Abstract：
OBJECTIVE: Rheumatoid arthritis can be classified according to ACPA and RF status. ACPA status may be associated with other pathophysiological differences, e.g., the cytokines driving inflammation. Obesity influences the course of RA, likely involving leptin; the exact mechanisms are not completely understood. This study investigates BMI influence on RA cytokine profiles and the possibility of predicting ACPA status and disease activity measured by Power-Doppler sonography (PDS).
METHODS: Patients were examined using a multi-biomarker disease assay and PDS examination of wrists and MCP and PIP joints and stratified according to ACPA status and BMI, using prediction precision to determine BMI cutoff. Analysis was performed using elastic net regularization of logistic and multiple regression. We then attempted to predict ACPA status/PDS activity based on a bootstrap approach.
RESULTS: A total of 120 measurements from 95 patients were performed. ACPA status prediction peaked at BMI 26 kg/m2, with AUC 0.82. PDS activity prediction had a mean average error of < 1.6 PDS points for all groups. In obese patients, cytokine profiles appear to align in ACPA-positive and -negative patients, with leptin playing a greater role in predicting PDS activity, but with some remaining differences.
CONCLUSIONS: When stratified according to BMI, cytokine patterns can predict ACPA status and PDS activity in RA with a high degree of precision. This indicates that studies into the pathophysiology of RA should take BMI into account, to differentiate between disease- and obesity-associated phenomena. The underlying pathological processes of ACPA-negative and -positive RA appear different. Multi-cytokine evaluations may provide a deeper understanding of disease processes. Key Points • A multi-cytokine approach combined with ultrasonography and modern mathematical methods can contribute to a deeper understanding of the relationship between systemic and joint inflammation. • BMI influences cytokine profiles in rheumatoid arthritis and appears to \"override\" disease-specific processes. • Using cytokines only, and adjusting for BMI, it is possible to predict the ACPA status and joint inflammation with considerable precision.
METHODS: Patients were examined using a multi-biomarker disease assay and PDS examination of wrists and MCP and PIP joints and stratified according to ACPA status and BMI, using prediction precision to determine BMI cutoff. Analysis was performed using elastic net regularization of logistic and multiple regression. We then attempted to predict ACPA status/PDS activity based on a bootstrap approach.
RESULTS: A total of 120 measurements from 95 patients were performed. ACPA status prediction peaked at BMI 26 kg/m2, with AUC 0.82. PDS activity prediction had a mean average error of < 1.6 PDS points for all groups. In obese patients, cytokine profiles appear to align in ACPA-positive and -negative patients, with leptin playing a greater role in predicting PDS activity, but with some remaining differences.
CONCLUSIONS: When stratified according to BMI, cytokine patterns can predict ACPA status and PDS activity in RA with a high degree of precision. This indicates that studies into the pathophysiology of RA should take BMI into account, to differentiate between disease- and obesity-associated phenomena. The underlying pathological processes of ACPA-negative and -positive RA appear different. Multi-cytokine evaluations may provide a deeper understanding of disease processes. Key Points • A multi-cytokine approach combined with ultrasonography and modern mathematical methods can contribute to a deeper understanding of the relationship between systemic and joint inflammation. • BMI influences cytokine profiles in rheumatoid arthritis and appears to \"override\" disease-specific processes. • Using cytokines only, and adjusting for BMI, it is possible to predict the ACPA status and joint inflammation with considerable precision.
摘要：
目的：类风湿性关节炎可以根据ACPA和RF状态进行分类。ACPA状态可能与其他病理生理差异有关，例如，驱动炎症的细胞因子。肥胖影响RA的病程，可能涉及瘦素；确切的机制尚未完全了解。这项研究调查了BMI对RA细胞因子谱的影响，以及通过功率多普勒超声（PDS）预测ACPA状态和疾病活动的可能性。
方法：使用多生物标志物疾病检测和腕关节和MCP和PIP关节的PDS检查对患者进行检查，并根据ACPA状态和BMI进行分层，使用预测精度来确定BMI截止值。使用Logistic和多元回归的弹性网络正则化进行分析。然后，我们尝试基于引导方法来预测ACPA状态/PDS活动。
结果：共对95例患者进行了120次测量。ACPA状态预测在BMI为26kg/m2时达到峰值，AUC为0.82。对于所有组,PDS活性预测具有<1.6PDS点的平均平均误差。在肥胖患者中，细胞因子谱在ACPA阳性和阴性患者中似乎一致，瘦素在预测PDS活性方面发挥更大的作用，但有一些剩余的差异。
结论：根据BMI进行分层时，细胞因子模式可以高精度预测RA中的ACPA状态和PDS活性。这表明RA的病理生理学研究应考虑BMI，区分疾病和肥胖相关现象。ACPA阴性和阳性RA的潜在病理过程有所不同。多细胞因子评估可以提供对疾病过程的更深入的理解。要点•多细胞因子方法结合超声检查和现代数学方法可以有助于更深入地了解全身和关节炎症之间的关系。•BMI影响类风湿性关节炎中的细胞因子谱，并且似乎“超越”疾病特异性过程。•仅使用细胞因子，并调整BMI，可以相当精确地预测ACPA状态和关节炎症。
方法：使用多生物标志物疾病检测和腕关节和MCP和PIP关节的PDS检查对患者进行检查，并根据ACPA状态和BMI进行分层，使用预测精度来确定BMI截止值。使用Logistic和多元回归的弹性网络正则化进行分析。然后，我们尝试基于引导方法来预测ACPA状态/PDS活动。
结果：共对95例患者进行了120次测量。ACPA状态预测在BMI为26kg/m2时达到峰值，AUC为0.82。对于所有组,PDS活性预测具有<1.6PDS点的平均平均误差。在肥胖患者中，细胞因子谱在ACPA阳性和阴性患者中似乎一致，瘦素在预测PDS活性方面发挥更大的作用，但有一些剩余的差异。
结论：根据BMI进行分层时，细胞因子模式可以高精度预测RA中的ACPA状态和PDS活性。这表明RA的病理生理学研究应考虑BMI，区分疾病和肥胖相关现象。ACPA阴性和阳性RA的潜在病理过程有所不同。多细胞因子评估可以提供对疾病过程的更深入的理解。要点•多细胞因子方法结合超声检查和现代数学方法可以有助于更深入地了解全身和关节炎症之间的关系。•BMI影响类风湿性关节炎中的细胞因子谱，并且似乎“超越”疾病特异性过程。•仅使用细胞因子，并调整BMI，可以相当精确地预测ACPA状态和关节炎症。
