Magnetic Resonance Imaging 32 (2014) 428–432
Contents lists available at ScienceDirect
Magnetic Resonance Imaging journal homepage: www.mrijournal.com
Cervical spondylosis: Evaluation of microstructural changes in spinal cord white matter and gray matter by diffusional kurtosis imaging Masaaki Hori a,⁎, Satoshi Tsutsumi b, Yukimasa Yasumoto b, Masanori Ito b, Michimasa Suzuki a, Fumine S. Tanaka a, Shinsuke Kyogoku c, Masanobu Nakamura d, Takashi Tabuchi d, Issei Fukunaga a, e, Yuriko Suzuki a, f, Koji Kamagata a, Yoshitaka Masutani g, Shigeki Aoki a a
Department of Radiology, Juntendo University School of Medicine, Tokyo, Japan Department of Neurological Surgery, Juntendo University Urayasu Hospital, Chiba, Japan c Department of Radiology, Juntendo University Urayasu Hospital, Chiba, Japan d Radiology, Yaesu Clinic, Tokyo, Japan e Department of Health Science, Graduate School of Human Health Sciences, Tokyo Metropolitan University, Tokyo, Japan f Philips Electronics Japan, Tokyo, Japan g Department of Radiology, The University of Tokyo Hospital, Tokyo, Japan b
a r t i c l e
i n f o
Article history: Received 11 August 2013 Revised 20 January 2014 Accepted 24 January 2014 Keywords: Cervical spondylosis Diffusional kurtosis Fractional anisotropy Apparent diffusion coefficient Spinal cord Gray matter
a b s t r a c t Introduction: We investigated microstructural changes in the spinal cord, separately for white matter and gray matter, in patients with cervical spondylosis by using diffusional kurtosis imaging (DKI). Methods: We studied 13 consecutive patients with cervical myelopathy (15 affected sides and 11 unaffected sides). After conventional magnetic resonance (MR) imaging, DKI data were acquired by using a 3 T MR imaging scanner. Values for fractional anisotropy (FA), apparent diffusion coefficient (ADC), and mean diffusional kurtosis (MK) were calculated and compared between unaffected and affected spinal cords, separately for white matter and gray matter. Results: Tract-specific analysis of white matter in the lateral funiculus showed no statistical differences between the affected and unaffected sides. In gray matter, only MK was significantly lower in the affected spinal cords than in unaffected spinal cords (0.60 ± 0.18 vs. 0.73 ± 0.13, P = 0.0005, Wilcoxon’s signed rank test). Conclusions: MK values in the spinal cord may reflect microstructural changes and gray matter damage and can potentially provide more information beyond that obtained with conventional diffusion metrics. © 2014 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-SA license (http://creativecommons.org/licenses/by-nc-sa/3.0/).
1. Introduction Cervical spondylosis is a common degenerative disease that causes several types of motor and sensory dysfunction. Routine clinical magnetic resonance (MR) imaging (for example, T2-weighted imaging) to evaluate pathological changes in this disease is of limited use because the correlation between the MR findings and clinical symptoms is weak [1]. Diffusion tensor imaging (DTI) has been added to conventional MR imaging in many investigations of the spinal cord to evaluate microstructural changes [2]. Changes in DTI signals depend on the diffusivity of water molecules in a particular environment. DTIderived quantitative metrics such as the fractional anisotropy (FA) ⁎ Corresponding author at: Department of Radiology, School of Medicine, Juntendo University, 2-1-1 Hongo, Bunkyo-ku, Tokyo 113–8421, Japan. Tel.: +81 3 3813 3111; fax: +81 3 3816 0958. E-mail address: [email protected] (M. Hori).
and apparent diffusion coefficient (ADC) show promise as biomarkers in evaluating the microstructural pathology of the cervical spinal cord. For example, reduced FA and increased ADC have been reported at damaged spinal cord regions regardless of whether abnormal signal intensity in the spinal cord was observed on conventional MR images [3–5]. Moreover, a recently introduced extension of the DTI technique called diffusional kurtosis imaging (DKI) [6,7] has shown greater promise than DTI in evaluating the microstructure and pathologic condition of neuronal tissue [8–14], especially gray matter [15,16]. For evaluation of the spinal cord, DKI can provide a more comprehensive characterization of lesions and changes of white or gray matter in patients with multiple sclerosis [17]. Furthermore, as noted in a recent study [18], the mean diffusional kurtosis (MK) can provide additional information on the spinal cord in patients with cervical spondylosis. However, white matter and gray matter were not analyzed separately in the report and were treated as a single unit when setting regions of interest (ROIs). Tract-specific analysis by using white matter tractograms
0730-725X/$ – see front matter © 2014 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-SA license (http://creativecommons.org/ licenses/by-nc-sa/3.0/). http://dx.doi.org/10.1016/j.mri.2014.01.018
M. Hori et al. / Magnetic Resonance Imaging 32 (2014) 428–432 Table 1 Demographic characteristics of subjects. Sex (male: female) Mean Age (SD) Symptoms⁎ Muscle weakness Numbness Pain Hypalgesia Apraxia Neck stiffness
5: 8 54.9 (10.8) 12 11 9 4 3 1
⁎ Multiple symptoms were reported by some patients.
enables more precise measurements and better anatomical localization of white matter. The purpose of this study was to investigate the use of MK to estimate changes in the spinal cord, separately for white matter and gray matter, in patients with early cervical spondylosis. 2. Materials and methods 2.1. Participants Thirteen consecutive patients diagnosed with cervical myelopathy by clinical signs and symptoms participated in this study. Their demographic characteristics are summarized in Table 1. Prior to the study, the research protocol was approved by the institutional review board, and informed consent was obtained from each patient. The exclusion criteria were as follows: the presence of other intraspinal diseases such as tumors, a history of neck surgery for any disease, or unsatisfactory image quality for calculating diffusion metrics.
2.2. Image acquisition All images were acquired on a 3 T MR scanner (Achieva; Philips Medical Systems, Best, The Netherlands). The imaging parameters for DKI were as follows: repetition time/echo time, 10758/88 ms; number of excitations, two; slice thickness/gap, 4/0 mm; number of slices, 32; field of view, 64 × 64 mm; matrix, 128 × 128 reconstructed; imaging time, approximately 13 min; and four b-values (0, 700, 1400, and 2100 s/mm2) with diffusion encoding in 6 directions for each b-value. The gradient length (δ) and time between the two leading edges of the diffusion gradient (Δ) were 9.8 and
429
44.1 ms, respectively. A reduced field-of-view technique was used to improve image quality [19,20]. Before DKI, conventional turbo spinecho T1- and T2-weighted sagittal and axial images were obtained. The imaging parameters for sagittal images were as follows: repetition time/echo time, 400/10 ms for T1-weighted imaging (T1WI) and 3246/128 ms for T2-weighted imaging (T2WI); echo train length, 4 for T1WI and 36 for T2WI; number of excitations, two; slice thickness/gap, 3/0.3 mm; number of slices, 11; field of view, 250 × 250 mm; and matrix, 512 × 512. Imaging parameters for the axial images were as follows: repetition time/echo time, 726/10 ms for T1WI and 6196/ 93 ms for T2WI; echo train length, 5 for T1WI and 36 for T2WI; number of excitations, two; slice thickness/gap, 4/0.4 mm; number of slices, 24; field of view, 160 × 160 mm; and matrix, 512 × 512.
2.3. Analyses of DTI, tractography, and DKI Analyses of DTI, tractography, and DKI were performed by using the free software dTV II FZRx and Volume-One 1.81 (Image Computing and Analysis Laboratory, Department of Radiology, The University of Tokyo Hospital, Tokyo, Japan) [21] on an independent Windows PC. First, maps for FA, ADC, and MK were calculated. The FA and ADC maps were established on the basis of a conventional monoexponential model that assumes a Gaussian probability diffusion function, by using data at b-values of 0 and 700 s/mm 2. Next, we performed diffusion tensor tractography of the bilateral lateral funiculus with threshold values for the termination of fiber tracking set to FA N 0.18. We recognized the advantages of the use of multiple b-values or DKI tractography [22]; however, such advanced fiber tracking was not implemented in our software. Identification of fiber tracts was initiated by placing a seed ROI of 2 pixels in diameter in the lateral funiculus on axial FA maps at spinal canal levels C3–C4 (Fig. 1). A tractographic image of the lateral funiculus was then generated for each patient (Fig. 2). The tract was divided into spinal canal levels C1–C2, C2–C3, C3–C4, C4–C5, C5–C6, and C6–C7 by manually by referring to T1- and T2-weighted images, and each segment of the tractogram was voxelized. The ADC, FA, and MK values in coregistered voxels were then calculated and compared between the affected and unaffected sides, as diagnosed on the basis of clinical symptoms and findings. A subgroup analysis was also performed for 7 patients in whom the damaged spinal level and affected side were clearly identified for the corresponding clinical symptoms. ROIs that conformed to the size and shape of the gray matter on T2-weighted images were placed manually on the gray matter near the tractogram of the
Fig. 1. Placement of ROIs in the lateral funiculus. A representative FA map with ROI (left) and the corresponding MK map (right).
430
M. Hori et al. / Magnetic Resonance Imaging 32 (2014) 428–432
lateral funiculus on the FA map itself (Fig. 3), because the T2weighted images could not be overlaid on the FA map owing to differences in resolution at the damaged spinal level. Diffusion
Fig. 3. Placement of ROIs for comparison of affected and unaffected sides. A gray matter ROI (blue area) is shown near the tractographic image (orange lines) on the spinal cord.
metrics including ADC, FA, and MK of the gray matter were compared between the affected and unaffected sides. 2.4. Statistical analysis Statistical comparisons were performed with Wilcoxon’s signed rank test by using IBM SPSS Statistics software (version 19.0; SPSS, Chicago, IL). The level of statistical significance was set at P b 0.05. 3. Results In all patients, DKI data of good image quality were successfully obtained. Moreover, white matter tractography of the bilateral lateral funiculus was successful, and values for FA, ADC, and MK were obtained (Table 2). There were 15 affected and 11 unaffected sides in 13 patients. Tract-specific analysis of the lateral funiculus showed no statistical differences between the affected and unaffected sides (Wilcoxon’s signed rank test). Values (mean ± standard deviation) of FA, ADC (10 −3 mm 2/s), and MK for gray matter on the unaffected side were 0.55 ± 0.11, 1.19 ± 0.12, and 0.73 ± 0.13, respectively. The corresponding values for gray matter on the affected side were 0.50 ± 0.08, 1.15 ± 0.18, and 0.60 ± 0.18, respectively (Fig. 4). Only MK of the gray matter was significantly lower on the affected side than on the unaffected side (P = 0.0005, Wilcoxon’s signed rank test). 4. Discussion
Fig. 2. Tractographic image and division of the tract. A representative tractographic image of the lateral funiculus (a). The tract was divided by spinal canal level on the diffusion metric maps (b, ADC map) in reference to a conventional T2-weighted image (c).
In patients with cervical spondylosis, previous studies with diffusion metrics showed results, in which FA decreased and ADC increased in the affected spinal cord [3,4]. However, our tractspecific analysis of white matter showed no statistical difference between affected and unaffected sides in the cervical cord. Equivocal evidence in the literatures suggests that diffusion metrics for white matter are sensitive to other factors. For example, normal variations may also affect the FA values: in a previous study, diffusion metrics of the normal spinal cord were different at each spinal level [23], and a decrease of FA with age has been reported for the cervical spinal cord [24], which is vulnerable to aging. This suggests that the variation in diffusion metrics due to pathologic changes in the white matter of the spinal cord may be smaller than the variation across spinal cord levels and aging. Hence, a larger sample size may be required to detect abnormality due to pathologic changes.
M. Hori et al. / Magnetic Resonance Imaging 32 (2014) 428–432
431
Table 2 Comparison of diffusion metrics in the white matter of spinal cords between unaffected and affected sides. C1/C2
C2/C3
C3/C4
C4/C5
C5/C6
C6/C7
Unaffected side FA ADC MK
0.74 ± 0.09 0.90 ± 0.08 1.01 ± 0.08
0.74 ± 0.05 0.89 ± 0.06 0.98 ± 0.17
0.72 ± 0.06 0.9 ± 0.07 1.01 ± 0.15
0.64 ± 0.09 0.98 ± 0.09 0.91 ± 0.08
0.54 ± 0.10 0.99 ± 0.11 0.87 ± 0.12
0.53 ± 0.05 0.93 ± 0.09 0.89 ± 0.12
Affected side FA ADC MK
0.73 ± 0.07 0.89 ± 0.08 1.09 ± 0.15
0.70 ± 0.08 0.87 ± 0.07 1.03 ± 0.17
0.69 ± 0.11 0.91 ± 0.08 0.97 ± 0.17
0.63 ± 0.13 0.95 ± 0.10 0.94 ± 0.19
0.52 ± 0.09 0.95 ± 0.12 0.90 ± 0.12
0.51 ± 0.07 0.94 ± 0.10 0.86 ± 0.07
Values represent mean ± SD. Units for FA and MK are dimensionless. ADC values are given in 10−3 mm2/s.
The reduction of MK values in affected gray matter can be explained by a microcirculatory disturbance in the spinal cord. Although this explanation is speculative, a histological study [25] has shown abnormalities predominantly within the gray matter, whereas axonal degeneration and obvious demyelination have rarely been seen in cervical myelopathy. These findings suggest that microcirculatory disturbance is an important contributor to spinal cord damage in patients with cervical spondylosis. We found no statistical differences in FA and ADC values in the gray matter, consistent with other reports that have shown advantages of MK over FA and ADC in evaluating gray matter in the brain and spinal cord [15,17]. Therefore, MK offers advantages over FA for assessing the cervical spinal cord, particularly gray matter. A potential limitation of this study is the relatively low maximum b-value (b = 2100 s/mm 2 ) compared with those typically used for DKI in the brain. We chose these settings because using higher b-values in clinical settings leads to severe image degradation in spinal cord imaging. In fact, in a past report, DKI data for maximum b values of 2000 s/mm 2 in 15 out of 50 patients were excluded from analysis because of degraded image quality [18]. Although the maximum b-value used here may be insufficient for extracting the full non-Gaussian effect in the data, we presume that a portion of the effect was extracted because the post-processing procedure revealed a non-mono-exponential curve fit. Clinical considerations overrode the theoretical method in this study. Another limitation is the small number of motion probing gradient (MPG) directions. We used 6 directions to reduce the scan time in clinical use. Jensen et al. have suggested that at least 15 (but ideally more than 30) different MPG directions are required to measure MK [6]. Diffusion metrics such as axial kurtosis or radial kurtosis derived from DKI data
with 15 or more MPG directions may also provide more detailed information on the microstructure of white matter tissue. However, in a report on diffusional kurtosis estimation in multiple sclerosis, others have argued that 6 directions may be sufficient [26]. Although we recognize the usefulness of a greater number of diffusion MPG directions, we considered the lower number to be the more practical option given the limited scan time in clinical use. A third limitation is the heterogeneity of both age and disease severity among the patients. As described above, diffusion metrics including FA and ADC in the cervical spinal cord may be influenced by age-related changes. Moreover, different symptoms (e.g., paralysis and pain) show different abnormalities in diffusional metrics at each spinal cord location [5]. However, here we only compared the diffusion metrics between affected and unaffected sides and not across spinal cord levels. Therefore, longitudinal studies, a larger sample size, and clinical correlations with diffusional metrics are needed in the future to control for the influence of age-related changes and to establish diffusion metrics as clinical biomarkers. In conclusion, MK in the spinal cord may reflect microstructural changes and damage of the spinal cord gray matter. Although further studies of the imaging–pathology relationship are needed, MK has the potential to provide new information beyond that provided by conventional diffusion metrics such as ADC and FA, which are based on the mono-exponential model.
Acknowledgment This work was supported by JSPS KAKENHI Grant Number 25461847.
Fig. 4. Statistical comparison of affected and unaffected sides of the gray matter. Box-and-whisker plots of FA, ADC, and MK for spinal cord gray matter in affected and unaffected sides.
432
M. Hori et al. / Magnetic Resonance Imaging 32 (2014) 428–432
References [1] Singh A, Crockard HA, Platts A, Stevens J. Clinical and radiological correlates of severity and surgery-related outcome in cervical spondylosis. J Neurosurg 2001;94:189–98. [2] Wheeler-Kingshott C, Stroman PW, Schwab JM, Bacon M, Bosma R, Brooks J, et al. The current state-of-the-art of spinal cord imaging — applications. NeuroImage 2013. http://dx.doi.org/10.1016/j.neuroimage.2013.07.014. [3] Demir A, Ries M, Moonen CT, Vital JM, Dehais J, Arne P, et al. Diffusion-weighted MR imaging with apparent diffusion coefficient and apparent diffusion tensor maps in cervical spondylotic myelopathy. Radiology 2003;229:37–43. [4] Sato T, Horikoshi T, Watanabe A, Uchida M, Ishigame K, Araki T, et al. Evaluation of cervical myelopathy using apparent diffusion coefficient measured by diffusion-weighted imaging. AJNR Am J Neuroradiol 2012;33:388–92. [5] Yoo WK, Kim TH, Hai DM, Sundaram S, Yang YM, Park MS, et al. Correlation of magnetic resonance diffusion tensor imaging and clinical findings of cervical myelopathy. Spine J 2013;13:867–76. [6] Jensen JH, Helpern JA, Ramani A, Lu H, Kaczynski K. Diffusional kurtosis imaging: the quantification of non-gaussian water diffusion by means of magnetic resonance imaging. Magn Reson Med 2005;53:1432–40. [7] Tabesh A, Jensen JH, Ardekani BA, Helpern JA. Estimation of tensors and tensorderived measures in diffusional kurtosis imaging. Magn Reson Med 2011;65:823–36. [8] Jensen JH, Falangola MF, Hu C, Tabesh A, Rapalino O, Lo C, et al. Preliminary observations of increased diffusional kurtosis in human brain following recent cerebral infarction. NMR Biomed 2011;24:452–7. [9] Raab P, Hattingen E, Franz K, Zanella FE, Lanfermann H. Cerebral gliomas: diffusional kurtosis imaging analysis of microstructural differences. Radiology 2010;254:876–81. [10] Nakanishi A, Fukunaga I, Hori M, Masutani Y, Takaaki H, Miyajima M, et al. Microstructural changes of the corticospinal tract in idiopathic normal pressure hydrocephalus: a comparison of diffusion tensor and diffusional kurtosis imaging. Neuroradiology 2013;55:971–6. [11] Yoshida M, Hori M, Yokoyama K, Fukunaga I, Suzuki M, Kamagata K, et al. Diffusional kurtosis imaging of normal-appearing white matter in multiple sclerosis: preliminary clinical experience. Jpn J Radiol 2013;31:50–5. [12] Falangola MF, Jensen JH, Tabesh A, Hu C, Deardorff RL, Babb JS, et al. NonGaussian diffusion MRI assessment of brain microstructure in mild cognitive impairment and Alzheimer's disease. Magn Reson Imaging 2013;31:840–6. [13] Grossman EJ, Jensen JH, Babb JS, Chen Q, Tabesh A, Fieremans E, et al. Cognitive impairment in mild traumatic brain injury: a longitudinal diffusional kurtosis and perfusion imaging study. AJNR Am J Neuroradiol 2013;34:951–7.
[14] Kamagata K, Tomiyama H, Motoi Y, Kano M, Abe O, Ito K, et al. Diffusional kurtosis imaging of cingulate fibers in Parkinson disease: comparison with conventional diffusion tensor imaging. Magn Reson Imaging 2013;31:1501–6. [15] Helpern JA, Adisetiyo V, Falangola MF, Hu C, Di Martino A, Williams K, et al. Preliminary evidence of altered gray and white matter microstructural development in the frontal lobe of adolescents with attention-deficit hyperactivity disorder: a diffusional kurtosis imaging study. J Magn Reson Imaging 2011;33:17–23. [16] Cheung MM, Hui ES, Chan KC, Helpern JA, Qi L, Wu EX. Does diffusion kurtosis imaging lead to better neural tissue characterization? A rodent brain maturation study. NeuroImage 2009;45:386–92. [17] Raz E, Bester M, Sigmund EE, Tabesh A, Babb JS, Jaggi H, et al. A better characterization of spinal cord damage in multiple sclerosis: a diffusional kurtosis imaging study. AJNR Am J Neuroradiol 2013. http://dx.doi.org/10.3174/ ajnr.A3512. [18] Hori M, Fukunaga I, Masutani Y, Nakanishi A, Shimoji K, Kamagata K, et al. New diffusion metrics for spondylotic myelopathy at an early clinical stage. Eur Radiol 2012;22:1797–802. [19] Wilm BJ, Svensson J, Henning A, Pruessmann KP, Boesiger P, Kollias SS. Reduced field-of-view MRI using outer volume suppression for spinal cord diffusion imaging. Magn Reson Med 2007;57:625–30. [20] Wilm BJ, Gamper U, Henning A, Pruessmann KP, Kollias SS, Boesiger P. Diffusionweighted imaging of the entire spinal cord. NMR Biomed 2009;22:174–81. [21] Hori M, Fukunaga I, Masutani Y, Taoka T, Kamagata K, Suzuki Y, et al. Visualizing non-Gaussian diffusion: clinical application of q-space imaging and diffusional kurtosis imaging of the brain and spine. Magn Reson Med Sci 2012;11:221–33. [22] Paydar A, Fieremans E, Nwankwo JI, Lazar M, Sheth HD, Adisetiyo V, et al. Diffusional kurtosis imaging of the developing brain. AJNR Am J Neuroradiol 2013. http://dx.doi.org/10.3174/ajnr.A3764. [23] Mamata H, Jolesz FA, Maier SE. Apparent diffusion coefficient and fractional anisotropy in spinal cord: age and cervical spondylosis-related changes. J Magn Reson Imaging 2005;22:38–43. [24] Agosta F, Laganà M, Valsasina P, Sala S, Dall'Occhio L, Sormani MP, et al. Evidence for cervical cord tissue disorganisation with aging by diffusion tensor MRI. NeuroImage 2007;36:728–35. [25] al-Mefty O, Harkey HL, Marawi I, Haines DE, Peeler DF, Wilner HI, et al. Experimental chronic compressive cervical myelopathy. J Neurosurg 1993;79:550–61. [26] Lätt J, Nilsson M, Wirestam R, Johansson E, Larsson EM, Stahlberg F, et al. In vivo visualization of displacement-distribution-derived parameters in q-space imaging. Magn Reson Imaging 2008;26:77–87.
Contents lists available at ScienceDirect
Magnetic Resonance Imaging journal homepage: www.mrijournal.com
Cervical spondylosis: Evaluation of microstructural changes in spinal cord white matter and gray matter by diffusional kurtosis imaging Masaaki Hori a,⁎, Satoshi Tsutsumi b, Yukimasa Yasumoto b, Masanori Ito b, Michimasa Suzuki a, Fumine S. Tanaka a, Shinsuke Kyogoku c, Masanobu Nakamura d, Takashi Tabuchi d, Issei Fukunaga a, e, Yuriko Suzuki a, f, Koji Kamagata a, Yoshitaka Masutani g, Shigeki Aoki a a
Department of Radiology, Juntendo University School of Medicine, Tokyo, Japan Department of Neurological Surgery, Juntendo University Urayasu Hospital, Chiba, Japan c Department of Radiology, Juntendo University Urayasu Hospital, Chiba, Japan d Radiology, Yaesu Clinic, Tokyo, Japan e Department of Health Science, Graduate School of Human Health Sciences, Tokyo Metropolitan University, Tokyo, Japan f Philips Electronics Japan, Tokyo, Japan g Department of Radiology, The University of Tokyo Hospital, Tokyo, Japan b
a r t i c l e
i n f o
Article history: Received 11 August 2013 Revised 20 January 2014 Accepted 24 January 2014 Keywords: Cervical spondylosis Diffusional kurtosis Fractional anisotropy Apparent diffusion coefficient Spinal cord Gray matter
a b s t r a c t Introduction: We investigated microstructural changes in the spinal cord, separately for white matter and gray matter, in patients with cervical spondylosis by using diffusional kurtosis imaging (DKI). Methods: We studied 13 consecutive patients with cervical myelopathy (15 affected sides and 11 unaffected sides). After conventional magnetic resonance (MR) imaging, DKI data were acquired by using a 3 T MR imaging scanner. Values for fractional anisotropy (FA), apparent diffusion coefficient (ADC), and mean diffusional kurtosis (MK) were calculated and compared between unaffected and affected spinal cords, separately for white matter and gray matter. Results: Tract-specific analysis of white matter in the lateral funiculus showed no statistical differences between the affected and unaffected sides. In gray matter, only MK was significantly lower in the affected spinal cords than in unaffected spinal cords (0.60 ± 0.18 vs. 0.73 ± 0.13, P = 0.0005, Wilcoxon’s signed rank test). Conclusions: MK values in the spinal cord may reflect microstructural changes and gray matter damage and can potentially provide more information beyond that obtained with conventional diffusion metrics. © 2014 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-SA license (http://creativecommons.org/licenses/by-nc-sa/3.0/).
1. Introduction Cervical spondylosis is a common degenerative disease that causes several types of motor and sensory dysfunction. Routine clinical magnetic resonance (MR) imaging (for example, T2-weighted imaging) to evaluate pathological changes in this disease is of limited use because the correlation between the MR findings and clinical symptoms is weak [1]. Diffusion tensor imaging (DTI) has been added to conventional MR imaging in many investigations of the spinal cord to evaluate microstructural changes [2]. Changes in DTI signals depend on the diffusivity of water molecules in a particular environment. DTIderived quantitative metrics such as the fractional anisotropy (FA) ⁎ Corresponding author at: Department of Radiology, School of Medicine, Juntendo University, 2-1-1 Hongo, Bunkyo-ku, Tokyo 113–8421, Japan. Tel.: +81 3 3813 3111; fax: +81 3 3816 0958. E-mail address: [email protected] (M. Hori).
and apparent diffusion coefficient (ADC) show promise as biomarkers in evaluating the microstructural pathology of the cervical spinal cord. For example, reduced FA and increased ADC have been reported at damaged spinal cord regions regardless of whether abnormal signal intensity in the spinal cord was observed on conventional MR images [3–5]. Moreover, a recently introduced extension of the DTI technique called diffusional kurtosis imaging (DKI) [6,7] has shown greater promise than DTI in evaluating the microstructure and pathologic condition of neuronal tissue [8–14], especially gray matter [15,16]. For evaluation of the spinal cord, DKI can provide a more comprehensive characterization of lesions and changes of white or gray matter in patients with multiple sclerosis [17]. Furthermore, as noted in a recent study [18], the mean diffusional kurtosis (MK) can provide additional information on the spinal cord in patients with cervical spondylosis. However, white matter and gray matter were not analyzed separately in the report and were treated as a single unit when setting regions of interest (ROIs). Tract-specific analysis by using white matter tractograms
0730-725X/$ – see front matter © 2014 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-SA license (http://creativecommons.org/ licenses/by-nc-sa/3.0/). http://dx.doi.org/10.1016/j.mri.2014.01.018
M. Hori et al. / Magnetic Resonance Imaging 32 (2014) 428–432 Table 1 Demographic characteristics of subjects. Sex (male: female) Mean Age (SD) Symptoms⁎ Muscle weakness Numbness Pain Hypalgesia Apraxia Neck stiffness
5: 8 54.9 (10.8) 12 11 9 4 3 1
⁎ Multiple symptoms were reported by some patients.
enables more precise measurements and better anatomical localization of white matter. The purpose of this study was to investigate the use of MK to estimate changes in the spinal cord, separately for white matter and gray matter, in patients with early cervical spondylosis. 2. Materials and methods 2.1. Participants Thirteen consecutive patients diagnosed with cervical myelopathy by clinical signs and symptoms participated in this study. Their demographic characteristics are summarized in Table 1. Prior to the study, the research protocol was approved by the institutional review board, and informed consent was obtained from each patient. The exclusion criteria were as follows: the presence of other intraspinal diseases such as tumors, a history of neck surgery for any disease, or unsatisfactory image quality for calculating diffusion metrics.
2.2. Image acquisition All images were acquired on a 3 T MR scanner (Achieva; Philips Medical Systems, Best, The Netherlands). The imaging parameters for DKI were as follows: repetition time/echo time, 10758/88 ms; number of excitations, two; slice thickness/gap, 4/0 mm; number of slices, 32; field of view, 64 × 64 mm; matrix, 128 × 128 reconstructed; imaging time, approximately 13 min; and four b-values (0, 700, 1400, and 2100 s/mm2) with diffusion encoding in 6 directions for each b-value. The gradient length (δ) and time between the two leading edges of the diffusion gradient (Δ) were 9.8 and
429
44.1 ms, respectively. A reduced field-of-view technique was used to improve image quality [19,20]. Before DKI, conventional turbo spinecho T1- and T2-weighted sagittal and axial images were obtained. The imaging parameters for sagittal images were as follows: repetition time/echo time, 400/10 ms for T1-weighted imaging (T1WI) and 3246/128 ms for T2-weighted imaging (T2WI); echo train length, 4 for T1WI and 36 for T2WI; number of excitations, two; slice thickness/gap, 3/0.3 mm; number of slices, 11; field of view, 250 × 250 mm; and matrix, 512 × 512. Imaging parameters for the axial images were as follows: repetition time/echo time, 726/10 ms for T1WI and 6196/ 93 ms for T2WI; echo train length, 5 for T1WI and 36 for T2WI; number of excitations, two; slice thickness/gap, 4/0.4 mm; number of slices, 24; field of view, 160 × 160 mm; and matrix, 512 × 512.
2.3. Analyses of DTI, tractography, and DKI Analyses of DTI, tractography, and DKI were performed by using the free software dTV II FZRx and Volume-One 1.81 (Image Computing and Analysis Laboratory, Department of Radiology, The University of Tokyo Hospital, Tokyo, Japan) [21] on an independent Windows PC. First, maps for FA, ADC, and MK were calculated. The FA and ADC maps were established on the basis of a conventional monoexponential model that assumes a Gaussian probability diffusion function, by using data at b-values of 0 and 700 s/mm 2. Next, we performed diffusion tensor tractography of the bilateral lateral funiculus with threshold values for the termination of fiber tracking set to FA N 0.18. We recognized the advantages of the use of multiple b-values or DKI tractography [22]; however, such advanced fiber tracking was not implemented in our software. Identification of fiber tracts was initiated by placing a seed ROI of 2 pixels in diameter in the lateral funiculus on axial FA maps at spinal canal levels C3–C4 (Fig. 1). A tractographic image of the lateral funiculus was then generated for each patient (Fig. 2). The tract was divided into spinal canal levels C1–C2, C2–C3, C3–C4, C4–C5, C5–C6, and C6–C7 by manually by referring to T1- and T2-weighted images, and each segment of the tractogram was voxelized. The ADC, FA, and MK values in coregistered voxels were then calculated and compared between the affected and unaffected sides, as diagnosed on the basis of clinical symptoms and findings. A subgroup analysis was also performed for 7 patients in whom the damaged spinal level and affected side were clearly identified for the corresponding clinical symptoms. ROIs that conformed to the size and shape of the gray matter on T2-weighted images were placed manually on the gray matter near the tractogram of the
Fig. 1. Placement of ROIs in the lateral funiculus. A representative FA map with ROI (left) and the corresponding MK map (right).
430
M. Hori et al. / Magnetic Resonance Imaging 32 (2014) 428–432
lateral funiculus on the FA map itself (Fig. 3), because the T2weighted images could not be overlaid on the FA map owing to differences in resolution at the damaged spinal level. Diffusion
Fig. 3. Placement of ROIs for comparison of affected and unaffected sides. A gray matter ROI (blue area) is shown near the tractographic image (orange lines) on the spinal cord.
metrics including ADC, FA, and MK of the gray matter were compared between the affected and unaffected sides. 2.4. Statistical analysis Statistical comparisons were performed with Wilcoxon’s signed rank test by using IBM SPSS Statistics software (version 19.0; SPSS, Chicago, IL). The level of statistical significance was set at P b 0.05. 3. Results In all patients, DKI data of good image quality were successfully obtained. Moreover, white matter tractography of the bilateral lateral funiculus was successful, and values for FA, ADC, and MK were obtained (Table 2). There were 15 affected and 11 unaffected sides in 13 patients. Tract-specific analysis of the lateral funiculus showed no statistical differences between the affected and unaffected sides (Wilcoxon’s signed rank test). Values (mean ± standard deviation) of FA, ADC (10 −3 mm 2/s), and MK for gray matter on the unaffected side were 0.55 ± 0.11, 1.19 ± 0.12, and 0.73 ± 0.13, respectively. The corresponding values for gray matter on the affected side were 0.50 ± 0.08, 1.15 ± 0.18, and 0.60 ± 0.18, respectively (Fig. 4). Only MK of the gray matter was significantly lower on the affected side than on the unaffected side (P = 0.0005, Wilcoxon’s signed rank test). 4. Discussion
Fig. 2. Tractographic image and division of the tract. A representative tractographic image of the lateral funiculus (a). The tract was divided by spinal canal level on the diffusion metric maps (b, ADC map) in reference to a conventional T2-weighted image (c).
In patients with cervical spondylosis, previous studies with diffusion metrics showed results, in which FA decreased and ADC increased in the affected spinal cord [3,4]. However, our tractspecific analysis of white matter showed no statistical difference between affected and unaffected sides in the cervical cord. Equivocal evidence in the literatures suggests that diffusion metrics for white matter are sensitive to other factors. For example, normal variations may also affect the FA values: in a previous study, diffusion metrics of the normal spinal cord were different at each spinal level [23], and a decrease of FA with age has been reported for the cervical spinal cord [24], which is vulnerable to aging. This suggests that the variation in diffusion metrics due to pathologic changes in the white matter of the spinal cord may be smaller than the variation across spinal cord levels and aging. Hence, a larger sample size may be required to detect abnormality due to pathologic changes.
M. Hori et al. / Magnetic Resonance Imaging 32 (2014) 428–432
431
Table 2 Comparison of diffusion metrics in the white matter of spinal cords between unaffected and affected sides. C1/C2
C2/C3
C3/C4
C4/C5
C5/C6
C6/C7
Unaffected side FA ADC MK
0.74 ± 0.09 0.90 ± 0.08 1.01 ± 0.08
0.74 ± 0.05 0.89 ± 0.06 0.98 ± 0.17
0.72 ± 0.06 0.9 ± 0.07 1.01 ± 0.15
0.64 ± 0.09 0.98 ± 0.09 0.91 ± 0.08
0.54 ± 0.10 0.99 ± 0.11 0.87 ± 0.12
0.53 ± 0.05 0.93 ± 0.09 0.89 ± 0.12
Affected side FA ADC MK
0.73 ± 0.07 0.89 ± 0.08 1.09 ± 0.15
0.70 ± 0.08 0.87 ± 0.07 1.03 ± 0.17
0.69 ± 0.11 0.91 ± 0.08 0.97 ± 0.17
0.63 ± 0.13 0.95 ± 0.10 0.94 ± 0.19
0.52 ± 0.09 0.95 ± 0.12 0.90 ± 0.12
0.51 ± 0.07 0.94 ± 0.10 0.86 ± 0.07
Values represent mean ± SD. Units for FA and MK are dimensionless. ADC values are given in 10−3 mm2/s.
The reduction of MK values in affected gray matter can be explained by a microcirculatory disturbance in the spinal cord. Although this explanation is speculative, a histological study [25] has shown abnormalities predominantly within the gray matter, whereas axonal degeneration and obvious demyelination have rarely been seen in cervical myelopathy. These findings suggest that microcirculatory disturbance is an important contributor to spinal cord damage in patients with cervical spondylosis. We found no statistical differences in FA and ADC values in the gray matter, consistent with other reports that have shown advantages of MK over FA and ADC in evaluating gray matter in the brain and spinal cord [15,17]. Therefore, MK offers advantages over FA for assessing the cervical spinal cord, particularly gray matter. A potential limitation of this study is the relatively low maximum b-value (b = 2100 s/mm 2 ) compared with those typically used for DKI in the brain. We chose these settings because using higher b-values in clinical settings leads to severe image degradation in spinal cord imaging. In fact, in a past report, DKI data for maximum b values of 2000 s/mm 2 in 15 out of 50 patients were excluded from analysis because of degraded image quality [18]. Although the maximum b-value used here may be insufficient for extracting the full non-Gaussian effect in the data, we presume that a portion of the effect was extracted because the post-processing procedure revealed a non-mono-exponential curve fit. Clinical considerations overrode the theoretical method in this study. Another limitation is the small number of motion probing gradient (MPG) directions. We used 6 directions to reduce the scan time in clinical use. Jensen et al. have suggested that at least 15 (but ideally more than 30) different MPG directions are required to measure MK [6]. Diffusion metrics such as axial kurtosis or radial kurtosis derived from DKI data
with 15 or more MPG directions may also provide more detailed information on the microstructure of white matter tissue. However, in a report on diffusional kurtosis estimation in multiple sclerosis, others have argued that 6 directions may be sufficient [26]. Although we recognize the usefulness of a greater number of diffusion MPG directions, we considered the lower number to be the more practical option given the limited scan time in clinical use. A third limitation is the heterogeneity of both age and disease severity among the patients. As described above, diffusion metrics including FA and ADC in the cervical spinal cord may be influenced by age-related changes. Moreover, different symptoms (e.g., paralysis and pain) show different abnormalities in diffusional metrics at each spinal cord location [5]. However, here we only compared the diffusion metrics between affected and unaffected sides and not across spinal cord levels. Therefore, longitudinal studies, a larger sample size, and clinical correlations with diffusional metrics are needed in the future to control for the influence of age-related changes and to establish diffusion metrics as clinical biomarkers. In conclusion, MK in the spinal cord may reflect microstructural changes and damage of the spinal cord gray matter. Although further studies of the imaging–pathology relationship are needed, MK has the potential to provide new information beyond that provided by conventional diffusion metrics such as ADC and FA, which are based on the mono-exponential model.
Acknowledgment This work was supported by JSPS KAKENHI Grant Number 25461847.
Fig. 4. Statistical comparison of affected and unaffected sides of the gray matter. Box-and-whisker plots of FA, ADC, and MK for spinal cord gray matter in affected and unaffected sides.
432
M. Hori et al. / Magnetic Resonance Imaging 32 (2014) 428–432
References [1] Singh A, Crockard HA, Platts A, Stevens J. Clinical and radiological correlates of severity and surgery-related outcome in cervical spondylosis. J Neurosurg 2001;94:189–98. [2] Wheeler-Kingshott C, Stroman PW, Schwab JM, Bacon M, Bosma R, Brooks J, et al. The current state-of-the-art of spinal cord imaging — applications. NeuroImage 2013. http://dx.doi.org/10.1016/j.neuroimage.2013.07.014. [3] Demir A, Ries M, Moonen CT, Vital JM, Dehais J, Arne P, et al. Diffusion-weighted MR imaging with apparent diffusion coefficient and apparent diffusion tensor maps in cervical spondylotic myelopathy. Radiology 2003;229:37–43. [4] Sato T, Horikoshi T, Watanabe A, Uchida M, Ishigame K, Araki T, et al. Evaluation of cervical myelopathy using apparent diffusion coefficient measured by diffusion-weighted imaging. AJNR Am J Neuroradiol 2012;33:388–92. [5] Yoo WK, Kim TH, Hai DM, Sundaram S, Yang YM, Park MS, et al. Correlation of magnetic resonance diffusion tensor imaging and clinical findings of cervical myelopathy. Spine J 2013;13:867–76. [6] Jensen JH, Helpern JA, Ramani A, Lu H, Kaczynski K. Diffusional kurtosis imaging: the quantification of non-gaussian water diffusion by means of magnetic resonance imaging. Magn Reson Med 2005;53:1432–40. [7] Tabesh A, Jensen JH, Ardekani BA, Helpern JA. Estimation of tensors and tensorderived measures in diffusional kurtosis imaging. Magn Reson Med 2011;65:823–36. [8] Jensen JH, Falangola MF, Hu C, Tabesh A, Rapalino O, Lo C, et al. Preliminary observations of increased diffusional kurtosis in human brain following recent cerebral infarction. NMR Biomed 2011;24:452–7. [9] Raab P, Hattingen E, Franz K, Zanella FE, Lanfermann H. Cerebral gliomas: diffusional kurtosis imaging analysis of microstructural differences. Radiology 2010;254:876–81. [10] Nakanishi A, Fukunaga I, Hori M, Masutani Y, Takaaki H, Miyajima M, et al. Microstructural changes of the corticospinal tract in idiopathic normal pressure hydrocephalus: a comparison of diffusion tensor and diffusional kurtosis imaging. Neuroradiology 2013;55:971–6. [11] Yoshida M, Hori M, Yokoyama K, Fukunaga I, Suzuki M, Kamagata K, et al. Diffusional kurtosis imaging of normal-appearing white matter in multiple sclerosis: preliminary clinical experience. Jpn J Radiol 2013;31:50–5. [12] Falangola MF, Jensen JH, Tabesh A, Hu C, Deardorff RL, Babb JS, et al. NonGaussian diffusion MRI assessment of brain microstructure in mild cognitive impairment and Alzheimer's disease. Magn Reson Imaging 2013;31:840–6. [13] Grossman EJ, Jensen JH, Babb JS, Chen Q, Tabesh A, Fieremans E, et al. Cognitive impairment in mild traumatic brain injury: a longitudinal diffusional kurtosis and perfusion imaging study. AJNR Am J Neuroradiol 2013;34:951–7.
[14] Kamagata K, Tomiyama H, Motoi Y, Kano M, Abe O, Ito K, et al. Diffusional kurtosis imaging of cingulate fibers in Parkinson disease: comparison with conventional diffusion tensor imaging. Magn Reson Imaging 2013;31:1501–6. [15] Helpern JA, Adisetiyo V, Falangola MF, Hu C, Di Martino A, Williams K, et al. Preliminary evidence of altered gray and white matter microstructural development in the frontal lobe of adolescents with attention-deficit hyperactivity disorder: a diffusional kurtosis imaging study. J Magn Reson Imaging 2011;33:17–23. [16] Cheung MM, Hui ES, Chan KC, Helpern JA, Qi L, Wu EX. Does diffusion kurtosis imaging lead to better neural tissue characterization? A rodent brain maturation study. NeuroImage 2009;45:386–92. [17] Raz E, Bester M, Sigmund EE, Tabesh A, Babb JS, Jaggi H, et al. A better characterization of spinal cord damage in multiple sclerosis: a diffusional kurtosis imaging study. AJNR Am J Neuroradiol 2013. http://dx.doi.org/10.3174/ ajnr.A3512. [18] Hori M, Fukunaga I, Masutani Y, Nakanishi A, Shimoji K, Kamagata K, et al. New diffusion metrics for spondylotic myelopathy at an early clinical stage. Eur Radiol 2012;22:1797–802. [19] Wilm BJ, Svensson J, Henning A, Pruessmann KP, Boesiger P, Kollias SS. Reduced field-of-view MRI using outer volume suppression for spinal cord diffusion imaging. Magn Reson Med 2007;57:625–30. [20] Wilm BJ, Gamper U, Henning A, Pruessmann KP, Kollias SS, Boesiger P. Diffusionweighted imaging of the entire spinal cord. NMR Biomed 2009;22:174–81. [21] Hori M, Fukunaga I, Masutani Y, Taoka T, Kamagata K, Suzuki Y, et al. Visualizing non-Gaussian diffusion: clinical application of q-space imaging and diffusional kurtosis imaging of the brain and spine. Magn Reson Med Sci 2012;11:221–33. [22] Paydar A, Fieremans E, Nwankwo JI, Lazar M, Sheth HD, Adisetiyo V, et al. Diffusional kurtosis imaging of the developing brain. AJNR Am J Neuroradiol 2013. http://dx.doi.org/10.3174/ajnr.A3764. [23] Mamata H, Jolesz FA, Maier SE. Apparent diffusion coefficient and fractional anisotropy in spinal cord: age and cervical spondylosis-related changes. J Magn Reson Imaging 2005;22:38–43. [24] Agosta F, Laganà M, Valsasina P, Sala S, Dall'Occhio L, Sormani MP, et al. Evidence for cervical cord tissue disorganisation with aging by diffusion tensor MRI. NeuroImage 2007;36:728–35. [25] al-Mefty O, Harkey HL, Marawi I, Haines DE, Peeler DF, Wilner HI, et al. Experimental chronic compressive cervical myelopathy. J Neurosurg 1993;79:550–61. [26] Lätt J, Nilsson M, Wirestam R, Johansson E, Larsson EM, Stahlberg F, et al. In vivo visualization of displacement-distribution-derived parameters in q-space imaging. Magn Reson Imaging 2008;26:77–87.
