FC 074POOLED EFFICACY AND CARDIOVASCULAR SAFETY RESULTS OF 3 PLACEBO-CONTROLLED AND 1 DARBEPOETIN ALFA-CONTROLLED STUDIES OF ROXADUSTAT FOR TREATMENT OF ANAEMIA IN PATIENTS WITH NON-DIALYSIS-DEPENDENT CHRONIC KIDNEY DISEASE

Abstract

Abstract Background and Aims Roxadustat, an oral hypoxia-inducible factor prolyl hydroxylase inhibitor, induces a coordinated erythropoietic response by increasing endogenous erythropoietin levels and improving iron metabolism. This analysis was performed to examine the consistency of efficacy and cardiovascular safety results for roxadustat vs placebo or darbepoetin alfa (DA) in patients with non-dialysis-dependent (NDD) chronic kidney disease (CKD). Method Results from three phase 3, double-blind studies comparing roxadustat with placebo (ALPS, ANDES, OLYMPUS) with anaemia and stage 3-5 NDD CKD were pooled and assessed for consistency with results of an open-label study comparing roxadustat with DA (DOLOMITES) in predominantly European patients. The primary efficacy endpoint was haemoglobin (Hb) response, defined as (a) Hb ≥11.0 g/dL that changed from baseline (CFB) by ≥1.0 g/dL in patients with Hb >8.0 g/dL or (b) Hb CFB by ≥2.0 g/dL in patients with Hb ≤8.0 g/dL irrespective of and without rescue therapy for placebo-controlled (assessed for superiority) and DA-controlled (assessed for non-inferiority) studies, respectively. The secondary efficacy endpoint was Hb CFB to Weeks 28-36 using least squares mean difference (LSMD) without rescue therapy. The key safety endpoints of major adverse cardiovascular event (MACE, comprising death, myocardial infarction, and stroke) and MACE+ (MACE plus hospitalization with heart failure or unstable angina) were adjudicated in all studies. MACE and MACE+ for roxadustat vs placebo or DA were compared in the intention-to-treat sample using a Cox proportional hazards model. Results In total, 4886 patients were randomised in placebo-controlled (2386 roxadustat, 1884 placebo) and DA-controlled (323 roxadustat, 293 DA) studies. Baseline characteristics of patients treated with roxadustat vs placebo and roxadustat vs DA were similar: baseline Hb (placebo, 9.10 vs 9.10 g/dL; DA, 9.55 vs 9.55 g/dL) and iron repletion (placebo, 59.9% vs 59.8%; DA, 56.3% vs 51.9%). Roxadustat was superior to placebo for Hb response without rescue therapy (80.2% vs 8.7%; difference of proportion [DOP], 71.50%; 95% CI, 69.40-73.51) and noninferior to DA (89.5% vs 78.0%; DOP, 11.51%; 95% CI, 5.66-17.36). Mean Hb CFB (Weeks 28-36) achieved superiority in pooled analysis vs placebo (LS mean, 1.91 vs 0.14; LSMD, 1.77; 95% CI, 1.69-1.84; P<0.0001) and noninferiority vs DA (LS mean, 1.85 vs 1.84; LSMD, 0.02; 95% CI, -0.13 to 0.16). Risk for MACE or MACE+ was similar for roxadustat vs placebo (MACE, 480 [20.1%] vs 350 [18.6%]; HR, 1.10; 95% CI, 0.96-1.27; MACE+, 578 [24.2%] vs 432 [22.9%]; HR, 1.07; 95% CI, 0.94-1.21) and vs DA (MACE, 38 [11.8%] vs 41 [14.0%]; HR, 0.81; 95% CI, 0.52-1.25; MACE+, 54 [16.7%] vs 43 [18.1%]; HR, 0.90; 95% CI, 0.61-1.32). Conclusion Roxadustat corrected Hb more effectively than placebo and comparably to DA in patients with anaemia and stage 3-5 NDD CKD. Cardiovascular safety was comparable between roxadustat and DA and placebo.