#2700 Serum C3/C4 levels in renal antineutrophil cytoplasmic antibody-associated vasculitis: predictors of lung disease and worse prognosis

Abstract

Abstract Background and Aims Complement activation, specially the alternative pathway, has been implicated in the development of antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV). This was further corroborated with the efficacy of Avacopan, a selective C5a receptor inhibitor used as a corticosteroids reduction strategy in the treatment of AAV [1]. Low complement factor 3 (C3) levels without real hypocomplementemia are predictors of prognosis in AAV [2]. Low C4 levels have not yet been associated with worst outcomes. We aimed to evaluate the association between circulating C3 and C4 levels at diagnosis of AAV with clinical and biochemical features and with relevant outcomes: relapse, dialysis and death. Method This is a retrospective study including 63 patients with AAV with renal involvement followed between 2003-2023, with C3 and C4 levels measurements at diagnosis. Data was recorded from clinical registries. Patients were grouped according to the median C3 and/or C4 levels, as previously performed [2]. Groups were compared using the chi-square test for categorical variables and non-parametric tests for continuous variables. We created a logistic regression model for the outcomes, adjusting for covariates. Results Mean age at diagnosis was 61.1 ± 13.4 years old and most patients had ANCA MPO specificity (69.8%, n = 44). Only 3 patients had low C3 levels (< 83 mg/dL) and C4 hypocomplementemia (<12 mg/dL) was present in other 3 patients. Median C3 levels were 127.5 mg/dL (107.3-139.0) and median C4 levels were 29.5 mg/dL (22.0-35.0). Groups were defined according to these values: low C3 levels (C3 <127.5 mg/dL, n = 31), low C4 levels (C4 <29.5 mg/dL, n = 31) and low C3 and/or C4 levels, when either of these conditions were met (n = 40). There was a positive correlation between C3 and C4 levels (r = 0.57, p < 0.001). Patients with low C3 levels had more frequent lung involvement than patients with higher C3 levels (71% vs 31.3%, p = 002), as well as alveolar hemorrhage (38.7% vs 12.9%, p = 0.020). Regarding the outcomes, there was a tendency for more patients requiring acute dialysis (first month) and progressing to dialysis during follow-up in the low C3 subgroup. Relapse and death rates were similar between the two groups. Low C4 levels were associated with older age (75 ± 11.8 vs 67.5 ± 13.7 years, p = 0.043), higher relapse rates (50% vs 25.8%, p = 0.05) and faster progression to dialysis (1.0 (1.0-5.0) vs 9.5 (1.8-54) months, p = 0.034). When combining low C3 and/or C4 levels, lung involvement was again more frequent in patients with lower complement levels (62.5% vs 30.4%, p = 0.014). Time to dialysis was lower and relapse rates were higher in the subgroup with low complement levels, but not statistically significant. No association with death was found. Interestingly, acute dialysis requirement was more prevalent in the low C3/C4 levels subgroup (32.5% vs 8.7%, p = 0.033). After adjustment for baseline renal function, low C3/C4 levels showed an increase of 7 times the odds of a patient requiring acute dialysis (OR 6.98; 95% CI 1,196-40,839, p = 0.031, Table 1). Low C3/C4 levels independently predicted the odds of lung disease in AAV (OR = 5,069; 95% CI 1,475-17,612 p = 0.010, Table 2). Conclusion Complement activation in AAV has been previously associated to renal outcomes, but our results also highlight a possible role of complement in the pathophysiology of lung disease. Low C4 levels were associated with higher relapse rates and faster progression to dialysis, revealing that C4 activation impacts outcomes in AAV and that both classic and alternative complement pathways are involved. Combining low C3 and/or C4 levels identified patients at higher risk of acute dialysis requirement, independently of renal function at admission. Further insights in the specific effects of C3/C4 are needed to stipulate the value of monitoring and targeting complement factors in AAV.