Abstract
Objective To investigate the prevalence of facet joint ankylosis in the whole spine in axial spondyloarthritis (axSpA) using low-dose computed tomography (LDCT), and to identify factors associated with facet joint ankylosis.
Methods Whole spine LDCT images from 161 patients with axSpA were examined, and the presence of facet joint ankylosis was assessed (right and left, C2–S1) by 2 readers. Facet joint ankylosis was scored from 0 to 46. Structural damage of vertebral body was assessed using CT Syndesmophyte Score (CTSS). Factors associated with ankylosed facet joint scores for the whole spine were identified using a generalized linear model with a negative binomial distribution.
Results Seventy-nine patients (49%) and 70 patients (43%; reader 1 and reader 2, respectively) had ≥ 1 ankylosed facet joint. Facet joint ankylosis was most common in the thoracic spine. The mean score of facet joint ankylosis for the whole spine was 6.6 (SD 11.2) in reader 1 and 4.2 (SD 8.4) in reader 2. Whole spine facet joint ankylosis score positively correlated with Ankylosing Spondylitis Disease Activity Score (ASDAS) and CTSS. In multivariable analysis, the ankylosed facet joint score was associated with ASDAS, sacroiliitis, CTSS, and a history of uveitis in both readers. Uveitis history, ASDAS, and CTSS were associated with whole spine facet joint ankylosis score in subgroup analysis of only radiographic axSpA.
Conclusion The prevalence of ankylosed facet joints is high in axSpA, especially in the thoracic segment. The whole spine ankylosed facet joint score is significantly associated with a history of uveitis, ASDAS, sacroiliitis, and syndesmophyte score.
Axial spondyloarthritis (axSpA), which includes radiographic axSpA (r-axSpA; also known as ankylosing spondylitis [AS]) and nonradiographic axSpA (nr-axSpA), is a chronic inflammatory disease characterized by pain, functional and structural impairment, reduced mobility, and potential deformity of the axial skeleton.1 Chronic inflammation in axSpA can result in erosive destruction of the bones and joints as well as new bone formation, both of which lead to ankylosis in both the anterior and posterior parts of the vertebrae.2 This structural damage causes irreversible impairment of spinal mobility and contributes to the disease burden independently of inflammation-related pain and stiffness.3
Ossification at the annulus fibrosis in the periphery of the intervertebral disc in the form of syndesmophytes is a typical finding on the anterior aspect of the vertebrae.2 In addition, facet joints (also known as zygapophyseal joints) on the posterior aspect of the vertebrae are involved directly in the inflammatory process of axSpA3; ossification of facet joints fuses the posterior aspect of the vertebra. Ankylosed facet joints, in addition to syndesmophytes, are the predominant type of structural damage in patients with axSpA4,5 and are a crucial component that affects spinal range of motion.5,6
The 2015 European Alliance of Associations for Rheumatology guidelines for the use of imaging in SpA in clinical practice recommended that conventional radiography may be used for monitoring structural damage caused by axSpA.7 Syndesmophytes or bridging of the vertebral bodies can be assessed by conventional radiography. However, the curved structures of facet joints throw overlapping shadows on radiographs, making assessment difficult. As such, structural damage at the posterior aspect of vertebrae, including the facet joints, is usually overlooked during assessment by conventional radiography. Until now, the most validated method for assessing radiographic damage is the modified Stoke Ankylosing Spondylitis Spinal Score (mSASSS), which focused on structural change at the anterior aspect of the vertebrae2; however, this score does not include the facet joints.
Computed tomography (CT) provides clear images that enable assessment of bone proliferation, including at the posterior elements of the vertebrae that are difficult to assess on radiographs. Since CT examination detects bony erosion or ankylosis, it is the ideal method of assessing ankylosed facet joints.8 As such, low-dose CT (LDCT) is being increasingly used to assess whole spinal structural damage in axSpA.9-11
A few studies using CT show that facet joint ankylosis not only correlates strongly with impaired spinal mobility5,6 but also is independently associated with functional impairment after adjusting for the syndesmophyte score and disease activity.5 Despite the known association between ankylosed facet joints and clinical outcome, the prevalence of facet joint ankylosis on the whole spine in axSpA and the associated risk factors are unknown.
Therefore, the objectives of this study were to analyze the prevalence of facet joint ankylosis in the whole spine among patients with axSpA using LDCT and to identify factors associated with facet joint ankylosis.
METHODS
Study patients. The present analysis was based on consecutive patients with axSpA from the Incheon Saint Mary’s Axial Spondyloarthritis study (ISAXSPA), which is an ongoing, prospective, observational cohort study focused on identifying predictors of new bone formation and bone loss in the spine of patients with axSpA.12 All patients were aged ≥ 18 years and fulfilled the 2009 Assessment of SpondyloArthritis international Society criteria for axSpA,13 including patients with established AS.14
Whole spinal LDCT from patients enrolled in the ISAXSPA cohort were assessed. The present study analyzed 161 consecutive patients, all of whom underwent lateral radiographs of the cervical and lumbar spine within 3 months of LDCT. The study was performed in accordance with the Declaration of Helsinki. All study participants provided written informed consent prior to inclusion in the observational cohort. The study protocol was approved by the local ethics committee (study no. OC16OISI0138).
Clinical data. According to the ISAXSPA study protocol, the following baseline characteristics were collected: sex, age, time after symptom onset, HLA-B27 status, BMI, current smoking status, history of uveitis, and presence of peripheral arthritis. Use of medications and disease activity measures were investigated at the same time as examination of LDCT. The erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) levels were measured. Measures of disease activity were collected. The Ankylosing Spondylitis Disease Activity Score (ASDAS) were calculated as described previously.15 The Bath Ankylosing Spondylitis Functional Index (BASFI) was also assessed.16
Radiographic sacroiliitis on anterior-posterior views of the pelvis was defined according to the modified New York criteria.14 Radiographs were scored by a single trained expert who was blinded to the patient’s characteristics.
CT image acquisition and reconstruction. Whole spine LDCT images were acquired by a 320-detector CT scanner (Aquilion One, Canon Medical Systems). Patients were placed in the machine in a supine position, feet first. A volumetric CT scan was acquired from the superior endplate of C2 down to the inferior endplate of S1 using a multidetector scanner with the following parameters: tube voltage = 120 kVp; tube current = 60 mA, without automated tube current modulation; detector collimation = 100 × 0.5 mm; rotation time = 0.5 s; and pitch = 0.96. Subsequently, reconstructions were performed in 3 orthogonal planes, with 2-mm thick slices, using FC03 and FC36 convolutional kernels for soft tissue and bone, respectively. The CT dose index volume was 3.1 mGy, and the approximate effective radiation dose per LDCT was approximately 4 mSv.
CT scoring procedure. Images in the axial and sagittal planes were reformatted into 2-mm slices. Facet joint ankylosis was assessed independently by 2 readers (1 rheumatologist and 1 musculoskeletal radiologist) who were blinded to the patient’s characteristics. Bilateral facet joints from C2 to S1 were assessed for the presence of bony ankyloses. Ankylosis of a facet joint was defined as a bridge of bone connecting adjacent superior and inferior articular processes. The nonankylosed or ankylosed status of the left and right facet joints from C2 to S1 was determined. If at least 1 part of a joint was ankylosed, it was scored as ankylosed. CT scores were dichotomized as 0 (no ankylosis) or 1 (ankylosed).11 Joints from C2 to S1 were scored from 0 to 2 (total, 0-46).
Syndesmophytes were assessed in the sagittal and coronal planes using the CT Syndesmophyte Score (CTSS).9 The anterior and posterior quadrants from the bottom half of C2 to the top half of S1 were scored (23 vertebral units [VU] in total); 8 quadrants were scored per VU. The score assesses the height of a syndesmophyte relative to the intervertebral disc space (IDS). For each quadrant, the score was 0 if no syndesmophyte was present, 1 if a syndesmophyte was present but did not reach 50% of the IDS, 2 if the syndesmophyte reached or crossed 50% of the IDS, and 3 if the syndesmophyte bridged the IDS. The CTSS was scored by a single trained expert who was blinded to the patient’s characteristics. The total score ranged from 0 to 552.
Statistical analysis. Continuous data are expressed as the mean (SD) and categorical data are expressed as n (%). Clinical variables for patients with and without facet joint ankylosis were compared using a chi-square test (categorical variables) or a paired t test (continuous data).
The interobserver reliability with respect to scoring of ankylosed facet joints was calculated using ANOVA to yield an intraclass correlation coefficient (ICC) based on total summed scores for the whole spine and the individual cervical, thoracic, and lumbar segments. A 2-way random model for absolute agreement was used.
Spearman correlation coefficient was used to analyze correlations between ankylosed facet joint scores and clinical variables.
Ankylosed facet joint scores for the whole spine are typically nonnormally distributed, with a skew toward the right. Therefore, a generalized linear model with a negative binomial distribution was used to identify factors associated with the ankylosed facet joint score. All variables that were significant in the univariable negative binomial regression analyses (P < 0.05) were incorporated as explanatory variables into the multivariable analyses. The regression coefficients (β) and their 95% CI were calculated using the negative binomial regression models.
All statistical tests were 2-tailed and statistical significance was defined as P < 0.05. Statistical analyses were performed using PASW statistics 22 (SPSS).
RESULTS
Table 1 presents the demographic and clinical characteristics of the 161 patients. The mean age was 35 (SD 10) years. One hundred thirty-two (82%) were male. One hundred twenty-four (77%) had radiographic sacroiliitis that fulfilled the AS classification criteria.14
Characteristics of patients with axSpA.
The ICCs for scores from the 2 readers were 0.85, 0.89, 0.93 and 0.92 for the cervical, thoracic, and lumbar segments, and whole spine, respectively. Seventy-nine patients (49%) in reader 1 and 70 patients (43%) in reader 2 had ankylosis of more than 1 facet joint in the whole spine (Table 2). The prevalence of facet joint ankylosis in each spinal segment was as follows: cervical, 25/161 (16%); thoracic, 72/161 (45%); and lumbar, 39/161 (24%). The mean ankylosed facet joint scores were 1.0 (SD 2.6), 4.2 (SD 6.8), 1.5 (SD 3.2), and 6.6 (SD 11.2) at the cervical, thoracic, and lumbar segments, and whole spine, respectively.
Prevalence of facet joint ankylosis in patients with axSpA (N = 161).
Supplementary Table S1 (available with the online version of this article) using data from reader 1 and reader 2 shows a comparison of clinical characteristics between patients with facet joint ankylosis and those without. In the data from reader 1 only, patients with ankylosed facet joints were older and had longer symptom duration than those without. Also, HLA-B27 positivity and BASFI scores were higher in patients with facet joint ankylosis only in data from reader 1. In data from both reader 1 and reader 2, patients with ankylosed facet joints were more likely to have a history of uveitis and higher disease activity scores (ASDAS) than those without. Additionally, the proportion of increased inflammatory markers (ESR and CRP) were significantly higher in patients with ankylosed facet joints. Radiographic sacroiliitis was more common in patients with facet joint ankylosis. Patients with facet joint ankylosis had more syndesmophytes than those without. However, there was no significant difference between the groups with respect to current medication.
We used the r coefficients for the correlation between clinical variables and the ankylosed facet joint scores for the whole spine. In both readers, ankylosed facet joint scores correlated positively with symptom duration. Whole spine scores correlated significantly with the ASDAS and CTSS (P < 0.001 in both readers; data not shown).
Table 3 shows the results of univariable and multivariable analysis for identifying factors associated with ankylosed facet joint scores for the whole spine in each reader. Univariable analysis for reader 1 identified symptom duration, male sex, HLA-B27, history of uveitis, ASDAS, the presence of sacroiliitis on radiographs, and CTSS as being associated with facet joint ankylosis scores. Multivariable analysis revealed that the facet joint ankylosis score was independently associated with male sex, HLA-B27, a history of uveitis, ASDAS, radiographic sacroiliitis, and CTSS. Symptom duration, male sex, HLA-B27, smoking, history of uveitis, ASDAS, the presence of sacroiliitis on radiographs, and CTSS also showed significance in the univariable analysis for reader 2. In multivariable analysis for reader 2, uveitis, ASDAS, radiographic sacroiliitis, and CTSS were independently associated with facet joint ankylosis scores.
Negative binomial regression analysis facet joint ankylosis scores for the whole spine in patients with axSpA (N = 161).
Facet joint changes might be significantly more prevalent in r-axSpA.17 Therefore, we performed subgroup analysis of factors associated with the presence of facet joint ankylosis in 124 patients with r-axSpA (Table 4). Multivariable analysis for reader 1 revealed that among patients with r-axSpA, male sex, a history of uveitis, ASDAS, and CTSS were independently associated with the presence of facet joint ankylosis. In multivariable analysis for reader 2, a history of uveitis, ASDAS, and CTSS were significantly associated factors.
Regression analysis facet joint ankylosis scores for the whole spine in patients with r-axSpA (N = 124).
DISCUSSION
The present study using whole spine LDCT showed that approximately half of patients with axSpA had 1 or more ankylosed facet joint. Factors associated with the scores of facet joint ankylosis were a history of uveitis, high disease activity, radiographic sacroiliitis, and syndesmophyte scores.
Radiographic assessment of spinal structural damage has focused mainly on the anterior spine, so structures at the posterior spine (including facet joints) are often overlooked. This study shows that facet joint ankylosis is highly prevalent in all segments of the spine.
Facet joints, the part of the synovial joints in the spine, comprise hyaline articular cartilage overlying subchondral bone, a synovial membrane, and a joint capsule.18 The joints connect the superior and inferior articular processes of the vertebrae. Inflammatory involvement of the facet joints leads to bony erosion, joint space narrowing, and ankylosis.19,20 Earlier histomorphometric studies in axSpA show that facet joint ankylosis is initiated by fusion of both cartilaginous surfaces, followed by bony intraarticular ankylosis.19,21 Ankylosis of facet joints impairs normal joint function and compromises spinal mobility. Our previous study confirmed the effect of facet joint ankylosis on spinal mobility, independent of syndesmophyte number. Additionally, we found that facet joint ankylosis is related with functional impairment, after adjusting for disease activity and syndesmophyte score.5 Therefore, it is important to identify factors associated with facet joint ankylosis and prevent further progression of ankylosis in patients with axSpA.
To date, a few radiography-based studies have reported the prevalence of facet joint ankylosis but investigated only the cervical and/or lumbar spine.20,22 LDCT provides improved visualization of facet joints and allows more accurate readings, with relatively less radiation exposure, than conventional CT. A recent study using LDCT analyzed the presence of facet joint ankylosis in r-axSpA with 1 or more syndesmophytes on radiograph, but the number of patients was small (N = 60) and the study did not identify clinical factors related to facet joint ankylosis.11 The present study using LDCT investigated the prevalence of facet joint ankylosis in the whole spine among a large number of patients and included patients (60%) without syndesmophytes on radiograph, as well as those with syndesmophytes. Our results of a high prevalence of ankylosed facet joints, especially in the thoracic segment (approximately half of our cohort), is similar to the findings of studies based on CT analysis.6,11
We also identified factors associated with ankylosed facet joints in axSpA, including nr-axSpA. We found that structural damages, such as the presence of radiographic sacroiliitis and CTSS, were independently associated with facet joint ankylosis. Of 37 patients with nr-axSpA, 5 (14%) had facet joint ankylosis, compared with 60% of patients with r-axSpA (data not shown). Radiographic sacroiliitis is associated with more severe and advanced disease, which is caused by persistent inflammation of the axial joints, including the facet joints. The current study showed the presence of radiographic sacroiliitis is significantly associated with the score of facet joint ankylosis.
Previous studies report an association between facet joint ankylosis and syndesmophytes.6,20,22 The results of previous studies are inconsistent with respect to the order in which facet joint ankylosis and bridging syndesmophytes develop, but there is a definite relationship between them. Our results were more consistent in that syndesmophytes were strongly associated with facet joint ankylosis. Syndesmophyte score assessed using CT showed a positive association with the presence of facet joint ankylosis as well as the total score of facet joint ankylosis in both readers. A longitudinal analysis of the cervical spine shows that baseline syndesmophytes are indeed associated with radiographic progression of vertebral bodies, but not with progression of facet joint disease.22 On the contrary, in a recent study of 51 patients with r-axSpA, it was confirmed that the presence of syndesmophytes and facet joint ankylosis at baseline affects each other’s occurrence after 2 years.23 Thus, the presence of syndesmophytes appears to be related to the prevalence of facet joint ankylosis as well as to be a predictor of future development of facet joint ankylosis.
The present study also showed the association between disease activity and facet joint ankylosis. The results are consistent with those of a previous study.22 As is the case for structural damage to the vertebral body,24,25 high disease activity is independently associated with whole spine score of facet joint ankylosis. According to a study of facet joint ankylosis using CT in r-axSpA (with or without syndesmophytes) and patients with nr-axSpA, facet joint ankylosis was more prevalent in r-axSpA with syndesmophytes.17 Similarly, in our subgroup analysis of 124 patients with r-axSpA, disease activity was associated with facet joint ankylosis scores. These findings suggest that strict control of inflammation may reduce future facet joint ankylosis in young patients with high disease activity.
Interestingly, we found that a history of uveitis was associated with the ankylosed facet joint score for the whole spine. This finding is consistent with those of earlier studies.20,22 One of these previous studies showed that patients with AS with facet joints ankylosis in the cervical spine are more likely to have a history of uveitis.22 Also, a cross-sectional study of 50 patients with AS showed that 37% of patients with facet ankylosis had a history of uveitis.20 Although the proportion of patients with a history of uveitis increases with age and disease duration,26 we found that a history of uveitis showed a significant association with facet joint ankylosis after adjusting for confounding factors, including symptom duration. In a recent retrospective study comparing the clinical features of patients with SpA with and without uveitis, a longer disease duration and more common radiological damage (sacroiliitis, syndesmophytes, and bamboo spine) were shown in patients with uveitis.27 Considering this, our results showing the association between uveitis and facet joint ankylosis might be because uveitis is more common in patients with advanced and severe disease rather than the direct effect of inflammation in facet joint associated with uveitis.
This study has several limitations. First, it was a cross-sectional study; thus, although we could assess associations, we could not determine a cause-and-effect relationship. Predictors of facet joint ankylosis in the whole spine should be investigated in further prospective studies using CT. Second, we assessed facet joints as ankylosed or nonankylosed. Abnormalities such as bony erosion or joint space narrowing were not analyzed. This was due to the limited resolution of LDCT and inconsistencies in the interpretation of joint space narrowing or irregularities, as mentioned in earlier studies.6,11 Third, we did not measure abnormalities in combination with clear degenerative changes. A previous study suggested that facet joint damage is nonspecific for AS since the facet joints can be affected by degenerative changes.28 Facet joint ankylosis on imaging can overlap with degenerative changes due to facet joint osteoarthritis, especially in old age.
In conclusion, assessment of the whole spine using LDCT showed that facet joint ankylosis is common in axSpA. The ankylosed facet joint score for the whole spine is independently associated with a history of uveitis, as well as with disease activity, radiographic sacroiliitis, and CTSS. Strict control of inflammation may reduce future facet joint ankylosis in patients with high disease activity.
Footnotes
The authors declare no conflicts of interest relevant to this article.
- Accepted for publication November 17, 2022.
- Copyright © 2023 by the Journal of Rheumatology
