Optimization of CD34+ collection for autologous transplantation using the evolution of peripheral blood cell counts after mobilization with chemotherapy and G-CSF

Background: Peripheral blood progenitor cells (PBPC) collection after high dose chemotherapy can be influenced by several factors. We searched for parameters that may predict the best day to start harvesting of PBPC in order to collect most CD34+ cells with the least number of aphereses. Methods: We studied patients who underwent mobilization chemotherapy for autologous transplantation. The influence of age, sex, diagnosis, number of previous chemotherapy cycles, peripheral blood (PB) counts at day of mobilization (D0), day of neutrophils < 1.0 x 10(9) l(-1) and day of nadir and interval between both (delta) on harvesting was investigated. Multivariate linear correlation models were built to predict the best harvesting with principles of parsimony. In patients where sequential CD34+ cell count was performed, the theoretical day of peak was calculated by interpolation in polynomial regression. Results: One hundred and thirty four patients entered the analysis: 36 Hodgkin's lymphoma (HL), 65 B-large cell lymphoma (NHL) and 33 multiple myeloma (MM). Day of harvesting correlated with nr CHT, hemoglobin on DO, day of granulocytes < 1.0 x 10(9) l(-1), delta and dosis of mobilization therapy. The day of CD34+ peak could be calculated by the formula = (-0.41) x Hemoglobin D0 + (day peripheral CD34+ cells = 10 X 10(6) mu l(-1)) x 0.99 + 7.8. This model could explain 81% of the variance of the peak day and was stable by bootstrap resampling. Day of peripheral CD34+ cells = 10 X 10(6) mu l(-1) preceded the calculated peak by 3-9 days. Conclusions: Although the day of best collection can be predicted using only sequential PB counts after mobilization chemotherapy, a model of prediction using peripheral CD34+ cell count is important especially for optimizing collection in poor mobilizing patients. (c) 2005 Elsevier Ltd. All rights reserved.