& 2007 Nature Publishing Group All rights reserved /07 $

Transcription

1 (7), & 7 Nature Publishing Group All rights reserved /7 $. ORIGINAL ARTICLE Patients mobilizing large numbers of CD34 þ cells ( super mobilizers ) have improved survival in autologous stem cell transplantation for lymphoid malignancies BJ Bolwell 1, B Pohlman 1, L Rybicki 2, R Sobecks 1, R Dean 1, J Curtis 1, S Andresen 1, A Koo 1, VMineff 1, M Kalaycio 1 and JW Sweetenham 1 1 Department of Hematologic Oncology and Blood Disorders, Taussig Cancer Center, The Cleveland Clinic Foundation, Cleveland, OH, USA and 2 Department of Quantitative Health Sciences, Taussig Cancer Center, The Cleveland Clinic Foundation, Cleveland, OH, USA The cellular composition of an autologous graft may influence autologous stem cell transplantation (ASCT) outcome. Etoposide (VP) plus filgrastim (G) frequently mobilizes high numbers of CD34 þ cells for autologous transplantation. We investigated whether patients collecting high numbers of CD34 þ cells ( super mobilizers ) have a better outcome than other patients. We reviewed 3 consecutive adult patients with NHL or Hodgkin s lymphoma receiving an ASCT from January 1994 to December 5, mobilized with VP þ G. Super mobilizers were defined as collecting a minimum of 8 6 CD34 þ cells/kg. Two hundred and three patients were super mobilizers, while 147 collected between 2. and 7.95 CD34 þ cells/kg. Super mobilizers were younger and more likely to have received two or fewer prior chemotherapy regimens ( versus 63%, Po.1). Median CD34 þ cell dose for the super mobilizing group was versus /kg in the standard collecting group. The super mobilizer group had a superior overall survival (P ¼.6). In multivariable analysis, favorable disease status and younger age at transplant, and super mobilization were associated with improved survival. We conclude that patients had an improved ASCT outcome if large numbers of CD34 þ cells were mobilized and infused. The explanation for this observation is unknown. (7), ; doi:.38/sj.bmt.15763; published online 9 July 7 Keywords: stem cells; mobilization; stem cell transplant; lymphoma Introduction Prognostic factors in hematopoietic stem cell transplantation are well described. 1 3 Disease status at the time of Correspondence: Dr BJ Bolwell, The Cleveland Clinic Foundation, Euclid Avenue, Desk R32, Cleveland, OH 44195, USA. MINEFFV@ccf.org Received March 7; revised 11 May 7; accepted 4 June 7; published online 9 July 7 transplant, International Prognostic Index and other variables are associated with outcomes in autologous stem cell transplantation (ASCT). Degree of HLA match, and age, among others, are associated with outcome in allogeneic transplantation. Sparse data exists examining whether the cellular composition of the infused product affects clinical transplant outcome. Total nucleated cell dose correlates with outcome in umbilical cord blood transplants, 4 and increased T cells infused in peripheral blood progenitor cell (PBPC) allogeneic transplants probably contribute to greater and more severe chronic graft-versus-host disease (GVHD). 5 The infused absolute lymphocyte count has been reported to correlate with clinical outcome in ASCT for patients with non-hodgkin s lymphoma. 6 Additionally, increasing dendritic cell count in the autologous hematopoietic cell graft have been associated with improved survival in ASCT for diffuse large B-cell lymphoma. 7 No study has examined whether total number of PBPCs influences survival. There is no optimal PBPC mobilizing strategy. However, etoposide (VP) plus filgrastim (G-CSF) has been reported to be an effective mobilizing regimen. 8,9 Indeed, this is the most common mobilizing regimen that our institution uses in preparation of ASCT for patients with lymphoid malignancies. We have anecdotally noted that a significant percentage of patients achieve extremely high yields of CD34 þ cells with VP þ G. Given that the cellular composition of the autologous stem cell product might influence outcome, we asked the straightforward question of whether or not patients mobilizing large numbers of CD34 þ cells had a more favorable clinical outcome than did those with lower CD34 þ cell yields. Materials and methods Patient population Six hundred and ninety three adult patients (X18 years old) were identified from the Unified Transplant Database of the Cleveland Clinic Taussig Cancer Center who received an ASCT from January 1994 to December 5 with either Hodgkin s lymphoma or non-hodgkin s lymphoma. All

2 438 were treated with a uniform preparative regimen of busulfan, cyclophosphamide and etoposide. Of these 693 patients, 3 were primed with a combination of VP þ G and comprised the study population. All patients were treated on protocols approved by the Institutional Review Board of the Cleveland Clinic and gave written informed consent. PBPC mobilizing regimen All patients received filgrastim (G-CSF) plus etoposide (2 g/m 2 ) plus G-CSF for PBPC mobilization. Leukapheresis was begun once total WBC recovery was 4 per mm 3. PBPC collection was performed with a COBE spectra leukapheresis machine (COBE, Denver, CO, USA). Patients underwent apheresis for at least 2 days, with a minimum collection goal of 2. 6 CD3 þ cells/kg. Apheresis continued for up to 5 days, or when 7. 6 CD34 þ cells were collected, whichever came first. CD34 determination by flow cytometry has been previously described. For the purpose of this analysis, patients were split into two groups based on CD34 þ mobilization yield. Patients collecting and infusing X8 6 cells/kg are deemed super mobilizers. Patients collecting 42 and o8 6 CD34 þ cells/kg are deemed normal mobilizers. All patients collecting o8 6 CD34 þ cells/kg received all collected cells; all patients collecting 48 6 CD34 þ cells/kg received at least 8 6 CD34 þ cells/kg. A small number of patients collected CD34 þ cells/kg and had a portion of cells cryopreserved, but all received well over 8 6 CD34 þ cells and thus remained in the super mobilizer group. Statistical analysis Recursive partitioning analysis with a log-rank splitting method was used to identify the cut point in CD34 þ collection that best correlated with survival. This cut point was used to define super mobilizers and normal mobilizers, and was used for all subsequent analyses. Continuous baseline characteristics were compared between super mobilizers and normal moblizers using either the t-test or Wilcoxon rank-sum test, as appropriate. Categorical characteristics were compared using the w 2 test. Survival was calculated from the date of transplant to the date of last follow-up or to the date of death. Freedom from relapse was calculated from the date of transplant to the date relapse or to the date of last follow-up. Relapse-free survival was calculated from the date of transplant to the date of relapse, death or last follow-up. Among patients who relapsed, survival following relapse was calculated from the date of relapse to the date of last follow-up or to the date of death. Outcomes were estimated using the Kaplan Meier method and compared between mobilizing groups using the log-rank test. Univariable Cox proportional hazards analysis was used to assess the prognostic importance of nine variables: age at transplant, gender, race, primary diagnosis, number of prior chemotherapy regimens, prior radiation therapy, disease status at transplant, LDH at transplant and mobilizing group. Multivariable Cox proportional hazards analysis was then performed to determine the prognostic importance of mobilization after adjusting for any variable which either differed between mobilizing groups, or any variable which was a significant univariable prognostic factor for survival. All analyses were done using SAS software (SAS Institute Inc., Cary, NC, USA). All statistical tests were two-sided, and Po.5 was used to indicate statistical significance. Results Patient characteristics are shown in Table 1. In addition to variables shown, pre-transplant performance status, left ventricular ejection fraction and pulmonary function (percent predicted forced expiratory volume in 1 second (FEV1) and percent predicted diffusion capacity (DLC)) were similar between the two groups. The super mobilizer group required fewer days of pheresis for stem cell harvest (median 2 versus 5 days, Po.1). The median CD34 þ cell dose of the super mobilizer group was versus CD34 þ cells/kg in the normal collecting group. Super mobilizers had more rapid engraftment of neutrophils and platelets. Median neutrophil engraftment was days in the super mobilizing group versus 11 days in the normal mobilizing group (Po.1); median platelet engraftment was 12 days in the super mobilizing group versus 16 days in the normal mobilizing group (Po.1). Overall survival was superior in the super mobilizing group, as shown in Figure 1. The freedom from relapse and relapse free survival curves are shown in Figures 2 and 3, respectively. Risk of non-relapse mortality is shown in Figure 4. One hundred and thirty-four (38%) patients have relapsed in the entire group; there is no difference in the incidence of relapse between the two mobilizing groups (P ¼.27; log-rank test). Of 136 patients who have died, relapse is the most common reason in both groups. Kaplan Meier estimates of survival after relapse and relapse-related mortality are shown in Figures 5 and 6, respectively. These show lower relapse-related mortality, and more prolonged survival after relapse in the super mobilizer group. Among the nine variables examined in univariable analysis, four were found to be prognostic for improved survival: younger age at transplant (P ¼.17), fewer prior chemotherapy regimens (P ¼.8), disease which is sensitive to chemotherapy (P ¼.7) and super mobilization (P ¼.6). After adjusting for mobilizing group differences and univariable prognostic factors, super mobilization remained prognostic for improved survival (hazard ratio ¼.69, 95% confidence interval ¼.49.98, P ¼.38). Younger age (P ¼.33) and disease which is sensitive to chemotherapy (P ¼.4) also remained prognostic for improved survival. Discussion Much of the literature concerning PBPC mobilization is limited by heterogeneity of patients and treatment regimens. Most efficacy trials target an optimal dose of 5 6 CD34 þ cells/kg, as this threshold value has been

3 Table 1 Patient characteristics 439 Variable All patients (3) CD34+ o8 (147) CD34+ X8 (3) P-value N % N % N % Age (years) Mean7s.d * Median (range) 51 (18 77) 53 (18 77) 49 (19 75) Primary diagnosis NHL HD NHL grade Low Intermediate High Other Number of prior chemotherapy regimens o.1* Prior radiation therapy Yes No Disease status at transplant Sensitive disease Relapse/refractory LDH p Abbreviation: HD, Hodgkin s disease. *Significantly different between groups (Po.5). Overall survival (%) Figure 1 P = /kg (n=3) <8 6 /kg (n=147) Overall survival. Freedom from relapse (%) Figure 2 P= /kg (n=3) <8 6 /kg (n=147) Freedom from relapse. associated with more rapid platelet engraftment than has lower values. 11 There are also abundant data concerning characteristics associated with poor PBPC mobilization, in which much of the data suggests that increasing prior exposure to chemotherapy is associated with problematic mobilization. 12,13 The composition of the infused cellular product has long been known to potentially influence outcome in allogeneic transplantation. For example, T-cell depletion is known to reduce the risk of GVHD at the expense of disease relapse. 2 However, few reports examine the potential influence of the cellular infused product in autologous transplantation.

4 4 Relapse-free survival (%) Figure /kg (n=3) <8 6 /kg (n=147) P= Relapse free survival. Survival after relapse (%) Figure 5 <8 6 /kg (n=59) P= /kg (n=75) Survival after relapse. Non-relapse mortality (%) Figure 4 <8 6 /kg (n=147) 8 6 /kg (n=3) P= Risk of non-relapse mortality. Relapse-related mortality (%) Figure 6 P=.12 <8 6 /kg (n=147) 8 6 /kg (n=3) Relapse-related mortality. Recent reports suggest that the absolute lymphocyte count and the dendritic cell count in the infused product correlate with outcome in autografting in patients with lymphoid malignancies. We have demonstrated that a higher number of CD34 þ cells is associated with improved survival in a cohort of patients with lymphoid malignancies treated with uniform mobilization and transplant preparative regimens. In view of the retrospective nature of this study, we are unable to investigate whether the presence of high numbers of CD34 þ cells correlates with numbers of other cell populations including lymphocytes and dendritic cells. This requires further prospective evaluation. Our results show a correlation between mobilization of high numbers of CD34 þ cells and overall survival. The observed correlation between super mobilizers, lower age, disease status at transplant and fewer prior chemotherapy regimens suggests that the improved overall survival in this group is explained in part by the fact that the super mobilizer group includes more patients with relatively favorable underlying disease. However, freedom from relapse and relapse-free survival did not correlate with high CD34 þ cell numbers, suggesting that other factors may also have contributed. Earlier neutrophil and platelet engraftment was observed in the super mobilizer group, but although this was statistically significant, its clinical significance is unclear. Neutrophil engraftment was 1 day earlier in the super mobilizer group, which did not have a major impact on the rate of death from sepsis. Part of the explanation for the superior long-term survival in the super mobilizer group relates to the course of the disease from the point of relapse after ASCT. As Figures 5 and 6 demonstrate, despite similar rates of relapse in the two groups, the relapse mortality and survival after relapse differ strikingly with lower relapse-related mortality and more prolonged survival in the super mobilizer group. We investigated the possibility that this relates to an overrepresentation of indolent histologic subtypes among those who have prolonged survival after relapse, but this was not the case. This raises the possibility that other cellular components of the reinfused autologous product may inhibit subsequent disease progression. From a practical perspective, while our observations concerning the CD34 þ cell content of the infused product are interesting, few data exist describing strategies, which can either increase the CD34 þ yield above standard

5 target numbers or manipulate the cellular composition in other ways. Generally, patients are treated using mobilization strategies that vary between centers, aimed at collecting a target threshold dose of CD34 þ cells, which are most commonly around 5 6 /kg. Few reports, if any, demonstrate that collecting extremely large numbers of CD34 þ cell collection is even possible. This report has demonstrated that the combination of VP þ G can generate a very large CD34 þ cell yield. A theoretic increased risk of secondary malignancies after transplant when using etoposide as part of a mobilization strategy has not been realized. 14 Therefore, one potential therapeutic strategy as a result of this data might be to treat more patients with etoposide-based mobilization regimens. Additionally, the new agent AMD 3 has been shown to generate robust CD34 þ cell yields when used in combination with G-CSF. 15,16 Therefore, if the present data are confirmed in a prospective study, the use of AMD 3, or other agents, to generate larger numbers of CD34 þ cell collections might improve clinical outcome. The use of high-dose therapy and ASCT has been a standard of care for patients with relapsed lymphoid malignancy for almost years. However, despite improvements in supportive care for patients receiving this therapy, the outcome for these patients has not improved dramatically over the past 15 years. Our results suggest that altering the cellular composition of the infused cellular product of the graft is a potential treatment strategy to improve outcomes, which warrants additional prospective study. In summary, this retrospective study has demonstrated that the collection and reinfusion of large numbers of CD34 þ cells, in excess of 8 6 /kg is associated with improved survival in patients undergoing ASCT for lymphoid malignancies. Although this is probably partly explained by the preponderance of patients with favorable disease-related factors in the super mobilizer group, the failure to demonstrate a difference in relapse rate suggests that other mechanisms are also involved. Our results suggest that further prospective studies investigating the use of improved mobilization regimens to maximize CD34 þ yields are needed. Whether the observed improvement in survival is related directly to CD34 þ content, or to other cellular components of the reinfused product requires further prospective evaluation. References 1 Bolwell BJ. Are predictive factors clinically useful in bone marrow transplantation? Bone Marrow Transplant 3; 32: Copelan EA. Hematopoietic stem cell transplantation. New Engl J Med 6; 354: Bolwell BJ. Factors predicting success or failure associated with common types of transplants. Pediatr Transplant 5; 9: Gluckman E, Rocha V, Arcese W, Michel G, Sanz G, Chan K et al. Factors associated with outcomes of unrelated cord blood transplant: Guidelines for donor choice. Exp Hematol 4; 32: Stem Cell Trialists Collaborative Group. Allogeneic peripheral blood stem-cell compared with bone marrow transplantation in the management of hematologic malignancies: an individual patient data meta-analysis of nine randomized trials. J Clin Oncol 5; 23: Porrata LF, Litzow MR, Inwards DJ, Gastineau DA, Moore SB, Pineda AA et al. Infused peripheral blood autograft absolute lymphocyte count correlates with day 15 absolute lymphocyte count and clinical outcome after autologous peripheral hematopoietic stem cell transplantation in non-hodgkin s lymphoma. Bone Marrow Transplant 4; 33: Dean R, Masci P, Pohlman B, Andresen S, Serafino S, Sobecks R et al. Dendritic cells in autologous hematopoietic stem cell transplantation for diffuse large B-cell lymphoma: graft content and post transplant recovery predict survival. Bone Marrow Transplant 5; 36: Copelan EA, Ceselski SK, Ezzone SA, Lasky LC, Panza SL, Bechtel TP et al. Mobilization of peripheral-blood progenitor cells with high-dose etoposide and granulocyte colony-stimulating factor in patients with breast cancer, non-hodgkin s lymphoma, and Hodgkin s disease. J Clin Oncol 1997; 15: Kanfer EJ, McGuigan D, Samson D, Abboudi Z, Abrahamson G, Apperley JF et al. High-dose etoposide with granulocyte colony-stimulating factor for mobilization of peripheral blood progenitor cells: efficacy and toxicity at three dose levels. Br J Cancer 1998; 78: Bolwell BJ, Goormastic M, Yanssens T, Dannley R, Baucco P, Fishleder A. Comparison of G-CSF with GM-CSF for mobilizing peripheral blood progenitor cells and for enhancing marrow recovery after autologous bone marrow transplant. Bone Marrow Transplant 1994; 14: Siena S, Schiavo R, Pedrazzoli P, Carlo-Stella C. Therapeutic relevance of CD34+ cell dose in blood cell transplantation for cancer therapy. J Clin Oncol ; 18: Boeve S, Strupeck J, Creech S, Stiff PJ. Analysis of remobilization success in patients undergoing autologous stem cell transplants who fail an initial mobilization: risk factors, cytokine use and cost. Bone Marrow Transplant 4; 33: Ford CD, Green W, Warenski S, Petersen FB. Effect of prior chemotherapy on hematopoietic stem cell mobilization. Bone Marrow Transplant 4; 33: Copelan E, Hoshaw-Woodard S, Elder P, Penza S, Farag S et al. Therapy-related myelodysplasia and leukemia occur infrequently following VP-16 priming and autotransplantation without total body irradiation. Bone Marrow Transplant 4; 34: Broxmeyer HE, Orschell CM, Clapp DW, Hangoc G, Cooper S, Plett PA et al. Rapid mobilization of murine and human hematopoietic stem and progenitor cells with AMD3, a CXCR4 antagonist. J Exp Med 5; 1: Devine SM, Flomenberg N, Vesole DF, Liesveld J, Weisdorf D, Badel K et al. Rapid mobilization of CD34+ cells following administration of the CXCR4 antagonist AMD3 to patients with multiple myeloma and non- Hodgkin s lymphoma. J Clin Oncol 4; 22: