Plerixafor (a CXCR4 antagonist) following myeloablative allogeneic hematopoietic stem cell transplantation enhances hematopoietic recovery

Despite its curative potential, myeloablative allogeneic hematopoietic stem cell transplantation (HSCT) is associated with a high incidence of morbidity and mortality [1, 2]. Approximately 25 % of patients will die within 1 year of transplant-related complications. Many of these complications arise during the period of bone marrow aplasia that precedes donor cell engraftment. Infection is the most common cause of morbidity and mortality immediately following HSCT. The speed of hematopoietic recovery and donor cell engraftment has been correlated with the length of the hospital stay, the severity of illness during the stay, incidence of infection, and the patient’s survival [3]. Reduction in the period of bone marrow aplasia would be expected to reduce transplant-related mortality. Hence, there is urgent need to develop novel approaches for enhancing hematological recovery in HSCT.

The interaction between stromal cell-derived factor 1α (SDF-1, also called CXCL12) and the CXCR4 chemokine receptor plays a critical role in homing of hematopoietic stem cells to the bone marrow niche. SDF-1 is secreted by endothelial cells, osteoblasts, and other stromal cells and is present in the bone marrow (BM) extracellular matrix. CXCR4 is expressed on hematopoietic stem/ progenitor cells (HSCs) and many other cells. The interaction between SDF-1 in the BM extracellular matrix and the CXCR4 receptors on HSCs has dual effects: tethering HSCs in the niche and arresting the cycling of very primitive HSCs [4, 5, 6]. By keeping HSCs in a quiescent state, this proliferation inhibitory effect of CXCR4/SDF-1 interaction may delay the reconstitution and recovery of donor HSCs in the early phase following HSCT.

We thus hypothesized that blocking CXCR4/SDF-1 interaction after transplant would remove this inhibitory effect of the CXCR4/SDF-1 interaction on HSCs, resulting in enhance donor cell reconstitution following HSCT. Additionally, we reasoned that because CXCR4 is expressed on T cells, dendritic cells, and many other cells that produce inflammatory cytokines, CXCR4/SDF-1 blockade would attenuate transplant-related cytokine storm. The massive cytokine storm from conditioning has deleterious effects on hematopoiesis [7, 8, 9, 10]. Consistent with these hypotheses, we demonstrated in our preclinical murine study that administration of AMD3100 (an CXCR4 antagonist, also called plerixafor or Mozobil® Sanofi USA) following myeloablative stem cell transplantation resulted in improved animal survival, reduced pro-inflammatory cytokine/chemokine production, and enhanced recovery of all donor cell lineages (myeloid, T and B lymphocytes, erythrocytes, and platelets) [11]. Here, we report our phase I/II study aimed at determining the safety and the efficacy of post-transplant administration of plerixafor in patients receiving myeloablative allogeneic HSCT.

We performed a phase I/II clinical study aimed at determining the safety and efficacy of plerixafor (a specific and reversible CXCR4 antagonist) after allogeneic HSCT to enhance hematopoietic cell recovery in patients receiving myeloablative allogeneic HSCT. Plerixafor was extensively explored as a stem cell mobilizer and approved by the FDA to be used in the pre-transplant setting for stem cell collection [12, 13, 14, 15]. Contradictory to the dogma related to CXCR4 blockade, we previously found that post-transplant administration of plerixafor led to enhanced recovery of donor-derived cells in a murine transplant model [11], and our results were confirmed by others using SDF-1 conditional knockout mice [16]. In the current study, we demonstrate that patients treated with plerixafor have a shorter time to platelet recovery compared to contemporaneous controls (18 versus 19 days) and a higher cumulative incidence of platelet recovery compared to the control cohort. In addition, there is a trend toward enhanced neutrophil recovery in plerixafor-treated patients (p = 0.07). Unplanned subgroup analyses in fact demonstrate a statistically significant difference when the two study patients who received BM grafts are removed (p = 0.04). Our preclinical and clinical data demonstrates that blockade of the CXCR4-SDF1 axis alters the BM niche and promotes stem cell proliferation [11, 17]. This is the first study reporting the beneficial effects of post-transplant administration of plerixafor in enhancing donor cell engraftment following myeloablative allogeneic HSCT.

Adverse events attributable to plerixafor were modest, expected (GI upset, nausea, diarrhea, fatigue) and difficult to distinguish from effects of the myeloablative conditioning. There were no dose-limiting toxicities of primary or secondary graft failure or mortality associated with plerixafor administration. Taken together, this study provides evidence that administration of plerixafor following allogeneic HSCT is safe and feasible.

Post-transplant administration of plerixafor enhances hematologic recovery via several mechanisms. We have previously shown that giving plerixafor after HSCT mobilized residual recipient cells into circulation and increased niche availability for donor HSCs to engraft [11]. Furthermore, we found that post-transplant administration of plerixafor induced HSC cell division in vivo [11]. CXCR4/SDF-1 interaction has two major functions: tethering HSCs in niches and maintaining HSCs in a quiescent state. However, in HSCT, rapid expansion and proliferation of HSCs are crucial and desired especially during the early stage of hematological recovery [18, 19]. Plerixafor has a terminal half-life of 5.3 h [20]. Thus, giving plerixafor once every other day results in transient and pulsed blockade of SDF-1/CXCR4, inducing HSC expansion and proliferation without the risk of HSC exhaustion [11, 16].

Transplant preparatory conditioning induces massive cell death and tissue damage, resulting in a dramatic increase in the levels of secreted chemokines, cytokines, and proteolytic enzymes [21]. These cytokines/chemokines in turn further exacerbate inflammation and vascular/tissue injury, leading to multi-organ toxicity (cytokine storm). Additionally, several of these cytokines/chemokines including TNF [8, 9, 10], TGF-β [22], IL-3 [23], and IFN-γ [7] have negative effects on hematopoiesis. In mouse models, we previously found that post-transplant administration of plerixafor led to a significant reduction in the levels of nine pro-inflammatory cytokines/chemokines [11]. Additionally, we demonstrated that in the early phase of transplant, the marrow microenvironment of transplant recipient mice contains soluble factors that inhibit hematopoiesis and that post-transplant administration of plerixafor likely reduces these inhibitors and/or increases factors that promote hematopoiesis [17]. Consistent with our preclinical studies, our correlative inflammatory cytokine studies demonstrated a reduction in MIP-1β expression at day 30 when compared to patient samples from the historical control cohort. MIP-1β is an inflammatory cytokine with various affects under physiologic conditions. Our current studies are limited to the measurement of only four cytokines (IL-12, TNF-α, IFN-γ, and MIP-1β). These four cytokines were selected based on our preclinical studies and are important mediators in inflammatory process. It would be important to expand the measurement to many other cytokines. Nevertheless, our finding supports the role of CXCR4 antagonism in attenuating inflammatory cytokine storm following allogeneic HSCT.

Compared to what was observed in our preclinical mouse studies, the augmentation of engraftment kinetics by post-transplant plerixafor observed in this clinical trial was modest. There are several potential reasons for this. In our animal study, plerixafor was given at 5 mg/kg body weight whereas in our clinical trial, plerixafor was given at 240 μg/kg body weight. Second, in our animal study, plerixafor was given every other day until day +56 post-HSCT whereas in our clinical trial, plerixafor was administrated every other day until neutrophil engraftment, which came at a median of 16 days following transplantation. This may have dampened the beneficial effect of plerixafor on platelet recovery. Third, in our animal model, the recipient mice were transplanted with a small dose of donor stem cells (250 stem cells per mouse). In contrast, all our patients received on average 5 × 10⁶ CD34 cells/kg (ranging from 1.25 × 10⁶ to 8.5 × 10⁶ CD34 cells/kg). Such a high dose of donor stem cells likely masked the beneficial effects of CXCR4 blockade on stem cell engraftment and proliferation. Finally, although we chose contemporaneous control patients for comparison, there are differences between these two groups of patients. For instance, 36 % of plerixafor-treated patients were ALL patients versus 11 % in control cohort. Forty percent of plerixafor-treated patients received conditioning with TBI + cyclophosphamide compared to only 21 % in control cohort. Furthermore, because of the small patient size, our analysis was not adjusted for CD34 cell dose or ideal body weight. These variables could potentially impact the effects of plerixafor seen in this clinical study.

Our preclinical studies demonstrated an improvement in immune reconstitution among mice treated with plerixafor compared to those given PBS. This finding was not observed in this study. This may be secondary to a multitude of factors including confounding by immunosuppressive medications, clinical status at the time of blood draw, and other homeostatic differences between individual patients.

Currently, recombinant G-CSF is the only agent approved by FDA for enhancing neutrophil recovery following HSCT. G-CSF, however, impacts only myeloid recovery and improves the time to neutrophil recovery following HSCT by only a few days [24, 25]. There is an unmet medical need to identify agents that improve the kinetics of recovery of multiply hematopoietic lineages. This study suggests that plerixafor may be such an agent. It is also noteworthy that plerixafor does not appear to negatively impact OS or disease relapse: the 1-year OS in plerixafor-treated patients was 65 % compared to 57 % in control group, and the 2-year disease relapse rate was 26 % in plerixafor-treated patients versus 34 % in control cohort. We believe that our study has potentially important ramifications in improving hematopoietic recovery in myeloablative HSCT.

Our study provides the foundation for the efficacy of manipulating the CXCR4-SDF1 interaction to provide clinical benefit to recipients of HSCT. Moving forward, we plan to study plerixafor administration following umbilical cord blood transplantation, where the stem cell dose is lower and hematopoietic recovery is significantly slower. Second, a large study controlled for CD34+ cell dose, ideal body weight, and disease characteristics will help further determine the beneficial effects of plerixafor in enhancing hematopoietic recovery. Third, plerixafor has a short half-life (a few hours), and in this study, it was given every other day. Therefore, more frequent dosing higher dose more consistent with what was used in the preclinical studies will be explored. Finally, other CXCR4-SDF1 antagonists with greater affinity to the site of action and with longer half-lives may result in a more clinically meaningful effect and deserve further study.

In summary, blockade of the CXCR4-SDF1 axis post-allogeneic HSCT is safe. While modest, it enhances hematopoietic recovery in patients who have received matched HLA transplants. Further confirmatory studies are justified, potentially in populations with anticipated delays in hematopoietic recovery. Newer CXCR4 antagonists are in varying stages of development and may have utility in this setting in the future.