Journal of Neuroradiology (2010) 37, 60—63
ORIGINAL ARTICLE
Interictal arterial spin-labeling MRI perfusion in intractable epilepsy IRM interictale de perfusion par marquage des protons dans l’épilepsie pharmacorésistante N. Pendse a, M. Wissmeyer b, S. Altrichter a, M. Vargas a, J. Delavelle a, M. Viallon c, A. Federspiel d, M. Seeck e, K. Schaller f, K.O. Lövblad a,∗ a
Department of Neuroradiology, Geneva University Hospital, Switzerland Department of Nuclear Medicine, Geneva University Hospital, Switzerland c Department of Radiology, Geneva University Hospital, Switzerland d Department of Psychiatry, University of Bern, Switzerland e Department of Neurology, Geneva University Hospital, Switzerland f Department of Neurosurgery, Geneva University Hospital, Switzerland b
Available online 11 August 2009
KEYWORDS Arterial spin-labeling; MRI; Perfusion; Epilepsy; Temporal lobe
Summary Introduction. — Magnetic resonance imaging (MRI) is required for the investigation of surgically intractable epilepsy. In addition to the standard MRI techniques, perfusion sequences can be added to improve visualization of underlying pathological changes. Arterial spin-labeling (ASL) MRI perfusion does not require contrast administration and, for this reason, may have advantages in these patients. Methods. — We report here on 16 patients with epilepsy who underwent MRI of the brain with ASL and positron emission tomography (PET). Results. — Despite a slightly reduced resolution with ASL, we found a correlation between ASL, PET and electrophysiological data, with hypoperfusion on ASL that corresponded with hypoperfusion on interictal PET. Conclusion. — Given the correlation between ASL and PET and electrophysiology, perfusion with ASL could become part of the standard work-up in patients with epilepsy. © 2009 Elsevier Masson SAS. All rights reserved.
Introduction ∗ Corresponding author. Service neuro-diagnostique et neurointerventionnel, DISIM, hôpitaux universitaires de Genève, 24, rue Micheli-du-Crest, 1211 Genève 4, Switzerland. E-mail address:
[email protected] (K.O. Lövblad).
Imaging plays a major role in the management of patients with epilepsy [1,2]. Seizures are a frequent symptom in such patients and, while pharmaceutical therapy remains the first line of approach, patients can develop pharmacologically
0150-9861/$ – see front matter © 2009 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.neurad.2009.05.006
Interictal arterial spin-labeling MRI perfusion in intractable epilepsy
61
refractive seizures that may necessitate surgery. Thus, it is important to further improve the diagnostic armamentarium that can be used in these patients. Perfusion techniques are usually employed as an adjunct to anatomical sequences in Magnetic resonance imaging (MRI). As arterial spin-labeling (ASL) requires no contrast use, it is of considerable interest [3—6]. We report here on patients who underwent MRI with additional ASL.
Patients and methods In the present study, which was approved by our local ethics committee (study number: 08-097 R [NAC 08-031R]), we examined 16 patients (10 males, six females; ages 7—55 years) with intractable epilepsy who were referred to our epilepsy presurgical work-up unit for evaluation of intractable seizures. In this unit, patients are seen by a multidisciplinary team comprising neurologists, neurosurgeons, neuroradiologists and pediatricians specializing in the workup of patients with epilepsy. In addition to a full clinical examination, our 16 patients underwent non-invasive electroencephalography (EEG). On the basis of their clinical and electrophysiological findings, nine patients had temporal lobe epilepsy, two had post-traumatic scar lesions, three had focal seizures and two had dysplastic lesions with secondary epilepsy. MRI was performed using a 3.0-T Magnetom Trio (Siemens; Erlangen, Germany) (Figs. 2 and 4). ASL was performed with a pulsed sequence, using a QUIPSII perfusion mode and the following parameters: 16 slices; voxel size: 3.4 × 3.4 × 6 mm; TA = 5:55 min; lambda = 0.9 mL/g; alpha = 95%; and TE/TR/TI1/TI2/T1 (blood 3T) = 15/5000/ 700/1800/1496,19 ms. Relative cerebral blood flow (relCBF) maps for ASL were calculated online by the MRI scanner, and offline for contrast-enhanced perfusion-weighted imaging (cePWI) using Syngo Perfusion (MR) software. An epilepsy protocol was used, comprising the following sequences: axial T2-weighted images (TE: 101 ms, TR 4000 ms, 26 slices, 4-mm thickness, 372 × 510 matrix); sagittal T1-weighted multiplanar reconstructed (MPR) (TE: 2.32 ms, TR: 1900 ms, 512 × 512 matrix) and sagittal threedimensional fluid-attenuated inversion recovery (3D FLAIR) (TE: 420 ms, TR: 6000 ms, 256 × 256 matrix, 162 1-mm thick contiguous images). Susceptibility-weighted imaging (SWI) was performed, using 3D acquisition with an in-plane resolution of 1 × 1 × 1 mm, as well as diffusion-weighted imaging (DWI) with 30-direction scanning. Interictal [18 F]-2-fluoro-deoxy-d-glucose positron emission tomography with computed tomography (18 F-FDG PET/CT) was used for comparison with ASL as it is known to have better spatial resolution and greater sensitivity for correct localization of epileptogenic regions compared with interictal nuclear-medicine perfusion imaging techniques such as 99m-technetium ethylcysteinate dimer singlephoton emission computed tomography (99m Tc-ECD SPECT). Thirty minutes after intravenous injection of a mean dose of 169 ± 27 MBq of 18 F-FDG, integrated PET/CT was acquired, using a Siemens Biograph 16, after CT without contrast enhancement (2-mm slices) for anatomical co-registration, attenuation and scatter correction. As well as visual
Figure 1 From a 53-year-old man with left temporal lobe epilepsy (TLE): A. Coronal FLAIR shows left hippocampal atrophy and hyperintensity; B. ASL map shows left hippocampal hypoperfusion (arrow); C. Axial PET shows left hippocampal hypoperfusion; D. Coronal PET shows left hippocampal hypoperfusion.
evaluation, the FDG PET scans were compared with a normal series from 12 healthy subjects to identify clusters of voxels showing statistically significant (P < 0.05) hypometabolism.
Results The ASL perfusion maps showed multiple areas of hypoperfusion that corresponded well with the hypometabolic areas seen on the PET images. All cases with temporal lobe epilepsy (TLE) showed hypoperfusion in the affected temporal lobe on both the interictal PET and ASL maps (Figs. 1—3). In those cases where a cortical lesion was visible on imaging and corresponded with an alteration seen on PET, the same result was also seen on ASL perfusion (Fig. 4). In general, although the ASL images showed slightly poorer spatial resolution, they provided an equal amount of information.
Discussion The perfusion maps we obtained with ASL corresponded well with the PET perfusion and EEG results in patients with intractable epilepsy. These patients were also candidates for eventual surgery; this is not surprising because, while PET with FDG represents metabolic activity, ASL perfusion provides maps of cerebral blood flow, another parameter of perfusion. Whenever patients with intractable epilepsy — seizures that cannot be controlled even with complex pharmacological combination therapies — are investigated, it is necessary to consider surgery as a possible option. In such patients, the precise localization of seizures is carried out by both electrophysiology as well as neuroimaging studies. This has conferred on MRI an important role in
62
Figure 2 From a 49-year-old man with left TLE: A. Coronal T2-weighted image shows left hippocampal atrophy; B. Axial FLAIR shows left hippocampal hyperintensity; C. Axial ASL shows slight hippocampal hypoperfusion; D. Axial PET shows distinct hippocampal hypoperfusion on the left.
the anatomical evaluation of these lesions. While traditional MRI techniques can provide anatomical information, nuclear-medicine techniques are relied upon to visualize brain perfusion both ictally and interictally. This means that new MRI techniques, such as diffusion and perfusion, are now becoming part of the standard care of such patients. In the present study, we found a correspondence between interictal and ictal findings: the interictal lesions showing hypoperfusion on PET also showed hypoperfusion on ASL perfusion. Indeed, ictal ASL findings were found to correlate with ictal PET in a study of a patient with hemimegalen-
Figure 3 From a 40-year-old man with left TLE: A. Coronal FLAIR shows left hippocampal hypersignaling and atrophy; B. ASL map shows hypoperfusion in the left hippocampus (arrow); C. Gadolinium (Gd) perfusion map shows the same hypoperfusion on cerebral blood flow maps; D. Gd perfusion map shows a decrease in mean transit time (MTT) in the left temporal lobe.
N. Pendse et al.
Figure 4 From an 11-year-old boy with frontal dysplasia: A. Axial FLAIR shows hyperintensity and alterations of the frontal white and gray matter; B. Coronal T2-weighted image shows frontal dysplasia on the right; C. ASL map shows hypoperfusion in the right frontal cortex (arrow); D. PET shows hypoperfusion in the right frontal cortex.
cephaly and epilepsy [7]. In addition, we also show that ASL can be used for the evaluation of interictal lesions of various types — not just in TLE, but in developmental disorders as well [8]. For these reasons, imaging plays a crucial role in any investigations [8], and adding ictal and interictal perfusion information has been shown to be essential in the evaluation of epilepsy, where nuclear-medicine methods — especially the SISCOM (subtraction ictal SPECT co-registered to MRI) technique — are the standard imaging procedures. As for interictal perfusion imaging, its findings have already been well established. Given that ASL offers visualization of flow values and not metabolism, it may represent a better means of visualizing interictal hypoperfusion in patients with epilepsy. In addition, ASL does not require the use of intravenous contrast agents, which is an advantage in the case of young patients and/or in those who may have renal impairment. It is also advantageous for patients in whom standard ictal perfusion imaging cannot be performed, and in those cases where interictal examinations may have to be repeated often. The latter is of particular interest, given the current concerns over the potential for the development of systemic nephrogenic fibrosis with exposures to some contrast media. Thus, we believe that adding brain perfusion investigations to the MRI protocol in patients with intractable epilepsy may have a number of advantages.
References [1] Urbach H. Imaging of the epilepsies. Eur Radiol 2005;15(3): 494—500. [2] Heiniger P, el-Koussy M, Schindler K, Lövblad KO, Kiefer C, Oswald H, et al. Diffusion and perfusion MRI for the localisation
Interictal arterial spin-labeling MRI perfusion in intractable epilepsy of epileptogenic foci in drug-resistant epilepsy. Neuroradiology 2002;44(6):475—80. [3] Wong EC, Buxton RB, Frank LR. Quantitative perfusion imaging using arterial spin labeling. Neuroimaging Clin N Am 1999;9(2):333—42. [4] Wong EC. Quantifying CBF with pulsed ASL: technical and pulse sequence factors. J Magn Reson Imaging Dec 2005;22(6):727—31. [5] Detre JA, Alsop DC. Perfusion magnetic resonance imaging with continuous arterial spin labeling: methods and clinical applications in the central nervous system. Eur J Radiol 1999;30(2):115—24.
63
[6] Deibler AR, Pollock JM, Kraft RA, Tan H, Burdette JH, Maldjian A. Arterial spin-labeling in routine clinical practice. Part 1: technique and artifacts. AJNR Am J Neuroradiol 2008;29: 1228—34. [7] Altrichter S, Pendse N, Wissmeyer M, Jägersberg M, Federspiel A, Viallon M, et al. Arterial spin-labeling demonstrates ictal cortical hyperperfusion in epilepsy consecutive to hemimegalencephaly. J Neuroradiol 2009 [in press]. [8] Widjaja E, Wilkinson ID, Griffiths PD. Magnetic resonance perfusion imaging in malformations of cortical development. Acta Radiol Oct 2007;48(8):907—17.
