Document resource
Objectives Early reduction in proteinuria is an important predictor of favorable outcomes in patients with Lupus Nephritis (LN). To date, few studies have explored the longitudinal kinetics of 24-hour proteinuria (24hr UP) in real-world settings. We aimed to identify distinct 24hr UP trajectories over 24 months in a contemporary LN cohort and determine baseline predictors, as well as clinical outcomes associated with each trajectory.Methods A multicenter cohort study of patients with biopsy-proven LN was established. Repeated 24hr UP measurements were analyzed using latent class trajectory modelling (LCTM). Baseline clinical, laboratory, and histological characteristics were compared across latent classes by multinomial logistic regression, and an elastic-net regularized approach with cross-validation was applied to identify independent baseline predictors of class assignment. Outcomes at 24 months were compared across trajectories.Results 174 LN patients with a total of 998 study visits were included. The analysis revealed three distinct 24hr UP trajectories: (a) rapid responders (n=70, class 1), (b) suboptimal responders (n=63, class 2), and (c) non-responders (n=41, class 3) ( figure 1a). In multivariate multinomial regression analysis, with class 1 as the reference group, higher baseline proteinuria (Class 2: OR 1.94, p< 0.001; Class 3: OR 1.52, p= 0.006), greater SDI (Class 2: OR 2.81, p= 0.046), and higher crescent count (Class 2: OR 1.46, p= 0.032; Class 3: OR 1.49, p= 0.024) independently predicted suboptimal or no response of proteinuria.At 24 months, estimated glomerular filtration rate (eGFR) varied significantly across trajectory classes, with the lowest mean eGFR observed among non-responders (- 12.6 mL/min/1.73 m2 vs. rapid responders, p= 0.047), consistent with worse renal outcomes in patients with persistent proteinuria (figure 1b).Abstract S9:04 Figure 1Conclusions Three distinct trajectories of proteinuria were identified in a real-world multicenter LN cohort. Higher baseline proteinuria, increased damage, and histologic activity independently predicted inclusion in the non-response trajectory and identified patients at greater risk of GFR compromise at 24 months.