Image Assessment Of Atlantoaxial Joint And Its Correlation With Clinical Features In Patients With Craniocervical Junction Lesions

Background:The craniocervical region is a complicated anatomical complex.The atlantoaxial joint is the most mobilizable segment of the spine and is one of the most important joints in the region.The changes of the skeletal structure of this joint caused by upper cervical lesions can lead to the structural compression of the neural structures within the spinal canal,leading to death or disability.Currently most recognized indexes includeatlanto-dentalinterval,space availablefor cord,etc.Because of the differences of race,gender,age and other confounding factors,results of these indexes are not well matched,showing a great variation.Therefore,it needs to be some newindexes to weaken or eliminate the impact of confounding factors,which shall have a smaller variation in normal population and be suitable for comparison,and can be suggestive for patients and the clinical symptoms and the choice of operation method.Objectives:(1)To eliminate the influence of the confounding factors such as age and gender on the direct measurement parameters,two new imaging parameters were introduced-relative spaceavailableforcordand relativeatlas height.To define the normal boundaries of two indexes.(2)To examine the correlation between these new parametersand the incidence of clinical symptoms.(3)To characterize the imaging patterns of unstable and malformation group during preoperative,postoperative and follow-up using these two new parameters.To evaluate the improvement rate of different surgical methods with these parameters.Methods:(1)Data of 102 cases with no obvious upper cervicalanomaly was collected from the image database system.Clivus-canal angle,atlanto-dentalinterval,space availablefor cord,anteroposterior diameterofforamen magnumandatlas,vertical height of atlas and axis,bony height of posterior cranial fossa were measured from neutral plate of cervical CT scan.Bony width and anteroposterior interval of posterior cranial fossa were measured from horizontal transverse section of petrous bone in cervical CT scan.Relative spaceavailableforcordand relativeatlas height were calculated with above parameters.Normal boundaries of the relative space available for spinal cord and the atlas relative height was calculated.(2)A total of 88 cases of upper cervical anomaly between April 2014 and April 2017were collected.Clivus-canal angle,atlanto-dentalinterval,space availablefor cord,anteroposterior diameterofforamen magnumandatlas,vertical height of atlas and axis,bony height of posterior cranial fossa were measured from neutral plate of cervical CT scan.Bony width and anteroposterior interval of posterior cranial fossa were measured from horizontal transverse section of petrous bone in cervical CT scan.Relative spaceavailableforcordand relativeatlas height were calculated with above parameters.The relevance of clinical symptoms and imaging changes was explored.(3)All 88 cases of upper cervical anomaly between April 2014 and April 2017were divided into unstable group and malformation group according to the diagnosis.Imaging patterns of these two groups were characterized using relative spaceavailableforcordand relativeatlas height.The difference of pre-and post-operative imaging and clinical symptoms of patients was also explored with above imaging parameters.The patients were followed up and compared between the posterior fixation and transoral release with posterior fixation.An ROC analysis was performed to define the cut-off value for surgical methods suggestion.Results:(1)The relative space available for spinal cord and the relative height of the atlas could decrease the influence of gender and other confounding factors,and the normal lower limit is 0.50 and 0.61 respectively.(2)In patients' population,relative space available for spinal cord was 0.39 ± 0.12 and relative height of the atlas was 0.53±0.15,presenting no relevance with gender and age.The incidence of occipital and neck pain in patients with relative space available for spinal cordsevere change was significantly lower than normal group and mild change group(P = 0.011,P = 0.003),incidence of upper limb paresthesia in severe change group was significantly higher than normal group(P = 0.004).The incidence of occipital and neck pain in patients withrelative height of the atlas severe change was significantly lower than normal group and mild change group(P = 0.007,P = 0.001).(3)In unstable group,the relative space available for spinal cord was 0.40 ±0.13 and relative height of the atlas was 0.61 ± 0.10,significantly smaller than that of healthy group(P=0.0000 P=0.000).In malformation group therelative space available for spinal cord was 0.38 ± 0.11 and relative height of the atlas was 0.49 ± 0.15,significantly smaller than that of healthy group(P=0.000,P=0.000).The relative height of the atlas of malformation group was still smaller than that of unstable group(P=0.007).The incidence of neck limitation of motion in malformation group was lower than that in unstable group(P=0.050),and the incidence of upper limb paresthesiain malformation group was higher than that in unstable group(P=0.007).Transoral release with posterior fixation groupshowed a vertical imaging recovery rate of 0.64(0.44,0.44),significantly better than posterior fixation0.27(0.06,0.37)(P=0.000).The horizontal imaging recovery rate of transoral release with posterior fixation group was 0.29(0.18,0.46)similar to posterior fixation group 0.27(0.15,0.37).The cut-off value of above parameters was 0.35and 0.53respectively.Conclusions:(1)The relative space available for spinal cord and the relative height of the atlas were able to decrease the influence of gender and other confounding factors,accomplishing the goal of design,and the normal lower limit was 0.50 and 0.61 respectively.(2)The incidence of occipital and neck painwas higher in patients with mild change in relative space available for spinal cord and relative height of the atlas.Theincidence of upper limb paresthesia wasnegatively correlated with relative space available for spinal cordand potentially negatively correlated relative height of the atlas.(3)Both unstable and malformation groups showed smaller relative space available for spinal cord and relative height of the atlas.But malformation group had a much smaller relative height of the atlas.This difference could not be noted with traditional imaging parameters such as altano-dental interval but only in these two new parameters.The patients who received the transoral release with posterior fixation surgery showed a better relativeatlas height recovery rate.Changes in relative spaceavailableforcord and relativeatlas height were parallel to changes of clinical symptoms,suggesting these two parameters reflect surgical effects.We recommended patients with relative spaceavailableforcord<0.35 or relativeatlas height<0.53to receive transoral release with posterior fixation in order to obtain a better imaging recovery and steady in follow-up.