Comparative effectiveness of hypoxia-inducible factor prolyl hydroxylase inhibitors versus erythropoiesis-stimulating agents on prognosis in non-dialysis chronic kidney disease: a propensity-matched cohort study

Abstract

Renal anemia in patients with stage-IV chronic kidney disease (CKD) is commonly treated with erythropoiesis-stimulating agents (ESAs), which are effective but associated with cardiovascular risks, variable responsiveness, and inflammatory complications. Hypoxia-inducible factor prolyl hydroxylase inhibitors (HIF-PHIs) have recently emerged as an alternative therapy that enhances endogenous erythropoietin production and improves iron metabolism. This retrospective cohort study utilized the TriNetX Global Collaborative Network to compare clinical outcomes between HIF-PHI and ESA users in non-dialysis stage IV CKD patients (estimated glomerular filtration rate 15-30 mL/min/1.73 m²). After 1:1 propensity score matching, 493 patients were included in each group and followed for up to three years. HIF-PHI therapy was associated with significantly lower all-cause mortality (hazard ratio [HR] 0.39; 95% confidence interval [CI] 0.26-0.58; p < 0.0001) and reduced sepsis risk (HR 0.32; 95% CI 0.14-0.74; p = 0.01) compared with ESA therapy. Subgroup analyses demonstrated consistent mortality benefit across major comorbidities, with the greatest advantage observed in patients with ferritin 100-299 ng/mL (HR 0.41; 95% CI 0.23-0.74; p = 0.002) and in comparisons against short-acting ESAs (HR 0.55; p = 0.0304). No survival difference was observed between HIF-PHI and long-acting ESA users, indicating that the observed benefit of HIF-PHIs was driven exclusively by differences versus short-acting ESAs. These findings suggest that HIF-PHIs may offer a safer and more physiologically adaptive approach to correcting anemia in advanced CKD, particularly when compared with short-acting ESA therapy. In contrast, long-acting ESAs achieve outcomes comparable to those of HIF-PHIs.

Keywords: Chronic kidney disease; erythropoiesis-stimulating agents; hypoxia-inducible factor prolyl hydroxylase inhibitors; renal anemia; sepsis.

Figure 1.

Flow chart of study cohort selection and matching process. Patients with stage IV chronic kidney disease (CKD, eGFR 15–30 mL/min/1.73m²) were identified from the TriNetX platform between Jan 1, 2020, and Dec 24, 2024. After applying exclusion criteria, 501 HIF-PH inhibitor users and 37,699 ESA users were identified. Following 1:1 propensity score matching, 493 pairs were included for outcome analysis.

Figure 2.

Kaplan–Meier curve comparing all-cause mortality between HIF-PHI and ESA users. Propensity score–matched patients with stage IV CKD were followed for up to 3 years. HIF-PHI users (blue line) showed significantly lower mortality compared to ESA users (red line) (log-rank p < 0.0001).

Figure 3.

Hazard ratios for major clinical outcomes: HIF-PHI vs ESA. Adjusted hazard ratios for all-cause mortality, sepsis, cardiovascular events, malignancy, and cognitive impairment were calculated after propensity score matching. HIF-PHI use was associated with significantly reduced risks of death (HR 0.39) and sepsis (HR 0.32).

Figure 4.

Subgroup analysis of all-cause mortality comparing HIF-PHI and ESA users. Mortality hazard ratios were stratified by comorbidities (diabetes, hypertension) and showed consistent benefit for HIF-PHI in all subgroups. Notably, the HR was 0.25 in patients without diabetes (p < 0.0001).

Figure 5.

Subgroup analysis of sepsis risk in HIF-PHI vs ESA users. Most comorbidity subgroups showed lower sepsis risk with HIF-PHI use, although the difference in hypertensive patients did not reach statistical significance (HR: 0.42, p = 0.09).

Figure 6.

Outcomes stratified by ferritin level: comparison of HIF-PHI and ESA. Outcomes were analyzed by ferritin level: (A) <100 ng/mL, (B) 100–299 ng/mL. HIF-PHI users in group B had significantly lower mortality (HR 0.41), sepsis (HR 0.09), and dementia (HR 0.18).