Acute Leukemia

Comparable outcomes post allogeneic hematopoietic cell transplant for patients with de novo or secondary acute myeloid leukemia in first remission

Article metrics


Secondary AML (sAML) has a poor prognosis with conventional chemotherapy alone. Allogeneic hematopoietic cell transplantation (HCT) is beneficial for high-risk AML. Data comparing outcomes of transplants for patients with de novo and sAML are limited. We compared outcomes of patients transplanted for de novo and sAML in first complete remission and investigated the effect of age, HCT comorbidity index (HCT-CI) and karyotype in both groups. A total of 264 patients with de novo (n=180) and sAML (n=84) underwent allogeneic HCT between 1999 and 2013. Median age at transplant was 51 years (range 18–71), median follow-up of survivors was 77 months. Evaluation of all patients demonstrated no significant difference between de novo and sAML for overall survival (P=0.18), leukemia-free survival (P=0.17), cumulative incidence of relapse (P=0.51) and non-relapse mortality (P=0.42). Multivariable and propensity score analyses confirmed the comparable outcomes between de novo and sAML post transplant. Although sAML demonstrates outcomes inferior to de novo AML treated with chemotherapy alone, outcomes following allogeneic HCT are comparable between the two groups.


AML is a heterogeneous disease with diverse clinical presentation, molecular pathogenesis and prognosis. The development of AML may occur de novo or subsequent to a previous diagnosis of solid tumors, hematopoietic malignancies or myelodysplasia.1, 2, 3, 4 The occurrence of AML following a previous malignancy may be the consequence of cytoreductive therapy for the primary disease, and thus can be considered treatment-related AML (tAML). However, AML may also develop without cytotoxic management as for instance reported for cases where primary solid tumors were treated with surgery alone.3, 5 The emergence of AML under these conditions likely reflects different mechanisms (non-tAML) such as genetic instability that already may have been associated with the development of the primary malignancy. The concept of progression to AML in the absence of cytotoxic measures is well accepted for myelodysplastic syndrome (MDS) and other hematopoietic malignancies. Studies in the past have often merged patients of both categories (secondary tAML and secondary non-tAML), characterizing them all as secondary AML (sAML).5, 6, 7, 8 Treated conservatively, a worse prognosis was commonly reported for patients with sAML compared with de novo AML.4, 9, 10, 11, 12, 13, 14 It remains debatable whether or not the decreased rate of survival reflects biological differences in the disease process, differences in the frequency of well-recognized high-risk factors such as unfavorable karyotype, or differences in management. In contrast to de novo AML, the role of allogeneic hematopoietic cell transplantation (HCT) as part of the management of sAML is not well established and the effect on the above described subtypes of sAML (secondary tAML and secondary non-tAML) is unknown.

It is the purpose of the current study to examine these issues in a retrospective comparison of patients with de novo and sAML transplanted in first complete remission (CR1). A separate subgroup analysis was performed for sAML patients with primary solid tumors, primary MDS and other primary hematopoietic malignancies that have emerged with or without exposure to cytotoxic therapies.

Materials and methods


The study population included 264 consecutive patients with de novo (n=180) or sAML (n=84) in CR1 aged 18–71 years who had received an allogeneic HCT between 1 January 1999 and 30 May 2013 at the Princess Margaret Cancer Centre, University Health Network, Toronto, Canada. All patients consented to the transplant and the study was approved by the Cancer Registry Data Access Committee and the Research Ethics Board of the University Health Network/Princess Margaret Cancer Centre.


The data included a number of pre-transplant variables such as age at transplant, intensity of the conditioning regimen, hematopoietic progenitor cell source, related or unrelated donor choice and year the transplants were performed. Disease-related variables comprised characterization as de novo or sAML and cytogenetic risk at diagnosis. Patients with sAML were sub-grouped as having emerged from solid tumors, MDS or other hematologic malignancies, and the management of the primary malignancy was recorded. All patients underwent HCT in CR1. The HCT-comorbidity index (HCT-CI) was calculated for each patient and stratified for a low-risk (0–2) or high-risk score (≥3).15

Cytogenetic information at diagnosis was available for 232 patients (88%). Risk was assigned according to the revised Medical Research Council (MRC)16 and the recently published Center for International Blood and Marrow Transplant Research (CIBMTR) criteria.17

Grafts were obtained from 175 related and 89 unrelated donors. Before 2005, HLA-typing of related and unrelated donor–recipient pairs comprised A, B and DR determinants, the majority of typing studies were performed at low resolution. Since 2005, HLA-C and DQ were added and since 2008 HLA-typing was performed at high resolution. Ten of the related donors were mismatched at one antigen, and these transplants occurred before the year 2005. The unrelated donors included 17 with one antigen or allelic mismatch, all of these transplants were performed since 2008.

Conditioning regimens and GvHD prophylaxis

Patients received myeloablative or reduced-intensity (RIC) conditioning regimens. The administration of RIC was based on patient age (>60 years) or the presence of co-morbidities.18

The intensity of conditioning regimens was classified based on criteria published by CIBMTR.19 Myeloablative regimens used from 1999 to 2006 included BuCy (busulfan 3.2 mg/kg × 4 days, cyclophosphamide 60 mg/kg × 2 days) and CyTBI (cyclophosphamide 60 mg/kg × 2 days, total body irradiation (TBI) 1200 cGy). Since 2006 patients were prepared with fludarabine 50 mg/m2 × 4 days, busulfan 3.2 mg/kg × 4 days, TBI 400 cGy in 2 fractions (FBT400). RIC regimens used between 1999 and 2006 included combinations of fludarabine 30 mg/m2 × 4–5 days with either busulfan 3.2 mg/kg × 2 days or TBI 200 cGy. Since 2006 patients were conditioned with fludarabine 30 mg/m2 × 4 days, busulfan 3.2 mg/kg × 2 days and TBI 200 cGy in a single fraction (FBT200).

GvHD prophylaxis comprised cyclosporin A combined with methotrexate (15 mg/m2 on day+1 and 10 mg/m2 on days +3, +6 and +11) or mycophenolate mofetil (15 mg/kg q8h × 30 days). In vivo T-cell depletion (alemtuzumab or ATG) was used in 84 patients. Unrelated donors were identified through Canadian Blood Services (OneMatch Stem Cell and Marrow Network).

Definitions of clinical endpoints

CR was defined morphologically as a bone marrow with <5% blasts and count recovery. CR without platelet recovery (CRp) was documented when patients in CR demonstrated a platelet count of ≤100 000/μl. Relapse criteria included a marrow aspirate with ≥5% blasts, occurrence of blasts in the peripheral blood or development of extra-medullary leukemia following transplant. Overall survival (OS) times were measured from the date of HCT until death from any cause. Alive patients were censored on the date of their last follow-up. Leukemia-free survival (LFS) was defined as time from transplantation to relapse or death from any cause. Non-relapse mortality (NRM) was calculated as death without evidence of disease relapse.

Statistical analysis

The data were updated as of 30 May 2014, representing a minimum follow-up of patients of at least 1 year. Patient demographics and treatment-related outcomes were reported using descriptive statistics. Categorical variables were shown as counts and percentages. The continuous variable of age at transplant was displayed as median with range.

Median follow-up of survivors was calculated using the method of Korn.20 The main outcome variables of interest included OS, LFS, cumulative incidence of relapse (CIR) and cumulative incidence of NRM. OS and LFS rates were calculated using the Kaplan–Meier product-limit method and the impact of variables assessed by the log-rank test. CIR and NRM rates were determined using the competing risk method. For CIR, death was accounted for as competing risk, whereas for NRM relapse was accounted as competing risk.21 Multivariable analyses were performed using Cox proportional hazards regression for the outcome of OS and LFS. For CIR and NRM, the Fine-Grey’s method was performed taking into account competing risk of each endpoint.22 The following clinically meaningful variables were entered into the multivariable model: AML subtype (de novo versus sAML), age at transplant, HCT-CI score, donor status, cytogenetic risk at diagnosis of AML, conditioning regimen intensity and time period of transplant. The impact of variables was considered significant when P<0.05. The influence of cytogenetic risk was evaluated in separate multivariable analyses using either the cytogenetic risk classification by MRC16 or CIBMTR.17

To address the possibility of association of AML subtype (de novo vs sAML) with variables studied in the analysis, we implemented the propensity score matching method23, 24 to obtain comparable subpopulations of patients with de novo and sAML matched for age at HCT, conditioning regimen intensity, HCT-CI score, cytogenetic risk and donor status. For 62 out of the 84 patients with sAML (74%) a suitably matched de novo AML patient was identified. This was followed by univariate comparison of the matched groups for OS and LFS.

Data analysis was performed using SAS Version 9.3 (SAS Institute, Inc., Cary, NC, USA) and the open source statistical software R version 3.0 (R Foundation for Statistical Computing, Vienna, Austria).


Patient and transplant characteristics

Baseline patient, disease and transplant-related characteristics and their distribution for patients with de novo or sAML are listed in Table 1. A total of 264 patients with AML in CR1 were transplanted with a median age of 51 years (range 18–71 years) and equal gender distribution. Median follow-up of survivors was 77 months. Transplants were performed for 180 (68%) patients with de novo AML and 84 with sAML (32%). Peripheral blood stem cells were used in 217 patients (82%) and marrow in 47 (18%) patients. In all, 119 patients were aged <50 years at transplant, 145 were ≥50 years old. In all, 176 patients (67%) were prepared with myeloablative and 88 (33%) with RIC regimens. Incomplete platelet recovery was documented for 48 patients, 27 patients (15%) with de novo AML and 21 patients (25%) with sAML (Table 1). Cytogenetic data at diagnosis, evaluated by the MRC criteria, were available for 232 patients (88%) including 12 patients (5%) with favorable, 167 (72%) with intermediate and 53 (23%) with unfavorable risk karyotypes.16 CIBMTR criteria defined 4 patients (2%) with favorable, 200 (86%) with intermediate and 28 (12%) with unfavorable risk.17 Among the intermediate risk patients, 107 patients demonstrated normal karyotype and molecular studies were performed on 53 of those patients. Fms-like tyrosine kinase 3-internal tandem duplication (FLT3-ITD) mutations were present in 19 of those patients. During the time period 1999–2004, 91 patients (34%) underwent HCT and 173 patients (66%) underwent HCT from 2005 onwards, whereas the distribution of patients with de novo or sAML in both time groups was not significantly different (Table 1).

Table 1 Baseline patient, disease and transplant-related characteristics are listed combined and separately for patients presenting with de novo or secondary AML

Patients with de novo or sAML differed with respect to certain risk indicators. Patients with sAML were more likely to be older, had a higher HCT-CI score and were prepared with RIC. Cytogenetic assessment with both the MRC and CIBMTR scoring system showed no significant differences in the distribution of risk categories for de novo and sAML (Table 1).

The 84 patients transplanted for sAML included 73 (87%) with a previously diagnosed hematological malignancy and 11 (13%) with previous solid tumors. All 11 patients with solid tumors were treated surgically and with cytotoxic therapy. The 73 patients with a previous hematological malignancy comprised 40 patients with MDS and 33 with other hematopoietic disorders. One of the MDS patients received systemic treatment in the form of autologous transplant, 19 of the 33 patients with other hematopoietic malignancies underwent chemotherapy. Details concerning primary diagnosis and respective treatment preceding the diagnosis of sAML are listed in Table 2.

Table 2 Primary disease and treatment-related characteristics of patients developing secondary AML subsequent to a previous malignancy

OS and LFS

The OS (P=0.18) and LFS (P=0.17) of patients with de novo and sAML did not differ when all patients transplanted during the complete study period were evaluated by univariate analysis (Figure 1 and Table 3). However, patients with sAML transplanted before 2005 demonstrated inferior OS and LFS compared with patients with de novo AML. A difference in OS and LFS was not observed between patients with de novo and sAML transplanted since 2005 (Table 3). An evaluation and outcome comparison of patients with sAML was performed taking into account the subgroups tAML and non-tAML. Neither the subtype of primary malignancy nor its management impacted OS (Table 4). A subgroup evaluation of patients with sAML following a previous hematological malignancy (n=73) compared with a primary solid tumor (n=11) also failed to demonstrate a significant difference in OS (P=0.88, hazard ratio (HR)=0.94, 95% confidence interval (CI)=0.41–2.15, Figure 2 and Table 4). Incomplete platelet recovery pre-transplant did not influence OS (P=0.49) or LFS (P=0.44).

Figure 1

Univariate comparison of OS for patients with de novo (n=180) or sAML (n=84) after allogeneic HCT.

Table 3 Comparison of OS and LFS of patients with de novo or secondary AML by age, comorbidity score, cytogenetic risk as defined by MRC and CIBMTR criteria and transplant period (univariate analysis)
Table 4 Univariate comparison between sAML subgroups for OS
Figure 2

OS of patients after allogeneic HCT for sAML. Comparison between patients with previous hematological malignancies (n=73) or solid tumors (n=11).

Univariate analysis for OS of patients stratified by cytogenetic risk group16 demonstrated a 3-year OS of 75%, 51% and 46% for favorable, intermediate and unfavorable risk at diagnosis, respectively. Patients with de novo and sAML demonstrated comparable OS and LFS when subcategories of age, HCT-CI score and cytogenetic risk groups by MRC or CIBMTR criteria were taken into account (univariate analyses, Table 3).

A subgroup analysis was performed for the 34 patients with normal karyotype found to be FLT3-ITD mutation negative, 27 of which were de novo and 7 of which were sAML. Univariate analysis for OS demonstrated no significant difference between the two groups (P=0.66).

Multivariable analysis for OS and LFS (Table 5) demonstrated no significant difference between de novo and sAML (P=0.56 and P=0.66, respectively). Increasing age, higher HCT-CI score, unrelated donor status and time period of the transplant before the year 2005 identified as contributory variables to inferior OS and LFS (Table 5). A further interrogation of the data revealed that the poorer OS and LFS of patients transplanted from unrelated donors pertained to the time period before 2005 (P=0.03 for OS, P=0.01 for LFS). OS and LFS were not different for recipients of related and unrelated grafts since 2005 (P=0.13 for OS, P=0.08 for LFS). OS and LFS were not influenced by cytogenetic risk as determined by MRC criteria (overall P-value 0.29 and 0.18, respectively). A multivariable analysis exchanging the MRC with CIBMTR criteria yielded a similar result (data not shown).

Table 5 The effect of patient, disease and transplant-related characteristics on OS, LFS, CIR and NRM by multivariable analysis

A multivariable analysis for OS stratified for the time period of transplantation before and since 2005 was performed to examine whether or not a time-dependent difference existed between de novo and sAML. The HR between sAML and de novo AML before 2005 was 3.1, the HR since 2005 was 0.69. A comparison of both HRs yielded a P-value of 0.04.

Propensity score-matched analysis

A propensity score-matched analysis was performed on a subset of 62 pairs of matched patients with de novo and sAML as described in the Methods. The examined clinical characteristics noted in the Methods were proportionally distributed between the de novo and sAML groups that were generated (data not shown). Of note, approximately 75% of matched pairs for both groups were transplanted since 2005.

Both groups demonstrated a comparable OS (P=0.54, HR=0.86 for sAML, 95% CI=0.53–1.39, Figure 3a) and LFS (P=0.40, HR=0.82 for sAML, 95% CI=0.51–1.30, Figure 3b).

Figure 3

Univariate comparison between de novo and sAML following propensity score-matched analysis (62 pairs, a) for OS and (b) for LFS.

CIR, NRM and subsequent outcome

Patients with de novo or sAML did not show any difference with respect to CIR post transplant (P=0.51). A multivariable analysis (Table 5) confirmed the observation (P=0.88). None of the other variables investigated in the multivariable analysis influenced CIR.

Fifty-six patients (21% of total) relapsed following the transplant. Median time to relapse post-HCT was 6 months (range 1–93). Their median survival post relapse was 2 months (range 1–41 months) and the 2-year survival 9%. The prognosis of patients relapsing more than 6 months after the transplant was somewhat improved with a 2-year survival of 15% (P=0.01, HR=2.34 for relapse post transplant ≤6 months, 95% CI=1.20–4.56). There was no significant difference in survival post-relapse for patients with de novo and sAML (P=0.64, HR for de novo AML=0.85, 95% CI=0.43–1.68).

The study did not reveal a difference between de novo and sAML with respect to NRM in univariate (P=0.42) and multivariable analysis (P=0.85). The influence of the other tested variables on the multivariable analysis is outlined in Table 5.


In this retrospective study of patients undergoing HCT for AML in CR1 at our center, we compared the post-transplant outcomes of patients presenting with de novo AML and those with sAML.

The data were collected from patients undergoing transplants during a period of 14.5 years starting from the year 1999. During this prolonged time interval, a number of management strategies were introduced that may have influenced transplant outcomes over time. To reduce the impact of these changes, the analyses were performed with the time of transplant as a covariate. In addition to other well-recognized risk factors such as age, HCT-CI score and cytogenetic risk, we were also cognisant of potential difficulties comparing the outcome of patients with sAML emerging from different subtypes (secondary tAML and secondary non-tAML). The subgroup analysis did not demonstrate any significant differences. Based on this observation it is reasonable to group the described subtypes as sAML to evaluate the other risk factors.

Univariate and multivariable assessment strategies did not show any significant difference in OS, LFS, CIR and NRM after HCT for all examined patients with de novo or sAML. This observation was made in spite of a risk factor imbalance with respect to a higher proportion of patients in the sAML group presenting with older age, increased HCT-CI score and preparation with RIC. To address the imbalance, we applied a propensity score matching approach to create a data set of 62 matched pairs and repeated the analysis for OS and LFS. The latter strategy also failed to demonstrate a difference.

One may speculate on the reasons why the survival differences for patients with de novo or sAML transplanted before 2005 were not sustained in the more recent time period. The difference may have resulted from the relatively small number of patients with sAML transplanted before 2005. The similar survival since 2005 between de novo and sAML may reflect improvement in the HLA-typing methodology and standard application of high-resolution techniques since 2008, which would also explain the improved outcomes of unrelated transplants during the same time period. Additional contributing factors may include advances in supportive strategies such as the use of alemtuzumab or ATG as GvHD prophylaxis for recipients of unrelated allografts since 2005.

Allogeneic transplants thus represent a treatment option for sAML patients resulting in outcomes comparable to those seen for patients transplanted with de novo AML. These observations are consistent with previous reports.25 Age continues to be a predominant risk factor, whereas the impact of risk related to intermediate and unfavorable karyotype at diagnosis appears to be substantially modified and improved through allogeneic transplant strategies.

Using the MRC and CIBMTR scoring systems in our data set we were unable to demonstrate in our albeit limited data set a significant difference in OS and LFS for patients with intermediate and unfavorable cytogenetics presenting with de novo or sAML.

Allogeneic transplants are usually not recommended in CR1 for patients with favorable cytogenetics. The question remains debated whether or not to transplant patients with more complex risk profiles but favorable cytogenetics in CR1. Grimwade et al. demonstrated equivalent outcomes of conventional treatment for patients with favorable cytogenetics and sAML compared with de novo AML.26 This observation is not supported by other studies27, 28 in which patients with sAML have a less favorable outcome compared with patients with de novo AML, independent of cytogenetic risk. The latter studies include patients of more advanced age which is considered to represent an important risk factor.

In summary, a number of previous studies have concluded that patients with sAML represent a high-risk group. The poorer outcome compared with patients with de novo AML is likely a reflection of risk factors such as age and karyotype, rather than the nature of a secondary malignancy. Allogeneic transplants appear to result in outcomes comparable to those seen in patients with de novo AML.


  1. 1

    Leone G, Mele L, Pulsoni A, Equitani F, Pagano L . The incidence of secondary leukemias. Haematologica 1999; 84: 937–945.

  2. 2

    Witherspoon RP, Deeg HJ . Allogeneic bone marrow transplantation for secondary leukemia or myelodysplasia. Haematologica 1999; 84: 1085–1087.

  3. 3

    Pagana L, Pulsoni A, Tosti ME, Avvisati G, Mele L, Mele M et al. Clinical and biological features of acute myeloid leukaemia occurring as second malignancy: GIMEMA archive of adult acute leukaemia. Br J Haematol 2001; 112: 109–117.

  4. 4

    Vardiman JW, Thiele J, Arber DA, Brunning RD, Borowitz MJ, Porwit A et al. The 2008 revision of the World Health Organization (WHO) classification of myeloid neoplasms and acute leukemia: rationale and important changes. Blood 2009; 114: 937–951.

  5. 5

    Abdelhameed A, Pond GR, Mitsakakis N, Brandwein J, Chun K, Gupta V et al. Outcome of patients who develop acute leukemia or myelodysplasia as a second malignancy after solid tumors treated surgically or with strategies that include chemotherapy and/or radiation. Cancer 2008; 112: 1513–1521.

  6. 6

    Anderson JE, Gooley TA, Schoch G, Anasetti C, Bensinger WI, Clift RA et al. Stem cell transplantation for secondary acute myeloid leukemia: evaluation of transplantation as initial therapy or following induction chemotherapy. Blood 1997; 89: 2578–2585.

  7. 7

    Ostgård LS, Kjeldsen E, Holm MS, Brown Pde N, Pedersen BB, Bendix K et al. Reasons for treating secondary AML as de novo AML. Eur J Haematol 2010; 85: 217–226.

  8. 8

    Preiss BS, Bergmann OJ, Friis LS, Sørensen AG, Frederiksen M, Gadeberg OV et al. Cytogenetic findings in adult secondary acute myeloid leukemia (AML): frequency of favorable and adverse chromosomal aberrations do not differ from adult de novo AML. Cancer Genet Cytogenet 2010; 202: 108–122.

  9. 9

    Dann EJ, Rowe JM . Biology and therapy of secondary leukaemias. Best Pract Res Clin Haematol 2001; 14: 119–137.

  10. 10

    Josting A, Wiedenmann S, Franklin J, May M, Sieber M, Wolf J et al. Secondary myeloid leukemia and myelodysplastic syndromes in patients treated for Hodgkin's disease: a report from the German Hodgkin's Lymphoma Study Group. J Clin Oncol 2003; 21: 3440–3446.

  11. 11

    Kayser S, Döhner K, Krauter J, Köhne CH, Horst HA, Held G et al. The impact of therapy-related acute myeloid leukemia (AML) on outcome in 2853 adult patients with newly diagnosed AML. Blood 2011; 117: 2137–2145.

  12. 12

    Della Porta MG . Prognosis of secondary acute myeloid leukemia. Leuk Res 2013; 37: 857–858.

  13. 13

    Park SH, Chi HS, Cho YU, Jang S, Park CJ . Evaluation of prognostic factors in patients with therapy-related acute myeloid leukemia. Blood Res 2013; 48: 185–192.

  14. 14

    Ornstein MC, Mukherjee S, Mohan S, Elson P, Tiu RV, Saunthararajah Y et al. Predictive factors for latency period and a prognostic model for survival in patients with therapy-related AML. Am J Hematol 2014; 89: 168–173.

  15. 15

    Sorror ML, Sandmaier BM, Storer BE, Maris MB, Baron F, Maloney DG et al. Comorbidity and disease status based risk stratification of outcomes among patients with acute myeloid leukemia or myelodysplasia receiving allogeneic hematopoietic cell transplantation. J Clin Oncol 2007; 25: 4246–4254.

  16. 16

    Grimwade D, Hills RK, Moorman AV, Walker H, Chatters S, Goldstone AH et al. Refinement of cytogenetic classification in acute myeloid leukemia: determination of prognostic significance of rare recurring chromosomal abnormalities among 5876 younger adult patients treated in the United Kingdom Medical Research Council trials. Blood 2010; 116: 354–365.

  17. 17

    Armand P, Kim HT, Zhang MJ, Perez WS, Dal Cin PS, Klumpp TR et al. Classifying cytogenetics in patients with acute myelogenous leukemia in complete remission undergoing allogeneic transplantation: a Center for International Blood and Marrow Transplant Research study. Biol Blood Marrow Transplant 2012; 18: 280–288.

  18. 18

    Khabori MA, El-Emary M, Xu W, Guyatt G, Galal A, Kuruvilla J et al. Impact of intensity of conditioning therapy in patients aged 40-60 years with AML/myelodysplastic syndrome undergoing allogeneic transplantation. Bone Marrow Transplant 2011; 46: 516–522.

  19. 19

    Bacigalupo A, Ballen K, Rizzo D, Giralt S, Lazarus H, Ho V et al. Defining the intensity of conditioning regimens: working definitions. Biol Blood Marrow Transplant 2009; 15: 1628–1633.

  20. 20

    Korn EL . Censoring distributions as a measure of follow-up in survival analysis. Stat Med 1986; 5: 255–260.

  21. 21

    Pepe MS, Mori M . Kaplan-Meier marginal or conditional probability curves in summarizing competing risks failure time data? Stat Med 1993; 12: 737–751.

  22. 22

    Fine J, Gray R . A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 1999; 94: 496–509.

  23. 23

    D'Agostino RB Jr . Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group. Stat Med 1998; 17: 2265–2281.

  24. 24

    Liem YS, Wong JB, Hunink MM, de Charro FT, Winkelmayer WC . Propensity scores in the presence of effect modification: a case study using the comparison of mortality on hemodialysis versus peritoneal dialysis. Emerg Themes Epidemiol 2010; 7: 1.

  25. 25

    Gyurkocza B, Storb R, Storer BE, Chauncey TR, Lange T, Shizuru JA et al. Nonmyeloablative allogeneic hematopoietic cell transplantation in patients with acute myeloid leukemia. J Clin Oncol 2010; 28: 2859–2867.

  26. 26

    Grimwade D, Walker H, Oliver F, Wheatley K, Harrison C, Harrison G et al. The importance of diagnostic cytogenetics on outcome in AML: analysis of 1,612 patients entered into the MRC AML 10 trial. The Medical Research Council Adult and Children's Leukaemia Working Parties. Blood 1998; 92: 2322–2333.

  27. 27

    Schoch C, Kern W, Schnittger S, Hiddemann W, Haferlach T . Karyotype is an independent prognostic parameter in therapy-related acute myeloid leukemia (t-AML): an analysis of 93 patients with t-AML in comparison to 1091 patients with de novo AML. Leukemia 2004; 18: 120–125.

  28. 28

    Borthakur G, Lin E, Jain N, Estey EE, Cortes JE, O'Brien S et al. Survival is poorer in patients with secondary core-binding factor acute myelogenous leukemia compared with de novo core-binding factor leukemia. Cancer 2009; 115: 3217–3221.

Download references


We thank Otsuka Pharmaceuticals Inc for their generous financial support used for data collection and analysis. Funding provided by Otsuka Pharmaceuticals Inc for data collection and analysis.

Author information

Correspondence to H A Messner.

Ethics declarations

Competing interests

Funding provided by Otsuka Pharmaceuticals Inc for data collection and analysis. The authors have no other conflict of interest or financial disclosures to declare.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Further reading