It is known that the crosstalk between adhesion molecules, bone marrow microenvironment, and cytokines facilitates the multi step process of stem cell mobilization from bone marrow to peripheral blood. A combination of G-CSF plus AMD3100 - a CXCR-4 antagonist - has been shown to be safe and efficient in stem cell mobilization of healthy donors and cancer patients. Nevertheless, data predicting the efficacy of this approach are still missing. The present study investigated the correlation of the expression of CXCR-4 (CD184) and adhesion molecules with the kinetics and efficacy of stem cell mobilization in nine patients with Multiple Myeloma (MM) or NHL, respectively. Steady-state mobilization was performed using a combination of G-CSF (Filgrastim, 10μg/kg/d, 8 am) for 4 days followed by AMD3100 (240μg/kg) on day 4 at 10pm. Autologous aphereses were started on day 5. Bone marrow and peripheral blood (PB) before AMD3100 application (day 4) and PB on day 5 were investigated with a 4-color flow cytometric procedure. Bone marrow aspirates of healthy donors (n=20) served as control. The qualitative (%) and quantitative (mean fluorescence intensity, [MFI]) antigen expression of CXCR-4 in relation to CD34 was assessed as well as the expression of certain adhesion molecules including LFA-1, PECAM-1, VLA-1, L-selectin and CD44. First, the median percentage of CXCR-4 surface expression in healthy bone marrow was significantly higher (92%; range: 52 – 99%) than in patients bone marrow (70%; 30 – 88%; p=0.002), PB before AMD3100 (87%; 35 – 97%; p=0.050) and on day 5 (17%; 2 – 74%; p<0.001), whereas cytoplasmic expression was comparable (91%; 53 – 95%) in all cell compartments. The median quantitative CXCR-4 surface expression was significantly decreased in PB on day 5 compared to pre AMD3100 (14 vs. 95; p=0.003). Furthermore, the qualitative expression of LFA-1 and the quantitative expression of LFA-1, PECAM-1, VLA-1, and CD44 were also downregulated in response to AMD3100 (p<0.010). Second, a median of 63/μl (range: 15 – 132/μl) CD34+ cells was measured in the PB on day 5. Thus, a high absolute count of CD34+ cells in the PB on day 5 significantly correlated with lower qualitative and quantitative CXCR-4 expression in the same material (r=0.833; p=0.015). Evaluating CXCR-4 expression in bone marrow, PB before AMD3100 and on day 5 no significant correlation to CD34+ counts could be detected. However, there was one very poor mobilizing patient (15/μl CD34+ cells on day 5) in whom the quantitative CXCR-4 expression in the bone marrow was significantly higher than the median of all patients (MFI 95 vs. 26). Furthermore, some of the adhesion molecules (L-selectin, VLA-4, and CD44) showed a rather positive correlation with CD34 count. In summary, these preliminary data suggest that the amount of CD34+ cells in the peripheral blood after G-CSF plus AMD3100 application seems to be negatively correlated with CXCR-4 expression. A higher quantitative CXCR-4 expression in the bone marrow pre AMD3100 might predict a lower mobilization efficacy.

Author notes

Corresponding author