This individual patient data meta-analysis concluded that allogeneic peripheral blood stem-cell transplantation (PBSCT) was associated with faster engraftment, decreased relapse rates and improved overall and disease-free survival, but an increased risk of chronic graft-versus-host disease compared with bone marrow transplantation in patients with haematologic malignancies. The authors' conclusions appear to be reliable.

To evaluate the differences between allogeneic peripheral blood stem-cell transplantation (PBSCT) in comparison with bone marrow transplantation (BMT) for the management of haematologic malignancies.

MEDLINE, EMBASE, LILACS, CANCERLIT and The Cochrane Library were searched without language restriction from 1990 to 2002. Search terms were reported (see Other Publications of Related Interest 1.). Relevant conference abstracts were searched and experts in the field contacted for additional data. Periodic searches were also performed up to August 2003.

Randomised controlled trials (RCTs) of PBSCT in comparison with bone marrow transplantation for the treatment of adult patients with hematologic malignancies and human leukocyte antigen (HLA) -matched sibling donors were eligible for inclusion.

Interventions in the included studies compared PBSC with bone marrow transplantation, G-PBSCT with G-BMT and T-cell depleted (TCD) PBSCT with TCD bone marrow transplantation. Donors in the included studies were all matched siblings. The age of participants in the included studies ranged from seven years to 65 years. The proportion of males ranged from 48.2% to 76.6%. Participants in the included studies were diagnosed with various hematological malignancies including acute lymphoblastic leukemia, acute myeloid leukemia, chronic myeloid leukemia, myelodysplastic syndrome, primary myelofibrosis, non-Hodgkin's lymphoma and refractory anemia with excess blast in transformation (RAEBT). Some patients had undergone total-body irradiation. Participants received granulocyte colony stimulating factor (G-CSF) for PBSCT mobilisations using Filgrastim (10μg/kg either four or five times daily) or Lenograstim (10μg/kg either four or five times daily).

The outcomes included in the review were overall survival, relapse or progression, graft-versus-host disease (GVHD), disease-free survival, death in remission and engraftment.

The authors stated neither how studies were selected for the review nor how many reviewers performed the selection.

Data were checked for inconsistencies. Trial investigators were contacted to resolve any discrepancies. Raw data were also checked against the results of the trial publications. Checks for imbalances between the randomised groups, differences in accrual, length of follow up and any differences in the numbers in subgroups were performed.

Principal investigators of identified studies were contacted for updated individual patient data (IPD). Data were extracted for each individual patient on: time to neutrophil and platelet engraftment; date of relapse or disease progression; date of onset and grade of acute graft-versus-host disease and chronic graft-versus-host disease; and date of last follow-up or death. Time was calculated from the date of randomisation (except for acute graft-versus-host disease, where it was calculated from the date of randomisation and date of transplant and chronic graft-versus-host disease where it was 100 days after transplantation). Data on patients were separated into separate categories for early-stage and late-stage disease by consensus of all reviewers.

The individual log-rank statistics were combined to provide an overall estimate of the estimate of the effect of PBSCT compared to bone marrow transplantation for the outcomes of interest. Odds ratios were calculated, together with corresponding 95% confidence intervals (CI). Comparisons were based on an intention-to-treat analysis. Statistical heterogeneity was assessed using the Χ² test. Sub-group analyses of evaluation of data by disease group were defined a priori.

Individual patient data from 12 RCTs (n=1,318) were included in the review and individual patient data from nine RCTs (n=1,111) were included in the analyses.

Overall survival and disease free survival: Overall survival was statistically significantly improved only in patients with late stage disease treated with PBSCT at three years (odds ratio 0.64, 95% CI: 0.46 to 0.90, p=0.01). PBSCT was associated with significant improvements in disease-free survival compared with bone marrow transplantation at three years (odds ratio 0.80; 95% CI: 0.67 to 0.97, p=0.02). This was more pronounced in patients with late stage disease (odds ratio 0.63, 95% CI: 0.45 to 0.87, p=0.01).

Relapse, relapse mortality and non-relapse mortality: There was a significant decrease at three years in relapse rates (odds ratio 0.71, 95% CI: 0.54 to 0.93, p=0.01), late-stage disease (odds ratio 0.59, 95% CI: 0.38 to 0.93, p=0.03) and early-stage disease (odds ratio 0.69, 95% CI: 0.49 to 0.98, p=0.04) patients treated with PBSCT. There was evidence of statistical heterogeneity for analysis of relapse (p=0.04). There were no statistically significant differences between groups for non-relapse mortality.

Acute or chronic graft-versus-host disease: There was a significant increased risk for the development of grade 3 to 4 acute graft-versus-host disease (odds ratio 1.39, 95% CI: 1.03 to 1.88, p=0.03), extensive graft-versus-host disease (odds ratio 1.89, 95% CI: 1.47 to 2.42, p<0.00001) and overall chronic graft-versus-host disease (68% versus 52% at three years, odds ratio 1.92, 95% CI: 1.47 to 2.49, p<0.00001) in patients treated with PBSCT. There were no significant differences between groups for grades 2 to 4 active graft-versus-host disease.

Engraftment: PBSCT was associated with faster neutrophil engraftment (odds ratio 0.31, 95% CI: 0.25 to 0.38, p<0.00001) and platelet engraftment (odds ratio 0.52, 95% CI: 0.44 to 0.61, p<0.00001). There was evidence of statistical heterogeneity for these analyses (for neutrophil engraftment p=0.000008 and for platelet engraftment p=0.00005).

Subgroup analyses of disease groups found that PBSCT was associated with improvements for patients with chronic myeloid leukemia for relapse (odds ratio 0.34, 95% CI: 0.2 to 0.58), improvement in disease free survival (odds ratio 0.28, 95% CI: 0.11 to 0.73), overall survival (odds ratio 0.31, 95% CI: 0.12 to 0.80). PBSCT was also associated with improvements for acute myeloid leukemia patients with late stage disease for disease free survival (odds ratio 0.39, 95% CI: 0.21 to 0.72), overall survival (odds ratio 0.45, 95% CI: 0.24 to 0.85).

PBSCT was associated with faster engraftment, a decrease in relapse rate for hematologic malignancies and an improvement in overall and disease-free survival compared with bone marrow transplantation in patients with late-stage hematologic malignancies. PBSCT was also associated with a significant increased risk of extensive chronic graft-versus-host disease.

The review addressed a clear question defined in terms of intervention, comparator, participants, outcomes and study design. Several relevant sources were searched and some efforts were made to locate unpublished data. It was unclear whether efforts were made to reduce language bias. A collaborative group of trial investigators was established to maximise retrieval of the individual patient data and to review results and conduct the meta-analysis. Full details of the analysis methods were not reported; for example, were the studies combined using a fixed-effect model, random-effects model or a stratified log-rank analysis? All of the results were reported as odds ratios and not hazard ratios, which appeared to contradict the analysis methods. The authors' conclusions appear to be reliable.

Practice: The authors stated that when making decisions on which stem cell source to utilise, physicians and patients should consider the greater risk of disease recurrence with bone marrow transplantation against the long-term consequences of chronic graft-versus-host disease.

Research: The authors stated that future research should try and determine the harmful effects of chronic graft-versus-host disease as well as the potential beneficial effects. Research should pay particular regard to the impact of milder clinical presentations of graft-versus-host disease for clinical outcomes in different patient subgroups. Also, future trials should report long-term follow up of patients to gain a fuller understanding of the role of stem cell source on outcomes.

NIH/NHLBI grant no 1RO1HL71650-01. The Jose Carreras Foundation Against Leukemia, NCI CA18029, CA18221. The Swiss National Research Foundation. The French Ministry of Health (Programme Hospitalier de Recherché Clinique 1996). The Ligue Nationale de Lutte Contre le Cancer.

1. Clark L, Clark O, Wheatley K, et al. Allogeneic peripheral blood stem cells transplantation versus bone marrow transplantation for the therapy of hematological malignancies (Protocol for a Cochrane Review), in The Cochrane Library, Issue 4, 2003. Chichester, UK, John Wiley & Sons Ltd, 2003.

2. Bessinger W. Individual patient data meta-analysis of allogeneic peripheral blood stem cell transplant vs bone marrow transplant in the management of hematological malignancies: indirect assessment of the effect of day 11 methotrexate administration. Bone Marrow Transplantation 2006;38:539-546.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.