A new Leukemia Prognostic Scoring System for refractory ... - Nature

Leukemia (2011) 25, 939–944 & 2011 Macmillan Publishers Limited All rights reserved 0887-6924/11 www.nature.com/leu

ORIGINAL ARTICLE A new Leukemia Prognostic Scoring System for refractory/relapsed adult acute myelogeneous leukaemia patients: a GOELAMS study P Chevallier1, M Labopin2, P Turlure3, T Prebet4, A Pigneux5, M Hunault6, K Filanovsky7, P Cornillet-Lefebvre8, I Luquet8, L Lode9, S Richebourg9, O Blanchet10, N Gachard11, N Vey4, N Ifrah6, N Milpied5, J-L Harousseau1, M-C Bene12, M Mohty1 and J Delaunay1 1 Service d’He´matologie Clinique, CHU and Universite´ de Nantes, INSERM U892, Nantes, France; 2ALWP, EBMT-Paris Office, Hoˆpital Saint Antoine AP-HP, Universite´ Pierre et Marie Curie Paris 6, Paris, France; 3Service d’He´matologie Clinique, CHU, Limoges, France; 4De´partement d’He´matologie, Institut Paoli-Calmettes, Marseille, France; 5Service d’He´matologie Clinique, CHU, Bordeaux, France; 6Service d’He´matologie Clinique, CHU, Angers, France; 7Department of Hematology, Kaplan Medical Center, Rehovot, Israel; 8Service de Biologie/He´matologie, CHU, Reims, France; 9Service de Biologie/He´matologie, CHU, Nantes, France; 10Servive de Biologie/He´matologie, CHU, Angers, France; 11Service de Biologie/He´matologie, CHU, Limoges, France and 12Service de Biologie/He´matologie, CHU, Nancy, France

A simplified prognostic score is presented based on the multivariate analysis of 138 refractory/relapsed acute myeloid leukaemia (AML) patients (median age 55 years, range: 19–70) receiving a combination of intensive chemotherapy þ Gemtuzumab as salvage regimen. Overall, 2-year event-free survival (EFS) and overall survival (OS) were 29±4% and 36±4%, respectively. Disease status (relapse o12 months, including refractory patients), FLT3-ITD-positive status and high-risk cytogenetics were the three strongest independent adverse prognostic factors for OS and EFS in this series. We then defined three subgroups with striking different outcomes at 2 years: no adverse factor (favourable, N ¼ 36): OS 58%, EFS 45%; one adverse factor (intermediate, N ¼ 54): OS 37%, EFS 31%; two or three adverse factors (poor, N ¼ 43): OS 12%, EFS 12% (Po104, P ¼ 0.001). This new simplified Leukemia Prognostic Scoring System was then validated on an independent cohort of 111 refractory/relapsed AML patients. This new simplified prognostic score, using three clinical and biological parameters routinely applied, allow to discriminate around two third of the patients who should benefit from a salvage intensive regimen in the setting of refractory/relapsed AML patients. The other one third of the patients should receive investigational therapy. Leukemia (2011) 25, 939–944; doi:10.1038/leu.2011.25; published online 18 February 2011 Keywords: AML; refractory; relapse; score; NPM1; FLT3

Ozogamicin (GO), intermediate dose Ara-C and Mitoxantrone (MIDAM regimen)7 represents a valid therapeutic option in this setting, allowing to achieve an overall response (OR) rate of 63%, and a disease-free and OS rates of 41 and 53%, respectively, at 2 years. Moreover, we have also shown that normal karyotype (NK) AML patients, receiving the MIDAM protocol and who had a favourable molecular profile (that is, mutated nucleophosmin gene (NPM1 þ ) without internal tandem duplications of the fms-like tyrosine kinase 3 gene (FLT3-ITD)), retained a good prognosis compared with other subgroups with NK.8 If well-known prognostic factors are identified for refractory/ relapsed AML outcomes such as age, performans status, karyotype, duration of first CR or previous stem cell transplant,5 currently, there is no series having studied the prognostic impact of the new molecular parameters compared with others factors in this setting. The aim of this study was to define a predictive score integrating the new molecular parameters in the setting of refractory/relapsed AML patients in order to propose appropriate therapeutic strategies to these patients.

Patients and methods

Study design and patients selection criteria Introduction Relapsed or refractory acute myeloid leukaemia (AML) patients have still a very dismal outcome.1–3 Patients who could proceed to an allogeneic transplantation, either after a sequential conditioning regimen in case of refractory disease4 or after a standard or reduced intensity conditioning regimen in case of second complete remission (CR),3,5 are likely to have the best chances to achieve long-term overall survival (OS). At present, there is no widely approved salvage regimen for refractory/relapsed AML.6 However, we have recently shown that a chemo-immunotherapy combining Gemtuzumab Correspondence: Dr P Chevallier, Service d’He´matologie Clinique, Centre Hospitalier Universitaire de Nantes, Place Alexis Ricordeau, 44093 Nantes Cedex 01, France. E-mail: [email protected] Received 16 November 2010; revised 23 December 2010; accepted 10 January 2011; published online 18 February 2011

This was a retrospective multicenter study which included 138 adult patients (female n ¼ 75; male n ¼ 63) with refractory (n ¼ 57) or relapsed (n ¼ 81) AML and treated in six institutions (Nantes, Limoges, Marseille, Bordeaux, Angers and Rehovot) in France and Israel, between 2001 and 2009. The patients were treated in different therapeutic trials or programs at the corresponding institution. Written informed consent was obtained from each patient. The studies were approved by the local ethical committee and performed according to institutional guidelines. No selection criteria other than ‘relapsed/refractory AML treated with GO and intensive chemotherapy as salvage regimen’ was used for the purpose of this analysis. The outcome of previously published patients7–9 was updated as of March 2010. Patients o60 years of age with no CR after a double induction chemotherapy (including one with intermediate or high dose Ara-C), patients over 60 years of age with no CR after one induction chemotherapy and patients who relapsed within 6 months after CR were considered as refractory AML.7 The median age in this cohort was 55 (range, 19–70) years old

Prognostic score for refractory/relapsed AML P Chevallier et al

940 Table 1

Characteristics of patients

Patients Sex: female/male Median age (range)

N ¼ 138 75 (54%)/63 (46%) 55 (19–70)

Median leukocytes counts (range)

10.8 109/L (1–294)

AML: de novo/secondary

117 (85%)/21 (15%)

FAB 0/1/2/4/5/6/7 Unknown Median CD33 expression (range)

9/26/36/25/21/4/4 13 98% (40–100)

Karyotype Favourable t(8;21) inv (16)

10 5 5

Intermediate Normal karyotype Others

101 84 17

High risk

27

Molecular status NPM1+ NPM1 Unknown

37 96 5

FLT3-ITD+ FLT3-ITD Unknown

37 96 5

NPM1+/FLT3-ITD NPM1+/FLT3-ITD+ NPM1/FLT3-ITD+ NPM1/FLT3-ITD Unknown

23 14 23 73 5

Status Refractory Relapse 6–12 months Relapse X12 months

57 (41%) 25 (18%) 56 (41%)

Previous autograft Previous allograft

21 (15%) 11 (8%)

Salvage regimen MIDAM Intensive regimen+GO

128 (93%) 10 (7%)

Abbreviations: AML, acute myeloid leukaemia; FAB, French-AmericanBritish classification; GO, Gemtuzumab Ozogamicin.

at time of AML diagnosis. All patients had CD33-positive blasts (median: 98% CD33 þ blasts). As salvage regimen, 128 patients received the previously published MIDAM regimen (GO: 9 mg/m2 at day 4 þ Ara-C 1 g/m2/12 h at days 1–5 þ Mitoxantrone 12 mg/m2/day at days 1–3).7 The remaining 10 patients received other intensive GO-based combinations (Ara-C, VP16 and GO: n ¼ 3; Ara-C, Idarubicine and GO: n ¼ 2; Ara-C, Amsacrine and GO: n ¼ 2; Ara-C and GO: n ¼ 3). In this series, all but five patients could be screened at time of diagnosis in peripheral blood or bone marrow samples for internal tandem duplication in the FLT3 gene (FLT3-ITD) and mutation of the NPM1 gene as previously described.10,11 CEBPa mutations screening was not investigated in this study because of insufficient material. Cytogenetics data at diagnosis were Leukemia

available for all patients. Patients’ characteristics are summarized in Table 1. To determine its impact in the setting of refractory/relapsed AML, patients from this cohort were also classified according to the new European Leukemia Net (ELN) prognosis subsets, proposed recently by an expert panel.3 This standardized report for the correlation of cytogenetics and molecular data in AML patients at diagnosis integrates the favourable molecular profile (NPM1 þ /FLT3-ITD) in the previously defined favourable cytogenetics group and separates the previous intermediate cytogenetics group in two subgroups: the intermediate-1 subgroup which includes all NK-AML except those with the favourable molecular profile while the intermediate-2 subgroup integrates patients with t(9;11) or cytogenetic abnormalities not classified as favourable or adverse. Two patients with an NK and unknown molecular status were classified in the intermediate-1 group.

Endpoints definition CR was defined as normalization of peripheral blood counts (neutrophils 41 109/l and platelets 4100 109/l) and bone marrow with p5% blasts after salvage therapy, CR with incomplete platelets recovery (CRp) was defined like CR but with platelet transfusion independence and platelet counts remaining o100 109/l. CR þ CRp was defined as OR. All other types of responses were considered as failure. Survivals were calculated from the date of the documentation of relapse or refractory disease until death or to last follow-up for OS, or to failure, relapse, death or last follow-up for event-free survival (EFS).

Statistical analysis Descriptive statistics are reported as frequencies, or medians and range. The w2 test was used to test for factors associated with OR. OS and EFS were estimated by the Kaplan–Meier product limit method. The log-rank test was used for univariate comparisons. The following parameters were taken into account for analyses of prognostic factors for survivals: age, sex, FrenchAmerican-British classification (M5 M6 M7 vs others), type of AML (de novo or secondary), cytogenetics and molecular status, status at time of salvage regimen (relapsed or refractory AML), duration of CR1 for relapsed patients (46 months and p12 months; 412 months), previous stem cell transplantation, type of salvage regimen (MIDAM vs other). Associations of patient and disease characteristics with outcomes are evaluated in multivariate analyses, using Cox proportional hazards for OS, and logistic regression for response rate. All factors associated with a P-value o0.20 by univariate analysis were forced in multivariate analyses, except the NPM1 molecular status as it was strictly associated with intermediate cytogenetics and French-American-British classification as it was also associated with cytogenetics (P ¼ 0.03). Adjusted hazard ratios (HRs) were calculated for each factor. Integer weights for the construction of the score were derived from Cox proportional hazards modeling with OS over the first 2 years as the outcome. Adjusted HRs were converted to integer weights according to the following: factors with adjusted HR of p1.4 were dropped from consideration, factors with adjusted HR of 1.5–2.0 were assigned a weight of 1. The score was the sum of these integer weights. Because of small numbers, the original score was reduced to three risk groups: 0 (low risk), 1 (intermediate risk) and 2–3 (high risk). The scores were applied to patients of the validation set to test their ability to predict for HR of 2-year survival. The final analysis was performed as of end of July 2010 using SPSS Inc., 18.0 (Chicago, IL, USA).


941 Results

Classification of patients and outcome according to the ELN subgroups Thirty-one (22%), 63 (46%), 17 (12%) and 27 (20%) patients were classified in the favourable, intermediate-1, intermediate-2 and high-risk group, respectively. Considering the favourable group, 20 patients had an NK with a favourable molecular status (NPM1 þ /FLT3-ITD) and 10 patients had t(8;21) or inv(16), including 3 patients NPM1/FLT3-ITD þ , 6 patients NPM1/ FLT3-ITD (unknown molecular status n ¼ 2). Considering the intermediate-1 group (all with NK by definition), 14, 30 and 18 patients had the following molecular status: NPM1 þ /FLT3ITD þ , NPM1/FLT3-ITD and NPM1/FLT3-ITD þ , respectively (unknown molecular status n ¼ 1). Considering the intermediate-2 group, 13 patients were NPM1/FLT3-ITD and 3 patients were NPM1 þ /FLT3-ITD (unknown molecular status n ¼ 1). Finally, considering the high-risk group, 24 patients were NPM1/FLT3-ITD and 2 patients were NPM1/ FLT3-ITD þ (unknown molecular status n ¼ 1). The 2-year EFS for the four groups was as follows: favourable 55±9%, intermediate-1 24±6%, intermediate-2 28±12% and high-risk 19±7%, P ¼ 0.03. The 2-year OS for the four groups was as follows: favourable 59±9%, intermediate-1 30±6%, intermediate-2 44±14% and high risk 19±9%, P ¼ 0.02. When considering the 82 patients with NK and known molecular status, 2-year EFS and OS were significantly higher in cases with the favourable molecular status (NPM1 þ /FLT3-ITD) vs others status: 50±12% vs 22±6%, P ¼ 0.007; 62±11% vs 28±6%, P ¼ 0.008. Two-year EFS and OS were comparable in patients with favourable cytogenetics and patients with NK and favourable molecular status (P ¼ 0.49).

OR, EFS and OS In this analysis, the median follow-up for the whole cohort was 18 months (range, 1–89). The OR rate was 64% (N ¼ 88) including 78 patients achieving CR and 10 patients with CRp. In univariate analysis, factors associated with a higher OR were disease status (refractory: 49% vs relapse between 6 and 12 months: 68% vs relapse 412 months: 77%, P ¼ 0.008), ELN classification (favourable: 77%; intermediate-1: 59%, intermediate-2: 88%, adverse: 44%, P ¼ 0.008), type of AML (de novo: 69% vs secondary: 33%), NPM1 status ( þ : 78% vs : 58%, P ¼ 0.03) and cytogenetics (good risk: 77% vs intermediate risk: 65% vs poor risk: 44%, P ¼ 0.04). In multivariate analysis, independent factors associated with OR were disease status (relative risk: 0.34, 95% confidence interval: 0.15–0.79, P ¼ 0.01), type of AML (relative risk: 0.24, 95% confidence interval: 0.08–0.71, P ¼ 0.01) and NPM1 status (relative risk: 2.78, 95% confidence interval: 1.08–7.13, P ¼ 0.03). Two-years EFS and OS were 29±4% and 36±4%, respectively. In univariate analysis, prognostic factors associated with a significantly improved EFS and OS are listed in Table 2. Among 88 patients achieving CR þ CRp, 42 relapsed (47.7%) during follow-up. At last follow-up, 94 patients (68%) died with leukaemia being the cause of death in 70 cases. The other causes of deaths included sepsis (n ¼ 8), graft-versus-host disease (acute n ¼ 1; chronic n ¼ 4), veno-occlusive disease (n ¼ 5), multiple organ failure (n ¼ 2), haemorrhage (n ¼ 3) or secondary malignancy (n ¼ 1; renal carcinoma). Considering the 94 patients who died, 10 have died early after the salvage regimen because of toxicity (veno-occlusive disease n ¼ 5; haemorrhage n ¼ 2; sepsis n ¼ 2 and multiple organ failure n ¼ 1).

Table 2

Univariate analyses

Overall population Sex Male Female

2-year EFS

P-value 2-year OS P-value

29±4%

36±4%

33±6% 25±5%

0.35

39±7% 33±6%

0.44

Age o60 years X60 years

29±5% 30±7%

0.86

34±5% 40±8%

0.50

FAB M5 M6 M7 Others

11±9% 33±5%

0.06

11±9% 40±5%

0.03

Type of AML De novo AML Secondary AML

33±5% 10±6%

0.003

40±5% 15±8%

0.03

Status Refractory Relapse 6–12 months Relapse X12 months

18±6% 15±8% 44±7%

0.02

25±7% 19±9% 53±7%

0.02

Cytogenetic Favourable Intermediate High risk

50±16% 30±5% 19±8%

0.08

50±16% 39±5% 19±9%

0.09

Molecular status FLT3-ITD+ FLT3-ITD

21±7% 32±5%

0.16

23±7% 40±5%

0.03

Age o60 FLT3-ITD+ FLT3-ITD

17±7% 32±6%

0.02

17±8% 38±7%

0.003

Age X60 FLT3-ITD+ FLT3-ITD

29±14% 31±8%

0.65

35±15% 43±10%

0.99

FLT3-ITD+a FLT3-ITDa

18±7% 36±7%

0.07

20±8% 47±8%

0.02

NPM1+ NPM1

37±8% 26±5%

0.09

46±9% 31±5%

0.09

NPM1+a NPM1a

35±9% 25±7%

0.08

42±9% 33±8%

0.2

ELN classification Favourable Intermediate-1 Intermediate-2 Adverse

55±9% 24±6% 28±12% 19±7%

0.03

59±9% 30±6% 44±14% 19±9%

0.02

0.9

43±11% 34±5%

0.83

0.88

36±5% 27±15%

0.4

Previous stem cell transplantation Yes 33±10% No 28±5% MIDAM Other regimen

29±4% 30±14%

Abbreviations: AML, acute myeloid leukaemia; EFS, event-free survival; ELN, European Leukemia Net; FAB, French-American-British AML classification; OS, overall survival. Including refractory AML patients+patients in relapse between 6 and 12 months. a Considering only AML patients with normal karyotype.

After salvage therapy, 47 patients underwent an allogeneic stem cell transplant (myeloablative conditioning: n ¼ 1; reduced intensity conditioning regimen: n ¼ 46). The majority of patients Leukemia


942 (n ¼ 40) received the allo-transplant because they achieved CR after the salvage treatment while seven unresponsive patients received the allo-transplant because of the availability of a human leukocyte antigen-matched identical donor and good performans status without major comorbidities. The 2-year OS for these patients was 45±9%.

prospective GOELAMS LAM-2001 trial received various intensive chemotherapies as salvage regimen. The HR and 2-year OS are summarized by scores for the training and validation sets in Table 5.

Discussion

Prognostic score for relapsed/refractory adult AML patients The aim of this study was to develop a predictive scoring system in the setting of refractory/relapsed AML patients integrating new molecular prognostic parameters. Thus, we applied multiple multivariate Cox regression analyses to detect independent prognostic factors for OS. Finally, three independent prognostic factors were significantly associated with higher OS (Table 3): the molecular status (FLT3ITD vs þ ), cytogenetics (favourable and intermediate vs poor) and the disease status (relapse 412 months vs relapse between 6 and 12 months and refractory patients). Three prognostic groups (favourable (0 adverse prognostic factor), intermediate (1 adverse prognostic factor) and poor (2 or 3 adverse prognostic factors)) could then be identified for OS as described in Table 4. OS and EFS curves according to the three subgroups are shown in Figures 1 and 2. Given the striking differences obtained when using the above scoring system, we thought to test its validity on an independent cohort of 111 refractory/relapsed AML patients initially included in the prospective GOELAMS LAM-2001 trial12,13 as first-line therapy. Status disease (refractory, relapse between 6 and 12 months and relapse 412 months), cytogenetics and molecular status were available for all 111 patients. Patients from this Table 3

Hazard ratio

95% confidence interval

P

1.95

1.2–3.17

0.007

0.8

0.6

0.4

1.86 1.75

1.15–3.03 1.03–2.98

0.012 0.04

0.0

Abbreviation: CR, complete remission. a Including refractory patients and patients in relapse between 6 and 12 months from CR1. b High risk vs favourable+intermediate. Table 4

1.0

0.2

Multivariate analyses for overall survival

Relapse p12 monthsa vs 412 months FLT3+ vs FLT3 Cytogeneticb

This is the first large series reporting on the outcome of refractory/relapsed AML patients taking into account, beside well-recognized prognostic factors, such as cytogenetics, disease status or duration of first CR, the new molecular parameters, that is, NPM1 mutation and FLT3-ITD. We confirm here that patients with NK and favourable molecular status have a very good prognosis,8 sharing similar EFS and OS compared with patients with favourable cytogenetics. These results are in favour of the fusion of this two subgroups in a same entity of good prognosis in the setting of refractory/relapsed AML,

0

1

Years

2

3

Figure 1 Two-year OS according to the three prognostic subgroups: favourable (0 adverse factor, N ¼ 36) 59%, intermediate (1 adverse factor, N ¼ 54) 37%, high risk (2 or 3 adverse factors, N ¼ 43) 12%. Factors: relapse o12 months, including refractory patients ; FLT3ITD þ status or high-risk cytogenetics. Abscissa: years. Ordinate: % of survival.

Characteristics of the three prognostic groups

Number of patients Relapse o12 monthsa FLT3-ITD+ High-risk cytogenetics Allograft post-GO OR after GO Two-year EFS Two-year OS

Group 1 Favourable No adverse factor

Group 2 Intermediate One adverse factor

Group 3 High risk Two or three adverse factors

36 0 0 0 14 (39%) 26 (72%) 45±9% 58±8%

54 36 (67%) 11 (20%) 7 (13%) 17 (31%) 41 (76%) 31±7% 38±7%

43 43 (100%) 26 (60%) 19 (44%) 14 (33%) 18 (42%) 12±5% 12±6%

P

0.75 0.001 0.001 o104

Abbreviations: CR, complete remission; EFS, event-free survival; GO, Gemtuzumab Ozogamicin; OR, overall response (CR+CRp); OS, overall survival. The total number of patients is 133, as FLT3-ITD status was unknown in five patients. Factor: relapse o12 months, including refractory patients; FLT3-ITD+ status or high-risk cytogenetics. a Including refractory patients and patients in relapse between 6 and 12 months from CR1. Leukemia


943 as recommended already at diagnosis by the new prognostic classification proposed recently by an international expert panel.3 Also, we were able to demonstrate for the first time, in this setting, that FLT3-ITD was indeed associated with adverse EFS and OS. The negative impact of FLT3-ITD has previously been demonstrated on first-line therapy.3 Following the ELN classification, patients in the intermediate-1 subgroup, where FLT3-ITD þ cases were concentrated (86% of FLT3-ITD þ patients in this series) shared similar outcomes compared with patients classified in the adverse group, demonstrating the strong negative impact of FLT3-ITD. Thus, if, as documented here, the new ELN classification kept its prognostic significance in the setting of refractory/relapsed AML, in term of EFS and OS, there might be a need to improve this prognosis index regarding to FLT3-ITD. We were then interested by the elaboration of a new prognostic score integrating the new molecular parameters. As NPM1 þ status was strictly associated with intermediate cytogenetic, we made the choice to not consider this marker in multivariate analyses. Thus, the independent variables with

1.0

0.8

0.6

0.4

0.2

0.0 0

1

2

3

Years Figure 2 Two-year EFS according to the three prognostic subgroups: favourable (0 adverse factor, N ¼ 36) 46%, intermediate (1 adverse factor, N ¼ 54) 30%, high risk (2 or 3 adverse factors, N ¼ 43) 12%. Factors: relapse o12 months, including refractory patients ; FLT3ITD þ status or high-risk cytogenetics. Abscissa: years. Ordinate: % of survival. Table 5

the highest impact on OS were the cytogenetic (combining patients with favourable and intermediate cytogenetic because of similar outcomes vs high risk), the duration of first CR (combining refractory patients þ patients in relapse between 6 and 12 months because of similar outcomes vs patients in relapse 412 months) and still the FLT3-ITD status. The two first parameters can be easily documented, but investigation of the FLT3-ITD status is usually performed only in patients with NK. This study pleads towards analysing FLT3-ITD in all AML patients at diagnosis or relapse, especially those with favourable cytogenetics. Because of the characterization of these three strong prognostic factors for OS, we were able to establish a new simplified prognostic score for refractory/relapsed AML patients, dubbed the Leukemia Progression Scoring System. Currently, and as recommended by the international expert panel,3 the score of Breems et al,5 including four parameters (length of relapse-free interval after CR1, cytogenetics at diagnosis, age at relapse and previous stem cell transplantation), validated by Giles et al,14 should be considered for all studies with relapsed AML. Nevertheless, this study published in 2005 has several limitations regarding new recent advances in the setting AML. For example, it did not include patients 460 years old, or with refractory disease and especially new molecular parameters. Also, our score was validated on an independent cohort of patients receiving chemotherapy as salvage regimen, demonstrating the pertinence of the Leukemia Progression Scoring System. This is of crucial importance as the US Food and Drug Administration has recently announced the withdrawal of the GO because of concerns about the product’s safety and the clinical benefit to patients enrolled in trials. Although we are still convinced, with others,15,16 that GO may have some interest in particular subgroups of AML patients, it seems that our new prognostic method for refractory/relapsed AML is broadly applicable to all intensive regimens, whether or not they contain GO. Finally, it seems that the new prognostic score proposed here could be more adapted to predict outcome and tailor therapeutic strategies. Thus, patients in the favourable/intermediate groups (0 or 1 adverse prognostic factor) should be considered for an intensive salvage regimen followed by an allogeneic transplantation in case of available donor, while patients in the adverse group (2 or 3 adverse prognostic factors) should be considered for investigational therapies, for example an anti-FLT3 inhibitor in case of FLT3-ITD17 or hypomethylating agents.18 In conclusion, despite its retrospective feature, this series allowed to establish a new simplified prognostic score for refractory/relapsed AML patients including three clinical and biological parameters currently routinely applied. Although this was validated here by means of an independent cohort, retrospective and prospective studies are still needed to assess the reproducibility of this new prognostic score for refractory/ relapsed AML patients.

New score and prediction for 2-year overall survival (2y OS): study vs validation score Study group, N ¼ 133

Score

Good ¼ 0 factor Intermediate ¼ 1 factor Poor ¼ 2–3 factors

Validation group, N ¼ 111

HRs

95% CI

2y OS

HRs

95% CI

2y OS

1 1.34 3.53

0.73–2.55 1.9–6.5

58±8% 38±7% 12±6%

1 1.44 3.2

0.83–2.49 1.84–5.57

35±8% 23±6% 0%

Abbreviations: CI, confidence interval; HRs, hazard ratios. Factors: relapse o12 months, including refractory patients; FLT3-ITD+ status or high-risk cytogenetics. Leukemia


944 Conflict of interest The authors declare no conflict of interest.

Acknowledgements We would like to acknowledge the continuous support of the cell banking facility (‘tumurotheque’) of the CHU de Nantes, Angers, Limoges, Bordeaux, Marseille, Reims to their ongoing basic and clinical research work. We also thank all the data managers of the CHU de Nantes, Angers, Limoges, Bordeaux, Marseille, Reims, Rehovot and of the GOELAMS.

Author contributions PC conceived and designed the study, analysed data, recruited patients, provided clinical care, performed bibliographic search and wrote the manuscript. JD recruited patients, provided clinical care, analysed data, performed the statistical analyses and helped in writing the manuscript. ML performed the statistical analyses. MM recruited patients, provided clinical care and helped in writing the manuscript. PT, TP, AP, MH, KF, IL, PCL, LL, SR, OB, NG, NV, NI, NM, JLH and MCB recruited patients, provided clinical care or performed biological analyses and commented on the manuscript. All the above authors approved the manuscript for publication purposes.

References 1 Lo¨wenberg B, Downing J, Burnett A. Acute myeloid leukemia, review article. N Engl J Med 1999; 341: 1051–1062. 2 Milligan DW, Grimwade D, Cullis JO, Bond L, Swirsky C, Craddock C et al. Guidelines on the management of acute myeloid leukaemia in adults. Br J Haematol 2006; 135: 450–474. 3 Dohner H, Estey EH, Amadori S, Appelbaum FR, Buchner T, Burnett AK et al. Diagnosis and management of acute myeloid leukemia in adults: recommendations from an international expert panel, on behalf of the European Leukemia Net. Blood 2010; 115: 453–474. 4 Schmid C, Schleuning M, Schwerdtfeger R, Hertenstein B, Mischak-Weissinberg 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. 5 Breems DA, Van Putten W, Huijgens PC, Ossenkoppele GJ, Verhoef GE, Verdonck LF et al. Prognosis index for adult patients with acute myeloid leukemia in first relapse. J Clin Oncol 2005; 23: 1969–1978. 6 Craddock C, Tauro S, Moss P, Grimwade D. Biology and management of relapsed acute myeloid leukaemia. Br J Haematol 2005; 129: 18–34.

Leukemia

7 Chevallier P, Delaunay J, Turlure P, Pigneux A, Hunault M, Garand R et al. Long-term disease-free survival after Gemtuzumab, intermediate-dose cytarabine and Mitoxantrone in patients with CD33+ primary resistant or relapsed acute myeloid leukaemia. J Clin Oncol 2008; 26: 5192–5197. 8 Chevallier P, Prebet T, Pigneux A, Hunault M, Delaunay J, Perry F et al. Influence of NPM1 and FLT3-ITD status on outcome in relapsed/refractory AML patients receiving salvage therapy including Gemtuzumab Ozogamicin. Leukemia 2010; 24: 467–469. 9 Filanovski K, Shvidel L, Shtalrid M. Advantage of MIDAM protocol in treatment of the elderly patients with refractory and relapsing acute myeloid leukemia. J Clin Oncol 2010; 28: e115. 10 Boissel N, Renneville A, Biggio V, Philippe N, Thomas X, Cayuela JM et al. Prevalence, clinical profile, and prognosis of NPM mutations in AML with normal karyotype. Blood 2005; 106: 3618–3620. 11 Kiyoi H, Naoe T, Yokota S, Hamagushi M, Ohno R, Saito H et al. Internal tandem duplication of FLT3 associated with leukocytosis in acute promyelocytic leukemia. Leukemia Study Group of the Ministry of Health and Welfare (Kohseisho). Leukemia 1997; 11: 1447–1452. 12 Lioure B, Delaunay J, Blaise D, Milpied N, Guardiola P, Cahn JY et al. Allogeneic stem cell transplantation (SCT) with non myeloablative conditioning regimen (NST) following intensive consolidation may be equivalent to conventional alloSCT and superior to autoSCT for patients over 50 with acute myeloid leukemia (AML) in 1st CR: first results of the AML 2001 trial. Blood (ASH Annual Meeting Abstracts) 2006; 108, abstract 319. 13 Chevallier P, Fornecker L, Lioure B, Bene MC, Pigneux A, Recher C et al. Tandem versus single autologous peripheral blood stem cell transplantation as post-remission therapy in adult acute myeloid leukemia patients under 60 in first complete remission: results of the multicenter prospective phase III GOELAMS LAM-2001 trial. Leukemia 2010; 24: 1380–1385. 14 Giles F, Verstovsek S, Garcia-Manero G, Thomas D, Ravandi F, Wierda W et al. Validation of the European Prognosis Index for younger adult patients with acute myeloid leukaemia in first relapse. Br J Haematol 2006; 134: 58–60. 15 Clarke WT, Marks PW. Gemtuzumab Ozogamicin: is there room for salvage? Blood 2010; 116: 2618–2619. 16 Burnett AK, Hills RK, Milligan D, Kjeldsen L, Kell J, Russell NH et al. Identification of patients with acute myeloblastic leukemia who benefit from the addition of Gemtuzumab Ozogamicin: results of the MRC AML15 trial. J Clin Oncol 2010; 29: 369–377. 17 Metzelder S, Wang Y, Wollmer E, Wanzel M, Teichler S, Chaturvedi A et al. Compassionate-use of sorafenib in Flt3-ITD positive acute myeloid leukemia: sustained regression prior and post-allogeneic stem cell transplantation. Blood 2009; 113: 6567–6571. 18 Fenaux P, Mufti GJ, Hellstro¨m-Lindberg E, Santini V, Gattermann N, Germing U et al. Azacitidine prolongs overall survival compared with conventional care regimens in elderly patients with low bone marrow blast count acute myeloid leukemia. J Clin Oncol 2010; 28: 562–569.