MO293NEUTROPHIL-TO-LYMPHOCYTE RATIO AND OUTCOME IN CRESCENTIC GLOMERULONEPHRITIS

Background and Aims Crescent formation is a nonspecific response to severe injury to the glomerular capillary wall, which may be seen with any form of inflammatory glomerulopathy. Despite improved therapeutic interventions, patients with crescentic glomerulonephritis (CGN) still have a severe kidney prognosis and high mortality. Since neutrophil-to-lymphocyte ratio (NLR) is an inflammatory marker linked to worse outcomes in patients with malignancies, chronic kidney disease, myocardial infarction, and other clinical settings, we aimed to assess if NLR can predict kidney outcome and mortality in subjects with CGN. Method Eighty-four adults with biopsy-proven CGN between 1st Jan. 2008 and 31st Dec. 2017 [age 56 (95%CI 53 to 59) years, 50% males, eGFR 9.3 (95%CI 7.8-12.7) mL/min] were retrospectively enrolled in this single-centre study. Subjects were followed for a median of 31 (95%CI 6 to 56) months, until 31st May 2018. Seven subjects with inadequate biopsy sample and insufficient data were excluded. Demographic (age, gender), clinical and laboratory data at the time of biopsy were obtained from medical records. Kaplan-Meier method was used to evaluate kidney and patient survival. Variables related to kidney outcome were evaluated in a multivariate Cox proportional hazard (CPH) model. According to median NLR value (4.3, 95%CI 3.5 to 5.2) subjects were clustered in low NLR group (≤4.3; n=38 pts.) and high NLR group (>4.3; n=38 pts.). The primary endpoints were time to renal replacement therapy (RRT) initiation and all-cause mortality. Results The most common CGN subtype was pauci-immune GN (76.3%; i.e. myeloperoxidase-ANCA vasculitis - 48.7%, PR3-ANCA vasculitis - 15.8% and ANCA-negative vasculitis - 11.8%) followed by. anti-GBM antibody and immune complex related GN, with similar frequencies (11.8% each). According to kidney biopsy (KB) findings, half of the subjects had fibro-cellular crescents (55.3%), while cellular and fibrous crescents were found in 35.5% and 9.2%, respectively. Almost all subjects received corticosteroids (97.4%) and 82.9% received cyclophosphamide. Baseline eGFR was lower in the high-NLR group (8.5 vs. 11.6 mL/min, p=0.04), but no other differences in laboratory findings at baseline between the two groups were found. In bivariate analysis, NLR was negatively associated with serum albumin (rs=-0.26, p=0.02). NLR was not associated with other inflammation markers, Charlson comorbidity score, nor with the type of crescents at KB. During the follow-up period 53.9% started RRT and 19.7% died. There were no differences regarding mortality between the two groups. The mean kidney survival time was 47.6 (95%CI 33.5, 61.7) months. Kidney survival at 12, 24, 48 and 60 months were 44, 41, 38, and 33% respectively. In univariate time-dependent analysis (Figure) patients with low-NLR (68.95%CI 48 to 88 months) had better kidney survival than those with high-NLR (25, 95%CI 13 to 38 months; log rank p=0.004). Moreover, the kidney survival advantage remained (OR 1.06, 95%CI 1.002 to 1.16) after adjusting for eGFR, proteinuria, C reactive protein, immunosuppressive treatment and CGN etiology. Lower eGFR-ul was also associated with poor kidney survival (OR 0.96, 95%CI 0.88 to 0.97). Conclusion In adults with biopsy-proven crescentic glomerulonephritis and advanced kidney function decline, a higher neutrophil-to-lymphocyte ratio seems to predict worse kidney survival, but not the risk of mortality.