I read with interest the article, “For Children with Leukemia, Transplants with Bone Marrow Found Safer than with Allogeneic Peripheral Blood Stem Cells,” in the December 10 issue, reporting on a study published in the Journal of Clinical Oncology.1
For that analysis, data on 143 HLA-identical peripheral blood progenitor cell and 640 bone marrow transplants performed between 1995 and 2000 collected from centers worldwide were compared.
There was a significant difference between the groups for age, disease status, period in which transplanted, infused nucleated cell dose, and growth-factor administration post-transplant for engraftment; patients in the peripheral blood progenitor cell (PBPC) group were older, more were transplanted in relapse (22% vs 11%) and less in first transplant (42% vs 52% complete response rate), received remarkably high nucleated cell doses, were more recently transplanted, and had more frequently received growth factors.
There was no difference in gender, disease type, and conditioning regimen, and graft-versus-host disease (GVHD) prophylaxis in Cox proportional hazard models were used to adjust for patient, disease, and transplant-related variables; center effects were evaluated, but the method was not described.
Advanced disease and use of growth factor were significantly associated with increased risk of relapse after both types of transplants and with higher mortality after transplant.
Disease status is a known risk factor for disease outcome after transplant.2 The disease status was considerably more advanced in the PBPC group. Thus, the risk factors more likely than the graft type induced the negative outcome.
Concern is expressed about the high nucleated cell dose used in the PBPC group and the lack of knowledge about the effect of nucleated cell dose on outcome.
There is no justification to administer excessive cell doses; the higher nucleated cell dose induced significantly more chronic GVHD, undefined associated side effects, and therewith risk of treatment-related mortality; causes of death are not given. The relationship between cell dose and chronic GVHD warrants studying.
Furthermore, the discrepancy of number of transplants in each group is remarkable; although the registry's database contained data from 351 PBPC transplants of eligible patients, data from only 143 patients were used.
In the BMT arm, this was. respectively, 1,532 and 630. Bias in data selection may have occurred; it is notable that only 40% of reported transplants of both stem cell sources were used.
If there were that many BMT transplants reported it is not clear why the groups did not better match for disease status. If data were missing these could have been requested from centers.
Finally, growth factor post-transplant is given to accelerate engraftment; PBPC grafts contain in general higher cell dose, and growth factor after PBPC transplants is not routinely justified. The more frequent growth factor use in the PBPC induced a negative selection bias
Investigators should not have been too much surprised with the outcome of the analysis. Moreover, to discourage PBPC transplantation based on the results is not justified.
Standards for nucleated cell dose used in children and elimination of routine post-transplant growth factor seem warranted.
Marlies Van Hoef, MD, PhD
Amsterdam, The Netherlands