SWiTCH 2012.

Study characteristics
Methods	Multicentre randomised controlled non‐inferiority trial conducted in 26 paediatric sickle cell centres in the USA. Total duration of trial treatment was 30 months after randomisation, with a final trial visit scheduled 6 months after discontinuation of trial treatments.
Participants	Inclusion criteria: paediatric participants with severe forms of SCA (HbSS, HbS/βº‐thalassemia, HbS/OArab); age range of 5.0 to 18.9 years, inclusive, at the time of enrolment; initial (primary) completed overt clinical stroke after the age 12 months with documented infarction on brain CT or MRI; at least 18 months of chronic monthly erythrocyte transfusions since primary stroke; transfusional iron overload, defined as a previously documented liver iron concentration ≥ 5.0 mg Fe per g of dry weight liver or serum ferritin ≥ 500 ng/mL on 2 independent measurements; adequate monthly erythrocyte transfusions with average HbS ≤ 45% (the upper limit of the established academic community standard) for the past 6 months before enrolment; parent or guardian willing and able to provide informed consent with verbal or written assent from the child (< 18 years of age), and subject willing and able to provide informed consent (≥ 18 years of age); ability to comply with trial related treatments, evaluations, and follow‐up. Exclusion criteria: inability to receive or tolerate chronic RBC transfusion therapy; inability to take or tolerate daily oral hydroxyurea; clinical and laboratory evidence of hypersplenism (temporary); abnormal laboratory values at initial evaluation (temporary); current participation in other therapeutic clinical trials; current use of other therapeutic agents for sickle cell disease (e.g. arginine, decitabine, magnesium); any condition or chronic illness, such as a positive tuberculin (PPD) test, which in the opinion of the investigator makes participation ill‐advised; inability or unwillingness to complete required screening studies, including blood tests, brain MRI/MRA, and liver biopsy; a sibling enrolled in SWiTCH. Participant flow: screened: N = 202; enrolled: N = 161; randomised: N = 134 Transfusion + chelation: N = 66 (one moved before starting trial treatment) Sex: Male: 31 (47%) Age: Mean (SD): 13.3 (3.8) Phenotype: HbSS: 66 (100%) Previous recurrent stroke: 4 (6%) History of TIA: 11 (17%) Infarction: 65 (98%) Vasculopathy: 54 (82%) Moya‐moya: 5 (8%) Liver iron content (LIC), mg Fe/g dw liver median (IQR): 14.5 (9.5 to 23.3) Serum ferritin, ng/mL median (IQR): 3282.0 (2321.0 to 4306.0) HbS% median (IQR): 27.0 (21.2 to 38.6) HbF% median (IQR): 1.7 (1.0 to 2.5) Alpha thalassaemia: not reported Hydroxyurea + phlebotomy: N = 67 Sex: male: 41 (61%) Age: Mean (SD): 13.0 (4.0) Phenotype: HbSS: 66 (99%) Previous recurrent stroke: 10 (15%) History of TIA: 10 (15%) Infarction: 65 (98%) Vasculopathy: 53 (79%) Moya‐moya: 11 (16%) Liver iron content (LIC), mg Fe/g dw liver median (IQR): 13.9 (8.7 to 22.9) Serum ferritin, ng/mL median (IQR): 3346.0 (2202.0 to 4682.0) HbS% median (IQR): 30.3 (23.8 to 39.6) HbF% median (IQR): 1.4 (0.8 to 2.2) Alpha thalassaemia: not reported
Interventions	Standard treatment (transfusion + chelation): N = 66 For standard treatment (blood transfusion + iron chelation) participants received monthly blood transfusions designed to maintain 30% HbS, with local discretion regarding transfusion type (e.g., simple or erythrocytapheresis). Red cell component: not reported. Red cell matching: not reported. Iron chelation: daily iron chelation. Hydroxyurea + phlebotomy: N = 67 Participants randomised to hydroxyurea + phlebotomy commenced hydroxyurea at 20 mg/kg/d with stepwise escalation to MTD. Transfusions continued for 4 to 9 months during an overlap phase designed to protect against recurrent stroke during hydroxyurea dose escalation. Once MTD was reached and transfusions were discontinued, phlebotomy commenced with a target of 10 mL/kg (maximum volume, 500 mL) blood removed monthly to reduce iron burden.
Outcomes	Primary outcome: composite primary endpoint of secondary stroke recurrence rate and quantitative liver iron concentration. Secondary outcomes: non‐stroke neurological events, non‐neurological sickle cell clinical events, quality of life evaluation, and measures of organ function.
Notes	Funding: supported by National Heart, Lung, and Blood Institute grants U01‐HL078787 (R.E.W.) and U01‐HL078987 (R.W.H.). Declarations of interest: the authors declare no competing financial interests. Trial registration: ClinicalTrials.gov NCT00122980. Mean length of follow‐up: 6‐month follow‐up. Power calculation: not stated. Analysis: because reduction in LIC was not superior on hydroxyurea/phlebotomy, the DSMB concluded that the composite primary trial end point would not be met and recommended trial closure. NHLBI closed SWiTCH N = 40 did not complete treatment phase in transfusion/iron chelation arm and N = 43 did not complete treatment phase in hydroxyurea/phlebotomy arm.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of sequence generation not reported.
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment not reported.
Blinding of participants and personnel (performance bias) All outcomes	High risk	By the nature of the trial treatments used to prevent recurrent stroke (blood transfusions vs hydroxyurea), it is impractical to make SWiTCH a blinded (masked) trial.
Blinding of outcome assessment (detection bias) Stroke and TIA	Low risk	The inclusive independent stroke adjudication process for all suspected new neurological events is a novel feature of the trial. Stroke recurrence is a primary trial endpoint but also is a critical safety endpoint for the SWiTCH trial. Accordingly, it was necessary to develop an inclusive process by which all potential stroke events were recognized and systematically adjudicated using a standardized protocol and masked consultants. Participants who develop any acute neurological change are promptly evaluated for possible stroke. In addition, site personnel are provided with a written script to use at each interval clinic visit, to ensure that subjects and families are asked each month about any signs and symptoms of stroke. After a new neurological event is suspected, the stroke adjudication process begins. The clinical history and neurological exam are reviewed by 3 independent neurologists without knowledge of the imaging findings. Simultaneously, the radiological evaluation is reviewed by 3 independent masked neuroradiologists without knowledge of the clinical history or neurological examination. Only after their independent consensus opinions are formed are these two opinions reconciled into a final stroke adjudication decision; a diagnosis of stroke requires new neurological findings with corresponding radiological changes.
Blinding of outcome assessment (detection bias) All outcomes apart from stroke or all‐cause mortality	High risk	The SWiTCH principal investigator was masked to all treatment‐specific results, including laboratory tests and clinical events. In addition, all investigators at the peripheral clinical sites are masked to trial treatment results outside of their own clinical centre. Unblinded trial.
Incomplete outcome data (attrition bias) All outcomes	Low risk	The primary statistical analyses of efficacy and safety will be performed on the ITT population, which consists of all participants who were randomised to a trial treatment and for whom outcome data are available.
Selective reporting (reporting bias)	High risk	Several secondary outcomes not reported (i.e. quality of life, growth and development, organ damage, transfusion‐related, chelation‐related and phlebotomy related complications)
Other bias	Unclear risk	the DSMB concluded that the composite primary trial end point would not be met and recommended trial closure. NHLBI closed SWiTCH ‐ N = 40 did not complete treatment phase in transfusion/iron chelation arm and N = 43 did not complete treatment phase in hydroxyurea/phlebotomy arm. More participants had moya‐moya in the hydroxyurea arm (11 participants) than the transfusion arm (5 participants), it was not known if there was a difference between treatment arms in the number of participants with other types of severe vasculopathy.