Cervical spine taper ratios: Normal tolerance limits

Abstract

Background

Spinal canal tapering, which can be measured as taper ratios, affects cerebrospinal fluid flow dynamics. We calculated the tolerance interval for normal cervical spine taper ratios to facilitate the detection of abnormal taper ratios.

Methods

We collected a series of patients who had cervical spine magnetic resonance studies reported as normal. We measured anteroposterior diameters of the cervical spine and calculated C1–C4, C4–C7, and C1–C7 taper by standard methodology. We calculated the normal tolerance limits for taper ratios and compared results of this study with data in previous reports on taper ratios.

Results

We enrolled 78 patients aged 2–55 years. The 99% tolerance limits for the taper ratios for C1–C4, C4–C7, and C1–C7 were −0.31 to +0.09, −0.11 to +0.14, and −0.13 to +0.05 cm/level, respectively. Age and sex were not significant variables for taper ratios. Taper ratios in this study agreed with those reported for controls in previous studies. Patients with syringomyelia in previous reports tended to have taper ratios outside the normal tolerance limits.

Conclusion

Normal limits of the cervical taper ratios are reported.

Introduction

According to prevailing theory, syringomyelia results from abnormal cerebrospinal fluid (CSF) dynamics. Many studies have investigated the possibility that obstruction in the foramen magnum creates abnormal CSF flow because ectopic cerebellar tonsils obstructing the foramen magnum (Chiari I deformity) often have associated syringomyelia. Research on syringomyelia includes measurements of cervical spinal dimensions as the cervical spinal canal modulates CSF flow velocities below the foramen magnum. The normal cervical spine tapering between the foramen magnum and C4 accelerates CSF flow and increases CSF flow complexity.^1,2 Cervical spine tapering is often measured in terms of taper ratios, the slope of a line fit to selected spinal canal diameters, usually C1–C7, C1–C4, or C4–C7. Taper ratios are steeper in patients with Chiari deformity than in controls,^3,4 and are steeper in those patients with Chiari I deformity with syrinx than in those without syrinx.⁵ Taper ratios reportedly differ between patients with distended and non-distended syringes.⁶ Taper ratios therefore have a role in identifying patient groups that may have abnormal CSF dynamics.

Taper ratios have been measured using small groups of normal participants as controls for the patient groups. We designed a study to collect taper ratios in a large series of normal participants, with the goal of calculating the tolerance limits, i.e., the range of taper ratios for the normal general population. These limits may be useful for identifying patient groups or possibly individual patients who have abnormal CSF dynamics, and to validate conclusions in previous studies with small control groups.

Materials and methods

The institutional review board at the University of Wisconsin School of Medicine and Public Health (approval number: 2013-0459-CR005) approved this retrospective study and waived the requirement for written informed consent. This study conforms to the standards set out in the USA Health Insurance Portability and Accountability Act.

A consecutive series of magnetic resonance (MR) studies of the cervical spine was collected by selecting an arbitrary date (January 1, 2012) and searching chronologically for MR studies of the cervical spine that were reported as normal in patients referred for symptoms or signs unrelated to the spinal column. MR studies were performed on 1.5 or 3 T magnets with neck coils. The MR acquisition in each patient included at least one T2-weighted, spin echo sagittal sequence with 3000–4000/100–120 TR/TE, 4 mm slice thickness, 320 × 320 field of vision, 10% gap, and 2 excitations averaged. One reader, a neuroradiology fellow, reviewed each case to exclude studies with poor image quality or artifacts, pathology of the vertebral column, syringomyelia, scoliosis, previous spine surgery, spinal degenerative changes, spinal anomalies, or tonsillar ectopia. Age and sex of the patients was also recorded.

The reviewer selected the sagittal section or sections in each study that best showed midline structures, such as the spinous processes or the dens. The reviewer measured the anteroposterior diameter of the subarachnoid space from C1 to C7 on the selected midline sagittal image, using a previously described method.⁴ The reader placed a line perpendicular to the spinal canal axis at the level halfway between the discs, identified the points where the line crossed from the CSF to epidural tissue anterior and posterior to the subarachnoid space, and measured the distance between these two points with the measurement tool in the PACS viewer (Figure 1). At C1, the cranio-caudal midpoint of the anterior arch of C1 was defined, the line was drawn perpendicular to the spinal axis at this level, and diameter of the subarachnoid space measured on this line.

Figure 1.

Sagittal magnetic resonance image showing the method for calculating anteroposterior diameters from which taper ratios were calculated.

The C1–C7, C1–C4, and C4–C7 taper ratios were calculated for each patient by least squares fitting of a line to the appropriate diameters, by means of the LINEST macro in Microsoft Excel.⁴ The mean and the SDs for the taper ratios for the entire population, and for females and males separately were calculated. The taper ratios were plotted against age and the linear trendline (least squares fit) was calculated for each taper ratio.

Statistical methods

Normality of the data was assessed with quantile-quantile plotting. We screened for outliers using the generalized extreme studentized deviate test.⁷ The association of taper ratio with age and sex was tested with the multiple linear regression. The k factor was calculated,^8,9 and then the 99% confidence interval was calculated and adjusted by means of the k factor to calculate the 99% tolerance limits.

Results

Seventy-eight patients, including 39 males and 39 females were included in the study. Typically, the MR studies were ordered because of suspected demyelinating disease, acute ligamentous injury, myelitis, or metastasis to the spinal cord. On average, antero-posterior diameters diminished from C1 to C4 by 0.31 cm (negative tapering) and increased from C4 to C7 by 0.09 cm (positive or reverse tapering) (Figure 2).

Figure 2.

Average antero-posterior diameters for the seven cervical levels.

CL: Confidence limit; AVG: average.

Mean taper ratios for the patients were −0.11, 0.01, and −0.04 cm/level for C1–C4, C4–C7, and C1–C7, respectively (Table 1 and Figure 3). Men had steeper taper ratios than women by 0.015 cm/level for C4–C7 and <0.01 cm/level for C1–C4 and C1–C7. The three taper ratios varied per year by 0.0006 cm/level or less.

Table 1.

Mean taper ratios and tolerance limits for cervical spine (cm/level).

Taper ratio	Lower limit	Mean	Upper limit
C1–C7	−0.13	−0.04	0.05
C1–C4	−0.31	−0.11	0.09
C4–C7	−0.11	0.01	0.14

Figure 3.

Tolerance limits and means for C1–C4, C4–C7, and C1–C7 taper ratios.

The quantile-quantile plot showed the data distributed normally. No outliers were present. The 99% tolerance limits for the taper ratios for C1–C4, C4–C7, and C1–C7 were −0.31 to +0.09, −0.11 to +0.14, and −0.13 to +0.05 cm/level, respectively (Table 1 and Figure 3). Age, sex, and the product of age and sex were not significant variables for taper ratios (Figure 4). P values for the effect of age for the C1-C4, C4-C7, and C1-C7 taper ratios were 0.7, 0.997, and 0.8, respectively; p values for differences between sexes were 0.9, 0.5, and 0.06, respectively.

Figure 4.

C1–C4, C4–C7, and C1–C7 cervical spine taper ratios in the 78 participants with their linear trendlines plotted against age. The three taper ratios varied less than 0.0006 cm/level per year (insignificant, P = 0.7–0.997). The upper tolerance limit for C4–C7 and the lower tolerance limit for C1–C4 are shown. See Figure 3 for both tolerance limits for each taper ratio.

Discussion

In patients between the ages of 2 and 55 years, with no clinical evidence of spinal column pathology and with normal appearing MR scans, the mean C1–C4, C4–C7, and C1–C7, taper ratios were −0.11, 0.01, and −0.04 cm/level, respectively. Neither age nor sex had a significant effect of the taper ratios, although the magnitude of the male C4–C7 taper ratio exceeded the female ratio by an almost significant amount. The tolerance limits calculated for the C1–C4, C4–C7, and C1–C7 taper ratios were −0.31 to 0.09, −0.11 to 0.14, and −0.13 to 0.05, respectively.

Taper ratios in this study were in agreement with those previously reported for normal patients. A previous MR study of 1211 Japanese individuals¹⁰ reported mean spinal diameters of C2–C7, from which we calculated the mean C4–C7 taper ratio to be 0.021 within the tolerance limits in our study. In other published studies, midsagittal anteroposterior cervical spinal canal diameters from C2 to C7 were measured by computed tomography images¹¹ or by analogue or digital lateral radiography.^12–16 The calculated C4–C7 taper ratio in the participants in these studies averaged 0.06 mm/level, which is within our tolerance interval.

The number or percentage of patients with taper ratios exceeding normal tolerance intervals has not yet been calculated. In some previous studies, the mean taper ratios for patients with syrinx or presyrinx exceeded the tolerance limits in this study, indicating that at least 50% of the participants in these studies exceeded the tolerance limits (Table 2). The mean C4–C7 taper ratios in patients with Chiari I and syrinx (0.038 cm/level) exceeded the normal tolerance interval, and 41 of the 60 patients had C4–C7 taper ratios outside the normal tolerance interval.⁵ Patients with presyrinx had a mean C4–C7 taper ratio exceeding tolerance intervals.¹⁷ Patients with idiopathic syringomyelia had steeper than normal positive tapering between C3 and C7.¹⁸

Table 2.

Mean cervical spinal canal taper ratios for patient groups in recent publications (data from this study is included for comparison).

Study Patients	n	C1–C4 taper ratio, mm/level	C4–C7 taper ratio, mm/level	C1–C7 taper ratio, mm/level
This study	78 Normal	−1.11 (0.09 to −0.31)	0.01 (0.14 to −0.11)	−0.04 (0.05 to −0.13)
Hammersley et al.4	22 with scoliosis with Chiari	−2.2	No data	−0.8
Hammersley et al.4	32 with scoliosis without Chiari	−1.1	No data	−0.4
Thompson et al.5	49 with Chiari with syrinx	−1.22	0.28	−0.35
Thompson et al.5	101 with Chiari without syrinx	−1.34	0.11	−0.5
Struck et al.18	50 with idiopathic syrinx	−0.98	0.21	−0.30
Zhu et al.6	44 with Chiari with distended syrinx	−0.98	0.24	−0.34
Zhu et al.6	33 with Chiari with non-distended syrinx	–1.61	–0.04	–0.73
Gadde et al.17	21 with presyrinx	–1.31	0.32	No data

Bolded values are means that fall outside our tolerance intervals. Values in parentheses show 95% confidence intervals (95% CIs).

Our study has limitations. We cannot exclude a bias due to our selection of patients with normal MR studies rather than healthy volunteers. The results of this study cannot be generalized to all ages. A reader bias is unlikely because the reader did not consider age or sex in the measurements and reader error for taper ratios is reported to be <0.005 mm/level in 95% of cases.⁴ We report tolerance limits rather than confidence limits or SDs to provide ranges for the general population rather than for the statistical sample.

Conclusion

Normal tolerance limits for tapering of the upper cervical spinal canal or the entire cervical spinal canal are −0.31 to +0.09 cm/level and −0.13 to +0.05 mm/level, respectively. For the lower cervical spinal canal, they are −0.11 to +0.014 cm/level. Age and gender are not significant variables for the tolerance limits.

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.