Time-Varying Proteinuria and Progression of IgA Nephropathy: A Cohort Study

Rationale & Objective Proteinuria is a surrogate endpoint for predicting long-term kidney outcomes in IgA nephropathy (IgAN) with levels <1 g/d identified as a therapeutic target. However, this threshold has not been sufficiently studied. We aimed to quantify the associations of progression of IgAN with various levels of proteinuria. Study Design Observational cohort study. Setting & Participants: 1530 patients with IgAN and at least 12 months of follow-up at Peking University First Hospital. Exposure Proteinuria levels updated over time (TV-P). Outcomes A composite kidney outcome of a 50% reduction in the eGFR or end-stage kidney disease (ESKD). Analytical Approach Marginal structural models (MSMs). Results After a median follow-up of 43.5 [27.2, 72.8] months, 254 (16.6%) patients developed the composite kidney outcome. A graded association was observed between TV-P and composite kidney outcomes with higher risk among those with proteinuria ≥0.5 g/d. Compared with TV-P <0.3 g/d, the HRs (95% CI) for proteinuria levels of 0.3 - <0.5 g/d, 0.5 - <1.0 g/d, 1.0 - <2.0 g/d and ≥ 2.0 g/d were 2.22 (0.88-5.58), 4.04 (1.93-8.46), 8.46 (3.80-18.83) and 38.00 (17.62-81.95), respectively. The trend was more pronounced in patients with baseline proteinuria ≥1.0 g/d among whom a higher risk was observed with TV-P of 0.3 to <0.5 g/d compared to TV-P < 0.3 g/d (HR 3.26, 95% CI 1.07-9.92; P=0.04). However, in patients with baseline proteinuria levels <1 g/d, the risk of composite kidney outcome only began to increase when TV-P was ≥1.0 g/d (HR 3.25, 95% CI 1.06-9.90). Limitations Single-center observational study, selection bias and unmeasured confounders Conclusions This study showed that patients with IgAN and proteinuria levels >0.5 g/d, have an elevated risk of kidney failure especially among patients with proteinuria levels ≥1.0 g/d before initiating treatment. These data may serve to inform the selection of proteinuria targets in the treatment of IgAN.