Original Article

pASL versus DSC perfusion MRI in lateralizing temporal lobe epilepsy

Acta Radiologica 2015, Vol. 56(4) 477–481 ! The Foundation Acta Radiologica 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0284185114531128 acr.sagepub.com

A Yusuf Oner1, Bulent Eryurt1, Murat Ucar1, Irem Capraz2, Gokhan Kurt3, Erhan Bilir2 and Turgut Tali1

Abstract Background: Accurate lateralization of the epileptogenic focus in temporal lobe epilepsy (TLE) is crucial. Pulsed arterial spin labeling (pASL) has the capability of quantifying local relative cerebral blood flow (rCBF) by measuring the inflow of electromagnetically labeled arterial blood into the target area, and can be used in the presurgical workup of refractory TLE. Purpose: To evaluate pASL in detecting mesial temporal lobe (mTL) perfusion asymmetry for the lateralization of the epileptogenic focus in patients with refractory TLE and to compare it with dynamic susceptibility contrast enhanced (DSC) magnetic resonance imaging (MRI) technique. Material and Methods: This study was approved by the local ethical committee, and written informed consent was obtained in each patient. Thirty-six patients with medically refractory TLE and 11 healthy volunteer was enrolled in this study. Following brain MRI, pASL and DSC perfusion were performed in all subjects at 3T. rCBF measurements with two different perfusion MRI technique were compared between the patient and healthy volunteers. Lateralization based on perfusion asymmetry index (AI) were also evaluated and compared with clinical lateralization. Results: rCBF ratios measured in healthy volunteers by two different perfusion technique did not show any statistically significant difference. In TLE patients rCBF ratio of the ipsilateral (affected) side was found to be significantly lower than the contralateral (unaffected) side with both technique. The AI in the patient group was 8.86  3.88 with pASL and 8.39  4.06 with DSC. Correlation coefficient between clinical laterality and perfusion AI were 0.86 for pASL and 0.83 for DSC. Conclusion: pASL can successfully detect interictal asymmetry in patients with TLE and can readily be combined with routine structural assessment for lateralization, providing an alternative to DSC perfusion.

Keywords Epilepsy, arterial spin labeling, perfusion MRI Date received: 26 March 2013; accepted: 17 March 2014

Introduction Temporal lobe epilepsy (TLE) is the most frequent type of focal refractory epilepsy (1). The seizure outcome of epilepsy surgery in refractory TLE has been shown to be superior to medical therapy, and the patients in the surgical group also showed improvement in quality of life parameters (2). Success of epilepsy surgery not only relies on a good surgical technique but also an adequate presurgical workup for the lateralization of the epileptogenic focus. This workup generally includes surface or invasive electroencephalography (EEG), magnetic resonance imaging (MRI), and positron emission tomography (PET) (3,4). With the advances in MR technology, functional MRI techniques are more widely used to complement structural findings in the lateralization of TLE. Among those functional

methods, perfusion MRI plays an important role in detecting mesial temporal hypoperfusion. Perfusion asymmetry is described to be present not only in association with morphological findings of TLE, but also in patients with normal appearing mesial temporal lobe (mTL), supporting dissociation between structure and 1 Department of Radiology, Gazi University School of Medicine, Ankara, Turkey 2 Department of Neurology, Gazi University School of Medicine, Ankara, Turkey 3 Department of Neurosurgery, Gazi University School of Medicine, Ankara, Turkey

Corresponding author: A Yusuf Oner, Department of Radiology, Gazi University School of Medicine, Ankara, Turkey. Email: [email protected]

Downloaded from acr.sagepub.com at The University of Auckland Library on June 14, 2015

478

Acta Radiologica 56(4)

function. Nowadays several MRI methods, including dynamic susceptibility contrast enhanced (DSC) and arterial spin labeling (ASL), can be used to study local cerebral perfusion changes in TLE, with the advantages of being non-invasive and delivering high spatial resolution (5–8). ASL, which is a developing MR perfusion technique, has the capability of quantifying local relative cerebral blood flow (CBF) by measuring the inflow of electromagnetically labeled arterial blood into the target area and does not require contrast medium administration (9). Earlier applications of ASL were unable to reach wide usage because of systemic errors in CBF quantification and contamination by intravascular signal (10). With the introduction of newer pulsed ASL techniques, these limitations are reduced, and this non-invasive MR perfusion method is getting increased usage in different areas of neuroradiology, including TLE imaging (11,12). However, as yet there have been few systematic studies assessing the value of pulsed ASL (pASL) technique in localizing the seizure focus in patients with TLE. The purpose of this study was to evaluate pulsed ASL in detecting mTL perfusion asymmetry for the lateralization of the epileptogenic focus in patients with refractory TLE and to compare it with DSC MRI technique.

Material and Methods Subject selection We included 11 healthy volunteer (6 men and 5 women; mean age, 34 years; range, 18–55 years) and 36 patients with intractable TLE (17 men and 19 women; mean age, 34 years; range, 20–51 years) referred to our department for imaging workup (Table 1). All patients underwent MRI, interictal EEG, and ictal scalp video EEG monitoring. Twenty of the patients did also undergo interictal PET imaging. Patients with evidence of structural lesions on MRI (except mesial temporal sclerosis and hippocampal atrophy) were excluded. None of the patients had a seizure for at least 72 h prior to perfusion imaging, confirmed by an EEG recording preceding the MRI examination. Video EEG monitoring and PET findings when present, were used for clinical lateralization of the epileptogenic foci. Informed consent following the explanation of the procedures including the possible side effects of contrast media injection was obtained from all subjects and this prospective study was approved by our local ethical committee.

MRI In all cases, conventional MRI, DSC, and pASL perfusion MRI were performed using a 3T superconducting

Table 1. Patient clinical data and pASL versus DSC perfusion AI in lateralization of epileptogenic focus. Patient

Sex

Age

Video EEG

MRI

PET

pASL AI

DSC AI

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

M M M M M M M M M M M M M M M M M F F F F F F F F F F F F F F F F F F F

21 28 40 44 29 43 46 32 55 42 20 23 24 32 44 35 46 22 41 24 26 32 32 52 36 42 26 38 24 54 36 35 38 22 28 29

R L L L L L L R L R R L L R L L R R L R L R L L L L R L L L L R L R L L

R MTS L HA N L MTS N L MTS L MTS R HA L MTS N R MTS L HA LMTS N L MTS N R MTS R HA L HA R MTS N N L MTS L MTS L MTS L HA R MTS N L MTS L HA L MTS R HA L MTS R MTS L MTS N

R L – L L – L – L – R – L R L – R R – R – R – L – – R – – – L – L – L L

8.85 8.36 8.78 8.79 9.01 4.98 8.88 9.02 9.04 8.83 7.73 8.89 8.93 12.74 9.03 8.84 8.85 8.89 6.79 8.92 8.75 8.99 8.87 8.85 8.65 8.84 10.80 9.69 8.99 8.86 8.98 8.90 8.83 8.86 8.89 8.98

8.38 8.56 8.77 8.23 8.01 4.33 8.23 8.56 8.90 9.67 7.65 8.12 8.23 12.45 7.06 7.98 8.89 8.90 6.98 8.98 8.65 8.32 8.43 8.88 8.01 8.98 8.76 8.33 8.17 8.80 8.11 7.09 9.67 8.18 8.56 8.17

AI ¼ 100 [mTL (left)  mTL (right)]/[mTL (left) þ mTL (right)], where mTL is mesial temporal lobe. DSC, dynamic susceptibility contrast-enhanced, HA, hyppocampal atrophy; MRI, lateralization at conventional MRI; MTS, mesial temporal sclerosis; N, normal; pASL, pulsed arterial spin labeling; PET, lateralization at PET; Video EEG, lateralization of the epileptogenic focus at video EEG.

magnet (Siemens Magnetom Verio, Erlangen, Germany) with a multichannel head coil in supine position. Conventional MRI included axial and sagittal T1weighted (T1W), axial T2-weighted (T2W), coronal

Downloaded from acr.sagepub.com at The University of Auckland Library on June 14, 2015

Oner et al.

479

fluid attenuated inversion recovery (FLAIR), and 3D Inversion Recovery images. T2W oblique axial images through the long axis of both hippocampi consisting of 20 slices was also obtained with 3 mm slice thickness, 0.75 mm interslice gap. Both DSC and pASL images were acquired in the same oblique axial plane with the same slice thickness and interslice gap. pASL scan consisted of a series of 101 echo planar images (EPI) in which blood was alternately tagged and untagged. pASL was performed with a pulse sequence, using a quantitative imaging of perfusion using a single subtraction technique. The parameters were as follows: slices, 20; slice thickness, 3 mm; interslice gap, 0.75 mm; repetition time (TR)/echo time (TE)/first inversion time (TI1)/ second inversion time (TI2), 2500/11/600/1500 ms; flip angle (FA), 90 ; matrix, 64  64; field of view (FOV), 192  192; and total acquisition time, 4 min 22 s. DSC MRI was performed following a bolus of Gadodiamide (OmniscanÕ , GE Healthcare, Cork, Ireland) (0.1 mmol/ kg) administered intravenously via a large bore catheter in the antecubital fossa by a mechanical injector at a rate of 5 mL/s with the following parameters: slices, 20; slice thickness, 3 mm; interslice gap, 0.75 mm; TR/TE, 1500/ 30 ms; FA, 90 ; matrix, 128  128; FOV, 230  230 mm; and total acquisition time, 1 min 38 s.

Image analysis Image datasets were transferred to a separate workstation and cerebral blood flow maps for both perfusion techniques were generated using a commercially available software package (Syngo version B17, Siemens, Erlangen, Germany). Separate regions of interest (ROIs) covering left and right mTL were manually drawn by a single investigator, blinded to patient data, on a mid-hippocampal slice of the oblique axial T2 image. mTL ROI included uncus, amygdala, hippocampus, and parahippocampus. No attempt was made to separate these structures, with attention given to avoid adjacent vascular structures. All ROIs drawn on anatomical image were then applied to both sets of CBF map, using the same software. A perfusion AI was calculated from the mean flow values in these ROI using the equation: AI ¼ 100 [mTL (left)  mTL (right)]/[mTL (left) þ mTL (right)] and corresponding mean CBF values were recorded. A second ROI was drawn at the mesencephalon for normalization purposes. Normalized mTL CBF ratios (rCBF) were generated by dividing mean CBF in each mTL ROI by the mean flow values measured at the mesencephalon. Lateralized side of hippocampus in all TLE patients were defined as ‘‘ipsilateral side’’ (n ¼ 36). The other side of hippocampus in TLE patients and both hippocampus of controls were defined as ‘‘contralateral side’’ (n ¼ 58).

Statistical analysis All descriptive statistics between control and TLE groups were tested using independent sample t-test and Mann-Whitney U test. The level for determining statistical significance was set at P < 0.05. The AI obtained from pASL perfusion MRI was statistically compared with that of DSC MRI and the laterality of temporal lobe epilepsy, as determined clinically. The correlation of AI values between pASL and DSC, together with clinical lateralization was examined by Spearman correlation coefficients.

Results Twenty-four patients had left and 12 patients had right temporal lobe epilepsy. Twenty-seven of 36 patients showed abnormal findings on conventional MR sequences under the form of signal and/or parenchymal volume abnormalities in the mTL on the side eventually chosen as the seizure side (mesial temporal sclerosis in 19 and hippocampal atrophy in 8). Figs. 1 and 2 show representative examples of pASL and DSC MRI of two different patients with left mesial temporal hypoperfusion. mTL rCBF for the volunteer group was 1.02  0.22 on the right and 1.02  0.26 on the left with pASL MRI (P > 0.05). mTL rCBF for the volunteer group was 1.02  0.12 on the right and 1.02  0.20 on the left with DSC MRI (P > 0.05). rCBF measured with two different perfusion technique in patients is summarized in Table 2. rCBF of the ipsilateral (affected) side was found to be significantly lower than the contralateral (unaffected) side both perfusion MRI technique. As patients with both right and left temporal lobe epilepsy were present in the patient sample, absolute values of asymmetry indices were used for comparison. In healthy volunteers the mean absolute asymmetry index  SD was 1.98  0.38 with pASL and 0.34  0.08 with DSC. In patients the mean absolute asymmetry index  SD was 8.86  3.88 with pASL and 8.39  4.06 with DSC (Table 1). The difference of mean absolute AI between healthy volunteers and patients was statistically significant with both techniques. Correlation coefficient between clinical laterality and perfusion AI were 0.86 for pASL and 0.83 for DSC. This reflected a good agreement between AI obtained with both perfusion techniques and clinical laterality. Twenty-three of the patients have since proceeded to an anterior temporal lobectomy and amygdalohippocampectomy. The pathologic examination demonstrated changes of mesial temporal sclerosis in 18 patients and non-specific cell loss and gliosis in the other. All patients achieved a good outcome with respect to seizures during a postoperative follow-up period in the range of 6–18 months, with 20 remaining completely seizure-free and three patients having only spontaneous seizures.

Downloaded from acr.sagepub.com at The University of Auckland Library on June 14, 2015

480

Acta Radiologica 56(4)

Fig. 1. Oblique axial T2W image (a) from a patient with left temporal lobe epilepsy shows a representative mTL ROI placement. pASL (b) and DSC (c) perfusion MR images at the same anatomic levels show hypoperfusion in the anterior left mTL (white arrow).

Fig. 2. 3D coronal inversion recovery image (a) from a patient with left temporal lobe epilepsy with no structural abnormality depicted fails to lateralize the epileptogenic focus. pASL (b) and DSC (c) perfusion MR images at the same anatomic levels show hypoperfusion in the anterior left mTL (white arrow) which is well correlated with clinical lateralization.

Table 2. mTL rCBF values with pASL and DSC techniques of the ipsilateral and contralateral sides in the TLE patient group.

rCBF (pASL) rCBF (DSC)

Contralateral side (n ¼ 58)

Ipsilateral side (n ¼ 36)

P

1.02  0.25 1.02  0.16

0.82  0.23 0.82  0.17

0.001 0.001

mTL rCBF are given as mean  SD. DSC, dynamic contrast-enhanced; mTL, mesial temporal lobe; pASL, pulsed arterial spin labeling; rCBF, normalized CBF ratio.

Discussion Measurements of cerebral blood flow and metabolism in TLE has widely been investigated (13,14). Nuclear medicine techniques including 18FDG-PET, H215O PET, and SPECT have been used for this purpose. Although interictal CBF studies using 18FDG-PET is reported to have a sensitivity of 60–90%, SPECT and H215O PET scan have not provided reliable localizing information (15–17). The main limiting factor in those

current nuclear medicine studies is their relatively poor spatial resolution. Perfusion MR techniques have important advantages over SPECT and PET, since they are non-invasive and provide higher spatial resolution. Among those, ASL is a totally non-invasive, developing perfusion MRI technique. This technique measures the inflow of magnetically labeled arterial blood into the target region (3,7). There are few reports investigating ASL perfusion MR studies in the setting of TLE. Wolf et al. (7) used a continuous ASL technique for lateralization of the epileptogenic focus in 12 TLE patients and 12 control subjects. Although he reported correct lateralization with AI in 11 patients, there was considerable overlap with the range for the control subjects. Liu et al. (18) evaluated eight TLE patients with FAIR-HASTE perfusion imaging and found good correlation with H215O PET. However, his study data were not in concordance with EEG lateralization and lacked comparison with control subjects. Recently Lim et al. (5) investigated pASL in a small study group consisting of 10 TLE patients and 11 control subjects. He reported good correlation with H215O PET in terms of localization but

Downloaded from acr.sagepub.com at The University of Auckland Library on June 14, 2015

Oner et al.

481

correctly lateralized the epileptogenic focus in only four patients. The above presented study differs from previous perfusion MR studies by the large number of TLE patients included and the evaluation of a pASL technique, which is accepted to be superior to continuous ASL in different neuroradiological settings (5), in the lateralization of the epileptogenic focus. Another differing point is the provided comparison of pASL with the DSC technique for lateralization purposes. In the presented study, rCBF values recorded in healthy subjects with both techniques were in accordance with previous reports and did not show any statistical difference (P > 0.05) or significant asymmetry (7). In TLE patients, rCBF of the ipsilateral mesial temporal lobe was found to be significantly decreased compared with the contralateral side, by both techniques. In addition, mesial temporal blood flow asymmetry was significantly more asymmetric in patients than in normal participants, resulting in a good correlation with clinical lateralization with both perfusion techniques. Although this study included a large number of patients with TLE, none of the subjects suffered from bilateral disease. One can assume that in cases of bilateral TLE, interpretation of differences in perfusion asymmetry index may be more complicated. To address this question, assessment of absolute perfusion, normalized to global perfusion in a study including bilateral TLE subjects may provide a better objective measure of abnormality. Another future prospective study may also investigate whether data provided by pASL perfusion MRI provide any additional lateralizing information to that of established functional imaging techniques such as MR spectroscopy and 18FDG-PET. Implementation of a voxel based analysis in that same prospective study can also show a larger area of perfusion alteration affecting not only the mTL, but also the adjacent structures. In conclusion, pASL perfusion MRI can be an inexpensive and completely non-invasive alternative to the DSC technique in successfully detecting interictal mTL perfusion asymmetry consistent with clinical lateralization in patients with TLE. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References 1. Wieser HG ILAE Commission on Neurosurgery of Epilepsy. ILAE Commission Report. Mesial temporal lobe epilepsy with hippocampal sclerosis. Epilepsia 2004; 45:695–714.

2. Wiebe S, Blume WT, Girvin JP, et al. Effectiveness and Efficiency of Surgery for Temporal Lobe Epilepsy Study Group. A randomized, controlled trial of surgery for temporal-lobe epilepsy. N Engl J Med 2001;345:311–318. 3. Fisher R, Stein A, Karis J. Epilepsy for the neuroradiologist. Am J Neuroradiol 1997;18:851–863. 4. Duncan J. Imaging and epilepsy. Brain 1997;120: 339–377. 5. Lim YM, Cho YW, Shamim S, et al. Usefulness of pulsed arterial spin labeling MR imaging in mesial temporal lobe epilepsy. Epilepsy Res 2008;82:183–189. 6. Alsop DC, Connelly A, Duncan JS, et al. Diffusion and perfusion MRI in epilepsy. Epilepsia 2002;43:69–77. 7. Wolf RL, Alsop DC, Levy-Reis I, et al. Detection of mesial temporal lobe hypoperfusion in patient with temporal lobe epilepsy by use of arterial spin labeled perfusion MR imaging. Am J Neuroradiol 2001;22: 1334–1341. 8. O’Brien TJ, David EP, Kilpatrick CJ, et al. Contrast enhanced perfusion and diffusion MRI accurately lateralize temporal lobe epilepsy: A pilot study. J Clin Neurosci 2007;14:841–849. 9. Barbier EL, Lamalle L, Decorps M. Methodology of brain perfusion imaging. J Magn Reson Imaging 2001; 13:496–520. 10. Buxton RB, Frank LR, Wong EC, et al. A general kinetic model for quantitative perfusion imaging with arterial spin labeling. Magn Reson Med 1998;40:383–396. 11. Detre JA, Alsop D. Perfusion magnetic resonance imaging with continuous arterial spin labeling: methods and clinical applications in the central nervous system. Eur J Radiol 1999;30:115–124. 12. Wu RH, Bruening R, Noachtar S, et al. MR measurement of regional relative cerebral blood volume in epilepsy. J Magn Reson Imaging 1999;9:435–440. 13. Engel J Jr, Henry TR, Risinger MW, et al. Presurgical evaluation for partial epilepsy: relative contributions of chronic depth-electrode recordings versus FDG-PET and scalp-sphenoidal ictal EEG. Neurology 1990;40: 1670–1677. 14. Duncan R. Epilepsy, cerebral blood flow, and cerebral metabolic rate. Cerebrovasc Brain Metab Rev 1992;4: 105–121. 15. Spencer SS. The relative contributions of MRI, SPECT and PET imaging in epilepsy. Epilepsia 1994;35(Suppl 6): S72–S89. 16. Van Paesschen W, Dupont P, Sunaert S, et al. The use of SPECT and PET in routine clinical practice in epilepsy. Curr Opin Neurol 2007;20:194–202. 17. Theodore WH, Gaillard WD, Satos, et al. Positron emission tomographic measurement of cerebral blood flow and temporal lobectomy. Ann Neurol 1983;14:429–437. 18. Liu HL, Kochunv P, Hou J, et al. Perfusion weighted imaging of interictal hypoperfusion in temporal lobe epilepsy using FAIR-HASTE: comparison with H215O PET. Magn Reson Med 2001;45:431–435.

Downloaded from acr.sagepub.com at The University of Auckland Library on June 14, 2015

pASL versus DSC perfusion MRI in lateralizing temporal lobe epilepsy.

Accurate lateralization of the epileptogenic focus in temporal lobe epilepsy (TLE) is crucial. Pulsed arterial spin labeling (pASL) has the capability...
1MB Sizes 0 Downloads 4 Views