Radiographic involvement of cervical facet joints in ankylosing spondylitis: a longitudinal analysis in correlation with vertebral body lesions

Abstract

Background: The inability to assess structural changes in facet joints is a limitation of established radiographic scoring systems for ankylosing spondylitis (AS). We compared radiographic evidence of ankylosis in cervical facet joints and cervical vertebral bodies in patients with AS.

Methods: We analysed longitudinal data collected from 1106 AS patients and assessed 4984 spinal radiographs obtained up to 16 years of follow-up. Comparisons between cervical facet joints and cervical vertebral bodies focused on the presence of ankylosis, which was defined by at least one facet joint exhibiting complete ankylosis (according to the method of de Vlam) or at least one vertebral body with a bridging syndesmophyte (according to the modified Stoke Ankylosing Spondylitis Spinal Score [mSASSS]). Ankylosis was assessed over time using spinal radiographs collected during follow-up periods stratified in 4-year increments.

Results: Patients with cervical facet joint ankylosis had higher cervical mSASSS, sacroiliitis grades, and inflammatory markers, with more prevalent hip involvement and uveitis. Overall, the numbers of spinal radiographs indicating ankylosis were comparable between cervical facet joints (17.8%) and cervical vertebral bodies (16.8%), and they usually presented together (13.5%). We observed similar proportions of radiographs with ankylosis only in cervical facet joints (4.3%) and cervical vertebral bodies (3.3%). As damage progressed, configurations with both cervical facet joint ankylosis and bridging syndesmophytes became more predominant with longer follow-up times, while configurations with cervical facet joint ankylosis only or bridging syndesmophytes only were less frequently observed.

Conclusions: Evidence of cervical facet joint ankylosis appears as often as bridging syndesmophytes on routine AS spinal radiographs. Presence of cervical facet joint ankylosis should be considered because it may have a higher disease burden.

Keywords: Ankylosing spondylitis; Facet joint; Radiography; Syndesmophyte.

Fig. 1

A lateral radiograph of the cervical spine obtained from a 39-year-old ankylosing spondylitis patient with a disease duration of 18 years. At C3–C4 level, cervical facet joint with complete ankylosis is seen (arrowhead, a score of 3 according to de Vlam et al. [14]). Partial ankylosis and joint space narrowing is seen at C2–C3, C4–C5, and C5–C6 (white arrows, a score of 2) and C6–C7 level (dashed arrow, a score of 1), respectively. On the anterior side of the spine, bridging syndesmophtyes are seen

Fig. 2

Flowchart of included ankylosing spondylitis patients with spinal radiographic data. Several spinal radiographs within the given follow-up period were included per patient. (^aSpinal radiographs were collected during any time point within the corresponding follow-up period. ^bCervical facet joints that were not visible on lateral radiographs or could not be evaluated due to spinal surgery)

Fig. 3

Radiographic damage scores over time at the patient level. Each line represents the progression in cervical mSASSS or cervical facet joint score for each respective patient. (A) Trends in cervical facet joint score over time as assessed among patients with or without at least one bridging syndesmophyte that occurred during any follow-up period. (B) Trends in cervical mSASSS over time as assessed among patients with or without at least one facet joint ankylosis that occurred during any follow-up period mSASSS, modified Stoke Ankylosing Spondylitis Spine Score

Fig. 4

Configuration of cervical facet joint ankylosis and bridging syndesmophytes as assessed during the entire or stratified follow-up periods based on (A) total spinal radiographs and (B) spinal radiographs in the presence of either cervical facet joint ankylosis or bridging syndesmophytes. Each bar shows the frequency of the presence and/or absence of at least one ankylosis in facet joints or vertebral bodies at a given time period for the C2–C3 to C6–C7 vertebral level Y, follow-up year

Fig. 5

Distribution of cervical facet joint ankylosis and bridging syndesmophytes per individual vertebral level as assessed during the entire or stratified follow-up periods. Each bar shows the frequency of facet joint ankylosis or bridging syndesmophytes at a single vertebral level for a given time period Y, follow-up year