Sagittal alignment of the cervical spine: radiographic analysis of 111 asymptomatic adolescents, a retrospective observational study

A total of 111 adolescent subjects (44 males and 67 females), aged 15–20 years (mean age 16.12 ± 2.40 years), who visited our clinic for spinal deformity screening in 2014–2019, were included in this study. All the subjects were in Risser Grade 3–5, had no clinical neck and back symptoms, and X-rays confirmed no spinal deformities such as scoliosis or kyphosis. The study was approved by the Ethics Committee of Shanghai Tenth People’s Hospital, also informed consent was obtained from each subject or their legal guardians.

Standard standing whole-spine anteroposterior and later radiographs were obtained for each subject. All the subjects were asked to stand in a natural position, look straightforward, and keep their hands at the level of their clavicles according to the standard position recommended for adults [15, 16]. Regional and global sagittal alignment and spinopelvic alignment parameters were measured, including C7 slope, C2-7 Cobb, C2-S1 sagittal vertical axis (C2-S1 SVA), C7-S1 sagittal vertical axis (C7-S1 SVA), C2-7 plumb line (C2-7PL), T5-12 Cobb, T10-L2 Cobb, L1-S1 Cobb, pelvic incidence (PI), pelvic tilt (PT) and sacral slope (SS). All the radiographic parameters were measured by two experienced spinal surgeons independently.

C7 slope was formed by the horizontal plane and the upper endplate of C7 (Fig. 1a). C2-7 Cobb angle was measured from the inferior endplate of C2 to the inferior endplate of C7 (Fig. 1b). C2 plumb line and C7 plumb line were defined as the vertical line (plumb line) drawn from the middle of the C2 or C7 vertebral body (Fig. 1a). C2-S1 SVA and C7-S1 SVA were defined as the horizontal distances from the C2 or C7 plumb line to the posterior superior corner of the sacrum (S1) (Fig. 1a). The deviation of the C2 plumb line and C7 plumb line was defined as C2-7PL (Fig. 1a). T5-12 Cobb angle was measured from the superior endplate of T5 to the inferior endplate of T12 (Fig. 1b). T10-L2 Cobb angle was measured from the superior endplate of T10 to the inferior endplate of L2 (Fig. 1b). L1-S1 Cobb angle was measured from the superior endplate of L1 to the superior endplate of S1 (Fig. 1b). Pelvic parameters including PI, PT, and SS were measured according to the method described in the published paper (Fig. 1a) [6].

Besides, cervical morphology was classified by two different methods. One is a modified method by Toyama et al. [17, 18]. In brief, a line AB was drawn from the midpoints of the inferior margin of C2 to the midpoints of the superior margin of C7. And, the morphology was classified into four types according to the relative positions of the centroids of C3-6 to line AB (Fig. 2). Another method is based on the C2-7 Cobb angle < 0° or ≥ 0°, the cervical morphology was classified as lordotic and non-lordotic.

Table 3 showed the correlation efficiency of sagittal alignment parameters. C7 Slope exhibited significant negative correlation with C2-7 Cobb (r = -0.544, P < 0.001), while significant positive correlation with C2-S1 SVA (r = 0.335, P < 0.001), C7-S1 SVA (r = 0.310, P = 0.001), T5-12 Cobb (r = 0.236, P = 0.013), respectively. C2-7 Cobb showed significant negative correlation with C7-S1 SVA (r = -0.255, P = 0.007) but positively related with C2-7PL (r = 0.319, P = 0.001). C2-7PL showed significant positive correlation with C2-S1 SVA (r = 0.405, P < 0.001). Cervical reginal sagittal alignment parameters including C7 Slope, C2-7 Cobb, C2-7 PL showed no significant correlation with lumbosacral parameters (L5-S1 Cobb, PI, PT, SS).

All the subjects were divided into three groups according to cervical morphology, Lordotic group, Straight or Sigmoid group, and Kyphotic group, and the data were shown in Table 4. According to the modified method, there were 37 lordotic (33.3%), 39 straight (35.1%), 16 sigmoid (14.4%) and 19 kyphotic cases (17.1%). The mean value of the C7 slope in the Lordotic group is 26.91 ± 5.40°, which is significantly different from the Straight or Sigmoid group (17.91 ± 8.84°, P < 0.05). C2-7 Cobb showed significant differences between the three groups (P < 0.05) and C2-C7PL showed a significant difference between Lordosis and Kyphotic group (P < 0.05).

All the subjects were divided into two groups according to C2-7 Cobb, the Lordotic group (C2-7 Cobb < 0, n = 80), and the Non-Lordotic group (C2-7 Cobb ≥ 0, n = 31), and the data was shown in Table 5. Intergroup analysis showed that the C7 Slope (Lordosis vs Non-lordotic: 22.52 ± 8.94°vs 15.14 ± 6.18°, P < 0.001), C2-7PL (Lordosis vs Non-lordotic: 11.86 ± 12.49 mm vs 18.05 ± 7.3 mm, P = 0.001) and C7-S1 SVA were significantly different between the two groups.

Although it is well accepted that lordosis is the natural cervical alignment of the cervical spine, the definition of “normal” cervical spine morphology remains controversial. Yu et al. [7] included 120 cases of asymptomatic subjects (mean age 23.2 ± 6.3 years) showed that only 28.3% (34/120) of the subjects with lordotic cervical alignment. Similarly, Kim et al. [2] showed around one-fourth (26.3%) of asymptomatic adult volunteers have kyphotic cervical alignment. These data are from asymptomatic adult volunteers, more interestingly, K. Abelin-Genevois et al. reported data from a normal pediatric Caucasian population, which also found a high prevalence of kyphotic or straight morphology [15]. In this study, we found there were only 33.3% of asymptomatic adolescent subjects with cervical lordosis according to the modified method, and around 72.1% of subjects with cervical lordosis according to C2-7 Cobb. More importantly, all the subjects we included were adolescence ranging from 12 to 20 years old with Risser sign ≥ 3, it allows us to better understand the morphological characteristics of the cervical spine after it has reached a stable state of natural development. It provides more reliable evidence that the cervical spine morphology of asymptomatic adolescents varies widely, from lordotic to kyphotic. Therefore, it is more reasonable to diagnose a cervical spine alignment that fails to achieve the horizontal gaze or causes symptoms like neck pain as pathological, rather than a kyphotic cervical alignment itself.

Notably, these findings do support there is a quiet percentage of kyphotic cervical alignment in asymptomatic populations, but do not suggest cervical surgery planning should allow the cervical alignment to be kyphotic after surgery. Villavicencio et al. [19] showed that those patients who are maintained or more lordotic of the fused segment alignment postoperatively had better surgical outcomes than those who became more kyphotic. Brooke et al. [20] showed increased NDI scores were correlated with cervical kyphosis.

It is gradually becoming a consensus that not only regional but global spinal alignment should be taken into account for the surgical treatment of the cervical spine. In clinical practice, however, sometimes only lateral radiographs of the cervical spine are obtained pre-operatively, rather than the global spine, especially in some developing countries. So, what cervical regional sagittal alignment parameters can reflect the global sagittal balance? In this study, we showed that the C7 slope showed a significant correlation with global spine sagittal alignment parameters C2-S1 SVA and C7-S1 SVA. And we provided evidence that C2-7PL is the key to restoring cervical lordosis.

It is worth thinking about what method to use to define and classify the morphology of the cervical spine. C2-7 Cobb has been considered as a key parameter in cervical sagittal alignment, being used as a measurement of lordosis because of its good intra- and interrater reliability and feasibility [11, 24]. We divided all the subjects into the lordotic group (C2-7 Cobb < 0°) and the non-lordotic group (C2-7 Cobb ≥ 0°). There were 73% of subjects with lordotic cervical sagittal alignment, this percentage was similar to what Ella Been et al. [14] showed in a previous study, that there were 71% of the children with lordotic cervical spine in the children’s group (aging from 6–19 years old). However, the percentage is much higher than that classified by the modified method of Toyama et al. [17, 18], in which only 33.3% of subjects in the Lordosis group. Most recently, Sohrab Virk et al. [25] developed a new method to better define the morphology of the cervical spine. All these indicate that combining different methods to evaluate the cervical sagittal alignment can get a more comprehensive understanding of cervical sagittal morphology.

There are differences in the cervical spine morphology between gender. Previously, Yasutsugu Yukawa et al. [26] reported that C3-7 cervical lordosis showed a significant difference in 20–29 years old asymptomatic individuals. Kuang-Ting Yeh et al. [27] found that C7 Slope and C2-7 SVA were significantly different in asymptomatic adults. In this study, we showed that C2-7 Cobb, and C7 Slope showed significant differences between the male and female groups. All these findings gave evidence that there are differences in the cervical sagittal alignment between males and females, indicating sex differences should be considered pre-operatively.

In addition, we found cervical regional sagittal alignment parameters including C7 Slope, C2-7 Cobb, and C2-7 PL showed no significant correlation with lumbosacral parameters (L5-S1 Cobb, PI, PT, SS). Le Huec et al. [21] also showed a poor correlation between the pelvic parameters and the cervical parameters.

There are several limitations of this study. The total number of subjects we included is relatively small. Another limitation is that we didn’t get the data like HRQOL and NDI from the subjects, which could better quantify the relation between sagittal alignment parameters and life quality. On the other hand, although all the subjects we included were currently asymptomatic, there were no following-up data to investigate the relationship between cervical morphology and degeneration. Also, there are some studies included some angles between the head and cervical spine like Occipito C2 angle, spino-cranial angle [15, 28]. Since the subjects in this study are all asymptomatic and can gain a horizontal gaze, so we didn’t include these parameters. However, given the fact that our data were from asymptomatic adolescent subjects, making it is worthy to better understand the morphology of cervical sagittal alignment.

In summary, our study demonstrated that there are different types of cervical morphology in asymptomatic adolescents, with a wide range of C2-7 Cobb angles, from lordosis to kyphosis. C7 slope is an important predictor of global sagittal balance and C2-7PL is a key parameter for restoring cervical lordosis, which should be considered pre-operatively and for conservative treatment. The classified method will make a different conclusion about cervical morphology, a more comprehensive understanding of cervical morphology can be gained by combining the different methods. The cervical regional sagittal alignment parameters are not correlated with lumbosacral parameters, and C2-7 Cobb, C7 Slope showed significant differences between males and females.