Limited lenalidomide exposure and access to rescue plerixafor can eliminate haemopoietic progenitor cell collection failure in multiple myeloma patients treated with bortezomib, lenalidomide and dexamethasone

Abstract

This retrospective single‐centre study assessed sequential strategies to improve the success rate for haemopoietic progenitor cell mobilisation in multiple myeloma. Transplant‐eligible multiple myeloma patients treated with bortezomib, lenalidomide and dexamethasone experience a mobilisation failure rate of 32%. We show that a combination of limiting lenalidomide exposure to less than 6 weeks and access to rescue plerixafor can eliminate haemopoietic progenitor cell collection failure without compromising other collection parameters.

Australian clinical practice guidelines for transplant‐eligible multiple myeloma (MM) patients recommend triplet therapy with bortezomib, lenolidamide and dexamethasone (VRD), followed by autologous haemopoietic progenitor cell (HPC) transplant. ¹ We identified an increase in HPC mobilisation failure when VRD replaced velcade, cyclophosphamide and bortezomib (VCD), possibly due to HPC toxicity from lenalidomide exposure. Impaired filgrastim HPC mobilisation with lenalidomide is reported to be related to the duration of drug exposure. ² A suboptimal HPC mobilisation, measured by CD34 on the day of collection, results in inefficient patient care due to an increase in apheresis collection days, failed HPC collections and transplant delays.

Common mobilisation strategies for MM in Australia are filgrastim only or cyclophosphamide and filgrastim to obtain sufficient HPC (>2 × 10⁶/kg CD34) to ensure engraftment. In Australia, plerixafor, a CXCR4 blocker, is restricted to second collection attempt under the Pharmaceutical Benefits Scheme (PBS) in filgrastim‐only HPC mobilisation. A rescue approach to HPC mobilisation by adding plerixafor to a failing filgrastim mobilisation is an efficient way to avoid collection failure but remains unfunded in Australia.

To reduce collection failure, we limited lenalidomide exposure to less than 6 weeks prior to collection and then added hospital‐funded rescue plerixafor in a sequential manner. The report is a retrospective analysis of first HPC mobilisation in VRD‐treated transplant‐eligible MM comparing three groups: PBS rules plerixafor and more than 6 weeks of lenalidomide leading into mobilisation (Group 1), PBS rules plerixafor and less than 6 weeks of lenalidomide exposure (Group 2) and less than 6 weeks of lenalidomide and rescue plerixafor as per study protocol (Group 3). This study assessed whether collection failure can be minimised by scheduling collection earlier to reduce lenalidomide exposure alone or with the addition of rescue plerixafor. As a chemotherapy‐free HPC mobilisation results in a higher pre‐collection white cell count, we assessed whether this impaired other HPC collection parameters. White cell counts higher than 30 × 10⁹/L can negatively impact collection efficiency ³ and final cryopreserved product volume. These factors were assessed along with collection HPC yield and engraftment to determine whether there was any impact on workflow and HPC product quality. In addition, parameters that may impact collection yield were assessed, including the degree of marrow involvement at diagnosis, platelet count at time of collection, co‐morbidity and exposure to cytotoxic drugs or radiotherapy treatment. ⁴

Ethics approval was granted by Austin Health (HREC/110168/Austin‐2024). The study participants were recruited from a single tertiary apheresis centre based in Melbourne, Australia from 2020 to 2023. MM patients at first‐attempt HPC mobilisation were eligible (n = 111). Patients with AL amyloidosis were excluded. Three sub‐groups were assessed to compare HPC collection outcomes: Group 1 – PBS rules plerixafor with >6 weeks of lenalidomide exposure (n = 46), Group 2 – limited lenalidomide to <6 weeks before HPC collection and PBS rules plerixafor (n = 28) and Group 3 – limited lenalidomide limited to <6 weeks before HPC collection, and an option for rescue plerixafor (n = 37).

Demographic data, including disease response at collection and prior radiotherapy exposure, were collected as outlined in Table 1. Performance status was based on Eastern Cooperative Oncology Group (ECOG) scale. ⁵ HPC mobilisation data collected were pre‐collection white cell count (WCC), CD34, number of collection days, HPC mobilisation failure defined as insufficient HPC for target number of transplants (2 × 10⁶/kg per transplant), CD34 yield, collection efficiency, product cryopreserved volume and post‐thaw CD34 bag viability. Engraftment days for platelets and neutrophils were compared using CIBMTR criteria. ⁶ The mobilisation schedule was filgrastim 10 mcg/kg 4 nights and pre‐count day 5. In Group 3, plerixafor 0.24 mg/kg was administered at 10 pm on day 5 collection if CD34 was <14 × 10⁶/L or the estimated first‐day CD34 yield was less than half the total target, with first collection day 6. In Groups 1 and 2, a failing filgrastim‐only HPC mobilisation could not be rescued; however, PBS plerixafor rules could be followed for cyclophosphamide and granulocyte colony‐stimulating factor (GCSF) chemomobilisation, meaning peripheral blood CD34 of 10 × 10⁶/L or less for administration. HPC collection proceeded if the CD34 was >18 × 10⁶/L in these groups, checked on day 5 in the G‐CSF‐only collections and day 11 onwards in the cyclophosphamide mobilised patients. In the groups with less than 6 weeks of lenalidomide exposure, HPC collection was scheduled after cycle 2 VRD 28‐day cycle or cycle 3 VRD 21‐day cycle. Although the main mobilising strategy was filgrastim only, clinician discretion was used to decide on cyclophosphamide 1.5 g/m² mobilisation at first attempt in any group, for example, if a higher collection target was required. Collection efficiency was calculated using CD34 collected/pre‐collection CD34 × blood volume processed. If more than one collection day was needed, the average collection efficiency over the days was taken. Melphalan 200 mg/m² was used for autograft conditioning. Disease response criteria were in accordance with International Myeloma Working Group consensus. ⁷ Medians of demographic data and collection outcomes were compared using Kruskal‐Wallis, the Mann–Whitney U test was used to compare a response of very good partial response (VGPR) or better to less than VGPR and CD34 yield, and group proportions were calculated with Fisher exact test contingency tables using R. ⁸

Table 1.

Demographic and collection data.

	Group 1 >6 weeks lenalidomide	Group 2 <6 weeks lenalidomide	Group 3 <6 weeks lenalidomide, rescue plerixafor	P‐value
Total attempting mobilisation (n)	46	28	37
Gender male (%)	70	61	51	0.168
Age (years)	63	63.5	63	0.995
Weight (kg)	81	80	82	0.761
Radiotherapy (%)
All sites (%)†	13	21.4	22	0.496
Pelvis, spine sternum (%)	10.6	17.9	22	0.391
Marrow CD138 (%)	65	60	50	0.216
Response (%)
CR‡	13	3.5	8.1	0.407
VGPR	19.6	17.9	35	0.191
PR	58.7	71.4	54	0.364
SD, PD	8.7	7.1	2.7	0.611
Collection target CD34 (× 106/kg)	6	6	6	0.938
Filgrastim‐only mobilisation (n, %)	32 (69.5)	20 (71.4)	14 (38)	0.005**
Plerixafor use (n, %)	4 (8.7)	0 (0)	23 (59)	<0.001***
Cyclophosphamide or other chemomobilisation§ (n, %)	14 (30)	7 (25)	1 (3)	0.002**
Day 1 peripheral blood CD34 (× 106/L)	29.5	22.5	46	<0.001***
WCC × 109/L	26.5	35.1	42.2	<0.001***
Platelet × 109/L	187	193	221	0.006**
eGFR (mL/min)	84.5	84.5	90	0.120
Collection failure or not proceeding to apheresis (n, %)	15 (32.6)	8 (32)	0 (0)	<0.001***

^†Other sites of radiotherapy: breast, skull.

^‡CR, complete response; VGRP, very good partial response; PR, partial response; SD, stable disease; PD, progressive disease; WCC, white cell count.

^§Chemomobilisation with 1.5 g/m² cyclophosphamide, one patient with dexamethasone, cyclophosphamide, etoposide and cisplatin (DCEP).

The three collection groups were comparable for demographic data, burden of marrow infiltration at diagnosis by CD138 immunohistochemistry, disease response and collection targets (Table 1). Platelet count at collection was highest in Group 3, 221 × 10⁹/L, compared to Groups 1 and 2 (P = 0.006), likely due to more G‐CSF‐only mobilisation. Groups were not statistically significantly different for degree of renal impairment (P = 0.120) and comorbidity that may affect mobilisation such as diabetes mellitus (P = 0.935) and autoimmune disease (P = 0.821). Performance status based on ECOG was predominantly 0–1 across all groups (P = 0.910). Presence of extramedullary disease was rare at 2% (abdomen), 3.6% (pleura) and 5.4% (paraspinal, carotid canal) across Groups 1–3 respectively (P = 0.821). Radiotherapy exposure to high‐marrow‐risk sites such as pelvis was not statistically different across groups (Table 1). A single patient in Groups 1 and 2 required paclitaxel and carboplatin for breast cancer more than 6 months prior to HPC mobilisation. No other patients had alkylator exposure preceding the MM diagnosis.

Group 1 had a median of 9 weeks of lenalidomide exposure, compared to 6 weeks in Groups 2 and 3. The mobilisation strategy varied in the groups (Table 1). Clinician decision to choose cyclophosphamide HPC chemomobilisation was more frequent with PBS rules plerixafor (Group 1 and 2) compared to rescue plerixafor (Group 3) (P = 0.002). A single patient in the unlimited lenalidomide group required dexamethasone, cyclophosphamide, etoposide and cisplatin (DCEP) salvage after VRD and was collected off this chemotherapy. In Group 3, plerixafor was employed in 23 of 37 (59%), with a median single dose given. The median number of filgrastim doses was five across all groups; however, the spread was significantly different, showing a skew to higher filgrastim days in Groups 1 and 2 compared to Group 3 (P < 0.001), consistent with the more frequent chemomobilisation in these groups.

Median day of collection of peripheral blood CD34 was higher in Group 3, where rescue plerixafor was available (46 × 10⁶/L) compared to Groups 1 and 2 (P < 0.001). Limiting lenalidomide without the use of rescue plerixafor in Group 2 did not improve the median CD34 pre‐count at day of collection, 22.5 × 10⁶/L, compared to Group 1, 29.5 × 10⁶/L (P = 0. 560). Collection day WCC was highest with rescue plerixafor (Group 3), 42.2 × 10⁹/L, compared to Group 1, 26.5 × 10⁹/L, and Group 2, 35.1 × 10⁹/L respectively (P < 0.001). Primary HPC mobilisation/collection failure rate was higher in Group 1, 15/46 (32.6%), and Group 2, 8/28 (32%) compared to Group 3, with no collection failure (P < 0.001).

The collection of outcome data is shown in Table 2. There was a median of two collection days in each group; however, the distribution of these medians was different (P = 0.026). A higher proportion of single‐day collections was achieved in Group 2, 6/16 (37.5%) and Group 3, 12/37 (32%), compared to Group 1, 2/35 (6%) (P = 0.004). A smaller blood volume was processed in Group 3, 19.8 L, compared to Group 1, 27 L, and Group 2, 25.3 L (P = 0.019). There was no statistically significant difference in collection efficiency between the groups. Disease response of VGPR or better at the time of collection showed a trend to superior HPC CD34 yield (median 1.75 × 10⁶/kg higher) in Group 1; however, this did not reach statistical significance (U = 288, P = 0.194). Median HPC yield in Group 2 with VGPR or better had a median CD34 of 0.2 × 10⁶/kg higher without reaching statistical significance (U = 65.5, P = 0.687). Paradoxically, in Group 3, patients with VGPR or better had a lower CD34 yield (median of 1.46 × 10⁶/kg lower) (U = 235.5, P = 0.037). Group 3 VGPR or better patients were less likely to require plerixafor rescue (50%) compared to those with a lower depth of response (76%), though the association did not meet statistical significance (P = 0.165).

Table 2.

First collection outcome.

	Group 1 >6 weeks lenalidomide PBS plerixafor	Group 2 <6 weeks lenalidomide, PBS plerixafor	Group 3 <6 weeks lenalidomide, Rescue plerixafor	P‐value
Number proceeding to collection	35	16	37
One day collection (n, %)	2 (6)	6 (37.5)	12 (32)	0.004
Total blood volume processed (L)	27	25.3	19.8	0.019
Total CD34 yield (× 106/kg)	6.01	5.58	7.44	0.002
Collection efficiency (%)	103	108	117	0.075
Product cryopreserved volume (mL)	341.5	263.5	335	0.079
Post‐thaw product viability (%)	87.3	86	85.5	0.720
Number proceeding to transplant	32	14	30
Transplant CD34 dose (× 106/kg)	2.75	3.13	3.74	0.001
Engraftment neutrophils (Days)	13	14	14	0.329
Engraftment platelet (Days)	16	18	16	0.245

Limiting analysis to patients who proceeded to collection, the median total CD34 yield was highest with Group 3, 7.44 × 10⁶/kg (Table 2), compared to other groups (P = 0.002). No statistically significant increase in yield was achieved by limiting lenalidomide exposure to less than 6 weeks without the addition of rescue plerixafor. There was no statistically significant difference in cryopreserved product volume or viability between the groups. Of the patients proceeding to melphalan HPC autograft, the median CD34 HPC transplant dose was highest in Group 3 at 3.74 × 10^6/kg (P = 0.001). The engraftment kinetics were similar in all groups.

Discussion

Lenalidomide is a highly effective component of multi‐agent up‐front therapy in MM prior to melphalan autograft. Lenalidomide is known to impair HPC mobilisation, ² with failure to collect sufficient HPC reported in up to 70%. ⁹ As a result, the HPC mobilisation strategy needs to be modified to ensure earlier mobilisation and to consider additional strategies to filgrastim alone. Collecting earlier in the disease course means the maximal depth of response will still be reached in many patients. Although there was a trend to higher CD34 HPC yield in patients with greater than a response of VGPR at collection in Groups 1 and 2, this did not reach statistical significance. In Group 3, where rescue plerixafor was available, there was a higher CD34 HPC collection in patients with less than VGPR, where there was a trend towards more plerixafor use. In our cohort, limiting lenalidomide to less than 6 weeks of treatment did not reduce HPC mobilisation failure rate. The addition of rescue plerixafor was able to eliminate HPC mobilisation failure without the need for cyclophosphamide. The improvement in HPC mobilisation outcome is consistent with other studies where plerixafor was administered routinely. ⁹

Clinician decision for cyclophosphamide HPC mobilisation was markedly reduced from 25%–30% to 3% by the availability of rescue plerixafor. The advantages of avoiding cyclophosphamide are the avoidance of chemotherapy toxicity, more predictable collection kinetics and a more favourable graft for immune reconstitution. Graft immune content appears to play a role in overall survival, ¹⁰ and an optimal graft is more commonly achieved in filgrastim‐only mobilisation. ¹¹ Favourable immune reconstitution after melphalan autograft is of importance for future lines of therapy in the era of immunotherapy with CAR‐T cells and bispecific antibodies.

Access to rescue plerixafor and earlier scheduling of HPC collection resulted in shorter apheresis collection procedure time and a smaller blood volume processed without compromising HPC yield. The shorter procedure time per collection improves overall apheresis unit capacity. As collection toxicity is related to procedure duration, ¹² processing a lower blood volume is expected to improve procedure tolerance, though these data were not specifically collected in the analysis.

Plerixafor resulted in a higher collection‐day WCC. High WCC may negatively impact collection efficiency. ³ In our unit, the prediction of volume to process for a given CD34 pre‐count is modified if the pre‐collection WCC is >20 × 10⁹/L. In our cohort, we found no significant difference in collection efficiency in the rescue plerixafor group, indicating prediction algorithms can overcome any potential negative impact of higher WCCs on collection yield. The higher rate of plerixafor mobilisation in the rescue plerixafor group (Group 3) had no statistically significant impact on product parameters such as product cryopreserved volume or viability.

The study limitations are the retrospective nature of the data and the low overall patient numbers. The more frequent use of plerixafor in the rescue group incurred an increase in drug cost for mobilisation. Although rescue plerixafor cost may be offset by capacity and activity increase, this was not quantified in our study.

Overall, in VRD‐treated MM, the combination of limiting lenalidomide to 6 weeks and access to rescue plerixafor is an effective way to eliminate HPC collection failure. The combination of measures was superior to limiting lenalidomide alone for reducing collection failure without compromising product quality or transplant engraftment.

Data availability statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.