FP331INITIATION OF ERYTHROPOEISIS STIMULATING AGENTS AND MORTALITY IN A LARGE REFERRED CKD COHORT

Abstract

Introduction and Aims: Treating anemia in CKD patients with erythropoiesis stimulating agents (ESA) reduces the need of blood transfusions and increases patient-related quality of life. However, trials comparing partial correction with normalization of hemoglobin (Hb) in CKD-patients have failed to demonstrate any beneficial effect on cardiovascular outcomes and death. Instead, they have indicated a higher risk of cardiovascular events and stroke. In the KDIGO guidelines from 2012 it was recommended not to initiate ESA at Hb >100 g/l, but no floor for ESA initiation was given. Few trials have been designed to investigate when to initiate ESA. Here, using novel statistical methods to overcome problems of confounding by indication and time-varying covariates, we investigated different strategies for ESA initiation in relation to mortality and thromboembolic events in a nationwide cohort of CKD-patients. Methods: Included were all patients >18 years of age in the Swedish Renal Registry - Chronic Kidney Disease (referred stage 3-5 CKD patients) between 2005-2011 who at the time of the study inclusion had been ESA-naïve for at least 3 months. Inclusion was set by the first registered visit where their eGFR by MDRD was <60 ml/min/1.73m2 and Hb level ≤120 g/l. Patients were then followed until death or Sept 30, 2013. Statistical analysis was based on inverse probability weighting and dynamic marginal structural models including a rich set of both time-fixed (age, sex, primary renal disease, co-morbidity status) and time-varying covariates (eGFR, laboratory measurements, medication including iron use, diet, blood pressure, and body mass index).We defined and investigated several treatment strategies: “never initiate ESA”, “always initiate ESA”, “use ESA discriminating on Hb level 90-120 g/l” and compared these strategies to the crude population mortality (current practice). Results: 6589 CKD patients were included (median age 72 years, 61% men). The median ESA epoetin equivalent dose at initiation was 4000 IU/week (25-75th percentile 3000; 6000) and during follow-up 4000 IU/week (2500; 6000). The Figure displays the 3-year survival probability by ESA initiation Hb. Survival probability increased when the Hb cut-off (the strategy to initiate ESA at a given Hb) rose from Hb = 90 to 110 g/l, and then decreases rapidly up to 120 g/l. ESA initiation between Hb 100-115 g/l and 110-130 g/l resulted in significantly better survival (HR 0.82 95% CI 0.79-0.86 and HR 0.93 95% CI 0.91-0.95 respectively) compared with “never initiate ESA”, “always initiate ESA” and “current practice”. Stratification by sex, age, diabetes, and previous malignancy gave similar results. ESA-initiation at Hb 100-115 g/l or 110-130 g/l did not associate with the risk of thromboembolic events Conclusions: This observational study investigating several treatment strategies for ESA initiation shows that an initiation level at Hb below 110 g/l is associated with better survival. (Figure Presented).