Newsletter: June 2006 | Issue 10: Volume 1

<< back to Newsletter Archive

PRIMARY PREVENTION OF STROKE IN CHILDREN WITH SCD

back to top

In 1997, a Clinical Alert was issued by NHLBI based on the results of a randomized clinical trial (RCT) called the Stroke Prevention in Sickle Cell Anemia — the STOP study. STOP was the first RCT testing any approach to stroke prevention in SCD, and the first RCT involving regular blood transfusion in SCD for any indication.

The key to this study was the selection of children at risk for first stroke using transcanial Doppler (TCD). TCD had been reported to identify a high risk subset with a >10%/year risk of stroke if not treated. This is compared to about a .5%/year risk of stroke in children with SCD generally based on the Cooperative Study of Sickle Cell Disease.

In STOP, children age 2-16 years of age with homozygous SCD and no history of stroke were screened with TCD. Children were eligible for randomization if they had a TCD indicating high stroke risk on two examinations. (TCD estimates blood flow in the arteries of the Circle of Willis using pulsed ultrasound; the higher the blood flow and/or the more narrow or stenotic the artery, the higher the blood flow velocity.) The selection criteria used was 200 cm/sec, which was found in about 10% of children with SCD. These 10% are presumed to be those at highest risk.

The children randomized were either given regular transfusion every 3-4 weeks (simple or exchange methods were acceptable with a Hb S target of 30% or less as a fraction of total Hb measured just before the next transfusion) or standard care, which was either no or only episodic transfusions for specific and accepted short term problems. The main outcome was stroke, as judged by a blinded panel based on history, examination and MRI taken after a possible stroke.

After the planned enrollment of 130 children had been followed for about 2 years, the study was halted by the Monitoring Committee when there were 11 strokes among 67 children randomized to standard care and only one of 63 randomized to transfusion. The stroke risk was 10%/year among those randomized to standard care but <1% in the transfusion arm (p=0.002). These robust results led to the recommendations by NHLBI and the American Stroke Association that all children with homozygous (or Sbeta 0 thal) be screened starting at age 2 years and thereafter at regular intervals (not clearly determined but suggested as every 6-12 month) to look for TCD results that meet the criteria for prophylactic blood transfusion.

In addition to the main finding, a number of secondary reports have been produced that:

1. elucidated rise in ferritin with transfusion;
2. reported low rates of alloimmunization using a matching protocol for blood;
3. showed that silent infarcts and significant non-neurological events were also reduced by regular transfusion;
4. showed that a minority of children with high risk TCD have markedly abnormal magnetic resonance angiograms of the Circle of Willis, and that those higher TCD velocities (>250 cm/sec) rarely drop significantly with treatment, while those with more normal appearing MRA studies may revert to normal TCD with prolonged treatment;
5. showed that transfusion was highly effective even if the child missed some transfusions and that those with Hb S of >30% but <40% were also protected from stroke¹.

 

CAN TRANSFUSION FOR PRIMARY STROKE PREVENTION BE SAFELY STOPPED

back to top

Despite the results of STOP, acceptance of transfusion as a highly effective stroke prevention strategy was limited by the uncertain duration of transfusion with iron overload and other related problems. A second study, again an RCT funded by NHLBI, looked at what happens when transfusion is removed in a low risk subset treated according to the STOP protocol for at least 30 months. In STOP II, 23 centers participated and 79 children of a planned 100 were enrolled before this study too was halted prematurely, this time for safety reasons. In STOP II, children 5 years of age or older who had initially abnormal TCD which reverted to low risk (defined as <170 cm/sec) during a minimum treatment period of 30 months of regular transfusion were randomized to continue transfusion or have it abruptly stopped. After randomization both groups had regular follow up with TCD, with the idea that those that could not tolerate withdrawal would first show return of high risk TCD before becoming at risk for stroke. The study used a composite endpoint which could either be a stroke or a TCD that showed return of high risk.

This surveillance strategy worked, but only partially. The study was halted when there were 16 events, all in the halt transfusion arm. Fourteen were reversions to high risk TCD without stroke and two were strokes. In the two patients with strokes, a high risk TCD was initially observed — but in the 3-4 weeks allowed by the protocol to obtain a second confirmatory high risk TCD before declaring an endpoint and restarting transfusion, two children developed stroke. Thus, while the TCD did give an indication of return of risk, the time to respond was short.

About 25% of those taken off transfusion appeared to tolerate staying off transfusion at least to 1-2 years. However, there were no distinctive features of these children, in that the only indicator of higher risk of reversion to abnormal TCD or stroke after stopping transfusion was a higher average TCD observed before ever initiating regular transfusion — and this was not a powerful indicator (p= 0.05). By one year after stopping transfusion, about half the children had been restarted on transfusion, either due to stroke or stroke risk, or other indications such as pain or acute chest. While the duration of transfusion remains to be determined, STOP II was interpreted as indicating ongoing risk in most cases even after prolonged transfusion, arguing for the need for continued treatment. The search is ongoing for a less intensive “maintenance” therapy that could substitute for transfusion, either as first or follow-on therapy.

CAN HYDROXUREA SUBSTITUTE FOR TRANSFUSION AS EITHER MAINTENANCE OR INITIAL THERAPY TO PREVENT STROKE?

back to top

The evidence base for stroke prevention based on RCTs continues to grow. This summer (2006), enrollment will begin in SWiTCH, a Phase III randomized clinical trial that will compare two treatments for the prevention of a recurrent stroke (secondary stroke prevention) and treatment of iron overload in children with sickle cell anemia. Standard treatment (monthly blood transfusions for stroke prevention and iron chelation for the prevention of iron overload) will be compared to hydroxyurea to prevent recurrent stroke and phlebotomy (the removal of blood) to treat iron overload.

The trial design is based on the experience of Russell Ware and colleagues, who in 2004 reported the outcome of this sort of approach in an open label uncontrolled study of 35 patients with SCD and stroke. At first hydroxuyrea was started two weeks after stopping transfusion, at an initial dose of 15-20 md/kg, then escalated by 5 mg/kg every 8 weeks as tolerated (or up to 30-35 mg/kg). The protocol was changed to incorporate an overlap of both treatments in which transfusion was continued until maximum hydroxyurea dosing was achieved. The latter approach appeared more effective and is being used in SWiTCH. Stroke recurrence was 5.7 events per 100 patients years (7 of 35 had strokes, all cerebral infarctions) — but when considering only the 20 patients with overlap of therapy, there were only two events (estimated rate of 3.6 per 100 patients years). The average HU dose was 26.7 mg/kg, and all but 5 patients were believed to have good compliance. As expected with regular phlebotomy, serum ferritin decreased markedly and liver biopsies were normal.

This group is to be applauded for this systematic open label study (adding to other smaller studies of HU in this setting such as Sumoza A, de Bisotti R, Sumoza D, Fairbanks V. Hydroxyurea (HU) for prevention of recurrent stroke in sickle cell anemia (SCA). Am J Hem. 2002;71:161-165)and for pursuing this approach in a randomized controlled Phase III study.

Also in 2004, the report by Gulbis et al appeared on the experience of using HU in 127 SCD patients in a treatment registry , including 18 at risk for stroke. The Belgian Registry began in 1993, enrolling children with SCD into open label treatment with HU if one or more risk factors for a severe course were present (99 for 3 or more recent vaso-occlusive crises, 21 for acute chest, 7 for stroke, and 1 for TIA) or if the patient had two abnormal TCD’s using the STOP criteria only. Mean age at entry was 6 years (range 8 months-19 years). In this registry the authors identify 34 patients at risk for first stroke based on abnormal TCD. In this group, follow up on HU of about 3 years on average was reported, with only one event (seizure) observed.

If these patients were comparable to the standard care arm of STOP, which they appear to be, there should have been at least 6 strokes if HU was no more effective than standard care. While it is indeed encouraging that primary stroke prevention with HU will work, and may even be comparable to transfusion, it would be highly desirable to see this result more adequately confirmed in a randomized controlled trial.

 

HYDROXYUREA AS MAINTENANCE THERAPY FOR PRIMARY STROKE PREVENTION

back to top

The evidence base for stroke prevention based on RCTs continues to grow. This summer (2006), enrollment will begin in SWiTCH, a Phase III randomized clinical trial that will compare two treatments for the prevention of a recurrent stroke (secondary stroke prevention) and treatment of iron overload in children with sickle cell anemia. Standard treatment (monthly blood transfusions for stroke prevention and iron chelation for the prevention of iron overload) will be compared to hydroxyurea to prevent recurrent stroke and phlebotomy (the removal of blood) to treat iron overload.

The trial design is based on the experience of Russell Ware and colleagues, who in 2004 reported the outcome of this sort of approach in an open label uncontrolled study of 35 patients with SCD and stroke. At first hydroxuyrea was started two weeks after stopping transfusion, at an initial dose of 15-20 md/kg, then escalated by 5 mg/kg every 8 weeks as tolerated (or up to 30-35 mg/kg). The protocol was changed to incorporate an overlap of both treatments in which transfusion was continued until maximum hydroxyurea dosing was achieved. The latter approach appeared more effective and is being used in SWiTCH. Stroke recurrence was 5.7 events per 100 patients years (7 of 35 had strokes, all cerebral infarctions) — but when considering only the 20 patients with overlap of therapy, there were only two events (estimated rate of 3.6 per 100 patients years). The average HU dose was 26.7 mg/kg, and all but 5 patients were believed to have good compliance. As expected with regular phlebotomy, serum ferritin decreased markedly and liver biopsies were normal.

This group is to be applauded for this systematic open label study (adding to other smaller studies of HU in this setting such as Sumoza A, de Bisotti R, Sumoza D, Fairbanks V. Hydroxyurea (HU) for prevention of recurrent stroke in sickle cell anemia (SCA). Am J Hem. 2002;71:161-165)and for pursuing this approach in a randomized controlled Phase III study.

Also in 2004, the report by Gulbis et al appeared on the experience of using HU in 127 SCD patients in a treatment registry , including 18 at risk for stroke. The Belgian Registry began in 1993, enrolling children with SCD into open label treatment with HU if one or more risk factors for a severe course were present (99 for 3 or more recent vaso-occlusive crises, 21 for acute chest, 7 for stroke, and 1 for TIA) or if the patient had two abnormal TCD’s using the STOP criteria only. Mean age at entry was 6 years (range 8 months-19 years). In this registry the authors identify 34 patients at risk for first stroke based on abnormal TCD. In this group, follow up on HU of about 3 years on average was reported, with only one event (seizure) observed.

If these patients were comparable to the standard care arm of STOP, which they appear to be, there should have been at least 6 strokes if HU was no more effective than standard care. While it is indeed encouraging that primary stroke prevention with HU will work, and may even be comparable to transfusion, it would be highly desirable to see this result more adequately confirmed in a randomized controlled trial.