MO540CARDIOVASCULAR OUTCOMES ASSOCIATED WITH ACHIEVED HAEMOGLOBIN LEVEL IN POOLED PHASE 3 STUDIES OF ROXADUSTAT IN NON-DIALYSIS-DEPENDENT PATIENTS WITH ANAEMIA

Background and Aims Randomised clinical trials evaluating anaemia correction using erythropoietin-stimulating agents (ESAs) in chronic kidney disease (CKD) patients have demonstrated an association between ESA use and increased risk of cardiovascular (CV) complications and mortality. Although the mechanism for this increased risk remains unclear, possibilities include rapid rates of rise (ROR) in haemoglobin (Hb), exposure to high ESA dose and dose escalation, a high Hb target range, and off-target effects of ESAs. Thus, current recommendations generally state that Hb levels should be maintained between 10 and 12 g/dL. Roxadustat, a hypoxia-inducible factor prolyl hydroxylase inhibitor (HIF-PHI), is used for the treatment of anaemia of CKD; roxadustat stimulates endogenous production of erythropoietin and enhances iron availability, thereby supporting red blood cell production. In this pooled analysis, we explored the relationship between 1) achieved Hb levels and 2) preliminary data for ROR of Hb and subsequent incidence rates of CV events in non–dialysis-dependent (NDD) CKD patients who received roxadustat. Method We analysed pooled data from four phase 3 studies in NDD patients (placebo-controlled: ALPS, ANDES, OLYMPUS; ESA-controlled: DOLOMITES) with anaemia of CKD who received any dose of roxadustat. The reported events occurred during the treatment period and within 7 days of the last dose of roxadustat. Incidence rates of adjudicated major adverse CV events (MACE: all-cause mortality [ACM], myocardial infarction, and stroke), MACE+ (MACE plus heart failure and unstable angina requiring hospitalisation), and ACM were examined relative to 1) most recent Hb level before the event and 2) the preliminary data for ROR of Hb within the 4 weeks of treatment immediately prior to the event. Results Overall, 2709 patients were randomised and received roxadustat. Incidence rates of MACE, MACE+, and ACM were 3- or 4-fold higher in patients with lower reported Hb levels (Hb 2 g/dL in Hb within the 4-week period prior to the event compared to those with a change of ≤2 g/dL within the same period (Table). However, it should be noted that patient-exposure years (PEY) were low for the subgroup of patients with ROR >2 g/dL (∼3.3% of overall PEY), and patients with a decreasing Hb were included in the ≤2 g/dL/4 weeks subgroup. Conclusion In this analysis of roxadustat-treated patients with NDD CKD, incidence rates of MACE, MACE+, and ACM were lower in patients who achieved target Hb levels of 10-12 g/dL compared with patients who achieved Hb