University of Manitoba, Winnipeg, Manitoba, Canada, and Institute of Health Policy Management & Evaluation, Toronto, Ontario, Canada.
The Hospital for Sick Children (SickKids), Toronto, Ontario, Canada.
Arthritis Care Res (Hoboken). 2018 May;70(5):750-757. doi: 10.1002/acr.23319. Epub 2018 Mar 25.
No previous study has studied the longitudinal disease course of childhood-onset systemic lupus erythematosus (cSLE). Our objectives are to assess distinguishable differences in disease activity trajectories in cSLE patients, determine baseline factors predictive of disease trajectory membership, and assess if the different disease activity trajectories are associated with different damage trajectories.
This is a retrospective, longitudinal inception cohort of cSLE patients. Patients were followed from diagnosis as children, until they were adults. SLE disease activity was modeled as a latent characteristic, jointly using the Systemic Lupus Erythematosus Disease Activity Index 2000 and prednisone in a Bayesian growth mixture model. Baseline factors were tested for membership prediction of the latent classes of disease trajectories. Differences in damage trajectories by disease activity classes were tested using a mixed model.
A total of 473 patients (82% females), with median age at diagnosis of 14.1 years, were studied. We studied 11,992 visits (2,666 patient-years). We identified 5 classes of disease activity trajectories. Baseline major organ involvement, number of American College of Rheumatology criteria, and age at diagnosis predicted memberships into different classes. A higher proportion of Asians was in class 2 compared to class 5. Class 1 was associated with the most accrual of damage, while class 5 was associated with no significant damage accrual, even after 10 years.
There are 5 distinct latent classes of disease trajectory in patients with cSLE. Membership within disease trajectories is predicted by baseline clinical and demographic factors. Membership in different disease activity trajectory classes is associated with different damage trajectories.
既往研究尚未探讨儿童起病系统性红斑狼疮(cSLE)的纵向疾病进程。我们的目标是评估 cSLE 患者疾病活动轨迹中的可区分差异,确定预测疾病轨迹分类的基线因素,并评估不同疾病活动轨迹是否与不同的损害轨迹相关。
这是一项回顾性、纵向发病队列的 cSLE 患者研究。患者从儿童期被诊断为狼疮起接受随访,直至成年。使用贝叶斯增长混合模型,将系统性红斑狼疮疾病活动指数 2000 和泼尼松共同作为潜在特征来对 SLE 疾病活动进行建模。对基线因素进行测试,以预测疾病轨迹潜在分类的成员身份。使用混合模型测试不同疾病活动类别的损害轨迹差异。
共纳入 473 例(82%为女性)患者,中位诊断年龄为 14.1 岁,研究期间共进行了 11992 次就诊(2666 患者年)。我们确定了 5 种疾病活动轨迹分类。基线时主要器官受累、美国风湿病学会标准数和诊断时年龄可预测不同分类的成员身份。与 class5 相比,class2 中亚洲人比例更高。class1 与损害积累最多相关,而 class5 与无显著损害积累相关,即使在 10 年后也是如此。
cSLE 患者存在 5 种不同的潜在疾病轨迹分类。疾病轨迹内的成员身份由基线临床和人口统计学因素预测。不同疾病活动轨迹类别的成员身份与不同的损害轨迹相关。