Volume 13, Issue 5 , Pages 601-607, May 2007
Equivalent Survival for Sibling and Unrelated Donor Allogeneic Stem Cell Transplantation for Acute Myelogenous Leukemia
Article Outline
Abstract
Recent studies have shown comparable survival outcomes for unrelated donor (URD) stem cell transplantation in chronic myelogenous leukemia compared to sibling donors. We compared outcomes for 105 patients aged 16 to 59 years undergoing URD transplants for acute myelogenous leukemia (AML) who were reported to the Australasian Bone Marrow Transplant Recipient Registry between 1992 and 2002, and a strictly selected matching set of 105 HLA-matched sibling donor (MSD) transplants. There was no significant difference between URD and MSD controls in the distributions of time from diagnosis to transplant, donor-recipient sex match, prior therapies, donor age, or performance status. The median follow-up of live URD patients was 4.4 years and for live MSD controls was 6.3 years. There were 18 good risk (complete remission [CR]1) and 87 poor risk (>CR1) recipients in both URD and sibling groups. Five-year disease-free survival (DFS) was not significantly different for good-risk URD and sibling donor recipients (62% versus 40%, P = .2), or poor-risk URD and sibling recipients (21% versus 25%, P = .2). In a stratified multivariate Cox regression model, the independent adverse risk factors for DFS were recipient cytomegalovirus positivity (P = .01) and the interaction of URD and earlier year of transplant (P = .006). Both neutrophil and platelet engraftment were significantly more rapid in the sibling group, but transplant-related mortality at 100 days was not significantly different. There was no difference in the cumulative incidence of acute graft-versus-host disease grade II or above at 100 days. Relapse occurred in 28% of good risk URD subjects and 16% of siblings (P = .3), and in poor risk subjects 39% and 29%, respectively (P = .2). Based on this data, URD allografts should be considered in AML patients without a matched sibling donor. This study provides a rationale for a larger prospective study of risk factors in allogeneic transplantation for AML and a guide on the subset of patients who may most benefit from an unrelated donor allograft in AML.
Key Words: Acute myelogenous leukemia, Hemopoietic stem cell transplantation
Introduction
Allogeneic hemopoietic stem cell transplantation (HSCT) is a potentially curative procedure in a wide range of hematologic malignancies [1]. It is generally believed that the major benefit of allogeneic HSCT is a reduction in relapse presumably on the basis of the graft-versus-leukemia (GVL) effect. Acute myelogenous leukemia (AML) is one hematologic malignancy that has been demonstrated to be susceptible to this GVL effect, making it the most common indication for allogeneic transplant in Australasian and International registries [2]. Allogeneic HSCT is often recommended for patients with intermediate risk AML in first complete remission (CR1) who have an HLA identical sibling match. The toxicity of the procedure, however, can offset the reduction in relapse making the decision to perform allografting in CR1 controversial [3]. In contrast, allogeneic HSCT is likely to be the only curative procedure for AML patients beyond CR1, and in this setting both unrelated donor (URD) and matched sibling donor (MSD) HSCT are widely used [4].
Given that only 30% of patients have a potential HLA MSD, it is likely that the use of histocompatible URD may increase the potential donor pool for patients that are considered eligible for allogeneic HSCT. URD HSCT has, however, historically been perceived to be associated with a higher transplant-related mortality (TRM) and lower disease-free survival (DFS) [5]. In contrast to chronic myelogenous leukemia (CML) where studies have demonstrated equivalent outcomes using MSD and MUD donors in certain circumstances [6], there is very little data on the use of MUD donors in AML. More recently, the use of peripheral blood stem cells (PBSC) and high-resolution HLA typing have made the use of alternative donors a realistic choice [7]. In this study we have directly compared the outcome of MSD and MUD donor allografts for AML in 7 Australian centers and have found that when matched for disease stage and age the outcomes are equivalent.
Patients and Methods
Study Design
The study was a retrospective matched case-historic control design. Transplant recipients were selected for this study using the Australasian Bone Marrow Transplant Recipient Registry, which captures details on more than 95% of HSCT carried out in Australia and New Zealand each year. Study subjects (n = 105) were aged between 16 and 59 years at transplant, and were treated with a first allograft for AML from HLA-A, -B, and -DR identical unrelated volunteer donors. In general, serologic typing was used for HLA-A and -B, whereas molecular typing was used for HLA-DR B1. Subjects received unmanipulated transplants between the years of 1992 and 2002 inclusive, at 1 of the 7 Australian hospitals that participated in this study. Study controls were selected, 1 per subject, from 450 patients receiving their first allogeneic transplant for AML from HLA-A, -B, and -DR identical sibling donors from 1992 to 2002 at 1 of the same centers. Study controls were matched with subjects using a hierarchy of disease stage at transplant, recipient age, and sex. Age and disease stage at transplant were selected as stratifying factors because of their strong and well documented effects on survival [2, 8, 9, 10, 11]. Sex was chosen as the third stratifying factor to minimize any chance of unintentional bias in selection of the controls. Where there was a choice of more than 1 control, a random algorithm was used to select 1. The close-out date was December 31, 2004. An initial request for data was made in 2003, with data analyzed in June 2005.
Definitions
Definitions used in this study follow IBMTR guidelines unless otherwise specified. Good risk disease was defined as patients in CR1 at the time of transplantation, whereas poor risk disease was defined as patients beyond CR1, including patients in CR2, first relapse, and other disease stages as outlined in Table 1. Cytogenetic abnormalities at diagnosis were classed as good, t(8;21)(q22;q22), inv(16)(p13q22), t(15;17)(q22;q12), intermediate (normal karyotype and all other), or poor risk (monosomy of 5 or 7, 3q, or complex karyotype). Patients were considered evaluable for engraftment and acute graft-versus-host disease (aGVHD) if they survived at least 21 days after HSCT. Patients who survived longer than 100 days posttransplant were considered evaluable for chronic GVHD (cGVHD). Day of neutrophil engraftment was defined as the first of 3 consecutive days when the blood neutrophil count was >0.5 × 109/L. Day of platelet engraftment was defined as the first day when the blood platelet count was >20 × 109/L and there had been no platelet transfusions in the previous 7 days. TRM was defined as deaths in the first 100 days posttransplant from all causes other than relapse or persistent disease. GVHD was classified as acute when occurring up to 100 days posttransplant, and chronic after this time. Grading of aGVHD was carried out according to criteria used by IBMTR. Performance status at transplant was defined as good (Karnofsky performance scale of 80% or greater, or ECOG performance status of 0-1), or poor otherwise.
Table 1. Characteristics of Subjects and Controls
| Unrelated Donor | Sibling Donor | P | |
|---|---|---|---|
| n | 105 | 105 | — |
| Good risk (CR1 at transplant) (n) | 18 | 18 | ns |
| Poor risk (>CR1 at transplant) (n) | 87 | 87 | |
 CR2 | 23 | 16 | |
| 1st Relapse | 25 | 41 | |
| Other | 39 | 30 | |
| Age: median (range) | 35 (16-59) | 35 (16-58) | ns |
| Sex (n) | |||
| Male | 56 | 56 | ns |
| Female | 49 | 49 | |
| Transplant 1992-1996 (n) | 33 | 49 | 0.03 |
| Transplant 1997-2002 (n) | 72 | 56 | |
| Cytogenetic abnormalities | |||
| Good | 10 | 17 | ns |
| Intermediate | 32 | 38 | |
| Poor | 42 | 29 | |
| Not done | 21 | 21 | |
| Source of stem cells (n) | |||
| Bone marrow | 99 | 59 | <.0005 |
| PBSC | 5 | 41 | |
| Marrow+PBSC | 1 | 5 | |
| Donor-recipient CMV status (n) | |||
| Neg-Neg | 25 | 20 | .02 |
| Neg-Pos | 37 | 24 | |
| Pos-Neg | 14 | 10 | |
| Pos-Pos | 29 | 51 | |
| Pretransplant conditioning (n) | |||
| Cy+TBI | 98 | 18 | <.0005 |
| Bu+Cy | 3 | 74 | |
| Other | 4 | 13 | |
| GVHD prophylaxis (n) | |||
| CSP+MTX | 87 | 79 | .002 |
| CSP only | 0 | 13 | |
| CSP+MTX+other | 17 | 11 | |
| CSP+other | 1 | 2 | |
| Nucleated cell dose × 108/kg | 2.9 (0.17-19.7) | 3.2 (0.08-23.3) | <.0005 |
| CD34+ dose × 106/kg | 3.1 (0.9-10) | 4.3 (0.5-14.7) | .02 |
Analysis Methods and Software Used
Differences between groups were assessed using either a Mann-Whitney U-test for continuous variables, or in the case of categoric variables, chi-squared test, or Fisher’s exact test as appropriate. aGVHD and relapse incidence were determined using cumulative incidence curves, treating death as a competing risk. Overall survival (OS) and DFS were calculated using Kaplan-Meier product limit estimates, and differences between groups were assessed with the log-rank test. Differences between groups in incidence of GVHD in specific organs were assessed using chi-squared tests.
Multivariate analysis was carried out to find the significant factors affecting OS, DFS, incidence of relapse, and incidence of aGVHD. The following factors were considered as independent variables in an initial univariate analysis: donor type (unrelated/sibling), cytogenetic abnormalities at diagnosis, length of time to remission postinitial treatment, length of time between diagnosis and transplant, year of transplant, donor sex, donor and recipient cytomegalovirus (CMV) status, stem cell source, pretransplant conditioning regimen, GVHD prophylaxis, and cell doses [8, 9, 11, 12]. Interaction terms for donor type (sibling/unrelated) with all other independent variables were also tested.
The significance of the above factors on OS, DFS, incidence of relapse, and incidence of acute GVHD was assessed using stratified multivariate Cox regression analysis, with disease stage at transplant, recipient age, and sex as the strata. Independent variables with nonsignificant effects were progressively eliminated from the model by backward selection using likelihood ratio tests. In all tables, the P-value has been displayed as <.0005 when statistical output displayed it as 0, and >.9 when displayed as 1. All tests were 2 tailed. Power calculations were carried out using PASS 2005 statistical software. Survival curves were produced using Graphpad Prism Version 3.0. Cumulative incidence curves were produced using NCSS 2004 statistical software. Statistical difference tests and multivariate Cox regressions were carried out using SPSS Version 12.0 statistical software.
Results
Patient and Disease Characteristics
The final study group comprised 105 subjects (URD recipients) and 105 controls (sibling donor recipients). The subjects and controls were well matched for disease stage, age, and sex, with no statistically significant differences in any of these parameters (Table 1). At this time, the median follow-up time for live URD subjects was 4.4 years (range: 369 days-9.2 years), and for live MSD controls, 6.3 years (range: 162 days-10.9 years). There were no statistically significant differences between subjects and controls in the distributions of cytogenetic abnormalities at diagnosis, length of time from diagnosis to transplant, donor-recipient sex match, prior therapies, donor age, or performance status at transplant (Table 1).
URD transplants were more likely to be after 1997, and bone marrow was more commonly used as the stem cell source (P < .0005). The proportion of CMV negative donors was significantly higher among URD transplants (P = .02). The proportion of cyclophosphamide-total body irridiation (Cy-TBI) conditioning among URD transplants was higher than among siblings (P < .0005). Cyclosporin alone was more commonly used in sibling donor recipients (P = .002). The median cell doses for both total nucleated cells and CD34+ cells were significantly lower for URD recipients than for siblings (P < .0005).
Transplant Outcome
Both neutrophil and platelet engraftment were significantly more rapid in the sibling group (Table 2). TRM at 100 days was not significantly different between groups for patients in CR1, or beyond CR1. There were also no significant differences between the groups in the incidence of other adverse events such as interstitial pneumonitis, veno-occlusive disease, hemorrhagic cystitis, or CMV infection posttransplant. There was no difference in the cause of death, including relapse, GVHD, and infection, in the first year post transplant between the unrelated and sibling donor groups.
Table 2. Transplant Outcomes
| Unrelated Donor (n = 105) | Sibling Donor (n = 105) | P | |
|---|---|---|---|
| Days to neutrophil engraftment: median (range) | 18(10-34) | 17(9-41) | .04 |
| Days to platelet engraftment: median (range) | 28(8-55) | 19(10-45) | <.0005 |
| Transplant related mortality (n, %) | |||
| CR1 | 0(0%) | 3(3%) | ns |
| Not in CR1 | 25(24%) | 17(16%) | ns |
| Overall survival at 5 years | |||
| CR1 | 69% | 40% | ns |
| Not in CR1 | 24% | 31% | ns |
| Disease-free survival at 5 years | |||
| CR1 | 62% | 40% | ns |
| Not in CR1 | 21% | 25% | ns |
Incidence of Acute GVHD
The cumulative incidence of aGVHD Grade II-IV in evaluable patients at day 100 posttransplant was 59% for good-risk URDs versus 50% for good-risk siblings (P = .5), and 55% for poor-risk URDs compared with 44% for poor-risk sibs (P = .1). There was a significantly increased incidence of GVHD Grade II-IV of the skin in the URD recipients (P < .0005, Table 3). A stratified Cox regression analysis was carried out on the incidence of aGVHD testing the independent variables and interactions as listed in the Methods section. No combination of factors had a significant effect on the incidence of aGVHD.
Table 3. GVHD for Patients Who Survived >21 Days Posttransplant
| MUD (n = 96) | Sibling (n = 96) | P | |
|---|---|---|---|
| Acute GVHD | |||
| Skin | 74(77%) | 52(54%) | .001 |
| Liver | 26(27%) | 24(25%) | ns |
| Gut | 24(25%) | 30(31%) | ns |
| Skin stage 2 or higher | 48(50%) | 21(22%) | <.0005 |
| Liver stage 2 or higher | 17(18%) | 17(18%) | ns |
| Gut stage 2 or higher | 11(11%) | 20(21%) | ns |
| Chronic GVHD | |||
| Limited | 12(12.5) | 8(8.3) | ns |
| Extensive | 40(41.7) | 47(48.9) | ns |
Incidence of Chronic GVHD
The cumulative incidence of cGVHD at 1 year posttransplant was 94% for good-risk URD transplants and 61% for good-risk sibs (P = .04), and 68% for poor-risk URD compared with 79% for poor-risk sibs (P = .4). The rates of cGVHD among patients alive at day 100 posttransplant is shown in Table 3.
Incidence of Relapse
The cumulative incidence of hematologic relapse at 2 years posttransplant for patients in CR1 at the time of transplant was 28% for URD and 16% for siblings (P = .3), whereas for those beyond CR1, 39% and 29% (P = .2). A stratified Cox regression analysis was carried out on the incidence of hematologic relapse testing the effects of independent variables as listed in the Methods section. No combination of factors had a significant effect on the incidence of hematologic relapse.
Overall and DFS
Kaplan-Meier curves for all URD and MSD patients are illustrated in Figure 1. At 5 years the DFS for all MSD and URD patients was 31.5% and 29%, respectively (P = .6). For patients in CR1 at the time of transplant, the 5-year OS posttransplant was 69% for URD and 40% for siblings (P = .2). Because of the low numbers (n = 18) in these 2 groups the power of this comparison is low (40%) so the finding of nonsignificance is expected. and results can only be taken as indicative. For those not in CR1, the 5-year OS posttransplant was 24% for URD and 31% for siblings (P = .2) (Table 2). Stratified Cox multivariate regressions were carried out on OS and DFS testing the effects of the independent variables listed above. The only significant adverse risk factors were recipient CMV positivity and the interaction factor of URD and earlier year (1992 to 1996) of transplantation (Table 4).
Figure 1.
Kaplan-Meier curve of DFS (a) and overall survival (b) in URD and MSD allogeneic HSCT for AML.
Table 4. Significant Independent Risk Factors for Overall and Disease-Free Survival—Results of Multivariate Analysis
| Factor | RR | 95% CI | P |
|---|---|---|---|
| Overall survival | |||
| CMV status: recipient positive | 1.741 | (1.189,2.549) | .004 |
| Interaction: unrelated donor × year of transplant 1992-1996 | 1.711 | (1.121,2.613) | .01 |
| Disease-free survival | |||
| CMV status: recipient positive | 1.602 | (1.109,2.313) | .01 |
| Interaction: unrelated donor × year of transplant 1992-1996 | 1.800 | (1.188,2.728) | .006 |
Discussion
This retrospective matched case-control led study has demonstrated similar outcomes for allogeneic HSCT in AML using either a sibling or URD. This is one of the few studies to compare these different donor sources in the context of AML and has implications for the use of URD donors in AML. Results from the multivariate analysis suggest that the use of a CMV positive recipient and transplantation from 1992-1996 using an URD are the major adverse factors for both DFS and OS. Although this study predominantly assessed patients with advanced AML, it appears that the use of a URD particularly in a CMV-negative recipient in recent years is a valid therapeutic option in the treatment of AML patients.
There have been few studies directly comparing the use of adult URD and sibling donors in HSCT for leukemia. Weisdorf et al [6] examined this question in the context of CML using the NMDP database and demonstrated that the outcome was similar between MSD and URD when patients <30 years received an allograft within 1 year of diagnosis while in the chronic phase. This study was one of the first to confirm a role for URDs for this disease. More recently, the use of imatinib [13] has decreased CML as an indication for HSCT, leaving AML as the major indication for unrelated and sibling donor transplantation. In this AML study there was no difference in survival between both good- and poor-risk URD and MSD recipients. The study was, however, not powered to assess the difference in CR1 because of the small numbers. Larger numbers are required to confirm the findings in this analysis, particularly for those patients in CR1.
In 1997, Szydlo et al [5] compared IBMTR data from MSD and URD transplants in acute lymphoblastic leukemia (ALL), AML, and CML patients. The data from this trial compared outcomes only from patients transplanted before 1991 and showed a higher TRM and a lower DFS in the unrelated setting. This study is difficult to compare to this current AML study given that many patients were mismatched at both Class I and II antigens, and there was no matching for age or disease stage. An Austrian study also assessed donor source in allogeneic HSCT for AML [11]. This was not a matched cohort study and contained only 40 URD patients; however, a multivariate analysis did confirm that donor source was not a major prognostic factor for DFS. Eapen et al [14] confirmed that the use of URD donors in infants with AML in CR1 was justified given the equivalent survival of MSD and URD allografts. The role of MUD donors in AML has also been assessed in the nonmyeloblative setting by Bertz et al [15]; however. there was no comparative sibling cohort. Recently, the French Society of Bone Marrow Transplantation and Cell Therapy have demonstrated similar survival for URD and MSD in CML, myelodysplastic syndromes (MDS), and acute leukemia [16]. This study also confirmed the importance of positive recipient CMV serology in outcome.
The major limitation of this study is the retrospective cohort design. The MSD and URD donors were carefully matched for age and disease stage, which were believed to be important prognostic factors based on the CML allogeneic HSCT data [6] and previous AML studies [9]. The lack of cytogenetics in 21% of patients is of uncertain importance but may have had some effect on outcome. There were other significant differences between the 2 groups, in that URD HSCT more frequently used Cy/TBI conditioning, bone marrow as the stem cell source, and CMV-negative donors. The latter 2 factors are determined by the physician’s preference or guidelines for the use of granulocyte-colony stimulating factor (G-CSF) in normal donors in some registries. It is difficult to determine whether these factors may have affected the results of this analysis. Large randomized trials have previously confirmed the equivalence of Cy/TBI and Cy/Bu conditioning regimens in AML [17]. Likewise, bone marrow and peripheral blood stem cells have previously been compared in Australian centers [18] with no clear advantage found, suggesting that these factors would not have had a major impact on the conclusions from this data.
The multivariate analysis confirmed that CMV serostatus of the recipient is an important adverse prognostic factor in the allogeneic HSCT setting. Several other studies have also reported this observation in allogeneic HSCT trials [19]; however, the data is more definitive in the T cell-depleted setting [20]. The second adverse factor outlined in the multivariate analysis is the interaction of URD and transplantation in the first half of the study period (1992-1996). It is unclear why this is the case, but it is possible that better supportive care, including newer antifungals and the increased use of high-resolution HLA typing may have improved URD transplantation since 1996. Further studies will be required to clarify this finding.
Although this study only has significant power to make conclusions about advanced AML, it would appear that transplantation from both a URD and a matched sibling have equivalent outcomes in this patient group. The results of this analysis suggest that URD HSCT should be considered early in the decision-making process for patients with AML. Further studies are required to confirm subgroups of patients who may most benefit from URD allografting in AML.
References
- . Bone marrow transplantation. N Engl J Med. 1994;330:827–838
- . Haemopoietic stem cell transplantation in Australia and New Zealand, 1992-2001: progress report from the Australasian Bone Marrow Transplant Recipient Registry. Int Med J. 2005;35:18–27
- The value of allogeneic bone marrow transplant in patients with acute myeloid leukaemia at differing risk of relapse: results of the UK MRC AML 10 trial. Br J Haematol. 2002;118:385–400
- Long-term outcome after allogeneic hematopoietic stem cell transplantation for advanced stage acute myeloblastic leukemia: a retrospective study of 379 patients reported to the Societe Francaise de Greffe de Moelle (SFGM). Bone Marrow Transplant. 2000;26:1157–1163
- Results of allogeneic bone marrow transplants for leukemia using donors other than HLA-identical siblings. J Clin Oncol. 1997;15:1767–1777
- Allogeneic bone marrow transplantation for chronic myelogenous leukemia: comparative analysis of unrelated versus matched sibling donor transplantation. Blood. 2002;99:1971–1977
- Impact of HLA class I and class II high-resolution matching on outcomes of unrelated donor bone marrow transplantation: HLA-C mismatching is associated with a strong adverse effect on transplantation outcome. Blood. 2004;104:1923–1930
- Prognostic factors of patients with acute myeloid leukemia (AML) allografted in first complete remission: an analysis of the EORTC-GIMEMA AML 8A trial. Bone Marrow Transplant. 1996;17:993–1001
- Transplantation of peripheral blood stem cells as compared with bone marrow from HLA-identical siblings in adult patients with acute myeloid leukemia and acute lymphoblastic leukemia. J Clin Oncol. 2002;20:4655–4664
- . Summary slides 2003-Part II. IBMTR/ABMTR newsletter. 2004;Volume 10(2):6
- Factors affecting long-term outcome after allogeneic haematopoietic stem cell transplantation for acute myelogenous leukaemia: a retrospective study of 172 adult patients reported to the Austrian Stem Cell Transplantation Registry. Br J Haematol. 2002;117:914–923
- Late cytomegalovirus disease and mortality in recipients of allogeneic hematopoietic stem cell transplants: importance of viral load and T-cell immunity. Blood. 2003;101:407–414
- . Imatinib mesylate—the new gold standard for treatment of chronic myeloid leukemia. N Engl J Med. 2003;348:1048–1050
- Comparable long-term survival after unrelated and HLA-matched sibling donor hematopoietic stem cell transplantations for acute leukemia in children younger than 18 months. J Clin Oncol. 2006;24:145–151
- . Allogeneic stem-cell transplantation from related and unrelated donors in older patients with myeloid leukemia. J Clin Oncol. 2003;21:1480–1484
- Allogeneic marrow stem-cell transplantation from human leukocyte antigen-identical siblings versus human leukocyte antigen-allelic-matched unrelated donors (10/10) in patients with standard-risk hematologic malignancy: a prospective study from the French Society of Bone Marrow Transplantation and Cell Therapy. J Clin Oncol. 2006;24:5695–5702
- Comparison of outcome following allogeneic bone marrow transplantation with cyclophosphamide-total body irradiation versus busulphan-cyclophosphamide conditioning regimens for acute myelogenous leukaemia in first remission. Br J Haematol. 2002;119:1115–1124M
- Allogeneic haemopoietic cell transplants in Australia, 1996—a multi-centre retrospective comparison of the use of peripheral blood stem cells with bone marrow. Bone Marrow Transplant. 2001;28:21–27
- . The impact of cytomegalovirus serostatus of donor and recipient before hematopoietic stem cell transplantation in the era of antiviral prophylaxis and preemptive therapy. Blood. 2003;103:2003–2008
- Cytomegalovirus seropositivity adversely influences outcome after T-depleted unrelated donor transplant in patients with chronic myeloid leukaemia: the case for tailored graft-versus-host disease prophylaxis. Br J Haematol. 2001;112:228–236
PII: S1083-8791(07)00118-8
doi:10.1016/j.bbmt.2007.01.073
© 2007 American Society for Blood and Marrow Transplantation. Published by Elsevier Inc. All rights reserved.
Volume 13, Issue 5 , Pages 601-607, May 2007
