Magnetic Resonance Imaging 33 (2015) 95–101
Contents lists available at ScienceDirect
Magnetic Resonance Imaging journal homepage: www.mrijournal.com
In vivo evaluation of rabbit sciatic nerve regeneration with diffusion tensor imaging (DTI): correlations with histology and behavior Tetsuro Yamasaki a, 1, Hiroyoshi Fujiwara a, 1, Ryo Oda a,⁎, Yasuo Mikami a, 1, Takumi Ikeda a, 1, Masateru Nagae a, 1, Toshiharu Shirai a, 1, Shinsuke Morisaki a, 1, Kazuya Ikoma a, 1, Miwako Masugi-Tokita b, 2, Kei Yamada c, 3, Mitsuhiro Kawata b, 2, Toshikazu Kubo a, 1 a b c
Department of Orthopaedics, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kawaramachi-Hirokoji, Kamigyo-ku, Kyoto, 602-8566, Japan Department of Anatomy and Neurobiology, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kawaramachi-Hirokoji, Kamigyo-ku, Kyoto, 602-8566, Japan Department of Radiology, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kawaramachi-Hirokoji, Kamigyo-ku, Kyoto, 602-8566, Japan
a r t i c l e
i n f o
Article history: Received 8 March 2014 Revised 23 August 2014 Accepted 22 September 2014 Keywords: Peripheral nerve Regeneration Diffusion tenor imaging Rabbit Fractional anisotropy Toe-spreading reflex
a b s t r a c t Diffusion tensor imaging (DTI) is widely used in the study of the central nervous system. DTI represents a potential diagnostic tool for the peripheral nerve. However, more detailed information is needed for application of DTI in the clinical setting. In this study, peripheral degeneration and regeneration were evaluated using DTIbased analyses in a rabbit model. The changes in DTI parameters were compared to histological and functional changes after nerve injury. We used a high magnetic field (7.04 T) MRI system. Japanese white male rabbits were used as the model of sciatic nerve crush injury. MR images were obtained before injury and at 2, 4, 6 and 8 weeks post-injury. The DTI parameters of fractional anisotropy (FA), axial diffusivity (λ||), and radial diffusivity (λ⊥) were calculated. Our results showed decreased FA and increased λ⊥ during the degenerative phase after sciatic nerve injury. In contrast, increased FA and decreased λ⊥ were observed during the regenerative phase. FA changes were correlated with axon number and with motor function recovery, assessed with the toe-spreading index. This study clearly demonstrates the validity of applying DTI parameters to the in vivo evaluation of peripheral nerve regeneration. Furthermore, results suggest that DTI can be a potent tool for predicting the extent of functional recovery after peripheral nerve injury. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Peripheral nerve injures are common in the clinical field. They result from various circumstances such as trauma, inflammation, and peripheral neuropathies [1,2]. Accurate assessment of the location and severity of nerve injury is essential for selection of treatment protocols. Physical examination and electrophysiological experiments play a key role in assessing peripheral nerve injury [3,4]. However, the reliability of assessment using these methods is highly dependent on the clinician's experience. Additionally, electrophysiological examination, requiring insertion of needles
⁎ Corresponding author. Tel.: +81 75 251 5549; fax: +81 75 251 5841. E-mail address: [email protected] (R. Oda). 1 Tel.: +81 75 251 5549; fax: +81 75 251 5841. 2 Tel.: +81 75 251 5301; fax: +81 75 251 5306. 3 Tel.: + 81 75 251 5620; fax: +81 75 251 5840.
http://dx.doi.org/10.1016/j.mri.2014.09.005 0730-725X/© 2014 Elsevier Inc. All rights reserved.
into the muscles, is an invasive procedure. Therefore, a new method that can noninvasively and objectively evaluate peripheral nerve injury needs to be established. Magnetic resonance imaging (MRI) is an effective method for detecting various structural changes, such as edema and necrosis, within soft tissues. A number of studies have also indicated that MRI can be utilized for the assessment of peripheral nerves [5–9]. In the nervous system, diffusion tensor imaging (DTI) has been accepted as a suitable tool for detecting nerve fiber orientation, and can provide quantitative evaluation based on the diffusional movement of water molecules through tissue [10]. DTI can accurately depict nerve fibers and microstructure within the central nervous system (CNS) [11–13]. DTI is expected to be utilized as a new diagnostic tool for peripheral nerve injury, as it can evaluate nerve degeneration and regeneration noninvasively, objectively, and quantitatively. Several animal studies have been conducted to validate DTI for use in the peripheral nervous system (PNS) [14–17]. However, it remains unclear whether the evaluation of peripheral nerve injury by DTI can be applied to humans. More detailed foundational data must be collected for application of DTI in the clinical setting. In this study, we investigated the validity of
96
T. Yamasaki et al. / Magnetic Resonance Imaging 33 (2015) 95–101
DTI for evaluating peripheral nerve degeneration and regeneration in a rabbit model by comparing DTI parameters with histological and motor function data after sciatic nerve injury. 2. Materials and methods 2.1. Animal model
was 2 mm. The most proximal image plane includes silicone tube surrounding the sciatic nerve (Fig. 1b). The most distal slice included the sciatic nerve 4 mm distal to the crush lesion (Fig. 1e). The sciatic nerve was identified by tracing these serial images. Imaging plane for DTI was set at the same sites as PDWI (Fig. 1f, g, h). DTI parameter was assessed in the most distal slice to exclude the effect of acute inflammatory reaction at crush lesion. 2.4. DTI data analyses
In all of the studies, male Japanese white rabbits (Shimizu Laboratory Supplies, Kyoto, Japan), weighing 3.0–3.5 kg, were used. The rabbits were maintained on a 12-hour light/dark schedule with free access to food and water. All of the surgical and experimental procedures were approved by the committee for animal research at our institution and were performed in accordance with the guidelines of the National Institutes of Health regarding animal care. Six rabbits underwent serial MRI scans and behavioral evaluation longitudinally before and after sciatic nerve injury. Histological evaluation was performed on 3 rabbits in each time points (before injury and 2, 4, 6, 8 weeks after injury) separately. Thus the total of rabbits used for histological analysis was 15.
The six independent elements of the diffusion tensor were calculated from each diffusion-weighted image. The resulting tensor element maps were used to derive the eigenvalues of the tensor (λ1, λ2 and λ3) by matrix diagonalization. The regions of interest (ROIs) were manually placed on the sciatic nerve as identified on the PDWI (Fig. 1e). The mean size of ROIs was 1.3 ± 0.2 mm 2. The signal-tonoise ratios measured in the images with minimum b value (b = 0) were greater than 19 for all experiments. Fractional anisotropy (FA), axial diffusivity (λ||), and radial diffusivity (λ⊥) were calculated using the following equations:
2.2. Surgical procedures
λjj ¼ λ1
ð1Þ
All of the rabbits were deeply anesthetized with a mixture of 10% oxygen, air and 2.0% isoflurane (Abbott, Osaka, Japan). The right sciatic nerve was exposed [18] and ligated at the proximal thigh level with a 3–0 silk suture to produce a crush injury. Ligation was performed with a force of 0.7 N applied for 5 minutes. To identify the injury appropriately on MR images, the lesion site was marked with a silicone tube (Sogo Laboratory Glass Works, Kyoto, Japan). The silicone tube (3 mm in diameter and 2 mm in length) was cut longitudinally and was gently wrapped around the sciatic nerve 2 mm proximal to the injury site. The tube was fixed onto the epineurium with a 10–0 nylon suture.
λ⊥ ¼ ðλ2 þ λ3 Þ=2
ð2Þ
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi rffiffiffi 2 2 2 3 ðλ1 −MDÞ þ ðλ2 −MDÞ þ ðλ3 −MDÞ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi FA ¼ 2 λ1 2 þ λ2 2 þ λ3 2
ð3Þ
mean diffusivity ðMDÞ ¼ ðλ1 þ λ2 þ λ3 Þ=3
ð4Þ
2.3. Magnetic resonance imaging The rabbits were anesthetized with a mixture of 10% oxygen, air and isoflurane (5.0% for induction and 3.0% for maintenance). After obtaining the appropriate level of anesthesia, the rabbits were placed in the prone position. DTI is greatly affected by movements such as breathing, so to diminish these effects, the rabbits were tightly fixed with their legs stretched to avoid inclusion of the trunk in the field of view. Serial MRI procedures were performed at various time points, including before the crush injury (defined as week 0) and at 2, 4, 6 and 8 weeks post-injury. Images were obtained using a high magnetic field MRI unit designed for animal experiments (Varian MRI system 7.04 T, Agilent Technology, Palo Alto, California, USA) with a 210 mm gradient coil (Agilent Technology). The volume resonator coil with 140 mm in diameter was used. The MRI protocol included fast spin-echo proton density-weighted imaging (PDWI) and DTI. In this study, we used fast spin echo sequence with diffusion probing gradients for acquisition of DTI. The scanning parameters for DTI were as follows: repetition time, 3 seconds; effective echo time, 26 ms; echo train length, 4; echo space, 9.8 ms; field of view, 100 mm × 100 mm; pixel matrix, 128 × 128; slice thickness, 2 mm; numbers of average, 4; and b-value, 0 s/mm 2 and 700 s/mm 2. DELTA, 14 ms; delta, 6 ms; diffusion probing gradient orientations were applied along six directions: [Gx,Gy,Gz] = [1,1,0], [1,0,1], [0,1,1], [− 1,1,0], [0,−1,1] and [1,0,−1]. Total scan time for acquiring DTI was 44 minutes 54 seconds. Before running the DTI sequence, PDWI was obtained to confirm the sciatic nerve orientation. The imaging plane was prepared perpendicular to the sciatic nerve. Four imaging planes were set for detecting sciatic nerve (Fig. 1b, c, d, e). The thickness of each slice
2.5. Histological analyses To evaluate the histological changes in the injured sciatic nerves, we used light microscopy to examine cross-sections of the sciatic nerves at multiple time points including: before crush injury and at 2, 4, 6 and 8 weeks post-injury. We performed paraffin-embedded myelin sheath staining with osmium tetroxide [19] for the assessment of histological change after sciatic nerve injury. Rabbits were sacrificed with a pentobarbital overdose. The sciatic nerves were washed with phosphate-buffered saline (pH 7.4). Segments from the region 4 mm distal to the lesion site were collected. The nerves were fixed with 4% paraformaldehyde and 1% glutaraldehyde and were then post-fixed in 1% osmium tetroxide (OsO4) for 2 hours. After dehydration through graded ethanol solutions (10 minutes each at 50%, 70%, 80%, 90% and 95% concentrations, and 15 minutes at 100% concentration), the specimens were soaked in propylene oxide (Nissin EM, Tokyo, Japan). After infiltration in a 1:1 mixture of propylene oxide and Epon 812 (Nissin EM) for 1 hour then a 1:2 mixture for 1 hour, the samples were polymerized in 100% Epon 812 for 1 hour. The samples were then embedded in plastic capsules and were polymerized overnight at 37 °C and then for 2 days at 60 °C. Semi-thin transverse sections 2 μm in thickness were collected onto glass slides and were stained with Toluidine blue dye. Sciatic nerve samples obtained from the region 4 mm distal to the lesion site were examined using a light microscope. Images were obtained using a BX50 microscope (Olympus, Tokyo, Japan) and a DP21 color digital camera (Olympus). For quantitative analysis, we evaluated the histological changes in terms of the number of axons, as well as the area of the myelinated axons, to reflect changes in both the axons and myelin, respectively. Microscopic visual fields were randomly selected in each sample and were observed using a 100× oil immersion objective lens (Olympus). Three microscopic fields from each sample were randomly captured. The obtained images were
T. Yamasaki et al. / Magnetic Resonance Imaging 33 (2015) 95–101
97
Fig. 1. Representative images of the right thighs of the rabbits are shown. Right sciatic nerve was injured. Sagittal image of right thigh including silicone tube was shown (a). Four imaging plane was set for detecting sciatic nerve in PDWI (b, c, d, e). The ROI was manually placed on the sciatic nerve. A red circle on the PDWI slice 4(e) indicated ROI. ROIs on the FA image and images of the eigenvalues (λ1, λ2 and λ3) were set by superimposing PDWI images onto them (h). Images acquired with minimum b value (b = 0) (f) and maximum b value (b = 700) (g) were shown. Red arrow heads indicate silicone tube. Red arrows indicate sciatic nerve. D, dorsal; V, ventral; L, lateral; M, medial. Scale bar = 5 mm.
digitized in grayscale using an automatic threshold tool linked to morphometry software (Wayne Rasband, ImageJ, U.S. National Institutes of Health, Bethesda, Maryland, USA). The number of axons within a field (area: 1.40 × 10 −2 mm 2) were counted. For statistical analyses, the number of axons was calculated as the number within 1.0 × 10 −2 mm 2. The area of the myelinated axons was expressed as the area (%) comprised of axons and myelin divided by the total area of the nerve fibers.
2.6. Behavioral analyses Motor function assessment of the right hind paw was performed before the crush injury and at 2, 4, 6 and 8 weeks after injury using the toe-spreading reflex. The toe-spreading reflex is a reliable, repeatable and non-invasive method for assessing recovery of peroneal nerve function after injury [20–23]. The procedure was as follows: the rabbits were held by the loose skin of their backs then were suddenly lowered in the air without letting them contact a surface. When functional re-innervation had occurred, the rabbits reflexively spread their second, third and fourth toes. The movement of these toes was captured using a digital video camera (Panasonic, Tokyo, Japan) and then was graded following the toe-spreading index (TSI) described by E. Gutmann (Table 1) [20,21].
2.7. Statistical analyses The diffusion tensor parameters obtained from the injured side (right side) and non-injured side (left side, control) were compared using the t-test. The changes in diffusion tensor parameters on each injured and control side were tested using nonparametric one-way analysis of variance (ANOVA). Nonparametric one-way ANOVA was also used to evaluate changes in the number of axons and the ratio of the myelinated axon areas at each time point. The Tukey–Kramer test was used for the post-hoc analysis. All of the results are expressed as the means ± standard deviation of the mean (SD). To evaluate the correlations between DTI parameters and TSI, we used Spearman's rank-correlation coefficient. Pearson's correlation coefficients were used to evaluate the correlations between DTI
Table 1 The toe-spreading index for behavioral analyses. Grade
Symptoms
1 2 3 4
Just visible spreading of the 4th toe alone (also 2nd and 3rd) Slight spreading of all three toes Spreading of all three toes (less forceful than normal) Full spreading of all three toes (equal to normal)
98
T. Yamasaki et al. / Magnetic Resonance Imaging 33 (2015) 95–101
parameters and histological data. We performed all of the statistical analyses using SPSS software (IBM, Japan). 3. Results 3.1. Quantitative measurement of the DTI parameters The time courses of FA, λ⊥ and λ|| are shown in Fig. 2. The baseline value of FA was 0.39 ± 0.01. FA decreased significantly at 2 and 4 weeks post-injury, compared with the controls (P b 0.05). At 6 weeks post-injury, the FA of the injured side increased to 0.36 ± 0.01, and there was no significant difference compared with the controls (Fig. 2a). The baseline value of λ⊥ was 1.33 ± 0.08 μm 2/ms. The λ⊥ increased to 1.49 ± 0.05 μm 2/ms at 2 weeks post-injury and was significantly higher than the control. At 4 weeks post-injury, λ⊥ decreased to 1.34 ± 0.03 μm 2/ms, and there was no significant
difference compared with the controls (Fig. 2c). The baseline value of λ|| was 2.32 ± 0.06 μm 2/ms. λ|| showed no significant differences among any of the time points on the control side (Fig. 2b). Comparing DTI parameters at each time point on the crushed side, the FA at 2 and 4 weeks post-injury showed a significant decrease, compared with week 0 (P b 0.01). The FA at 4, 6 and 8 weeks significantly increased, compared with 2 weeks post-injury (P b 0.01). There was no significant difference in the FA values between week 8 and week 0. The λ⊥ was significantly increased at 2 weeks post-injury, compared with week 0. The λ⊥ value at 2 weeks post-injury was significantly greater than at 4, 6 and 8 weeks post-injury (P b 0.01, for all). λ|| showed no significant differences among the time points in each of the groups. 3.2. Histological changes Histological changes were evaluated at the same distance from the lesion site in each sample (Fig. 3). Almost all of the axons showed Wallerian degeneration, a breakdown of myelinated structure, and degeneration of the axons at 2 weeks post-injury. At 4 weeks postinjury, regenerating nerve fibers appeared. We performed quantitative morphometric evaluation (Fig. 4), and at 2 weeks post-injury, the number of axons and the ratio of the myelinated axon area had significantly decreased, compared with week 0. At 6 weeks postinjury, the number of axons and the ratio of the myelinated axon area increased to 95.2 ± 3.7 and 30.9 ± 2.8%, respectively; this change was significant compared with the values at 2 and 4 weeks post-injury (P b 0.001). We evaluated the correlation between DTI parameters and quantitative histological analysis. FA showed a strong, positive correlation with axon number, and the ratio of the myelinated axon area (Fig. 5a, b). λ⊥ was also correlated with axon number (Fig. 5e). However, there were no significant correlations between λ|| and quantitative histological analysis (Fig. 5c, d). 3.3. Functional recovery Motor function recovery was assessed by TSI at 0, 2, 4, 6 and 8 weeks post-injury. At 2 and 4 weeks post-injury, most of the rabbits had a decreased TSI score. Six weeks post-injury, the TSI scores gradually increased. At 8 weeks post-injury, 4 of 6 rabbits were graded as a 4 (Table 2). There was a positive correlation between FA and TSI (R = 0.71, P b 0.001), while λ⊥ and λ|| showed no correlation with TSI (Table 3). 4. Discussion This study investigated the validity of using DTI parameters to evaluate peripheral nerve degeneration and regeneration in vivo in a rabbit model. We performed DTI of the sciatic nerve using a 7.04 T MRI system. Our results revealed that DTI parameters, especially FA and λ⊥, changed significantly with the progression of the regeneration process after crush injury. Furthermore, these changes were correlated with histological and functional changes. 4.1. DTI analyses
Fig. 2. The time course of FA (a), λ|| (b), and λ⊥ (c) is shown in the sciatic nerve, 4 mm distal to the crush site. FA decreased at 2 weeks post-injury and subsequently increased at 4 to 8 weeks post-injury. There were significant differences between the crushed side and the control side at 2 and 4 weeks. λ|| decreased at 2 and 4 weeks post-injury but not to a significant degree. λ⊥ increased at 2 weeks post-injury and showed a significant difference from the control side. Each value is expressed as the mean ± standard deviation of the mean (SDM). * indicates a significant difference between the crushed side and control (P b 0.05).
Nerve fiber tracts serve as a barrier against the diffusion of water molecules across them, but they also act as a diffusion pathway along the tract via the myelin sheath [24,25]. FA is a quantitative index used to characterize the degree of diffusion anisotropy. In this study, the decreased FA observed at 2 and 4 weeks post-injury represents lowered anisotropic diffusion by the loss of nerve integrity. The increased FA observed from 4 weeks post-injury resulted from the regeneration of myelin sheaths, which induces anisotropic diffusion of water molecules. λ⊥, the index of water diffusivity perpendicular
T. Yamasaki et al. / Magnetic Resonance Imaging 33 (2015) 95–101
99
Fig. 3. Micrographs of transverse sections of the sciatic nerves were obtained 0 to 8 weeks after crush injury. The arrows indicate the degenerated axon and myelin. The arrowheads indicate regenerating axons. At 2 weeks post-injury, few normal axons were observed. At 4 weeks post-injury, regenerating axons appeared and subsequently increased. Scale bar = 10 μm.
to the nerve fiber, significantly increased at 2 weeks post-injury and subsequently decreased at 4 and 6 weeks post-injury. This change in λ⊥ suggests the decay of myelin structure. Although there were no significant changes in λ|| after injury, λ|| had a tendency to decrease at 2 and 4 weeks post-injury. λ|| is the index of water diffusivity along the nerve fiber. This change in λ|| and λ⊥ suggested that the change in water diffusivity perpendicular to nerve fiber largely affected to the change of FA. These DTI parameters showed a similar tendency after injury in a previous in vivo rat study [17]. Quite recently, changes in DTI parameters after nerve injury were also reported using a rabbit model [16]. The results of these studies support our current findings, and this consensus demonstrates the validity of DTI for diagnosing peripheral nerve degeneration and regeneration. In the previous rat study, DTI parameters did not return to preinjury levels within the 4 weeks of the observation period. In our study, however, by setting the observational period up to 8 weeks, we were able to confirm that the FA values returned to pre-injury levels. In the field of DTI parameter analysis in CNS, Brennan et al reported the change of DTI parameters after contusive spinal cord injury using rat model [26]. They showed that FA decreased rapidly after injury. λ|| mainly affected to decrease in FA on acute phase. Subsequently, gradual increase in λ⊥ resulted in continuous decrease in FA. In our study, increase in FA to pre-injury levels was observed after nerve injury. This result agrees to the fact that unlike CNS, peripheral nervous system has powerful regeneration capacity. Additionally Jirjis et al performed the correlation between the severity of spinal cord injury and the change of DTI parameter [27]. Further investigation about relationship between DTI parameter and severity or type of injury on peripheral nerve is needed.
4.2. Comparison of DTI with histological changes To demonstrate that the change in DTI parameter reflects the regeneration of the nerve, we compared DTI with histological parameters. The nerve regeneration phase was observed from 4 to 8 weeks post-injury. Both FA and λ⊥ showed correlations with the quantitative histological assessments. Although several studies have suggested that λ⊥ is sensitive to myelin injury [28–33], λ⊥ did not show a correlation with the ratio of the myelinated axon area. The process of peripheral nerve degeneration and regeneration is quite complex. Some studies have suggested that myelin might not be the only factor affecting λ⊥ [34–36]. The changes observed in DTI parameters reflected the microstructural changes in degenerating and regenerating nerve fibers. This result suggests that DTI parameters are potential markers of degeneration and regeneration of the peripheral nerve. Furthermore, FA showed an increase from 4 weeks post-injury, when nerve regeneration began to appear histologically. Thus, DTI parameters—especially FA—can detect early signs of peripheral nerve regeneration. 4.3. Comparison of DTI with motor function recovery Motor function recovery was assessed using the toe-spreading reflex. TSI recovery was observed beginning at 6 weeks post-injury, whereas FA was increased from 4 weeks post-injury. Thus, functional recovery was preceded by the recovery of DTI parameters. Similarly, functional and histological recovery were attained at about 4 weeks after crush injury in rats [37], which was preceded by the recovery of DTI parameters [17]. The time lag between the change in
Fig. 4. Quantitative analysis of histological changes after crush injury: the number of axons per 1 × 10−2 mm2 (a) and the ratio of the myelinated axon area are expressed as percentages (%) (b). Axon number significantly decreased at 2, 4, and 6 weeks post-injury, compared with week 0. Axon number subsequently increased significantly at 8 weeks post-injury, compared with 2, 4, and 6 weeks post-injury. There was no significant difference between week 0 and week 8 post-injury. The ratio of the myelinated axon area decreased at 2 weeks post-injury. At 4, 6, and 8 weeks post-injury, the ratio of the myelinated axon area increased, but the percentage was significantly lower than at week 0. *P b 0.001 vs. 0 weeks; **P b 0.001 vs. 8 weeks.
100
T. Yamasaki et al. / Magnetic Resonance Imaging 33 (2015) 95–101
Fig. 5. The DTI parameters at each time point (0, 2, 4, 6, and 8 weeks post-injury) were compared with the number of axons (a, c, e) and the ratio of the myelinated axon area (b, d, f) at the same time points. Scatterplots (a, e) show that the changes in axon numbers were strongly correlated with FA and λ⊥ (R = 0.98, P b 0.05 and −0.75, P b 0.05, respectively). A scatterplot (b) shows that the changes in the ratio of the myelinated axon area were strongly correlated with FA (R = 0.85, P b 0.05). λ|| had no significant correlation with the quantitative histological data (c, d).
DTI parameters and motor function recovery reflected the time needed for regeneration and re-innervation. Furthermore, the FA measured in the region distal to the lesion site showed a correlation with TSI. This result demonstrates that changes in DTI parameters measured at a point distal to the lesion, not at the lesion point, could be an indicator of motor function recovery after peripheral nerve injury. These findings suggest
Table 2 The time course of motor function recovery assessed by TSI. Grade
0 weeks
2 weeks
4 weeks
6 weeks
8 weeks
1 2 3 4
0 0 0 6
6 0 0 0
5 1 0 0
2 0 4 0
0 0 2 4
The numbers in each row are the numbers of rabbits graded according to TSI at each time point. At 2 weeks post-injury, all of the rabbits were graded as 1. At 4 weeks post-injury, only one rabbit was graded as 2, and the remainder was graded as 1. At 6 weeks post-injury, four rabbits were graded as 3. At 8 weeks post-injury, the motor function of four rabbits had recovered to a normal level.
that the FA value could be used to determine the prognosis of motor function after peripheral nerve injury. 4.4. Limitations In order to serially evaluate in vivo degeneration and regeneration, the rabbits had to be deeply anesthetized during the MRI procedure. To prevent rabbit mortality due to prolonged exposure to anesthesia, the scan time had to be shortened. Consequently, the resolution of the images was low in order to improve the SNR.
Table 3 Correlation coefficients between TSI and DTI parameters. Parameters
Correlation coefficient
P value
FA λ|| λ⊥
0.71 0.26 −0.31
P b 0.001 NS NS
FA showed a strong correlation with TSI.
T. Yamasaki et al. / Magnetic Resonance Imaging 33 (2015) 95–101
In this study, we assessed DTI parameters at sites 4 mm distal to lesion site. Indeed, assessment at further distal parts of the nerve would provide further information; there is difficulty for fine detection of the nerve at distal parts. The spatial variation of the DTI measures along the length of the nerve needs further investigation. 5. Conclusion We utilized a medium-sized animal model to demonstrate the effective evaluation of peripheral nerve injury by DTI for future clinical applications. The results of this study showed that DTI could be used as a non-invasive and quantitative method to evaluate peripheral nerve degeneration and regeneration. References [1] Noble J, Munro CA, Prasad VS, Midha R. Analysis of upper and lower extremity peripheral nerve injuries in a population of patients with multiple injuries. J Trauma 1998;45(1):116–22. [2] Kouyoumdjian JA. Peripheral nerve injuries: a retrospective survey of 456 cases. Muscle Nerve 2006;34(6):785–8. [3] Derr JJ, Micklesen PJ, Robinson LR. Predicting recovery after fibular nerve injury: which electrodiagnostic features are most useful? Am J Phys Med Rehabil 2009; 88(7):547–53. [4] Marciniak C, Armon C, Wilson J, Miller R. Practice parameter: utility of electrodiagnostic techniques in evaluating patients with suspected peroneal neuropathy: an evidence-based review. Muscle Nerve 2005;31(4):520–7. [5] Bendszus M, Wessig C, Solymosi L, Reiners K, Koltzenburg M. MRI of peripheral nerve degeneration and regeneration: correlation with electrophysiology and histology. Exp Neurol 2004;188(1):171–7. [6] Keen NN, Chin CT, Engstrom JW, Saloner D, Steinbach LS. Diagnosing ulnar neuropathy at the elbow using magnetic resonance neurography. Skeletal Radiol 2012;41(4):401–7. [7] Filler AG, Howe FA, Hayes CE, Kliot M, Winn HR, Bell BA, et al. Magnetic resonance neurography. Lancet 1993;341(8846):659–61. [8] Filler AG, Kliot M, Howe FA, Hayes CE, Saunders DE, Goodkin R, et al. Application of magnetic resonance neurography in the evaluation of patients with peripheral nerve pathology. J Neurosurg 1996;85(2):299–309. [9] Stanisz GJ, Midha R, Munro CA, Henkelman RM. MR properties of rat sciatic nerve following trauma. Magn Reson Med 2001;45(3):415–20. [10] Basser PJ, Pierpaoli C. Microstructural and physiological features of tissues elucidated by quantitative-diffusion-tensor MRI. J Magn Reson B 1996;111(3): 209–19. [11] Mori S, Zhang J. Principles of diffusion tensor imaging and its applications to basic neuroscience research. Neuron 2006;51(5):527–39. [12] Le Bihan D, Mangin JF, Poupon C, Clark CA, Pappata S, Molko N, et al. Diffusion tensor imaging: concepts and applications. J Magn Reson Imaging 2001;13(4): 534–46. [13] Kim JH, Budde MD, Liang HF, Klein RS, Russell JH, Cross AH, et al. Detecting axon damage in spinal cord from a mouse model of multiple sclerosis. Neurobiol Dis 2006;21(3):626–32. [14] Lehmann HC, Zhang J, Mori S, Sheikh KA. Diffusion tensor imaging to assess axonal regeneration in peripheral nerves. Exp Neurol 2010;223(1):238–44. [15] Takagi T, Nakamura M, Yamada M, Hikishima K, Momoshima S, Fujiyoshi K, et al. Visualization of peripheral nerve degeneration and regeneration: monitoring with diffusion tensor tractography. Neuroimage 2009;44(3):884–92.
101
[16] Li X, Chen J, Hong G, Sun C, Wu X, Peng MJ, et al. In vivo DTI longitudinal measurements of acute sciatic nerve traction injury and the association with pathological and functional changes. Eur J Radiol 2013;82(11):e707–14. [17] Morisaki S, Kawai Y, Umeda M, Nishi M, Oda R, Fujiwara H, et al. In vivo assessment of peripheral nerve regeneration by diffusion tensor imaging. J Magn Reson Imaging 2011;33(3):535–42. [18] Schmitz HC, Beer GM. Muscle-sparing approach to the peroneal nerve of the rabbit. Lab Anim 2001;35(4):334–9. [19] Di Scipio F, Raimondo S, Tos P, Geuna S. A simple protocol for paraffin-embedded myelin sheath staining with osmium tetroxide for light microscope observation. Microsc Res Tech 2008;71(7):497–502. [20] Gutmann E. Factors affecting recovery of motor function after nerve lesions. J Neurol Psychiatry 1942;5(3–4):81–95. [21] Gutmann E, Guttmann L, Medawar PB, Young JZ. The rate of regeneration of nerve. J Exp Biol 1942;19:14–44. [22] Schmitz HC, Beer GM. The toe-spreading reflex of the rabbit revisited–functional evaluation of complete peroneal nerve lesions. Lab Anim (NY) 2001;35(4): 340–5. [23] Beer GM, Seifert B, Schneller M, Iscru G, Schmitz HC. Relevance of muscle weight as a functional parameter for the regeneration of the peroneal nerve in rabbits. J Reconstr Microsurg 2008;24(1):11–9. [24] Nilsson M, van Westen D, Stahlberg F, Sundgren PC, Latt J. The role of tissue microstructure and water exchange in biophysical modelling of diffusion in white matter. MAGMA 2013:1–26. [25] Beaulieu C. The basis of anisotropic water diffusion in the nervous system–a technical review. NMR Biomed 2002;15(7–8):435–55. [26] Brennan FH, Cowin GJ, Kurniawan ND, Ruitenberg MJ. Longitudinal assessment of white matter pathology in the injured mouse spinal cord through ultra-high field (16.4 T) in vivo diffusion tensor imaging. Neuroimage 2013;82:574–85. [27] Jirjis MB, Kurpad SN, Schmit BD. Ex vivo diffusion tensor imaging of spinal cord injury in rats of varying degrees of severity. J Neurotrauma 2013;30(18): 1577–86. [28] Klawiter EC, Schmidt RE, Trinkaus K, Liang HF, Budde MD, Naismith RT, et al. Radial diffusivity predicts demyelination in ex vivo multiple sclerosis spinal cords. Neuroimage 2011;55(4):1454–60. [29] Song SK, Sun SW, Ju WK, Lin SJ, Cross AH, Neufeld AH. Diffusion tensor imaging detects and differentiates axon and myelin degeneration in mouse optic nerve after retinal ischemia. Neuroimage 2003;20(3):1714–22. [30] Song SK, Sun SW, Ramsbottom MJ, Chang C, Russell J, Cross AH. Dysmyelination revealed through MRI as increased radial (but unchanged axial) diffusion of water. Neuroimage 2002;17(3):1429–36. [31] Sun SW, Liang HF, Le TQ, Armstrong RC, Cross AH, Song SK. Differential sensitivity of in vivo and ex vivo diffusion tensor imaging to evolving optic nerve injury in mice with retinal ischemia. Neuroimage 2006;32(3):1195–204. [32] Budde MD, Xie M, Cross AH, Song SK. Axial diffusivity is the primary correlate of axonal injury in the experimental autoimmune encephalomyelitis spinal cord: a quantitative pixelwise analysis. J Neurosci 2009;29(9):2805–13. [33] Nair G, Tanahashi Y, Low HP, Billings-Gagliardi S, Schwartz WJ, Duong TQ. Myelination and long diffusion times alter diffusion-tensor-imaging contrast in myelin-deficient shiverer mice. Neuroimage 2005;28(1):165–74. [34] Zollinger LV, Kim TH, Hill K, Jeong EK, Rose JW. Using diffusion tensor imaging and immunofluorescent assay to evaluate the pathology of multiple sclerosis. J Magn Reson Imaging 2011;33(3):557–64. [35] Drobyshevsky A, Derrick M, Wyrwicz AM, Ji X, Englof I, Ullman LM, et al. White matter injury correlates with hypertonia in an animal model of cerebral palsy. J Cereb Blood Flow Metab 2007;27(2):270–81. [36] Ludeman NA, Berman JI, Wu YW, Jeremy RJ, Kornak J, Bartha AI, et al. Diffusion tensor imaging of the pyramidal tracts in infants with motor dysfunction. Neurology 2008;71(21):1676–82. [37] Raducan A, Mirica S, Duicu O, Raducan S, Muntean D, Fira-Mladinescu O, et al. Morphological and functional aspects of sciatic nerve regeneration after crush injury. Rom J Morphol Embryol 2013;54(3 Suppl.):735–9.
Contents lists available at ScienceDirect
Magnetic Resonance Imaging journal homepage: www.mrijournal.com
In vivo evaluation of rabbit sciatic nerve regeneration with diffusion tensor imaging (DTI): correlations with histology and behavior Tetsuro Yamasaki a, 1, Hiroyoshi Fujiwara a, 1, Ryo Oda a,⁎, Yasuo Mikami a, 1, Takumi Ikeda a, 1, Masateru Nagae a, 1, Toshiharu Shirai a, 1, Shinsuke Morisaki a, 1, Kazuya Ikoma a, 1, Miwako Masugi-Tokita b, 2, Kei Yamada c, 3, Mitsuhiro Kawata b, 2, Toshikazu Kubo a, 1 a b c
Department of Orthopaedics, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kawaramachi-Hirokoji, Kamigyo-ku, Kyoto, 602-8566, Japan Department of Anatomy and Neurobiology, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kawaramachi-Hirokoji, Kamigyo-ku, Kyoto, 602-8566, Japan Department of Radiology, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kawaramachi-Hirokoji, Kamigyo-ku, Kyoto, 602-8566, Japan
a r t i c l e
i n f o
Article history: Received 8 March 2014 Revised 23 August 2014 Accepted 22 September 2014 Keywords: Peripheral nerve Regeneration Diffusion tenor imaging Rabbit Fractional anisotropy Toe-spreading reflex
a b s t r a c t Diffusion tensor imaging (DTI) is widely used in the study of the central nervous system. DTI represents a potential diagnostic tool for the peripheral nerve. However, more detailed information is needed for application of DTI in the clinical setting. In this study, peripheral degeneration and regeneration were evaluated using DTIbased analyses in a rabbit model. The changes in DTI parameters were compared to histological and functional changes after nerve injury. We used a high magnetic field (7.04 T) MRI system. Japanese white male rabbits were used as the model of sciatic nerve crush injury. MR images were obtained before injury and at 2, 4, 6 and 8 weeks post-injury. The DTI parameters of fractional anisotropy (FA), axial diffusivity (λ||), and radial diffusivity (λ⊥) were calculated. Our results showed decreased FA and increased λ⊥ during the degenerative phase after sciatic nerve injury. In contrast, increased FA and decreased λ⊥ were observed during the regenerative phase. FA changes were correlated with axon number and with motor function recovery, assessed with the toe-spreading index. This study clearly demonstrates the validity of applying DTI parameters to the in vivo evaluation of peripheral nerve regeneration. Furthermore, results suggest that DTI can be a potent tool for predicting the extent of functional recovery after peripheral nerve injury. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Peripheral nerve injures are common in the clinical field. They result from various circumstances such as trauma, inflammation, and peripheral neuropathies [1,2]. Accurate assessment of the location and severity of nerve injury is essential for selection of treatment protocols. Physical examination and electrophysiological experiments play a key role in assessing peripheral nerve injury [3,4]. However, the reliability of assessment using these methods is highly dependent on the clinician's experience. Additionally, electrophysiological examination, requiring insertion of needles
⁎ Corresponding author. Tel.: +81 75 251 5549; fax: +81 75 251 5841. E-mail address: [email protected] (R. Oda). 1 Tel.: +81 75 251 5549; fax: +81 75 251 5841. 2 Tel.: +81 75 251 5301; fax: +81 75 251 5306. 3 Tel.: + 81 75 251 5620; fax: +81 75 251 5840.
http://dx.doi.org/10.1016/j.mri.2014.09.005 0730-725X/© 2014 Elsevier Inc. All rights reserved.
into the muscles, is an invasive procedure. Therefore, a new method that can noninvasively and objectively evaluate peripheral nerve injury needs to be established. Magnetic resonance imaging (MRI) is an effective method for detecting various structural changes, such as edema and necrosis, within soft tissues. A number of studies have also indicated that MRI can be utilized for the assessment of peripheral nerves [5–9]. In the nervous system, diffusion tensor imaging (DTI) has been accepted as a suitable tool for detecting nerve fiber orientation, and can provide quantitative evaluation based on the diffusional movement of water molecules through tissue [10]. DTI can accurately depict nerve fibers and microstructure within the central nervous system (CNS) [11–13]. DTI is expected to be utilized as a new diagnostic tool for peripheral nerve injury, as it can evaluate nerve degeneration and regeneration noninvasively, objectively, and quantitatively. Several animal studies have been conducted to validate DTI for use in the peripheral nervous system (PNS) [14–17]. However, it remains unclear whether the evaluation of peripheral nerve injury by DTI can be applied to humans. More detailed foundational data must be collected for application of DTI in the clinical setting. In this study, we investigated the validity of
96
T. Yamasaki et al. / Magnetic Resonance Imaging 33 (2015) 95–101
DTI for evaluating peripheral nerve degeneration and regeneration in a rabbit model by comparing DTI parameters with histological and motor function data after sciatic nerve injury. 2. Materials and methods 2.1. Animal model
was 2 mm. The most proximal image plane includes silicone tube surrounding the sciatic nerve (Fig. 1b). The most distal slice included the sciatic nerve 4 mm distal to the crush lesion (Fig. 1e). The sciatic nerve was identified by tracing these serial images. Imaging plane for DTI was set at the same sites as PDWI (Fig. 1f, g, h). DTI parameter was assessed in the most distal slice to exclude the effect of acute inflammatory reaction at crush lesion. 2.4. DTI data analyses
In all of the studies, male Japanese white rabbits (Shimizu Laboratory Supplies, Kyoto, Japan), weighing 3.0–3.5 kg, were used. The rabbits were maintained on a 12-hour light/dark schedule with free access to food and water. All of the surgical and experimental procedures were approved by the committee for animal research at our institution and were performed in accordance with the guidelines of the National Institutes of Health regarding animal care. Six rabbits underwent serial MRI scans and behavioral evaluation longitudinally before and after sciatic nerve injury. Histological evaluation was performed on 3 rabbits in each time points (before injury and 2, 4, 6, 8 weeks after injury) separately. Thus the total of rabbits used for histological analysis was 15.
The six independent elements of the diffusion tensor were calculated from each diffusion-weighted image. The resulting tensor element maps were used to derive the eigenvalues of the tensor (λ1, λ2 and λ3) by matrix diagonalization. The regions of interest (ROIs) were manually placed on the sciatic nerve as identified on the PDWI (Fig. 1e). The mean size of ROIs was 1.3 ± 0.2 mm 2. The signal-tonoise ratios measured in the images with minimum b value (b = 0) were greater than 19 for all experiments. Fractional anisotropy (FA), axial diffusivity (λ||), and radial diffusivity (λ⊥) were calculated using the following equations:
2.2. Surgical procedures
λjj ¼ λ1
ð1Þ
All of the rabbits were deeply anesthetized with a mixture of 10% oxygen, air and 2.0% isoflurane (Abbott, Osaka, Japan). The right sciatic nerve was exposed [18] and ligated at the proximal thigh level with a 3–0 silk suture to produce a crush injury. Ligation was performed with a force of 0.7 N applied for 5 minutes. To identify the injury appropriately on MR images, the lesion site was marked with a silicone tube (Sogo Laboratory Glass Works, Kyoto, Japan). The silicone tube (3 mm in diameter and 2 mm in length) was cut longitudinally and was gently wrapped around the sciatic nerve 2 mm proximal to the injury site. The tube was fixed onto the epineurium with a 10–0 nylon suture.
λ⊥ ¼ ðλ2 þ λ3 Þ=2
ð2Þ
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi rffiffiffi 2 2 2 3 ðλ1 −MDÞ þ ðλ2 −MDÞ þ ðλ3 −MDÞ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi FA ¼ 2 λ1 2 þ λ2 2 þ λ3 2
ð3Þ
mean diffusivity ðMDÞ ¼ ðλ1 þ λ2 þ λ3 Þ=3
ð4Þ
2.3. Magnetic resonance imaging The rabbits were anesthetized with a mixture of 10% oxygen, air and isoflurane (5.0% for induction and 3.0% for maintenance). After obtaining the appropriate level of anesthesia, the rabbits were placed in the prone position. DTI is greatly affected by movements such as breathing, so to diminish these effects, the rabbits were tightly fixed with their legs stretched to avoid inclusion of the trunk in the field of view. Serial MRI procedures were performed at various time points, including before the crush injury (defined as week 0) and at 2, 4, 6 and 8 weeks post-injury. Images were obtained using a high magnetic field MRI unit designed for animal experiments (Varian MRI system 7.04 T, Agilent Technology, Palo Alto, California, USA) with a 210 mm gradient coil (Agilent Technology). The volume resonator coil with 140 mm in diameter was used. The MRI protocol included fast spin-echo proton density-weighted imaging (PDWI) and DTI. In this study, we used fast spin echo sequence with diffusion probing gradients for acquisition of DTI. The scanning parameters for DTI were as follows: repetition time, 3 seconds; effective echo time, 26 ms; echo train length, 4; echo space, 9.8 ms; field of view, 100 mm × 100 mm; pixel matrix, 128 × 128; slice thickness, 2 mm; numbers of average, 4; and b-value, 0 s/mm 2 and 700 s/mm 2. DELTA, 14 ms; delta, 6 ms; diffusion probing gradient orientations were applied along six directions: [Gx,Gy,Gz] = [1,1,0], [1,0,1], [0,1,1], [− 1,1,0], [0,−1,1] and [1,0,−1]. Total scan time for acquiring DTI was 44 minutes 54 seconds. Before running the DTI sequence, PDWI was obtained to confirm the sciatic nerve orientation. The imaging plane was prepared perpendicular to the sciatic nerve. Four imaging planes were set for detecting sciatic nerve (Fig. 1b, c, d, e). The thickness of each slice
2.5. Histological analyses To evaluate the histological changes in the injured sciatic nerves, we used light microscopy to examine cross-sections of the sciatic nerves at multiple time points including: before crush injury and at 2, 4, 6 and 8 weeks post-injury. We performed paraffin-embedded myelin sheath staining with osmium tetroxide [19] for the assessment of histological change after sciatic nerve injury. Rabbits were sacrificed with a pentobarbital overdose. The sciatic nerves were washed with phosphate-buffered saline (pH 7.4). Segments from the region 4 mm distal to the lesion site were collected. The nerves were fixed with 4% paraformaldehyde and 1% glutaraldehyde and were then post-fixed in 1% osmium tetroxide (OsO4) for 2 hours. After dehydration through graded ethanol solutions (10 minutes each at 50%, 70%, 80%, 90% and 95% concentrations, and 15 minutes at 100% concentration), the specimens were soaked in propylene oxide (Nissin EM, Tokyo, Japan). After infiltration in a 1:1 mixture of propylene oxide and Epon 812 (Nissin EM) for 1 hour then a 1:2 mixture for 1 hour, the samples were polymerized in 100% Epon 812 for 1 hour. The samples were then embedded in plastic capsules and were polymerized overnight at 37 °C and then for 2 days at 60 °C. Semi-thin transverse sections 2 μm in thickness were collected onto glass slides and were stained with Toluidine blue dye. Sciatic nerve samples obtained from the region 4 mm distal to the lesion site were examined using a light microscope. Images were obtained using a BX50 microscope (Olympus, Tokyo, Japan) and a DP21 color digital camera (Olympus). For quantitative analysis, we evaluated the histological changes in terms of the number of axons, as well as the area of the myelinated axons, to reflect changes in both the axons and myelin, respectively. Microscopic visual fields were randomly selected in each sample and were observed using a 100× oil immersion objective lens (Olympus). Three microscopic fields from each sample were randomly captured. The obtained images were
T. Yamasaki et al. / Magnetic Resonance Imaging 33 (2015) 95–101
97
Fig. 1. Representative images of the right thighs of the rabbits are shown. Right sciatic nerve was injured. Sagittal image of right thigh including silicone tube was shown (a). Four imaging plane was set for detecting sciatic nerve in PDWI (b, c, d, e). The ROI was manually placed on the sciatic nerve. A red circle on the PDWI slice 4(e) indicated ROI. ROIs on the FA image and images of the eigenvalues (λ1, λ2 and λ3) were set by superimposing PDWI images onto them (h). Images acquired with minimum b value (b = 0) (f) and maximum b value (b = 700) (g) were shown. Red arrow heads indicate silicone tube. Red arrows indicate sciatic nerve. D, dorsal; V, ventral; L, lateral; M, medial. Scale bar = 5 mm.
digitized in grayscale using an automatic threshold tool linked to morphometry software (Wayne Rasband, ImageJ, U.S. National Institutes of Health, Bethesda, Maryland, USA). The number of axons within a field (area: 1.40 × 10 −2 mm 2) were counted. For statistical analyses, the number of axons was calculated as the number within 1.0 × 10 −2 mm 2. The area of the myelinated axons was expressed as the area (%) comprised of axons and myelin divided by the total area of the nerve fibers.
2.6. Behavioral analyses Motor function assessment of the right hind paw was performed before the crush injury and at 2, 4, 6 and 8 weeks after injury using the toe-spreading reflex. The toe-spreading reflex is a reliable, repeatable and non-invasive method for assessing recovery of peroneal nerve function after injury [20–23]. The procedure was as follows: the rabbits were held by the loose skin of their backs then were suddenly lowered in the air without letting them contact a surface. When functional re-innervation had occurred, the rabbits reflexively spread their second, third and fourth toes. The movement of these toes was captured using a digital video camera (Panasonic, Tokyo, Japan) and then was graded following the toe-spreading index (TSI) described by E. Gutmann (Table 1) [20,21].
2.7. Statistical analyses The diffusion tensor parameters obtained from the injured side (right side) and non-injured side (left side, control) were compared using the t-test. The changes in diffusion tensor parameters on each injured and control side were tested using nonparametric one-way analysis of variance (ANOVA). Nonparametric one-way ANOVA was also used to evaluate changes in the number of axons and the ratio of the myelinated axon areas at each time point. The Tukey–Kramer test was used for the post-hoc analysis. All of the results are expressed as the means ± standard deviation of the mean (SD). To evaluate the correlations between DTI parameters and TSI, we used Spearman's rank-correlation coefficient. Pearson's correlation coefficients were used to evaluate the correlations between DTI
Table 1 The toe-spreading index for behavioral analyses. Grade
Symptoms
1 2 3 4
Just visible spreading of the 4th toe alone (also 2nd and 3rd) Slight spreading of all three toes Spreading of all three toes (less forceful than normal) Full spreading of all three toes (equal to normal)
98
T. Yamasaki et al. / Magnetic Resonance Imaging 33 (2015) 95–101
parameters and histological data. We performed all of the statistical analyses using SPSS software (IBM, Japan). 3. Results 3.1. Quantitative measurement of the DTI parameters The time courses of FA, λ⊥ and λ|| are shown in Fig. 2. The baseline value of FA was 0.39 ± 0.01. FA decreased significantly at 2 and 4 weeks post-injury, compared with the controls (P b 0.05). At 6 weeks post-injury, the FA of the injured side increased to 0.36 ± 0.01, and there was no significant difference compared with the controls (Fig. 2a). The baseline value of λ⊥ was 1.33 ± 0.08 μm 2/ms. The λ⊥ increased to 1.49 ± 0.05 μm 2/ms at 2 weeks post-injury and was significantly higher than the control. At 4 weeks post-injury, λ⊥ decreased to 1.34 ± 0.03 μm 2/ms, and there was no significant
difference compared with the controls (Fig. 2c). The baseline value of λ|| was 2.32 ± 0.06 μm 2/ms. λ|| showed no significant differences among any of the time points on the control side (Fig. 2b). Comparing DTI parameters at each time point on the crushed side, the FA at 2 and 4 weeks post-injury showed a significant decrease, compared with week 0 (P b 0.01). The FA at 4, 6 and 8 weeks significantly increased, compared with 2 weeks post-injury (P b 0.01). There was no significant difference in the FA values between week 8 and week 0. The λ⊥ was significantly increased at 2 weeks post-injury, compared with week 0. The λ⊥ value at 2 weeks post-injury was significantly greater than at 4, 6 and 8 weeks post-injury (P b 0.01, for all). λ|| showed no significant differences among the time points in each of the groups. 3.2. Histological changes Histological changes were evaluated at the same distance from the lesion site in each sample (Fig. 3). Almost all of the axons showed Wallerian degeneration, a breakdown of myelinated structure, and degeneration of the axons at 2 weeks post-injury. At 4 weeks postinjury, regenerating nerve fibers appeared. We performed quantitative morphometric evaluation (Fig. 4), and at 2 weeks post-injury, the number of axons and the ratio of the myelinated axon area had significantly decreased, compared with week 0. At 6 weeks postinjury, the number of axons and the ratio of the myelinated axon area increased to 95.2 ± 3.7 and 30.9 ± 2.8%, respectively; this change was significant compared with the values at 2 and 4 weeks post-injury (P b 0.001). We evaluated the correlation between DTI parameters and quantitative histological analysis. FA showed a strong, positive correlation with axon number, and the ratio of the myelinated axon area (Fig. 5a, b). λ⊥ was also correlated with axon number (Fig. 5e). However, there were no significant correlations between λ|| and quantitative histological analysis (Fig. 5c, d). 3.3. Functional recovery Motor function recovery was assessed by TSI at 0, 2, 4, 6 and 8 weeks post-injury. At 2 and 4 weeks post-injury, most of the rabbits had a decreased TSI score. Six weeks post-injury, the TSI scores gradually increased. At 8 weeks post-injury, 4 of 6 rabbits were graded as a 4 (Table 2). There was a positive correlation between FA and TSI (R = 0.71, P b 0.001), while λ⊥ and λ|| showed no correlation with TSI (Table 3). 4. Discussion This study investigated the validity of using DTI parameters to evaluate peripheral nerve degeneration and regeneration in vivo in a rabbit model. We performed DTI of the sciatic nerve using a 7.04 T MRI system. Our results revealed that DTI parameters, especially FA and λ⊥, changed significantly with the progression of the regeneration process after crush injury. Furthermore, these changes were correlated with histological and functional changes. 4.1. DTI analyses
Fig. 2. The time course of FA (a), λ|| (b), and λ⊥ (c) is shown in the sciatic nerve, 4 mm distal to the crush site. FA decreased at 2 weeks post-injury and subsequently increased at 4 to 8 weeks post-injury. There were significant differences between the crushed side and the control side at 2 and 4 weeks. λ|| decreased at 2 and 4 weeks post-injury but not to a significant degree. λ⊥ increased at 2 weeks post-injury and showed a significant difference from the control side. Each value is expressed as the mean ± standard deviation of the mean (SDM). * indicates a significant difference between the crushed side and control (P b 0.05).
Nerve fiber tracts serve as a barrier against the diffusion of water molecules across them, but they also act as a diffusion pathway along the tract via the myelin sheath [24,25]. FA is a quantitative index used to characterize the degree of diffusion anisotropy. In this study, the decreased FA observed at 2 and 4 weeks post-injury represents lowered anisotropic diffusion by the loss of nerve integrity. The increased FA observed from 4 weeks post-injury resulted from the regeneration of myelin sheaths, which induces anisotropic diffusion of water molecules. λ⊥, the index of water diffusivity perpendicular
T. Yamasaki et al. / Magnetic Resonance Imaging 33 (2015) 95–101
99
Fig. 3. Micrographs of transverse sections of the sciatic nerves were obtained 0 to 8 weeks after crush injury. The arrows indicate the degenerated axon and myelin. The arrowheads indicate regenerating axons. At 2 weeks post-injury, few normal axons were observed. At 4 weeks post-injury, regenerating axons appeared and subsequently increased. Scale bar = 10 μm.
to the nerve fiber, significantly increased at 2 weeks post-injury and subsequently decreased at 4 and 6 weeks post-injury. This change in λ⊥ suggests the decay of myelin structure. Although there were no significant changes in λ|| after injury, λ|| had a tendency to decrease at 2 and 4 weeks post-injury. λ|| is the index of water diffusivity along the nerve fiber. This change in λ|| and λ⊥ suggested that the change in water diffusivity perpendicular to nerve fiber largely affected to the change of FA. These DTI parameters showed a similar tendency after injury in a previous in vivo rat study [17]. Quite recently, changes in DTI parameters after nerve injury were also reported using a rabbit model [16]. The results of these studies support our current findings, and this consensus demonstrates the validity of DTI for diagnosing peripheral nerve degeneration and regeneration. In the previous rat study, DTI parameters did not return to preinjury levels within the 4 weeks of the observation period. In our study, however, by setting the observational period up to 8 weeks, we were able to confirm that the FA values returned to pre-injury levels. In the field of DTI parameter analysis in CNS, Brennan et al reported the change of DTI parameters after contusive spinal cord injury using rat model [26]. They showed that FA decreased rapidly after injury. λ|| mainly affected to decrease in FA on acute phase. Subsequently, gradual increase in λ⊥ resulted in continuous decrease in FA. In our study, increase in FA to pre-injury levels was observed after nerve injury. This result agrees to the fact that unlike CNS, peripheral nervous system has powerful regeneration capacity. Additionally Jirjis et al performed the correlation between the severity of spinal cord injury and the change of DTI parameter [27]. Further investigation about relationship between DTI parameter and severity or type of injury on peripheral nerve is needed.
4.2. Comparison of DTI with histological changes To demonstrate that the change in DTI parameter reflects the regeneration of the nerve, we compared DTI with histological parameters. The nerve regeneration phase was observed from 4 to 8 weeks post-injury. Both FA and λ⊥ showed correlations with the quantitative histological assessments. Although several studies have suggested that λ⊥ is sensitive to myelin injury [28–33], λ⊥ did not show a correlation with the ratio of the myelinated axon area. The process of peripheral nerve degeneration and regeneration is quite complex. Some studies have suggested that myelin might not be the only factor affecting λ⊥ [34–36]. The changes observed in DTI parameters reflected the microstructural changes in degenerating and regenerating nerve fibers. This result suggests that DTI parameters are potential markers of degeneration and regeneration of the peripheral nerve. Furthermore, FA showed an increase from 4 weeks post-injury, when nerve regeneration began to appear histologically. Thus, DTI parameters—especially FA—can detect early signs of peripheral nerve regeneration. 4.3. Comparison of DTI with motor function recovery Motor function recovery was assessed using the toe-spreading reflex. TSI recovery was observed beginning at 6 weeks post-injury, whereas FA was increased from 4 weeks post-injury. Thus, functional recovery was preceded by the recovery of DTI parameters. Similarly, functional and histological recovery were attained at about 4 weeks after crush injury in rats [37], which was preceded by the recovery of DTI parameters [17]. The time lag between the change in
Fig. 4. Quantitative analysis of histological changes after crush injury: the number of axons per 1 × 10−2 mm2 (a) and the ratio of the myelinated axon area are expressed as percentages (%) (b). Axon number significantly decreased at 2, 4, and 6 weeks post-injury, compared with week 0. Axon number subsequently increased significantly at 8 weeks post-injury, compared with 2, 4, and 6 weeks post-injury. There was no significant difference between week 0 and week 8 post-injury. The ratio of the myelinated axon area decreased at 2 weeks post-injury. At 4, 6, and 8 weeks post-injury, the ratio of the myelinated axon area increased, but the percentage was significantly lower than at week 0. *P b 0.001 vs. 0 weeks; **P b 0.001 vs. 8 weeks.
100
T. Yamasaki et al. / Magnetic Resonance Imaging 33 (2015) 95–101
Fig. 5. The DTI parameters at each time point (0, 2, 4, 6, and 8 weeks post-injury) were compared with the number of axons (a, c, e) and the ratio of the myelinated axon area (b, d, f) at the same time points. Scatterplots (a, e) show that the changes in axon numbers were strongly correlated with FA and λ⊥ (R = 0.98, P b 0.05 and −0.75, P b 0.05, respectively). A scatterplot (b) shows that the changes in the ratio of the myelinated axon area were strongly correlated with FA (R = 0.85, P b 0.05). λ|| had no significant correlation with the quantitative histological data (c, d).
DTI parameters and motor function recovery reflected the time needed for regeneration and re-innervation. Furthermore, the FA measured in the region distal to the lesion site showed a correlation with TSI. This result demonstrates that changes in DTI parameters measured at a point distal to the lesion, not at the lesion point, could be an indicator of motor function recovery after peripheral nerve injury. These findings suggest
Table 2 The time course of motor function recovery assessed by TSI. Grade
0 weeks
2 weeks
4 weeks
6 weeks
8 weeks
1 2 3 4
0 0 0 6
6 0 0 0
5 1 0 0
2 0 4 0
0 0 2 4
The numbers in each row are the numbers of rabbits graded according to TSI at each time point. At 2 weeks post-injury, all of the rabbits were graded as 1. At 4 weeks post-injury, only one rabbit was graded as 2, and the remainder was graded as 1. At 6 weeks post-injury, four rabbits were graded as 3. At 8 weeks post-injury, the motor function of four rabbits had recovered to a normal level.
that the FA value could be used to determine the prognosis of motor function after peripheral nerve injury. 4.4. Limitations In order to serially evaluate in vivo degeneration and regeneration, the rabbits had to be deeply anesthetized during the MRI procedure. To prevent rabbit mortality due to prolonged exposure to anesthesia, the scan time had to be shortened. Consequently, the resolution of the images was low in order to improve the SNR.
Table 3 Correlation coefficients between TSI and DTI parameters. Parameters
Correlation coefficient
P value
FA λ|| λ⊥
0.71 0.26 −0.31
P b 0.001 NS NS
FA showed a strong correlation with TSI.
T. Yamasaki et al. / Magnetic Resonance Imaging 33 (2015) 95–101
In this study, we assessed DTI parameters at sites 4 mm distal to lesion site. Indeed, assessment at further distal parts of the nerve would provide further information; there is difficulty for fine detection of the nerve at distal parts. The spatial variation of the DTI measures along the length of the nerve needs further investigation. 5. Conclusion We utilized a medium-sized animal model to demonstrate the effective evaluation of peripheral nerve injury by DTI for future clinical applications. The results of this study showed that DTI could be used as a non-invasive and quantitative method to evaluate peripheral nerve degeneration and regeneration. References [1] Noble J, Munro CA, Prasad VS, Midha R. Analysis of upper and lower extremity peripheral nerve injuries in a population of patients with multiple injuries. J Trauma 1998;45(1):116–22. [2] Kouyoumdjian JA. Peripheral nerve injuries: a retrospective survey of 456 cases. Muscle Nerve 2006;34(6):785–8. [3] Derr JJ, Micklesen PJ, Robinson LR. Predicting recovery after fibular nerve injury: which electrodiagnostic features are most useful? Am J Phys Med Rehabil 2009; 88(7):547–53. [4] Marciniak C, Armon C, Wilson J, Miller R. Practice parameter: utility of electrodiagnostic techniques in evaluating patients with suspected peroneal neuropathy: an evidence-based review. Muscle Nerve 2005;31(4):520–7. [5] Bendszus M, Wessig C, Solymosi L, Reiners K, Koltzenburg M. MRI of peripheral nerve degeneration and regeneration: correlation with electrophysiology and histology. Exp Neurol 2004;188(1):171–7. [6] Keen NN, Chin CT, Engstrom JW, Saloner D, Steinbach LS. Diagnosing ulnar neuropathy at the elbow using magnetic resonance neurography. Skeletal Radiol 2012;41(4):401–7. [7] Filler AG, Howe FA, Hayes CE, Kliot M, Winn HR, Bell BA, et al. Magnetic resonance neurography. Lancet 1993;341(8846):659–61. [8] Filler AG, Kliot M, Howe FA, Hayes CE, Saunders DE, Goodkin R, et al. Application of magnetic resonance neurography in the evaluation of patients with peripheral nerve pathology. J Neurosurg 1996;85(2):299–309. [9] Stanisz GJ, Midha R, Munro CA, Henkelman RM. MR properties of rat sciatic nerve following trauma. Magn Reson Med 2001;45(3):415–20. [10] Basser PJ, Pierpaoli C. Microstructural and physiological features of tissues elucidated by quantitative-diffusion-tensor MRI. J Magn Reson B 1996;111(3): 209–19. [11] Mori S, Zhang J. Principles of diffusion tensor imaging and its applications to basic neuroscience research. Neuron 2006;51(5):527–39. [12] Le Bihan D, Mangin JF, Poupon C, Clark CA, Pappata S, Molko N, et al. Diffusion tensor imaging: concepts and applications. J Magn Reson Imaging 2001;13(4): 534–46. [13] Kim JH, Budde MD, Liang HF, Klein RS, Russell JH, Cross AH, et al. Detecting axon damage in spinal cord from a mouse model of multiple sclerosis. Neurobiol Dis 2006;21(3):626–32. [14] Lehmann HC, Zhang J, Mori S, Sheikh KA. Diffusion tensor imaging to assess axonal regeneration in peripheral nerves. Exp Neurol 2010;223(1):238–44. [15] Takagi T, Nakamura M, Yamada M, Hikishima K, Momoshima S, Fujiyoshi K, et al. Visualization of peripheral nerve degeneration and regeneration: monitoring with diffusion tensor tractography. Neuroimage 2009;44(3):884–92.
101
[16] Li X, Chen J, Hong G, Sun C, Wu X, Peng MJ, et al. In vivo DTI longitudinal measurements of acute sciatic nerve traction injury and the association with pathological and functional changes. Eur J Radiol 2013;82(11):e707–14. [17] Morisaki S, Kawai Y, Umeda M, Nishi M, Oda R, Fujiwara H, et al. In vivo assessment of peripheral nerve regeneration by diffusion tensor imaging. J Magn Reson Imaging 2011;33(3):535–42. [18] Schmitz HC, Beer GM. Muscle-sparing approach to the peroneal nerve of the rabbit. Lab Anim 2001;35(4):334–9. [19] Di Scipio F, Raimondo S, Tos P, Geuna S. A simple protocol for paraffin-embedded myelin sheath staining with osmium tetroxide for light microscope observation. Microsc Res Tech 2008;71(7):497–502. [20] Gutmann E. Factors affecting recovery of motor function after nerve lesions. J Neurol Psychiatry 1942;5(3–4):81–95. [21] Gutmann E, Guttmann L, Medawar PB, Young JZ. The rate of regeneration of nerve. J Exp Biol 1942;19:14–44. [22] Schmitz HC, Beer GM. The toe-spreading reflex of the rabbit revisited–functional evaluation of complete peroneal nerve lesions. Lab Anim (NY) 2001;35(4): 340–5. [23] Beer GM, Seifert B, Schneller M, Iscru G, Schmitz HC. Relevance of muscle weight as a functional parameter for the regeneration of the peroneal nerve in rabbits. J Reconstr Microsurg 2008;24(1):11–9. [24] Nilsson M, van Westen D, Stahlberg F, Sundgren PC, Latt J. The role of tissue microstructure and water exchange in biophysical modelling of diffusion in white matter. MAGMA 2013:1–26. [25] Beaulieu C. The basis of anisotropic water diffusion in the nervous system–a technical review. NMR Biomed 2002;15(7–8):435–55. [26] Brennan FH, Cowin GJ, Kurniawan ND, Ruitenberg MJ. Longitudinal assessment of white matter pathology in the injured mouse spinal cord through ultra-high field (16.4 T) in vivo diffusion tensor imaging. Neuroimage 2013;82:574–85. [27] Jirjis MB, Kurpad SN, Schmit BD. Ex vivo diffusion tensor imaging of spinal cord injury in rats of varying degrees of severity. J Neurotrauma 2013;30(18): 1577–86. [28] Klawiter EC, Schmidt RE, Trinkaus K, Liang HF, Budde MD, Naismith RT, et al. Radial diffusivity predicts demyelination in ex vivo multiple sclerosis spinal cords. Neuroimage 2011;55(4):1454–60. [29] Song SK, Sun SW, Ju WK, Lin SJ, Cross AH, Neufeld AH. Diffusion tensor imaging detects and differentiates axon and myelin degeneration in mouse optic nerve after retinal ischemia. Neuroimage 2003;20(3):1714–22. [30] Song SK, Sun SW, Ramsbottom MJ, Chang C, Russell J, Cross AH. Dysmyelination revealed through MRI as increased radial (but unchanged axial) diffusion of water. Neuroimage 2002;17(3):1429–36. [31] Sun SW, Liang HF, Le TQ, Armstrong RC, Cross AH, Song SK. Differential sensitivity of in vivo and ex vivo diffusion tensor imaging to evolving optic nerve injury in mice with retinal ischemia. Neuroimage 2006;32(3):1195–204. [32] Budde MD, Xie M, Cross AH, Song SK. Axial diffusivity is the primary correlate of axonal injury in the experimental autoimmune encephalomyelitis spinal cord: a quantitative pixelwise analysis. J Neurosci 2009;29(9):2805–13. [33] Nair G, Tanahashi Y, Low HP, Billings-Gagliardi S, Schwartz WJ, Duong TQ. Myelination and long diffusion times alter diffusion-tensor-imaging contrast in myelin-deficient shiverer mice. Neuroimage 2005;28(1):165–74. [34] Zollinger LV, Kim TH, Hill K, Jeong EK, Rose JW. Using diffusion tensor imaging and immunofluorescent assay to evaluate the pathology of multiple sclerosis. J Magn Reson Imaging 2011;33(3):557–64. [35] Drobyshevsky A, Derrick M, Wyrwicz AM, Ji X, Englof I, Ullman LM, et al. White matter injury correlates with hypertonia in an animal model of cerebral palsy. J Cereb Blood Flow Metab 2007;27(2):270–81. [36] Ludeman NA, Berman JI, Wu YW, Jeremy RJ, Kornak J, Bartha AI, et al. Diffusion tensor imaging of the pyramidal tracts in infants with motor dysfunction. Neurology 2008;71(21):1676–82. [37] Raducan A, Mirica S, Duicu O, Raducan S, Muntean D, Fira-Mladinescu O, et al. Morphological and functional aspects of sciatic nerve regeneration after crush injury. Rom J Morphol Embryol 2013;54(3 Suppl.):735–9.
