The manuscript has been prepared as per CONSORT extension for non-inferiority and equivalence trial.
Study design and patient selection
This was a prospective phase III, randomized, open label, two-arm, parallel group, multi-center, active-controlled, noninferiority clinical study performed at 14 nephrology centers in India during Sept 2012–May 2014. Permuted block randomization schedule with block of size 4 and ratio of 1:1 in the two groups was generated using SAS® software version 9.1. The study was not blinded because of highly varying dosing schedules in the treatment groups. In this study, clinically stable patients (M/F) aged 18–65 years who were on hemodialysis or peritoneal dialysis for at least 4 weeks and had baseline Hb levels < 12 g/dL, were enrolled. Study included patients who were on EPO (not within 1 week before screening) or those who were EPO - naive, and had adequate serum ferritin (≥ 200 ng/mL) and transferrin saturation (≥ 20%) levels. Major exclusion criteria were congestive heart failure, history of uncontrolled hypertension (not amenable two standard drugs over 2 weeks of screening period), severe hyperparathyroidism, pregnant women or lactating mothers, diabetes with HbA1C ≥ 10%, systemic hematological diseases, liver disease, reported hypersensitive to any active study drug substances, or infections. During the study Vitamin B12 and red cell folate concentrations were tested at screening, week 12, end of correction phase and end of maintenance phase. Iron tests i.e. Serum Ferritin, Iron, TIBC, and TSAT were measured at screening, weeks 4, 8, 12, 16, 20, 24 and end of maintenance. Inflammatory marker (CRP) is measured at screening.
The study treatment comprised correction phase (12 to 24 weeks) and maintenance phase (24 to 36 weeks). During 12-24 weeks of correction phase, at baseline, the patients having Hb levels < 10 g/dL with EPO, were switched to either DA-α once weekly (0.45 μg/kg subcutaneous injection manufactured by Hetero Biopharma Limited, India) or EPO thrice weekly (50 IU/kg, Eprex® manufactured by Cilag AG, Switzerland) in allocation ratio of 1:1. The patients with treatment failure, i.e those who had Hb level < 10 g/dL at the end of correction phase (EOC), were discontinued from the study. Following EOC phase, the patients with Hb levels ≥10 g/dL were switched to DA-α for 12 weeks of maintenance phase. If the dialysis patients had Hb level ≥ 10 g/dL at baseline, they directly entered the maintenance phase and were randomized (1:1) to receive DA-α (0.45 μg/kg) once weekly or EPO (50 IU/kg) thrice weekly. In both of the treatment arms, appropriate dose adjustments were made to achieve and maintain patients’ Hb level within the target range, i.e ≥ 1 g/dL increase from baseline Hb, and within the range of 10–12 g/dL during the 36-week study period. Dosage was increased by 25% if a patient’s Hb remained < 10 g/dL even after achieving the target range during the correction phase. In the maintenance phase, if the patient’s Hb levels went above the target range (≥ 11.5 g/dL) for two consecutive weekly assessments, the dosage was decreased by 25%. The dosage was increased by 25%, if after achieving the target range (10–12 g/dL), patient’s Hb level was below 10 g/dL. There was no specific rescue therapy defined in the study protocol. However, Iron supplements (oral/IV) were allowed as concomitant medication to prevent apparent iron deficiency. For patients with serum ferritin values < 100 μg/L or ≥ 100 μg/L, the IV iron dosing regimen was determined per the individual center’s treatment protocol.
Efficacy and safety assessments
Hemoglobin (Hb) levels were assessed throughout the correction and maintenance phases. Primary efficacy endpoint included the mean change in Hb level from baseline at the first evaluation visit (EOC) in patients treated with EPO versus DA-α. Secondary efficacy endpoints were Hb variability during correction phase, mean change in Hb levels from baseline to week-4 and week-36 (EOM), proportion of patients achieving the Hb target (defined as Hb increase ≥ 1 g/dL from baseline) at EOC, mean DA-α dose, proportion of patients who could maintain the target Hb of 10–12 g/dL at EOM, and time to initial achievement of Hb target. Safety assessment included serum chemistry, complete blood count, and urinalysis measured at baseline and after treatment. Adverse events (AEs) were monitored at each study visit. Safety endpoints were the incidence of treatmentemergent adverse events (TEAEs) and immunogenicity as assessed by anti-drug antibody titers using ELISA methods. Samples for immunogenicity to DA-α/EPO were withdrawn within 1 h of dosing before initial dosing, i.e Day 1 of week 1 and on day 1 of weeks 5, 13, 25 and EOM. Since the screening time and evaluation periods of the various patient subgroups were different, the time points for immunogenicity varied accordingly.
Statistical analyses
Assuming an estimated difference between the treatments of − 0.5 g/dL in mean change in Hb levels with a noninferiority margin at − 1.0 g/dL, to achieve 80% power for the non-inferiority test, a total of 100 patients are required to be allocated in the ratio of 1:1 to either DA-α (n = 50) or EPO (n = 50). A final sample size of 126 patients (63 patients in each treatment arm) was derived assuming dropout rate at 20%. A designated statistician (Sristek, India) generated the allocation sequence and assigned participants to their groups, and investigators at 14 nephrology sites enrolled participants according to this sequence. Efficacy analysis included both the intent to treat (ITT) and per protocol (PP) population. ITT population comprised randomized patients who received at least one dose of the study drug, with baseline and at least one efficacy assessment available during the evaluation period. PP set included patients who completed all the study visits as defined in the protocol without major protocol deviations. Safety population comprised all randomized patients who received at least one dose of the study drug. The between group difference in the mean Hb change from baseline to first evaluation visit was analyzed using analysis of covariance (ANCOVA). Treatment was considered as main effect and baseline Hb levels as the covariate in the model. The two-sided 95% CI for difference in mean Hb for treatments was calculated to assess the non-inferiority, and if it was above the non-inferiority margin of − 1.0 g/dL, the non-inferiority was accepted. Two sample t-tests at 5% level of significance were performed to compare the difference of change in Hb level from baseline to week-4 and week-36 EOM. For all tests, p-value of ≤ 0.05 was considered as statistically significant. Logistic regression was carried out to compare the proportion of patients achieving the Hb target (≥ 1 g/dL from baseline and Hb concentration of 10–12 g/dL) at the end of first evaluation visit and EOM. The time to the initial achievement of Hb target at the EOC was estimated using Kaplan-Meier. SAS® Version 9.4 (SAS Institute Inc., USA) was used to perform all statistical analyses.