- Email: [email protected]
Contribution of the apparent diffusion coefficient in perilesional edema for the assessment of brain tumors
Journal of Neuroradiology (2008) 35, 224—229 Disponible en ligne sur www.sciencedirect.com
ORIGINAL ARTICLE
Contribution of the apparent diffusion coefficient in perilesional edema for the assessment of brain tumors Apport du coefficient de diffusion apparent de l’œdème périlésionnel pour l’étude des tumeurs cérébrales R. Guzman a,b, S. Altrichter c, M. El-Koussy d, J. Gralla d, J. Weis e,f, A. Barth a, R.W. Seiler a, G. Schroth c, K.O. Lövblad c,∗ a
Department of Neurosurgery, Inselspital, University of Berne, Berne, Switzerland Department of Neurosurgery, Stanford University, Palo Alto, USA c Department of radiology, Neuroradiology Section, Geneva University Hospital, Geneva, 24, rue Micheli-du-Crest, 1211 Geneva 14, Switzerland d Department of Neuroradiology, Inselspital, University of Berne, Bern, Switzerland e Department of Neuropathology, University of Berne, Berne, Switzerland f RWTH Aachen University Hospital, Aachen, Germany b
Available online 16 April 2008
KEYWORDS Diffusion; Brain tumors; Gliomas; MRI-ADC
∗
Summary Objectives. — Diffusion-weighted MRI is sensitive to molecular motion and has been applied to the diagnosis of stroke. Our intention was to investigate its usefulness in patients with brain tumor and, in particular, in the perilesional edema. Methods. — We performed MRI of the brain, including diffusion-weighted imaging and mapping of the apparent diffusion coefficient (ADC), in 16 patients with brain tumors (glioblastomas, lowgrade gliomas and metastases). ADC values were determined by the use of regions of interest positioned in areas of high signal intensities as seen on T2-weighted images and ADC maps. Measurements were taken in the tumor itself, in the area of perilesional edema and in the healthy contralateral brain. Results. — ADC mapping showed higher values of peritumoral edema in patients with glioblastoma (1.75 × 10−3 mm2 /s) and metastatic lesions (1.61 × 10−3 mm2 /s) compared with those who had low-grade glioma (1.40 × 10−3 mm2 /s). The higher ADC values in the peritumoral zone were associated with lower ADC values in the tumor itself.
Corresponding author. E-mail address: [email protected] (K.O. Lövblad).
0150-9861/$ – see front matter © 2008 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.neurad.2008.02.003
Contribution of the apparent diffusion coefficient in perilesional edema
225
Conclusions. — The higher ADC values in the more malignant tumors probably reflect vasogenic edema, thereby allowing their differentiation from other lesions. © 2008 Elsevier Masson SAS. All rights reserved.
MOTS CLÉS Diffusion ; Tumeurs cérébrales ; Gliomes ; IRM-ADC
Résumé Objectif. — L’IRM de diffusion est sensible aux mouvements moléculaires et a été appliquée avec succès pour le diagnostic des accidents vasculaires cérébraux. Notre objectif était d’étudier son intérêt chez les patients présentant des tumeurs cérébrales, avec une attention particulière au niveau de l’œdème périlésionnel. Méthodes. — Une IRM cérébrale comportant des séquences de diffusion et des cartographies du coefficient de diffusion apparent (ADC) était réalisée chez 16 patients présentant des tumeurs cérébrales (glioblastomes, gliomes de bas grade et métastases). Les valeurs d’ADC étaient déterminées à l’aide de régions d’intérêt définies au niveau des zones de signal élevé visibles sur les images T2 et les cartographies ADC. Les mesures étaient réalisées au niveau de la tumeur elle-même, dans l’œdème périlésionnel et dans le cerveau sain controlatéral. Résultats. — Les cartographies ADC ont montré des valeurs plus élevées au niveau de l’œdème péritumoral chez les patients présentant un glioblastome (1,75 × 10−3 mm2 /s) ou une lésion métastatique (1,61 × 10−3 mm2 /s) comparées à ceux présentant un gliome de bas grade (1,40 × 10−3 mm2 /s). Ces valeurs plus élevées en zone péritumorale étaient associées à un ADC bas dans la tumeur elle-même. Conclusions. — L’augmentation des valeurs d’ADC au niveau de l’œdème périlésionnel des tumeurs plus malignes reflète probablement un œdème vasogénique plus important permettant de les différencier des autres lésions. © 2008 Elsevier Masson SAS. All rights reserved.
Introduction Diffusion-weighted imaging (DWI) has become a standard routine sequence in most clinical investigations [1,2]. It involves a simple modification of a spin-echo sequence, where two gradient sensitizing pulses are applied before and after the 180 pulse. Diffusion imaging is sensitive to the molecular motion of tissues, allowing it to be used with great success in the setting of stroke [3,4], where it is thought to be able to differentiate between cytotoxic and vasogenic edema. The occurrence of acute ischemia is accompanied by a decrease in the so-called apparent diffusion coefficient (ADC), which allows quantification of the diffusion characteristics of a given tissue. This has been applied to other pathologies such as brain abscess [5] and some tumors, and brain development [6] and aging [7], as well as for monitoring interventions [8]. The possibility of mapping the ADC allows the quantification of water movement. This has already been applied to tumors of the brain, head and neck [9,10]. Previous studies have shown that ADC values reflect both cellularity and edema [11]. The present study aimed to address its capacity to determine the type of intracranial tumor based on the ADC values from the perilesional edema. To do this, we investigated a series of patients who had different types of intracranial tumors.
Materials and methods Patients Sixteen consecutive patients (nine men, seven women; ages 35 to 82, mean age 58 years) were prospectively included, having been recruited from the neurosurgical clinic of our hospital over a period of six months (January to July 2001).
All patients had undergone brain MRI because of an intracranial tumor. They were informed of the examination they were about to undergo and had given their consent. The patients all subsequently underwent craniotomy and biopsy/surgery to obtain samples of tumor for histopathological examination.
Imaging studies Imaging was performed using a commercial 1.5-T clinical system equipped with a head coil (Siemens Vision, Erlangen, Germany) capable of echo-planar imaging, and with a gradient booster and gradient amplitude of 35 mT/m, a rise time of 150 ms and a slew rate of 170 T/m. First, routine axial T1- (TE: 12 ms; TR: 528 ms; matrix 96 × 200; NEX: 2; 5-mm-thick slices) and T2- (TE: 98/16 ms; TR: 3.176 ms; matrix: 180 × 256; 5-mm-thick slices) weighted imaging was done with postcontrast T1-weighted imaging in three planes after administration of gadolinium (Gd)-DTPA (Magnevist, Schering, Germany) via a cubital-vein cannula, using a power injector (Medrad). Axial isotropic single-shot whole-brain multislice DWI (TR: 4000 ms; TE 36 ms; 18 5-mm-thick slices) was performed, and ADC maps derived from the b0 and bmax images (b values of 0 and 1000 s/mm2 ) by pixel-by-pixel analysis. The images were then transferred to a dedicated workstation. Signal anomalies on DWI and ADC images were determined by two experienced neuroradiologists (KOL, MEK). Regions of interest (ROI) on the ADC maps were also selected in: (i) the areas seen with high signal intensities on T2-weighted images and ADC maps that corresponded to perilesional edema (Figs. 1 and 2); (ii) the tumor itself; and (iii) the contralateral healthy brain. Thus, there were three measurements per patient. Mean ROI size was 1.6 mm2 .
226
R. Guzman et al.
Figure 1 In a patient with a high-grade glioma of the left hemisphere, the peripheral edema can be seen on T2 images and ADC maps. Figure 1 Patient présentant un gliome de haut grade. L’œdème est visualisé sur les séquences T2 et les cartographies ADC.
Statistical analysis Student’s t test was used to compare the two groups of patients with those who had low-grade glioma.
Results On histology, eight patients showed high-grade gliomas, four had cerebral metastases and four had low-grade gliomas (Fig. 3).
ADC values in peritumoral edema The absolute values in the peritumoral edema of the glioblastomas were: median: 1.75 × 10−3 mm2 /s, SD: 13.76;
range: 1.55—1.95 × 10−3 mm2 /s; 25th percentile: 1.65, 75th percentile: 1.88; lower 95% CI of mean: 1.674, upper 95% CI of mean: 1.826 (Fig. 3B). The absolute values in the metastases were: median: 1.61 × 10−3 mm2 /s, SD: 19.13; range: 1.21—1.93 × 10−3 mm2 /s; 25th percentile: 1.525, 75th percentile: 1.89; lower 95% CI of mean: 1.454, upper 95% CI of mean: 1.748. The absolute values in the low-grade gliomas were: median: 1.40 × 10−3 mm2 /s, SD: 20.42; range: 1.08—1.60 × 10−3 mm2 /s; 25th percentile: 1.10, 75th percentile: 1.55; lower 95% CI of mean: 1.17, upper 95% CI of mean: 1.547. The difference was significant for the perilesional edema of glioblastomas compared with low-grade gliomas (P = 0.0071), but not between metastases and low-grade gliomas (P = 0.0811).
Figure 2 In this patient with a metastasis in the left frontal region, there is a central contrast-enhancing nodule on T1 images. The ADC map shows the ROIs in the lesion, in the edema and in the contralateral brain. Figure 2 Patient présentant une métastase de la région frontale gauche. Un nodule prenant le contraste en T1 est entouré d’un œdème visible sur les images ADC sur lesquelles sont placées les régions d’intérêt.
Contribution of the apparent diffusion coefficient in perilesional edema
227
Figure 3 ADC values in three types of brain tumor: (A): relationships between the pathological compared with normal tissues; (B): in the areas of perilesional edema; and (C): in the lesions themselves. Met: metastases; LGG: low-grade gliomas; GBM: glioblastomas. Figure 3 Valeurs ADC dans les différents types de tumeurs cérébrales : (A) : rapport de l’ADC entre le côté pathologique et le côté normal ; (B) : valeurs d’ADC dans l’œdème périlésionnel ; (C) : valeurs d’ADC dans les lésions elles-mêmes. Mets : métastases ; LGG : gliomes de bas grade ; GBM : glioblastomes.
Ratios of edema compared with contralateral healthy brain The ratios for the glioblastomas were: median: 2.38, SD: 0.41; range: 1.7—2.9; 25th percentile: 1.8, 75th percentile: 2.6; lower 95% CI of mean: 2.013, upper 95% CI of mean: 2.476 (Fig. 3A). The ratios for the metastases were: median: 2.143, SD: 0.27; range: 1.66—2.57; 25th percentile: 1.86, 75th percentile: 2.25; lower 95% CI of mean: 1.9, upper 95% CI of mean: 2.3. The ratios for the low-grade gliomas were: median: 1.57, SD: 0.27; range: 1.24—2.05; 25th percentile: 1.3, 75th percentile: 1.74; lower 95% CI of mean: 1.32, upper 95% CI of mean: 1.83.
ADC values in the tumors themselves The intratumoral values for the glioblastomas were: median: 1.30 × 10−3 mm2 /s, SD: 16.75; range: 1.02—1.53 × 10−3 mm2 /s; 25th percentile: 1.16, 75th percentile: 1.39; lower 95% CI of mean: 1.124, upper 95% CI of mean: 1.433 (Fig. 3C). The values for the metastases were: median: 1.44 × 10−3 mm2 /s, SD: 21.73; range: 110—165 × 10−3 mm2 /s; 25th percentile: 119, 75th percentile: 163; lower 95% CI of mean: 119.2, upper 95% CI of mean: 164.8. Values for the low-grade gliomas were: median: 1.95 × 10−3 mm2 /s, SD: 10.78; range: 1.82—2.08 × 10−3 mm2 /s; 25th percentile: 1.83, 75th percentile: 2.035; lower 95% CI of mean: 1.802, upper 95% CI of mean: 2.07. The differences were significant for glioblastomas compared with low-grade gliomas (P = 0.0001) as well as for metastases compared with low-grade gliomas (P = 0.0011). The ratios for brain tumors compared with the contralateral healthy brain were: 0.95 for glioblastomas; 1.06 for metastases; and 1.43 for low-grade gliomas.
Discussion This study showed that both high-grade gliomas and metastatic brain tumors have higher ADC values in the per-
ilesional edema than do low-grade gliomas, indicating a higher water content and greater tissue displacement due to vasogenic edema, and probably secondary to a more aggressive histological reaction. Also, there is a tendency for the more malignant lesions to have lower ADC values within the tumor itself, probably due to necrosis. Therefore, DWIderived ADC maps can be helpful in differentiating tumor malignancy, and provide information not just concerning the tumor, but also about tumor edema extent. Regarding highgrade brain tumors, the study by Castillo et al. [12] found that such tumors have specific values that, however, could not be differentiated from those in the peripheral edema. Other, earlier studies have postulated that diffusion imaging could differentiate the tumor itself from edema [13]. As regards perilesional edema, similar results have been reported in the literature: Lu et al. [14] reported that ADC measurements can differentiate cytotoxic from vasogenic edema, and also between metastases and gliomas: they found increased diffusibility within the vasogenic edema surrounding both high-grade gliomas and metastatic tumors, reflecting increased extracellular water. However, the authors also found that peritumoral fractional anisotropy (FA) demonstrated no statistically significant differences. The FA changes observed around gliomas, therefore, can be attributed not only to increased water content, but also to tumor infiltration [15]. There have also been studies regarding tumor cellularity and diffusibility: increased cellularity is associated with a lower ADC, and lower ADC values appear to be associated with higher-grade tumors [16]. In a study of ADC values and T2 values, Oh et al. [17] found that ADCs for gliomas were similar to those for meningiomas and metastases in all regions. However, tumor T2 values for gliomas were significantly higher than those for meningiomas and metastases. Localized-edema T2 values for meningiomas and metastases were significantly higher than those for gliomas. Peripheraledema T2 values for gliomas were similar to those for meningiomas and metastases. In addition, they found that using a combination of ADC and T2 values resulted in improved tissue characterization of these lesions. They also found significant correlations between ADC and T2 values. As with this study and that of Baehring et al. [18], we found a decrease in the central ADC of glioblastomas.
228 Diffusion imaging has been used in the appraisal of tumors and may even have the potential to differentiate between tumor subtypes. However, on reviewing the literature, it is clear that there is overlap between values [19]. Thus, a certain degree of care should be taken before relying on this parameter too strictly [20]. Nevertheless, its findings can certainly be used when carrying out multimodality imaging [21]. Toh et al. [22] did a study of the ADC in the peritumoral edema of metastases and meningiomas, and found that all values were significantly different between the two types of tumors. Provenzale et al. [23] examined the FA and mean ADC values in peritumoral hyperintense regions, and found that mean FA values were significantly higher for meningiomas than for gliomas, while mean ADCs in peritumoral normal-looking white matter were not significantly different. Overall, this means that ADC mapping is a further means for obtaining improved tissue characterization. This, in addition to the information provided by other techniques such as spectroscopy [24], perfusion [25,26] and diffusion tensor imaging [23], it should improve the initial diagnosis and work-up. Indeed, because of the often nonhomogeneous nature of tumors such as glioblastomas, a single measurement in a single area will not necessarily be helpful; it is the use of many parameters that will aid in improving our understanding of the nature of any lesion.
Conclusion ADC mapping of brain tumors and assessment of the ADC in the peritumoral lesion appears to be promising in the differentiation of tumors and in the assessment of local invasiveness. Glioblastomas, while having a low central ADC, have a high ADC in the periphery, reflecting a central high cellularity and peripheral vasogenic edema.
References [1] Le Bihan D, Breton E, Lallemand D, Grenier P, Cabanis E, Laval-Jeantet M. MR imaging of intravoxel incoherent motions: application to diffusion and perfusion in neurologic disorders. Radiology 1986;161(2):401—7. [2] Le Bihan D, Breton E, Lallemand D, Desbleds MT, Aubin ML, Vignaud J, et al. Contribution of intravoxel incoherent motion (IVIM) imaging to neuroradiology. J Neuroradiol 1987;14(4):295—312. [3] Lovblad KO, Laubach HJ, Baird AE, Curtin F, Schlaug G, Edelman RR, et al. Clinical experience with diffusion-weighted MR in patients with acute stroke. AJNR Am J Neuroradiol 1998;19(6):1061—6. [4] Moseley ME, Cohen Y, Mintorovitch J, Chileuitt L, Shimizu H, Kucharczyk J, et al. Weinstein PR Early detection of regional cerebral ischemia in cats: comparison of diffusionand T2-weighted MRI and spectroscopy. Magn Reson Med 1990;14:330—46. [5] Guzman R, Barth A, Lövblad KO, El-Koussy M, Weis J, Schroth G, et al. Use of diffusion-weighted magnetic resonance imaging in differentiating purulent brain processes from cystic brain tumors. J Neurosurg 2002;97(5):1101—7. [6] Lövblad KO, Schneider J, Ruoss K, Steinlin M, Fusch C, Schroth G. Isotropic apparent diffusion coefficient mapping of postnatal cerebral development. Neuroradiology 2003;45(6):400—3.
R. Guzman et al. [7] Lövblad KO, Delavelle J, Wetzel S, Kelekis AD, Assal F, Palmesino M, et al. ADC mapping of the aging frontal lobes in mild cognitive impairment. Neuroradiology 2004;46(4): 282—6. [8] Lövblad KO, Plüschke W, Remonda L, Gruber-Wiest D, Do DD, Barth A, et al. Diffusion-weighted MRI for monitoring neurovascular interventions. Neuroradiology 2000;42(2):134—8. [9] Matsushima N, Maeda M, Takamura M, Takeda K. Apparent diffusion coefficients of benign and malignant salivary gland tumors. Comparison to histopathological findings. J Neuroradiol 2007;34(3):183—9. [10] Suzuki C, Maeda M, Hori K, Kozuka Y, Sakuma H, Taki W, et al. Apparent diffusion coefficient of pituitary macroadenoma evaluated with line-scan diffusion-weighted imaging. J Neuroradiol 2007;34(4):228—35. [11] Rodallec M, Colombat M, Krainik A, Kalamaridès M, Redondo A, Feydy A. Diffusion-weighted MR imaging and pathologic findings in adult cerebellar medulloblastoma. J Neuroradiol 2004;31(3):234—7. [12] Castillo M, Smith JK, Kwock L, Wilber K. Apparent diffusion coefficients in the evaluation of high-grade cerebral gliomas. AJNR Am J Neuroradiol 2001;22(1):60—4. [13] Eis M, Els T, Hoehn-Berlage M, Hossmann KA. Quantitative diffusion MR imaging of cerebral tumor and edema. Acta Neurochir Suppl (Wien) 1994;60:344—6. [14] Lu S, Ahn D, Johnson G, Cha S. Peritumoral diffusion tensor imaging of high-grade gliomas and metastatic brain tumors. AJNR Am J Neuroradiol 2003;24(5):937—41. [15] Lu S, Ahn D, Johnson G, Law M, Zagzag D, Grossman RI. Diffusion-tensor MR imaging of intracranial neoplasia and associated peritumoral edema: introduction of the tumor infiltration index. Radiology 2004;232(1):221—8. [16] Sugahara T, Korogi Y, Kochi M, Ikushima I, Shigematu Y, Hirai T, et al. Usefulness of diffusion-weighted MRI with echo-planar technique in the evaluation of cellularity in gliomas. J Magn Reson Imaging 1999;9(1):53—60. [17] Oh J, Cha S, Aiken AH, Han ET, Crane JC, Stainsby JA, et al. Quantitative apparent diffusion coefficients and T2 relaxation times in characterizing contrast enhancing brain tumors and regions of peritumoral edema. J Magn Reson Imaging 2005;21(6):701—8. [18] Baehring JM, Bi WL, Bannykh S, Piepmeier JM, Fulbright RK. Diffusion MRI in the early diagnosis of malignant glioma. J Neurooncol 2007;82(2):221—5. [19] Provenzale JM, Mukundan S, Barboriak DP. Diffusion-weighted and perfusion MR imaging for brain tumor characterization and assessment of treatment response. Radiology 2006;239(3):632—49. [20] Tozer DJ, Jäger HR, Danchaivijitr N, Benton CE, Tofts PS, Rees JH, et al. Apparent diffusion coefficient histograms may predict low-grade glioma subtype. NMR Biomed 2007;20(1): 49—57. [21] Schneider JF, Viola A, Confort-Gouny S, Ayunts K, Le Fur Y, Viout P, et al. Infratentorial pediatric brain tumors: the value of new imaging modalities. J Neuroradiol 2007;34(1): 49—58. [22] Toh CH, Wong AM, Wei KC, Ng SH, Wong HF, Wan YL. Peritumoral edema of meningiomas and metastatic brain tumors: differences in diffusion characteristics evaluated with diffusiontensor MR imaging. Neuroradiology 2007;49(6):489—94. [23] Provenzale JM, McGraw P, Mhatre P, Guo AC, Delong D. Peritumoral brain regions in gliomas and meningiomas: investigation with isotropic diffusion-weighted MR imaging and diffusiontensor MR imaging. Radiology 2004;232(2):451—60. [24] Alimenti A, Delavelle J, Lazeyras F, Yilmaz H, Dietrich PY, de Tribolet N, et al. Monovoxel 1H magnetic resonance spectroscopy in the progression of gliomas. Eur Neurol 2007;58(4): 198—209.
Contribution of the apparent diffusion coefficient in perilesional edema [25] Cotton F, Ongolo-Zogo P, Luois-Tisserand G, Streichenberger N, Hermier M, Jouvet A, et al. IRM de diffusion—perfusion dans l’évaluation des lymphomes cérébraux. J Neuroradiol 2006;33:220—8.
229
[26] Guyer S, Kosel M, Altrichter S, El-Koussy M, Haemmig R, Fisch HU, et al. Pattern of regional cerebral blood-flow changes induced by acute heroin administration — a perfusion MRI study. J Neuroradiol 2007;34(5):322—9.
ORIGINAL ARTICLE
Contribution of the apparent diffusion coefficient in perilesional edema for the assessment of brain tumors Apport du coefficient de diffusion apparent de l’œdème périlésionnel pour l’étude des tumeurs cérébrales R. Guzman a,b, S. Altrichter c, M. El-Koussy d, J. Gralla d, J. Weis e,f, A. Barth a, R.W. Seiler a, G. Schroth c, K.O. Lövblad c,∗ a
Department of Neurosurgery, Inselspital, University of Berne, Berne, Switzerland Department of Neurosurgery, Stanford University, Palo Alto, USA c Department of radiology, Neuroradiology Section, Geneva University Hospital, Geneva, 24, rue Micheli-du-Crest, 1211 Geneva 14, Switzerland d Department of Neuroradiology, Inselspital, University of Berne, Bern, Switzerland e Department of Neuropathology, University of Berne, Berne, Switzerland f RWTH Aachen University Hospital, Aachen, Germany b
Available online 16 April 2008
KEYWORDS Diffusion; Brain tumors; Gliomas; MRI-ADC
∗
Summary Objectives. — Diffusion-weighted MRI is sensitive to molecular motion and has been applied to the diagnosis of stroke. Our intention was to investigate its usefulness in patients with brain tumor and, in particular, in the perilesional edema. Methods. — We performed MRI of the brain, including diffusion-weighted imaging and mapping of the apparent diffusion coefficient (ADC), in 16 patients with brain tumors (glioblastomas, lowgrade gliomas and metastases). ADC values were determined by the use of regions of interest positioned in areas of high signal intensities as seen on T2-weighted images and ADC maps. Measurements were taken in the tumor itself, in the area of perilesional edema and in the healthy contralateral brain. Results. — ADC mapping showed higher values of peritumoral edema in patients with glioblastoma (1.75 × 10−3 mm2 /s) and metastatic lesions (1.61 × 10−3 mm2 /s) compared with those who had low-grade glioma (1.40 × 10−3 mm2 /s). The higher ADC values in the peritumoral zone were associated with lower ADC values in the tumor itself.
Corresponding author. E-mail address: [email protected] (K.O. Lövblad).
0150-9861/$ – see front matter © 2008 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.neurad.2008.02.003
Contribution of the apparent diffusion coefficient in perilesional edema
225
Conclusions. — The higher ADC values in the more malignant tumors probably reflect vasogenic edema, thereby allowing their differentiation from other lesions. © 2008 Elsevier Masson SAS. All rights reserved.
MOTS CLÉS Diffusion ; Tumeurs cérébrales ; Gliomes ; IRM-ADC
Résumé Objectif. — L’IRM de diffusion est sensible aux mouvements moléculaires et a été appliquée avec succès pour le diagnostic des accidents vasculaires cérébraux. Notre objectif était d’étudier son intérêt chez les patients présentant des tumeurs cérébrales, avec une attention particulière au niveau de l’œdème périlésionnel. Méthodes. — Une IRM cérébrale comportant des séquences de diffusion et des cartographies du coefficient de diffusion apparent (ADC) était réalisée chez 16 patients présentant des tumeurs cérébrales (glioblastomes, gliomes de bas grade et métastases). Les valeurs d’ADC étaient déterminées à l’aide de régions d’intérêt définies au niveau des zones de signal élevé visibles sur les images T2 et les cartographies ADC. Les mesures étaient réalisées au niveau de la tumeur elle-même, dans l’œdème périlésionnel et dans le cerveau sain controlatéral. Résultats. — Les cartographies ADC ont montré des valeurs plus élevées au niveau de l’œdème péritumoral chez les patients présentant un glioblastome (1,75 × 10−3 mm2 /s) ou une lésion métastatique (1,61 × 10−3 mm2 /s) comparées à ceux présentant un gliome de bas grade (1,40 × 10−3 mm2 /s). Ces valeurs plus élevées en zone péritumorale étaient associées à un ADC bas dans la tumeur elle-même. Conclusions. — L’augmentation des valeurs d’ADC au niveau de l’œdème périlésionnel des tumeurs plus malignes reflète probablement un œdème vasogénique plus important permettant de les différencier des autres lésions. © 2008 Elsevier Masson SAS. All rights reserved.
Introduction Diffusion-weighted imaging (DWI) has become a standard routine sequence in most clinical investigations [1,2]. It involves a simple modification of a spin-echo sequence, where two gradient sensitizing pulses are applied before and after the 180 pulse. Diffusion imaging is sensitive to the molecular motion of tissues, allowing it to be used with great success in the setting of stroke [3,4], where it is thought to be able to differentiate between cytotoxic and vasogenic edema. The occurrence of acute ischemia is accompanied by a decrease in the so-called apparent diffusion coefficient (ADC), which allows quantification of the diffusion characteristics of a given tissue. This has been applied to other pathologies such as brain abscess [5] and some tumors, and brain development [6] and aging [7], as well as for monitoring interventions [8]. The possibility of mapping the ADC allows the quantification of water movement. This has already been applied to tumors of the brain, head and neck [9,10]. Previous studies have shown that ADC values reflect both cellularity and edema [11]. The present study aimed to address its capacity to determine the type of intracranial tumor based on the ADC values from the perilesional edema. To do this, we investigated a series of patients who had different types of intracranial tumors.
Materials and methods Patients Sixteen consecutive patients (nine men, seven women; ages 35 to 82, mean age 58 years) were prospectively included, having been recruited from the neurosurgical clinic of our hospital over a period of six months (January to July 2001).
All patients had undergone brain MRI because of an intracranial tumor. They were informed of the examination they were about to undergo and had given their consent. The patients all subsequently underwent craniotomy and biopsy/surgery to obtain samples of tumor for histopathological examination.
Imaging studies Imaging was performed using a commercial 1.5-T clinical system equipped with a head coil (Siemens Vision, Erlangen, Germany) capable of echo-planar imaging, and with a gradient booster and gradient amplitude of 35 mT/m, a rise time of 150 ms and a slew rate of 170 T/m. First, routine axial T1- (TE: 12 ms; TR: 528 ms; matrix 96 × 200; NEX: 2; 5-mm-thick slices) and T2- (TE: 98/16 ms; TR: 3.176 ms; matrix: 180 × 256; 5-mm-thick slices) weighted imaging was done with postcontrast T1-weighted imaging in three planes after administration of gadolinium (Gd)-DTPA (Magnevist, Schering, Germany) via a cubital-vein cannula, using a power injector (Medrad). Axial isotropic single-shot whole-brain multislice DWI (TR: 4000 ms; TE 36 ms; 18 5-mm-thick slices) was performed, and ADC maps derived from the b0 and bmax images (b values of 0 and 1000 s/mm2 ) by pixel-by-pixel analysis. The images were then transferred to a dedicated workstation. Signal anomalies on DWI and ADC images were determined by two experienced neuroradiologists (KOL, MEK). Regions of interest (ROI) on the ADC maps were also selected in: (i) the areas seen with high signal intensities on T2-weighted images and ADC maps that corresponded to perilesional edema (Figs. 1 and 2); (ii) the tumor itself; and (iii) the contralateral healthy brain. Thus, there were three measurements per patient. Mean ROI size was 1.6 mm2 .
226
R. Guzman et al.
Figure 1 In a patient with a high-grade glioma of the left hemisphere, the peripheral edema can be seen on T2 images and ADC maps. Figure 1 Patient présentant un gliome de haut grade. L’œdème est visualisé sur les séquences T2 et les cartographies ADC.
Statistical analysis Student’s t test was used to compare the two groups of patients with those who had low-grade glioma.
Results On histology, eight patients showed high-grade gliomas, four had cerebral metastases and four had low-grade gliomas (Fig. 3).
ADC values in peritumoral edema The absolute values in the peritumoral edema of the glioblastomas were: median: 1.75 × 10−3 mm2 /s, SD: 13.76;
range: 1.55—1.95 × 10−3 mm2 /s; 25th percentile: 1.65, 75th percentile: 1.88; lower 95% CI of mean: 1.674, upper 95% CI of mean: 1.826 (Fig. 3B). The absolute values in the metastases were: median: 1.61 × 10−3 mm2 /s, SD: 19.13; range: 1.21—1.93 × 10−3 mm2 /s; 25th percentile: 1.525, 75th percentile: 1.89; lower 95% CI of mean: 1.454, upper 95% CI of mean: 1.748. The absolute values in the low-grade gliomas were: median: 1.40 × 10−3 mm2 /s, SD: 20.42; range: 1.08—1.60 × 10−3 mm2 /s; 25th percentile: 1.10, 75th percentile: 1.55; lower 95% CI of mean: 1.17, upper 95% CI of mean: 1.547. The difference was significant for the perilesional edema of glioblastomas compared with low-grade gliomas (P = 0.0071), but not between metastases and low-grade gliomas (P = 0.0811).
Figure 2 In this patient with a metastasis in the left frontal region, there is a central contrast-enhancing nodule on T1 images. The ADC map shows the ROIs in the lesion, in the edema and in the contralateral brain. Figure 2 Patient présentant une métastase de la région frontale gauche. Un nodule prenant le contraste en T1 est entouré d’un œdème visible sur les images ADC sur lesquelles sont placées les régions d’intérêt.
Contribution of the apparent diffusion coefficient in perilesional edema
227
Figure 3 ADC values in three types of brain tumor: (A): relationships between the pathological compared with normal tissues; (B): in the areas of perilesional edema; and (C): in the lesions themselves. Met: metastases; LGG: low-grade gliomas; GBM: glioblastomas. Figure 3 Valeurs ADC dans les différents types de tumeurs cérébrales : (A) : rapport de l’ADC entre le côté pathologique et le côté normal ; (B) : valeurs d’ADC dans l’œdème périlésionnel ; (C) : valeurs d’ADC dans les lésions elles-mêmes. Mets : métastases ; LGG : gliomes de bas grade ; GBM : glioblastomes.
Ratios of edema compared with contralateral healthy brain The ratios for the glioblastomas were: median: 2.38, SD: 0.41; range: 1.7—2.9; 25th percentile: 1.8, 75th percentile: 2.6; lower 95% CI of mean: 2.013, upper 95% CI of mean: 2.476 (Fig. 3A). The ratios for the metastases were: median: 2.143, SD: 0.27; range: 1.66—2.57; 25th percentile: 1.86, 75th percentile: 2.25; lower 95% CI of mean: 1.9, upper 95% CI of mean: 2.3. The ratios for the low-grade gliomas were: median: 1.57, SD: 0.27; range: 1.24—2.05; 25th percentile: 1.3, 75th percentile: 1.74; lower 95% CI of mean: 1.32, upper 95% CI of mean: 1.83.
ADC values in the tumors themselves The intratumoral values for the glioblastomas were: median: 1.30 × 10−3 mm2 /s, SD: 16.75; range: 1.02—1.53 × 10−3 mm2 /s; 25th percentile: 1.16, 75th percentile: 1.39; lower 95% CI of mean: 1.124, upper 95% CI of mean: 1.433 (Fig. 3C). The values for the metastases were: median: 1.44 × 10−3 mm2 /s, SD: 21.73; range: 110—165 × 10−3 mm2 /s; 25th percentile: 119, 75th percentile: 163; lower 95% CI of mean: 119.2, upper 95% CI of mean: 164.8. Values for the low-grade gliomas were: median: 1.95 × 10−3 mm2 /s, SD: 10.78; range: 1.82—2.08 × 10−3 mm2 /s; 25th percentile: 1.83, 75th percentile: 2.035; lower 95% CI of mean: 1.802, upper 95% CI of mean: 2.07. The differences were significant for glioblastomas compared with low-grade gliomas (P = 0.0001) as well as for metastases compared with low-grade gliomas (P = 0.0011). The ratios for brain tumors compared with the contralateral healthy brain were: 0.95 for glioblastomas; 1.06 for metastases; and 1.43 for low-grade gliomas.
Discussion This study showed that both high-grade gliomas and metastatic brain tumors have higher ADC values in the per-
ilesional edema than do low-grade gliomas, indicating a higher water content and greater tissue displacement due to vasogenic edema, and probably secondary to a more aggressive histological reaction. Also, there is a tendency for the more malignant lesions to have lower ADC values within the tumor itself, probably due to necrosis. Therefore, DWIderived ADC maps can be helpful in differentiating tumor malignancy, and provide information not just concerning the tumor, but also about tumor edema extent. Regarding highgrade brain tumors, the study by Castillo et al. [12] found that such tumors have specific values that, however, could not be differentiated from those in the peripheral edema. Other, earlier studies have postulated that diffusion imaging could differentiate the tumor itself from edema [13]. As regards perilesional edema, similar results have been reported in the literature: Lu et al. [14] reported that ADC measurements can differentiate cytotoxic from vasogenic edema, and also between metastases and gliomas: they found increased diffusibility within the vasogenic edema surrounding both high-grade gliomas and metastatic tumors, reflecting increased extracellular water. However, the authors also found that peritumoral fractional anisotropy (FA) demonstrated no statistically significant differences. The FA changes observed around gliomas, therefore, can be attributed not only to increased water content, but also to tumor infiltration [15]. There have also been studies regarding tumor cellularity and diffusibility: increased cellularity is associated with a lower ADC, and lower ADC values appear to be associated with higher-grade tumors [16]. In a study of ADC values and T2 values, Oh et al. [17] found that ADCs for gliomas were similar to those for meningiomas and metastases in all regions. However, tumor T2 values for gliomas were significantly higher than those for meningiomas and metastases. Localized-edema T2 values for meningiomas and metastases were significantly higher than those for gliomas. Peripheraledema T2 values for gliomas were similar to those for meningiomas and metastases. In addition, they found that using a combination of ADC and T2 values resulted in improved tissue characterization of these lesions. They also found significant correlations between ADC and T2 values. As with this study and that of Baehring et al. [18], we found a decrease in the central ADC of glioblastomas.
228 Diffusion imaging has been used in the appraisal of tumors and may even have the potential to differentiate between tumor subtypes. However, on reviewing the literature, it is clear that there is overlap between values [19]. Thus, a certain degree of care should be taken before relying on this parameter too strictly [20]. Nevertheless, its findings can certainly be used when carrying out multimodality imaging [21]. Toh et al. [22] did a study of the ADC in the peritumoral edema of metastases and meningiomas, and found that all values were significantly different between the two types of tumors. Provenzale et al. [23] examined the FA and mean ADC values in peritumoral hyperintense regions, and found that mean FA values were significantly higher for meningiomas than for gliomas, while mean ADCs in peritumoral normal-looking white matter were not significantly different. Overall, this means that ADC mapping is a further means for obtaining improved tissue characterization. This, in addition to the information provided by other techniques such as spectroscopy [24], perfusion [25,26] and diffusion tensor imaging [23], it should improve the initial diagnosis and work-up. Indeed, because of the often nonhomogeneous nature of tumors such as glioblastomas, a single measurement in a single area will not necessarily be helpful; it is the use of many parameters that will aid in improving our understanding of the nature of any lesion.
Conclusion ADC mapping of brain tumors and assessment of the ADC in the peritumoral lesion appears to be promising in the differentiation of tumors and in the assessment of local invasiveness. Glioblastomas, while having a low central ADC, have a high ADC in the periphery, reflecting a central high cellularity and peripheral vasogenic edema.
References [1] Le Bihan D, Breton E, Lallemand D, Grenier P, Cabanis E, Laval-Jeantet M. MR imaging of intravoxel incoherent motions: application to diffusion and perfusion in neurologic disorders. Radiology 1986;161(2):401—7. [2] Le Bihan D, Breton E, Lallemand D, Desbleds MT, Aubin ML, Vignaud J, et al. Contribution of intravoxel incoherent motion (IVIM) imaging to neuroradiology. J Neuroradiol 1987;14(4):295—312. [3] Lovblad KO, Laubach HJ, Baird AE, Curtin F, Schlaug G, Edelman RR, et al. Clinical experience with diffusion-weighted MR in patients with acute stroke. AJNR Am J Neuroradiol 1998;19(6):1061—6. [4] Moseley ME, Cohen Y, Mintorovitch J, Chileuitt L, Shimizu H, Kucharczyk J, et al. Weinstein PR Early detection of regional cerebral ischemia in cats: comparison of diffusionand T2-weighted MRI and spectroscopy. Magn Reson Med 1990;14:330—46. [5] Guzman R, Barth A, Lövblad KO, El-Koussy M, Weis J, Schroth G, et al. Use of diffusion-weighted magnetic resonance imaging in differentiating purulent brain processes from cystic brain tumors. J Neurosurg 2002;97(5):1101—7. [6] Lövblad KO, Schneider J, Ruoss K, Steinlin M, Fusch C, Schroth G. Isotropic apparent diffusion coefficient mapping of postnatal cerebral development. Neuroradiology 2003;45(6):400—3.
R. Guzman et al. [7] Lövblad KO, Delavelle J, Wetzel S, Kelekis AD, Assal F, Palmesino M, et al. ADC mapping of the aging frontal lobes in mild cognitive impairment. Neuroradiology 2004;46(4): 282—6. [8] Lövblad KO, Plüschke W, Remonda L, Gruber-Wiest D, Do DD, Barth A, et al. Diffusion-weighted MRI for monitoring neurovascular interventions. Neuroradiology 2000;42(2):134—8. [9] Matsushima N, Maeda M, Takamura M, Takeda K. Apparent diffusion coefficients of benign and malignant salivary gland tumors. Comparison to histopathological findings. J Neuroradiol 2007;34(3):183—9. [10] Suzuki C, Maeda M, Hori K, Kozuka Y, Sakuma H, Taki W, et al. Apparent diffusion coefficient of pituitary macroadenoma evaluated with line-scan diffusion-weighted imaging. J Neuroradiol 2007;34(4):228—35. [11] Rodallec M, Colombat M, Krainik A, Kalamaridès M, Redondo A, Feydy A. Diffusion-weighted MR imaging and pathologic findings in adult cerebellar medulloblastoma. J Neuroradiol 2004;31(3):234—7. [12] Castillo M, Smith JK, Kwock L, Wilber K. Apparent diffusion coefficients in the evaluation of high-grade cerebral gliomas. AJNR Am J Neuroradiol 2001;22(1):60—4. [13] Eis M, Els T, Hoehn-Berlage M, Hossmann KA. Quantitative diffusion MR imaging of cerebral tumor and edema. Acta Neurochir Suppl (Wien) 1994;60:344—6. [14] Lu S, Ahn D, Johnson G, Cha S. Peritumoral diffusion tensor imaging of high-grade gliomas and metastatic brain tumors. AJNR Am J Neuroradiol 2003;24(5):937—41. [15] Lu S, Ahn D, Johnson G, Law M, Zagzag D, Grossman RI. Diffusion-tensor MR imaging of intracranial neoplasia and associated peritumoral edema: introduction of the tumor infiltration index. Radiology 2004;232(1):221—8. [16] Sugahara T, Korogi Y, Kochi M, Ikushima I, Shigematu Y, Hirai T, et al. Usefulness of diffusion-weighted MRI with echo-planar technique in the evaluation of cellularity in gliomas. J Magn Reson Imaging 1999;9(1):53—60. [17] Oh J, Cha S, Aiken AH, Han ET, Crane JC, Stainsby JA, et al. Quantitative apparent diffusion coefficients and T2 relaxation times in characterizing contrast enhancing brain tumors and regions of peritumoral edema. J Magn Reson Imaging 2005;21(6):701—8. [18] Baehring JM, Bi WL, Bannykh S, Piepmeier JM, Fulbright RK. Diffusion MRI in the early diagnosis of malignant glioma. J Neurooncol 2007;82(2):221—5. [19] Provenzale JM, Mukundan S, Barboriak DP. Diffusion-weighted and perfusion MR imaging for brain tumor characterization and assessment of treatment response. Radiology 2006;239(3):632—49. [20] Tozer DJ, Jäger HR, Danchaivijitr N, Benton CE, Tofts PS, Rees JH, et al. Apparent diffusion coefficient histograms may predict low-grade glioma subtype. NMR Biomed 2007;20(1): 49—57. [21] Schneider JF, Viola A, Confort-Gouny S, Ayunts K, Le Fur Y, Viout P, et al. Infratentorial pediatric brain tumors: the value of new imaging modalities. J Neuroradiol 2007;34(1): 49—58. [22] Toh CH, Wong AM, Wei KC, Ng SH, Wong HF, Wan YL. Peritumoral edema of meningiomas and metastatic brain tumors: differences in diffusion characteristics evaluated with diffusiontensor MR imaging. Neuroradiology 2007;49(6):489—94. [23] Provenzale JM, McGraw P, Mhatre P, Guo AC, Delong D. Peritumoral brain regions in gliomas and meningiomas: investigation with isotropic diffusion-weighted MR imaging and diffusiontensor MR imaging. Radiology 2004;232(2):451—60. [24] Alimenti A, Delavelle J, Lazeyras F, Yilmaz H, Dietrich PY, de Tribolet N, et al. Monovoxel 1H magnetic resonance spectroscopy in the progression of gliomas. Eur Neurol 2007;58(4): 198—209.
Contribution of the apparent diffusion coefficient in perilesional edema [25] Cotton F, Ongolo-Zogo P, Luois-Tisserand G, Streichenberger N, Hermier M, Jouvet A, et al. IRM de diffusion—perfusion dans l’évaluation des lymphomes cérébraux. J Neuroradiol 2006;33:220—8.
229
[26] Guyer S, Kosel M, Altrichter S, El-Koussy M, Haemmig R, Fisch HU, et al. Pattern of regional cerebral blood-flow changes induced by acute heroin administration — a perfusion MRI study. J Neuroradiol 2007;34(5):322—9.