BACKGROUND: Allogeneic peripheral blood stem cell (PBSC) transplantation is widely performed as a curative therapy for hematopoietic malignancies. Donors for PBSC harvest (PBSCH) are usually healthy subjects and undergo granulocyte-colony-stimulating factor treatment and apheresis procedures. A considerable proportion of donors experience poor mobilization, necessitating additional harvesting or marrow collection or remobilization. Although some characteristics have been reported to correlate with poor mobilization, they may not be taken into account in selecting PBSC donors. To protect healthy donors, it is preferable to predict the number of apheresis procedures needed for PBSCH before the procedure is initiated. STUDY DESIGN AND METHODS: A retrospective cohort study of 83 subjects was conducted, using statistical models to predict the probability of obtaining a sufficient number of CD34+ cells (>or=2.0 x 10(6)/kg) in the first to the third apheresis procedures and the probability of failure to obtain sufficient cells within three apheresis sessions. This study explored potential candidate factors in an ordinal probit regression analysis. RESULTS: Significant factors predicting successful PBSCH were donor age, donor sex, and body weight difference between donor and recipient. The predictive model showed good agreement with the observed number of apheresis sessions. Simulation tables are presented with this model. CONCLUSION: The statistical model developed to predict the number of apheresis procedures for PBSCH may be useful for planning PBSCH in clinical practice.
BACKGROUND: Allogeneic peripheral blood stem cell (PBSC) transplantation is widely performed as a curative therapy for hematopoietic malignancies. Donors for PBSC harvest (PBSCH) are usually healthy subjects and undergo granulocyte-colony-stimulating factor treatment and apheresis procedures. A considerable proportion of donors experience poor mobilization, necessitating additional harvesting or marrow collection or remobilization. Although some characteristics have been reported to correlate with poor mobilization, they may not be taken into account in selecting PBSC donors. To protect healthy donors, it is preferable to predict the number of apheresis procedures needed for PBSCH before the procedure is initiated. STUDY DESIGN AND METHODS: A retrospective cohort study of 83 subjects was conducted, using statistical models to predict the probability of obtaining a sufficient number of CD34+ cells (>or=2.0 x 10(6)/kg) in the first to the third apheresis procedures and the probability of failure to obtain sufficient cells within three apheresis sessions. This study explored potential candidate factors in an ordinal probit regression analysis. RESULTS: Significant factors predicting successful PBSCH were donor age, donor sex, and body weight difference between donor and recipient. The predictive model showed good agreement with the observed number of apheresis sessions. Simulation tables are presented with this model. CONCLUSION: The statistical model developed to predict the number of apheresis procedures for PBSCH may be useful for planning PBSCH in clinical practice.