Efficacy and Safety of Roxadustat for Anemia in Patients Receiving Chemotherapy for Nonmyeloid Malignancies: A Randomized, Open-Label, Active-Controlled Phase III Study

Abstract

Purpose: We evaluated the efficacy and safety of roxadustat, a first-in-class hypoxia-inducible factor prolyl hydroxylase inhibitor, for chemotherapy-induced anemia (CIA) in patients with nonmyeloid malignancies receiving multicycle treatments of chemotherapy.

Patients and methods: In this open-label, noninferiority phase III study conducted at 44 sites in China, 159 participants age ≥18 years with CIA nonmyeloid malignancy and CIA were randomly assigned (1:1) to oral roxadustat or subcutaneous recombinant human erythropoietin-α (rHuEPO-α) three times a week for 12 weeks. Roxadustat starting dosages were 100, 120, and 150 mg three times a week for participants weighing 40-<50, 50-60, and >60 kg, respectively. rHuEPO-α starting dosage for all participants was 150 IU/kg three times a week. Both roxadustat and rHuEPO-α dosages could be modified. The primary end point was least-squares mean (LSM) change in hemoglobin (Hb) concentration from baseline to the concentration averaged over weeks 9-13.

Results: Of the 159 participants randomly assigned, 140 were included in the per-protocol set (roxadustat, n = 78; rHuEPO-α, n = 62). The LSM (95% two-sided CI) change from baseline to weeks 9-13 in Hb concentration was 17.1 (13.58 to 20.71) g/L with roxadustat and 15.4 (11.34 to 19.50) g/L with rHuEPO-α (mean difference [95% CI], 1.7 [-3.39 to 6.84]). The lower bound of the one-sided 97.5% CI for the treatment difference (‒3.4 g/L) was greater than the predefined noninferiority margin of ‒6.6 g/L, establishing noninferiority. Noninferiority was supported by five of six key secondary end points. Rates of adverse events were generally comparable between treatments and consistent with previous findings.

Conclusion: Roxadustat was noninferior to rHuEPO-α in treating CIA in participants with nonmyeloid malignancies receiving multicycle treatments of myelosuppressive chemotherapy. The oral formulation of roxadustat may potentially increase compliance.

Trial registration: ClinicalTrials.gov NCT05301517.

FIG 1.

Patient disposition (all screened patients). Percentages were based on the number of patients randomly assigned. AE, adverse event; Discont, discontinued; FAS, full analysis set; PPS, per-protocol set; rHuEPO-α, recombinant human erythropoietin-α; SAE, serious adverse event; SAF, safety analysis set.

FIG 2.

Mean (SE) of absolute Hb and Hb change from baseline by scheduled visit (per-protocol set). Baseline Hb was defined as the average of the last three available central laboratory Hb values before the first doses of study drug. For patients without central laboratory hemoglobin assessments from screening period to day 1, the Hb values from local laboratory were used to define baseline in lieu of central laboratory values. Hb, hemoglobin; rHuEPO-α, recombinant human erythropoietin-α; SE, standard error.

FIG 3.

MMRM analysis of mean change in Hb (g/L) from baseline averaged over weeks 9-13 by subgroups (per-protocol set). CRP, C-reactive protein; EPO, erythropoietin; Hb, hemoglobin; LSM, least-squares mean; MMRM, mixed model repeated measures; rHuEPO-α, recombinant human erythropoietin-α; TSAT, transferrin saturation; ULN, upper limit of normal.