Comparison between two fludarabine-based reduced-intensity ...

Leukemia (2007) 21, 2109–2116 & 2007 Nature Publishing Group All rights reserved 0887-6924/07 $30.00 www.nature.com/leu

ORIGINAL ARTICLE Comparison between two fludarabine-based reduced-intensity conditioning regimens before allogeneic hematopoietic stem-cell transplantation: fludarabine/melphalan is associated with higher incidence of acute graft-versus-host disease and non-relapse mortality and lower incidence of relapse than fludarabine/busulfan A Shimoni, I Hardan, N Shem-Tov, A Rand, C Herscovici, R Yerushalmi and A Nagler The Division of Hematology and Bone Marrow Transplantation, Chaim Sheba Medical Center, Tel-Hashomer, Israel

Reduced-intensity conditioning (RIC) regimens are increasingly used in allogeneic stem-cell transplantation (SCT). There are no data whether any of these regimens has advantage and in what setting. We retrospectively analyzed SCT outcomes in 151 patients given fludarabine-based RIC for various hematological malignancies; 72 conditioned with fludarabine and intravenousbusulfan (FB) and 79 with fludarabine and melphalan (FM). FM was more myelosuppressive. Grade III–IV organ toxicity occurred in 31 and 53% of FB and FM recipients (P ¼ 0.005) and acute graft-versus-host disease grade II–IV in 33 and 53%, respectively (P ¼ 0.01). Non- relapse mortality rate (NRM) was 16 and 40%, respectively (P ¼ 0.003). Active disease (HR 2.2, P ¼ 0.003) and prior autologous SCT (HR 1.7, P ¼ 0.04) predicted inferior overall survival (OS). Among patients transplanted in remission, OS was 72 and 36% after FB and FM, respectively (P ¼ 0.03) due to increased NRM with FM. Similarly, patients transplanted in active disease experienced higher NRM with FM, however lower relapse rates resulted in equivalent OS. In conclusion, there are marked differences in outcome between RIC regimens that are theoretically dose-equivalent. The FM regimen is more myelosuppressive and toxic but controls disease better. FB was associated with improved survival in patients transplanted in remission. These observations merit further study in prospective studies. Leukemia (2007) 21, 2109–2116; doi:10.1038/sj.leu.2404886; published online 9 August 2007 Keywords: stem-cell transplantation; reduced-intensity conditioning; fludarabine; busulfan; melphalan; non-relapse mortality

Introduction Allogeneic hematopoietic stem-cell transplantation (SCT) is an effective, potentially curative therapy in a variety of hematological malignancies.1 However, SCT may be associated with a high rate of treatment-related complications due to organ toxicity, related to the conditioning regimen, graft-versus-host disease (GVHD) and immunodeficiency state after SCT. Standard myeloablative conditioning may be associated with increased toxicity in elderly, medically infirm or heavily pretreated patients, such as those already failing a prior autologous SCT. Over the last decade, a variety of reducedintensity conditioning (RIC) has been designed attempting at reducing toxicity and exploiting the graft-versus-malignancy effect as the primary curative approach.2–4 These regimens allowed the extension of allogeneic SCT to a much wider patient population, previously excluded from SCT, but the relative Correspondence: Dr A Shimoni, Department of Bone Marrow Transplantation, Chaim Sheba Medical Center, Tel-Hashomer, Israel. E-mail: [email protected] Received 8 April 2007; revised 14 June 2007; accepted 2 July 2007; published online 9 August 2007

merits of myeloablative and RIC regimens in various SCT settings are still under investigation.5 Conditioning regimens can now be viewed as part of a spectrum of regimens differing in their immunosuppressive and myelosuppressive intensity.4 These regimens can be tailored to patient and disease characteristics at SCT. On one end of the spectrum are the classical myeloablative regimens such as the combination of high-dose cyclophosphamide with high-dose busulfan or total body irradiation (TBI). These regimens produce irreversible myeloablation and are associated with rapid achievement of complete chimerism and with significant extramedullary organ toxicity. On the opposite end are the truly nonmyeloablative regimens, such as the combination of fludarabine and low-dose TBI (F-TBI), designed in Seattle,6 or the combination of fludarabine with cyclophosphamide (FC).7 These regimens have been administered without stem-cell support or are not expected to produce significant myeloablation, such that autologous reconstitution is prompt if the graft is rejected. Organ toxicity is minimal and mixed chimerism is often observed in the first few months after SCT. Reduced intensity regimens are in the middle. They have not been used without stem-cell support, and autologous reconstitution is expected to be delayed, if at all. Complete chimerism is prompt and organ toxicity may be more significant, although to a lesser degree than following myeloablative regimens. The most commonly used RIC regimens worldwide are the fludarabine/melphalan regimen (FM) originally designed in Houston8 and the fludarabine/busulfan regimen (FB) designed in Jerusalem9 or modifications of the two, differing mostly by the use of additional serotherapy. These two fludarabine-based regimens are considered of equivalent myelo- and immunosuppressive intensity in the spectrum of conditioning regimens. There are no defined data as to whether one of these RIC regimens has any advantage over the other, and in what setting. In this study, we compare SCT outcomes following these two regimens. We show that although they are theoretically similar, they are associated with significant difference in the major SCT outcomes including engraftment kinetics, organ toxicity, GVHD, relapse and survival after SCT, and that the differences are also dependent on the status of disease at SCT.

Patients and Methods

Patient eligibility This study is a retrospective analysis of SCT outcomes of 151 consecutive patients with hematological malignancies treated with the most commonly used RIC regimens, FM and FB. Patients were eligible to be treated on these protocols if they

Fludarabine-based RIC in hematological malignancies A Shimoni et al

2110 were eligible for allogeneic SCT by standard criteria, but were at high-risk for nonrelapse mortality (NRM) with standard myeloablative conditioning due to advanced age (older than 55 years in SCT from HLA-matched sibling and older than 50 years for SCT from unrelated or mismatched SCT), extensive prior therapy (including three or more lines of prior chemotherapy or a prior autologous SCT), organ dysfunction, a recent fungal infection or poor performance status. Patients enrolled in the studies of RIC, but who were otherwise eligible for standard myeloablative conditioning, such as young patients with chronic myeloid leukemia (CML), were not included in this analysis. Patients achieving less than partial response to prior chemotherapy by standard hematological criteria were considered to have active disease at SCT. In patients with leukemia and myelodysplastic syndrome (MDS) this group included patients with more than 10% marrow blasts at SCT.5 Patients had to have an HLAcompatible related or unrelated donor willing to donate G-CSF mobilized peripheral blood stem cells (the preferred requested option) or bone marrow. All patients gave written informed consent, and the study was approved by the Institutional Review Board.

Conditioning regimens Patients were given one of two fludarabine-based RIC regimens; the FB regimen consisted of fludarabine (30 mg/m2 on days 6 to 2) combined with intravenous busulfan (3.2 mg/kg daily on days 4 to 3). The FM regimen consisted of fludarabine (30 mg/m2 on days 6 to 2) combined with melphalan (50– 70 mg/m2 on days 3 to 2). Patients with an unrelated or mismatched donor were also given antithymocyte globulin (ATG, Fresenius, Bad Homburg, Germany, 5 mg/kg given on days 3 to 1). Protocol assignment was not randomized and was affected to some degree by disease type, prior therapy, protocol priorities at the time of SCT and also by the attending physician discretion. Patients with myeloid malignancies were preferentially assigned to the FB regimen, while patients with multiple myeloma were more often assigned to FM. However, both protocols were open for all hematological malignancies. Phenytoin was administered before and until 24 h after the completion of busulfan administration. Prophylaxis against GVHD consisted of cyclosporine A and a short course of methotrexate (15 mg/m2 on day 1 and 10 mg/m2 on days 3 and 6). GVHD prophylaxis was administered for 3 months and tapered afterwards in patients with no active GVHD. GVHD prophylaxis did not differ between the two regimens. As a role, cyclosporine dose was adjusted by plasma levels but not by creatinin. G-CSF was administered from day þ 7 until engraftment. A standard regimen of antibiotic prophylaxis was used to prevent bacterial, viral, fungal and Pneumocystis carini infections.

Evaluation of response Neutrophil and platelet engraftment were defined as the first of 3 days with absolute neutrophil count (ANC) 40.5 109/l and the first of 7 days with an untransfused platelet count 420 109/l, respectively. Toxicity after SCT was graded by the National Cancer Institute (NCI) common toxicity criteria (NCI, Bethesda, MD, USA). Acute and chronic GVHDs were graded and staged by standard criteria. Chimerism was tested using FISH with X and Y chromosome probes in sex-mismatched transplants and with PCR analysis of microsatellite markers in sex-matched transplants.10 Chimerism was tested in whole blood/marrow cells. Chimerism analysis within cell subsets was not performed. Leukemia

Mixed chimerism was defined as less than 99% donor chimerism. Response and relapse were determined by standard hematological criteria.

Study end points The primary end points of analysis were treatment-related toxicity and mortality and incidence and severity of GVHD. Since by study design, the study group included a heterogeneous group of patients, overall survival (OS), and relapse incidence were considered secondary end points. SCT data were collected prospectively during patient follow-up and recorded in an institutional database.

Statistical analysis Overall survival was calculated from the day of SCT until death or last follow-up. The probabilities of OS was estimated using the Kaplan–Meier method.11 Relapse and NRM rates were estimated using cumulative incidence analysis and considered as competing risks.12 In the analysis of the cumulative incidence of GVHD, relapse was considered a competing risk. In the analysis of different causes of NRM, GVHD and organ toxicity related deaths were considered competing risks. The effect of various patient and disease categorical variables on survival probabilities was studied with the log-rank test. A Cox proportional hazard model was used to determine the significance of multiple variables in determining these outcomes.

Results

Patient and donor characteristics The study included 151 consecutive patients with hematological malignancies given allogeneic SCT with two RIC regimens, FM and FB, in a single institution, over a 5-year period. During this period, 367 patients had their first allogeneic SCT in our institution; 205 with myeloablative regimens (56%), 151 with the RIC regimen studies (41%), and 11 (3%) with other RIC regimens (such as FC or F-TBI). Patient and donor characteristics are outlined in Table 1. All patients were considered not to be good candidates for standard myeloablative therapy, due to advanced age (n ¼ 83), extensive prior therapy (n ¼ 89), including 74 patients with a history of prior autologous SCT (33 of them within the previous 1 year), organ dysfunction or a recent fungal infection (n ¼ 13) and poor performance status (n ¼ 7); some patients had more than one criteria for the selection of RIC. The median age was 53 years (range, 16–70), and 60% of the patients had active disease at the time of SCT. Ninety donors were HLA-matched siblings, 59 were unrelated volunteers and 2 were 1-antigen mismatched relatives (analyzed with the unrelated transplants). One-hundred and forty-seven donors donated peripheral-blood stem cells and only four donors donated bone marrow. Seventy-two patients were given conditioning with fludarabine and intravenous busulfan (FB group), while 79 patients were conditioned with fludarabine and melphalan (FM group). The median age of FB group was older, and this group included a larger proportion of patients with acute myeloid leukemia (AML) and MDS, and more patients with unrelated donors. The FM group included a larger proportion of patients with multiple myeloma and patients with a prior autologous SCT (Table 1).

Engraftment and chimerism One-hundred and thirty-five patients achieved primary engraftment. Thirteen patients died before engraftment and three had


2111 Table 1

Patient characteristics

Median age (years, range) Gender Male/female Diagnosis AML MDS CML MM NHL HD CLL ALL PNH

Table 2

All (n ¼ 151)

FB (n ¼ 72)

FM (n ¼ 79)

P-valuea

53 (16–70)

56 (23–70)

51 (16–66)

0.002

84/67 41 7 7 40 31 9 8 6 2

(27%) (5%) (5%) (26%) (21%) (6%) (5%) (4%) (1%)

40/32 34 7 4 5 12 6 2

(47%) (10%) (6%) (7%) (17%) (8%) (3%) F 2 (3%)

44/35 7 (9%) F 3 (4%) 35 (40%) 19 (24%) 3 (4%) 6 (8%) 6 (8%) F

NS 0.001 0.001

Donor Sibling Mud/mm

90 (60%) 61 (40%)

35 (49%) 37 (51%)

55 (70%) 24 (30%)

0.009

Prior autoSCT Within 1 year

74 (49%) 33 (22%)*

25 (35%) 10 (14%)

49 (62%) 23 (29%)

0.008 0.02

Disease activityb Active Remission

90 (60%) 61 (40%)

41 (60%) 31 (40%)

49 (62%) 30 (38%)

NS

Abbreviations: ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; autoSCT, autologous SCT given before the current allogeneic SCT; CLL, chronic lymphocytic leukemia; CML, chronic myeloid leukemia; FB, fludarabine and intravenous busulfan; FM, fludarabine and melphalan; HD, Hodgkin’s disease; MDS, myelodysplastic syndrome; MM, multiple myeloma; mud/mm, matched unrelated or mismatched donor; NHL, non-Hodgkin’s lymphoma; PNH, paroxysmal nocturnal hemoglobinuria; SCT, stem cell transplantation. *Included patients with MM (n ¼ 15), NHL (n ¼ 9), HD (n ¼ 4), AML (n ¼ 4), CML (n ¼ 1). a The two regimens were compared to each other by t-test (for age) or w2-analysis and comparisons with statistical significance are presented. b Active disease defined as not achieving at least a partial response by standard hematological criteria.

primary graft failure, two after FB and one after FM (all with unrelated donors). The two FB recipients had autologous recovery. The median time to ANC 0.5 109/l was 13 days (range, 9–30 days). The median time to platelet 20 109/l was 14 days (range, 7–39 days). Time to engraftment was not different between FB and FM groups. After FB, the mean first day of neutropenia, in patients with normal neutrophil counts at the start of chemotherapy, was day þ 5 (range 2 to þ 13) and five patients never dropped ANC level below 0.5 109/l. The mean first day of neutropenia after FM was earlier than after FB, on day þ 2 (range, 2 to þ 6, P ¼ 0.03) and all patients became neutropenic. This resulted in a mean of 8 days of neutropenia (range, 3–16 in patients becoming neutropenic) and 11 days (range, 5–22) after FB and FM, respectively (P ¼ 0.01). Similarly 11 patients never dropped platelet count below 20 109/l after FB, while all FM recipients became thrombocytopenic during SCT. There was also significant difference in chimerism kinetics. At 1 month after SCT, 15 of 66 (23%) evaluable FB recipients had mixed chimerism; five of them below 90% donor. Among 70 evaluable FM recipients, only three patients (4%) were initially mixed chimera, and only one had less than 90% donor cells

Organ toxicities and GVHD All (n ¼ 151)

Organ toxicity Any gradeX3 toxicityb (excluding mucositis) Severe mucositis Hepatic toxicity VOD Cardiovascular Neurological Renal Pulmonary/DAH TMA

FB (n ¼ 72)

FM (n ¼ 79)

Pvaluea

64 (42%)

22 (31%)

42 (53%)

0.005

60 50 7 5 5 17 12 9

21 (29%) 20 (28%) 3 (4%) 0 1 (1%) 3 (4%) 2 (3%) 1 (1%)

39 30 4 5 4 14 10 8

0.01 NS 0.03 NS 0.009 0.03 0.02

(40%) (33%) (5%) (3%) (3%) (11%) (8%) (6%)

(49%) (38%) (5%) (6%) (5%) (18%) (13%) (10%)

Death due to organ toxicityc Cumulative incidenced

n ¼ 25 17 (12–24)

n¼7 10 (5–20)

n ¼ 18 23 (16–35)

0.04

Acute GVHD grade II–IVd Acute GVHD grade III–IVd Chronic GVHDd

43 (35–53) 29 (21–40) 62 (51–74)

33 (23–48) 17 (9–32) 61 (47–79)

53 (42–68) 42 (30–58) 62 (48–80)

0.01 0.003 NS

Death due to GVHDe Cumulative incidenced

n ¼ 16 11 (7–18)

n¼4 6 (2–15)

n ¼ 12 17 (10–28)

0.04

Abbreviations: See Table 1. DAH, diffuse alveolar hemorrhage; GVHD, graft-versus-host disease; TMA, thrombotic microangiopathy; VOD, veno-occlusive disease of the liver. a The two regimens were compared to each other by w2-analysis (categorical variables) or log-rank test (for incidence variables). b Includes deaths due to organ toxicity or infections not in the setting of active GVHD. c Death rates due to organ toxicity, GVHD, and relapse were estimated using cumulative incidence analysis and were considered competing risks. Rates are presented with 95% confidence intervals. d Includes deaths related to complications of acute or chronic GVHD. e Toxicity is graded by NCI common toxicity criteria. Only patients with X3 toxicity are presented. Some patients had toxicity in more than one organ system.

(P ¼ 0.003). There was no difference between the regimens in chimerism at 3 months after SCT. In all, of 18 patients with mixed chimerism, 7 converted spontaneously into complete chimerism, 7 relapsed, 2 died of nonrelapse causes, and only 2 remained with near complete chimerism.

Toxicity and GVHD National Cancer Institute Grade III–V organ toxicity (excluding mucositis) occurred in 64 patients (42%), 22 recipients of FB (31%) and 42 recipients of FM (53%, P ¼ 0.005). Mucositis, cardiovascular, pulmonary and renal toxicity, as well as transplant-related thrombotic microangiopathy were all more common in the FM group (Table 2). NCI grade XIII elevation of bilirubin and/or transaminases was a common finding, occurring in 33% of all patients with no significant difference between the two groups. In only seven patients (5%) could a clinical diagnosis of veno-occlusive disease be made by the Jones criteria,13 again, with no difference between the two groups. Grade II–IV and III–IV acute GVHD occurred in 52 and 31 patients, with cumulative incidence of 43 and 29%, respectively. There was significant difference in the rates of GVHD among recipients of the two regimens with significantly higher incidence of acute GVHD after FM (Table 2, Figure 1). The major difference was related to higher incidence of GVHD of the gastrointestinal tract among FM recipients (data not shown). Chronic GVHD occurred in 49 of 86 evaluable patients, Leukemia


2112 Table 3 after SCT

1.000

Category

acute GVHD III-IV

0.750

0.500

FM

FB

0.000 0.0

20.0

40.0

60.0

80.0

NRM (No.)

NRM (CI)

P-value

151

41a

28 (22–37)

48 103

16 25

36 (24–53) 25 (18–35)

NS

Gender Male Female

84 67

25 16

31 (22–43) 25 (16–38)

NS


90 61

27 14

31 (23–43) 24 (15–38)

NS

55

7

13 (7–27)

0.003

40 48

10 20

26 (15–44) 43 (31–60)

8

4

61 (32–100)

74 77

23 18

32 (23–44) 25 (17–38)

NS

Prior autoSCT within 1 year Yes 33 12

37 (24-58)

0.12

1.5 (0.8–2.9) P ¼ NS

0.01

1.8 (0.9–3.9) P ¼ 0.10

0.003

2.5 (1.2–4.9) P ¼ 0.01

All patients Age (years) o50 X50

0.250

100.0

days after transplantation

1.000

Disease AML/MDS/ CML MM Lymphoma/ CLL Othersc

0.750

Prior autoSCT Yes No

Figure 1 Cumulative incidence of acute graft-versus-host disease (GVHD) grade III–V after stem-cell transplantation (SCT); incidence is higher following fludarabine and melphalan (FM) conditioning than following fludarabine and intravenous busulfan (FB).

non-relapse mortality

Analysis of factors predicting for non-relapse mortality

0.500

FM

0.250

No

FB

No.

118

29

26 (19–35)

Status of disease at SCT 90 31 Actived

35 (26–46)

Remission

61

10

19 (11–33)

Conditioning FM

79

30

40 (30–53)

72

11

16 (9–27)

MVA HR (P-value)

1.9 (1.0–3.73) P ¼ 0.04b

P=0.003 0.000 0.0

10.0

20.0

30.0

40.0

50.0

60.0

months after transplantation

Figure 2 Cumulative incidence of non-relapse mortality (NRM) after stem-cell transplantation (SCT). The incidence was higher after fludarabine and melphalan (FM) conditioning (P ¼ 0.003). Relapse was considered a competing risk in this analysis.

26 extensive, 23 limited (cumulative incidence 62%) with no difference between the two groups in overall incidence or severity. In all, NRM occurred in 41 patients with a cumulative incidence of 28% (95 CI 22–37%) at 2 years after SCT. Twentyone patients died of organ toxicities, 16 patients died of complications related to acute GVHD, and four died of infections not in the context of active GVHD. There was a higher cumulative incidence of NRM among patients given FM than in patients given FB, 40 and 16%, respectively ((P ¼ 0.003, Figure 2). Excess NRM among FM recipients was related to both increased risk of organ toxicity and acute GVHD (Table 2). Patients with lymphoma/chronic lymphocytic leukemia (CLL) and patients with active disease at the time of SCT had a higher risk of NRM (Table 3). Prior high-dose chemotherapy with autologous SCT tended to associate with increased NRM, especially when given within 1 year of the current allogeneic SCT, however not reaching the statistical significance. The FM regimen included twice as many patients having a recent autologous SCT. These autografts were predominantly given for patients with myeloma (Table 1). However, these differences Leukemia

FB

Abbreviations: See Tables 1 and 2. HR, hazard ratio with 95% confidence intervals; MVA, multivariable analysis, factors with at least borderline statistical significant in the univariant analysis (Po0.2) were included in a Cox proportional hazard model; NRM, non-relapse mortality at 2-years (No., total number; CI, cumulative incidence, rates are given with 95% confidence intervals, relapse was considered as a competing risk). a 21 patients died of organ toxicities, 16 of complications related to GVHD and 4 of infectious complications not in the context of GVHD (see Table 2). b HR for patients with Lymphoma/CLL compared with all others. c Six patients with ALL and two patients with PNH, not included in the other disease categories. d Active disease as in Table 1.

could not explain the increased NRM with FM. When excluding the 33 patients with recent autologous SCT from the analysis, NRM rates remained similar, 40 and 13% after FM and FB, respectively (P ¼ 0.004). Similarly, the association with increased rates of acute GVHD grade III–IV remained similar, 42 and 16% after FM and FB, respectively (P ¼ 0.002). Excluding myeloma patients from analysis also did not impact results (data not shown). Interestingly, age category and an unrelated donor were also not found to be the risk factors for NRM. A multivariable analysis determined that the most significant risk factors for NRM were conditioning with FM, and a diagnosis of lymphoma/CLL with hazard ratios of 2.5 and 1.9, respectively


2113 (Table 3). NRM tended to be higher after FM in all major disease categories, however, not reaching statistical significance due to smaller number of patients in disease subgroups (data not shown).

Table 4

Survival

With a median follow-up of 26 months (range, 1–60), 66 patients are alive and 85 have died. Forty-one patients died of treatment-related complications and 43 of disease relapse. The actuarial 2-year OS is 34% (95 CI 25–43%). Table 4 outlines the univariable and multivariable analysis of factors influencing the outcome. The status of disease at SCT was the most significant predictor of outcome. Patients in remission had an OS of 54% (95 CI 38–70%), whereas patients with active disease had an OS of 24% (95 CI 15–34%, P ¼ 0.0003, Figure 3). Patients with lymphoma/CLL and patients with a prior autologous SCT within 1 year of allogeneic SCT had poorer outcome. The type of conditioning was not predictive of the outcome (Figure 3). Multivariable analysis confirmed the adverse effect of active disease (HR 2.2, P ¼ 0.003) and a recent prior autologous SCT (HR 1.7, P ¼ 0.04) as independent adverse factors for survival. However, when the analysis of outcome is limited to patients in remission at SCT the conditioning regimen used had a

0.750

FB

0.500

0.250 FM p=NS 0.000 0.0

10.0

20.0 30.0 40.0 50.0 months after transplantation

60.0

1.000

0.750 Survival

Outcome

1.000

0.500

remission

Analysis of factors predicting for overall survival after SCT

Category

No. Alive

OS

Pa-value

151

66

34 (25–43)

All patients

0.250

Active disease

MVA HR (P-value)

p=0.0003 0.000 0.0

Age (years) o50 X50

48 103

17 23 (8–37) 49 41 (30–52)

0.15

Gender Male Female

84 67

34 35 (23–46) 32 35 (20–49)

NS


90 61

41 37 (26–49) 25 32 (18–46)

NS

Disease AML/MDS/CML MM Lymphoma/CLL Others

55 40 48 8

27 19 16 4

(27–57) (18–55) (9–37) (0–78)

0.02b

Prior autoSCT Yes No

74 77

26 27 (15–39) 40 44 (33–56)

0.11

42 37 23 39

1.0 (0.6–1.7) P ¼ NS

1.2 (0.8–2.0) P ¼ NSb

Prior autoSCT within 1 year Yes 33 8 16 (1–30) No 118 58 41 (31–51)

0.01

1.7 (1.0–2.7) P ¼ 0.04

Status of disease at SCT Active 90 26 24 (15–34) Remission 61 40 54 (38–70)

0.0003

2.2 (1.3–3.7) P ¼ 0.003

0.12

1.3 (0.8–1.9) NS

Conditioning FM FB

79 72

30 28(17–40) 36 41 (27–55)

Abbreviations: See Table 1–3. MVA, multivariable analysis as in Table 3;OS, overall survival at 2 years (estimated by Kaplan–Meier product limit test and given with 95% confidence intervals). a Univariant analysis using log rank test. b HR for patients with Lymphoma/CLL compared with all others.

10.0

20.0

30.0

40.0

50.0

60.0

months after transplantation

Figure 3 Overall survival after stem-cell transplantation (SCT). Patients conditioned with fludarabine and melphalan (FM) and fludarabine and intravenous busulfan (FB) had similar survival (a). Patients with active disease at SCT had inferior outcome in comparison to patients given SCT in remission (b, P ¼ 0.0003).

significant impact. OS of patients in remission given FB was 72% (95 CI 53–91%) compared to 36% (95 CI 14–57%, P ¼ 0.03, Figure 4) in FM recipients. This difference was related to marked difference in the cumulative incidence of NRM, but no significant difference in the incidence of relapse mortality, (Table 5). There were marked differences in disease categories between the two groups; the FB regimen was given predominantly to patients with myeloid malignancies while the FM regimen included a larger proportion of patients with myeloma. However, these differences could not explain the observed difference in the outcome between FM and FB. The same results were observed when excluding patients with myeloma from analysis (data not shown). A multivariable analysis in this patient subset showed that conditioning with FM was the only significant factor (HR 4.9 (1.2–19.1), P ¼ 0.03) while age, disease and donor type and prior autografting were not significant. Among patients given SCT in active disease, recipients of FM had a higher incidence of NRM, but lower incidence of relapse, such that OS was equivalent to recipients of FB (data not shown).

Discussion Our data confirm that allogeneic SCT with RIC regimens is a feasible and effective therapy for a wide range of hematological Leukemia


non-relapse mortality

2114 1.000

Table 5 Characteristics and outcomes of patients in remission at the time of SCT

0.750

Category

FM 0.250 FB p=0.03

relapse mortality

10.0


FM (n ¼ 30)

FB (n ¼ 31)

Age (median) 56 (27–70) 52 (27–66) 59 (37–70) Gender 27/34 13/17 14/17 (male/female) Donor 33/28 18/12 15/16 Sibling/mud/mm

0.500

0.000 0.0

All (n ¼ 61)

60.0

Disease AML/MDS/CML MM Lymphoma/CLL ALL

1.000

Prior autoSCT All/ within 1 year

0.750

OS NRM Relapse

P-value 0.001 NS NS

30 18 9 4

5 16 5 4

25 2 4 0

o0.001 o0.001 NS 0.04

29/10

21/9

8/1

o0.001/0.005

54 (38–70) 36 (14–57) 72 (53–90) 19 (11–33) 32 (17–57) 7 (2–6) 28 (16–47) 33 (17–62) 22 (9–49)

0.03 0.03 NS

Abbreviations: See Table 1–4. OS, overall survival at 2 years (estimated by Kaplan–Meier product limit test and given with 95% confidence intervals); NRM and relapse at 2 years (estimated using cumulative incidence analysis and considered competing risks).

0.500 FM 0.250 FB p=NS 0.000 0.0

10.0


60.0

1.000

0.750 Survival

FB 0.500

FM

0.250

p=0.03 0.000 0.0

10.0


60.0

Figure 4 Stem-cell transplantation (SCT) outcomes of patients given SCT in remission. Patients conditioned with fludarabine and intravenous busulfan (FB) had a lower cumulative incidence of non-relapse mortality (NRM) (a, P ¼ 0.03), a similar cumulative incidence of relapse (b) and longer overall survival (c, P ¼ 0.03).

malignancies in patients not eligible for standard myeloablative conditioning due to extensive prior therapy, advanced age or comorbidities.2–4 In this study we compared retrospectively the SCT outcomes following the two most commonly used fludarabine-based RIC regimens worldwide, the FM and FB regimens. Although these are considered theoretically of equivalent myelo- and immunosuppressive intensity in the spectrum of conditioning regimens,4 we observed significant difference in the major SCT outcomes including engraftment kinetics, organ toxicity, GVHD, relapse and survival after SCT. Leukemia

The differences were also dependent on the status of disease at SCT. The FM regimen was associated with more intense myeloablation. All patients became aplastic and required transfusions, while some patients conditioned with FB did not become severely pancytopenic or had delayed onset of neutropenia. Two patients conditioned with FB had spontaneous autologous reconstitution after graft failure, while FM is usually associated with irreversible myelosuppression.14 Melphalan has broad stem-cell toxicity to both primitive and committed stem cells while busulfan may spare committed stem cells resulting in relatively early onset and a prolonged duration of neutropenia with FM.15 In addition FM recipients had rapid achievement of complete chimerism, while 23% of FB recipients initially had mixed chimerism. Thus, engraftment kinetics associated with the FM regimen is more similar to myeloablative conditioning. There was marked difference in NRM between the two regimens. The FB regimen has a relatively safe toxicity profile with reduced incidence of both organ toxicity and GVHD, while the FM regimen is more toxic. The incidence of organ toxicity was increased in all organ systems after FM and, in particular, the incidence of severe mucositis was significantly increased. Only hepatotoxicity was equivalent in the two regimens as busulfan most common toxicity is hepatic. The 2-year NRM rate associated with the FB regimen in the current study was 16%, similar to that associated with truly nonmyeloablative conditioning.6,7,16 NRM rate was 10–20% in most FB series.2,5,9,17–22 The use of intravenous busulfan in the current study may have further limited regimen toxicity.17 On the other hand, the NRM rate associated with the FM regimen in the current study was 40%, very similar to the original report by Giralt et al.,8 reaching 45%, 2 years after SCT, and was 20–40% in most other series.23–30 These rates are more similar to those associated with myeloablative conditioning in patients eligible for such therapy. The high NRM was related to increase in the rate of organ toxicity, but to a larger extent due to increase in the incidence of acute GVHD. Studies using FM with alemtuzumab had a lower NRM, similar to that seen with FB, due to marked reduction in


the incidence of acute GVHD.24,31,32 High-dose ATG may have similar effects.30 GVHD is the major limitation for successful RIC transplants. Couriel et al. reported that the incidence of acute and chronic GVHD is higher after myeloablative regimens.33 Interestingly the FM regimen was analysed with the myeloablative regimens in that study, as it had similar incidence. We have shown that acute GVHD is more common after myeloablative conditioning than after fludarabine-based conditioning.5 The Seattle group showed that acute GVHD after nonmyeloablative conditioning was not reduced but rather delayed in time after SCT.34 Acute GVHD is related in part to cytokine release from damaged gastrointestinal tissues which amplifies the allogeneic response.35,36 Limitation of mucosal injury with RIC may, therefore, reduce the incidence of acute GVHD.37 Severe mucositis was common after FM and directly associated with high incidence of severe GVHD, mostly of the gastrointestinal tract. The more common occurrence of mixed chimerism early after SCT with FB may also have limited the evolvement of acute GVHD.10 However, most patients with FB achieved complete chimerism rapidly, and we could not show statistically significant difference in the incidence of acute GVHD between the minority of patients with initial mixed chimerism and the majority achieving complete chimerism more rapidly. Ultimately, chronic GVHD incidence was similar between the groups. Similarly, there is no difference in chronic GVHD after myeloablative and RIC in some,34 although not all studies.5,33 Although acute GVHD is the major predicting factor for chronic GVHD, chronic GVHD is a different pathological entity and much less dependant on cytokine release during the initial conditioning. Importantly, analyzing all GVHD-related deaths still shows marked increase in GVHD-related mortality after FM, suggesting that GVHD is a major limitation of this regimen. The FM regimen was more myelosuppressive and toxic, but probably also more cytoreductive against a variety of hematological malignancies. Chemosensitivity was the single most important factor predicting outcome. The 2-year survival of patients having SCT in at least partial remission was 54% compared to 24% for patients given SCT in active disease. The difference was related to both increased incidence of relapse and NRM in patients with active disease. Patients given SCT in remission had similar relapse rates with both regimens. Thus, there was a survival advantage for SCT with the less toxic regimen. Estimated 2-year survival was 72% after FB and 36% after FM. However, when the underlying malignancy was not in remission at the time of SCT, there was advantage in disease control for FM recipients, but since toxicity was also markedly increased, OS was similar. The effect of dose intensity on SCT outcomes was studied in comparisons of different myeloablative regimens and in comparisons of myeloablative and RIC regimens. More intensive myeloablative conditioning is associated with a reduced risk for relapse after SCT, but not translating to improved survival due to increased transplantrelated toxicity.38 A few recent nonrandomized comparisons suggested that outcome after RIC is grossly equivalent to that after myeloablative SCT.18,39 These studies showed lower NRM, but higher relapse rate, and equivalent survival with RIC. We have shown in a group of patients with AML/MDS that outcome was similar after myeloablative and RIC regimens when disease was in remission, but there was significant advantage for myeloablative conditioning in patients with active leukemia at SCT.5 RIC requires induction of graft-versus-leukemia effect (GVL) as the curative therapeutic tool, and this may require several months after SCT to evolve. Patients with active disease may progress early after SCT outpacing the development of

2115 effective GVL. However, a more intensive cytoreduction as achieved with myeloablative conditioning may allow sufficient time for GVL. The current study has a few limitations. This is a heterogeneous group of patients with various hematological malignancies. Selection of conditioning regimen was not randomized and may have caused some biases. The FM group had a larger proportion of patients with myeloma, who are known to be more susceptible to NRM.29 This group also included more patients with a prior autologous SCT. On the other hand the FB group, despite having a lower observed GVHD rate, included older patients and more patients having unrelated donor SCT, which are known risk factors for GVHD. In particular, relapse and survival data should be interpreted with caution as differences between the groups may cause bias. Although observations were similar across the wide range of hematological malignancies studied, relapse potential in relation to disease activity and the regimen used may be different in different diagnoses. However, patient group size did not allow stratification by disease type. Notwithstanding these limitation a few conclusions can be made, and may be used to direct future studies. Different RIC regimens that are considered dose equivalent may result in substantial difference in SCT outcomes. The FM regimen at the doses used in this study is more myelosuppressive and cytoreductive. In a group of patients not eligible for standard myeloablative conditioning it is significantly inferior to FB in safety, and in particular, associated with significantly higher incidence of GVHD. Thus, inferior results may be achieved in patients in remission at SCT. When disease is active at SCT, there is advantage in a more intensive regimen for disease control, compensating for the increased toxicity. These conclusions may not apply to all disease categories, and disease-specific randomized studies are needed to better define the role of each regimen in different malignancies and SCT settings. RIC regimens are not all equivalent and need to be tailored to patient and disease parameters at the time of SCT.

References 1 Thomas ED, Storb R, Clift RA, Fefer A, Johnson FL, Neiman PE et al. Bone marrow transplantation. N Engl J Med 1975; 292: 832–843. 2 Barrett AJ, Savani BN. Stem cell transplantation with reducedintensity conditioning regimens: a review of ten years experience with new transplant concepts and new therapeutic agents. Leukemia 2006; 20: 1661–1672. 3 Shimoni A, Nagler A. Nonmyeloablative stem cell transplantation: lessons from the first decade of clinical experience. Curr Hematol Rep 2004; 3: 242–248. 4 Champlin R, Khouri I, Shimoni A, Gajewski J, Kornblau S, Molldrem J et al. Harnessing graft-versus-malignancy: nonmyeloablative preparative regimens for allogeneic haematopoietic transplantation, an evolving strategy for adoptive immunotherapy. Br J Haematol 2000; 111: 18–29. 5 Shimoni A, Hardan I, Shem-Tov N, Yeshurun M, Yerushalmi R, Avigdor A et al. Allogeneic hematopoietic stem-cell transplantation in AML and MDS using myeloablative versus reducedintensity conditioning: the role of dose intensity. Leukemia 2006; 20: 322–328. 6 McSweeney PA, Niederwieser D, Shizuru JA, Sandmaier BM, Molina AJ, Maloney DG et al. Hematopoietic cell transplantation in older patients with hematologic malignancies: replacing highdose cytotoxic therapy with graft-versus-tumor effects. Blood 2001; 97: 3390–3400. 7 Childs R, Chernoff A, Contentin N, Bahceci E, Schrump D, Leitman S et al. Regression of metastatic renal-cell carcinoma after nonmyeloablative allogeneic peripheral-blood stem-cell transplantation. N Engl J Med 2000; 343: 750–758. Leukemia


2116 8 Giralt S, Thall PF, Khouri I, Wang X, Braunschweig I, Ippolitti C et al. Melphalan and purine analog-containing preparative regimens: reduced-intensity conditioning for patients with hematologic malignancies undergoing allogeneic progenitor cell transplantation. Blood 2001; 97: 631–637. 9 Slavin S, Nagler A, Naparstek E, Kapelushnik Y, Aker M, Cividalli G et al. Nonmyeloablative stem cell transplantation and cell therapy as an alternative to conventional bone marrow transplantation with lethal cytoreduction for the treatment of malignant and nonmalignant hematologic diseases. Blood 1998; 91: 756–763. 10 Shimoni A, Nagler A. Non-myeloablative stem cell transplantation (NST): Chimerism testing as guidance for immune-therapeutic manipulations. Leukemia 2001; 15: 1967–1975. 11 Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assos 1958; 53: 457–481. 12 Gooley TA, Leisenring W, Crowley J, Storer BE. Estimation of failure probabilities in the presence of competing risks: new representations of old estimators. Stat Med 1999; 18: 695–706. 13 Jones RJ, Lee KS, Beschorner WE, Vogel VG, Grochow LB, Braine HG et al. Venoocclusive disease of the liver following bone marrow transplantation. Transplantation 1987; 44: 778–783. 14 Van Besien K, Smith S, Anastasi J, Larson R, Thirman M, Odenike T et al. Irreversible myelosuppression after fludarabine-melphalan conditioning: observations in patients with graft rejection. Blood 2004; 103: 4373–4374. 15 Westerhof GR, Ploemacher RE, Boudewijn A, Blokland I, Dillingh JH, McGown AT et al. Comparison of different busulfan analogues for depletion of hematopoietic stem cells and promotion of donortype chimerism in murine bone marrow transplant recipients. Cancer Res 2000; 60: 5470–5478. 16 Diaconescu R, Flowers CR, Storer B, Sorror ML, Maris MB, Maloney DG et al. Morbidity and mortality with nonmyeloablative compared with myeloablative conditioning before hematopoietic cell transplantation from HLA-matched related donors. Blood 2004; 104: 1550–1558. 17 Shimoni A, Kroger N, Zabelina T, Ayuk F, Hardan I, Yeshurun M et al. Hematopoietic stem-cell transplantation from unrelated donors in elderly patients (age 455 years) with hematologic malignancies: older age is no longer a contraindication when using reduced intensity conditioning. Leukemia 2005; 19: 7–12. 18 Alyea EP, Kim HT, Ho V, Cutler C, Gribben J, DeAngelo DJ et al. Comparative outcome of nonmyeloablative and myeloablative allogeneic hematopoietic cell transplantation for patients older than 50 years of age. Blood 2005; 105: 1810–1814. 19 Blaise DP, Michel Boiron J, Faucher C, Mohty M, Bay JO, Bardoux VJ et al. Reduced intensity conditioning prior to allogeneic stem cell transplantation for patients with acute myeloblastic leukemia as a first-line treatment. Cancer 2005; 104: 1931–1938. 20 Crawley C, Szydlo R, Lalancette M, Bacigalupo A, Lange A, Brune M et al. Outcomes of reduced-intensity transplantation for chronic myeloid leukemia: an analysis of prognostic factors from the Chronic Leukemia Working Party of the EBMT. Blood 2005; 106: 2969–2976. 21 Ho AY, Pagliuca A, Kenyon M, Parker JE, Mijovic A, Devereux S et al. Reduced-intensity allogeneic hematopoietic stem cell transplantation for myelodysplastic syndrome and acute myeloid leukemia with multilineage dysplasia using fludarabine, busulphan, and alemtuzumab (FBC) conditioning. Blood 2004; 104: 1616–1623. 22 Kroger N, Bornhauser M, Ehninger G, Schwerdtfeger R, Biersack H, Sayer HG et al. Allogeneic stem cell transplantation after a fludarabine/busulfan-based reduced-intensity conditioning in patients with myelodysplastic syndrome or secondary acute myeloid leukemia. Ann Hematol 2003; 82: 336–342. 23 Van Besien K, Devine S, Wickrema A, Jessop E, Amin K, Yassine M et al. Regimen-related toxicity after fludarabine-melphalan conditioning: a prospective study of 31 patients with hematologic malignancies. Bone Marrow Transplant 2003; 32: 471–476. 24 Perez-Simon JA, Kottaridis PD, Martino R, Craddock C, Caballero D, Chopra R et al. Nonmyeloablative transplantation with or

Leukemia

25

26

27

28

29

30

31

32

33

34

35 36

37

38

39

without alemtuzumab: comparison between 2 prospective studies in patients with lymphoproliferative disorders. Blood 2002; 100: 3121–3127. Dasgupta RK, Rule S, Johnson P, Davies J, Burnett A, Poynton C et al. Fludarabine phosphate and melphalan: a reduced intensity conditioning regimen suitable for allogeneic transplantation that maintains the graft versus malignancy effect. Bone Marrow Transplant 2006; 37: 455–461. de Lima M, Anagnostopoulos A, Munsell M, Shahjahan M, Ueno N, Ippoliti C et al. Nonablative versus reduced-intensity conditioning regimens in the treatment of acute myeloid leukemia and high-risk myelodysplastic syndrome: dose is relevant for long-term disease control after allogeneic hematopoietic stem cell transplantation. Blood 2004; 104: 865–872. Anderlini P, Saliba R, Acholonu S, Okoroji GJ, Donato M, Giralt S et al. Reduced-intensity allogeneic stem cell transplantation in relapsed and refractory Hodgkin’ disease: low transplant-related mortality and impact of intensity of conditioning regimen. Bone Marrow Transplant 2005; 35: 943–951. Alvarez I, Sureda A, Caballero MD, Urbano-Ispizua A, Ribera JM, Canales M et al. Nonmyeloablative stem cell transplantation is an effective therapy for refractory or relapsed hodgkin lymphoma: results of a spanish prospective cooperative protocol. Biol Blood Marrow Transplant 2006; 12: 172–183. Giralt S, Aleman A, Anagnostopoulos A, Weber D, Khouri I, Anderlini P et al. Fludarabine/melphalan conditioning for allogeneic transplantation in patients with multiple myeloma. Bone Marrow Transplant 2002; 30: 367–373. Kroger N, Perez-Simon JA, Myint H, Klingemann H, Shimoni A, Nagler A et al. Relapse to prior autograft and chronic graft-versushost disease are the strongest prognostic factors for outcome of melphalan/fludarabine-based dose-reduced allogeneic stem cell transplantation in patients with multiple myeloma. Biol Blood Marrow Transplant 2004; 10: 698–708. Morris E, Thomson K, Craddock C, Mahendra P, Milligan D, Cook G et al. Outcomes after alemtuzumab-containing reduced-intensity allogeneic transplantation regimen for relapsed and refractory non-Hodgkin lymphoma. Blood 2004; 104: 3865–3871. van Besien K, Artz A, Smith S, Cao D, Rich S, Godley L et al. Fludarabine, melphalan, and alemtuzumab conditioning in adults with standard-risk advanced acute myeloid leukemia and myelodysplastic syndrome. J Clin Oncol 2005; 23: 5728–5738. Couriel DR, Saliba RM, Giralt S, Khouri I, Andersson B, de Lima M et al. Acute and chronic graft-versus-host disease after ablative and nonmyeloablative conditioning for allogeneic hematopoietic transplantation. Biol Blood Marrow Transplant 2004; 10: 178–185. Mielcarek M, Martin PJ, Leisenring W, Flowers ME, Maloney DG, Sandmaier BM et al. Graft-versus-host disease after nonmyeloablative versus conventional hematopoietic stem cell transplantation. Blood 2003; 102: 756–762. Antin JH, Ferrara JL. Cytokine dysregulation and acute graft-versushost disease. Blood 1992; 80: 2964–2968. Hill GR, Crawford JM, Cooke KR, Brinson YS, Pan L, Ferrara JL. Total body irradiation and acute graft-versus-host disease: the role of gastrointestinal damage and inflammatory cytokines. Blood 1997; 90: 3204–3213. Johansson JE, Brune M, Ekman T. The gut mucosa barrier is preserved during allogeneic, haemopoietic stem cell transplantation with reduced intensity conditioning. Bone Marrow Transplant 2001; 28: 737–742. Clift RA, Buckner CD, Appelbaum FR, Bearman SI, Petersen FB, Fisher LD et al. Allogeneic marrow transplantation in patients with acute myeloid leukemia in first remission: a randomized trial of two irradiation regimens. Blood 1990; 76: 1867–1871. Valcarcel D, Martino R, Sureda A, Canals C, Altes A, Briones J et al. Conventional versus reduced-intensity conditioning regimen for allogeneic stem cell transplantation in patients with hematological malignancies. Eur J Haematol 2005; 74: 144–151.