81 Identifying and predicting novel classes of long-term disease trajectories for patients with juvenile dermatomyositis using growth mixture models

Objectives Uncertainty around clinical heterogeneity and outcomes for patients with juvenile dermatomyositis (JDM) represents a major burden of disease and a challenge for clinical management. We sought to identify novel classes of patients having similar temporal patterns in disease activity and clinical features at baseline that predict class membership. Methods Data were obtained for n=519 patients, including demographic and clinical features at baseline, and the physician’s global assessment of disease (PGA) and the modified disease activity score for skin at all recorded visits. Growth mixture models (GMM) were fit to identify classes of patients with similar trajectories. Baseline predictors of class membership were analysed using lasso regression. Results GMM analysis of global disease activity identified 2 classes of patients. Patients in Class 1 (88.6%) tended to improve their PGA over time, while patients in Class 2 (11.6%) tended to have more ongoing disease. Lasso regression analysis identified abnormal respiration, lipodystrophy and an interaction between them as baseline predictors of Class 2 membership, with odds ratios of 1.71, 1.91 and 1.66, respectively. GMM analysis of skin disease activity identified 3 classes of patients. Patients in Classes 1 (14%) and 2 (11%) had higher levels of skin disease at diagnosis that steadily improved or remained high, respectively. Patients in Class 3 (66%) began with lower levels of skin disease that improved more quickly. Conclusions GMM analysis identified novel JDM sub-phenotypes based on longitudinal global and skin disease activities. Abnormal respiration and lipodystrophy at baseline predicted the group of patients with ongoing disease.