The Impact of Stem Cell Mobilization Regimen on Outcomes after Autologous Transplant for Multiple Myeloma

The infusion of adequate peripheral blood stem cells (PBSC) is critical for the success of autologous stem cell transplant (ASCT), which remains part of standard of care therapy for patients with multiple myeloma (MM). Different regimens mobilize stem cells from bone marrow into the blood where they can be collected by apheresis. Common mobilization strategies in multiple myeloma include chemotherapy (usually cyclophosphamide), granulocyte colony stimulating factor (G-CSF), and plerixafor (Mozobil). Plerixafor is a chemokine receptor antagonist that blocks the interaction between CXCR4 on stem cells and its ligand CXCL12, which serves to anchor stem cells in the marrow. Chemo-mobilization with cyclophosphamide can be used in patients with poor disease control to augment MM response as well as to increase stem cell yield compared to G-CSF alone.

In a single center retrospective analysis of 480 patients who underwent ASCT for MM between 2011 to 2016, we compared transfusion needs, relapse and non-relapse mortality, and immune reconstitution between groups who received different mobilization regimens. Plerixafor was given to 187 patients who were statistically older (63.8 years IQR 58.0, 69.2) compared to the patients who didn’t get plerixafor (62.1 years IQR 55.2, 66.5) (p<0.01). Of the patients who got plerixafor, only 23 (12%) also got cyclophosphamide compared to 58% of patients who did not get plerixafor (p<0.01). Patients who got plerixafor were more likely to be in complete remission (CR) compared to patients who did not (30% vs 16%, p<0.01). The median number of months from diagnosis to transplant differed significantly between the groups with patients who got plerixafor having a longer time from diagnosis to transplant. (8.0 months IQR 5.6, 12.1 compared to 7.0 months IQR 5.6, 10.8, p<0.01).

The graft obtained in patients who received plerixafor differed significantly with lower PBSC (4.3 x 10⁶ IQR 3.6, 5.4 compared to 6.0 x 10⁶ IQR 4.3, 8.6 p<0.01). Despite containing a lower number of PBSC, the grafts obtained by plerixafor had a significantly higher total nucleated cell count (TNC) in the graft (13.0 x 10⁸ IQR 8.0, 18.2 compared to 5.0 x 10⁸ IQR 2.9, 8.9 p<0.01).

The age of the 194 patients who received cyclophosphamide as part of their mobilization did not differ significantly from those that did not. Fewer patients had Karnofsky performance status (KPS) of <90 in the group who got cyclophosphamide (26% vs 36%, p=0.03) although HCT Co-morbidity index (HCT-CI) did not differ significantly. Given its anti-myeloma activity there were unsurprisingly more patients in partial remission (PR) in the group who got cyclophosphamide than in the group that did not (57% vs 21%, p<0.01). There were fewer patients with high-risk cytogenetics in the cyclophosphamide group (34% vs 49%, p<0.01). Months from diagnosis to transplant was also higher in the cyclophosphamide group, likely due from the added time of chemo-mobilization (8.3 IQR 6.2, 12.5 in patients who got cyclophosphamide compared to 6.7 months IQR 5.4, 10.4 p<0.01). The opposite effect was seen in TNC/CD34 ratio compared to the patients who got plerixafor. Patients who got cyclophosphamide had a significantly higher CD34 content in the graft (7.1 x 10⁶ IQR 5.1, 10.4 compared to 4.3 x 10⁶ IQR 3.6, 5.5, p<0.01) but also had significantly lower TNC (3.5 x 10⁸, IQR 2.4,5.5 compared to 10.8 x 10⁸ IQR 7.0,16.0, p<0.01).

Given differences in graft composition we hypothesized that immune reconstitution might differ. We compared leukocyte subsets (absolute numbers and percentages) at day 28 after transplant but saw no differences between patients based on mobilization regimen. We did see differences in transfusion needs with patients who got plerixafor needing more platelet transfusions before day 30 (2.0 IQR 1.0, 4.0 compared to 1.0 IQR 1.0, 2.0 p<0.01) whereas patients who got cyclophosphamide needed more red blood cell transfusions before day 30 (2.0 IQR 1.0, 3.0 compared to 1.0 IQR 0, 2.0 p<0.01).

We also compared relapse and non-relapse mortality in the ten years following transplant and found no significant differences based on mobilization. These results suggest that differences in graft composition do not translate to clinically significant long-term differences in disease control. Mobilization with G-CSF alone is associated with the lowest transfusion needs but all the methods compared here led to similar immune reconstitution and disease control.