Factors influencing collection of peripheral blood stem cells in patients with multiple myeloma

This study was performed to determine the factors influencing the collection of autologous peripheral blood stem cells (PBSC) in patients with multiple myeloma (MM) who had disease which had progressed after an initial response or who had refractory disease. Fifty-seven patients with MM underwent PBSC collections following recombinant human granulocyte colony stimulating factor (G-CSF) alone (n = 19) (16 micrograms/kg/day), cyclophosphamide (CY) (4 gm/m2 x 1) with either G-CSF (10 micrograms/kg/day) (n = 7) or granulocyte-macrophage colony-stimulating factor (GM-CSF) (500 micrograms/m2/day) (n = 7) or cyclophosphamide (4 gm/m2 x 1) and etoposide (200 mg/m2/day x 3) (CE) with G-CSF (10 micrograms/kg/day) (n = 24). The goal was to collect 5 x 10(6) CD34+ cells/kg. Fifty of 57 patients underwent autologous transplantation with PBSC alone (n = 39) or PBSC + marrow (n = 11). The median yield of CD34+ cells was 7 x 10(6)/kg (range 0-178.3). Thirty-nine of 57 patients (68%) achieved the target level of 5 x 10(6) CD34+ cells/kg in a median of three (range 1-8) collections. Eighteen (32%) patients yielded < 5 x 10(6) CD34+ cells/kg with the first collections. Thirteen of these 18 patients yielded < 2.5 x 10(6) and five yielded 2.5-4.95 x 10(6) CD34+ cells/kg. Of the 18 patients with less than optimal CD34+ cell yields, five with CD34+ yields of 2.5-4.95 x 10(6)/kg received PBSC alone at transplant, six underwent marrow storage to augment the PBSC dose and received PBSC plus marrow and seven patients underwent secondary collections. Of seven patients who underwent second (n = 5) or third (n = 2) cycles of PBSC collections using G-CSF 16 (n = 4) or 32 (n = 3) micrograms/kg/day, > 2.5 x 10(6) CD34+ cells/kg were collected in four patients. Two patients achieved < 0.18 CD34+ cells following three cycles of mobilization. In a linear regression model, an increased percentage of marrow involvement and prior radiotherapy (RT) were statistically significantly associated with a low CD34+ cell collection yield (P = 0.003, and 0.01, respectively). A mobilization regimen of CE plus G-CSF was associated with a significantly higher yield of CD34+ cells as compared to patients receiving G-CSF alone (P = 0.02). CY with G or GM-CSF was not significantly different than G-CSF alone (P = 0.49). Twenty-two of 24 (92%) patients receiving CE with G-CSF achieved a target level of 5 x 10(6) CD34+ cells/kg or more as compared to 11 of 19 (58%) patients receiving G-CSF alone (P = 0.01) and six of 14 (43%) patients receiving CY with G or GM-CSF (P = 0.001). These data suggest that percentage of marrow involvement, prior radiotherapy, and number of prior chemotherapy regimens are important predictors of PBSC yield in patients with MM. These data also suggest that CE plus G-CSF is superior to G-CSF alone or CY plus G/GM-CSF based on mean daily CD34+ cell collection yield. Higher doses of G-CSF (16-32 micrograms/kg/day) can result in adequate CD34+ cell collections in some secondary attempts in patients with MM failing an initial mobilization regimen.