Background Hydroxyurea is widely used in the first-line treatment of sickle cell anaemia; however, its mildly myelosuppressive effects when used at therapeutic doses have led to suggestions that treatment with hydroxyurea might increase the risk of infection. To address this question, we have investigated the incidence of clinically diagnosed infections among children within the Realizing Effectiveness Across Continents with Hydroxyurea (REACH) trial during various phases of hydroxyurea treatment. Methods REACH (NCT01966731 and NCT06171217) is an ongoing, phase 1/2, open-label, multicentre treatment trial of children with sickle cell anaemia with 2 months of screening followed by oral hydroxyurea at a fixed dose for 6 months and then dose escalation to the maximum tolerated dose (MTD). The current international, retrospective cohort study included children aged 1–10 years with sickle cell anaemia who were recruited and enrolled into REACH from four hospitals in Angola, DR Congo, Kenya, and Uganda and who had previously been followed up within the trial. Clinically diagnosed infections were characterised by likely aetiology and severity and incidence rate ratios (IRR) were then computed across time and during dosing phases as protocol-specified outcomes. The primary outcome of the current analysis was the incidence of clinically diagnosed infections, which was stratified by hydroxyurea treatment phase and assessed in all children who started fixed-dose hydroxyurea treatment. Rates of each clinically diagnosed infection (per 100 person-years of observation [PYO]) were calculated by age group and dosing phase. Findings Between July 4, 2014, and Nov 11, 2016, 635 children were enrolled in the trial. 606 children (296 [49%] female and 310 [51%] male) completed screening and began treatment with hydroxyurea. Overall, 5493 infections occurred across 5638 PYO, with an incidence rate of 97·4/100 PYO. Among non-malarial infections, rates declined from the screening to fixed-dose phase (adjusted IRR 0·81 [95% CI 0·67–0·98]) and decreased further from the screening to the MTD phase (0·51 [0·43–0·61]; p<0·0001). Statistically significant decreases across the fixed-dose, dose-escalation, and MTD phases, respectively, were observed in the adjusted IRRs for upper respiratory tract infections (IRR 0·84 [95% CI 0·62–1·14], 0·77 [0·58–1·03], and 0·64 [0·47–0·88]; p=0·0051), lower respiratory tract infections (1·26 [0·69–2·31], 0·60 [0·32–1·12], and 0·45 [0·23–0·89]; p=0·021), bacterial skin or nail infections (0·58 [0·26–1·30], 0·32 [0·15–0·68], and 0·31 [0·14–0·66]; p=0·0025), and non-specific infectious syndromes (0·80 [0·58–1·11], 0·47 [0·35–0·63], and 0·39 [0·28–0·53]; p<0·0001). A similar decrease was observed for the adjusted IRR of grade 3 or higher infections across the fixed-dose, dose-escalation, and MTD phases, respectively (0·65 [0·42–1·00], 0·20 [0·13–0·32], and 0·48 [0·32–0·73]; p=0·0006). Interpretation In the large REACH cohort, the incidence rate or severity of clinically diagnosed infections in children with sickle cell anaemia living in sub-Saharan Africa, was not increased during the hydroxyurea treatment phases of this trial. Our results provide reassurance that, even when administered at MTD, hydroxyurea does not put children with sickle cell anaemia at increased risk of illness or death from infections. Funding US National Heart, Lung, and Blood Institute. Translations For the French, Portuguese and Swahili translations of the abstract see Supplementary Materials section.
