Patients and the control group
One hundred and thirty-two symptomatic patients (78 women and 54 men; age range 18–76 years, mean 53.58 years) with different stages of cervical spondylosis observed on plain MR images and 25 control subjects (14 women and 11 men; age range 27–80 years, mean age 45.78 years) were enrolled in the study. The inclusion criteria for the patients were clinical and radiological signs and symptoms of degenerative cervical spine disease. Patients with congenital spinal canal narrowing, central spinal canal widening, previous spinal surgery and with any incidental findings on plain MR images suggestive of neurological disorder that could bias the results (e.g. inflammatory changes) were excluded from the study.
All control subjects were asymptomatic volunteers with no history of neurological disorders, consisting mainly of the hospital staff.
The study was conducted in accordance with the guidelines of the local University Ethics Committee for conducting research involving humans. Each patient provided their signed informed consent to participate in the examination.
MR imaging and DTI protocol
The MR examinations were performed with a 1.5-T MR unit (Signa Hdx, GE Medical Systems) with 33mT/m maximum gradient strength, using a sixteen-channel coil dedicated for head and spine imaging. The MR protocol consisted of sagittal T1-weighted images (TR/TE 555/10 ms), sagittal and axial T2-weighted images (TR/TE 3,580/111 ms), sagittal T2-weighted FAT SAT images (TR/TE 3,380/118 ms), followed by axial DTI sequence.
DTI acquisition was based on single-shot spin-echo echo-planar imaging (SE/EPI) with the following parameters: TR/TE 10.000/99 ms, 160 × 160 mm field of view, matrix 96 × 96, in plane image resolution 1.6 × 1.6 mm with a 4-mm-thick axial slices parallel to the intervertebral disk space, no gap, number of acquisitions: 2. In each study, the examination frame was adjusted to cover the length of the spinal cord from the second to seventh cervical vertebrae. Diffusion was measured along 15 non-collinear directions with two b values: 0 and 1,000 mm2/s resulting in 1 image without and 14 with diffusion weighting. The acquisition time of DTI ranged from 5 to 7 min.
Plain MR images of the cervical spine of 132 patients and 25 control subjects were analyzed by two independent investigators (A.B. and J.B.). Each spine segment for each disc level from C2/C3 to C5/C6 was evaluated separately in terms of evidence of degenerative spine disease (DSD). A C6/C7 level was excluded from the analysis due to a large amount of artifacts occurring at this level on diffusion tensor (DTI) images. Three hundred and forty-nine spine segments were selected altogether (268 from the patients` group and 81 from the controls) and subsequently divided into five groups as follows (Table 1):
Control group with no signs of DSD
Compression of the dura with obliteration of the subarachnoid space
Spinal cord surface scalloping but without signs of compression
Spinal cord compression without any changes in signal intensity of the spinal cord
Marked spinal cord compression with hyperintense signal changes on T2-weighted images.
The number of groups representing different stages of stenosis (A–E) at investigated cervical spine levels is shown in Fig. 1. The level most commonly affected by degenerative changes was C5/C6, while the degenerative changes in the earliest stage were observed most often at C2/C3. As far as the control group is concerned, only spine levels without any spinal canal obliteration visible on plain MR images were included in the study, which gave the total number of 81 spinal cord segments.
The anteroposterior (AP) diameter of the spinal canal and of the spinal cord was measured on the axial T2-weighted images at each evaluated level. Space available for the cord (SAC) value was determined by subtracting the sagittal spinal cord diameter from the corresponding sagittal spinal canal diameter, according to the equation: SAC = APSCaD − APSCoD (space available for the cord = anteroposterior spinal canal diameter − anteroposterior spinal cord diameter).
Image post-processing was done using the Functool software (GE ADW 4.4 workstation). Apparent diffusion coefficient (ADC) and fractional anisotropy (FA) transverse maps were generated and FA and ADC values were measured at the 268 levels corresponding to the selected spine segments presenting different stages of DSD, and at 81 reference spinal cord levels derived from healthy controls. The regions of interest (ROI) were drawn manually over the entire axial spinal cord cross section, according to the most accurate axial B0 image, as shown in Fig. 2. Special attention was paid to avoid partial volume effects, magnetic susceptibility effects and motion artifacts. In seven cases, ROI positioning was difficult to perform because of massive compression of the spinal cord (group E). These cases were not included in the statistical analysis.
Comparisons of the FA, ADC, AP spinal canal diameter (APSCaD), AP spinal cord diameter (APSCoD) and SAC values among all groups were made using a Student t test, ANOVA and LSD test. Comparisons were done between parameters derived from the same spinal cord levels. Correlations between spinal canal stenosis measurements (APSCaD, APSCoD and SAC values) and DTI parameters (FA, ADC) were estimated using Pearson’s correlation coefficients. Due to the small sample size, group E was excluded from statistical analysis. The Statistica10 software package was used for statistical calculations and p < 0.05 was considered statistically significant.