Autologous stem cell transplantation (ASCT) in multiple myeloma (MM) remains a valuable treatment modality in younger patients as a means of deepening response, extending progression-free survival (PFS) and improving overall survival (OS) following initial cytoreductive therapy. Threshold doses of CD34+ cells are required for successful haemopoietic re-constitution when peripheral blood stem cells (PBSCs) are used, but the contribution of other haematopoietic cell types to clinical outcomes is unknown. The importance of re-infused lymphocyte dose (LY-DO) was first suggested by Porrata et al (2004), who observed that patients receiving LY-DO of ≥0·5 × 109/kg had improved PFS and OS. This effect was linked to faster lymphocyte recovery post-ASCT. In contrast, a report in patients receiving upfront ASCT found that LY-DO, although predictive of faster lymphocyte recovery, had no impact on PFS or OS (Hiwase et al, 2008). Faster lymphocyte recovery post-ASCT may directly influence clinical outcome due to faster immunological recovery, mediating, perhaps, an anti-tumour response. Alternatively, it may simply reflect LY-DO, itself a surrogate marker for disease status and/or amount of prior therapy. We performed a retrospective review of patients undergoing ASCT for MM at a single centre. Importantly our cohort included 50 patients receiving CD34-selected PBSCs, thus acting as a control group for a low LY-DO, independent of disease status or prior therapy.
Between 1993 and 2008, 251 patients with MM underwent ASCT at University College Hospital, London, as part of planned upfront therapy (Table 1). The majority (n = 234) were harvested following cyclophosphamide (1·5 g/m2) and granulocyte colony-stimulating factor (GCSF), while 17 were harvested following ESHAP (etoposide, methylprednisolone, cytarabine, cisplatin). Conditioning was with melphalan (250 patients), melphalan/total body irradiation (TBI; n = 33) or cyclophosphamide/TBI (n = 1). Fifty patients received CD34-selected PBSCs (Table 1). Response was defined as per International Myeloma Working Group (IMWG) uniform response criteria for MM (Durie et al, 2006). For statistical analysis of outcome predictors, univariate and multivariate analysis was undertaken using Kaplan–Meier survival estimation. At the time of analysis, the median follow-up for living patients was 39 months; 54 patients were alive and progression-free, 79 had progressed or relapsed and 118 had died. The acute transplant-related mortality was 3·6%. At 6 months following ASCT, 134 (53%) had achieved complete response/very good partial response, 89 (36%) partial response, 10 (4%) stable disease and 18 (7%) had progressed or died. There was no significant difference in disease responses between patients receiving CD34-selected PBSCs and those receiving unselected PBSCs. Median PFS for the whole cohort was 22·3 months and median OS, 62·1 months. Median LY-DO infused to patients receiving unselected PBSCs was 0·09 × 109/kg (range 0·00–3·36 × 109/kg).
|Characteristic||CD34-selected PBSCs (n = 50)||Unselected PBSCs (n = 201)||P-value|
|Age at diagnosis (years; mean ± SEM)||51·57 ± 1·01||56·05 ± 0·52||0·0001|
|Male patients (%)||34 (68)||120 (60)||0·2809|
|Chain isotype (%)|
|IgG||29 (58)||120 (60)||0·96|
|IgA||10 (20)||40 (20)|
|LC||9 (18)||33 (16)|
|BM infiltration at diagnosis (%; mean ± SEM)||45·6 ± 0·04||45·8 ± 0·02||0·9689|
|Number of lines of induction treatment (%)|
|1 line||33 (66)||135 (67)||0·8756|
|2 or more lines||17 (34)||66 (33)|
|Best response pre-HDT (%)|
|CR/VGPR||11 (22)||45 (22)||0·9950|
|PR||31 (62)||122 (61)|
|SD||8 (16)||32 (16)|
|Time from diagnosis to ASCT (d; mean ± SEM)||455·6 ± 73·46||445·9 ± 25·83||0·8794|
|Age at ASCT (years; mean ± SEM)||52·8 ± 1·0||57·2 ± 0·5||0·0002|
There was no significant difference in PFS or OS (P = 0·31 and 0·16 respectively) between patients receiving CD34-selected PBSCs and those receiving unselected PBSCs. For patients receiving unselected PBSCs, univariate analysis of LY-DO (as a continuous variable, Cox regression model) showed no influence on PFS and OS (P = 0·10 and P = 0·36 respectively). Patients receiving unselected PBSCs were then divided into two groups around the median: LY-DO < 0·09 (n = 89) and LY-DO ≥ 0·09 (n = 112). A three-way comparison performed between these two groups and the CD34-selected group showed no significant difference in PFS (Fig 1) or OS (data not shown). Overall, factors that influenced outcome of ASCT were sex, depth of response pre-ASCT (PFS) and progressive disease following induction or post-ASCT (both PFS and OS). None of these factors differed significantly between low and high LY-DO patient groups.
Thus, in a series of patients with MM receiving ASCT as part of planned up-front therapy, and within the range of lymphocyte doses infused in our centre, we report no correlation between LY-DO and clinical outcome. This is supported by the observation that patients receiving CD34-selected PBSCs (i.e. with the lowest LY-DO secondary to patient-independent, ex-vivo modification of the graft) had similar outcome to the patients receiving unselected grafts. The inclusion of the CD34-selected group is validated by reports that CD34-selection does not impact PFS or OS (Vescio et al, 1999; Morineau et al, 2000; Bourhis et al, 2007). Our series differed significantly from that reported by Porrata et al (2004) in the much lower LY-DO infused. This was not due to more intensive mobilizing regimens (LY-DO following ESHAP were equivalent to those following cyclophosphamide priming). The majority of patients in both studies were mobilized using cyclophosphamide and growth factors. However a notable distinction lies in the characteristics of the patients studied. The study reported by Porrata et al (2004) included many patients (around 50%) with relapsed disease who were presumably receiving salvage ASCT. Such patients may have required multiple aphereses to achieve adequate CD34+ cell doses, resulting in higher LY-DO compared with our study and the study of Hiwase et al (2008). The poorer disease status of the patients in the Porrata study would also account for the lower PFS (median 22 months in 85 patients in high LY-DO group, and 15 months in the remaining 149 patients) and OS, when compared with other reports of ASCT in MM (O'Shea et al, 2006), including ours. Indeed, outcomes in the earlier study (Porrata et al, 2004) may be even worse, as these authors defined progression by a 50% increase in M-band (vs. 25% as per IMWG, and our study). In contrast, the majority of our patients underwent ASCT as planned upfront treatment. To investigate whether LY-DO might influence outcome in patients receiving ASCT later in the disease course, we compared outcomes for patients transplanted >1 year from diagnosis with those transplanted earlier, and found no difference in clinical outcomes, or LY-DO. In neither group was LY-DO prognostic for PFS or OS.
In conclusion we find no evidence that the dose of infused lymphocytes has any independent impact on clinical outcomes after ASCT employed as part of planned upfront therapy for MM. The impact of infused lymphocyte dose on clinical outcome may differ in the setting of salvage ASCT, where multiple aphereses may be required, and in this context would deserve further study. This is relevant as several studies today are examining the benefit of delayed ASCT because of high response rates to induction regimens incorporating novel agents.