Hypocomplementemia is associated with more severe renal disease and worse renal outcomes in patients with ANCA-associated vasculitis: a retrospective cohort study

Abstract Background The complement system has been recently proposed to play an important role in the pathogenesis of ANCA-associated vasculitis (AAV). This study evaluated the value of serum and kidney deposited C3 in predicting renal outcomes in AAV. Methods This was a retrospective study of 47 patients with AAV, who were categorized according to their serum C3 levels as hypo- or normo-complementemic and to those with positive or negative kidney biopsy immunofluorescence (IF) for C3. Baseline characteristics as well as progression to end-stage renal disease (ESRD) between the 2 groups were compared. Results In total, 23% (11/47) were hypo-complementemic; these patients were older (74 vs. 65 years, p = 0.013), had higher creatinine levels (4.9 vs. 2.2 mg/dL, p = 0.006), were more often hemodialysis dependent (64% vs. 19%, p = 0.009) and progressed more often to ESRD (55% vs. 11%, p = 0.01) compared to normo-complementemic patients (n = 36). On multivariate analysis, serum creatinine at diagnosis (HR = 16.8, 95%CI: 1.354–208.62, p = 0.028) and low serum C3 (HR = 2.492; 95% CI: 1.537–11.567; p = 0.044) were independent predictors for ESRD. Among 25 patients with an available kidney biopsy, 56% had C3 deposition by IF and displayed more often a mixed histological pattern (72% vs. 27%, p = 0.033), low serum C3 levels (42% vs. 18%, p < 0.001) and serious infections during follow-up (57% vs. 18%, p = 0.047) compared to those with negative (n = 11) IF staining. Conclusion Almost one of four patients with AAV has low C3 levels at diagnosis which is associated with more severe renal disease and worse renal outcomes (ESRD). This should be taken into account in therapeutic and monitoring strategies.


Introduction
Anti-neutrophil cytoplasmic autoantibody (ANCA)-associated vasculitis (AAV) is a systemic small vessel vasculitis with multiorgan involvement and significant associated morbidity and mortality. The role of complement activation, mainly of the alternative pathway, in its pathogenesis has been recently suggested to be significant [1]. Although necrotizing and crescentic glomerulonephritis related to ANCAs is typically referred as a "pauci-immune" process, in certain cases there is immune complex and complement deposition in glomeruli as detected by immunofluorescence (IF) [2]. Kidney complements deposits of patients with AAV correlated with greater kidney damage, more significant proteinuria, and overall disease activity [3]. Nevertheless, data regarding the prognostic significance of low C3 levels and C3 kidney deposition in patients with AAV are limited.
Hence, the aim of this study was to investigate the potential correlation of C3 in the serum and kidneys with renal characteristics and prognosis in a cohort of patients with AAV.

Patients and methods
This was a retrospective cohort study of patients with AAV followed in a tertiary referral hospital (Nephrology Department and Clinical Immunology--Rheumatology Unit, 2nd Department of Medicine, Hippokration General Hospital, Athens, Greece) through the years 2002 and 2019. Among 62 patients of the Caucasian race with newly diagnosed AAV, 47 patients with available complement levels at diagnosis were included in the study. All patients with AAV satisfied the inclusion criteria of ACR 1990 or the Chapel Hill Consensus Conference definitions [4] for granulomatosis with polyangiitis (GPA), microscopic polyangiitis (MPA), and ANCA-limited renal vasculitis (LRV). Patients with other underlying diseases that cause secondary vasculitis, such as systemic lupus erythematosus, rheumatoid arthritis, IgA nephropathy, Sjogren's syndrome, malignancy and Behcet's disease were excluded clinically and serologically. There was no temporal relationship between clinically evident vasculitis and the administration of an offending drug. So, drug-induced vasculitis was also excluded. The patients, who included in the study, met one of the two following criteria: 1. Necrotizing vasculitis predominantly affecting small vessels and/or necrotizing glomerulonephritis with few or no immune deposits 2. positive ANCA with IF and ELISA. The study was approved by the Hippokration General Hospital's institutional review board (57/26-3-2018) and the consent of all patients for their retrospective chart review was obtained. Demographic data were collected for each patient (age, sex, type of AAV, and patient follow-up). Clinical characteristics at diagnosis, such as major organ involvement (lung: pulmonary hemorrhage, kidney: hemodialysis dependence) and induction treatment (with or without plasmapheresis), episodes of relapses, and serious infections during follow-up and eventually mortality.
As far as laboratory results, hemoglobin, creatinine level, 24-h proteinuria, hematuria, ANCA status (p-ANCA MPO, c-ANCA PR3, double positivity [anti-GBM with ANCA] or negative), serum C3, C4 and ANA, and anti-dsDNA were included. C3 and C4 levels were assessed by tholosimetry. Anti-neutrophil cytoplasmic antibodies (c-ANCA and p-ANCA) were determined by a direct IF technique and the specificity for PR3 and MPO by ELISA. The serum C3 reference range in our laboratory is 75-180 mg/dL. The patients were divided into 2 groups, those with low C3 (hypo-complementemic: <75 mg/dL), and those with normal C3 (normo-complementemic: 75-180 mg/dL). No one had C3 levels higher than 180 mg/dL.
A total of 40 patients presented with renal involvement, which was defined by the presence of 24-h urine proteinuria of >1 g, active urinary sediment (leukocyturia > 5 white blood cells/field, hematuria >5 erythrocytes/field, and/or cellular casts) and/or increase of creatinine values over 1.2 mg/dL. Kidney biopsy was performed in 25 patients (62.5%). Renal histology was evaluated using light microscopy (hematoxylin and eosin, periodic acid-Schiff, and Masson trichrome stains) and was classified into 4 histological subtypes according to the 2010 histologic classification of AAV [5]. Those presenting !50% of all glomeruli globally sclerosed were screlotic type and those with !50% normal glomeruli were classified as focal type. Those showing !50% of glomeruli with semilunar cells were classified as crescentic type and those not presenting the anterior criteria were classified as mixed type. The presence of immunoglobulin (IgG, IgM, IgA), complement (C3, C1q, C4), kappa and lambda light chains was evaluated by direct IF on the fresh frozen tissue. The staining of C3 was routinely performed using polyclonal rabbit antihuman antiserum, at dilution 1:50 (Dako, Copenhagen, Denmark). According to their C3 deposition for at least 1þ in a 0-4þ scale, patients were divided into those with (þ) and (-) IF. Electron microscopy was not performed.

Statistical analysis
Data were analyzed using Statistical Package for the Social Sciences (SPSS) software program, version 19.0 for Windows (SPSS, Chicago, IL, USA). Values are expressed as medians and interquartile ranges (IQR) for continuous variables and percentages for categorical variables. Comparison of variables was done using the independent samples t-Test for Continuous variables and Chi-square test or Fisher's exact test for categorical variables. A two-sided value p < 0.05 was considered as statistically significant. To address the independent predictive value of factors associated with the progression to end-stage renal disease (ESRD--defined as eGFR < 15 mL/min/1.73 m 2 or requiring hemodialysis) a univariate and multivariate Cox proportional hazards regression analysis were performed. We used the variables with p values of less than 0.1 between progressors to ESRD and non-progressors.
In 68% of the patients, who included, presented with lung involvement and among them 6 (13%) presented with pulmonary hemorrhage (Table 1).
Finally, 10 patients (21%) progressed to ESRD and 8 (17%) died (4 during the induction and 4 during the follow-up period). The causes of death were serious infections in 3 patients, lung cancer in 1 patient, heart disease in 1 and unknown causes in 4 patients. It is noteworthy that the two patients with double positivity (ANCA and anti-GBM) at diagnosis progressed to ESRD between 3-6 months of disease duration.

Differences between C3 kidney positive and negative patients
Kidney biopsies were available in 25/40 (62.5%) patients with renal involvement. Compared to patients without kidney biopsy (n ¼ 15), those with renal biopsy available had more frequently significant proteinuria and higher serum creatinine at diagnosis (Table 2).

Discussion
This study highlights a potential association between low serum C3 levels at diagnosis and renal severity as well as renal prognosis in patients with AAV. This is one of the few studies showing that serum hypocomplementemia at diagnosis may be an independent prognostic factor for ESRD progression in patients with AAV.
A role for complement activation in the pathogenesis of ANCA disease is supported by in vitro studies [6,7], histologic data from kidney biopsies showing complement deposition [8] as well as data from animal models and humans [9,10] indicating that inhibition of complement may be therapeutically efficacious in AAVs. These studies demonstrated that the complement activation by alternative pathway and particularly the receptor C5a (C5aR) is important in the maintenance of the inflammatory process. Two inhibitors of the complement pathway are in clinical development for ANCA vasculitis: avacopan, an oral C5a receptor inhibitor, that has demonstrated efficacy, safety and steroid sparing in two Phase II trials, being it a promising future strategy and eculizumab, a monoclonal antibody that targeted against complement C5, which inhibits the cleavage of C5 into C5a and C5b [9,10]. Furthermore, the positive effect of plasma exchange in severe AAV is hypothetically due, at least in part, to the removal of activated complement factors and of chemotaxins [10]. However, whether or not serum or kidney deposited C3 can be used as markers of disease severity or prognosis (ESRD/death) in patients with AAV is unclear.
Low serum C3 levels have been reported in 4-35% of patients with AAV [3,[11][12][13][14][15][16][17] (Table 5). In our cohort 23% of patients were hypocomplementemic at diagnosis. There are conflicting literature data regarding the potential association between C3 levels and severity of renal disease. Some studies did not find a significant correlation between hypocomplementemia and creatinine levels [3,13,[15][16][17] whereas others did [11,12,14]. Recently, Garcia et al. [14] in a cohort of 93 patients described low C3 levels in 11% and that was associated significantly not only with higher creatinine levels, but as well with proteinuria > 1g/day and active urinary sediment. Similarly, Villacorta et al. [16] on 111 patients with AAV also described a significant correlation between hypocomplementemia and the need for dialysis at diagnosis.
In our patient cohort, low C3 levels were associated with more severe renal disease (as evidenced by high serum creatinine and need for hemodialysis) and worse renal prognosis (progression to ESRD). Therefore, these findings support the hypothesis of systemic activation of the complement in a proportion of patients with AAV and this could have clinical and prognostic implications.  There has been some evidence from previous studies that serum creatinine at diagnosis could be related to worse prognosis in patients with AAV [18][19][20][21][22]. In our cohort, we found a strong association between serum creatinine at diagnosis and progression to ESRD. The novel finding of our study was that low serum C3 levels were an independent predictor of ESRD progression. Few studies have shown that hypocomplementemia at the presentation of the AAV was associated with worse renal prognosis [11][12][13]15,17]. Crnogorac et al. [17] included 75 patients with AAV and renal involvement and suggested that low serum C3 at diagnosis could be associated with worse renal prognosis (ESRD). On contrary, Augusto et al. [11] on 45 patients found a significant correlation between even lower normal C3 values at diagnosis with ESRD related mainly to histologic data (crescentic/mixed histologic form).
In our study, more than half of patients showed C3 renal deposition. These patients had more often low serum C3 levels, a mixed histologic pattern in kidney biopsies and presented more frequently serious infections during follow-up. However, there was no correlation between kidney C3 deposition and AAV disease severity and prognosis (ESRD).
Our study has certain limitations including its retrospective design, the small number of included patients and the missing data regarding kidney biopsies. On contrary, this is one of the few studies in the literature where complement serum levels and kidney deposition were examined as markers of disease severity and prognosis during long-term follow-up ($4.5 years).
Our findings are novel indicating that the subgroup of patients with AAV with low C3 at diagnosis has a more severe disease and a higher risk for progression to ESRD. Consequently, more multicentric research studies are needed to confirm our study findings, so this should be taken into account when the appropriate therapeutic and monitoring strategies are designed.

Ethical policy and institutional review board statement
This research was approved by the Hippokration General Hospital's institutional review board and consent of all patients for their retrospective chart review was obtained.