In this study, we determined the reference values for estimating the occipital-cervical distance in neutral, flexion, and extension positions via our novel OC4D measurement method, which may provide a comprehensive and accurate estimation for vertical reduction of the occiput-cervical region during OCF. Importantly, the OC4D—as a simple, convenient, and highly reliable measurement of occiput-cervical distance—is not occluded by implants. More importantly, our present study revealed that the OC4D is not affected by changes in neutral, flexion, and extension cervical positions.
Conceptually, the occipitocervical neutral position is the functional and balanced position of the head atop the cervical spine. We considered that patients should have a normal occipitocervical angle and occiput-cervical distance in this neutral position. Sherekar et al.  measured the occipito-C2 angle in 518 asymptomatic volunteers (261 male and 257 female subjects), and obtained values of 14.66 ± 9.5° in males and 15.59 ± 8.26° in females. Many researchers have reported that non-normal occipitocervical angles lead to poor postoperative fusion, and even severe dysphagia and/or dyspnea during OCF [3, 8,9,10]. However, it remains unknown whether dysphagia and/or dyspnea are mostly due to mechanical airway obstruction caused by a non-normal occipitocervical angle. We believe that surgeons should pay more attention to the lower cranial nerve stretch airway obstruction caused by over-distraction of the occiput-cervical vertical distance. Shigeto et al. reported that the mechanism of dysphagia is not simply associated with the O-C2 angle, but that it also involves global craniocervical alignment in an individual patient, including the occiput-cervical distance . Wang et al. reported that performing OCF in the over-distraction position to treat vertical atlantoaxial dislocation may caudally displace the brainstem relative to the cranial base, resulting in traction injury to the 9th, 10th, and 11th lower cranial nerves .
In 1999, Phillips et al. first measured the occiput-cervical distance of OCD by measuring the shortest distance from the most superior aspect of the C2 spinous process to the occipital protuberance in 30 asymptomatic subjects. In this initial study, the value of the OCD in the neutral position was 21.5 ± 1.22 mm, and it was significantly different from OCD values measured in flexion (28.0 ± 1.32 mm) and extension (14.8 ± 1.48 mm) . Seong et al. measured OCDs in 200 normal, sagittal balanced patients (100 male and 100 female patients), and the mean neutral OCD was 22.98 ± 5.10 mm (range, 9.88–38.64 mm), which was significantly different from those in flexion and extension positions . In our present study, the mean neutral, flexion, and extension OCDs were 23.0 ± 4.8 mm, 27.6 ± 6.0 mm and 13.8 ± 4.7 mm, respectively, and we also found that these OCDs were significantly different from one another in neutral, flexion, and extension positions. Unfortunately, correlations between OCD with height, weight, and BMI have not been reported in previous studies. In this study, there was a weak but significant correlation between OCD and height in neutral, flexion, and extension positions, but there was no significant correlation of OCD with weight and BMI. We measured the occiput-cervical distance via the OC4D, and the mean neutral OC4D was found to be 69.0 ± 6.9 mm. Importantly, this neutral OC4D value was not significantly different from those measured in flexion (68.9 ± 6.8 mm) or extension (68.1 ± 6.9 mm). Seong et al. found that the posterior border of C4 serves as a landmark in the apex of cervical lordosis, and that it is therefore the least affected by the cervical curve . We hypothesized that the C4 vertebral body, being the central point of the cervical sequence, is the least affected by motion of the cervical position. Hence, the shortest distance from the center of the C4 vertebral body to the McGregor’s line in each cervical position can be regarded as the radius of a circle positioned at the center of the C4 vertebral body and tangent to the McGregor’s line (Fig. 3). Additionally, in our present study, we found significant positive correlations between OC4D and height, as well as between OC4D and weight, among which OC4D had a stronger correlation with height compared to that with weight. In contrast, the correlation between OC4D and BMI was weak and was not statistically significant. And we also found there was no significant correlation between OC4Ds and OCDs in this study. On the one hand, it is due to the difference of measurement methods. On the other hand, it is more important that variation of C2 spinous process has great influence on the individual difference of measurement results .
We found that our novel OC4D measurement has unique advantages compared with those of the OCD in previously reported studies [4, 5, 13]. First, we found that the OC4D was a more accurate parameter compared to OCD in our present study. Additionally, in terms of the OCD, previous studies have demonstrated significant inter-individual morphologic variation in the C2 spinous process (including gender differences). Jiang et al. found that variations in the C2 spinous process may affect the OCD value, and that there was a significant difference in OCD values between male and female subjects . Additionally, the inter- and intra-observer reliabilities of OCDs had ICC values of only 0.651 and 0.754 in a previous study . In the present study, the ICC values of inter- and intra-observer reliabilities for OCDs were moderate to good based on evaluation using standard conventions (see Methods section). We found that the posterior margin of the hard palate, occipital bone, and C4 vertebra (with less bone variation) were clear on lateral radiographs. The ICC values of inter- and intra-observer reliabilities for OC4Ds were more than 0.93 in neutral, flexion, and extension positions, which were significantly higher than those for OCDs. Second, we found that the OC4D was less affected by different positions of the head and neck in neutral, flexion, or extension positions. The alignment of the subaxial spine can influence the occipitocervical alignment required to ensure a functional position of the occiput. However, this variable was not specifically measured in the current study. At present, only a few studies have reported OCD measurements and have shown that neutral OCDs are significantly different from those in flexion and extension positions [4, 5, 10]. In contrast, in our present study, there was no significant difference in the OC4Ds among neutral, flexion, and extension positions (whereas there was for OCDs). This finding may have clinical significance for the use of the OC4D in guiding reduction during operations when the occiput-cervical region is not in a neutral position. Third, the OC4D is not occluded by implants and may therefore represent a valuable intraoperative tool for designing of fusion implants and testing of restoration in the operating room. However, there are no reports showing that the C2 spinous process can be occluded by fixed implants during OCF and that the implants could affect OCD measurements (Fig. 4). Previous literature has stated that it may be difficult to visualize the tip of the dens on radiographs, or that the dens may be absent or fixed in an abnormal position in many conditions under which OCF is performed [14, 15]. Therefore, it may be difficult and inaccurate to evaluate vertical reduction of the occipitocervical region by the distance from the odontoid tip to the McGregor’s line during surgery. Wang et al. first described lower cranial nerve palsy following vertical over-distraction after OCF in four patients who had atlantoaxial dislocation with or without basilar invagination, and the symptoms of all patients were alleviated to different extents by releasing the screw cap and recovery to partial reduction of the occipitoatlantal anatomy . However, our novel OC4D method avoids the occlusion caused by implants and the uncertainty of bony landmarks on radiographs, and therefore may represent a useful tool for estimating and testing the restoration of occipitoatlantal anatomy via regulation of fixed implants.
Limitations of the present study included the demographic data not being matched for age, as well as our sample size being relatively small. In spite of these limitations, our study presented a new method for measurement of the occipital-cervical distance, which may have practical valuable for guiding and testing the restoration condition of the occipital-cervical region. Another limitation of this study is that only cervical spinal radiographs were analyzed, as there were no data regarding the overall sagittal alignment of the spine. Although previous studies have reported that cervical curvature can be affected by overall spinal sagittal imbalance [16,17,18,19], only normal subjects with a normal cervical curvature were included in our present study, and we found no difference in OC4Ds as a function of changes in cervical curvature in neutral, flexion, and extension positions. However, we also recognize that cervical curvature changes can accelerate cervical degeneration, which may affect the results of OC4D measurements. Thus, future studies are needed to obtain more reliable measurements regarding cervical and overall spine sagittal alignment parameters, and to further explore the effect of spinal sagittal parameters on the OC4D. In addition, this present study did not provide the OC4D in patients with craniocervical joint instabilities as a clinically relevant comparator. Hence, we will measure the OC4D values of patients with craniocervical malformation as well as analyze the effect of OC4D fixation selection on the clinical efficacy and patient complications during OCF in future research.