Introduction
Over the past few decades, increasing attention has been given to the role of the mutual compensation between the upper and lower cervical spinal alignment in maintaining the sagittal balance of the cervical spine and horizontal gaze [
1,
2]. The kyphosis of the upper cervical spine caused by different craniovertebral junction pathologies or surgery of the upper cervical spine recruits a compensatory mechanism to increase lower cervical lordosis [
3,
4]. Kyphosis of the lower cervical spine is compensated for by the increase in upper cervical lordosis. To evaluate the parameters of cervical sagittal alignment and analyze the compensation mechanism between upper and lower cervical sagittal alignment in the normal population and in patients with cervical spine disorders, many different evaluation parameters of cervical sagittal alignment have been proposed [
5]. The C0-1 angle, C0-2 angle, and C1-2 angle were the most commonly used parameters to evaluate the sagittal alignment of the upper cervical spine. The correlation between these three parameters and the lower cervical sagittal alignment (C2-7 angle) has been repeatedly reported in normal people and patients with upper or lower cervical spine diseases [
2,
5,
6].
In past studies, different reference lines on the skull were used to measure the C0-1 angle and C0-2 angle [
5,
7‐
9]. The reference line on C1 was also different when measuring the C0-1 angle and C1-2 angle [
8,
10,
11]. The Frankfort horizontal line (FHL), foramen magnum line (FML), and McGregor line (ML) were previously commonly used as the reference lines while measuring the C0-1 angle and C0-2 angle [
7‐
9]. In the evaluation of upper cervical sagittal alignment, it is more important to choose a measurement method with high consistency and repeatability. However, no study has compared the consistency and repeatability of these methods. Therefore, the purpose of this study was to compare the reliability of using FHL, FML, and ML to measure C0-1angle and C0-2 angle. We analyzed the correlation between measurement results of the three methods and sagittal alignment of the lower cervical spine, which is expected to facilitate the selection of a more ideal measurement method for upper cervical sagittal alignment.
In addition, cervical sagittal radiographic alignment changes with age and sex in asymptomatic adults [
2,
12,
13]. However, few studies have specifically evaluated changes in correlations of upper and lower cervical sagittal alignment in various age groups. Therefore, we also explored the correlations of sagittal parameters of the upper and lower cervical spine in different age groups and analyzed the effect of sex on correlations of upper and lower cervical sagittal parameters.
Discussion
Previous studies have measured the C0-1 angle, C0-2 angle, and C1-2 angle in normal or asymptomatic people to obtain the normal range of these parameters and guide the selection of the C0-2 angle and C1-2 angle in occipitocervical fixation and atlantoaxial fixation to avoid abnormal changes in the lower cervical curvature or other postoperative complications caused by inappropriate C0-2 and C1-2 angles [
1,
14,
15]. These parameters have also been used to analyze and study the correlation of sagittal alignment and the mechanism of mutual compensation of the upper and lower cervical spine. It has been reported that in occipitocervical fusion, a C0-2 angle that is too small may lead to pharyngeal stenosis and dysphagia, while a C2-7 angle will result in compensatory enlargement [
16]. Tang et al. also reported that a small C0-2 angle is associated with postoperative implant failure [
15]. If the C0-2 angle is too large, the lower cervical spine may have a compensatory decrease in lordosis or even kyphosis [
14]. A retrospective study by Yoshimoto et al. found that surgical fixation of the atlantoaxial joint in a hyperlordotic position will lead the lower cervical spine into a kyphotic sagittal alignment after C1-2 fixation and fusion [
11]. Biomechanical studies have also shown that C1-2 angle fixation in a neutral position adds 10° to increase intradiscal pressure in the lower cervical spine, which may lead to accelerated degeneration of the lower cervical spine [
17]. However, there was no significant difference in the intradiscal pressure of the lower cervical spine between fixation in a neutral position minus 10° and fixation in a neutral position. Therefore, it is recommended to fix the C1-2 angle in a neutral position or a relatively smaller angle. It has been proven that the range of motion of C0-1 in the sagittal plane plays a vital role in compensation after atlantoaxial fixation and cervical kyphosis surgery. Kim et al. revealed that a small ROM at the C0-1 segment is a main risk factor for lower cervical kyphotic change after C1-2 fixation [
6]. Miyamoto et al.‘s study demonstrated that the C0-1 angle is more important than the C1-2 angle in the compensatory mechanism for kyphotic deformities at the lower cervical spine [
8].
At present, the methods of measuring the C0-1 angle, C0-2 angle, and C1-2 angle are multitudinous in different studies [
7‐
9]. This is mainly shown in the different reference lines selected on the skull when measuring the C0-1 angle and C0-2 angle and the different reference lines selected on C1 when measuring the C0-1 angle and C1-2 angle. Matsunaga et al. measured the C0-2 angle formed by the McGregor line and the inferior surface of the axis in 240 healthy volunteers, which is also the most commonly used measurement method for the C0-2 angle [
1]. Guo et al. also used the McGregor line as a reference line to measure the C0-1 angle and C0-2 angle in asymptomatic volunteers. Their correlation analysis showed that the C0-1 angle and C0-2 angle were negatively correlated with the C2-7 angle [
18]. A negative correlation was also found between the C0-2 angle and the C2-7 angle when using the MacRae line as the reference line to measure the C0-1 angle and the C0-2 angle in patients without spinal deformities [
8]. In addition, Liu et al. used the Frankfort horizontal line as the reference line on the skull and defined the angle formed by the Frankfort horizontal line and the inferior surface of the axis as the Frankfort axial angle [
7]. This measurement method conducted by Liu et al. is similar to using the McGregor line or MacRae line as the reference line, which indicates an angle between the reference lines selected on the skull and the C2 vertebral body; therefore, we can also regard the Frankfort axial angle as the C0-2 angle. The present studies showed that the McGregor line, MacRae line, or Frankfort horizontal line were selected as the reference lines on the skull for measuring the C0-1 angle and C0-2 angle, and the consistency and repeatability were high [
7,
8,
18]. Few studies have been conducted comparing reliability and reproducibility among the three measurements of upper cervical sagittal parameters. However, the ICC values of interobserver and intraobserver reliabilities for using the ML as a reference line to measure the C0-1 angle and C0-2 angle were significantly higher than those for using the FHL and FML as reference lines. The ICC values of interobserver and intraobserver reliabilities for using the FHL as a reference line to measure the C0-1 angle and C0-2 angle were slightly higher than those for using the FML as a reference line. The accuracy of measurement appears to depend on the clarity of the measurement landmarks on given radiographs. Shoda et al. found that either the basion or opisthion was difficult to identify in some of the films; the lateral bony structure of the skull base seems to render the occipital border dull and obscure [
19]. We also found that the lateral bony structure of the skull base may obscure the orbitale and porion. Therefore, when we drew the McGregor line on the skull in a neutral lateral cervical radiograph, it was more stable and accurate than the MacRae line and Frankfort horizontal line. Due to the highest repeatability and consistency of the ML line as a reference line to measure upper cervical sagittal parameters, it is most advantageous to use this method for requiring stable measurement of upper cervical sagittal parameters, such as the determination of upper cervical spine sagittal parameters in upper cervical surgery and the comparison of upper cervical spine sagittal parameters before and after surgery. In addition, although using the FHL as a reference line yields a lower ICC value than using the ML when measuring the C0-1 angle and C0-2 angle, using the FHL as a reference line has a higher correlation between the C0-1 angle and C0-2 angle and the C2-7 angle than the other two lines. In the study of the compensation mechanism of sagittal alignment of the upper and lower cervical spine, use of the FHL line as the reference line for measuring the C0-1 angle and C0-2 angle is superior.
The C2-7 angle has been confirmed to be correlated with age and sex in previous studies [
2,
12,
20]. Iorio et al. found that the C2-7 angle increased with aging, while the C0-2 angle did not change with age in asymptomatic people [
2]. Our research also found that the correlation between the C0-1 angle and C0-2 angle measured by the three methods and the C2-7 angle gradually increased with aging. This is the result of long-term compensation of the upper and lower cervical spine to maintain horizontal visual balance. In all age groups, the correlation between the C0-1 angle and C2-7 angle of the FHL as the reference line was higher than that of the FML or ML as the reference line. The correlation between the C0-2 angle and C2-7 angle of the FHL as the reference line was also stronger than that of the FML or ML as the reference line. The correlation between the C0-2 angle and the C2-7 angle in females was higher than that in males in the study by Nojiri et al [
20]. In our study, the correlation between the C0-1 angle and C2-7 angle in the female group was higher than that in the male group. In addition, the correlation between the C0-2 angle or C1-2 angle and the C2-7 angle in the female group was more advanced than that in the male group. This shows that sex is an important parameter affecting the correlation of sagittal parameters of the upper and lower cervical regions.
In some studies, the line passing through both the anterior tip of the anterior arch and the posterior tip of the posterior arch of the atlas was used as the reference line to measure the C0-1 angle and C1-2 angle [
8]. However, some other studies instead take the inferior aspects of the atlas as the reference line for measuring the C0-1 angle and C1-2 angle [
10,
11,
21,
22]. The C0-1 angle and C1-2 angle are often used to study the changes in cervical sagittal alignment before and after C1-2 fixation in patients with atlantoaxial dislocation [
21]. Tan’s technique and Harms’ technique are commonly used in C1 screw implantation during atlantoaxial fixation [
23,
24]. However, the C0-1 angle and C1-2 angle are difficult to accurately measure using the previous measurement methods because the tail of the C1 screw will obscure the middle-lower part of the posterior arch of the atlas after atlantoaxial fixation surgery by Tan’s technique or Harms’ technique. Therefore, we chose the superior aspect of the atlas as the reference line to measure the C0-1 angle and C1-2 angle. The ICC values of interobserver and intraobserver reliability of the method for measuring the C1-2 angle were 0.934 and 0.940, respectively. Pearson correlation analysis also showed that the C0-1 angle and C1-2 angle were negatively correlated with the C2-7 angle.
The present study had some limitations. First, this study was a single-center retrospective analysis, and the number of cases was small. A multicenter prospective study is required to confirm the current conclusion. Second, the three measurement methods were used and compared only in asymptomatic people in our study. However, which method is more conducive to the evaluation of cervical sagittal alignment and the study of the compensation mechanism of the upper and lower cervical spine in patients with different cervical diseases needs to be further explored.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.