Background: Hypoxia-inducible factor prolyl hydroxylase inhibitors (HIF-PHIs) are oral alternatives to current standard-of-care treatments for anaemia in chronic kidney disease (CKD). We conducted network meta-analyses to indirectly compare clinical outcomes for three HIF-PHIs in dialysis and non-dialysis populations with anaemia in CKD.

Methods: The evidence base comprised phase III, randomised, controlled trials evaluating daprodustat, roxadustat, or vadadustat. Three outcomes were evaluated: efficacy [change from baseline in haemoglobin (Hgb)], cardiovascular safety [time to first major adverse cardiovascular event (MACE)] and quality of life [change from baseline in 36-Item Short Form Health Survey (SF-36) Vitality score]. Analyses were performed separately for all patients and for erythropoiesis-stimulating agent (ESA) non-users at baseline (non-dialysis population) or prevalent dialysis patients (dialysis population). Bayesian Markov Chain Monte Carlo methods with non-informative priors were used to estimate the posterior probability distribution and generate pairwise treatment comparisons. Point estimates (medians of posterior distributions) and 95% credible intervals (CrI) were calculated.

Results: Seventeen trials were included. In non-dialysis patients, there were no clinically meaningful differences between the three HIF-PHIs with respect to Hgb change from baseline [all patients analysis (total n = 7907): daprodustat vs. roxadustat, 0.09 g/dL (95% CrI -0.14, 0.31); daprodustat vs. vadadustat, 0.09 g/dL (-0.04, 0.21); roxadustat vs. vadadustat, 0.00 g/dL (-0.22, 0.22)] or risk of MACE [all patients analysis (total n = 7959): daprodustat vs. roxadustat, hazard ratio (HR) 1.16 (95% CrI 0.76, 1.77); daprodustat vs. vadadustat, 0.88 (0.71, 1.09); roxadustat vs. vadadustat, 0.76 (0.50, 1.16)]. Daprodustat showed a greater increase in SF-36 Vitality compared with roxadustat [total n = 4880; treatment difference 4.70 points (95% CrI 0.08, 9.31)]. In dialysis patients, Hgb change from baseline was higher with daprodustat and roxadustat compared with vadadustat [all patients analysis (total n = 11 124): daprodustat, 0.34 g/dL (0.22, 0.45); roxadustat, 0.38 g/dL (0.27, 0.49)], while there were no clinically meaningful differences in the risk of MACE between the HIF-PHIs [all patients analysis (total n = 12 320): daprodustat vs. roxadustat, HR 0.89 (0.73, 1.08); daprodustat vs. vadadustat, HR 0.99 (0.82, 1.21); roxadustat vs. vadadustat, HR 1.12 (0.92, 1.37)]. Results were similar in analyses of ESA non-users and prevalent dialysis patients.

Conclusions: In the setting of anaemia in CKD, indirect treatment comparisons suggest that daprodustat, roxadustat, and vadadustat are broadly clinically comparable in terms of efficacy and cardiovascular safety (precision was low for the latter), while daprodustat may be associated with reduction in fatigue to a greater extent than roxadustat.