Purging of G-CSF-mobilized peripheral autografts in acute leukemia with mafosfamide and amifostine to protect normal progenitor cells

Transcription

1 (2000) 25, Macmillan Publishers Ltd All rights reserved /00 $ Purging of G-CSF-mobilized peripheral autografts in acute leukemia with mafosfamide and amifostine to protect normal progenitor cells F Fauth 1, H Martin 1, S Sonnhoff 1, H Bialleck 2, M Wiesneth 3, B Mihanovic 4 and D Hoelzer 1 1 Klinikum der Johann Wolfgang Goethe Universität, Dept of Hematology and Oncology, Frankfurt; 2 Institut für Transfusion und Immunhämatologie, DRK Hessen, Frankfurt am Main; 3 Blutspendedienst des DRK Baden-Württemberg, Ulm; and 4 Klinikum der Universität Ulm, Dept of Hematology, Ulm, Germany Summary: In the present study the in vitro growth of CFU-GM from PBPC of patients with AML (n = 11), purged with mafosfamide alone or a combination of mafosfamide and amifostine, was compared to historical controls of mafosfamide-purged bone marrow (AML CR1, n = 16). Two patients were transplanted with mafosfamide and mafosfamide/amifostine pretreated PBPC autografts. The in vitro experiments demonstrated a significantly higher resistance of peripheral blood derived CFU-GM to mafosfamide (median ID g mafosfamide/ml) compared with bone marrow derived CFU-GM (median ID g/ml). Preincubation with amifostine significantly further increased the median ID 95 to 245 g/ml. The clinical results showed short recovery times for neutrophils 500/ l (9 and 13 days) and platelets / l (12 and 21 days) and stable long-term engraftment with one relapse at day 118 and one patient in CR at day 760 after transplantation. The in vitro results show a significant advantage of PBPC over bone marrow-derived progenitors for purging with mafosfamide. Furthermore, a protective effect from mafosfamide of amifostine on normal progenitors could be demonstrated. The clinical results demonstrate the clinical feasibility of using mafosfamide-purged autologous PBPCT without impairing the short-term and longterm repopulating capacities of the autografts. Bone Marrow Transplantation (2000) 25, Keywords: autologous bone marrow transplantation; acute myeloblastic leukemia; mafosfamide; amifostine; cyclophosphamide derivatives Autologous bone marrow transplantation (ABMT) is a therapeutic approach for the treatment of acute leukemia. The possibility of graft contamination with leukemic cells leading to relapse as shown by gene marker studies for bone marrow 1 led to the development of purging strategies for elimination of residual leukemic progenitors in bone marrow autografts. In several studies a positive effect of mafos- Correspondence: Dr F Fauth, Klinikum der Johann Wolfgang Goethe Universität, Medizinische Klinik III Hämatologie/Onkologie, Knochenmarktransplantation, Theodor-Stern-Kai 7, Frankfurt, Germany Received 23 June 1999; accepted 10 December 1999 famide purging on patient survival could be demonstrated for bone marrow transplantation. 2,3 Reports of possible mobilization of leukemic progenitors into PB-derived autografts stress the importance of an effective purging before peripheral progenitor cell transplantation. 4,5 For mafosfamide-purging of G-CSF-mobilized peripheral blood progenitor cells (PBPC) some in vitro data with leukemic cell lines added to bone marrow and PBPC are available. 6 However, there are hitherto no data regarding the clinical feasibility of autologous PBPCT with mafosfamide- and amifostine-purged autografts. We purged aliquots of leukapheresis products from patients with de novo (n = 8) or secondary AML (n = 3) with mafosfamide alone or mafosfamide and amifostine. Amifostine is known for its protective effect on bone marrow progenitor cells exposed to alkylating agents, and also for its ability to increase the toxicity of mafosfamide against leukemic progenitor cells. 1 The in vitro results were compared to a group of mafosfamide-purged bone marrow autografts from 16 patients with AML in first complete remission. 8 To investigate whether mafosfamide-purged PBPCT is feasible as an alternative treatment to ABMT in vivo we transplanted two patients with PBPC autografts purged with mafosfamide and mafosfamide and amifostine. The purposes of the study were: to compare the effects of mafosfamide purging on CFU-GM from bone marrow vs G-CSFmobilized PBPC; to evaluate the protective effect of amifostine on non-malignant PBPC in mafosfamide purging of PBPC; and to show the clinical feasibility of mafosfamide and amifostine-purged autologous PBPCT. Patients and methods Leukaphereses All autologous PBPC grafts were prepared in the local blood banks of the participating institutions. Patients We investigated aliquots of stem cell aphereses from 11 patients (6 male, 5 female, 8 AML, 3 secondary AML). The median age was 51.1 years. Statistics are summarized in Table 1.

2 832 Table 1 Patients characteristics Sex Age Diagn No. cycles Last cycle Time last cycle No. CD34 + yield before apheresis before apheresis first apheresis aphereses M 23 AML 3 MHD Ara-C/m-AMSA M 43 AML 3 MHD Ara-C/m-AMSA F 64 AML 3 DAV II F 25 AML 4 HD Ara-C/m-AMSA M 62 AML 3 MHD Ara-C/m-AMSA M 51 AML 3 MHD Ara-C/DNR F 41 AML 3 MHD Ara-C/DNR F 19 AML 3 MHD Ara-C/DNR M 57 sek.-aml 4 MHD Ara-C/m-AMSA M 63 sek.-aml 3 IVA III F 53 sek.-aml 3 MHD Ara-C/m-AMSA Chemotherapy All patients with de novo AML were diagnosed and treated with induction chemotherapy and consolidation therapy as described by Seipelt et al. 9 Patients with secondary AML underwent induction therapy with two courses of daunorubicin or doxorubicin, ara-c and etoposide followed by consolidation therapy with one or two courses of MHD ara-c and m-amsa or daunorubicin. Patient No. 6 did not respond to induction I and was given MHD-ara-C as a second course. She achieved a CR and was given DAV again as third course before leukaphereses followed by consolidation with MHD-ara-C. Patient No. 9 received three courses of idarubicin, ara-c and etoposide. His condition did not allow intensification after induction therapy. Informed consent was given by all patients before entering the study. CFU-GM A methylcellulose CFU-GM assay as described by Ottmann et al 10 was used. To 50% (v/v) methylcellulose 1.125% the following reagents were added: FCS (Hyclone, Logan, UT, USA) 30% (v/v) G-CSF 10 ng/ml (1% v/v) (Amgen, USA); GM-CSF 10 ng/ml (1% v/v) (Behring, Marburg, Germany); mercapto-ethanol (Gibco BRL, Grand Island, NY, USA); 0.1 mmol (1% v/v); glutamine (Gibco BRL) 4 mmol (1% v/v); cells ( MNC/ml; 10% v/v), basal Iscove s medium (Seromed; 6% v/v). Colonies were scored on day +14 with an inverted microscope. Determination of ID 95 The ID 95 for all samples was calculated as described by Martin et al. 11 The colony numbers at different mafosfamide concentrations were displayed on a semilogarithmic scale. Linear regression was calculated and the concentration value corresponding to 95% growth inhibition (ID 95 ) was extrapolated from the graph as a parameter for the toxicity of mafosfamide. In vitro purging of aliquots MNC were separated from aliquots from chemotherapy and G-CSF-mobilized leukapheresis products of TNC by a Ficoll density gradient. They were then washed in S- MEM (Gibco BRL) and 30% FCS (Hyclone) and resuspended to a concentration of MNC/ml. Amifostine (Ethyol; Essex Pharma, Munich, Germany) was added to 60% of the cells at a concentration of 3 mg/ml to a resulting cell concentration of MNC/ml. Cells were incubated at 37 C for 30 min. Incubation was followed by washing twice with S-MEM and resuspension to a concentration of MNC/ml (n = 8). Cells were then incubated with mafosfamide (Asta, Frankfurt, Germany) at concentrations of 0, 60, 90, 120, 150 and 180 g/ml at 37 C for 30 min. The cell concentration was MNC/ml. The reaction was stopped after 30 min by washing twice with cold S-MEM. Cells were resuspended and plated in the CFU-GM assay (n = 11). In vitro purging of transplants (n = 4) Only leukapheresis products with more than CD34 + cells/kg were eligible for this procedure since we expected loss of cells during the purging process and the threshold for autologous PBPCT in our institution is CD34 + cells/kg. The transplants for clinical use were purged as follows: a whole leukapheresis product was washed twice in saline and human serum albumin (HSA) in a concentration of 2% to eliminate plasma, which is known to have an inhibitory effect on mafosfamide; 12 after washing cells were resuspended in S-MEM medium (Gibco BRL) and 2% human serum albumin (DRK Blutspendedienst, Hessen, Germany) was added; amifostine was then added in a concentration of 3 mg/ml. Cells were incubated at 37 C for 30 min. After incubation amifostine was eliminated by washing twice with saline and 2% humane serum albumin. Cells were then resuspended in S-MEM and 2% HSA. For mafosfamide incubation, sterile filtered mafosfamide was added at the appropriate concentration. Since red cells

3 also influence mafosfamide toxicity, 13 the suspension was adjusted to a hematocrit of 5%. Due to the large volume of the cell suspension ( ml) stopping of the mafosfamide activity by adding cold saline was not possible because the total bag content was only 600 ml. Cells were therefore incubated at 37 C for 20 min, and the reaction was then slowed down on ice for 10 min before centrifugation in a cold (4 C) centrifuge for 12 min at 1000 U/min. Cells were then washed in saline and HSA 2% once before they were resuspended in 100 ml blood group identical fresh frozen plasma. Total nucleated cell counts (TNC) and CD34 + cell counts were measured at the beginning and at the end of purging. CFU-GM assays were plated as for the aliquots to monitor CFU-GM growth. Parallel to purging the graft, aliquots were purged as described above. Statistical analysis All data shown are median values. For statistical analysis, we used the Mann Whitney test for non-paired observations and the Wilcoxon matched pairs test. Informed consent All patients and donors gave informed consent for the use of aliquots of leukapheresis products for this study. The recipients of purged autologous PBPCT gave informed consent to receiving the purged autografts. Results Comparison of mafosfamide-purged PBPC with BM autografts The median ID 95 of mafosfamide-purged PBPC was 190 (range ) g/ml (n = 11) for all samples. It significantly exceeded the median value of 130 g/ml of a previously published cohort of CR1-AML-BM autografts (n = 16) (P = ) (Figure 1). Comparison of mafosfamide and amifostine-purged PBPC with mafosfamide-purged PBPC There was a significant difference between both PBPC groups (n = 8; P = ) in favor of the mafosfamide and amifostine group. The median ID 95 were 245 ( ) mg/ml for the mafosfamide and amifostine group and 190 (90 260) mg/ml for the mafosfamide group as shown in Figure 2. Autologous transplantation using mafosfamide-purged PBPC Four patients had a leukapheresis purged with mafosfamide, three were pretreated with amifostine. Two patients received an allogeneic transplant, two AML patients were transplanted with mafosfamide-purged PBPC. Patient No. 2 had an aberrant karyotype at time of initial diagnosis. This aberration was not detectable on day +78 after auto- ID95 ( µ g/ml) P = >330 BMT AML-CR1 Mafo PBPC n = 16 n = 11 Figure 1 The ID 50 (mafosfamide concentration leading to 95% growth inhibition of CFU-GM) of historical BM controls and aliquots from leukaphereses, purged with mafosfamide (Mafo) are shown. ID95 ( m g/ml) Mafo P = >330 Mafo + Amif. Figure 2 A direct comparison between the ID 50 of mafosfamide alone vs mafosfamide and amifostine in eight cases of AML is shown. logous PBPCT. Nevertheless the patient relapsed on day +118 after PBPCT and died on day Adapted-dose purging, as established in bone marrow purging needs days for evaluation of the CFU-GM assays. In the setting of PBPC mobilization a different approach is needed since patients are pancytopenic 2 weeks prior to peak mobilization. Thus standard dose purging was used in these patients and the ID 95 was evaluated retrospectively. Table 2 shows the general patient characteristics and the data regarding the grafts and regeneration. Discussion The cyclophosphamide derivatives mafosfamide and 4-hydroperoxy-cyclophosphamide (4-HC) have both been used for purging of bone marrow autografts to reduce contamination of tumor cells. Brenner et al 1 observed in a study with 20 children with AML or neuroblastoma, undergoing ABMT after marking the autografts with a retroviral vector, that in three cases of relapse the tumor cells contained the retroviral vector. This proved that cells originating from the graft can contribute to relapse in ABMT. 833

4 834 Table 2 Transplanted patients characteristics Sex Age Diagnosis Purging Regeneration OAS PB-ID 95 Mafo WBC PLT days ( g/ml) ( g/ml) /l /l M 23 AML M (CR) (+amif 3 mg/ml) M 57 AML M (PD) For PBPC, reports of relapse after PBPCT suggest that mobilized leukemic progenitor cells can contribute to relapse in a similar way to bone marrow. 13,14 Several groups have demonstrated the clinical feasibility of mafosfamide-purged ABMT with good clinical results. 13,15,16 Gorin 2,3 and Laporte et al 17 have shown in a series of 125 patients receiving mafosfamide-purged autografts that high CD34 + counts before purging correlates with a low transplant-related mortality and aggressive purging correlates with a low risk of relapse. It was further shown that the best results in vivo are achieved when the number of CD34 + cells in the graft is very high and the graft is then purged as aggressively as possible. A problem occurring with the use of these transplants is the prolonged regeneration times for neutrophils and platelets aggravating the well-known problem of delayed regeneration after autografting in AML. Straetmans et al 18 demonstrated a significantly decreased production of adherent and nonadherent progenitors in cryopreserved AML autografts in comparison with normal donors in long-term bone marrow cultures, although they could not correlate this to the delayed engraftment in vivo. Miggiano et al 19 observed prolonged regeneration of neutrophils (26 days, range ) and platelets (74 days, range ) in a series of 51 patients undergoing ABMT in AML in first CR. In our own institution cases of up to 697 days of platelet transfusion dependence have been observed after ABMT with mafosfamide-purged bone marrow (Fauth et al, unpublished data). This problem is aggravated by purging especially when the autograft contains low numbers of progenitor cells before purging. Demirer et al 20 observed a higher mortality for patients undergoing 4-HC-purged ABMT in comparison with patients undergoing unpurged ABMT by day +100 (50% vs 20%). The percentage of patients achieving granulocyte levels of more than /l and independence of platelet transfusions were lower (58 vs 91% and 42 vs 61%) at that time. Only 33/151 patients were in first CR, therefore these data are not directly comparable to the data of Laporte et al, nevertheless they show a risk of further prolongation of regeneration of peripheral blood counts by purging the autograft. In the present study, a clear advantage of peripheral blood progenitor cells over bone marrow is demonstrated in vitro in terms of resistance of CFU-GM to mafosfamide. The mafosfamide concentration sparing 5% of CFU-GM in mafosfamide-purged PBPC was 190 (range ) g/ml (n = 11) and significantly exceeded the median value of 130 g/ml of a previously published cohort of CR1-AML- BM autografts (n = 16) (P = ). Preincubation with amifostine further increased the median mafosfamide concentration to 245 ( ) g/ml (n = 8). Even in this small series, this effect could be demonstrated on a significant level (P = ). We tried to further enhance the purging efficacy of mafosfamide on PBPC autografts by preincubation of the cells with amifostine, which is known to protect normal hematopoietic progenitor cells in vitro while sensitizing leukemic progenitors to the toxicity of mafosfamide. 8 We were able to show a significant advantage in terms of CFU- GM survival for combined purging with mafosfamide and amifostine vs mafosfamide alone. This means that there is a significant protective effect on amifostine on normal hematopoiesis against mafosfamide at the CFU-GM level in a small series of samples. Capizzi et al 21 demonstrated a shorter duration of neutropenia and an increased nadir of peripheral neutrophils for amifostine preteatment in patients given 1.5 g/m 2 cyclophosphamide and in a second phase after 1 month given amifostine (740 mg/m 2 ) before the same dose of cyclophosphamide. As a result of more progenitor cells remaining after amifostine pretreatment, Shpall et al 22 observed a shortening of engraftment periods resulting in less platelet transfusions and shorter hospitalization in patients undergoing 4-HC-purged ABMT in breast cancer patients in comparison to patients undergoing ABMT purged with 4-HC alone. We have up to now transplanted two patients with mafosfamide-purged autografts from peripheral blood progenitor cells (Table 2). The duration of neutropenia below /l and thrombocytopenia below /l are in the observed range for unpurged peripheral blood stem cell autografts. Sanz et al 4 reported median times of 13 days for neutrophil recovery and 19 days for platelet recovery to more than /l in a series of 24 CR1 AML patients undergoing unpurged autologous PBPCT. Gondo et al 23 reported a median of 15 days for neutrophil recovery to more than /l in a series of 60 patients undergoing unpurged autologous PBPCT. The median for platelet recovery to the last platelet transfusion was 24 days (range 0 180). Our data do not allow general conclusions because of the small number of patients, nevertheless they show that short engraftment periods can be achieved after autografting with PB-derived PBPC purged with mafosfamide doses of 100 g/ml. However, since the applied mafos-

5 famide dose was below the ID 95 of both patients, further data are needed to evaluate this topic. The early relapse of patient No. 2 corresponds well with the high risk karyotype of the leukemic cells and to reports of mafosfamide-purged ABMT in patients with secondary AML. 24 Patient No. 1 is in complete remission, 760 days after PBPCT. Comparing these data with those from the literature, we conclude that by using peripheral blood progenitor cells as autografts it is possible to apply considerably higher doses of mafosfamide to the graft without impairing its repopulating capacity at the CFU-GM level. This enables us to more aggressively purge PBPC in order to further reduce contamination by residual tumor cells without increasing the transplant-related mortality by prolonging the regeneration of peripheral blood counts. As Douay et al 25 have demonstrated in vitro on bone marrow cells of normal donors and patients with hematological malignancies, the resistance to mafosfamide of early progenitor cells such as LTC-IC is even higher than that of CFU-GM. In two cases in this study no toxicity to these cells seriously impairing repopulating capacity of the autografts could be observed. This is in line with the data of Motta et al 6 who demonstrated recovery of early hematopoietic progenitor cells after incubation with up to 175 g/ml mafosfamide in long-term liquid cultures. These doses could effectively eliminate K562 cells added to the evaluated samples. They observed no difference between BM-derived and PB-derived progenitors. Nevertheless whether or not more aggessive purging of PB-derived autografts with mafosfamide will result in depletion of normal progenitor cells leading to prolonged regeneration as observed in purging of BM-derived autografts in AML needs further evaluation in a larger number of patients. References 1 Brenner MK, Rill DR, Moen RC et al. Gene-marking to trace origin of relapse after autologous bone marrow transplantation. Lancet 1993; 341: Gorin NC, Aegerter P, Auvert B et al. Autologous bone marrow transplantation for acute myelocytic leukemia in first remission: a European survey of the role of marrow purging. Blood 1990; 75: Gorin NC, Labopin M, Meloni G et al. Autologous bone marrow transplantation for acute myeloblastic leukemia in Europe: further evidence of the role of marrow purging by mafosfamide. European Co-operative Group for Bone Marrow Transplantation (EBMT). Leukemia 1991; 5: Sanz MA, de la Rubia J, Sanz GF et al. Busulfan plus cyclophosphamide followed by autologous blood stem cell transplantation for patients with acute myeloblastic leukemia in first complete remission: a report from a single institution (see comments). J Clin Oncol 1993; 11: Laporte JP, Gorin NC, Feuchtenbaum et al. Relapse after autografting with peripheral blood stem cells. lancet 1987; 2: 1393 (letter). 6 Motta MR, Mangianti S, Rizzi S et al. Pharmacological purging of minimal residual disease from peripheral blood stem cell collections of acute myeloblastic leukemia patients: preclinical studies. Exp Hematol 1997; 25: Douay L, Giarratana MC, Bouchet S, Conlon J, Capizzi RL, Gorin NC. Amifostine improves the antileukemic therapeutic index of mafosfamide: implications for bone marrow purging. Blood 1995; 86: Martin H, Bruecher J, Claude R, Hoelzer D. Cumulative chemotherapy increases mafosfamide toxicity for normal progenitor cells in AML patients: rationale for cryopreserving adapted-dose purged marrow early in first complete remission. Bone Marrow Transplant 1993; 12: Seipelt G, Hofmann WK, Martin H et al. Comparison of toxicity and outcome in patients with acute meyloid leukemia treated with high-dose cytosine arabinoside consolidation after induction with a regimen containing idarubicin or daunorubicin. Ann Hematol 1988; 76: Ottmann OG, Stella CC, Eder M et al. Regulation of early hematopoiesis in serum-deprived cultures of mafosfamidetreated and untreated CD34 enriched bone marrow cells. Exp Hematol 1991; 19: Martin H, Atta J, Zumpe P et al. Purging of peripheral blood stem cells yields BCR-ABL-negative autografts in patients with BCR-ABL-positive acute lymphoblastic leukemia. Exp Hematol 1995; 23: Giarratana MC, Gorin NC, Douay L. Plasma interacts with mafosfamide toxicity to normal haematopoietic progenitor cells: impact on in vitro marrow purging. Nouv Rev Franc Hematol 1995; 37: Herve P, Cahn JY. Autologous bone marrow transplantation for acute leukemia using transplant chemopurified with metabolite of oxazaphosporines (Asta Z 7557, INN mafosfamide). First clinical results. Invest New Drugs 1984; 2: Mehta J, Powles R, Singhal S, Treleaven J. Peripheral blood stem cell transplantation may result in increased relapse of acute myeloid leukaemia due to reinfusion of a higher number of malignant cells. Bone Marrow Transplant 1995; 15: Korbling M, Fliedner TM, Holle R et al. Autologous blood stem cell (ABSCT) vs purged bone marrow transplantation (pabmt) in standard risk AML: influence of source and cell composition of the autograft on hemopoietic reconstitution and disease-free survival. Bone Marrow Transplant 1991; 7: Dusenbery KE, Steinbuch M, McGlave PB et al. Autologous bone marrow transplantation in acute myeloid leukemia: the University of Minnesota experience. Int J Radiat Oncol Biol Phys 1996; 36: Laporte JP, Douay L, Lopez M et al. One hundred and twentyfive adult patients with primary acute leukemia autografted with marrow purged by mafosfamide: a 10-year single institution experience. Blood 1994; 84: Straetmans N, Ma DD, Herman P et al. Long-term culture of autologous transplanted bone marrow for acute myeloid leukaemia: evidence for an in vitro haemopoietic defect and lack of correlation with the speed of engraftment. Bone Marrow Transplant 1995; 15: Miggiano F, Gherlinzoni G, Rosti G et al. Autologous bone marrow transplantation in late first complete remission improves outcome in acute myelogenous leukemia. Leukemia 1996; 10: Demirer T, Gooley T, Buckner CD et al. Influence of total nucleated cell dose from marrow harvest on outcome in patients with acute myelogenous leukemia undergoing autologous transplantation. Bone Marrow Transplant 1995; 15: Capizzi RL, Scheffler BJ, Schein PS. Amifostine-mediated protection of normal bone marrow from cytotoxic chemotherapy (review). Cancer 1993; 72 (Suppl.): Shpall EJ, Stemmer SM, Hami L et al. Amifostine (WR-2721) 835

6 836 shortens the engraftment period of 4-hydroperoxycyclophosphamide-purged bone marrow in breast cancer patients receiving high-dose chemotherapy with autologous bone marrow support. Blood 1994; 83: Gondo H, Harada M, Miyamoto T et al. Autologous peripheral blood stem cell transplantation for acute myelogenous leukemia. Bone Marrow Transplant 1997; 20: Laporte JP, Isnard F, Lesage S et al. Autologous bone marrow transplantation with marrow purged by mafosfamide in seven patients with myelodysplastic syndromes in transformation (AML-MDS): a pilot study. Leukemia 1993; 7: Douay L, Giarratana MC, Labopin M et al. Characterization of late and early hematopoietic progenitor/stem cell sensitivity to mafosfamide. Bone Marrow Transplant 1995; 15: