Peripheral blood stem cell transplantation as an alternative to ... - Nature

Bone Marrow Transplantation, (1999) 23, 1279–1282  1999 Stockton Press All rights reserved 0268–3369/99 $12.00 http://www.stockton-press.co.uk/bmt

Peripheral blood stem cell transplantation as an alternative to autologous marrow transplantation in the treatment of acute myeloid leukemia? E Vellenga3, WLJ van Putten2, MA Boogaerts9, SMGJ Daenen3, GEG Verhoef9, A ¨Hagenbeek6, AR Jonkhoff4, PC Huijgens4, LF Verdonck6, J van der Lelie¨ 5, HC Schouten8, J Gmur10, P Wijermans7, A Gratwohl11, U Hess12, MF Fey13 and B Lowenberg1 Dutch-Belgian Hemato-Oncology Cooperative Group (HOVON) and Swiss Group for Clinical Cancer Research (SAKK): 1Daniel den Hoed Cancer Center Rotterdam, University Hospital Rotterdam; 2HOVON Data Center, Rotterdam; 3University Hospital Groningen; 4 University Hospital, Vrije Universiteit, Amsterdam; 5University Hospital Amsterdam; 6University Hospital Utrecht; 7Leijenburg ¨ Hospital, The Hague; 8University Hospital Maastricht, The Netherlands; 9University Hospital Leuven, Belgium; 10Universitatsspital ¨ Zurich; 11Kantonsspital Basel; 12Kantonsspital St Gallen; and 13University Hospital Bern, Switzerland

Summary: The clinical use of autologous marrow transplantation in acute myeloid leukemia (AML) has been hampered by the inability to collect adequate numbers of cells after remission induction chemotherapy and the notably delayed hematopoietic regeneration following autograft reinfusion. Here we present a study in which the feasibility of mobilizing stem cells was investigated in newly diagnosed AML. Among 96 AML patients, 76 patients (79%) entered complete remission. Mobilization was undertaken with low dose and high dose schedules of G-CSF in 63 patients, and 54 patients (87%) were leukapheresed. A median of 2.0 × 106 CD34+ cells/kg (range 0.1–72.0) was obtained in a median of three leukaphereses following a low dose G-CSF schedule (150 ␮g/m2) during an average of 20 days. Higher dose regimens of G-CSF (450 ␮g/m2 and 600 ␮g/m2) given during an average of 11 days resulted in 28 patients in a yield of 3.6 × 106 CD34+ cells/kg (range 0–60.3) also obtained following three leukaphereses. The low dose and high dose schedules of G-CSF permitted the collection of 2 × 106 CD34-positive cells in 46% and 79% of cases respectively (P = 0.01). Twenty-eight patients were transplanted with a peripheral blood stem cell (PBSC) graft and hemopoietic repopulation was compared with the results of a previous study with autologous bone marrow. Recovery of granulocytes (⬎0.5 × 109/l, 17 vs 37 days) and platelets (⬎20 × 109/l; 26 vs 96 days) was significantly faster after peripheral stem cell transplantation compared to autologous bone marrow transplantation. These results demonstrate the feasibility of PBSCT in the majority of cases with AML and the potential advantage of this approach with respect to hemopoietic recovery. Keywords: AML; G-CSF; peripheral stem cell mobilization and transplantation

High dose chemotherapy followed by autologous bone marrow transplantation (ABMT) is commonly applied as a post remission treatment modality in patients with acute myeloid leukemia (AML) as an alternative of allogeneic marrow transplantation. The results are comparable to those seen after HLA-matched allotransplantation.1 One of the major limitations of the applicability of ABMT is that a considerable proportion of patients who have entered complete remission have no access to transplantation in clinical practice. These candidates have to be withdrawn from ABMT prematurely for several reasons, eg the collection of an insufficient autograft or the hazards of a prolonged period of cytopenia, a consequence of the slow hematological regeneration typical of ABMT in AML.1–5 The role of autologous stem cell transplantation in AML would potentially be greatly advanced if the repopulation abilities of the grafts were more potent. One possibility for overcoming these limitations is the use of peripheral blood autografts, which might be easier to collect and which might also have a better potential for hematopoietic repopulation. In fact, in patients with non-myeloid hematological malignancies, peripheral blood stem cell transplantation (PBSCT) has been shown to permit early hematopoietic reconstitution.6 However, as yet, these observations have mainly been based on experiences in patients with solid tumors and lymphoid malignancies who receive a significantly less aggressive treatment than do patients with AML. In addition, residual normal hematopoietic stem cell function in patients with AML may not be intact.7 We set out to evaluate the feasibility of collecting peripheral blood stem cells in newly diagnosed patients with AML. These patients were submitted to two cycles of remission induction therapy after which an attempt was made to collect peripheral blood stem cells with varying mobilization regimens of G-CSF. Materials and methods

Correspondence: E Vellenga, University Hospital Groningen, Department of Hematology, Hanzeplein 1, 9713 GZ Groningen, The Netherlands Received 15 October 1998; accepted 26 January 1999

This was a combined study of the Dutch–Belgian HematoOncology Cooperative Group HOVON and the Swiss Cancer Group SAKK in which the feasibility of collecting per-

PBSCT in the treatment of AML E Vellenga et al

1280

ipheral stem cells was investigated in a pilot study. Ninetysix patients (49% males) with previously untreated AML were prospectively entered at age 17–65 years (mean 45 years). AML was diagnosed and classified according to the French–American–British (FAB) classification.8 Remission induction treatment consisted of two cycles of chemotherapy. The first cycle included idarubicin 12 mg/m2 i.v. on days 1, 2 and 3, cytarabine 200 mg/m2 by continuous ˆ infusion (c.i.) on days 1–7 and G-CSF (lenograstim; Rhone 2 Poulenc Rorer; Antony Cedex, France) at 150 ␮g/m subcutaneously (s.c.) from day 0 (1 day before the start of chemotherapy) to day 7. The second cycle started within 6 weeks after the first course and included cytarabine 1 g/m2 i.v. every 12 h on days 1–6, amsacrine 120 mg/m2 i.v. on days 4, 5 and 6, and G-CSF again during the cycle on days 0–6. Following the two induction cycles, patients were to receive a third cycle of chemotherapy, mostly within 6 weeks after the second chemotherapy course, including mitoxantrone 10 mg/m2 i.v. days 1–5 and etoposide 100 mg/m2 i.v. days 1–5, or an autograft or an HLAmatched allograft depending on the risk estimates according to protocol and the availability of an HLA-matched donor. Patients planned for autografting were scheduled for peripheral blood stem cell mobilization. Peripheral blood stem cell mobilization To study the feasibility of G-CSF to mobilize peripheral stem cells in AML, a first series of patients (n = 31) were mobilized with a low dose regimen. However, in view of the high failure rate and long durations of mobilization this scheme was empirically modified in three successive steps in an effort to improve the outcome of collecting peripheral stem cells. At the end of this study four different schedules of G-CSF administration were used. (1) low dose G-CSF: G-CSF was started immediately after the second chemotherapy course at a dose of 150 ␮g/m2 s.c. and continued until completion of pheresis (n = 31), but no more than 40 days; (2) low dose + boost G-CSF: G-CSF was given exactly as in (1), but when granulocytes rose to 0.2 × 109/l or more, the dose was escalated to 600 ␮g/m2 s.c. and continued until completion of pheresis (n = 7); (3) 450 G-CSF: G-CSF was applied at a dose of 450 ␮g/m2 s.c. beginning only at the point of granulocytes recovery (0.2 × 109/l) or the appearance of detectable numbers of CD34+ cells in the blood (n = 13); (4) 600 G-CSF: as schedule 3, but G-CSF was applied at a dose of 600 ␮g/m2 (n = 12). Leukapheresis was started when leukocytes had risen to 2 × 109/l and it was continued on subsequent days with the objective of collecting a minimum of 1 × 108/kg nucleated cells or 2 × 106/kg CD34+ cells. The number of CD34+ cells was assessed by flow cytometry and the number of colony-forming unit granulocyte–macrophage (CFU-GM) was measured in in vitro colony assays. Leukapheresis products were cryopreserved for later use. G-CSF mobilization was discontinued prematurely in case of evidence of progressive leukemia or persisting AML on marrow evaluation. In the analysis of stem cell collection, the results of the low dose mobilization schedule 1 were compared to the combined results of schedules 2, 3 and 4. Hematological recovery following re-infusion of PBSC was evaluated in compari-

son to the ABMT cases from the HOVON-4/4A study, the previous collaborative study of the HOVON and SAKK groups in adults with AML.2 Peripheral blood stem cell transplantation Peripheral blood stem cell transplantation was given to a limited number of patients following conditioning with busulfan (16 mg/kg) and cyclophosphamide (120 mg/kg). The results were compared with a group of patients transplanted with autologous bone marrow cells. Statistical analysis Wilcoxon’s rank sum test was applied to test for differences in harvest between patients mobilized with low and high dose G-CSF, and between PBSC harvest and bone marrow harvest. Median time to recovery after transplantation was calculated by the actuarial method of Kaplan and Meier. Patients who did not recover were censored at date of last contact or death. The log rank test was applied to test for differences in recovery between patients transplanted with autologous bone marrow and patients with peripheral blood stem cells. Results The feasibility of collecting peripheral blood stem cells was studied in 96 patients with AML The characteristics of the population studied are given in Table 1 with regard to age, FAB classification and karyotyping. Of 96 evaluable cases, 76 patients (79%) attained a CR. Mobilization with G-CSF was not undertaken in 33 patients, in most cases because of early treatment failure (toxic death or progressive leukemia and the inability to continue on protocol (n = 24)). Refusals (n = 5), immediate allogeneic or ABMT (n = 4) Table 1

Characteristics of the patient population studied

Patient characteristics No. of patients Male/female Age (years) median (range

96 47/49 44.5 (17–65)

FAB classification M0 M1 M2 M3 M4 M5 M6 unclassified

3 8 38 7 16 19 2 3

Karyotype normal inv16 t(8;21) t(15;17) other abnormalities no data or failure

41 3 5 7 25 15


were reasons for withdrawal from mobilization in the other patients. Thus in 63 patients G-CSF mobilization was initiated which resulted in leukapheresis in 54 patients (87% of cases). Patients on schedules 1 to 4 received GCSF for mobilization for a median of 20, 17, 8 and 8 days, respectively. The failure rates for accomplishing the phereses were distributed as follows: 5/31 for the low dose GCSF regimen (1) and 4/32 for the three higher dose regimens (2, 3 and 4). Inability of pheresis was due to the lack of a rise of circulating CD34+ cells (n = 3) or the decision to perform an allogeneic BMT (n = 3) or other reasons. The median number of phereses conducted was three per patient for low (1) and high (2, 3, 4) G-CSF schedules. G-CSF treatment was started after a median of 22 days (range 11– 39) in groups 3 and 4. The cell yields were significantly higher for the high dose G-CSF regimens. In patients on G-CSF schedule 1 (150 ␮g/m2), the median yields were 2.0 × 106/kg CD34+ cells. In patients on G-CSF schedules 2, 3 and 4, the harvests contained a median 3.6 × 106/kg CD34+ cells (Table 2). The number of CD34+ cells per individual group were: 2, 3.0 (0.1–5.5), median (range); n = 7; 3 5.0 (0.0–60.3), n = 11; 4, 3.2 (0.0–43.6), n = 12. In fact in only 46% of the patients on the low dose schedule 1, the required minimum number of 2 × 106/kg CD34+ cells could be obtained, whereas the target number of 2 × 106 CD34+ cells or more was successfully obtained in 79% of cases on schedules 2, 3 and 4 (P = 0.01). Recovery following PBSCT in comparison to ABMT Twenty-eight patients received the PBSC autograft following busulfan-cyclophosphamide conditioning. The median number of days to recover neutrophils (⭓0.5 × 109/l) was 17 (range 9–160) and platelets (towards 20 × 109/l) was 26 (range 4–428). There was a distinct correlation between the number of CD34+ cells/kg in the graft and the time of recovery. When the number of CD34+ cells was subdivided Table 2

Peripheral blood stem cell mobilization in 63 AML patients

Number of patients G-CSF dose (␮g/m2/day) days applied (median) range Number of leukaphereses Total harvest CD34 × 106/kg Mononuclear cells × 108/kg CFU-GM × 104/kg

1

2, 3, 4

31

32

146 20 (1–48) 3 (1–5)

485 8 (2–47) 3 (1–7)

(n = 26) 2.0 (0.1–72.0) 5.9 (0.6–194) 22 (0–841)

(n = 28) 3.6* (0–60.3) 6.5 (1.3–38) 64** (0–711)

1–4 reflects the type of mobilization after the second chemotherapy. Course 1: 150 ␮g/m2 G-CSF; 2, similar to 1 but at a granulocyte count of 0.2 × 109/l G-CSF was given at a dose of 600 ␮g/m2; 3, G-CSF (450 ␮g/m2) was given when the granulocytes recovered to 0.2 × 109/l; 4, similar to 3 except the dose of G-CSF (600 ␮g/m2). *P = 0.21; **P = 0.05

in three classes (1.1–2.9 CD34+/kg, n = 8; 3–12.1 CD34+/kg, n = 9 and 12.1–65.9 CD34+/kg, n = 9) a significantly faster recovery was observed in the third group compared to the first group with regard to granulocyte ⬎0.5 × 109/l (median 14 vs 19 days respectively, P = 0.04, test for trend) and platelets ⬎20 × 109/l (median 11 vs 55 days respectively, P = 0.001, test for trend). This association is not apparent with regard to the relation between numbers of CFU-GM or mononuclear cells in the graft and the time to recovery. The results were also compared to ABMT results of the HOVON-SAKK HOVON4/4A study that had almost employed the same chemotherapy except for the use of daunomycin (45 mg/m2) instead of idarubicin in cycle 1 and a third chemotherapy course before ABMT.9 In 41 ABMT cases, a significantly lower number of mononuclear cells and CFU-GM were infused compared to PBSCT (Table 3a). The median numbers of days to neutrophil and platelet recovery were significantly prolonged in the ABMT group compared to the PBSCT group, ie 37 days and 96 days, respectively compared to the PBSCT group (Table 3). Discussion The value of autologous stem cell transplantation in the treatment of patients with AML in first CR is still the subject of intensive investigation.1 First of all, various approaches of dose-intensified treatment based either on allogeneic stem cell transplantation, autologous marrow transplantation or seqential cycles of high dose chemotherapy without stem cell support have produced roughly similar overall survival results.4,5,10–12 Second, based on Table 3

Harvest characteristics and hematological recovery PBSCT (n = 28)

ABMT (n = 41)

P value

(a) Characteristics of the peripheral blood stem cell or bone marrow harvest studied 8.1 1.7 ⬍0.001 Mononuclear cells × 108/kg (median, range) (1.4–194) (0.3–4.0) 74 6.4 ⬍0.001 CFU-GM × 104/kg (median, range) (6–841) (0.5–98) 5.4 ND CD34+ × 106/kg (median, range) (1.1–66) (b) Hematological recovery after PBSCT and ABMT in patients with AML in first CR Recovery to ANC 0.5 × 109/l not recovered, number of cases 1 7 ⬍0.001 recovered, median number of days 17 37 Recovery to ANC 1 × 109/l not recovered, number of cases 1 9 ⬍0.001 recovered, median number of days 19 56 Recovery to PLTs 20 × 109/l not recovered number of cases 1 18 ⬍0.001 recovered median number of days 20 96 The results of hematological recovery of 27 patients in first CR transplanted with PBSCT (this study) were compared to the results of hematological recovery after ABMT in 41 patients in the previous HOVONSAKK AML study (HOVON4-trial). ANC = absolute neutrophil count; PLT = platelets.

1281


1282

retrospective analysis there is a growing awareness that good risk patients, ie individuals with a relatively low probability of relapse following CR, may not need aggressive post-remission therapy in first CR since the risks of this procedure may not outweigh the benefits. Third, at present, allogeneic BMT and ABMT probably offer the best antileukemic treatment as evidenced by significantly reduced relapse rates, but they do not necessarily provide better survival rates because of a higher mortality.4,5 The toxic mortality associated with ABMT is especially linked to the slow hematopoietic regeneration that is characteristic of this type of transplantation in patients with AML. Thus, one may assume that the ultimate outcome of ABMT would be improved if hematological recovery could be hastened and, as a result, morbidity and mortality due to infections and hemorrhagic complications reduced. The use of PBSCT, instead of AMBT, might be attractive in that it might permit faster hematopoietic recovery.6 In any case, in patients with malignancies other than AML, PBSCT has provided more rapid hematopoietic reconstitution in comparison to autologous marrow grafts. In addition, PBSCT, as an alternative source of stem cells, might also have the advantage of a more successful harvest of an appropriate size graft. Frequently it has appeared impossible to aspirate sufficient numbers of marrow cells to allow for transplantation which limits the number of patients scheduled for ABMT.2 The results of the feasibility study presented here support the notion that PBSC may be a better alternative to autologous bone marrow. It is evident from these results that in the majority of cases an adequate PBSC graft can be obtained following aggressive chemotherapy and appropriate G-CSF mobilization. This objective is achieved in most patients with no more than three leukaphereses. It is also clear from our analysis that the G-CSF mobilization schedule has a significant impact on the yield, as suggested by Demirer et al13 in a small pilot study. The three higher dose G-CSF schedules produced considerably higher yields. As a matter of fact, the regimens that employed G-CSF at 450 ␮g/m2 or 600 ␮g/m2 from the earliest signs of recovery (at the point of granulocyte regeneration of 0.2 × 109/l or the appearance of circulating CD34+ cells) permitted a successful harvest of more than 2.0 × 106 CD34+ cells in 79% of cases and a median total yield of 3.6 × 106/kg CD34+ cells per patient. These numbers also appear to exceed the numbers presented in a recent study by Jowitt et al14 where a third of the patients failed to yield a PBSC harvest above the minimum threshold of 5 × 104 CFU-GM/kg. The results of this study also show a rapid neutrophil and platelet recovery following reinfusion of PBSC after marrow ablative therapy that is unusual for autografting patients with AML. Within approximately 2 weeks, neutrophil and platelets begin to recover. The faster recovery after PBSCT might be due to the infusion of a larger number of CD34+ stem cells. This might have the disadvantage that a larger number of residual malignant stem cells are also infused in the patient resulting in a higher risk of leukemic relapse. However, the data of this pilot study do not support this contention so far since the overall survival was comparable to ABMT. A greater number of patients in prospective randomized studies and with sufficient length of follow-up will be needed to settle this issue definitively. Taken

together, the data in this study indicate a number of potential advantages of PBSCT over ABMT that may lead to an easier applicability in clinical practice, a greater availability to patients and perhaps better survival outcome in the end. Hence, these data warrant future prospective studies to critically evaluate the role of PBSCT in the treatment of patients with AML and to establish whether the promise of PBSCT autografts translates into real clinical benefits. References ¨ 1 Lowenberg B. Post-remission treatment of acute myelogenous leukemia. New Engl J Med 1995; 332: 260–262. ¨ 2 Lowenberg B, Verdonck LJ, Dekker AW et al. Autologous bone marrow transplantation in acute myeloid leukemia in first remission: results of a Dutch prospective study. J Clin Oncol 1990; 8: 287–294. 3 Cassileth PA, Andersen J, Lazarus HM et al. Autologous bone marrow transplant in acute leukemia in first remission. J Clin Oncol 1993; 11: 314–319. 4 Zittoun R, Mandelli F, Willemze R et al. Autologous or allogeneic bone marrow transplantation compared with intensive chemotherapy in acute myelogenous leukemia. New Engl J Med 1995; 332: 217–223. 5 Burnett AK, Goldstone AH, Stevens RMF et al. Randomised comparison of addition of autologous bone-marrow transplantation to intensive chemotherapy for acute myeloid leukaemia in first remission: results of MRC AML10 trial. Lancet 1998; 351: 700–707. 6 Schmitz N, Linch DC, Dreger P et al. Randomised trial of filgrastim-mobilised peripheral blood progenitor cell transplantation versus autologous bone-marrow transplantation in lymphoma patients. Lancet 1996; 347: 353–357. 7 Vellenga E, Sizoo W, Hagenbeek A et al. Different repopulation kinetics of erythroid (BFU-E), myeloid (CFU-GM) and T lymphocyte (TL-CFU) progenitor cells after autologous and allogeneic bone marrow transplantation. Br J Haematol 1997; 65: 137–142. 8 French–American–British (FAB) Cooperative Group. Proposed revised criteria for the classification of acute myeloid leukemia. Ann Intern Med 1985; 103: 620–630. ¨ 9 Lowenberg B, Boogaerts MA, Daenen SM et al. Value of different modalities of granulocyte–macrophage colony-stimulating factor applied during or after induction therapy of acute myeloid leukemia. J Clin Oncol 1997; 15: 3496–3506. 10 Mayer RJ, Davis RB, Schiffer CA et al. Intensive postremission chemotherapy in adults with acute myeloid leukemia. New Engl J Med 1994; 331: 896–903. 11 Bishop JF, Matthews JP, Young GA et al. A randomized study of high-dose cytarabine in induction in acute myeloid leukemia. Blood 1996; 87: 1710–1717. 12 Appelbaum FR, Fisher LD, Thomas ED et al. Chemotherapy versus marrow transplantation for adults with acute nonlymphocytic leukemia: a five-year follow-up. Blood 1988; 72: 178–184. 13 Demirer T, Buckner CD, Appelbaum FR et al. Rapid engraftment after autologous transplantation utilizing marrow and recombinant granulocyte colony-stimulating factor-mobilized peripheral blood stem cells in patients with acute myelogenous leukemia. Bone Marrow Transplant 1995; 15: 915– 922. 14 Jowitt SN, Chang J, Morgenstern GR et al. Factors which affect the CFU-GM content of the peripheral blood haematopoietic progenior cell harvest in patients with acute myeloid leukaemia. Br J Haematol 1998; 100: 688–694.