Vadadustat cuts hospitalisations 10% versus ESA in VOICE trial
U.S. Renal Care (USRC) has presented results from the Vafseo Outcomes In Center Experience (VOICE) trial at the American Society of Nephrology Kidney Week 2026, showing that vadadustat, Akebia Therapeutics' oral HIF-PHI, outperformed standard erythropoiesis-stimulating agents (ESAs) on a composite endpoint of all-cause mortality and hospitalisation in dialysis patients.
The randomised, open-label, active-controlled non-inferiority trial enrolled 2,116 adult in-centre haemodialysis patients. The primary composite endpoint produced a win odds of 1.16 (95% CI 1.06–1.28, p=0.0016), meeting both non-inferiority and superiority thresholds versus ESA.
Trial findings
The reduction in the composite was driven principally by hospitalisation rather than mortality. Vadadustat-treated patients recorded 111.1 hospitalisations per 100 patient-years against 123.1 in the ESA arm, an incidence rate ratio of 0.90 (95% CI 0.82–0.99), representing a 10% relative reduction. All-cause mortality rates were identical between arms at 8.8 per 100 patient-years, with an incidence rate ratio of 1.00 (95% CI 0.72–1.39), indicating no statistically significant difference.
Exposure-adjusted serious adverse events were also lower with vadadustat: 121.5 events per 100 patient-years versus 134.2 for ESA. Haemoglobin control was notably tighter on vadadustat, with a 64% reduction in the odds of haemoglobin exceeding 12 g/dL compared with ESA (29.6% versus 51.2%, p<0.0001), and a 57% reduction above 13 g/dL. The odds of requiring ESA rescue were 28% lower in the vadadustat arm (44.6% versus 52.8%, p=0.0006).
Geoffrey Block, Associate Chief Medical Officer and Senior Vice President, Clinical Research and Medical Affairs at USRC, said vadadustat "resulted in a statistically significant reduction in the primary composite endpoint of all-cause mortality and hospitalization, with the results driven by a reduction in hospitalization."
Market and regulatory context
Vadadustat, marketed as Vafseo, received FDA approval in March 2024 for anaemia due to chronic kidney disease in adults on dialysis for at least three months, and became commercially available in the United States in January 2025. Akebia states it is the only oral HIF-PHI approved or in active US development for this indication, giving it a period of clear differentiation in what has historically been an injectable ESA-dominated market.
The HIF-PHI class has faced a challenging regulatory environment in some geographies. Roxadustat, a competing HIF-PHI from AstraZeneca and FibroGen, received a narrower-than-hoped label in the US following cardiovascular safety concerns, and dalprodustat from GSK gained approval in 2023 for dialysis-dependent CKD anaemia but primarily competes outside the US. The VOICE data, generated in a real-world-adjacent dialysis network rather than a traditional industry-sponsored trial, are likely to carry weight with US nephrologists and payers who have been cautious about the class. USRC operates more than 37,000 patients across 32 states, giving the dataset meaningful external validity.
Physicians and formulary committees will weigh the hospitalisation reduction against Vafseo's boxed warning for thrombotic vascular events, including MACE, and its prescribing restrictions around haemoglobin targets above 11 g/dL. The VOICE data showing fewer high haemoglobin excursions may help address some of those concerns in practice, though the open-label design is a limitation that reviewers are likely to note. Full trial data, including subgroup analyses, are expected to be published in a peer-reviewed journal following the ASN presentation.