From the Departments of Radiodiagnosis.
Paediatrics.
J Clin Rheumatol. 2023 Dec 1;29(8):402-407. doi: 10.1097/RHU.0000000000002032. Epub 2023 Oct 2.
To determine the utility of whole-body magnetic resonance imaging (WB MRI) to predict relapse in children with juvenile idiopathic arthritis (JIA) in clinical remission.
Consecutive patients with JIA who fulfilled the Wallace criteria for remission were recruited into this longitudinal pilot study and underwent WB MRI. A radiological score was devised, incorporating synovitis, bone marrow edema, sacroiliitis, enthesitis, and bone erosions. Two readers independently scored the MR data sets. The same score was calculated for both knee joints individually and correlated with outcome for that joint. Score-based models incorporating clinical and laboratory variables were generated. Logistic regression analysis was done to determine predictors for relapse. Receiver operating characteristic curve was drawn for significant variables.
Twenty-two children (median age, 12 years; interquartile range, 9.5-14.25 years) were included in the final analysis. At 24 months' follow-up, 15 joints in 5 children relapsed; knee was the most common site. Seven knee joints had disease relapse. On univariate analysis, synovitis and total score on WB MRI were significant predictors of relapse at follow-up, with odds ratios of 9.46 (bias-corrected 95% confidence interval, 3.07-29.13) and 2.8 (bias-corrected 95% confidence interval, 1.23-6.39) respectively. Two models, which included a higher number of joints involved at presentation and abrupt drug withdrawal strategy as predictor variables, were also statistically significant (odds ratio, approximately 1.9). On multivariate analysis of the predictors variables in models where p < 0.6, it was found that only synovitis score and total score were near statistical significance ( p = 0.06); no clinical or laboratory variables were significant. The areas under the receiver operating characteristic curve for relapse prediction were approximately 0.82, 0.87, 0.79, and 0.81 for synovitis score, total MRI score, and both models, respectively.
Synovitis on WB MRI is the strongest independent predictor for disease relapse in children with JIA in remission.
确定全身磁共振成像（WB MRI）在预测处于缓解期的幼年特发性关节炎（JIA）儿童疾病复发中的作用。
本纵向试点研究纳入了符合 Wallace 缓解标准的连续 JIA 患者，并进行了 WB MRI。制定了一种包含滑膜炎、骨髓水肿、骶髂关节炎、附着点炎和骨侵蚀的放射学评分。两名读者分别独立评分 MRI 数据。分别计算每个膝关节的相同评分，并将其与该关节的结果相关联。生成包含临床和实验室变量的基于评分的模型。进行逻辑回归分析以确定复发的预测因素。绘制有意义变量的受试者工作特征曲线。
最终分析纳入了 22 名儿童（中位年龄 12 岁；四分位距 9.5-14.25 岁）。在 24 个月的随访中，5 名儿童的 15 个关节复发；膝关节是最常见的部位。7 个膝关节疾病复发。单变量分析中，WB MRI 滑膜炎和总评分是随访时疾病复发的显著预测因素，优势比分别为 9.46（偏倚校正 95%置信区间，3.07-29.13）和 2.8（偏倚校正 95%置信区间，1.23-6.39）。还包括预测变量的两个模型，即就诊时受累关节数量较高和突然停药策略，也具有统计学意义（优势比约为 1.9）。在对模型中 p < 0.6 的预测变量进行多变量分析时，发现只有滑膜炎评分和总评分接近统计学意义（p = 0.06）；没有临床或实验室变量有意义。滑膜炎评分、总 MRI 评分和两个模型的预测疾病复发的受试者工作特征曲线下面积分别约为 0.82、0.87、0.79 和 0.81。
缓解期 JIA 儿童 WB MRI 滑膜炎是疾病复发的最强独立预测因素。
