MO535: Associations of Haemoglobin Values and Rate of Changes With MACE in the ASCEND-ND Randomised Clinical Trial

Abstract

Abstract BACKGROUND AND AIMS Rapid changes in haemoglobin (Hb) following treatment with erythropoiesis-stimulating agents (ESAs) in patients with anaemia of chronic kidney disease (CKD) have been suggested to be associated with adverse outcomes [1–3]. This exploratory post-hoc analysis was performed to investigate the association between absolute Hb values or Hb changes over a 4-week period and the occurrence of first adjudicated major adverse cardiovascular event (MACE) in CKD patients not on dialysis who were treated with either daprodustat or darbepoetin. METHOD ASCEND-ND was an event driven, cardiovascular outcomes trial conducted in over 30 countries that randomized 3872 CKD patients not on dialysis with baseline Hb of 8–10 g/dL if not on a prior ESA, or 8–11 g/dL if receiving an ESA, to receive either oral, once-daily daprodustat (1937 patients) or subcutaneous darbepoetin (1935 patients). Available doses were daprodustat 1–24 mg once-daily and darbepoetin 20–400 µg total 4-weekly dose. The study was recently reported to have met the co-primary endpoints of non-inferiority for first occurrence of adjudicated MACE and mean Hb change from baseline to weeks 28 through 52 [4]. MACE was a composite of death from any cause, non-fatal myocardial infarction or non-fatal stroke, and events were adjudicated by an independent clinical events committee blinded to treatment assignment. In this exploratory post-hoc analysis, we examined the associations of post-randomization absolute Hb values and Hb changes categorized into quintiles (see Table 1) with first adjudicated MACE. Each patient's time in the study, prior to a first MACE or end of follow-up, was divided into distinct 4-week intervals, with each interval associated with a particular post-randomization Hb value and rate of change. Separately for each treatment group, these 4-week periods were grouped according to quintiles of Hb values, and MACE rates were calculated for each quintile. This analysis was repeated using quintiles derived from Hb rate of decrease and increase. MACEs that occurred prior to Week 4, the first scheduled post-randomization Hb collection, were not included in the analysis. RESULTS This analysis included 371 and 361 first occurrences of adjudicated MACE (‘first MACE’) in the daprodustat and darbepoetin treatment groups, respectively. When evaluating the rate of first MACE by absolute Hb value quintiles and irrespective of Hb change (‘All’ column in Figure 1), the MACE rate was higher in the low Hb quintile than in the high Hb quintile for both treatment groups (Figure 1). Evaluating the rate of first MACE by Hb change quintile and irrespective of absolute Hb value (‘All’ row in Figure 1), the rates of MACE were comparable across Hb change quintiles within each treatment group. When MACE risk was evaluated by both absolute Hb value and Hb change quintiles, the highest MACE rate was observed in the lowest Hb quintile (Q1 row) with the largest positive change (Q5 increase column), and this was more pronounced in the darbepoetin group. CONCLUSION This exploratory analysis suggests a possible association between Hb quintile and MACE outcome for both absolute and fluxes in Hb values. However, this analysis has several limitations including a small number of events in each quintile, possible confounding by severity of disease, and the choice of assessment window. Further studies will be needed to confirm these findings.