Randomized Controlled Trial of the Efficacy and Safety of Deferiprone in Iron-Overloaded Patients with Sickle Cell Disease or Other Anemias

Background: Patients with sickle cell disease (SCD) or other rare anemias whose care includes chronic blood transfusions must receive iron chelation to prevent the morbidity of iron overload. Currently, only deferoxamine (DFO) and deferasirox (DFX) are approved chelators in these patient populations. This randomized open-label trial evaluated if the efficacy of deferiprone (DFP) was non-inferior to DFO. DFO was used as the comparator product since DFX was not approved as first-line treatment for SCD at trial initiation. Methods: Participants at 27 sites in 8 countries were randomized in a 2:1 ratio to receive either DFP or DFO for up to 12 months. Those with lower transfusional iron input and/or less severe iron load were prescribed either DFP 25 mg/kg of body weight t.i.d. or DFO 20 mg/kg (children) or 40 mg/kg (adults); those with higher iron input and/or more severe iron load received either DFP 33 mg/kg t.i.d. or DFO up to 40 mg/kg (children) or 50 mg/kg (adults). Dosages could be adjusted over the course of the trial if necessary. Efficacy endpoints were the changes from baseline in liver iron concentration (LIC), cardiac iron, and serum ferritin (SF) at Month 12. The primary endpoint was based on LIC, and for the demonstration of non-inferiority of DFP to DFO, the upper limit of the 95% confidence interval for the difference between treatments had to be no more than 2 mg/g dry weight (dw). All patients had their neutrophil count monitored weekly, whereas other safety assessments and compliance with study therapy were evaluated monthly. Acceptable compliance was defined as taking 80% to 120% of the prescribed dosage. Results: A total of 228 of the targeted 300 patients were dosed with 152 receiving DFP and 76 receiving DFO, to assess non-inferiority. There were no significant differences between the groups in any demographic measures: in each treatment group, 84% of patients had SCD and the remainder had other, rarer forms of transfusion-dependent anemia. Mean age at enrollment was 16.9 years (± 9.6); 53.1% of patients were male; and 77.2% were white, 16.2% black, and 6.6% multi-racial. Over the course of the study, 69% of patients in the DFP group and 79% in the DFO group had acceptable compliance with treatment. Based on the Pocock's α spending function, a more stringent confidence level of 96.01% was applied to the calculation of confidence interval for the evaluation of non-inferiority. For the primary efficacy endpoint, the least squares (LS) mea