (insula, hippocampus, amygdala, cerebellum, prefrontal and cingulate cortices) [3, 8]. In the present study, conducted without noxious stimuli in hyperalgesic patients, we found significant hyperperfusion in regions known to be involved in the sensory dimension of pain. More surprisingly, in addition to the subcortical hypoperfusion previously reported [4, 5, 9], we found hypoperfusion in cortical areas assumed to be related to affective-attentional dimension of pain. Decreased synaptic activity has been observed in chronic pain conditions. Gracely et al. [2] found that, compared with a healthy group of subjects, pain stimuli in FM patients
represents the number of significant voxels in the particular cluster (P voxel level <0.001; P cluster level <0.05, corrected for multiple comparisons)

y (mm) 40 61 41 30 51 49 54 43 34 38 48 40 5 49 48 50 7 50

z (mm) 61 21 39 66 63 10 1 2 25 30 30 27 18 10 2 15 20 17

BA R BA3 L BA39 L BA40 L BA3 R BA7 R BA10 R BA10 L BA32

Localisation R PostC G L STG L IPL L PostC G R SPL R MFG R SFG L Ant Cing R Cerebellum R Cerebellum R Cerebellum L Cerebellum L Amygdala L MFG L PHG L Post Cing R PHG R Post Cing

Hypoperfusion

2,795

655

313 169 213 203 114 89

L BA10 L BA30 L BA30 R BA34 R BA29

R right, L left, PostC G post central gyrus, STG superior temporal gyrus, IPL inferior parietal lobulus, SPL superior parietal lobulus, MFG middle frontal gyrus, SFG superior frontal gyrus, Ant Cing anterior cingulum, PHG parahippocampal gyrus, Post Cing posterior cingulum

European Journal of Nuclear Medicine and Molecular Imaging Vol. 00, No. 0, Month 20xx

produced less activation in the anterior cingulate cortex and in regions involved in motor response (supplementary motor area and cerebellum). These authors suggested that FM patients develop a central pain processing mechanism of adaptation, probably partly noxious, that reduces affective appraisal and responsiveness, and may maintain the pain. In contrast, previous rest imaging studies in FM patients reported only limited subcortical abnormalities [4, 5], and preserved metabolism [9]. Many experimental differences could explain these discrepancies. First, the present work was not limited by region of interest-based analysis. In addition, previous perfusion studies have all used 99mTcHMPAO as the tracer. Uptake of 99mTc-ECD and 99mTcHMPAO in the brain follows a fixed distribution but these tracers have different pharmacokinetics and provide different image qualities. Both tracers underestimate true cerebral blood flow (CBF) in high-flow regions, but 99mTcECD uptake is known to reflect CBF more closely than 99m Tc-HMPAO and to be more sensitive for regional metabolic rate estimation, especially in the medial temporal lobe of the cortex [6]. Our population was also different from the two other studies. It comprised hyperalgesic FM patients hospitalised in a pain management department because of treatment failure. By contrast, Kwiatek et al. [5] have also studied patients with a high VAS for pain but with a larger dispersion around the mean (23 vs 4; p<0.001, with Fisher s Variance Ratio F test). It is possible that such dispersion has integrated less severely ill patients, without perfusion cortical abnormalities. SPECT findings in hyperalgesic FM patients concur with the supposed increase in nociceptive perception and decrease in affective-attentional response to pain. As current pharmacological and non-pharmacological therapies act differently on the two components of pain [10, 11], we hypothesise that individual SPECT could be a valuable and readily available tool to guide individual therapeutic strategy and perform an objective follow-up of pain processing recovery under treatment. However, although visual interpretation seems to effectively evaluate brain hypoperfusion, individual brain hyperperfusion is much more difficult to identify because of lower contrast, especially in cases of symmetrically increased uptake. Further imaging studies with 99mTc-ECD brain SPECT will be necessary in particular to investigate the feasibility and the clinical interest of voxel-by-voxel analysis of individual FM patients, in comparison with a large control group, as proposed in other diseases [12].

References
1. Wolfe F, Ross K, Anderson J, Russell IJ, Hebert L. The prevalence and characteristics of fibromyalgia in the general population. Arthritis Rheum 1995;38:1928. 2. Gracely RH, Petzke F, Wolf JM, Clauw DJ. Functional magnetic resonance imaging evidence of augmented pain processing in fibromyalgia. Arthritis Rheum 2002;46: 133343. 3. Peyron R, Garcia-Larrea L, Gregoire MC, Costes N, Convers P, Lavenne F, et al. Haemodynamic brain responses to acute pain in humans: sensory and attentional networks. Brain 1999;122:176580. 4. Mountz JM, Bradley LA, Modell JG, Alexander RW, TrianaAlexander M, Aaron LA, et al. Fibromyalgia in women. Abnormalities of regional cerebral blood flow in the thalamus and the caudate nucleus are associated with low pain threshold levels. Arthritis Rheum 1995;38:92638. 5. Kwiatek R, Barnden L, Tedman R, Jarrett R, Chew J, Rowe C, et al. Regional cerebral blood flow in fibromyalgia: singlephoton-emission computed tomography evidence of reduction in the pontine tegmentum and thalami. Arthritis Rheum 2000;43:282333. 6. Hyun Y, Lee JS, Rha JH, Lee IK, Ha CK, Lee DS. Different uptake of 99mTc-ECD and 99mTc-HMPAO in the same brains: analysis by statistical parametric mapping. Eur J Nucl Med 2001;28:1917. 7. Wolfe F, Smythe HA, Yunus MB, Bennett RM, Bombardier C, Goldenberg DL, et al. The American College of Rheumatology 1990 Criteria for the Classification of Fibromyalgia. Report of the Multicenter Criteria Committee. Arthritis Rheum 1990; 33:16072. 8. Casey KL. Forebrain mechanisms of nociception and pain: analysis through imaging. Proc Natl Acad Sci USA 1999; 96:766874. 9. Yunus MB, Young CS, Saeed SA, Mountz JM, Aldag JC. Positron emission tomography in patients with fibromyalgia syndrome and healthy controls. Arthritis Rheum 2004;51: 5138. 10. Gracely RH, Dubner R, McGrath PA. Narcotic analgesia: fentanyl reduces the intensity but not the unpleasantness of painful tooth pulp sensations. Science 1979;203:12613. 11. Donta ST, Clauw DJ, Engel CC Jr, Guarino P, Peduzzi P, Williams DA, et al. Cognitive behavioral therapy and aerobic exercise for Gulf War veterans illnesses: a randomized controlled trial. JAMA 2003;289:1396404. 12. Signorini M, Paulesu E, Friston K, Perani D, Colleluori A, Lucignani G, et al. Rapid assessment of regional cerebral metabolic abnormalities in single subjects with quantitative and nonquantitative [18F]FDG PET: a clinical validation of statistical parametric mapping. Neuroimage 1999;9:6380.

European Journal of Nuclear Medicine and Molecular Imaging Vol. 00, No. 0, Month 20xx