Outcome after allogeneic transplantation for adult acute myeloid leukemia patients exhibiting isolated or associated trisomy 8 chromosomal abnormality: a survey on behalf of the ALWP of the EBMT

Article metrics


The aim of this multicenter retrospective analysis was to carry out a survey of overall outcomes after allo-hematopoietic SCT of AML patients harboring trisomy 8 (+8) as the sole chromosomal abnormality or associated with other abnormalities. We have identified 182 de novo AML patients who underwent allo-hematopoietic SCT between 1990 and 2007 exhibiting isolated +8 (n=136) or +8 (n=46) associated with other favorable (n=8), intermediate (n=30), high-risk (n=7) or unknown (n=1) cytogenetic abnormalities reported to the European Group of Blood and Marrow Transplantation (EBMT). With a median follow-up of 48 months, 5-year non-relapse mortality, relapse rate, leukemia-free survival and OS were 25, 30, 45 and 47%, respectively. In a multivariate analysis, leukemia-free survival rate was improved when patients were female and transplanted in CR with an HLA-identical sibling donor. Five-year leukemia-free survival was 41, 88, 57 and 21% in patients bearing isolated +8 or +8 and other cytogenetic abnormalities of good, intermediate or poor-risk, respectively. Our retrospective data show that allo-hematopoietic SCT is an effective treatment for AML patients harboring +8. The accompanying cytogenetic abnormality to +8 seems to influence outcomes of these patients.


Trisomy 8 (+8) is one of the most common chromosomal abnormalities encountered in AML, occurring in about 10–15% of cases.1 The prognostic effect of +8 as a sole chromosomal abnormality in AML remains debated and was found in the literature to be associated either with an intermediate or a poor prognosis.1, 2, 3 The results of the Medical Research Council-AML10 trial has assigned AML patients with isolated +8 in the ‘intermediate’ cytogenetic abnormality risk group, showing a 3-year OS and relapse risk of 42 and 42%, respectively. This study suggested also a worsened outcome for these patients in comparison with other cytogenetic abnormality risk group categories within this heterogeneous ‘intermediate’ prognosis group.4 Alternatively, patients with +8 occurring concomitantly with other cytogenetic aberrations seem to have the prognosis conferred by the accompanying cytogenetic abnormality, at least in the favorable prognosis group ((t(8;21), t(15;17), inv(16)) or when it is observed concomitantly to additional 11q23 aberrations.1, 5, 6 Currently, data assessing specifically the role of allo- hematopoietic SCT (HSCT) in the setting of AML with +8 are still relatively sparse.5, 7 The aim of this multicenter retrospective analysis was to carry out a survey on overall outcomes after allo-HSCT of AML patients harboring +8 as a sole chromosomal abnormality or associated with other abnormalities.

Patients and methods

Data collection

EBMT registry database was used to collect the data on allo-HSCT of AML patients harboring isolated or associated +8 AML. This registry includes data from more than 450 transplantation centers, which are required to fulfill all consecutive HSCT and follow-up.

Inclusion criteria

This was a retrospective multicenter study assessing the results of allo-HSCT in patients with de novo AML with either isolated +8 or +8 in addition to other cytogenetic abnormalities (designated as ‘associated’ +8 hereinafter), and reported to the EBMT registry between January 1990 and December 2007. The study was approved by the Acute Leukemia Working Party scientific committee of the EBMT and carried out according to EBMT guidelines.

Patients and transplant procedures

Patients, disease and transplant characteristics are summarized in Table 1.

Table 1 Patients, disease and transplant characteristics; n=182, period: 1990–2007

During the considered period, the total number of AML patients in the EBMT database (with available cytogenetics data) was 4157. One hundred and eighty-two AML patients with trisomy 8 (male n=100) were identified representing 4.4% of the total AML patients. Unrelated HSCT included both HLA-matched or -mismatched donor transplantation. As patients were included from 1990 to 2007, the definition of HLA compatibility between donors or recipients was HLA-A or -B by low-resolution typing and HLA-DRB1 by high-resolution typing, thus 6/6. Therefore, HLA mismatched was mostly 5/6 considering Ag level of HLA-A or -B or Ag or allelic level of DRB1. The median year of transplantation was 2000. Only four patients were transplanted in 2007 (two patients alive at 3 months). Unfortunately, we were not able to analyze the outcomes of patients with extramedullary or central nervous system involvement on this series of patients. Supportive care was carried out according to each center's guidelines.

Statistical analysis

Clinical outcomes that were studied were 5-year leukemia-free survival (LFS), relapse rate (RR), non-relapse mortality (NRM) and OS. Acute GVHD was diagnosed and graded at each transplant center according to the Seattle criteria.8 Therefore, only patients with grade II or superior were considered as having GHVD. Chronic GVHD was defined according to the standard criteria.9

Characteristics considered were recipient's age, sex, disease features (AML FAB subtype, cytogenetics, and remission status at the time of allo-HSCT), type of donor and allo-HSCT characteristics (GVHD prophylaxis and stem cell source). LFS was defined as survival without evidence of relapse or progression. Relapse was defined as any event related to the re-occurrence of the disease. NRM was defined as death from any cause without earlier relapse or progression. Starting on day 1, acute GVHD was scored according to the standard criteria and counted only for grades II. Patients surviving more than 100 days after transplant with sustained donor hematopoiesis were considered at risk for the development of chronic GVHD.

Cumulative incidence curves were used for RR and NRM in a competing risks setting.10, 11 For the estimation of GVHD, death was considered as a competing event. Probabilities of OS and LFS were calculated using the Kaplan–Meier method. The log-rank test was used for univariate comparisons. For all prognostic analyses, the median value of continuous variables was used as a cutoff point. Factors associated with a P value <0.15 by univariate analysis were included in the final model.

Associations of patient and graft characteristics with outcomes were evaluated in multivariate analyses, using Cox proportional hazards for LFS, and proportional sub-distribution hazard regression model of Fine and Gray10, 11 for other outcomes. The small number of patients and events in each group of cytogenetic abnormality (good, intermediate and poor-risk) of the associated +8 group does not allow us to make any further analysis. For this reason, we have compared isolated +8 patients (n=136) with an ‘all’ associated +8 group, including all patients bearing +8 and another cytogenetic abnormality (n=46). We did not include the three (good, intermediate and poor-risk) associated +8 groups independently as univariate analysis or in multivariable model, but just described the outcomes according to each group of cytogenetic abnormality in the +8 associated group.

All tests are two-sided. The type I error rate was fixed at 0.05 for determination of factors associated with time to event outcomes. Statistical analyses were carried out with the SPSS (Inc., Chicago, IL, USA) and Splus (MathSoft, Inc, Seattle, WA, USA) software packages.


Engraftment and GVHD

Engraftment was observed in 171 patients (95%). At day 100, cumulative incidence function of acute GVHD (grade II–IV) was 38±4% (Grade II n=39, 22%, and Grade III–IV n=20, 11%). For those patients transplanted with an HLA-identical donor (n=115), the incidence of acute GVHD (II–IV) was 40±5%, whereas it was 35±7% for those transplanted with other donors. Seventy-one patients presented with chronic GVHD (limited n=35; extensive n=36). Cumulative incidence of chronic GVHD at 2 years was 43±5% for patients at risk. For patients receiving a BM or a peripheral blood graft chronic GVHD was 41±6 and 45±6%, respectively.

Relapse rate

Overall, the 5-year RR was 30±3%. 5-year RR according to the cytogenetic abnormality were as follows: isolated +(8): 31%, n=43/136; +(8) associated to other abnormality of good risk: 12.5%, n=1/8; +(8) associated to other abnormality of intermediate risk: 23%, n=7/30, +(8) associated to other abnormality of high risk: 43%, n=3/7. When considering matched unrelated donor (MUD) (6/6) (n=41) and mismatch related donor (n=14) transplants, 5 year-RR was 36±8 vs 25±15%; P=0.57.

In univariate analysis (Table 2), the cumulative incidence of RR at 5 years was significantly higher for patients with active disease at transplantation: 51±7% for advanced patients at transplant compared with 20±4% for patients in CR1 and 28±10% for patients in CR2/CR3 at transplant (P<0.0001). In a multivariate analysis (Table 3), the disease status at transplant (active disease vs CR) was the only factor associated with RR after allogeneic transplant in this cohort.

Table 2 Univariate analysis of prognostic factors for outcomes 5 years after allo-HSCT in AML patients harboring +8
Table 3 Significant independent prognostic factors according to multivariate analysis of 182 AML patients with +8

Non-relapse mortality

The cumulative incidence function of 5-year NRM was 25±3%. Five-year NRM according to the cytogenetic abnormality were as follows: isolated +(8): 28%, n=38/136, +(8) associated to other abnormality of good risk: 0% (n=0/8), +(8) associated to other abnormality of intermediate risk: 16% (n=5/30), +(8) associated to other abnormality of high risk: 28% (n=2/7). When considering MUD (6/6) (n=41) and mismatch related donor (n=14) transplants, 5-year-NRM was 34±8 vs 53±15%; P=0.13.

In univariate analysis (Table 2), there was decreased NRM after allograft from an HLA-identical sibling donor: 18±4 vs 37±6% for other donor type (P=0.01). In a multivariate analysis (Table 3), the risk of mortality was statistically significantly higher for patients grafted with other than an HLA-identical sibling donor and for patients receiving a myeloablative compared with those receiving a reduced-intensity conditioning.

Overall and leukemia-free survivals

With a median follow-up of 48 (range, 3–180) months for survivors, the 5-year OS and LFS were 47±1 and 45±4%, respectively (Figures 1a and b). Median OS was 29 months. Overall, the 5-year OS was similar when comparing AML patients with isolated or ‘all’ associated +(8) (44±5%, n=136 vs 56±8%, n=46, P=0.14). Also, there was no significant difference when considering AML patients with isolated or ‘all’ associated +8 both in CR1 (60±6%, n=82, vs 55±11%, n=26, P=0.73).

Figure 1

(a) OS for the all cohort (n=182). (b) Leukemia-free survival for the all cohort (n=182).

The 5-year LFS was not statistically different between patients with isolated vs ‘all’ associated +8 (41±4 vs 55±8%, P=0.11), with no significant differences when considering those patients in CR1 (55±6%, n=82, vs 55±11%, n=26, P=0.9) (Table 2). When considering AML patients according to cytogenetic abnormality, 5-year LFS was higher in patients bearing +8 and other abnormality of good prognosis: 88% (n=8) vs 57% (n=30), intermediate risk vs 21% (n=7) poor-risk.

When grafted in CR1, 5-year OS and LFS were similar between isolated +8 AML patients (n=82) and AML patients bearing +8 and other abnormalities of intermediate + high-risk prognosis (n=24): 57±6 vs 48±12%, (P=0.43), and 55±6 vs 48±12%; P=0.66).

When considering MUD (6/6) (n=41) and mismatch related donor (n=14) transplants, 5-year LFS was 30±8 vs 22±13%; P=0.52.

In univariate analysis (Table 2), significantly higher 5-year LFS was observed for patients in CR at transplant (CR1: 55±5% vs CR2/CR3: 43±11% vs active disease: 24±6%, P<0.0001), with a female sex (56±5 vs 35±5%, P=0.02) and with an HLA-identical sibling donor (55±5 vs 28±6%, P=0.005).

Thus, 5-year LFS was increased when using an HLA-identical sibling donor up to 62±5 and 64±21% for patients in CR1 and CR2/CR3 at transplant, whereas 3-year LFS for patients with a MUD was 30±8% overall (n=41) and 35±13% in CR1 (n=18).

In multivariate analysis (Table 3) for LFS (including age, sex, disease status, karyotype (isolated or ‘all’ associated t(8)), donor type, type of conditioning regimen and use of T-cell depletion), disease status at transplant (CR vs others) (P<0.0001, HR=3.3, 95% confidence interval (CI), 2.2–5), recipient female sex (P=0.03, HR=1.61, 95% CI, 1.05–2.48) and a sibling donor (P=0.02, HR=1.64, 95% CI, 1.08–2.49) were the most significant predictors for improved LFS.


The effect of isolated +8 or +8 associated to other cytogenetic abnormalities in AML patients receiving allo-HSCT has not been described. We reported here the results of the largest series of allogeneic transplantation in patients with +8 AML published, so far. Although there is a bias inherent to all retrospective studies carried out within a registry database, the results are within the range of allo-HSCT results already reported for AML,12 and suggest that allo-HSCT is potentially an effective treatment for AML patients with +8, especially for younger patients as our population showed a median age of 37 years. Also our results, with a median OS of 29 months, compared favorably with those observed with series reporting outcomes of +8 AML patients after mixed consolidation regimen (chemotherapy, autologous or allogeneic transplantation) in which median survival did not exceed 16 months.1

Only four patients had t (15,17) and +8 in our series. We kept these four patients for the analyses, as the results did not change (for example, 5-year LFS was 43% excluding those four patients).

As already known,12 CR status at transplant was required to achieve better LFS after allogeneic transplantation in this series, indicating the pejorative effect of a higher residual tumor burden. It is interesting that 5-year LFS was increased when using an HLA-identical sibling donor up to 62 and 64% for patients in CR1 and CR2/CR3 at transplant, respectively. This could be explained by a significant lower NRM incidence when using such a type of donor as shown by the multivariate analysis. Results showed here with unrelated donor are poor but it has to be kept in mind that HLA typing determination used during our inclusion period was not optimal for all patients. With determination at the allelic level, it has been shown that transplantation from MUD (10/10) led to outcomes similar to those from HLA-identical sibling donors.13 Then, it should be hypothesized that now-a-days allo-HSCT for AML +8 patients transplanted in CR1 or CR2/3 and carried out with an allelic-MUD may permit the same outcome as observed with a sibling donor.

The size of this study allowed considering independently the outcome of both isolated and associated +8 after allo-HSCT. Overall, patients with isolated +(8) or +(8) associated with other abnormality of intermediate prognosis seem to share similar outcomes. On the other hand, as already reported, patients bearing +8 in addition to another abnormality from the ‘good-risk’ or the ‘high-risk’ groups may have the prognosis conferred by the accompanying cytogenetic aberration.2 Nevertheless, because of the small number of patients involved, these results have to be taken with caution.

Thus far, only two studies addressed the outcome of allogeneic transplantation in patients with AML and +8. However, the number of patients grafted in both studies was lower than 20 patients, not allowing meaningful conclusions as to the true effect of allo-HSCT in AML with +8. Recently, Schaich et al.5 reported a series of 131 AML patients aged under 60 years and exhibiting +8 within a non-complex karyotype: the 3-year OS and LFS were 29 and 29%, respectively with no differences between patients with +8 as a sole aberration and those with +8 and one additional cytogenetic aberration. In the latter study, allo-HSCT carried out in 19 patients was an independent prognostic factor for longer LFS (49% at 3 years vs 23% for patients who underwent autologous transplantation vs 28% for patients who received chemotherapy alone, P<0.05).

In the series of Farag et al.7 combining 101 patients with isolated +8, +11, +13 and +21, the OS was around 10% at 5 years and allo-HSCT significantly improved LFS. Among patients under 60 years in CR1, only one out of seven relapsed after allo-HSCT as compared with 16/19 patients treated with chemotherapy alone. The 5-years OS for patients with isolated trisomy was not significantly different from that of AML patients with normal karyotype receiving allo-HSCT (69 vs 60%), suggesting that AML with +8 shares a similar intermediate prognosis. In addition to this observation, we have looked at the LFS after allo-HSCT for patients with AML with normal karyotype transplanted in EBMT centers in the same period of this study. LFS was 43±1% for 2956 patients. Of course, although this result is in favor of the hypothesis that isolated +8 share similar results of patients with normal karyotype, only retrospective comparisons with statistical adjustments for potential confounding factors between the two populations can confirm this speculation. Moreover, AML patients with normal karyotype represent a heterogeneous group including a variety of gene amplifications or mutations providing different outcomes. It is now established that patients with +8 and normal karyotype share at least the same proportion of fms-like tyrosine kinase 3 internal tandem duplication (30%), one of the most frequent molecular abnormalities detected in AML;14 however, one should highlight that data are not yet available regarding the association between +8 and the NPM1 mutation, which is found in 50% of AML patients with normal karyotype. It is interesting that, Thiede et al showed that NPM1 mutations were mainly associated in the group of AML patients with aberrant karyotype with single genetic abnormalities like trisomy and especially trisomy 8.15

From the molecular standpoint, gain of chromosome 8 does not seem to confer a particularly bad prognosis, thus, showing no evidence for a functional or molecular effect of this abnormality. Indeed, the pathogenic effect of +8 remains elusive. It is rather accepted that +8 is an important event, but not sufficient for leukemogenesis, as several studies have indicated that +8 is not the primary event involved in malignant transformation.2 It is also admitted that the Myc gene located at 8q24 is not a target of +8 as Myc in +8 AML is down-regulated.16 Moreover, other data indicated that +8 AML showed no specific gene expression signature.2

In conclusion, our retrospective data show that allo-HSCT is an effective treatment for AML patients harboring +8. The addition of other chromosomal abnormalities to +8 seems to influence the outcomes of these patients compared with AML patients with isolated +8.


  1. 1

    Wolman SR, Gundacker H, Appelbaum FR, Slovak ML . Impact of trisomy 8 (+8) on clinical presentation, treatment response, and survival in acute myeloid leukemia: a Southwest Oncology Group Study. Blood 2002; 100: 29–35.

  2. 2

    Paulsson K, Johansson B . Trisomy 8 as the sole chromosomal aberration in acute myeloid leukaemia and myelodysplastic syndromes. Pathol Biol 2007; 55: 37–48.

  3. 3

    Jaff N, Chelghoum Y, Elhamri M, Tigaud I, Michallet M, Thomas X . Trisomy 8 as a sole anomaly or with other clonal aberrations in acute myeloid leukemia: impact on clinical presentation and outcome. Leuk Res 2007; 31: 67–73.

  4. 4

    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 1612 patients entered into the MRC AML 10 Trial. Blood 1998; 92: 2322–2333.

  5. 5

    Schaich M, Schlenk RF, Al-Ali HK, Döhner H, Ganser A, Heil G et al. Prognosis of acute myeloid leukaemia patients up to 60 years of age exhibiting trisomy 8 within a non-complex karyotype: individual patient data-based meta-analysis of the German Acute Myeloid Leukemia Intergroup. Haematologica 2007; 92: 763–770.

  6. 6

    Schoch C, Haase D, Fonatsch C, Haferlach T, Loffler H, Schlegelberger B et al. The significance of trisomy 8 in de novo acute myeloid leukaemia: the accompanying chromosome aberrations determine the prognosis. Br J Haematol 1997; 99: 605–611.

  7. 7

    Farag SS, Archer KJ, Mrozek K, Vardiman JW, Caroll AJ, Pettenati MJ et al. Isolated trisomy of chromosomes 8, 11, 13 and 21 is an adverse prognostic factor in adults with de novo acute myeloid leukaemia: results from cancer and leukemia group B 8461. Int J Oncol 2002; 21: 1041–1051.

  8. 8

    Glucksberg H, Storb R, Fefer A, Buckner CD, Neiman PE, Clift RA et al. Clinical manifestations of graft-versus-host disease in human recipients of marrow from HLA-matched sibling donors. Transplantation 1974; 18: 295–304.

  9. 9

    Shulman H, Sullivan K, Weuiden P . Chronic graft-versus-host syndrome in man. A long-term clinicopathologic study of 20 Seattle patients. Am J Med 1980; 69: 204–217.

  10. 10

    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.

  11. 11

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

  12. 12

    Lowenberg B, Downing JR, Burnett A . Acute myeloid leukemia. N Engl J Med 1999; 341: 1051–1062.

  13. 13

    Yacoub-Agha I, Mesnil F, Kuentz M, Boiron JM, Ifrah N, Milpied N et al. 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.

  14. 14

    Giles FJ, Keating A, Goldstone AH, Avivi I, Willman CL, Kantarjian HM . Acute Myeloid Leukemia: Biologic and genetic risk assessment of AML in the genomic era by Willman CL. Hematology Am Soc Hematol Educ Program 2002; 73–110.

  15. 15

    Thiede C, Koch S, Creutzig E, Steudel C, Illmer T, Schaich M et al. Prevalence and prognostic impact of NPM1 mutations in 1485 adult patients with acute myeloid leukemia (AML). Blood 2006; 107: 4011–4020.

  16. 16

    Virtaneva K, Wright FA, Tanner S, Yuan B, Lemon WJ, Caligiuri MA et al. Expression profiling reveals fundamental biological differences in acute myeloid leukemia with isolated trisomy 8 and normal cytogenetics. Proc Natl Acad Sci USA 2001; 98: 1124–1129.

Download references


P Chevallier conceived and designed the study, analyzed the data, recruited the patients, provided clinical care, performed bibliographic search and wrote the paper. M Mohty and V Rocha recruited patients, provided clinical care, analyzed the data and performed bibliographic search, helped with the statistical analyses and in writing the paper. M Labopin performed data management, collection and statistical analyses. A Nagler, P Ljungman, L Verdonck, T Ruutu, A Zander, J Finke, G Socie, C Cordonnier and JL Harousseau recruited the patients, provided clinical care and commented on the paper. All the above authors approved the paper for publication purposes.

Author information

Correspondence to P Chevallier.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Chevallier, P., Labopin, M., Nagler, A. et al. Outcome after allogeneic transplantation for adult acute myeloid leukemia patients exhibiting isolated or associated trisomy 8 chromosomal abnormality: a survey on behalf of the ALWP of the EBMT. Bone Marrow Transplant 44, 589–594 (2009) doi:10.1038/bmt.2009.68

Download citation


  • allogeneic HSCT
  • AML
  • trisomy 8

Further reading