SPINART study: validation of the extended magnetic resonance imaging scale for evaluation of joint status in adult patients with severe haemophilia A using baseline data

Abstract

Most previous joint imaging scales for haemophilia have focused on earlier disease stages observed in younger patients. We sought to demonstrate that the 45-item extended magnetic resonance imaging (eMRI) scale is a valid instrument for measuring joint status in adults with severe haemophilia A. Six scale categories (effusion/haemarthrosis, synovial hypertrophy, hemosiderin, erosion, subchondral cysts and cartilage loss) in two domains (soft tissue [range = 0-9] and osteochondral [range = 0-36]) were evaluated for each joint. eMRI scores were derived using baseline data from a randomized, controlled, parallel-group clinical trial (SPINART). Quantitative analysis of linearity included assessments of the relationship between scores derived from the eMRI scale vs. the 17-point International Prophylaxis Study Group (IPSG) MRI scale and Colorado Adult Joint Assessment Scale (CAJAS). Patient eMRI scores correlated with age (r = 0.58), consistent with the expectation of arthropathy progression and more severe joint changes in older patients. eMRI scores demonstrated excellent between-reader agreement for overall patient score (intraclass correlation coefficient [ICC] = 0.88) and outstanding agreement for knee evaluations (ICC = 0.95). There was a strong linear relationship of the eMRI score with the CAJAS (r = 0.7). Adding individual bone evaluations for each joint increased the sensitivity of the instrument to detect change at low scores. There was minimal ceiling effect (5% maximum scores for all evaluated joints) compared with the IPSG MRI scale (20%). Internal reliability was good (overall Cronbach's alpha = 0.75). Overall, the eMRI scale represents a valid and reliable measure of joint status in adolescents and adults with severe haemophilia A.