Abstract 1044


Stroke With Transfusions Changing to Hydroxyurea (SWiTCH Clinical Trials.gov NCT00122980), an NHLBI-sponsored Phase III multicenter trial, compared chronic blood transfusions/chelation to hydroxyurea/phlebotomy for the reduction of recurrent stroke and improvement in iron overload management in children with sickle cell anemia (SCA) and history of overt stroke. To date, however, phlebotomy to manage iron overload has not been commonly performed in children, especially those with SCA.


To describe the experience with SWiTCH phlebotomy procedures, including success rate, associated adverse events, and effect on liver iron stores.


Quantitative liver iron concentration (LIC) was measured by liver biopsy at study entry. Only subjects with LIC > 5 mg Fe/gram dry weight liver (DWL) were eligible for randomization. Those randomized to hydroxyurea/phlebotomy received decreasing volumes of monthly transfusion during hydroxyurea dose escalation, which lasted 4–9 months. Phlebotomy was performed every 4±1 weeks after discontinuation of transfusions. The prescribed phlebotomy volume was 10 mL/kg (maximum 500 mL) for Hb ≥ 8.0 gm/dL, and 5 mL/kg for Hb 7.0–7.9 gm/dL. Phlebotomy was held if Hb was <7.0 gm/dL. Phlebotomy was performed over 30 minutes with immediate normal saline replacement, typically using peripheral venous access. Exit LIC by liver biopsy was obtained in those completing 30 months of therapy. Ferritin was monitored monthly in all subjects using a centralized laboratory.


Sixty-seven children (mean age 13.0 ± 4.0 years; range 5.2–19.0 years) with history of previous stroke and transfusion therapy for an average of 7.4 ± 3.8 years (range 1.5–15.5 years) were randomized to the hydroxyurea/phlebotomy arm. Most of them had also received chelation therapy: 47 (71%) with deferoxamine for an average of 4.8 ± 3.2 years, and 57 (86%) with deferasirox for 1.5 ± 0.8 years prior to study entry. Their average entry LIC was 16.5 ± 9.4 mg/gram DWL. Sixty of 67 children (90%) successfully transitioned to hydroxyurea after 7.2 ± 2.4 months of transfusion overlap; one subject had a stroke during overlap and six failed to demonstrate adequate response/compliance to hydroxyurea to safely discontinue transfusions. These 60 subjects received an average of 8 ± 3 transfusions providing 63 ± 44 mL/kg PRBCs before completing transition and commencing phlebotomy, and 3 ± 3 transfusions providing 19 ± 20 mL/kg PRBCs after starting phlebotomy, for various clinical indications. During the course of the study, a total of 935 phlebotomies were performed (mean 16 per subject) removing an average total volume of 127 ± 74 mL/kg per subject. The mean pre-phlebotomy Hb level on hydroxyurea (9.1 gm/dL) was not significantly different than the mean pre-transfusion Hb during the transfusion overlap period (9.0 gm/dL). Mean ferritin for these 60 subjects on the hydroxyurea/phlebotomy arm decreased from 3523 ± 2150 ng/mL at study entry to 2227 ± 1646 ng/mL (p<0.0001) at exit; and decreased in 50 of 60 subjects. For the 23 patients on the hydroxyurea/phlebotomy arm who completed 30 months of study treatment, the average LIC was unchanged (18.5 mg Fe/gram DWL at entry compared to 18.1 mg Fe/gram at exit, p=0.817). However, average ferritin level for these subjects was significantly lower at exit (4216 ± 2799 ng/mL vs 2356 ± 2032 ng/mL, respectively, p=0.0003). Of 968 protocol-directed phlebotomy procedures, 935 (97%) were performed; 94% of which were at full prescribed volume. Of the 33 phlebotomy procedures that were not performed, 11 were held due to Hb < 7.0 gm/dL and 9 due to poor venous access. There were only 33 grade 2 adverse events (3.5% prevalence) reported in 12 subjects and no serious adverse events. The most common complication was hypotension (9 events; 5 subjects) followed by dizziness, syncope, headache and weakness. Six subjects had a recurrent stroke but there was no temporal relationship to the phlebotomy procedures.


Therapeutic phlebotomies were well-tolerated and did not result in worsening anemia or stroke recurrence in this cohort of children with SCA and previous stroke switched to hydroxyurea. Although ferritin levels decreased significantly, we did not demonstrate an overall decrease in LIC in this heavily iron overloaded cohort, most likely due to continued iron loading with transfusions in the overlap period and subsequent short duration of phlebotomy.


Off Label Use: Use of hydroxyurea in children with sickle cell anemia.

This icon denotes a clinically relevant abstract

Author notes


Asterisk with author names denotes non-ASH members.