guedj et al ejnm 2007a

Upload: haddig8

Post on 04-Jun-2018

212 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/14/2019 Guedj Et Al EJNM 2007a

    1/5

    Short communication

    99mTc-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, France2 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 99mTc-HMPAO SPECT in fibromyalgia (FM) patients havereported only limited subcortical hypoperfusion. 99mTc-ECD SPECT is known to provide better evaluation ofareas of high cerebral blood flow and regional metabolicrate. We evaluated a homogeneous group of hyperalgesic

    patients with FM using 99m

    Tc-ECD SPECT. The aim ofthis study was to investigate brain processing associatedwith spontaneous pain in FM patients.

    Methods:Eighteen hyperalgesic FM women (mean age 49years, range 2563 years; American College of Rheuma-tology criteria) and ten healthy women matched for agewere enrolled in the study. A voxel-by-voxel groupanalysis was performed using SPM2 (p

  • 8/14/2019 Guedj Et Al EJNM 2007a

    2/5

    Materials and methods

    Patients and control subjects

    Eighteen consecutive hyperalgesic FM women (49 11 years, range2563) who fulfilled the American College of Rheumatology (ACR)criteria [7] were enrolled in the study. All patients underwent ageneral medical assessment by one investigator to confirm thediagnosis. All patients failed to respond optimally to non-opioidanalgesics or weak opioids and had to be managed in a painmanagement unit. Patients with psychiatric disease were excluded.No patient had any other significant medical illness. No change wasmade in treatment during the month preceding inclusion. No patientwas receiving treatment with a strong opioid, tricyclic antidepressant,selective serotonin reuptake inhibitor, benzodiazepine or anti-convulsant agent.

    For comparison of imaging findings, a control group of tenwomen matched for age was also included (527 years, p=0.21,using the Mann-Whitney U test). All subjects provided informedconsent according to institutional guidelines.

    SPECT protocol and statistical analysis

    All brain SPECT studies were performed before any change wasmade in therapy in the pain care unit. Patients were injected with740 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-injectionperiod, was 824 (range 7590). SPECT image acquisitions wereperformed using a double-headed rotating gamma camera (ECAM,Siemens) equipped with a fan-beam collimator. Thirty-two projec-tions of 40 s were collected per head, in a 128128 format.Tomographic 3D reconstruction was performed using a filtered back-projection algorithm (Butterworth filter of order 4 with a cut-offfrequency of 0.4 cm1) and Changs attenuation correction. Each

    brain SPECT study was first visually interpreted by two nuclearmedicine physicians (E.G. and C.D.L.). A voxel-by-voxel groupstudy was then performed using SPM2 (Welcome Department ofCognitive Neurology, University College, London, UK, running onMatlab 6.0 Mathworks Inc, Sherborn, MA, USA). Images wereinitially converted from the DICOM to the Analyze format usingMRIcro (http://www.mricro.com), and transferred to SPM2. The datawere then standardised with the Montreal Neurological Institute(MNI) atlas by using a 12-parameter affine transformation, followedby non-linear transformations and a trilinear interpolation. Dimen-sions of the resulting voxels were 222 mm. Standardised data werethen smoothed by a Gaussian filter (FWHM 12 mm). FM and controlgroups were compared using the compare-populations one scan/subject routine, which carries out a fixed-effects simple t test foreach voxel. Global normalisation was performed using proportionalscaling. The SPM{T} maps were initially obtained at a heightthreshold of p

  • 8/14/2019 Guedj Et Al EJNM 2007a

    3/5

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

  • 8/14/2019 Guedj Et Al EJNM 2007a

    4/5

    10 mm radius, and a height threshold ofp

  • 8/14/2019 Guedj Et Al EJNM 2007a

    5/5

    produced less activation in the anterior cingulate cortex andin regions involved in motor response (supplementarymotor area and cerebellum). These authors suggested thatFM patients develop a central pain processing mechanismof adaptation, probably partly noxious, that reducesaffective appraisal and responsiveness, and may maintainthe pain.

    In contrast, previous rest imaging studies in FM patients

    reported only limited subcortical abnormalities [4,5], andpreserved metabolism [9]. Many experimental differencescould explain these discrepancies. First, the present workwas not limited by region of interest-based analysis. Inaddition, previous perfusion studies have all used 99mTc-HMPAO as the tracer. Uptake of 99mTc-ECD and 99mTc-HMPAO in the brain follows a fixed distribution but thesetracers have different pharmacokinetics and providedifferent image qualities. Both tracers underestimate truecerebral blood flow (CBF) in high-flow regions, but99mTc-ECD uptake is known to reflect CBF more closely than99mTc-HMPAO and to be more sensitive for regionalmetabolic rate estimation, especially in the medial temporal

    lobe of the cortex [6]. Our population was also differentfrom the two other studies. It comprised hyperalgesic FMpatients hospitalised in a pain management departmentbecause of treatment failure. By contrast, Kwiatek et al. [5]have also studied patients with a high VAS for pain butwith a larger dispersion around the mean (23 vs 4;p