Few studies on rheumatoid arthritis (RA) have generated machine learning models to predict biologic disease-modifying antirheumatic drugs (bDMARDs) responses; however, these studies included insufficient analysis on important features. Moreover, machine learning is yet to be used to predict bDMARD responses in ankylosing spondylitis (AS). Thus, in this study, machine learning was used to predict such responses in RA and AS patients. Data were retrieved from the Korean College of Rheumatology Biologics therapy (KOBIO) registry. The number of RA and AS patients in the training dataset were 625 and 611, respectively. We prepared independent test datasets that did not participate in any process of generating machine learning models. Baseline clinical characteristics were used as input features. Responders were defined as those who met the ACR 20% improvement response criteria (ACR20) and ASAS 20% improvement response criteria (ASAS20) in RA and AS, respectively, at the first follow-up. Multiple machine learning methods, including random forest (RF-method), were used to generate models to predict bDMARD responses, and we compared them with the logistic regression model. The RF-method model had superior prediction performance to logistic regression model (accuracy: 0.726 [95% confidence interval (CI): 0.725–0.730] vs. 0.689 [0.606–0.717], area under curve (AUC) of the receiver operating characteristic curve (ROC) 0.638 [0.576–0.658] vs. 0.565 [0.493–0.605], F1 score 0.841 [0.837–0.843] vs. 0.803 [0.732–0.828], AUC of the precision-recall curve 0.808 [0.763–0.829] vs. 0.754 [0.714–0.789]) with independent test datasets in patients with RA. However, machine learning and logistic regression exhibited similar prediction performance in AS patients. Furthermore, the patient self-reporting scales, which are patient global assessment of disease activity (PtGA) in RA and Bath Ankylosing Spondylitis Functional Index (BASFI) in AS, were revealed as the most important features in both diseases. RF-method exhibited superior prediction performance for responses of bDMARDs to a conventional statistical method, i.e., logistic regression, in RA patients. In contrast, despite the comparable size of the dataset, machine learning did not outperform in AS patients. The most important features of both diseases, according to feature importance analysis were patient self-reporting scales.

中文翻译：

---

基于机器学习的类风湿性关节炎和强直性脊柱炎患者 bDMARDs 反应的预测模型

很少有关于类风湿性关节炎 (RA) 的研究生成了机器学习模型来预测生物疾病缓解抗风湿药物 (bDMARDs) 的反应；然而，这些研究对重要特征的分析不够充分。此外，机器学习尚未用于预测强直性脊柱炎（AS）中的 bDMARD 反应。因此，在本研究中，机器学习用于预测 RA 和 AS 患者的此类反应。从韩国风湿病生物制剂治疗学院 (KOBIO) 注册处检索数据。训练数据集中 RA 和 AS 患者的数量分别为 625 和 611。我们准备了独立的测试数据集，这些数据集不参与任何生成机器学习模型的过程。基线临床特征用作输入特征。反应者被定义为在第一次随访时分别满足 RA 和 AS 的 ACR 20% 改善反应标准 (ACR20) 和 ASAS 20% 改善反应标准 (ASAS20) 的人。多种机器学习方法，包括随机森林（RF 方法），被用于生成模型来预测 bDMARD 反应，我们将它们与逻辑回归模型进行了比较。RF 方法模型具有优于逻辑回归模型的预测性能（准确度：0.726 [95% 置信区间 (CI)：0.725–0.730] vs. 0.689 [0.606–0.717]，接收器操作特征的曲线下面积 (AUC)曲线 (ROC) 0.638 [0.576–0.658] vs. 0.565 [0.493–0.605]，F1 得分 0.841 [0.837–0.843] vs. 0.803 [0.732–0.828]，AUC vs.0.0732–0.828 精确度曲线的 AUC vs.0.826989 . 0.754 [0.714–0.789]) 与 RA 患者的独立测试数据集。然而，机器学习和逻辑回归在 AS 患者中表现出相似的预测性能。此外，患者自我报告量表，即 RA 患者疾病活动度总体评估 (PtGA) 和 AS 患者巴斯强直性脊柱炎功能指数 (BASFI)，被揭示为这两种疾病的最重要特征。RF 方法在 RA 患者中对 bDMARDs 对传统统计方法（即逻辑回归）的反应表现出优异的预测性能。相比之下，尽管数据集的大小相当，但机器学习在 AS 患者中的表现并不好。根据特征重要性分析，这两种疾病最重要的特征是患者自我报告量表。患者自我报告量表，即 RA 患者疾病活动度总体评估 (PtGA) 和 AS 患者巴斯强直性脊柱炎功能指数 (BASFI)，被揭示为这两种疾病的最重要特征。RF 方法在 RA 患者中对 bDMARDs 对传统统计方法（即逻辑回归）的反应表现出优异的预测性能。相比之下，尽管数据集的大小相当，但机器学习在 AS 患者中的表现并不好。根据特征重要性分析，这两种疾病最重要的特征是患者自我报告量表。患者自我报告量表，即 RA 患者疾病活动度总体评估 (PtGA) 和 AS 患者巴斯强直性脊柱炎功能指数 (BASFI)，被揭示为这两种疾病的最重要特征。RF 方法在 RA 患者中对 bDMARDs 对传统统计方法（即逻辑回归）的反应表现出优异的预测性能。相比之下，尽管数据集的大小相当，但机器学习在 AS 患者中的表现并不好。根据特征重要性分析，这两种疾病最重要的特征是患者自我报告量表。RF 方法在 RA 患者中对 bDMARDs 对传统统计方法（即逻辑回归）的反应表现出优异的预测性能。相比之下，尽管数据集的大小相当，但机器学习在 AS 患者中的表现并不好。根据特征重要性分析，这两种疾病最重要的特征是患者自我报告量表。RF 方法在 RA 患者中对 bDMARDs 对传统统计方法（即逻辑回归）的反应表现出优异的预测性能。相比之下，尽管数据集大小相当，但机器学习在 AS 患者中的表现并不好。根据特征重要性分析，这两种疾病最重要的特征是患者自我报告量表。