We have recently reported that the hematologic recovery of patients with non-Hodgkin's lymphoma (NHL) and Hodgkin's disease (HD) undergoing autologous bone marrow transplantation (BMT) is significantly faster when recombinant human interleukin-3 (rhIL-3) is combined with recombinant human granulocyte colony-stimulating factor (rhG-CSF) in comparison with patients receiving G-CSF alone. In this paper, we studied the kinetic response and concentration of BM progenitor cells of 17 patients with lymphoid malignancies submitted to autologous BMT and treated with the G-CSF/IL-3 combination. The results were compared with those of five lymphoma patients receiving the same pretransplant conditioning regimen followed by G-CSF alone. rhG-CSF was administered as a single subcutaneous (sc) injection at the dose of 5 micrograms/kg/d from day 1 after reinfusion of autologous stem cells; rhIL-3 was added from day 6 at the dose of 10 micrograms/kg/d sc (overlapping schedule). In both groups (G-CSF- and G-CSF/IL-3-treated patients), cytokine administration was discontinued when the absolute neutrophil count (ANC) was >0.5 x 10(9)/L of peripheral blood (PB) for 3 consecutive days. After treatment with the CSF combination, the percentage of marrow colony-forming units-granulocyte/macrophage (CFU-GM) and erythroid progenitors (BFU-E) in S phase of the cell cycle increased from 9.3 +/- 2% to 33.3 +/- 12% and from 14.6 +/- 3% to 35 +/- 6%, respectively (p < 0.05). Similarly, we observed an increased number of actively cycling megakaryocyte progenitors (CFU-MK and BFU-MK). Conversely, G-CSF augmented the proliferative rate of CFU-GM (22.6 +/- 0.6% compared to a baseline value of 11.5 +/- 3%; p < 0.05) but not of BFU-E, CFU-MK, or BFU-MK, and the increase of S-phase CFU-GM was significantly lower than that observed in the posttreatment samples of patients receiving IL-3 in addition to G-CSF. The frequency of hematopoietic precursors in the BM, expressed as the number of colonies formed per number of cells plated, was unchanged or slightly decreased in both groups of patients. Because of the increase in marrow cellularity, however, a significant augmentation of the absolute number of both CFU-GM (3605 +/- 712/mL BM vs. 2213 +/- 580/mL; p < 0.05) and BFU-E (4373 +/- 608/mL vs. 3027 +/- 516/mL; p < 0.05) was reported after treatment with G-CSF/IL-3 but not G-CSF alone. Similarly, administration of the cytokine combination resulted in a higher number of CD34+ cells/mL BM, and their concentration was significantly greater than that observed in the posttreatment samples of G-CSF patients. Finally, we investigated the responsiveness to CSFs, in vitro, of highly enriched CD34+ cells, collected after priming with G-CSF in vivo (i.e., after 5 days of G-CSF administration). Our results demonstrated that pretreatment with G-CSF modified the response of BM cells to subsequent stimulation with additional CSFs. The results presented in this paper indicate that in vivo administration of two cytokines increases the proliferative rate and concentration of BM progenitor cells to a greater degree than G-CSF alone. These results support the role of growth factor combinations for accelerating hematopoietic recovery after high-dose chemotherapy.