Hydroxyurea lowers TCD—and also stroke?

The management or, more accurately, the prevention of stroke in sickle cell disease (SCD) has evolved considerably since the 1980s, when regular transfusion programs were instituted to prevent recurrent stroke in children with SCD but there was no strategy to prevent the first stroke. Unfortunately, the treatment of stroke itself and the prevention of strokes in adults with SCD have not seen comparable progress.

In this issue of Blood, Zimmerman and colleagues add another plank to an expanding platform of work that may lead to further reductions in stroke in children with SCD, beyond those seen with regular red-cell transfusion.¹  Transcranial Doppler, or TCD, provides a simple risk estimation for stroke in children with SCD.²  The unit of measure is velocity in centimeters per second, estimated by Doppler ultrasound, from the higher of the 2 middle cerebral arteries (MCAs), and it represents a physiological marker of the speed of blood flow in the artery. Blood flow velocity can be increased by reduced lumen diameter, as in stenosis or vasospasm, and/or by increased-volume flow through the artery. In SCD, both factors are often in play. Despite its limitations, TCD provides a number, like LDL or CD4 in other settings, that has proven useful to guide patient treatment, in this case for the selection of children for prophylactic transfusion (as in the Stroke Prevention in Sickle Cell Anemia [STOP] trial)³  to prevent stroke. More recently, TCD has helped established the need for the maintenance of long-term transfusion in children already transfused for many months who were taken off transfusion on a trial basis (as in the Optimizing Stroke Prevention in Sickle Cell Disease [STOP II] trial).⁴  In SCD, TCD velocity has emerged as a reasonable surrogate marker of stroke risk.

Children with SCD already have higher velocities in the MCA than healthy children of similar age based on anemia, which forces an increase in cerebral blood flow and blood flow velocity to deliver more oxygen to the brain. The increase in velocity is roughly proportional to the decrease in total hemoglobin, but when arterial stenosis is also present, the blood flow velocity in the stenotic artery becomes elevated out of proportion to the reduction in hemoglobin. Arbitrary, but now commonly used, cutpoints include: 140 cm/sec—the approximate 50th percentile above which stroke risk starts to increase; 170 to 199 cm/sec—“conditional,” which carries an elevated risk of stroke but for which prophylactic treatment is yet untested; and 200 cm/sec or greater, or “abnormal”—the cutoff used in the STOP studies for testing the effect of regular transfusion. For reference, the velocity in the MCA of a healthy adult is approximately 60 cm/second, and in children without SCD the MCA velocity is 80 to 90 cm/second, depending on age. In the figure, the TCD technique is depicted, along with 2 actual TCD spectra before and after 9 months of regular transfusion in a 7-year-old with SCD.

At some point between 190 cm/sec and 240 cm/sec, probably depending on several variables, it becomes likely that a magnetic resonance angiogram will indicate significant arterial stenosis.⁵ 

Hydroxyurea (HU) is an antimetabolite chemotherapeutic agent known, among other effects, to stimulate fetal hemoglobin production. In the current study, this drug was used in a prospective phase 2 design (no control group) to determine the effect of HU given at maximum tolerated dose (MTD) on TCD velocities. The authors treated 37 children with elevated (> 50th percentile TCD values at baseline), 6 with abnormal TCD velocity who declined to accept the recommended therapy of chronic transfusion and 31 whose TCD velocity did not dictate a therapy based on current evidence. At an MTD of about 27 mg/kg/day, an approximate decrease of 26 to 31 cm/sec was observed at a median 8 months. With therapy, total hemoglobin increased 20%, fetal hemoglobin more than doubled, and overall TCD velocities dropped by about 17%. Although the authors state that there was no correlation between the magnitude of the TCD velocity decrease and the hematocrit increase, this is likely a result of the small sample, as a large body of evidence links TCD velocity to total hemoglobin or hematocrit in healthy or anemic patients.

This study is an important addition to a growing body of work that paves the way toward less invasive or intensive options (than bone marrow transplant or indefinite regular transfusion) for stroke prevention, at least for some patients. Although the STOP study suggested that regular transfusion lowered TCD velocity out of proportion to the increase in total hemoglobin that accompanied the transfusion protocol, this has been hard to determine with certainty, and the precise mechanisms whereby transfusion reduces stroke risk are still not known.

Likewise, from the current and other studies with HU, we do not yet know if HU lowers TCD velocity and, by inference, stroke risk beyond that expected from the increase in hemoglobin. In the end, it may not matter if it can be shown in a randomized controlled trial that, with HU therapy and the accompanying increase in hemoglobin (and fetal hemoglobin), along with a substantial decrease in TCD velocity, the stroke risk also goes down. The study by Zimmerman and colleagues advances the field and helps set the stage for a needed test of HU for primary stroke prevention in a randomized clinical trial, ideally with stratification based on baseline TCD velocity.