Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine, Baltimore, MD, USA.
Department of Radiology, Boston University School of Medicine, Boston, MA, USA.
Osteoarthritis Cartilage. 2019 Feb;27(2):278-285. doi: 10.1016/j.joca.2018.10.013. Epub 2018 Nov 14.
To determine the association between Insall-Salvati ratio (ISR), a measure of patella alta, and worsening of Magnetic Resonance Imaging (MRI)-based osteoarthritis (OA)-related patellofemoral joint structural damages over 24-month in participants of the Osteoarthritis Initiative (OAI).
Using weighted random sampling method, we selected a sample of 500 knees (from 1,677 knees with available baseline and 24-months MRI OA Knee Score (MOAKS) measurements), which is OAI-representative regarding knee OA-related factors (i.e., baseline age, sex, body mass index (BMI), and radiographic Kellgren-Lawrence grading). The ISR was measured in all enrolled knees using baseline sagittal 3T-MRI plane by three radiologists. Baseline and 24-month MOAKS variables for patellofemoral bone marrow lesions (BMLs), cartilage damages, and osteophytes were extracted, and the associations between ISR and 24-month worsening of these 3T-MRI features were evaluated using multivariable regression models. After computing receiver operating characteristic curves, the optimal cutoff point of ISR for indicating worsening of patellofemoral OA was determined. P-values were adjusted for multiple comparisons and false discovery rate (FDR) adjusted P-values were reported.
In this longitudinal analysis, 24-month worsening of BML (odds ratio [OR] (95% confidence interval [95% CI]):11.18 (3.35-39.6), adjusted-p-value:<0.001) and cartilage scores (OR:7.39 (1.62-34.71), adjusted-p-value:0.042) in lateral patella was associated with higher baseline ISR. However, higher ISR was not statistically associated with medial patellar or medial and lateral trochlear BML or cartilage scores worsening. We determined the optimal cutoff point of ISR≥1.14 (95% CI: 1.083-1.284) for predicting lateral patellofemoral OA-related structural damages worsening over 24-months (sensitivity:73.73%; specificity: 66.67%).
Given the uncertainly surrounding the results, our overall findings suggest that ISR could be considered as a predictor of lateral patellofemoral OA-related structural damages worsening with the optimal cutoff point of ≥1.14 using knee sagittal MRI measurements.
确定 Insall-Salvati 比值(ISR)与 patella alta 之间的关联,以及在 Osteoarthritis Initiative(OAI)参与者中,24 个月时基于磁共振成像(MRI)的骨关节炎(OA)相关髌股关节结构损伤的恶化。
我们使用加权随机抽样方法,从基线和 24 个月的 MRI 膝关节评分(MOAKS)测量值均可用的 1677 个膝关节中选择了 500 个膝关节(即 OAI 中与膝关节 OA 相关因素有关的样本,如基线年龄、性别、体重指数(BMI)和放射学 Kellgren-Lawrence 分级)。使用基线矢状 3T-MRI 平面,由三位放射科医生测量所有纳入膝关节的 ISR。提取基线和 24 个月 MOAKS 变量,包括髌股骨髓病变(BML)、软骨损伤和骨赘,并使用多变量回归模型评估 ISR 与 24 个月时这些 3T-MRI 特征恶化的关系。计算接收者操作特征曲线后,确定 ISR 指示髌股 OA 恶化的最佳截断点。对 P 值进行了多重比较调整,并报告了假发现率(FDR)调整后的 P 值。
在这项纵向分析中,24 个月时外侧髌骨 BML(比值比[OR](95%置信区间[95%CI]):11.18(3.35-39.6),调整后的 P 值:<0.001)和软骨评分(OR:7.39(1.62-34.71),调整后的 P 值:0.042)与较高的基线 ISR 相关。然而,较高的 ISR 与内侧髌骨或内侧和外侧滑车 BML 或软骨评分恶化无统计学相关性。我们确定了 ISR≥1.14(95%CI:1.083-1.284)的最佳截断点,用于预测 24 个月时外侧髌股 OA 相关结构损伤的恶化(敏感性:73.73%;特异性:66.67%)。
鉴于结果存在不确定性,我们的总体研究结果表明,使用膝关节矢状 MRI 测量,ISR 可被视为预测外侧髌股 OA 相关结构损伤恶化的指标,最佳截断点为≥1.14。