University of Alabama at Birmingham.
Brigham and Women's Hospital, Boston, Massachusetts.
Arthritis Care Res (Hoboken). 2021 Apr;73(4):471-480. doi: 10.1002/acr.24471. Epub 2021 Mar 13.
To use unbiased, data-driven, principal component (PC) and cluster analysis to identify patient phenotypes of rheumatoid arthritis (RA) that might exhibit distinct trajectories of disease progression, response to treatment, and risk for adverse events.
Patient demographic, socioeconomic, health, and disease characteristics recorded at entry into a large, single-center, prospective observational registry cohort, the Brigham and Women's Rheumatoid Arthritis Sequential Study (BRASS), were harmonized using PC analysis to reduce dimensionality and collinearity. The number of PCs was established by eigenvalue >1, cumulative variance, and interpretability. The resulting PCs were used to cluster patients using a K-means approach. Longitudinal clinical outcomes were compared between the clusters over 2 years.
Analysis of 142 variables from 1,443 patients identified 41 PCs that accounted for 77% of the cumulative variance in the data set. Cluster analysis distinguished 5 patient clusters: 1) less RA disease activity/multimorbidity, shorter RA duration, lower incidence of comorbidities; 2) less RA disease activity/multimorbidity, longer RA duration, more infections, psychiatric comorbidities, health care utilization; 3) moderate RA disease activity/multimorbidity, more neurologic comorbidity; 4) more RA disease activity/multimorbidity, shorter RA duration, more metabolic comorbidity, higher body mass index; 5) more RA disease activity/multimorbidity, longer RA duration, more hepatic, orthopedic comorbidity and RA-related surgeries. The clusters exhibited differences in clinical outcomes over 2 years of follow-up.
Data-driven analysis of the BRASS registry identified 5 distinct phenotypes of RA. These results illustrate the potential of data-driven patient profiling as a tool to support personalized medicine in RA. Validation in an independent data set is ongoing.
使用无偏倚、数据驱动的主成分（PC）和聚类分析来确定类风湿关节炎（RA）患者的表型，这些表型可能表现出不同的疾病进展轨迹、治疗反应和不良事件风险。
使用 PC 分析对记录在大型单中心前瞻性观察性登记队列——布里格姆妇女类风湿关节炎序贯研究（BRASS）中的患者人口统计学、社会经济、健康和疾病特征进行了调和，以降低维度和共线性。通过特征值>1、累积方差和可解释性来确定 PC 的数量。使用 K-均值方法对患者进行聚类分析。在 2 年内比较了不同聚类之间的纵向临床结局。
对 142 个变量（来自 1443 个患者）的分析确定了 41 个 PC，这些 PC 解释了数据集总方差的 77%。聚类分析将患者分为 5 个聚类：1）RA 疾病活动度/合并症较少，RA 病程较短，合并症发生率较低；2）RA 疾病活动度/合并症较少，RA 病程较长，感染、精神合并症、卫生保健利用较多；3）RA 疾病活动度/合并症中等，神经合并症较多；4）RA 疾病活动度/合并症较多，RA 病程较短，代谢合并症较多，体质量指数较高；5）RA 疾病活动度/合并症较多，RA 病程较长，肝、骨科合并症和 RA 相关手术较多。在 2 年的随访中，不同聚类的临床结局存在差异。
对 BRASS 登记处的数据进行的无偏倚分析确定了 5 种不同的 RA 表型。这些结果说明了数据驱动的患者分析作为支持 RA 个体化医学的工具的潜力。正在进行独立数据集的验证。
