Photo courtesy of ASH
ORLANDO, FL—Hydroxyurea (HU) is noninferior to chronic blood transfusions for reducing the risk of stroke in children with sickle cell disease (SCD), results of the TWiTCH trial suggest.
The trial showed that daily doses of HU lower the transcranial Doppler (TCD) blood velocity in children with SCD to a similar degree as blood transfusions, thereby decreasing the risk of stroke.
Because of these findings, the trial was terminated early, in November of last year.
Last week, results from TWiTCH were presented at the 2015 ASH Annual Meeting (abstract 3*) and published in The Lancet. The study was funded by the National Heart Lung and Blood Institute.
“Stroke . . . is one of the most severe and catastrophic clinical events that occurs in children with sickle cell, with serious motor and cognitive sequelae,” said study investigator and ASH presenter Russell E. Ware, MD, of Cincinnati Children’s Hospital Medical Center in Ohio.
“With the advent of TCD, we now have the ability to identify high-risk children and use chronic transfusion therapy to prevent primary stroke.”
Dr Ware noted that results of the STOP trial showed that chronic transfusion reduced the risk of stroke in high-risk children with SCD, but the transfusions could not be stopped. The STOP 2 trial confirmed this, showing that stopping transfusions led to an increase in TCD blood velocity and stroke risk.
Because transfusions must be continued indefinitely and are associated with morbidity, an alternative stroke prevention strategy is needed, Dr Ware said. He and his colleagues conducted the TWiTCH trial to determine if HU would fit the bill.
Study design
For this phase 3 study, the researchers compared 24 months of transfusions to HU in children with SCD and abnormal TCD velocities. Study enrollment began in September 2011 and ended in April 2013.
All eligible children had received at least 12 months of transfusions prior to enrollment. They were randomized 1:1 to continue receiving transfusions or to receive the maximum-tolerated dose (MTD) of HU.
In the transfusion arm, the goal was to keep hemoglobin S levels below 30%, and iron overload was managed with daily oral chelation.
In the HU arm, the drug was escalated to the MTD, and children continued receiving transfusions until the MTD was achieved. Iron overload was managed with monthly phlebotomy.
The study had a noninferiority design, and the primary endpoint was the 24-month TCD velocity (with a noninferiority margin of 15 cm/sec). TCD velocities were obtained every 12 weeks and reviewed centrally. Local researchers were masked to the results.
Results
In all, 121 children were randomized—61 to transfusions and 60 to HU. Patient characteristics—baseline TCD velocities, age, duration of transfusion, etc.—were well balanced between the treatment arms.
“The average age of the patients was 9 or 10 years old, with about 3 or 4 years of transfusions coming in to the study,” Dr Ware noted.
In the transfusion arm, the children maintained a hemoglobin level of about 9 g/dL and hemoglobin S levels of less than 30%. Most patients received chelation with deferasirox at 26 ±6 mg/kg/day.
In the HU arm, 57 of 60 patients reached the MTD, which was 27 ± 4 mg/kg/day, on average. The median transfusion overlap was 6 months, the average absolute neutrophil count was 3.5 ± 1.6 x 109/L, the average hemoglobin was about 9 g/dL, and fetal hemoglobin rose to about 25%. There were 756 phlebotomy procedures performed in 54 children.
“[In the HU arm,] very quickly after enrollment, the sickle hemoglobin rises, as the transfusions are weaned,” Dr Ware noted.
“Commensurately, the hemoglobin F rises as a protection. The neutrophil count and reticulocyte count drops, and those curves [counts in the HU and transfusion arms] diverge fairly quickly. The serum ferritin [curves] diverged as well.”
Early termination and noninferiority
Interim data analyses were scheduled to take place after one-third of the patients had exited the study and after two-thirds had exited. The first interim analysis demonstrated noninferiority, and the trial was closed early. An analysis was repeated after half of the patients had exited the study, and the trial was terminated.
At that point, 42 children had completed 24 months of treatment in the transfusion arm, 11 patients had truncated treatment, and 8 had early exits. Forty-one patients had completed 24 months of therapy in the HU arm, 13 had truncated treatment, and 6 had early exits.
The final TCD velocity (mean ± standard error) was 143 ± 1.6 cm/sec in the transfusion arm and 138 ± 1.6 cm/sec in the HU arm. The P value for noninferiority (in the intent-to-treat population) was 8.82 x 10-16. By post-hoc analysis, the P value for superiority was 0.023.
Secondary endpoints
There were 29 new neurological events during the trial—12 in the transfusion arm and 17 in the HU arm. There were no new strokes, but there were 6 new transient ischemic attacks—3 in each arm.
There were no new cerebral infarcts in either arm. But there was 1 new progressive vasculopathy in the transfusion arm. And 1 child in the transfusion arm was withdrawn from the study for increasing TCD (>240 cm/sec).
Iron overload improved more in the HU arm than the transfusion arm, with a greater average change in both serum ferritin (P<0.001) and liver iron concentration (P=0.001).
Serious adverse events were more common in the HU arm than the transfusion arm—23 events in 9 patients and 10 events in 6 patients, respectively. But none of these events were thought to be related to study treatment or procedures.
The most common serious adverse event in both groups was vaso-occlusive pain—11 events in 5 HU-treated patients and 3 events in 1 transfusion-treated patient.
Dr Ware noted that there were no secondary leukemias associated with HU in this trial, and there is “a cumulative body of evidence” spanning 20 years that suggests the drug is not carcinogenic in this patient population.
*Data in the abstract differ from data presented at the meeting.