#887 Complement biomarkers in IgA nephropathy and IgA vasculitis with renal involvement: can urine rather than blood be the answer?

Abstract Background and Aims IgA nephropathy (IgAN) is the most common glomerulonephritis worldwide. Dysregulation of the complement system is considered a part of the pathogenesis and can occur both in the circulation and locally in the kidney. Despite this, most studies have focused on complement biomarkers in the blood, rather than in urine. Our aim was to evaluate a wide range of complement-related proteins in blood and urine at diagnosis and their association with disease activity in biopsy findings, proteinuria, and outcome. We added PTX-3, a protein expressed in activated endothelial cells, which can be both produced by and activate mesangial cells in IgAN and has a function in all three complement pathways. Adult-onset IgA vasculitis with renal involvement (IgAVN) shares many similarities with IgAN and cannot currently be distinguished by kidney biopsy or any established biomarker. We therefore also aimed to compare the same biomarkers between these two disorders. Method In a Swedish cohort of 96 patients with IgAN (n = 65) and IgAVN (n = 31), who had been followed prospectively for a median of 10.8 years, we performed the following analysis in plasma and urine in samples stored at the time of kidney biopsy. Analyses in plasma (p-) included C3bc, C4c, C5b-9, MASP2, MASP3, MAP1, FCN 1-3, MBL, CL11 and PTX-3 and in urine (u-) C3bc, C4c, C5b-9, MASP3, FCN2, FCN3, MBL and PTX-3. An in-house developed sandwich ELISA was used for all analyses. Kidney biopsies were classified according to the Oxford MEST-C score, describing (simplified for this small cohort) the presence vs absence (1 vs 0) of mesangial proliferation (M), endocapillary proliferation (E), segmental sclerosis (S), tubular atrophy/interstitial fibrosis (T) and crescents (C). Severe outcome was defined by a combination of an eGFR slope > 2.5 ml/min/1.73 m2 per year or end stage kidney disease. For statistical analysis, Mann-Whitney U test was performed for comparison between independent groups, Chi squared test and Fisher's test for categorical variables and Spearman's ranks test for correlation analysis. Median values are presented without IQR for this abstract. Results Urine samples were available in 59 patients (IgAN n = 39, IgAVN n = 20). Patients with detectable u-MBL (28.9%) and u-C5b-9 (27.1%) had higher levels of proteinuria than those with undetectable levels (1.5 g/d vs 0.71g/d; p = 0.009 and 2.3 g/d vs 0.7 g/d respectively; p < 0.001). Lower eGFR (71 ml/min/1.73 m2 vs 88 ml/min/1.73 m2; p = 0.034) were observed if u-C5b-9 was measurable. Detectable levels of u-PTX-3 and u-MBL were more often found in the presence of M1 vs M0 (53% vs 20%; p = 0.022 and 47% vs 12.5%; p = 0.011), E1 vs E0 (55% vs 14%; p = 0.001 and 40% vs 11%; p = 0.020) and C1 vs C0 (56% vs 11%; p = 0.003 and 44% vs 11%; p = 0.012). No differences in proteinuria or eGFR were seen in patients with or without detectable u-PTX-3. The degree of proteinuria was positively correlated to the levels of u-C3bc (r = 0.33; p = 0.011) and u-C4c (r = 0.39; p < 0.001). U-C4c levels were higher in patients with M1 vs M0 (54.4 Au/mL vs 23.8 Au/mL; p = 0.037) and T1 vs T0 (53.2 Au/mL vs 23.8 Au/mL; p = 0.034). Concerning the analyses in plasma (n = 95), lower p-MASP3 and p-CL11 levels were correlated with lower eGFR (r = 0.29; p = 0.004 and r = 0.31, p = 0.002). Moreover, lower p-MASP3 levels were detected in patients with E1 vs E0 (2918 ng/mL vs 3404 ng/mL; p = 0.002) and C1 vs C0 (3053 ng/mL vs 3427 ng/mL; p = 0.015). Both p-C5b-9 (13.3 Au/mL vs 8.2 Au/mL; p = 0.001) and p-FCN 1 (395 ng/mL vs 289 ng/mL; p = 0.001) were higher in patients with C1 vs C0. No single complement-related protein was associated with severe outcome, and no statistically significant differences in plasma or urine levels between patients with IgAN and IgAVN were observed. Conclusion Increased urinary levels of MBL and PTX-3 could be potential biomarkers of disease activity in both IgAN and IgAVN, whereas excretion of C5b-9 might be more unspecific and associated with chronic damage. However, these results need to be confirmed in larger and more ethnically diverse cohorts. The impact of treatment on these potential biomarkers for disease activity need to be evaluated in longitudinal studies.