TWiTCH: Itching to Find an Alternative to Transfusions for Prevention of Stroke in Sickle Cell Anemia

Transcranial Doppler (TCD) With Transfusions Changing to Hydroxyurea (TWiTCH)

NCT01425307

Cincinnati Children’s Hospital Medical Center

National Heart, Lung, and Blood Institute (NHLBI)

This is a national trial that includes 25 participating centers.

160 patients

This is a phase III, randomized, non-inferiority trial. To be eligible, children ages 4 to 15.99 must have sickle cell anemia (Hb SS) or sickle-b⁰-thalassemia (Hb Sb⁰) and have been on a chronic transfusion program for at least 12 months for primary stroke prophylaxis due to abnormal transcranial Doppler (TCD) velocities (≥200 cm/sec). The main exclusion criteria include prior overt stroke, transient ischemic attack, and severe cerebral vasculopathy or moya-moya disease. Participants are randomized to 24 months of alternative therapy (hydroxyurea) or continued standard therapy (transfusions). The primary endpoint is the maximum TCD velocity on the index side, which is the cerebral hemisphere with the higher velocity. The non-inferiority margin is 15 cm/sec.

Without prevention, by 18 years of age, overt stroke occurs in 11 percent of children with Hb SS. Children with Hb SS who have the highest risk of stroke have abnormally increased cerebral arterial blood flow velocities, which can be measured by TCD. An abnormal TCD examination (≥200 cm/sec) confers an approximately 10 percent yearly risk of stroke for the three years following the test. The Stroke Prevention Trial (STOP) in sickle cell anemia, published in 1998, showed that chronic red blood cell transfusions decreased the rate of first stroke in children with an abnormal TCD by 92 percent compared with no transfusions (observation) (Adams RJ et al. N Engl J Med. 1998;339:5-11). Using TCD, most pediatric sickle cell programs in the United States now screen children with Hb SS or Hb Sb⁰ (usually 2-16 years of age) and initiate chronic transfusions for high-risk children. This strategy has proved effective, as the incidence of hospitalization for overt stroke in children with Hb SS in the United States has decreased by 45 percent in the decade following publication of the STOP trial (McCavit TL et al. Pediatr Blood Cancer. 2013;60:823-827). Although this management approach represents an important advance, most children with abnormal TCD velocities would not have a stroke even if untreated, and chronic transfusion therapy is burdensome, is indefinite in duration, and quickly necessitates iron chelation therapy. Accordingly, alternative management strategies are desirable and continue to be evaluated.

One such alternative is to simply stop transfusions after several years, in the hope that the high-risk period for stroke has passed. However, the STOP 2 trial showed that discontinuation of transfusions after ≥30 months resulted in a high rate of both reversion to abnormal TCD velocities and risk of stroke (Adams RJ et al. N Engl J Med. 2005;353:2769-2778). So, truncation of transfusion is not a reasonable strategy for most children. Another alternative is the possibility that hydroxyurea can be given instead of chronic transfusions for primary stroke prevention; this alternative is now being tested in the TWiTCH trial. The TWiTCH trial builds upon the recently completed SWiTCH trial, which compared alternative treatment (daily hydroxyurea for stroke prevention with monthly phlebotomy to manage iron overload) and standard treatment (chronic transfusions for stroke prevention and daily chelation for iron overload) for secondary stroke prevention. The SWiTCH trial was stopped early due to futility of reaching its composite primary endpoint, and there was an excess of recurrent strokes in the hydroxyurea/phlebotomy arm (Ware RE et al. Blood. 2012;119:3925-3932). The severity of baseline cerebral vasculopathy in patients enrolled in the SWiTCH trial (the study enrolled subjects with sickle cell anemia, previous stroke, and ≥ 18 months of transfusions with documented iron overload) may have reduced the effectiveness of hydroxyureafor secondary stroke prevention. In TWiTCH, primary stroke prevention, an arguably different goal, is being studied, and severe cerebral vasculopathy is an exclusion criterion.

TWiTCH is not designed to determine whether hydroxyurea is better than chronic transfusions as a classical superiority trial would. Nor is TWiTCH designed to determine whether hydroxyurea is not unacceptably different than chronic transfusions as an equivalence trial would. Rather, as a non-inferiority trial, TWiTCH is designed to determine whether hydroxyurea is not unacceptably worse (even possibly less efficacious, but within limits) than chronic transfusions for primary stroke prevention. What, then, is the rationale for conducting a trial to test a therapy that the investigators realize may not work as well? In an excellent review of non-inferiority trials, Schumi and Wittes write: “A non-inferiority trial is reasonable when a new treatment has some property sufficiently favorable that physicians, and their patients, would be willing to sacrifice some degree of benefit relative to … [currently accepted therapy]. The advantage could be reduced cost, improved ease of use or dosing schedule … or an improved safety profile. The benefit given up in exchange for these advantages, however, should not be so large that patients and physicians are not willing to use the new … [treatment].”¹ 

Hydroxyurea as the “new” treatment in this clinical setting has such favorable criteria. It is less costly, easier to use, and probably safer than lifelong chronic transfusions. What about the benefit given up? TCD velocity, which is a surrogate for risk of stroke, will increase for many patients when switching from chronic transfusions to hydroxyurea, in large part because the hemoglobin concentration will decrease once chronic transfusions are stopped. Hemoglobin concentration is a major determinant of TCD velocity. The TWiTCH investigators set as their non-inferiority margin an increase in TCD velocity of up to 15 cm/sec. That is, as long as the mean TCD velocity in the hydroxyurea arm is not higher than 15 cm/sec above the mean of the transfusion arm (measured pre-transfusion) at the end of the 24-month treatment period, hydroxyurea will be declared non-inferior to chronic transfusions. The margin of 15 cm/sec was chosen because it is the within-subject standard deviation of TCD velocities from past studies. The TWiTCH investigators argue that a change of this magnitude is within the limits of precision for TCD measurements, so the additional risk of stroke would be negligible.

One limitation of the trial’s primary endpoint, TCD velocity, is that it is a surrogate outcome. The outcome we are really interested in, however challenging to study in a clinical trial, is stroke. The predictive capacity of TCD velocity for overt stroke is well characterized in untreated Hb SS patients, but the meaning (e.g., the positive or negative predictive value for overt stroke) of any particular TCD velocity in a patient on hydroxyurea or chronic transfusions is not well characterized, and its absolute value will be lower than the untreated patient. So, when the TWiTCH study results are finally available, we need to remember that changes in TCD velocity do not always predict the risk of stroke. Hopefully, any sacrifice in benefit (stroke prevention) using hydroxyurea will be small compared with its many practical advantages over indefinite transfusions.