Adding Hydroxyurea to Transfusion Shows Early Promise for Sickle Cell Stroke Patients, But Evidence Remains Thin

Nine young people with sickle cell disease finished a full year of a treatment combination that had never before been tested this way in a prospective trial. All nine had been receiving regular simple blood transfusions to prevent stroke. During the trial year, they also took hydroxyurea, the main drug made specifically for sickle cell disease. The study, led by Nickel and colleagues, is the subject of a commentary in the Annals of Translational Medicine that calls the results cautiously reassuring and the evidence still thin.

The problem the trial addresses is a familiar one. Stroke is among the most devastating complications of sickle cell disease. Over three decades of landmark trials, the cumulative incidence of stroke in high-resource countries has fallen from 10% to 1%, largely because transcranial Doppler screening identifies children with abnormal cerebral blood flow velocity, who then start chronic transfusion. Some patients without high-grade vasculopathy and with a sustained response to hydroxyurea can switch from transfusion to the drug for primary prevention. For secondary prevention, in patients who have not had a stem cell transplant, transfusion remains the main approach.

Transfusion has limits, though. In the TWITCH study, three patients without underlying vasculopathy on the transfusion arm had a transient ischemic attack, and one developed new vasculopathy. Other studies show recurrent strokes or silent infarcts in patients with cerebral vasculopathy or Moyamoya disease, even when pre-transfusion hemoglobin S stays under 30%. Transfusion also carries alloimmunization and iron overload, along with cost and access problems in rich and poor settings alike. The commentary frames two questions: how to intensify non-curative therapy for patients on transfusion alone, and how to safely reduce how much blood they need.

The trial enrolled people aged 2 to 25 with SS or Sβ0 thalassemia who had been on simple transfusions for at least a year. Exclusions included poor adherence (hemoglobin S above 45% at any point in the prior year, or intervals longer than five weeks between transfusions), hydroxyurea use in the previous year, pregnancy, and abnormal labs at enrollment. The primary aims were safety, feasibility, and whether the combination cut transfusion needs. The target dose was one that brought the absolute neutrophil count below 4×10⁹/L, with pre-transfusion hemoglobin of 10–11 g/dL and hemoglobin S of 30–45%.

Of the nine who completed the year at target dose, 66% cut their transfused blood volume by more than 10 mL/kg per year compared with their earlier transfusion-only years. That reduction tracked with better hydroxyurea adherence and higher fetal hemoglobin. Pre-transfusion hemoglobin rose and hemolysis markers fell. Median pre-transfusion hemoglobin S was higher on the combination, reaching 31.7% (IQR 25.5 to 40.8), which the commentary attributes to smaller transfusion volumes and the looser 30–45% target. No new or worsening cerebral ischemic events or vasculopathy appeared, and the cohort included patients with vasculopathy, unlike earlier blood-conservation studies that mostly excluded them.

The commentary does not oversell this. A 10 mL/kg annual saving is roughly one transfusion session and a small share of total blood given, which may not justify tolerating higher hemoglobin S without further study. The worry is that higher levels could worsen cerebrovascular injury over time; whether the benefits of fetal hemoglobin offset that remains unknown. The authors also note that earlier retrospective combination studies were not aimed at cutting transfusion volume.

The inflammation findings are a second thread. Hydroxyurea lowers white cells, neutrophils, and platelets, reduces pro-inflammatory cytokines, and suppresses a pro-adhesive neutrophil subset. In the trial, adding it to transfusion lowered white cell and neutrophil counts and levels of BDNF, sVCAM-1, and PDGF-AA, cytokines linked to prior stroke and higher Doppler velocities. Prior evidence conflicts: Dembélé and colleagues found no cytokine or aged-neutrophil differences across treatment groups but did not adjust for age, race, neutrophil count, or iron overload, while Terrigno and colleagues, in matched cohorts and adjusting for iron burden, found lower neutrophil and cytokine levels with hydroxyurea than with transfusion. Separately, one report links concurrent hydroxyurea to lower venous thromboembolism risk in transfused patients with central lines.

The commentary judges the risks of adding hydroxyurea minimal and the approach worth serious consideration. It calls for prospective studies in patients with cerebral vasculopathy and for work on progressive heart damage despite single therapy. It recommends that early studies add hydroxyurea without changing hemoglobin S targets and use clinically meaningful endpoints. Two of the commentary's authors disclose consulting fees from Pfizer, among other sources.