#6035 PREDICTORS OF MORTALITY IN ANCA-ASSOCIATED VASCULITIS: WHAT IS THE ROLE OF SERUM BIOMARKERS?

Abstract

Abstract Background and Aims ANCA-Associated vasculitis (AAV) is an autoimmune disease with necrotizing inflammation of small vessels as the main manifestation. AAV is a severe disease with high level of mortality. Severity markers are needed to better tailor treatment. ANCA titers have not always been correlated with disease severity and mortality. Other inflammation markers have also been used such as C3 and C-reactive protein (CRP). The aim of this study is to correlate ANCA titers, C3 and CPR values ​​with mortality in patients with renal vasculitis. Method We retrospectively analyzed all adult patients between June 2006 and December 2022 with a renal AAV in the nephrology department of São Bernardo Hospital. Demographic, clinical and laboratory findings of all patients were recorded. ANCA titers at the time of diagnosis (ANCAd), remission (ANCArm), relapse (ANCArl) and last follow-up (ANCAf) dates were registered. C3 levels were measured at diagnosis (dC3) and at the last follow-up (fC3) dates. CRP was measured at diagnosis (CRPd), remission (CRPrm) and last follow-up (CRPf). ANCA titers was analysed as continuous and categorical variables. C3 and CRP were used as continuous variables. We analyse the association between ANCA titers, C3, CPR values and mortality with univariate analysis. Multivariable adjusted cox regression analysis was performed for assessing predictive variables associated with mortality. Results We included 58 AAV patients with kidney involvement, with mean age of 69±10 years old, 54.4% were male (n = 31), 82.7% (n = 48) MPO-ANCA+, 12.1% (n = 7) PR3-ANCA+ and 5.2% (n = 3) seronegative. Median follow-up was 32.2 months. 20 (34.5%) patients died. At admission, categorization by ANCA titer revealed 5.2% (3) 0-19 UI/mL, 10.3% (6) 20-39, 8.6% (5) 40-59, 17.2% (10) 60-99, 8.6% (5) 100-134 and 38% >134 (22). C3 median was 109 UI/L at admission, 96.5 UI/L at last follow-up, CPR median was 7.9 mg/dL at admission, 2.4 mg/dL at last follow-up. Univariate analysis did not reveal a difference between patients who survived and those who died, in relation to ANCAd (p = 0.221), ANCArm (p = 0.346), ANCArl (p = 0.180) and ANCAf (p = 0.072). However, it did show that dC3 where significantly lower in those patients who died (p = 0.023) and CRPf higher (p = 0.012), but not for fC3 (p = 0.372), CRPd (p = 0.408) and CRPrm (p = 0.384). Used as categorical variable, ANCAd did not correlate to patients’ death occurrence (p = 0.788). In multivariate analysis including age, gender, dC3 and CRPf, only CRPf was almost significantly (p = 0,06, model p = 0,05). In univariate cox regression analysis to evaluate factors related to survival, none where statistic significant to predict dead – ANCA d (p = 0,349), ANCAf (p = 0.136), dC3 (p = 0.082) and CRPf (p = 0.066). In multivariate cox regression analysis, including age (p = 0.039), gender (p = 0.658), dC3 (p = 0.227) and CRPf (p = 0.639), none factor predict dead (model p = 0,08). Conclusion In our cohort of patients with AAV, we did not found a serologic marker to predict dead. However, it appears to have a relationship with low serum C3 levels at the time of diagnosis. We also found that could be a relationship with CRP at the last follow-up, but that could be a bias due to higher levels in those how died from infection causes. The diagnostic value of the PR3- and MPO ANCA test is well established, however the relationship between mortality and titer levels in the various stages of the disease is not well established. In our study, there was no association between them. Further research should determine whether C3 levels and CPR are indeed an independent predictor of AAV outcome.