TCD with Transfusions Changing to Hydroxyurea (TWiTCH): Hydroxyurea Therapy As an Alternative to Transfusions for Primary Stroke Prevention in Children with Sickle Cell Anemia

Transcranial Doppler (TCD) screening in children with sickle cell anemia (SCA) identifies abnormally elevated cerebral artery flow velocities that confer an elevated risk for primary stroke. Chronic transfusions offer effective stroke prophylaxis in this setting, but must be continued indefinitely and lead to transfusional iron overload. An alternative treatment strategy that offers similar effective protection against primary stroke, and provides control of iron overload, is needed. TCD With Transfusions Changing to Hydroxyurea (TWiTCH, NCT01425307) was an NHLBI-funded Phase III multicenter randomized clinical trial comparing 24-months of standard treatment (transfusions) to alternative treatment (hydroxyurea) in children with SCA and abnormal TCD velocities. All eligible children had received at least 12 months of transfusions. TWiTCH had a non-inferiority trial design; the primary study endpoint was the 24-month TCD velocity obtained from a linear mixed model, controlling for baseline (enrollment) values, with a non-inferiority margin of 15 cm/sec. The transfusion arm maintained children at HbS <30%; an elevated liver iron concentration (LIC) identified by R2 MRI FerriScan® was managed with chelation. The hydroxyurea arm included an overlap period with transfusions until a stable maximum tolerated dose (MTD) of hydroxyurea was reached; transfusions were then replaced by serial phlebotomy to reduce iron overload. In both arms, TCD velocities were obtained every 12 weeks and reviewed centrally, with local investigators masked to the results. A centralized TCD alert algorithm monitored changes from enrollment velocities. A total of 159 children were enrolled but 38 failed screening due primarily to severe vasculopathy on brain MRA or inadequate TCD exams; 121 children were randomized (61 to transfusions, 60 to hydroxyurea) with balanced characteristics including enrollment maximum TCD velocities (145 ± 21 versus 145 ± 26 cm/sec), age, duration of transfusions, serum ferritin, and LIC. Study participants randomized to transfusions maintained an average HbS <30% throughout the study, while those on hydroxyurea reached MTD after 7 ± 2 months at an average dose of 27 mg/kg/day, with expected hematological changes including HbF ~25% throughout the treatment period. After 37% of the participants exited the study, a scheduled interim analysis suggested the primary study endpoint was likely to be achieved. NHLBI allowed the study to continue until 50% of the children exited, at which time the statistical analysis was confirmed and the study was terminated; all remaining participants moved to the exit phase. The final analysis included 42 on the transfusion arm who completed all treatment, 11 with truncated treatment, and 8 withdrawn; the hydroxyurea arm included 41 who completed all treatment, 13 with truncated treatment, and 6 withdrawn. The final calculated TCD velocities (mean ± standard error) in the transfusion and hydroxyurea arms were 143 ± 1.6 and 138 ± 1.6 cm/sec, respectively; by intention-to-treat analysis, the p-value for non-inferiority = 8.82 x 10^-16 and by post-hoc analysis the p-value for superiority = 0.046. Among 29 new neurological events, all centrally adjudicated by masked reviewers, there were no strokes but 6 transient ischemic attacks (3 in each arm). One child (transfusion arm) was withdrawn per the TCD alert algorithm after developing on-study TCD velocities >240 cm/sec. Exit brain MRI/MRA exams documented no new parenchymal abnormalities but one child (transfusion arm) developed new vasculopathy. Sickle cell related serious adverse events were more common in the hydroxyurea arm than the transfusion arm (23 to 15), but none was related to study treatment or study procedures. Iron overload improved more in the hydroxyurea arm than in the transfusion arm, with a greater average change in serum ferritin (-1085 compared to -38 ng/mL, p<0.001) and LIC (average -1.9 compared to +2.4 mg/g dry weight liver, p=0.001). In the multicenter Phase III TWiTCH trial, which treated children with SCA and abnormal TCD velocities but without severe MRA vasculopathy, hydroxyurea at MTD was non-inferior and possibly superior to chronic transfusions for maintaining TCD velocities. Serial phlebotomy effectively managed iron overload. Hydroxyurea may represent an effective alternative to indefinite transfusions for the prevention of primary stroke in this high risk population.