Introduction
For autologous stem cell transplantation (ASCT), harvesting of hematopoietic stem cells (HSC) is often conducted after a disease-specific mobilization chemotherapy regimen followed by treatment with granulocyte colony-stimulating factor (G-CSF). Different risk factors for insufficient HSC mobilization have been identified [1]: Mobilization failure often occurs due to prior exposure to myelotoxic therapy, including irradiation of bone marrow (BM) sites and, in particular, through pretreatment with dose-intense regimens and repetitive chemotherapy cycles, which recruit more HSCs into the cell cycle and in turn lead to severe BM damage and a reduction in HSC. In particular, the application of DNA cross-linking agents, such as alkylating and platinum-based drugs, as well as purine analogs, are suspected of causing harvest failure by potentially killing HSCs and BM niche cells [2]. Moreover, lenalidomide may affect HSC mobilization by suppressing motility [3]. Other predictors for poor mobilization are refractory disease, in particular, extensive BM infiltration by the underlying malignancy, decreased BM cellularity, low platelet count reflecting overall HSC reserve, and patient age older than 65 years. In addition, the negative impact from severe infections due to inhibiting effects of pro-inflammatory cytokines on HSC proliferation, diabetes mellitus and iron overload could be shown [4–6].
For safe engraftment after high-dose chemotherapy, the transplantation of at least 2106 CD34þ cells/kg body weight (bw) is needed, as lower doses may lead to delayed neutrophil and platelet engraftment [7]. Circulating CD34þ cells in the peripheral blood (PB) are a good predictor for adequate apheresis results [8]. Because CD34þ cell levels in PB normally begin to rise by day 10 and reach peak values by day 14 after mobilization chemotherapy, monitoring should be initiated at this time point. Apheresis is commonly started when CD34þcells reach >10/ll in the PB, as this cutoff is trichohepatoenteric syndrome considered the minimum for the successful collection of >2106 CD34þ cells/kg bw. In patients with <5 PB CD34þ cells/ll, adequate harvests are challenging and with 10–20 CD34þ cells/ll several collections for sufficient yields are required. Today, the majority of patients succeed in collecting adequate CD34þcell counts in one apheresis procedure without further stimulation. However, 5–40% of the patients destined for ASCT are considered ‘poor mobilizers’ [1,3,4,9–18]. To improve HSC collection by increasing the number of PB CD34+cells for this subgroup, new therapeutic approaches are needed. Several agents for the improvement of HSC mobilization have been identified, such as plerixafor, stem cell factor (SCF), growth hormone (GH), thrombopoietin (TPO), parathormone, T-140, T134, SDF-1 and anti-VLA4. However, with the exception of plerixafor, none of them has reached clinical relevance because of severe side effects, lack of benefit or pending analyses [19]. The interaction of stromal cell-derived factor-1 (SDF1), a chemokine produced by BM stromal cells, and its receptor CXC chemokine receptor 4 (CXCR4), which is an adhesion molecule produced by HSCs, is crucial for keeping HSCs in their BM niches [20]. Plerixafor (AMD 3100, MozobilVR) is a bicyclam small-molecule CXCR4 antagonist that blocks the binding of SDF-1 on HSCs, thus reducing their chemotaxis to BM and allowing their release into the PB [21]. Randomized phase III studies have demonstrated that the supplemental application of plerixafor is superior to mobilization strategies using G-CSF alone in patients with multiple myeloma (MM) and non-Hodgkin lymphoma (NHL): The combination of plerixafor with G-CSF allowed the collection of significantly more CD34+ cells within fewer apheresis days compared to G-CSF alone [10,11,22–24]. Moreover, phase II studies and retrospective analyses were conducted with ‘poor mobilizers’ diagnosed with lymphoma, confirming the effectiveness of the preemptive application of plerixafor in addition to G-CSF. Of note, successful HSC collection was possible in >70% of the cases [12,13,25–27]. However, in Europe, stem-cell mobilization is mostly performed using a combination of mobilization chemotherapy (Supplementary Table 1) and G-CSF and randomized trials in this setting are lacking. Plerixafor is approved for stem cell mobilization in combination with G-CSF in patients with MM and NHL, who mobilize poorly. The recommended daily dose is 240lg/kg by subcutaneous injection on the fourth day of G-CSF application, repeated 2 to 4 times. The maximum increase in circulating CD34+ cells can be expected 6 to 11h after application, therefore, it should be given the night before the scheduled leukapheresis [10,11,28]. Only transient, mild side effects have been described, including diarrhea, nausea, headache and local injection site reactions [10,11,29].
Plerixafor is costly and should be very thoughtfully administered. Therefore, at our institution we aimed to administer the drug only once in a single fixed-dose. We here very thoroughly analyzed 46 consecutive patients who underwent ASCT at our academic center between 2011 and 2015 and received chemomobilization, G-CSF and additional plerixafor due to poor mobilization. We examined CD34+ cell counts in PB and in apheresis products to identify those patients with sufficient CD34+ cell count for ASCT after a scheduled single application of plerixafor with an absolute dose of 24mg.
Material and methods
Patients
Approximately 100 allogeneic SCTs and 110 ASCTs are conducted annually at the Freiburg University Medical Center. Additional administration of plerixafor has been practiced since 2011 at our center with the goal of a single fixed-dose administration. From June 2011 until August 2015, 46 (7%) consecutive patients received plerixafor out of 655 total apheresis patients undergoing HSC mobilization for ASCT during this time period. These patients were either expected to be poor mobilizers having <20 CD34+cells/ll in PB samples prior to apheresis, or because of an insufficient yield at the first day of apheresis, or after insufficient harvest with previous mobilizing chemotherapy. In general, we start measuring CD34+ cells in PB following regeneration of white blood cells (WBC) >1000/lland platelets (PLT) >20,000/ll for several days to analyze the dynamic of the increase of CD34+cells in the blood. When patients regenerated to WBC >10,000/ll and PLT >50,000/ll without any sufficient increase of PB CD34 >10–20/ll, we considered administration of plerixafor. The patients were diagnosed with NHL (n= 17), Hodgkin lymphoma (n= 3), MM (n= 23) and solid tumors (n= 3). Equal numbers of male (n= 24) and female (n= 22) patients Vorinostat chemical structure with a median age of 62 years (range: 26–75) were included (Table 1). The study was conducted in accordance with the Declaration of Helsinki, and informed consent for mobilization and leukapheresis was obtained from each patient before the start of treatment.
Mobilization
The mobilization treatment protocol consisted of a particular chemotherapy regimen (Table 1) and daily G-CSF starting at day +6 to +9 of the treatment schedule: 5lg/kg bw G-CSF (filgrastim, Neupogen, Amgen, Munich, Germany) was given subcutaneously every evening for a minimum of 5 days and after initiation of apheresis twice daily. The single dosing was rounded to standard size vials of 300lg (<70kg bw) or 480lg G-CSF (>70kg bw). Plerixafor was administered subcutaneously as a fixed dose of 24mg/day without bw adjusting and the goal of a single administration for each patient. Patients underwent apheresis 12h after the administration of
Apheresis
HSC collection was performed with a Cobe SpectraVR or Spectra OptiaVR apheresis system from Terumo BCT (Belgium). Aphereses were mostly completed in 4–5h and were well tolerated in all the patients. The planned target blood volume to be processed in this time was the 2.5-fold total blood volume calculated according to patients’ weight and size. The total WBC count was measured using an automated cell counter. Following WBC >1000/ll and PLT >20,000/ll in PB on day 11– 15 after mobilization therapy (expected day of harvest) PB CD34+counts were analyzed as described [30] using single platform flow cytometry until a sufficient PB CD34+cell count ideally of >20/ll was reached at the peak of mobilization and apheresis could be initiated. For each ASCT, we aimed for target yields of at least 2106 and preferentially 4106 CD34+ cells/kg bw, respectively. Cryopreservation with DMSO was conducted according to our internal standard protocol.
Statistical analysis
An unpaired two-sided student’s t-test was used for statistical analyses. p-values <.05 were considered statistically significant. Results Patient characteristics A total of 46 (7% of the entire leukapheresis cohort) consecutive patients received plerixafor, most of them diagnosed with MM or other lymphoma (Table 1). In 89.1% plerixafor was administered because of an insufficient yield at the first day of apheresis and in 8.7% after an insufficient harvest with previous mobilizing chemotherapy. In only one patient (2.2%) it was a preemptive application because of <20 CD34+ cells/ll in PB prior to apheresis. In 44/46 evaluable patients, the median BM infiltration detected was 30% for the underlying disease, 59% showed no or low infiltration, whereas 41% showed a bone marrow infiltration>50% (Table 1). Most patients were heavily pretreated: on average, 2 preceding lines of chemotherapy had been conducted (range 1–8); moreover, 43.5% of the patients had been previously irradiated, and 13% had already undergone ASCT (Table 1). On average, plerixafor was given at day 16 (range 8–39) after mobilization therapy. In 68% of the patients it was administered only once, as planned. Only 32% received more than one application, mostly due to insufficient yields of <1106/kg bw CD34+ cells in the first apheresis after application. 8 of these 15 patients (53.3%) were treated in the first year after introduction of plerixafor at our center before we started scrutinizing the manufacturer‘s suggestions after gaining extended experience with plerixafor use. In 14/46 (30%) of patients more than one apheresis session was conducted after a single dose of plerixafor. As the median bw was 73.5kg (range 48– 112) the median recommended daily dose of 240lg/kg would have been 17mg/day. On average, these patients underwent 3 apheresis days (range 2–9) during one independent mobilization procedure/ apheresis cycle (range 1–4). The median PB CD34+ Ocular genetics cell count at WBC peak was 10.0 cells/ll (0.04%, range: 0–0.33%). The median amount of CD34+cells collected per single apheresis was 1.0106/kg bw and the median total apheresis yield was 3.9106/kg bw (Table 2a).
CD34+cell harvest
In 44/46 patients the median collected CD34+cell count was higher after additional plerixafor application than with G-CSF alone. In only one patient, it remained the same (#23), and in another patient, PBSC harvest was unsuccessful with this patient not mobilizing at all (PB CD34+ 0/ll), both with and without plerixafor (#38) (Figure 1(a)). We also analyzed, how many patients were able to collect >4,>3,>2 or>1106 CD34+ cells/kg bw.
In particular, we examined the number of patients who succeeded in collecting a minimum count of >2 根106 CD34+cells/kg bw and could, therefore, receive at least one single ASCT. We observed that 48% and 74% of patients reached total apheresis yields of >4 and >3106/kg bw CD34+cells, respectively. Of note, 83% collected>2106/kg bw CD34+ cells (Table 2b, Figure 1(b)). All the patients with >2106/kg bw CD34+cells proceeded to ASCT; additionally, five patients successfully underwent ASCT even after collection of <2106/kg bw CD34+cells and each of them engrafted in time (Figure 1(b)) with achievement of a WBC count>1.0103/ll after a median of 15 days (range: 10–23). Thus, a total of 87% of these poor mobilizers were successfully transplanted (Figure 1(b)).
CD34+ cell count in PB
The amount of CD34+cells/ll in PB on the day planned for leukapheresis correlated well with the yield obtained per single collection and total harvest. We found that higher PB CD34+cells led to increases in the apheresis product: In patients with a total PBSC harvest yield of>4, 3–3.9, 2–2.9, 1– 1.9 or<1106/kg bw CD34+ cells, a median of 17.8, 8.0, 6.1, 6.2 or 0.0 PB CD34+cells/ll were detected prior to apheresis, with a corresponding median single collection count of 1.4, 0.9, 0.5, 0.3 and 0.2106 CD34+ cells/kg bw, respectively (Table 2b, Figure 2(a,b)). Benefit of plerixafor administration We also analyzed, whether certain patient subgroups in particular benefited most from plerixafor: those patients with substantial (76%) vs. minor response (24%) showed higher WBC counts (median 19.4 vs. 13.4103/ll, p= .88) and higher PB CD34+cell counts (median 8.84 vs. 1.72/ll, p< .001) before receiving plerixafor, and achieved a 5-fold higher yield of CD34+ cells per single apheresis (median 2.25 vs. 0.43106 CD34+/kg bw, p< .001), and 3.3-fold higher yield of total CD34+ cells (median 4.6 vs. 1.4106 CD34+/kg bw, p< .001), respectively, after receiving We also analyzed the percentage of patients with PB CD34 cells lower than a certain cutoff who were able to collect a CD34þcell count >2 and >4 106/kg bw after receiving plerixafor. A potential PB CD34 cutoff of 5/ll was confirmed as 47%, 86%, 100% and 100% of patients with <5/ll,<10/ll, <15/ll and <20/ll collected more than 2106 CD34þ/kg bw and 0%, 21%, 44% and 67% more than 4106/kg bw, respectively. Of interest, only 32.6% of the patients received more than one plerixafor application. Thus, a single administration of a fixed plerixafor dose of 24mg in 67.4% of our patients was sufficient and very costefficient. The number of patients examined was too low to allow further subgroup analyses of the patients with substantial vs. minor response to plerixafor, e.g. in terms of pretreatment and BM-infiltration rates. We could show, however, that both groups did not significantly differ in age (median age 63 vs. 56 years, p=.159) and bw (median bw 73kg (range 48– 112) vs. 76kg (range 55–94), p= .24), respectively. The two patients with>100kg bw were in the benefit group. These would have been the only underdosed patients with respect of the recommended dose of 240lg/kg (Table 2c).
Safety and engraftment
No relevant toxicities were documented beside minor side effects such as local skin reaction at the injection side or mild diarrhea as noted in routine reports. For lack of standardized questionnaires in our retrospective analysis we cannot quantify these data in more detail, but plerixafor was generally very well tolerated in line with prior reports [31,32]. All 40 patients who have undergone ASCT to date (87%) have achieved a durable WBC engraftment after a median time of 11 days (range: 8–23). All the patients received G-CSF stimulation after HSC-retransfusion, which was started on day þ7.
Poor mobilizers without plerixafor administration
From June 2011 until August 2015, 47 of 655 patients (7.2%) collected less than 2106 CD34þ cells/kg bw in the first apheresis session (median 1.19106 CD34þ cells/kg bw) and successfully underwent ASCT after collection of a median of 3.54106 CD34þ cells/ kg bw in three sessions without receiving plerixafor. Six patients did not receive plerixafor despite insufficient harvests with a median total apheresis yield of 0.52106 CD34þ cells/kg bw (median 0.34106 CD34þ cells/kg bw per single apheresis) and for practical reasons, patient and/or physician decisions no second mobilization attempt including the use of plerixafor was performed: Two patients did not proceed to ASCT due to serious secondary disorders prohibiting high-dose chemotherapy. One patient died due to progressive disease, one underwent allogeneic SCT after identification of a compatible donor, and in one the therapy concept was changed to best supportive care because of rapid clinical deterioration. One patient underwent ASCT with less than 2106 CD34þcells/ kg bw after a dose reduction of conditioning regimen and achieved full engraftment thereafter.
Discussion
High-dose chemotherapy with subsequent ASCT is considered a standard therapy for many malignancies, especially MM and relapsed or high-risk lymphoma. HSC apheresis is primarily conducted during hematopoietic reconstitution after mobilization chemotherapy and stimulation with G-CSF. However, up to 40% of the patients may fail to collect adequate amounts of CD34þcells with this conventional approach and are subsequently not able to proceed to a potentially curative or long-term remission-inducing ASCT. New effective treatment options and postponement of ASCT to first relapse in MM patients will lead to more patients that are challenging to mobilize.
Recent papers have shown that for various reasons, including mobilization failure, only …60% of MM patients proceed to 2-line ASCT [33,34]. Therefore, an improvement of mobilization efficacy and reduction of the number of apheresis procedures is desired. It could be demonstrated that in poorly mobilizing patients PB CD34þ cells increase with the additional application of the CXCR4 inhibitor plerixafor [35]. A significant correlation between CD34þ cells in PB and the amount of cells collected in 1 to 3 apheresis procedures has been shown [8,36]. Therefore, analysis of CD34þ cells in PB can help to identify patients exhibiting rising WBC counts without mobilization of sufficient CD34þ cells as candidates for plerixafor administration.
One particular benefit of plerixafor has been demonstrated in patients unable to complete prior attempts of HSC collection [14,37]. However, the identification of patients who will benefit most from plerixafor is crucial. Moreover, a guideline for the most reasonable and efficient usage of plerixafor would be helpful, which would also limit substantial costs. Potential risk-adapted algorithms for its optimized administration have been developed [9,15,16,38], distinguishing between preemptive usage in expected poor mobilizers, rescue application for patients with suboptimal CD34þ cell levels in PB, or inadequate harvest on the first day of apheresis, and planned salvage administration in prior failed cases. Therefore, different cutoffs for PB CD34þcell counts have been proposed: successful mobilization is difficult in patients with <20/ll circulating CD34þ cells despite WBC counts of >10103/ll [1]. In other studies, poor mobilizers have been defined as those with <8– 10 CD34+cells/ll in PB [17,18]. Similarly, there are different recommendations concerning the use of plerixafor: Mostly, a PB CD34+ cell count of 20/ll is defined as the threshold for additional application of plerixafor on demand [39,40]. Andreola et al. suggest an additional application in patients with PB CD34+ cell counts between 8 and 20/ll [18], whereas D'Addio et al. observed the highest fold-increase in patients with <5 circulating CD34+ cells/ll, therefore suggesting an application in this subgroup [17]. These recommendations have been developed for the patients who failed to mobilize sufficiently following chemotherapy combined with G-CSF, a schedule that is mostly used in Europe. When using steady-state G-CSF mobilization, the addition of plerixafor seems to allow successful mobilization, even in the patients with <3.5 PB CD34+ cells/ll [41]. It has also been shown that steady-state mobilization with G-CSF plus plerixafor may be superior to chemotherapy plus G-CSF [31,42]. Therefore, cutoffs for PB CD34+cells/ll in steady state mobilization may not be comparable, and the effect of adding plerixafor under steady-state conditions may even be more pronounced. Different cost analyses concerning total charges related to leukapheresis have been performed: It could be shown that the preemptive use of plerixafor in patients expected to be poor mobilizers has costs similar to those incurred in poor mobilizers treated with G-CSF alone when a second mobilization attempt is needed [15,32,39,43] and that chemomobilization with G-CSF is associated with a higher financial burden than steady state mobilization with additional plerixafor [31]. Thus, an offset of costs for plerixafor by reduction of other expenses should be considered. For example, the requirement for additional apheresis sessions could be eliminated. Moreover, a second myelosuppressive mobilization therapy could be spared, thus preventing additional side effects like prolonged cytopenia with a risk for febrile neutropenia and its expensive treatment. However, prospective studies are still lacking. An interim analysis of a study comparing patients that received plerixafor on demand according a defined algorithm with a retrospective analyzed cohort showed an improvement in apheresis results with no increase in costs [43]. In our analysis, 83% of patients succeeded in collecting >2106 CD34+cells/kg bw with the additional application of plerixafor, which was therefore sufficient for at least one ASCT. In 48% of patients, plerixafor even enabled a total apheresis yield of >4106 CD34+cells/kg bw, the minimum count needed for tandem-ASCT, e.g. in patients with MM. Moreover, we are the first to show that a cost-effective single application of a fixed dose of plerixafor can be sufficient, as 67% of our patients underwent successful collection with the help of only a single dose. The most notable of our results was that 76% of mostly heavily pretreated patients in preemptive and rescue settings showed a benefit of plerixafor. We observed a significant difference in the CD34+ cell count in PB before the administration of plerixafor (8.84 vs. 1.72/ll) in this subgroup compared to the non-benefiting subgroup, enabling the collection of a 5-fold higher amount of CD34+ cells per single apheresis procedure (2.25 vs. 0.43106 CD34+/kg bw) after plerixafor applicaton.
In summary, treatment with plerixafor is a safe and effective rescue strategy in the patients who are scheduled to undergo ASCT and who have failed conventional HSC mobilization. As the number of patients with extensive pretreatment, including new immunomodulatory drugs, and with long-disease courses will increase in the future [33,34], it is highly important to develop selection criteria for patients expected to benefit most from respective therapies, also with regard to cost efficiency. Our data suggest that most patients with <5/ll CD34+ cells detectable in PB may fail to substantially benefit from a single application of plerixafor and should, therefore, be carefully selected for repeated plerixafor application. Those with >5/ll PB CD34+cells can, however, greatly benefit from a single dose, with median CD34+apheresis yield increases of 3.4 and median total apheresis collections of 4.6106/kg bw. Therefore, our data suggest that patients with 5– 10/ll PB CD34+cells should receive additional plerixafor and this could also be considered in the patients with 10–20/ll PB CD34+cells, when a second ASCT is scheduled, with the goal of harvesting at least twice the amount of 2106 CD34+cells/kg bw. Our data may greatly aid other transplant centers to cost efficiently use plerixafor for the mobilization of HSCs.