Powles 2002.

Methods	Randomised controlled trial with two arms: BMT compared with PBSCT in haematological malignancies Participating centre(s): 1 Institute of Cancer Research, Sutton, Surrey, United Kingdom Recruitment period: June 1995 to August 1997 41 participants randomised, 39 participants evaluated (2 participants excluded: 1 participants due to fulminant relapse, 1 donor withdrew after enrolment but before donating) Intervention arm: 19 participants evaluated Control arm: 20 participants evaluated Median follow‐up time: 33 months
Participants	Inclusion criteria: between 18 and 55 years old, malignant haematological disease (primary refractory disease) requiring allogeneic hematopoietic stem‐cell transplantation, HLA‐identical sibling donor Exclusion criteria: not reported Median age: Intervention: 37 years, control: 34 years Recipient/donor sex (male/female): not reported Underlying disease: malignant haematological disease (primary refractory disease) requiring allogeneic haematopoietic stem cell transplantation Stage of Disease: Intervention arm: low risk N = 12, high risk N = 7, acute myeloid leukaemia in first (N = 3) remission, acute lymphoid leukaemia in first (N = 2) and 2nd (N = 1) remission and first relapse (N = 1), chronic myeloid leukaemia in chronic (N = 6) phase, chronic lymphoid leukaemia (N = 1), non‐Hodgkin lymphoma (N = 1), myelodysplastic syndrome (N = 1), multiple myeloma (N = 1), acute biphenotypic leukaemia in primary refractory (N = 1) and first remission (N = 1); control arm: low risk N = 9, high risk N = 11, acute myeloid leukaemia in first (N = 4) and 2nd (N = 1) remission, acute lymphoid leukaemia in first (N = 1) and first relapse (N = 2), chronic myeloid leukaemia in chronic (N = 4) and accelerated phase (N = 2), chronic lymphoid leukaemia (N = 1), myelodysplastic syndrome (N = 1), multiple myeloma (N = 1), acute biphenotypic leukaemia primary refractory (N = 1) in first relapse (N = 2) HLA matching: all participants have HLA‐identical sibling donors
Interventions	Conditioning regimen: Intervention arm: 6 patients received busulphan and cyclophosphamide, 12 patients received melphalan and total body irradiation, 1 received etoposide and total body irradiation; control arm: 7 patients received busulphan and cyclophosphamide, 12 patients received melphalan and total body irradiation, 1 patient received etoposide and total body irradiation GvHD prophylaxis: all participants received cyclosporine started intravenously at 3 mg/kg daily and was switched to oral (12.5 mg/kg daily) just before discharge from hospital until day +180 without acute GvHD and methotrexate (15 mg/m² on day +1 and 10 mg/m² on days +3, +6, +11) Intervention arm: received bone marrow stem cells Control arm: received peripheral blood stem cells
Outcomes	Reported outcomes: time to neutrophil and platelet engraftment, incidence of acute GvHD grades II to IV, incidence of chronic GvHD, relapse, overall survival Non‐reported outcomes: incidence of extensive chronic GvHD, non‐relapse mortality, disease‐free survival, transplant‐related mortality
Notes	Source of funding: "The study was supported by Chugai Pharmaceuticals (Japan), Bud Flanagan Leukaemia Fund, and UK National Health Service" Conflict of interest: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Method of randomisation not mentioned, but we assume that central randomisation and allocation was performed using a computer random number generator
Allocation concealment (selection bias)	Unclear risk	Insufficient information to permit judgement
Blinding of participants and personnel (performance bias) All outcomes	Low risk	"The study was double‐blinded to eliminate the subjective aspects to assessment and care that could potentially affect the outcome of such a comparison." "Masking depended on the harvested stem cells being frozen in bags that made the source indistinguishable when frozen, thawed, and re‐infused"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	"The study was double‐blinded to eliminate the subjective aspects to assessment and care that could potentially affect the outcome of such a comparison." "Masking depended on the harvested stem cells being frozen in bags that made the source indistinguishable when frozen, thawed, and re‐infused"
Incomplete outcome data (attrition bias) All outcomes	Low risk	Two participants excluded: one due to fulminant relapse, one donor withdrew after enrolment but before donating
Selective reporting (reporting bias)	Low risk	"The study protocol was approved by the institutional review board (scientific and ethics committees) of the Royal Marsden Hospital, and all patients and donors provided informed consent"
Other bias	Low risk	The study appears to be free of other sources of bias