Acute Leukemia

Allogeneic hematopoietic cell transplant for AML: no impact of pre-transplant extramedullary disease on outcome

Article metrics


The impact of extramedullary disease (EMD) in AML on the outcomes of allogeneic hematopoietic cell transplantation (alloHCT) is unknown. Using data from the Center for International Blood and Marrow Transplant Research, we compared the outcomes of patients who had EMD of AML at any time before transplant, with a cohort of AML patients without EMD. We reviewed data from 9797 AML patients including 814 with EMD from 310 reporting centers and 44 different countries, who underwent alloHCT between and 1995 and 2010. The primary outcome was overall survival (OS) after alloHCT. Secondary outcomes included leukemia-free survival (LFS), relapse rate and treatment-related mortality (TRM). In a multivariate analysis, the presence of EMD did not affect either OS (hazard ratio 1.00, 95% confidence interval (CI) 0.91–1.09), LFS (0.98, 0.89–1.09), TRM (relative risk 0.92, 95% CI 0.80–1.16, P=0.23) or relapse (relative risk=1.03, 95% CI, 0.92–1.16; P=0.62). Furthermore, the outcome of patients with EMD was not influenced by the location, timing of EMD, or intensity of conditioning regimen. The presence of EMD in AML does not affect transplant outcomes and should not be viewed as an independent adverse prognostic feature.


Extramedullary disease (EMD) in AML refers to disease found in organs or tissue outside the blood or bone marrow. The most common manifestations of EMD include myeloid sarcomas, leukemia cutis and meningeal leukemia. Although the exact frequency is unknown, EMD has been estimated to occur in 3%–8% of patients with AML, and has been reported to be more common in patients with core-binding factor leukemia, FAB M2/M4/M5, high WBC count and increased age.1 Historically, the presence of EMD has been considered a poor prognostic feature in AML.2 The impact of EMD may depend on the site of EMD as well as cytogenetic and molecular features. In adult patients with t(8:21), complete remission (CR) rates (50% vs 92%) and overall survival (OS; 5.4 vs 59.5 months) were markedly worse in patients with EMD treated with standard 7+3 regimens.3 In a retrospective analysis of 434 Japanese patients with AML, myeloid sarcomas were associated with higher relapse rate and lower disease-free survival.4

Owing to its potent antitumor effects, it has been suggested that allogeneic hematopoietic cell transplantation (alloHCT) could overcome the potentially poor prognostic impact of EMD in AML. However, data supporting this approach are limited. A retrospective study from the Société Francaise de Greffe de Moelle et de Thérapie Cellulaire registry of 51 patients with myeloid sarcoma, who underwent allogeneic stem cell transplantation demonstrated an OS of 36% at 5 years, confirming that allogeneic stem cell transplantation is a valid therapeutic option.5 Isolated EMD relapses are common following alloHCT in patients with AML, indicating a relative lack of graft versus leukemia effect in EMD sites.6 Furthermore, reduced intensity conditioning regimens, T-cell-depleted grafts, or non-total body irradiation-based conditioning regimens have been associated with higher rates of EMD relapse and may reduce the effectiveness of alloHCT in AML with EMD disease.7, 8, 9, 10

As a prospective study to determine the impact of alloHCT for AML with EMD is not feasible, the Center for International Blood and Marrow Transplant Research (CIBMTR) database offers a comprehensive data set, to identify factors that influence the outcome of alloHCT for AML with EMD. In this study, we compared the outcomes of patients who had EMD of AML at any time before transplant with a cohort of AML patients without EMD. We also examined disease-, treatment- and transplant-related characteristics that affected the outcomes of patients with EMD.

Patients and methods

Data source

The CIBMTR, a voluntary working group of more than 500 transplant centers worldwide, contributed data on consecutive alloHCTs to a statistical center housed both at the Medical College of Wisconsin (Milwaukee, WI, USA) and the National Marrow Donor Program (Minneapolis, MN, USA). Observational studies conducted by CIBMTR are performed with a waiver of informed consent and in compliance with Health Insurance Portability and Accountability Act regulations as determined by the Institutional Review Board and the Privacy Officer of the Medical College of Wisconsin.

Patient selection

The study population consists of AML patients between 18 and 70 years of age, who underwent bone marrow or peripheral blood alloHCT from either an HLA-identical sibling or unrelated donor between 1995 and 2010. Patients with acute promyelocytic leukemia were excluded.

The site of EMD was determined by the reporting center in one of four categories: the central nervous sytem, soft tissue, testes or other. The ‘other’ category was further subdivided into clinically relevant categories such as the ‘skin’ and the ‘liver/spleen’. Pathologic or radiographic confirmation of EM disease was not required. Cytogenetics were classified according to Southwest Oncology Group/Eastern Cooperative Oncology Group (SWOG/ECOG) criteria.11 Conditioning regimens were classified as myeloablative (MA), reduced intensity conditioning, or non-MA.12, 13 CIBMTR classifications of unrelated donor matching were used to define well-matched, partially matched or mismatched categories.

Study end points and definitions

The primary outcome was OS after alloHCT (defined as the time from transplantation to death). Secondary end points included leukemia-free survival (LFS), relapse rate and treatment-related (non-relapse) mortality (TRM; defined as any death in the first 28 days after transplantation or any death after day 28 in continuous remission), incidence of grade II–IV acute GvHD and the presence of chronic GvHD. Surviving patients were censored at the time of last contact.

Statistical analysis

Patient-, disease- and treatment-related factors were compared between EMD and non-EMD groups, using the χ2-tests for categorical and Mann–Whitney test for continuous variables. The probability of LFS and OS were calculated using the Kaplan–Meier estimator, with the variance estimated by Greenwood’s formula. Values for other endpoints were calculated using cumulative incidence curves to accommodate competing risks.14

EMD and non-EMD groups were compared using proportional hazards regression models. Risk factors with significant level of P<0.05 in stepwise model building procedures were included in the outcome models. Potential interactions between the main effect (EMD status) and conditioning intensity, cytogenetic risk and other significant variables were examined.


Patient characteristics

In total, 9797 patients were identified from 310 reporting centers and 44 different countries: 814 with EMD before alloHCT (EMD group) and 8983 without EMD pre-transplant (non-EMD group). The median follow-up of survivors was 58 months (range: 3–191 months) for the EMD group and 60 months (range: 3–194 months) for the non-EMD group.

Table 1 lists patient-, disease-, treatment- and transplant-related variables for all patients. Patients with EMD tended to be younger (median age of 42 vs 46 years, P<0.001), were more likely to have a monocytic subtype (FAB M4–M5, 46 vs 29%, P<0.001), and a higher initial WBC at diagnosis (22 vs 9, P<0.001). The most common site of EMD was the central nervous sytem (n=283, 35%). For other sites, 155 (19%) had skin-only and 112 (14%) possessed lymph node-only EMD. An additional 69 (8%) reported multiple sites of EMD.

Table 1 Characteristics of patients between 18 and 70 years of age who underwent allogeneic transplant for AML between 1995 and 2010 reported to the CIBMTR

Transplant conditioning regimens differed between the two groups. In the EMD group, 82% received an MA preparative regimen with 47% receiving an MA conditioning with total body irradiation compared with 75% and 35%, respectively, in the non-EMD group (P<0.001 for both comparisons). Disease status before conditioning also differed between the non-EMD and EMD groups: primary induction failure 15% versus 12%, CR1 49% versus 37%, CR2 or beyond 20% versus 26%, active relapse 17% versus 24%, respectively (P<0.001). The duration of first remission was shorter for subjects transplanted in CR2 in the EMD group versus the non-EMD group, 9 versus 11 months (P<0.001) with 32% having a CR1 duration of <6 months compared with 19% (P<0.001) in the non-EMD group.

Univariate analysis of outcomes

Comparisons of outcomes between the EMD and non-EMD groups are listed in Table 2. There were no significant differences in LFS or OS, the primary endpoints of our study, between patients with and without EMD in univariate analysis (Figure 1 and Table 2). The 5-year LFS and OS for the EMD group was 33% (95% confidence interval (CI) 30–37%) and 36% (95% CI 32–39%), respectively. The relapse rate in the EMD groups was significantly higher at 1 year (33% versus 29%, P=0.012) and 3 years (39% versus 34%, P=0.022) post transplant compared with the non-EMD group. However, this risk was offset by lower rates of TRM post transplant at 3 years (24% versus 29%, P=0.009) and 5 years (26% versus 31%, P=0.009) in the EMD group.

Table 2 Univariate analysis of outcomes
Figure 1

Analysis of HCT outcome by EMD versus no-EMD. Probability of OS, LFS and cumulative incidence frequency (CIF) of TRM and relapse.

For the 61 patients who proceeded to transplant with active medullary and EMD, univariate analysis showed significantly higher 3-year relapse rate, when compared with the EMD group (62% versus 39%, P=<0.001), and significantly worse 3-year LFS (9% versus 37%, P=<0.001) and OS (10% versus 40%, P=<0.001; Table 2).

Leukemia was the most common cause of death in the non-EMD group (42%) and EMD group (49%), followed by infection and GvHD (Supplementary Table 1). For all patients who relapsed after transplant, there were significant differences in the site of relapse (Supplementary Table 2). In the EMD group, 26% (14% EM site, 12% peripheral blood)/bone marrow and EM site) of patients relapsed at extramedullary sites, whereas only 9% (5% EM site, 4% peripheral blood/bone marrow and EM site) of patients in the non-EMD group relapsed at extramedullary sites. For patients transplanted in CR2 or at relapse, 23% of the EM group relapsed solely at an EM site post transplant and 18% relapsed at both medullary and EM sites.

Multivariate analysis of outcomes

Because of the poor outcomes of subjects with both EMD and active marrow disease at the time of transplant, these patients were excluded from the multivariate analysis. Notably, the presence of EMD did not affect either OS (hazard ratio 1.00, 95% CI 0.91–1.09; P=0.91) or LFS (hazard ratio 0.98, 95% CI 0.89–1.09; P=0.74). In addition, differences in both TRM (relative risk 0.92, 95% CI 0.80–1.06, P=0.23) and relapse (relative risk=1.03, 95% CI, 0.92–1.16; P=0.62) failed to retain their significance in the multivariate model (Table 3 and Supplementary Table 4).

Table 3 Multivariate analysis of outcome

Outcome by site of EMD disease or by onset of EMD

No significant differences were observed in the rate of relapse (P=0.66; Figure 2) or OS based on by site of EMD (P=0.28; Table 4). We compared the outcomes of the 71 patients who had an isolated granulocytic sarcoma compared with the remaining 743 patients with both EMD and marrow involvement of their leukemia (Supplementary Table 3). We also examined the timing of EMD onset, whether EMD was present at the time of diagnosis or at the time of transplantation. Again, in each case there was no difference in OS between these two groups (Table 4).

Figure 2

Cumulative incidence of relapse based on anatomic location of EMD.

Table 4 Univariate analysis based on EMD site and time of EMD onset

Pretransplant conditioning

As the GvL effect might be weaker in EM sites,6 we also tested for any interaction between the presence of EMD and the intensity of the conditioning regimen on the risk of relapse, which might confound our analysis. No interaction was identified between MA and reduced intensity conditioning on the risk of relapse (P=0.1591). After MA conditioning, the relative risk of relapse was 1.09 (95% CI, 0.95–1.24; P=0.21) and for reduced intensity conditioning was 0.89 (95% CI, 0.70–1.14; P=0.36).

Acute and chronic GvHD

At day 100, the incidence of grade II–IV acute GvHD was similar between the EMD and non-EMD groups (35% versus 36%, P=0.60, Table 2). The incidence of chronic GvHD at 1, 3 and 5 years was also similar between the two groups.


This analysis of 814 patients with EM involvement of AML represents the largest and most comprehensive analysis of alloHCT outcomes in this patient population. Historically, EMD has been viewed as a poor prognostic factor, although data supporting this position are limited. For other risk factors in AML such as a monosomal karyotype and FLT3-ITD, the adverse prognostic impact persists despite alloHCT.15, 16 Unexpectedly, when compared with a non-EMD cohort, we did not identify any impact of EMD itself on either survival, disease relapse, or TRM. In addition, the location, timing of EMD, or intensity of conditioning regimen also did not affect transplant outcomes.

As with any large retrospective registry study, there are important limitations in our analysis. The presence and location of EMD was determined by the reporting center and did not require a confirmatory pathologic or radiographic diagnosis. This may result in underreporting of EMD in locations such as lymph nodes and deep soft tissue sites that can escape detection on routine clinical examination, and over reporting in areas such as the skin, where leukemia cutis may be confused with other dermatologic conditions. At the same time, this mirrors current clinical practice in which no guidelines or standards exist for the evaluation of EMD.

Because of the relatively uncommon nature of EMD, we are also unable to make meaningful conclusions about specific individual cytogenetic subtypes, that is, t(8:21) or mixed lineage leukemia (MLL) rearrangements, and more uncommon sites of EMD such as testicular involvement. Molecular genetics in AML is a rapidly evolving field and the CIBMTR data set for AML is limited by lack of uniform molecular characterization of cases. For example, nucleophosmin mutations in cytogenetically normal AML confer a favorable prognosis and have been identified in 15% of isolated myeloid sarcomas through aberrant cytoplasmic localization of the protein by immunohistochemistry.17

Based on our analysis of the CIBMTR, we found that the presence of EMD is not an independent risk factor for relapse, disease-free survival or OS in patients undergoing alloHCT. Furthermore, the outcome of patients with EMD was not influenced by the location, timing of EMD, or intensity of conditioning regimen.


  1. 1

    Byrd JC, Edenfield WJ, Shields DJ, Dawson NA . Extramedullary myeloid cell tumors in acute nonlymphocytic leukemia: a clinical review. J Clin Oncol 1995; 13: 1800–1816.

  2. 2

    Chang H, Brandwein J, Yi QL, Chun K, Patterson B, Brien B . Extramedullary infiltrates of AML are associated with CD56 expression, 11q23 abnormalities and inferior clinical outcome. Leuk Res 2004; 28: 1007–1011.

  3. 3

    Byrd JC, Weiss RB, Arthur DC, Lawrence D, Baer MR, Davey F et al. Extramedullary leukemia adversely affects hematologic complete remission rate and overall survival in patients with t(8;21)(q22;q22): results from Cancer and Leukemia Group B 8461. J Clin Oncol 1997; 15: 466–475.

  4. 4

    Shimizu H, Saitoh T, Hatsumi N, Takada S, Yokohama A, Handa H et al. Clinical significance of granulocytic sarcoma in adult patients with acute myeloid leukemia. Cancer Sci 2012; 103: 1513–1517.

  5. 5

    Chevallier P, Mohty M, Lioure B, Michel G, Contentin N, Deconinck E et al. Allogeneic hematopoietic stem-cell transplantation for myeloid sarcoma: a retrospective study from the SFGM-TC. J Clin Oncol 2008; 26: 4940–4943.

  6. 6

    Harris AC, Kitko CL, Couriel DR, Braun TM, Choi SW, Magenau J et al. Extramedullary relapse of acute myeloid leukemia following allogeneic hematopoietic stem cell transplantation: incidence, risk factors and outcomes. Haematologica 2013; 98: 179–184.

  7. 7

    Craddock C, Nagra S, Peniket A, Brookes C, Buckley L, Nikolousis E et al. Factors predicting long-term survival after T-cell depleted reduced intensity allogeneic stem cell transplantation for acute myeloid leukemia. Haematologica 2010; 95: 989–995.

  8. 8

    Schmid C, Schleuning M, Schwerdtfeger R, Hertenstein B, Mischak-Weissinger E, Bunjes D et al. Long-term survival in refractory acute myeloid leukemia after sequential treatment with chemotherapy and reduced-intensity conditioning for allogeneic stem cell transplantation. Blood 2006; 108: 1092–1099.

  9. 9

    Lee KH, Lee JH, Choi SJ, Kim S, Seol M, Lee YS et al. Bone marrow vs extramedullary relapse of acute leukemia after allogeneic hematopoietic cell transplantation: risk factors and clinical course. Bone Marrow Transplant 2003; 32: 835–842.

  10. 10

    Kogut N, Tsai NC, Thomas SH, Palmer J, Paris T, Murata-Collins J et al. Extramedullary relapse following reduced intensity allogeneic hematopoietic cell transplant for adult acute myelogenous leukemia. Leuk Lymphoma 2013; 54: 665–668.

  11. 11

    Slovak ML, Kopecky KJ, Cassileth PA, Harrington DH, Theil KS, Mohamed A et al. Karyotypic analysis predicts outcome of preremission and postremission therapy in adult acute myeloid leukemia: a Southwest Oncology Group/Eastern Cooperative Oncology Group Study. Blood 2000; 96: 4075–4083.

  12. 12

    Giralt S, Ballen K, Rizzo D, Bacigalupo A, Horowitz M, Pasquini M et al. Reduced-intensity conditioning regimen workshop: defining the dose spectrum. Report of a workshop convened by the center for international blood and marrow transplant research. Biol Blood Marrow Transplant 2009; 15: 367–369.

  13. 13

    Champlin R, Khouri I, Shimoni A, Gajewski J, Kornblau S, Molldrem J et al. Harnessing graft-versus-malignancy: non-myeloablative preparative regimens for allogeneic haematopoietic transplantation, an evolving strategy for adoptive immunotherapy. Br J Haematol 2000; 111: 18–29.

  14. 14

    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.

  15. 15

    Fang M, Storer B, Estey E, Othus M, Zhang L, Sandmaier BM et al. Outcome of patients with acute myeloid leukemia with monosomal karyotype who undergo hematopoietic cell transplantation. Blood 2011; 118: 1490–1494.

  16. 16

    Brunet S, Labopin M, Esteve J, Cornelissen J, Socie G, Iori AP et al. Impact of FLT3 internal tandem duplication on the outcome of related and unrelated hematopoietic transplantation for adult acute myeloid leukemia in first remission: a retrospective analysis. J Clin Oncol 2012; 30: 735–741.

  17. 17

    Falini B, Lenze D, Hasserjian R, Coupland S, Jaehne D, Soupir C et al. Cytoplasmic mutated nucleophosmin (NPM) defines the molecular status of a significant fraction of myeloid sarcomas. Leukemia 2007; 21: 1566–1570.

Download references


We acknowledge the following contributing co-authors for their contributions to this manuscript: Kirk R Schultz, Mark R Litzow, Philip L McCarthy, Mahmoud D Aljurf, Mitchell S Cairo, William A Wood, Celalettin Ustun, Thomas R Klumpp, Edwin M. Horwitz, Jyotishankar Raychaudhuri, Bruce M Camitta, Yi-Bin Chen, Peter H Wiernik, Tsiporah B Shore, Selina M Luger, Ashish Bajel, Harry C Schouten, Grace H Ku, Maxim Norkin, Faiz Anwer, Asmita Mishra, Attaphol Pawarode, Amelia Langston, Mitchell Sabloff, Ann E Woolfrey, Hans-Jochem Kolb, Edmund K Waller, Usama Gergis, John Koreth, Reinhold Munker, Joseph McGuirk, William R Drobyski, Bita Jalilizeinali and H Jean Khoury. Geoffrey L Uy’s contribution is supported by NCI grant K23 CA140707. The CIBMTR is supported by Public Health Service Grant/Cooperative Agreement U24-CA076518 from the National Cancer Institute (NCI), the National Heart, Lung and Blood Institute (NHLBI) and the National Institute of Allergy and Infectious Diseases (NIAID); a Grant/Cooperative Agreement 5U10HL069294 from NHLBI and NCI; a contract HHSH250201200016C with Health Resources and Services Administration (HRSA/DHHS); two grants N00014-12-1-0142 and N00014-13-1-0039 from the Office of Naval Research; and grants from *Actinium Pharmaceuticals, Allos Therapeutics, Inc. and *Amgen, Inc. Anonymous donation to the Medical College of Wisconsin, Ariad, Be the Match Foundation, *Blue Cross and Blue Shield Association, *Celgene Corporation, Chimerix, Inc., Fred Hutchinson Cancer Research Center, Fresenius-Biotech North America, Inc., *Gamida Cell Teva Joint Venture Ltd., Genentech, Inc.,*Gentium SpA, Genzyme Corporation, GlaxoSmithKline, Health Research, Inc., Roswell Park Cancer Institute, HistoGenetics, Inc., Incyte Corporation, Jeff Gordon Children’s Foundation, Kiadis Pharma, The Leukemia & Lymphoma Society, Medac GmbH, The Medical College of Wisconsin, Merck & Co, Inc., Millennium: The Takeda Oncology Co., *Milliman USA, Inc., *Miltenyi Biotec, Inc., National Marrow Donor Program, Onyx Pharmaceuticals, Optum Healthcare Solutions, Inc., Osiris Therapeutics, Inc., Otsuka America Pharmaceutical, Inc., Perkin Elmer, Inc., *Remedy Informatics, *Sanofi US, Seattle Genetics, Sigma-Tau Pharmaceuticals, Soligenix, Inc., St Baldrick’s Foundation, StemCyte, A Global Cord Blood Therapeutics Co., Stemsoft Software, Inc., Swedish Orphan Biovitrum, *Tarix Pharmaceuticals, *TerumoBCT, *Teva Neuroscience, Inc., *THERAKOS, Inc., University of Minnesota, University of Utah and *Wellpoint, Inc. *Corporate members. The views expressed in this article do not reflect the official policy or position of the National Institute of Health, the Department of the Navy, the Department of Defense, Health Resources and Services Administration (HRSA) or any other agency of the U.S. Government.

Author information

Correspondence to G L Uy.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Presented in part at the 2014 BMT Tandem Meetings, Grapevine, TX, 2 March 2014.

Supplementary Information accompanies this paper on Bone Marrow Transplantation website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Goyal, S., Zhang, M., Wang, H. et al. Allogeneic hematopoietic cell transplant for AML: no impact of pre-transplant extramedullary disease on outcome. Bone Marrow Transplant 50, 1057–1062 (2015) doi:10.1038/bmt.2015.82

Download citation

Further reading