The Preliminary Analysis Of Whole-spine MRI In Patients With Axial Spondyloarthritis

Background: Spondyloarthritis (SpA) is a group of connected diseases which havethe clinical characteristics in commom. Axial SpA is a newly defined conceptwhich is mainly characterized by axial manifestation. Magnetic Resonanceimaging(MRI) is the most sensitive imaging technique to detect the earlysacroiliitis by now. The new criteria about the SpA have included the MRI ofsacroiliac joint. The spine is the most common involving site of axial SpA. Butthere was no uniform criteria about the spinal MRI to diagnose the axial SpA. Thisstudy aimed to investigate the spinal MR imaging of axial SpA, explore itscorrelation with clinical findings and evaluate the diagnostic value of spinal MRIlesions in distinguishing the axial SpA patients from chronic mechanical low backpain patients.Methods: In the first part of our study, we established a cohort of33axial SpApatients who meet the the new ASAS criteria.(1) The history was collected by aquestionnaire, the BASDAI and BASFI were conducted.(2)The patients’ physicalexamination were conducted blindly, the BASMI and ASDAS were calculated.(3)The HLA-B27, ESR and CRP were tested.(4) Each patient were undertookwhole-spine MRI in the same MR scanner.(5) The MR imagings were read blindly.To investigate the spinal MR imaging of axial SpA, explore its correlation withclinical findings.In the second part of our study, cohorts of axial SpA and chronic mechanicallow back pain patients were established.(1)52patients meeting the ASAS criteriaabout the axial SpA were included.(2) Each patient were undertook whole-spineMRI in the same MR scanner.(3)52mechanical back pain patients with age andsex matched who underwent cervical spine, thoracic spine and lumbar spine MRI were selected as control group.(4) The MR imagings were read blindly. Thedifferences of spine MRI findings between patients from the two groups werecompared and the diagnostic value of spinal MRI lesions in distinguishing the axialSpA patients from chronic mechanical low back pain patients was evaluated.Results: In the first part of our study, we found that there were statisticallysignificant correlation between acute Romanus lesions, chronic Romanus lesions,chronic Andersson lesions, total number of spinal lesions and BASMI(r=0.440, P<0.05; r=0.483, P <0.05; r=0.421, P <0.05; r=0.589, P <0.05). There werepositive correlation between chronic Andersson lesions and BASFI (r=0.392, P<0.05). But no statistically significant correlation between spinal lesions and otherclinical findings were observed.In the second part of our study, Romanus lesion was more common in SpAgroup than in the control group(vertebral level χ2=132.147,P=0.000; lesion levelχ2=378.613, P=0.001). In patients with SpA, Romanus lesion was frequently seenat the lower thoracic spine, whereas Romanus lesion was commonly seen at thecervical spine in mechanical back pain patients. The sensitivity and positivelikelihood ratio of Romanus lesions were not sufficient to make them being thediagnostic criteria of axial SpA. But the acute Romanus lesions in the lowerthoracic spine show relatively higher diagnostic significance. There were nosignificant difference in the incidence of discitis between the two groups.Conclusions:(1) The MRI findings of spinal involvement in axial SpA includedRomanus lesions, Andersson lesions and Facet arthritis and syndesmophyte.(2)There were statistically significant correlations between acute Romanus lesions,chronic Romanus lesions, chronic Andersson lesions, total number of spinal lesionsand BASMI.(3) The sensitivity and positive likelihood ratio(LR) of Romanuslesions were not sufficient to make them being the diagnostic criteria of axial SpA.