CT-based comprehensive scoring system for assessing hip arthropathy in ankylosing spondylitis: feasibility study with clinical correlation

Abstract

Abstract Objective To develop a CT-based scoring system for assessment of hip arthropathy in AS. Methods All AS patients were prospectively recruited, consented and underwent whole-body stereoradiographs and pelvis CT, which were assessed by two independent radiologists. Stereoradiographs were assessed according to Kellgreen-Lawrence and BASRI-h. For the Hip arthropathy CT score in AS (HACTSAS), joints were divided into 7 segments and scored for joint space, osteophytes, subchondral cysts/erosions. Patients were clinically assessed for range of motion (ROM), pain and clinical scores (BASMI, BASFI, ASQol, BASDAI and ASDAS). Radiological scores correlations with clinical parameters were compared. ROM sensitivity and specificity for hip arthropathy (BASRI-h ≥ 2) were calculated. Results Sample included 112 patients, with 36/112 females and 76/112 males. Average age was 51.0 ± 11.2 years and mean duration of AS was 20.9 ± 9.6 years. Intraclass correlation coefficients (ICC) for HACTSAS, Kellgreen-Lawrence and BASRI-h were 0.89, 0.89 and 0.82, respectively. HACTSAS showed moderate absolute correlation with ROM (ρ = −0.41) and BASMI (ρ = 0.45), and weak with pain (ρ = 0.18) and BASFI (ρ = 0.25). BASRI-h and Kellgreen-Lawrence exhibited moderate correlation with ROM (ρ = −0.44 and ρ = −0.40, respectively), weak with pain (ρ = −0.27and ρ = −0.23, respectively) and BASFI (ρ = −0.16 and ρ = −0.18, respectively), but only weak with BASMI (ρ = −0.34 and ρ = −0.36, respectively). Internal rotation <15°, abduction <31° and intermalleolar distance <75cm were, respectively, 73%, 70% and 73% sensitivity and 81%, 65% and 68% specific for hip arthropathy. Conclusion HACTSAS exhibited higher correlation with BASMI and BASFI when compared with BASRI-h, but less correlation with pain and ROM. Internal rotation was the best clinical discriminator for hip arthropathy.