Abstract
Introduction
In 2013, bendamustine/rituximab (BR) replaced RCHOP as standard first line treatment for both transplant eligible and ineligible MCL patients (pts) in BC. Retrospective cohort studies report that bendamustine has no adverse effect on peripheral blood stem cell (SC) mobilization but this is discordant with local experience. We sought to compare rates of failed SC collection in MCL pts planned for high dose chemotherapy and autologous stem cell transplant (ASCT) after BR or RCHOP and identify risk factors for failed SC mobilization and collection.
Methods
We identified all pts with MCL in BC treated with BR or RCHOP as first line therapy who underwent SC mobilization from Jan. 1 2003-Dec. 31 2017 using the Leukemia/Bone Marrow Transplant Program of BC and Apheresis Database Standard mobilization was with G-CSF alone (G) until difficulties with SC collection were noted after BR. Different mobilization strategies were then used, including delaying SC collection 2-3 mos after BR, G + cyclophosphamide (G+C) mobilization and/or "rescue" plerixafor if Day 1 SC collection was inadequate. Failure of SC collection was defined as yield <1 x106 CD34+ cells/kg on apheresis Day 1 (D1fail). . All variables significant in univariate analysis (P<0.1) were entered into a multivariate analysis (MVA) logistic regression model to identify factors associated with D1 fail.
Results
A total of 152 pts were identified. 2 were excluded as they had pre-emptive plerixafor prior to apheresis D1 on a trial. Of the remaining 150 pts, 55 (37%) received BR, 95 (63%) RCHOP for a median of 6 cycles (range 1-6). Baseline characteristics were similar between groups (Table 1). Pts receiving BR had higher remission rates compared to RCHOP (CR 58% vs 35%, P=.004) and a longer duration from their last chemotherapy to apheresis D1 (BR 89 d vs RCHOP 39 d, P<0.), reflecting an intentional change in practice. For this same reason, use of G+C mobilization was higher in BR pts (45%) compared with RCHOP pts (1%).
Failure on D1 of SC collection was significantly higher after BR compared to RCHOP (45% vs. 10%, P<.001). This difference persisted when only pts mobilized with G alone were included: D1fail 60% BR vs 10% RCHOP (P<.001), Fig. 1. Among BR pts, D1fail was higher for those mobilized with G compared with G+C (60% vs. 28%, P=.03).
Of the 25 BR pts with D1fail, 20 pts received rescue plerixafor (P) (1 dose: 18, 2 doses: 1, 3 doses: 1) with 17 successfully collecting > 2.0 x106 CD34+ cells/kg, 2 successfully collected after a 2nd round of SC mobilization (Pt1: G+C+P, Pt2: G+P), and 1 pt did not undergo ASCT due to collection failure. Of the 5 pts who did not receive plerixafor, 2 successfully collected with 1-2 additional apheresis days and 3 successfully collected with a 2nd round of SC mobilization (2: G+P, 1: G+C). Among the 9 RCHOP pts with D1fail, 1 had rescue plerixafor with success, 3 successfully collected after 1-2 more apheresis days and 5 had a 2nd round of SC mobilization (1: G, 4: G+P), of which 1 failed and did not proceed to ASCT.
Univariate analysis identified the following associated with D1fail: frontline therapy (BR vs RCHOP, P=<.001), mobilization regimen (G+C vs G, P=.001), gender (P=.002) and D1 platelet count (P=<.001). Delaying SC mobilization in pts who received BR did not improve D1 yield (P=.31). In MVA, frontline therapy with BR had an increased risk of D1fail compared to RCHOP with an odds ratio (OR) of 7.83 (95% CI 2.7-23.1), P<.001. Using G+C significantly improved the odds of a successful collection compared to G alone (OR for D1fail 0.17 [95% CI 0.05-0.62], P=.007). Female gender and low platelet count also increased risk of D1fail in MVA (Table 2).
Conclusion
MCL pts who receive BR as first line therapy have an almost 8-times higher risk of failure of D1 SC collection compared to those who receive RCHOP. This failure can be overcome largely with "rescue" plerixafor and partially with planned G+C mobilization such that successful collection can be achieved in >90% of pts. However, these agents add cost and potential toxicity. Platelet count on D1 SC collection is a potential indicator of high risk of failure. Delaying SC mobilization in BR pts does not improve collection yield. The impact of bendamustine in pre-transplant chemotherapy regimens, not only for MCL but also for a broad range of lymphomas where it is being increasingly used, must be considered when planning SC mobilization strategies.
Disclosures
Sehn: Karyopharm: Consultancy, Honoraria; Janssen: Consultancy, Honoraria; Lundbeck: Consultancy, Honoraria; Merck: Consultancy, Honoraria; Abbvie: Consultancy, Honoraria; TG Therapeutics: Consultancy, Honoraria; Roche/Genentech: Consultancy, Honoraria; Seattle Genetics: Consultancy, Honoraria; Amgen: Consultancy, Honoraria; Morphosys: Consultancy, Honoraria; Celgene: Consultancy, Honoraria.