Juvenile Dermatomyositis Magnetic Resonance Imaging Score (JIS) does not correlate with criteria for clinically inactive disease: a single-centre retrospective evaluation

Abstract

AbstractThe Paediatric Rheumatology International Trials Organisation (PRINTO) criteria for clinically inactive disease (CID) and their proposal for glucocorticoid tapering do not consider MRI findings, despite the growing use of MRI and development of reliable MRI scoring tools. We aim to evaluate how CID correlates with MRI scores and physician decision making. We retrospectively used the Juvenile Dermatomyositis Imaging Score (JIS) to score MRIs of all children with JDM over a 10-year period. Demographic, diagnosis, treatment and core set measures data were collected. Correlation between CID and JIS was assessed as well as correlation with the physician treatment decision. There were 25 patients with 59 follow-up episodes to analyse correlation between physician treatment decision and JIS; and 50 episodes for the CID category and JIS correlation. JIS was not significantly associated with the CID category but did correlate with the physician decision. No significant association was found between clinical decision and CID category. The JIS area under the ROC curve (AUC) was 0.80 (95% CI 0.62–0.99) with a score ≥ 8 to predict an escalation. JIS sensitivity and specificity were both 78% with accuracy of 78%, compared to only 67%, 46% and 49%, respectively, for the CID criteria. Clinical criteria alone are not sufficient to assess disease activity status. Clinical decision trends correlated to MRI findings but not PRINTO CID criteria. Multi centre prospective studies are needed to replicate our findings and establish how to best use MRI as a biomarker of disease activity.