Impact on Short and Long-Term Hematopoiesis of G-CSF Administered Posttransplant in Function of the Amounts of Peripheral Blood CD34⁺38⁻ Cells Reinfused.

We previously suggested that a reinfused threshold dose of CD34⁺38⁻ cells = 5x10⁴/kg b.w. better predict both short- and long-term engraftment after PBSCT than total CD34⁺ cells assessment, and should thus avoid unecessary postransplant (Tx) G-CSF administration. Therefore, we have further conducted a prospective study comparing postTx data from cancer patients undergoing autologous PBSCT and were administered or not G-CSF depending on the amount of CD34⁺38⁻ cells reinfused. 48 patients (mean age 49y) were transplanted with, on average 2.5x10⁴ CD34⁺38⁻ cells/kg b.w. (range 1–49) and were consequently administered G-CSF 5 μg/kg daily from d5 to ANC = 10⁹/1 (Group-I). 46 patients (mean age 50y) received an average of 20.5x10⁴ CD34⁺38⁻ cells/kg (range 5.5–162) without postTx G-CSF (Group-II). These 2 groups were compared and paired two by two with 2 groups of "historical" patients referred as controls : 11 patients (mean age 44y) had received, on average, 2.5x10⁴ CD34⁺38⁻ cells/kg b.w. (range 1.1–4.8) without G-CSF (Group-III); 29 patients (mean age 51y) received an average of 15.2x5x10⁴ CD34⁺38⁻ cells/kg b.w. (range 5.5–60) systematically associated with G-CSF for protocolar reasons (Group-IV). PostTx trilineage hematopoietic engraftment (up to 2 years), clinical and economical parameters were systematically recorded for each group of patients and statistically compared. PostTx ANC recovery occurred sooner, was faster and reached higher levels in the G-CSF groups (II and IV) compared to the others; platelets recovery kinetics was significantly faster in Group-III compared to the others; reticulocytic recovery was not statistically different whichever the group. When age, sex, disease, TBI did not significantly influence trilineage engraftment, a multiparametric study showed strong positive impacts of total CD34⁺ cells reinfused on ANC kinetics and of CD38⁻ subset amounts on platelet kinetics, which was on the contrary slowered by G-CSF administration. Group-I patients received more transfusions, stayed longer hospitalized and costed more than those of the 3 other groups. Regarding long-term hematopoiesis, platelets and hemoglobin levels were globally higher in Group-III compared to the other groups, but still more dramatically compared to Group-IV from 1 to 9 months, which might be explained by differences in BM CD34⁺ and 38⁻ subset differentiation.

In conclusion, if postTx G-CSF certainly accelerates ANC recovery, it seems to be to the detriment of short- an d median-term platelets and hemoglobin recovery, even in case of reinjection of CD34⁺38⁻ cell doses = 5x10⁴/kg b.w., which appears thus to be significantly discriminant for G-CSF administration decision.