Spinal subarachnoid space tapering in patients with syringomyelia

Abstract

Background and purpose

Cervical spine tapering affects cerebrospinal fluid dynamics. Cervical spine taper ratios derived from anteroposterior diameters reportedly differ between patients with syringomyelia and controls. We attempted to verify the differences in diameter and to show differences in cross-sectional area between syringomyelia and controls.

Methods

Cervical spine magnetic resonance images in syringomyelia patients (idiopathic or Chiari I related) and control patients were examined. In each subject, the anteroposterior diameter of the spinal canal was measured at each cervical level, and C1–C4, C4–C7, and C1–C7 taper ratios were calculated. Differences in taper ratio between groups were tested for statistical significance with the t-test. Cross-sectional areas of the spinal canal were measured at each cervical spinal level, and tapering was calculated.

Results

Eighteen patients with idiopathic syringomyelia, 28 with Chiari I, and 29 controls were studied. Chiari and syringomyelia patients had significantly steeper diameter-based taper ratios than controls. The dural sac areas tapered proportionally with the diameter-based taper ratio in all groups.

Conclusions

Cervical spine anteroposterior diameter tapering and dural sac cross-sectional areas tapering differ between syringomyelia patients and controls.

Introduction

Syringomyelia occurs in patients with Chiari I, theoretically because of abnormal cerebrospinal fluid (CSF) dynamics. Syringomyelia also occurs in patients without spinal deformity, infection, tumor, or trauma (idiopathic syringomyelia), theoretically because of abnormal CSF dynamics. The pathogenesis of abnormal CSF dynamics in these cases remains obscure. One important factor may be that the abnormal tapering of the cervical spinal canal found in these patients alters spinal CSF dynamics. In a study of Chiari I patients, tapering of the C1–C7 and the C4–C7 spinal segments differed significantly between the subgroups with and without syringomyelia.¹ In a study of patients with idiopathic syringomyelia, C3–C7 tapering differed between patients and controls.² Chiari I patients with syringomyelia in another study had significantly different C1–C7 ratios compared to controls.³ Patients with a presyrinx, a condition presumably progressing to syringomyelia, had abnormal C4–C7 spinal canal tapering.⁴ Investigators found that taper ratios for the C1–C7, C1–C4, and C4–C7 cervical segments differed significantly between patients with distended syringes compared to those with non-distended syringes.⁵

The effect of spinal canal tapering on CSF dynamics and its role in the pathogenesis of syringomyelia is incompletely studied. Peak systolic CSF velocities increase on average by 30% between C1 and C4 due to spinal canal narrowing.⁶ Since flow velocity and the cross-sectional area are inversely related, CSF velocities differ between patients with different degrees of spinal canal tapering. Subjects with greater tapering theoretically have greater acceleration of CSF along the tapering segment.

The goal of this study was to corroborate the greater than normal taper ratios in patients with syringomyelia and to investigate the correlation of taper ratios to the tapering of cross-sectional area of the dural sac.

Materials and methods

The Institutional Review Boards at the University of Wisconsin School of Medicine and Public Health and Tufts Medical Center approved this retrospective study. The requirement for written informed consent was waived. The study conformed to the Health Insurance Portability Accountability Act standards.

Radiology magnetic resonance imaging (MRI) reports from December 1, 2004, through June 30, 2017, were searched to identify all cases that had syrinx, syringomyelia, or syringo-hydromyelia in any part of the spinal cord. Exclusion criteria included a Chiari I or history of prior spinal trauma, spinal surgery, or neurologic tumors or infections. Chiari I patients with syringomyelia were selected from a clinical database.¹ Inclusion criteria included ≥5 mm tonsillar herniation and otherwise, the same inclusion and exclusion criteria as for the idiopathic syringomyelia group. Controls were selected in the PACS database by identifying patients with reportedly normal cervical and thoracic spine MRIs. Age and sex matching between groups was not attempted.

One reader (E.R.) supervised by a senior staff neuroradiologist at Tufts Medical Center measured spinal canal diameters and dural cross-sectional areas at each cervical level. The reader used DICOM viewing software (OsiriX, Pixmeo) to place lines on the sagittal T2 image perpendicular to the spinal canal to bisect each cervical body from C2 to C7. At C1, the line was placed through the anterior arch of C1 perpendicular to the canal. Axial T2 images were reformatted in the plane perpendicular to the spinal canal at each level. The spinal canal diameter was measured as the distance from the anterior to the posterior dural boundary in the midline. The area of the spinal subarachnoid space was also measured by manually tracing the spinal subarachnoid space at each level and recording the area computed by the viewing software.

The length of the C1–C4 segment was measured on sagittal T2-weighted image as the distance between the posterior superior edge of the C1 vertebral body to the posterior-inferior edge of C4. It was measured similarly for C4–C7 as the distance from the posterior-superior edge of C4 to the posterior-inferior edge of C7. The length of C1–C7 was measured from the posterior-superior edge of C1 to the posterior-inferior edge of C7.

Analysis and statistics

Taper ratios for C1–C7, C1–C4, and C4–C7 were calculated by least squares linear fitting¹ (LINEST macro in Excel; Microsoft, Redmond, WA). Taper ratios were also calculated as a percent of the length of the relevant spinal segment, a dimensionless number. Taper ratios for the patient groups were compared to those for controls and tested for statistical significance by means of Student’s t-test (single-tailed), with significance set at 0.05. The normality of the data distributions was tested with the Shapiro–Wilk test. Mean cross-sectional areas were plotted against diameters for each level of the cervical spine and for each group separately, and trend lines and goodness of fit were calculated.

Results

Seventy-nine subjects were identified, of whom four were excluded because images were suboptimal for the study. The 75 individuals in the study included 18 patients with idiopathic syringomyelia, 28 with Chiari I-related syringomyelia, and 29 controls. The age range was 3–67 years, and the average age for each subgroup was 31–32 years. The male-to-female ratio ranged from 1:2.0 to 1:1.9 in the three groups. Spine segment lengths averaged 5.7 (±0.9), 5.5 (±1.5), and 10.2 (±1.5) cm for the C1–C4, C4–C7, and C1–C7 segments, respectively. Differences in segment lengths between groups were not significant.

Spinal canal diameters in the three groups diminished from C1 to C4 (Figure 1). Diameters from C4 to C7 tended to increase or to remain nearly constant depending on the group. On average, the C1–C7 and C1–C4 taper ratios had negative signs for all groups, indicating tapering of the spinal canal in a caudal direction, whereas the C4–C7 taper ratio on average had a positive sign for all groups, indicating a reverse tapering or flaring of the spinal canal in the caudal direction.

Figure 1.

Plot of anteroposterior diameters by cervical spine level in patients and in controls. Cases have steeper tapering than controls.

C1–C4 taper ratios calculated from anteroposterior diameter measurements were –0.130 cm/level for the idiopathic syrinx group, –0.123 cm/level for the Chiari syrinx group, and –0.117 cm/level for the control group (Figure 2). Average C4–C7 taper ratios were 0.04, 0.03, and 0.01 cm/level respectively, for the three groups. C1–C7 anteroposterior diameter taper ratios were −0.04, –0.04, and –0.05 cm/level, respectively, for the idiopathic syrinx, Chiari syrinx, and control groups. As a percentage of the segment length, the taper ratios for C1–C4 segment were –8.66, –8.77, and –8.20, respectively, and for the C4–C7 segment, they were 2.81, 2.03, and 0.98, respectively, for the three groups (Figure 3).

Figure 2.

Bar graph showing the taper ratios in controls and patient groups. The C1–C4 taper ratios were significantly steeper in the idiopathic syringomyelia and Chiari 1 patients with syringomyelia than in controls. C4–C7 taper ratios were significantly greater in both idiopathic syringomyelia and Chiari 1-related syringomyelia than in controls.

Figure 3.

Bar graph showing the taper ratios as a percentage of the segment length for the cases and controls. The relative magnitudes of the taper ratios in the three patient groups show similar relationships to those in Figure 2.

The differences in C4–C7 taper ratios between patients with idiopathic syrinx and controls (p = 0.007), and between Chiari with syrinx and controls (p = 0.012) were significant. The difference in the C1–C4 and C4–C7 taper ratios between the syrinx and control group were significant (p = 0.0185 and 0.0255). The data were normally distributed by analysis with the Shapiro–Wilk test.

Dural sac area decreased comensurately with diameter from C1 to C4 in each group (Figure 4). The goodness of fit was 0.92–0.94 for the three groups.

Figure 4.

Plot of dural sac areas by level in patients and controls. Areas decrease from C1 to C4 and increase from C4 to C7.

Discussion

Patients with idiopathic syrinx and Chiari-associated syrinx have significantly steeper C1–C4 and C4–C7 taper ratios than controls. Dural sac cross-sectional areas correlate with the taper ratios based on cervical spine diameters.

The taper ratios measured in the study agree with values reported in previous studies. The taper ratios in controls were, respectively, –0.12, 0.01, and −0.05 cm/level for C1–C4, C4–C7, and C1–C7 in this study, and –0.11, 0.01, and –0.04 cm/level in a published study of normal controls.⁷ In one study,² lower cervical spine diameters tapered more steeply in 50 patients with idiopathic syringomyelia than they did in control patients, as was the case in this study.² We found Chiari patients with syrinx had C1–C7 of −0.06 cm/level, in agreement with a previous study.³ Our data on dural sac area tapering in controls agree well with those in previous studies of normal subjects.^8–11

The study design has limitations. The process of selecting patients was not random, but it was arbitrary. Age and sex were not intentionally matched but were similar in each group. The technique of measuring diameters used in this study differed from previous studies. The use of axial reformatted images in this study may have reduced errors due to imperfect midsagittal section choice in previous studies. Geometric errors in reformatted images are minimized by the choice of a plane perpendicular to the spinal axis for reformatting. The measurement technique in this study did not result in important differences between these and previously published results.

Significantly, this study shows abnormal tapering of the cervical spinal canal in syringomyelia, either idiopathic or Chiari related, which may contribute to the pathogenesis of the syrinx.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of interest

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.