Background Radiographic damage in the cervical spine is associated with restricted spinal mobility in patients with ankylosing spondylitits (AS). The most characteristic radiographic changes are syndesmophyte formation and ankylosis of the vertebral bodies. However, the zygapophyseal (ZA) joints are also involved in the disease process.
Objectives To assess reliability of standardized scoring of the cervical ZA joints in AS patients. To compare radiographic damage of the ZA joints with radiographic damage of the vertebral bodies in AS patients with active disease before and after 4 years of TNF-α blocking therapy.
Methods Consecutive AS patients with active disease from the GLAS cohort with available lateral radiographs of the cervical spine at baseline and after 4 years of TNF-α blocking therapy were included. ZA joints of C2-C3 up to C6-C7 were scored according to the method of de Vlam et al. (0=normal, 1=joint space narrowing or erosion, 2=partial blurring or ankylosis, 3=complete blurring or ankylosis). The mSASSS was scored to assess radiographic damage of the vertebral bodies. Two readers were trained and after attaining good reliability radiographs were scored independently blinded to patient characteristics and time sequence. To compare damage of ZA joints and vertebral bodies, ZA joint involvement was defined as ≥1 ZA joint with score ≥1 and vertebral body involvement was defined as ≥1 vertebra with mSASSS ≥1. Linear weighted kappa statistics and percentage absolute agreement were used to analyze the reliability of the ZA joint scoring method.
Results Of the 108 included patients, 76% was male, mean age was 43±11 years, median symptom duration 17 (1-50) years, 84% was HLA-B27 positive, mean BASDAI 5.9±1.7, and mean ASDAS 3.8±0.8. The ZA joint scoring method showed good interobserver reliability with kappa's between 0.80-0.84 and high percentages of agreement between 80%>89%. At baseline, 49 (45%) patients had ZA joint involvement of which 22 (20%) had ankylosis in ≥1 ZA joint. In comparison with the vertebral bodies, 95 (88%) patients had mSASSS ≥1 of which 41 (38%) had mSASSS ≥3 (bridging syndesmophytes). In 4 (4%) patients, ZA joint involvement was present without vertebral body involvement. During 4 years of follow-up, 18 (14%) patients developed new damage in the ZA joints of which 3 (3%) developed ankylosis in ≥1 ZA joint. Development of new damage of vertebral bodies was present in 66 (61%) patients of which 7 (6%) developed bridging syndesmophytes. In 3 (3%) patients, new damage was present in ZA joints but not in vertebral bodies.
Conclusions Scoring radiographic damage of ZA joints in AS patients is reliable. In this observational cohort of AS patients with active disease, damage of vertebral bodies at baseline and during 4 years of TNF-α blocking therapy was more common than damage of the ZA joints. Radiographic damage assessed with scoring the ZA joints contributed in only 3-4% of the patients in addition to the mSASSS.
Acknowledgements The GLAS cohort was supported by an unrestricted grant from Pfizer. Pfizer had no role in the design, conduct, interpretation, or publication of this study.
Disclosure of Interest F. Maas: None declared, S. Arends Grant/research support from: Abbott, Pfizer, Wyeth, E. van der Veer: None declared, F. Wink: None declared, M. Efde: None declared, H. Bootsma: None declared, R. Chaudhry: None declared, E. Brouwer Grant/research support from: Abbott, Pfizer, Wyeth, A. Spoorenberg Grant/research support from: Abbott, Pfizer, Wyeth, Consultant for: Abbvie, Pfizer, UCB