Results from a nationwide retrospective cohort measure the impact of C3 and soluble C5b-9 levels on kidney outcomes in C3 glomerulopathy

Abstract

C3 glomerulopathy (C3G) is a rare complement-mediated disease. Specific treatments are not yet available and factors predictive of kidney survival such as age, kidney function and proteinuria are not specific to C3G. The prognostic value of biomarkers of complement activation, which are pathognomonic of the diseases, remains unknown. In a large cohort of 165 patients from the French National registry, we retrospectively assess the prognostic value of C3, soluble C5b-9 (sC5b-9), C3 nephritic factor, and rare disease-predicting variants in complement genes in predicting clinical outcome of patients. By multivariate analysis age (adult onset), reduced kidney function (defined by estimated glomerular filtration rate under 60ml/min) and presence of rare disease-predicting variants in complement genes predicted risk of progression to kidney failure. Moreover, by multivariate analysis, normal C3/high sC5b-9 levels or low C3/normal sC5b-9 levels remained independently associated with a worse kidney prognosis, with the relative risk 3.7- and 8-times higher, respectively. Subgroup analysis indicated that the complement biomarker profiles independently correlated to kidney prognosis in patients with adult but not pediatric onset. In this subgroup, we showed that profiles of biomarkers C3 and/or sC5b-9 correlated with intra glomerular inflammation and may explain kidney outcomes. In children, only the presence of rare disease-predicting variants correlated with kidney survival. Thus, in an adult population, we propose a three-point C3G prognostic score based on biomarker profiles at risk, estimated glomerular filtration rate at presentation and genetic findings, which may help stratify adult patients into subgroups that require close monitoring and more aggressive therapy.

Keywords: C3; C3 glomerulopathy; biomarker; sC5b-9.

Figure 1:. Patients’ recruitment.

Flowchart of inclusion of patients within the population of the French C3 Glomerulopathy registry (n=402). A total of 165 patients were eligible for enrollment in the study, based on the availability of C3 and sC5b-9 level at the time of kidney biopsy.

Figure 2:. Biomarkers of complement activation, C3 and sC5b-9 levels in C3 Glomerulopathy patients with onset as a child or as an adult.

Four groups of patients were identified using C3 and sC5b-9 levels including patients with low C3 and high sC5b9 levels, low C3 and normal sC5b-9 levels, patients with normal C3 and normal sC5b-9 levels and patients with normal C3 and high sC5b-9 levels.

Figure 3:. Renal survival of patients with C3 glomerulopathy according to complement activation biomarkers profiles

A- Kaplan-Meier patient survival analysis of patients with C3 glomerulopathy comparing patients according to age at diagnosis. Renal survival of patients with disease onset as a child was significantly better that those with disease onset as an adult (p=0.0018). B) Kaplan-Meier patient survival analysis comparing children patients with C3 glomerulopathy according to the 4 profiles of complement activation biomarkers. In children, renal survivals of patients with low C3/high sC5b-9, low C3/normal sC5b-9 or normal C3/high sC5b-9 were comparable to renal survival of patients with normal C3/normal sC5b-9 (p=0.96; p=0.96; p=0.19 respectively). Renal survival of children with low C3/high sC5b-9 tended to be better than patients with low C3/normal sC5b-9 (p=0.09) and was significantly better than patients with normal C3/high sC5b-9 (p=0.02). C) Kaplan-Meier patient survival analysis comparing adults patients with C3 glomerulopathy according to the 4 profiles of complement activation biomarkers. In adult only, renal survival between adults with normal C3/normal sC5b-9 was similar to that of patients with low C3/high sC5b-9 (p=0.18). Renal survival of patients with low C3/normal sC5b-9 or normal C3/high sC5b-9 were significantly lower compared to patients with normal C3/normal sC5b-9 (p=0.005 and p=0.02 respectively). Renal survival of patients with low C3/high sC5b-9 levels was similar to that of patients with low C3/normal sC5b-9 (p=0.17) and was significantly better than that of patients with normal C3/high sC5b-9 (p=0.036).

Figure 4:. Forest plot of the hazard for End Stage renal Disease progression.

A) Univariate models adjusted for sex gender (male), age at onset, onset <16 years old, renal failure (eGFR<60ml/min) at onset, level of proteinuria, nephrotic range proteinuria, immunosuppressive therapy, level of C3 or sC5b-9, presence of C3NeF or rare variant in complement genes factor H, factor I or C3 and presence of biomarkers of complement activation. B) Multivariate model adjusted for age onset (<16 y), renal failure at onset, C3NeF, rare variant in complement genes and presence of complement activation biomarkers.

Figure 4:. Forest plot of the hazard for End Stage renal Disease progression.

A) Univariate models adjusted for sex gender (male), age at onset, onset <16 years old, renal failure (eGFR<60ml/min) at onset, level of proteinuria, nephrotic range proteinuria, immunosuppressive therapy, level of C3 or sC5b-9, presence of C3NeF or rare variant in complement genes factor H, factor I or C3 and presence of biomarkers of complement activation. B) Multivariate model adjusted for age onset (<16 y), renal failure at onset, C3NeF, rare variant in complement genes and presence of complement activation biomarkers.

Figure 5:. Renal survival of adult patients according to C3G Progression Risk Score.

Kaplan-Meier patient renal survival analysis comparing C3G patients risk score of progression to end stage kidney disease of 0 to 3.