Clinical management of primary non-acute promyelocytic leukemia acute myeloid leukemia: practice Guidelines by the Italian Society of Hematology, the Italian Society of Experimental Hematology and the Italian Group for Bone Marrow Transplantation

Abstract

As many options are now available to treat patients with de novo acute myeloid leukemia, the Italian Society of Hematology and two affiliated societies (SIES and GITMO) commissioned project to an Expert Panel aimed at developing clinical practice guidelines for acute myeloid leukemia treatment. After systematic comprehensive literature review, the Expert Panel formulated recommendations for the management of primary acute myeloid leukemia (with the exception of acute promyelocytic leukemia) and graded them according to the supporting evidence. When evidence was lacking, consensus-based statements have been added. First-line therapy for all newly diagnosed patients eligible for intensive treatment should include one cycle of induction with standard dose cytarabine and an anthracycline. After achieving complete remission, patients aged less than 60 years should receive consolidation therapy including high-dose cytarabine. Myeloablative allogeneic stem cell transplantation from an HLA-compatible sibling should be performed in first complete remission: 1) in children with intermediate-high risk cytogenetics or who achieved first complete remission after the second course of therapy; 2) in adults less than 40 years with an intermediate-risk; in those aged less than 55 years with either high-risk cytogenetics or who achieved first complete remission after the second course of therapy. Stem cell transplantation from an unrelated donor is recommended to be performed in first complete remission in adults 30 years old or younger, and in children with very high-risk disease lacking a sibling donor. Alternative donor stem cell transplantation is an option in high-risk patients without a matched donor who urgently need transplantation. Patients aged less than 60 years, who either are not candidate for allogeneic stem cell transplantation or lack a donor, are candidates for autologous stem cell transplantation. We describe the results of a systematic literature review and an explicit approach to consensus techniques, which resulted in recommendations for the management of primary non-APL acute myeloid leukemia.

Introduction

The management of acute myeloid leukemia (AML), whose overall incidence is about 3–4/100,000 per year,1 with more than half of the cases occurring in patients aged 60 years or older, still represents a challenge to hematologists.2 Treatment options range from supportive care to intensive programs of chemotherapy, including autologous and allogeneic stem cell transplantation (SCT). Despite significant advances in the management of treatment-related complications, the high incidence of relapse and ineligibility of many elderly patients for intensive chemotherapy programs have so far meant low probability of both disease-free survival (DFS) and overall survival (OS).1 In this article, recommendations for the therapy of de novo AML, with the exception of acute promyelocytic leukemia (APL) which requires very specific therapeutic approaches, are presented. The guidelines are intended to support the clinical practice of hematologists, oncologists and internists who care for leukemia patients.

Design and Methods

Organization and design

The methodology used for developing SIE guidelines has been extensively reported elsewhere.3 The working group was composed of eight senior hematologists and two literature reviewers. Pubmed and the Cochrane Library were searched for relevant publications since 1995. Major hematology, oncology and general medicine journals (Blood, Journal of Clinical Oncology, British Journal of Haematology, Bone Marrow Transplantation, Haematologica, Lancet, Leukemia, New England Journal of Medicine) were scanned for publications since 1995. Proceedings of international hematology meetings held since 2005 were also scanned. The list of papers was updated in January 2008. Full papers were assigned an evidence level, according to the Scottish Intercollegiate Guideline Network.4 Based on the reviewed literature, the members of the Expert Panel formulated some evidence-based recommendations; expertise-based recommendations were added when relevant areas could not be addressed by the available evidence, but indirect evidence could support a statement. A first round of consensus for the proposed recommendations was obtained through paper questionnaires, according to the Delphi Panel technique. The full body of recommendations was definitively approved during a meeting held in Bologna on March 10, 2008. The guidelines were reported according to the COGS checklist by the Conference on Guideline Standardization. The present guidelines are expected to be updated in 2012.

Definitions

The present guidelines apply to patients with de novo AML, according to WHO classification, which updated and modified the FAB diagnostic criteria.5,6 Thereafter, the guidelines do not apply to patients with AML secondary to myelodysplastic syndrome or chronic myeloproliferative disorders, previous exposure to radiation therapy or alkylating agents or topoisomerase II inhibitors. Standard definitions for response were adopted7 (Table 1).

Guidelines

Pre-treatment evaluations

Bone marrow (BM) evaluation of an AML patient provides important prognostic information and identification of specific blast markers for disease monitoring.7 There is no consensus yet about the panel to be employed for flow cytometry characterization of myeloid blasts.8,9 Cytogenetics is, together with age and white blood cell count, one of the most important prognostic factors to stratify patients into groups at standard, intermediate and high risk of relapse (Table 2).10–14 Conventional cytogenetic analysis is not always informative, especially in cases of cryptic translocations leading to AML1/ETO, CBFB/MYH11 and MLL fusion transcript. In cases of normal or failed cytogenetics, fluorescence in situ hybridization (FISH) for MLL fusion transcript and PCR analysis for AML1/ETO and CBFB/MYH11 are recommended. FISH for monosomy 7 detection is also recommended in pediatric patients, due to the prognostic value of this aberration.15 It has recently been shown that the presence of FLT3 mutations, especially in the form of internal tandem duplication (ITD), correlates with a high risk of relapse.16–20 Conversely, mutation of nucleophosmin (NPM1) in the absence of FLT3 ITD21 identifies a favorable risk subgroup of patients with normal standard cytogenetics.20–24 Finally, mutations of CEBPA and abnormal expression of BAALC or MN1 genes confer favorable and adverse prognostic relevance respectively, and should be investigated.25 Within the group of patients with core binding factor (CBF) leukemia, a c-kit mutation identifies a subgroup at high risk of relapse.26,27 The availability of these new prognostic markers allows a more refined stratification of patients according to risk and may translate in a more rational treatment of AML patients.

Table 1.Response criteria in acute myeloid leukemia

Trephine BM biopsy, although required by WHO classification, is not routinely performed; it may be useful in cases of unsuccessful marrow aspiration, especially in patients with M7 AML, and to identify NPM mutated AML.28 The clinical assessment should include echocardiography and evaluation of co-morbidities or active infections that could preclude or postpone the start of intensive therapy.

Recommendations

An effort should be made to enrol all patients with primary AML into clinical trials.

Besides the diagnostic assays, all patients with primary (non-APL) AML should comply with the following requirements before starting treatment:

• flow cytometry characterization of immunophenotype of bone marrow blasts;
• standard cytogenetics on bone marrow cells;
• molecular analysis for established prognostic subgroups (see Table 2 on Cytogenetic-Molecular Risk). Detection of AML1/ETO and CBFB/MYH11 anomalies is a minimum requirement;
• storing of bone marrow cells; In specific patient subgroups, additional assessments should also be performed:
• patients who show core binding factor anomalies, (i.e. t(8;21) or inv(16)) should also be investigated for c-kit mutational analysis;
• metaphase FISH (preferentially on bone marrow blasts) for MLL gene translocations and mutational analysis for FLT3, NPM1, CEBPA are strongly recommended in cases with normal or failed cytogenetics;
• FISH for established aberrations is reserved for cases in which both cytogenetics and molecular analysis were not informative. FISH for CALM/AF10 should be done in cases with 11q rearrangements MLL-negative;
• in pediatric patients FISH for monosomy 7 is also useful.

HLA typing (high-resolution molecular typing of classes I and II) of the patient and, when available, of his/her siblings should be performed at diagnosis for patients aged less than 55 years, free of severe comorbidities and not affected by Down’s syndrome.

Induction therapy

The combination of cytarabine and an anthracycline is the standard induction chemotherapy for AML. The conventional two-drug regimen of daunorubicin plus cytarabine has been reported to result in a CR rate of approximately 65%.29–34 A comparison between anthracyclines given in association with cytarabine was investigated by different groups. Idarubicin appeared to be more effective than daunorubicin, though the doses of idarubicin and daunorubicin may not have been equivalent.29–33 A meta-analysis of randomized trials comparing idarubicin (usually at the dosage of 10–12 mg/m/day for three days) and daunorubicin (45–60 mg/m) showed that the use of idarubicin in association with cytarabine resulted in a higher CR rate, but provided only a slight survival benefit, that disappeared after longer follow-up.34 No significant difference between daunorubicin and mitoxantrone has been reported.35

A multicenter randomized trial (EORTC/GIMEMA AML10) of 2,157 patients reported a significantly shorter DFS after CR in patients receiving daunorubicin instead of mitoxantrone or idarubicin for induction and consolidation therapy; however, this study was only reported in abstract form.36 Some trials explored the potential advantage of adding a third drug, i.e. thioguanine, to the classical two-drug induction therapy, though little evidence is available to conclude that a three-drug regimen is a better therapy.37,38 One study suggested that the addition of etoposide during induction therapy may improve response duration.39 Induction therapy including fludarabine and cytarabine proved to be safe and feasible in elderly patients, but it did not improve outcome in comparison to the classical two-drug induction treatment.40–42

Table 2.Cytogenetic- molecular risk in acute myeloid leukemia

The role of high-dose cytarabine in induction therapy is controversial. Some randomized trials showed prolongation of DFS, especially in younger patients,43,44 while other trials did not confirm any clinical advantage45,46 with a higher toxicity and treatment-related mortality than conventionally dosed cytarabine-based induction chemotherapy.

As far as the number of induction courses is concerned, it is difficult to support the recommendation of two induction courses through the standard evidence system, because no recent trial has been conducted to formally demonstrate an advantage in terms of relapse-free and/or overall survival. However, it has been shown that the lack of achievement of an early response, defined as the clearance of blasts to a percentage lower than 10–15% in bone marrow at day +14–16 (a condition known as persistent leukemia) has an important prognostic value.47 Some authors recommend that patients with persistent leukemia at early assessment should be given a second course of induction therapy. In cases of a two-course induction strategy, a cycle including mitoxantrone and intermediate or high dose cytarabine (HAM) should be considered.47

Due to on the one hand their increased risk of developing treatment-related complications, including cardiac toxicity, and on the other hand to their favorable response to cytotoxic treatment, children with Down’s syndrome are eligible for specific protocols of induction therapy with reduced toxicity.48 In patients with Fanconi anemia, induction therapy is usually complicated by severe extra-medullary toxicity and prolonged period of aplasia, attributable to the impairment of the hematopoietic reservoir.49

An important challenge is represented by the treatment of elderly patients: front-line palliative care, without giving remission-induction chemotherapy, is associated with significantly reduced survival in patients older than 65 years, with no favorable impact on the number of days of hospitalization.50 Some authors recommend stratifying older patients, assigning to investigational treatments those with comorbidities or a poor prognosis due to unfavorable cytogenetics, and assigning the others to standard chemotherapy.51

A high white cell count (greater than 100×10/L) at diagnosis is generally regarded as a poor prognostic factor for early death. There are no randomized studies showing an advantage of leukapheresis. However, this procedure is generally safe and may be considered in patients with AML presenting with a high white cell count.52

Empirical broad spectrum antimicrobial therapy is mandatory for febrile patients who are profoundly neutropenic.53 Prophylactic oral antibiotics may be appropriate in patients with expected prolonged, profound granulocytopenia (lower than 100/mm for two weeks). Fluoroquinolones have been shown to decrease the incidence of gram-negative infection and time to first fever in randomized trials.54 Serial surveillance cultures may be helpful in such patients to detect the presence or acquisition of resistant organisms. As far as platelet transfusions are concerned, available evidence suggests that all patients with platelet counts lower or equal than 10×10/L must be given platelet transfusions. In those with a platelet count between 10 and 20×10/L, platelet transfusions should be administered in cases of fever and/or infection, while above 20×10/L the only indication for platelet transfusion is represented by clinically relevant hemorrhage.55

Placebo-controlled randomized studies evaluated post-induction chemotherapy administration of prophylactic myeloid growth factors, including granulocyte colony stimulating factor (G-CSF) and granulocyte macrophage colony stimulating factor (GM-CSF) in adult patients with de novo AML56–61 and found no significant differences in primary outcomes, despite a significant reduction in the number of days with neutropenia and/or fever, hospitalization and/or antibiotic/antifungal therapy. G-CSF priming showed a significant increase in CR rate only in the GIMEMA study,56 while an improvement in DFS and OS was reported in another multicenter randomized trial62 in standard risk patients aged 18–60 years.

Recommendations

Newly diagnosed patients with primary (non-APL) AML should receive, as soon as possible, one cycle of “standard induction therapy” including cytarabine (100–200 mg/m/day), administered by continuous seven day-long intravenous infusion and one of the following agents, administered for three days: daunorubicin (45–60 mg/m/day), idarubicin (10 mg/m/day), mitoxantrone (10 mg/m/day) [grade A]. Induction therapy including high-doses of cytarabine cannot be recommended for any specific subgroup of patients [grade B].

“Standard induction therapy” is not recommended for children with Down’s syndrome or Fanconi anemia developing AML, due to the high risk of life-threatening complications, and for adults who show at least one of the following characteristics: a) advanced age (older than 80 years), b) severe comorbidity, c) poor and not potentially reversible performance status. These latter patients should rather receive the best supportive therapy, cytoreductive therapy (attenuated doses and/or oral administration) and/or experimental therapies with significantly lower non-hematologic toxicities [grade D]. Due to the very poor long-term prognosis, patients with high risk cytogenetics (Table 2) and aged over 65 years who therefore cannot receive allogeneic SCT, are recommended not to receive “standard induction therapy”: they should rather receive experimental therapies with limited non-hematologic toxicities, cytoreductive agents and the best supportive therapy [grade C].

“Standard induction therapy” can be temporarily delayed in patients with documented active infection or a potentially reversible decline in performance status. Anti-infective and/or cytoreductive agents should be administered in the meanwhile [grade D]. Leukapheresis can be considered in patients with hyperleukocytosis (>100×10/L) [grade C]. Leukapheresis should be performed before starting any induction therapy in children with a leukocyte count above 200×10/L, especially when associated with life-threatening, either disseminated intravascular coagulation or tissue lysis syndrome [grade D].

An early bone marrow morphological and immunophenotypic evaluation on day 14–16 after the start of “standard induction therapy” can be performed within a clinical trial [grade D]. Patients with persistent leukemia at early evaluation should receive a further cycle of induction therapy as soon as possible when clinically eligible [grade D]. Assessment of response (Table 1) to induction therapy is recommended to be performed at the time of hematopoietic recovery and no later than day 30 after the start of the cycle of induction therapy [grade D]. In the context of clinical trials, immunophenotypic, cytogenetic (karyotype and/or FISH) and molecular evaluation is recommended to be performed along with morphological evaluation for the assessment of response [grade D].

After the first induction course, all children who achieve a complete or partial response and the adults who achieve a partial response should receive a second induction course [grade B]. Response should be re-evaluated also after the second induction course [grade D].

The use of myeloid growth factors during induction therapy to induce sensitization of leukemic cells to chemotherapy (priming effect) can be considered although it cannot be routinely recommended [grade A]. G-CSF administration after induction therapy is recommended, especially in elderly patients and in patients with post-chemotherapy febrile neutropenia, in order to reduce the duration of neutropenia and the related complications [grade B]. The scheduling of G-CSF should adhere to local protocols.

Consolidation chemotherapy

A single induction course is virtually always followed by a 100% relapse rate;63 therefore, post-remission therapy is routinely used in patients with AML. Usually, consolidation chemotherapy associates cytarabine at different dosages with other drugs; however, there is no clear advantage in the use of one regimen compared to others and of combination therapy compared to high-dose cytarabine alone.64,65

The effect of cytarabine dose intensity has been investigated: large randomized clinical trials have shown better results with high-dose cytarabine in comparison with standard or intermediate doses. The advantage associated with more intensive doses of cytarabine (3 g/m×6 doses) was found to be particularly significant for patients who had not previously received high-dose cytarabine and for those with low cytogenetic risk: in such patients, consolidation with high-dose cytarabine was associated with an outcome similar to that obtained treating patients with an autograft.66–68 Conversely, higher-dose therapy had no benefit in the post-remission management of elderly patients (aged 60 years or older) with de novo AML, since the clinical benefit was jeopardized by toxicity.66 AML in Down’s syndrome patients is extremely sensitive to chemotherapy consolidation therapies, especially those including high-dose Ara-C.69

The optimal number of consolidation courses has been specifically investigated: for patients who are not candidates for transplantation, the duration of post-remission consolidation therapy should not exceed 3–4 cycles.66 Patients eligible for either autologous or allogeneic transplant, should receive a shorter consolidation, e.g. 1–2 cycles with high-dose chemotherapy. This is also useful for peripheral blood stem cell mobilization in patients who are candidates for autologous transplantation and is often used as a bridge to allogeneic transplantation.70

All patients with a given marker identified at diagnosis should be monitored for persistence of Minimal Residual Disease (MRD) after consolidation and before proceeding to the subsequent planned treatment. The approach to patients with documented persistence of MRD has not yet been standardized, so that the level of evidence is still low.

The role of post-consolidation maintenance therapy has not been clearly defined: in pediatric patients it has been employed by the German group71 without clear evidence of an advantage in terms of leukemia-free survival. Furthermore, in children, maintenance therapy may induce chemotherapy resistance and reduce response to salvage therapy in patients who experience leukemia relapse.72

A favorable effect on DFS has been observed with several schedules of maintenance chemotherapy for adult and elderly patients not submitted to SCT, even if there is not enough evidence to support such a recommendation.46,73,74 More recently, in a multicenter randomized trial, an improvement in DFS and OS was demonstrated with maintenance in elderly patients in first complete remission after intensive induction chemotherapy.75

Recommendations

Patients in first complete remission should receive a consolidation treatment, as soon as the hematologic recovery from induction therapy has occurred [grade B].

Children are candidates for post-remissional, consolidation therapy either alone or associated (in cases of intermediate/high risk cytogenetics) with SCT. Patients with Down’s syndrome have a particularly favorable response to consolidation therapy including high-dose Ara-C [grade D].

Adult patients aged under 60 years should receive post-remission consolidation chemotherapy based on high-dose cytosine arabinoside (3 g/m×6 doses); the number of cycles should not exceed 3–4 [grade A].

Potential candidates for allogeneic SCT should receive a shorter intensive consolidation including intermediate/high dose cytosine arabinoside in order to spare undue toxicity [grade D]. Potential candidates for autologous SCT should receive at least one intensive consolidation cycle including intermediate/high dose cytosine arabinoside before collecting stem cells and performing autograft [grade D].

Elderly patients (over 60 years) should not receive high-dose cytosine arabinoside-based consolidation therapy and no more than 2 consolidation cycles [grade C].

G-CSF administration is recommended after consolidation chemotherapy in order to reduce the duration of neutropenia and the related complications [grade C].

All patients with cytogenetic and/or molecular markers identified at diagnosis should be monitored for persistence of MRD after consolidation therapy has been completed and before proceeding to the subsequent planned treatment [grade D].

For patients who are not candidates to SCT, maintenance chemotherapy cannot be recommended [grade D].

Allogeneic stem cell transplantation

Allogeneic transplantation from an HLA matched sibling donor has been used for more than two decades and continues to be considered an optimal approach for prevention of relapse after remission, due to the graft-versus-leukemia effect. However, given the still significant acute transplant-related mortality (TRM) and long-term sequelae, its use remains an object of debate and ongoing investigation. Clinical trials comparing allogeneic SCT with other post-remissional therapies have produced inconsistent results.

After careful assessment of risks and benefits, intermediate and high risk patients (considering WBC counts, cytogenetics, and time to achieve remission as risk factors) are considered suitable candidates for allogeneic SCT from an HLA identical sibling.76,77 However, data derived from a literature review show a complex and variable scenario. Two metanalyses of randomized trials78,79 exploring treatment options for adult and pediatric AML patients in first CR, and employing both natural randomization based on donor availability and intention-to-treat analysis, showed a significant improvement in DFS and OS (hazard ratio 1.4) with allogeneic SCT from an HLA-identical sibling donor. The improvement was, however, limited to adult patients with high risk cytogenetics and to pediatric patients in the high- and intermediate-risk groups.78–80 A subsequent large naturally randomized trial demonstrated a significantly lower incidence of relapse and better DFS in patients belonging to the intermediate-and poor-risk cytogenetic group receiving allogeneic SCT; the benefit was even greater in patients under 40 years of age.81

Other risk factors can identify categories of patients who benefit from allogeneic transplantation, such as those with late achievement of CR.70 The effect of allogeneic SCT in other risk classes, defined by new genetic risk factors, has been retrospectively investigated: so far, there is no strong evidence that FLT3 status (the most common genetic mutation identified) should be considered an indicator for transplantation.82 Prospective randomized clinical trials with subgroup analysis for FLT3 mutation, as for other genetic risk factors, are warranted to address this question. Notwithstanding, there are convincing data suggesting that the subgroup of patients with intermediate cytogenetics and mutant NPM1 without FLT3-ITD mutations have more favorable prognosis and, therefore, allogeneic SCT is not a rational procedure for such patients.25

In the absence of an HLA-identical family donor, it seems reasonable to offer allogeneic SCT from a matched unrelated donor (MUD) to patients with poor-risk disease, either for biological features or for late achievement of CR.76 With the use of MUD transplantation, a long-term OS comparable to that obtained with a sibling donor, and far exceeding that observed with autotransplant, has been observed by some authors in this high-risk cohort.83 However, the evidence supporting an advantage with MUD-SCT in high-risk patients without a sibling donor is weak.84 T-cell depleted allogeneic SCT from an HLA-haplotype mismatched relative emerges as a viable, alternative option for AML patients without matched donors and/or those who urgently need transplantation, especially when the donor shows alloreactivity of natural killer cells towards the recipient.85

Cord blood from an unrelated donor represents a further alternative source of stem cells for pediatric and adult AML patients without matched donors and/or those who urgently need transplantation.86,87 Promising results in terms of reduction of TRM have been reported in adults given two different cord blood units.88 Moreover, it has been reported that even 2 HLA disparities between donor and recipient can be tolerated for cord blood transplant.89 Alternative donor SCT should be performed in centers with an active program in the field, since the procedure requires special expertise.

The choice of the stem cell source to be used cannot be based on the results of prospective comparative studies, which are lacking. Retrospective analyses did not show a survival advantage for either peripheral blood or bone marrow derived progenitor cells90 except in patients receiving a high dose of bone marrow stem cells.91

The introduction of reduced intensity conditioning (RIC) regimens has enabled the use of allogeneic transplantation in the elderly. Evidence for an advantage with RIC SCT in patients over 50 years with intermediate-high risk disease in first CR is weak: it relies on retrospective analyses, without genetic randomization and within cohorts with a short follow-up.92–94 Data show that in patients over 50 years, there was no statistical difference in DFS and OS after RIC transplant, compared with myeloablative SCT, irrespective of disease status. In multivariate analysis, the advantage in terms of significantly lower acute GVHD and TRM was offset by a higher relapse rate, 3-year DFS and OS being similar. In patients over 50 years receiving a low dose total body irradiation-based RIC regimen, a comparison between related and unrelated SCT showed no statistical difference in 2-year OS.95

Recommendations

Myeloablative allogeneic SCT from a fully matched sibling donor is recommended to be performed in first complete remission for all children with intermediate-high risk cytogenetics and for adults with high-risk cytogenetics (Table 2), provided that they are aged under 55 years and do not carry severe comorbidities [grade A]. Myeloablative allogeneic SCT from a fully matched sibling donor is recommended to be performed in first complete remission also for adult patients with intermediate-risk cytogenetics with the exception of NPM1 mutant and FLT3-ITD negative cases, provided that they are aged under 40 years and do not carry severe comorbidities [grade C]. Myeloablative allogeneic SCT from a fully matched sibling donor is recommended to be performed for patients who achieved a first complete remission only after having received a second course of induction therapy, irrespectively of their cytogenetic risk, provided that they are aged under 55 years and do not carry severe comorbidities [grade D].

No source of allogeneic stem cells (peripheral blood or bone marrow) can be recommended to be preferred for myeloablative allogeneic SCT [grade D].

If no fully matched sibling donor is available, it is recommended to consider allogeneic SCT from an unrelated donor for all adult patients in first complete remission aged under 30 years with high-risk cytogenetics, or who achieved first complete remission only after a second course of induction therapy [grade D]. Myeloablative allogeneic SCT from an unrelated donor is not recommended in patients older than 50 years who achieved complete remission after induction therapy [grade D]. Children with M7 AML, a complex karyotype, monosomy of chromosome 7, or high levels of MRD measured by flow cytometry after consolidation therapy are eligible for SCT from an unrelated volunteer [grade D].

Alternative donor (i.e. mismatched-related, cord blood) SCT should be performed only by centers with an active program in the field, in high cytogenetic risk AML adult patients without a matched (related or unrelated) donor and/or who urgently need transplantation [grade D].

Alternative stem cell donors should also be considered for all the children who are candidates for transplantation from an unrelated donor, but who lack such a matched donor or urgently need the allograft [grade D].

Allogeneic SCT with a RIC regimen should be considered in high-risk patients aged over 55 years or with severe comorbidities [grade D].

Autologous stem cell transplantation

Although data about the superiority of autologous SCT in first CR over conventional consolidation chemotherapy are controversial, both in the adult and pediatric population, the decrease in toxicity has made autologous SCT a feasible option for younger patients (and for 20% of elderly ones) who lack an HLA matched donor. A 45% 5-year OS rate in high-risk (with 31% DFS) and 64% in good-risk patients have been observed in a large cohort of patients with a median follow-up of 9.5 years after autologous SCT.96 In other trials, a significant reduction in the incidence of relapse has been observed in good- and intermediate-risk patients.97 Conversely, in 2 meta-analyses of 6 controlled trials conducted up to 1996 involving adult patients, autologous SCT was shown to improve event-free survival by about 25%, without any effect on OS, compared with standard dose consolidation.98,99 Subsequently published randomized trials confirmed these results.67,100 Nonetheless, retrospective analyses comparing autologous SCT to MUD transplant in patients in first CR lacking an HLA identical family donor showed an advantage with the use of autologous SCT.101 The decrease in toxicity has made autologous SCT a feasible option also in 20% of elderly patients: comorbidity, extreme age, or poor mobilization are the main reasons for not performing autologous SCT.102 A meta-analysis of a pediatric population79 failed to demonstrate a survival benefit for autologous SCT in comparison to conventional post-remissional chemotherapy.

In detail, 2 randomized trials showed a significant reduction in relapse in children given autologous SCT, which did not translate into an advantage in terms of survival due to an increased mortality (POG 8891)103 or to a higher chance of being rescued by a second line treatment for patients given chemotherapy as consolidation treatment (MRC AML10).104 A more recent randomized study comparing conventional post-remissional chemotherapy and autologous SCT showed a comparable relapse incidence.105

Stem cell harvest is usually performed after the last consolidation chemotherapy cycle. Different non-randomized trials provided compelling evidence in favor of a reduced relapse rate when ex vivo purging with a cyclophosphamide derivative, such as mafosfamide, is employed.106–108 Patients with persistent first CR for more than six months probably do not need autograft as they have a high chance of being already cured.

Recommendations

Consolidation autologous SCT is recommended for patients eligible for high-dose chemotherapy who are not candidate for allogeneic SCT from a fully HLA matched donor [grade B]. Children with Down’s syndrome should not be considered for autologous SCT [grade B]. Patients are recommended to receive autologous SCT within six months of achievement of first CR [grade D]. Patients with persistent first CR for more than six months should not receive autologous SCT [grade D]. Stem cell harvesting should be performed when the best “in vivo” purging has been completed, i.e. after the last consolidation chemotherapy cycle [grade D]. Peripheral stem cell should be mobilized with the administration of myeloid growth factors (usually G-CSF) after the consolidation chemotherapy [grade D]. The adequate number of peripheral stem cells to be reinfused is 2.5×10 per kilogram of patient weight [grade D]. Response to autologous SCT should be assessed after recovery of at least 500 neutrophils/μL [grade D]. There is no evidence to support maintenance chemotherapy after autologous SCT in patients with a first CR [grade D].

Discussion

The Expert Panel provided recommendations by answering the most relevant clinical questions regarding AML patients: which is the diagnostic and prognostic value of pre-treatment evaluations¿ Which patients are candidates for intensive chemotherapy¿ Which is the best induction regimen in eligible patients¿ Which are the post-remissional treatment options, including autologous or allogeneic transplantation¿

The most interesting advance in AML treatment strategy is the opportunity provided by the availability of clinical and biological prognostic factors, to stratify patients according to their probability of relapse and thus to optimize and tailor post-remissional treatment. Significant prognostic information can be obtained at diagnosis, searching for cytogenetic or molecular aberrations, and during the treatment course, monitoring persistent leukemia early after induction treatment and MRD after consolidation therapy. Although risk stratification still relies on conventional cytogenetics, in the future knowledge about the prognostic impact of genetic aberrations will provide the basis for an individualized and tailored treatment, and hopefully for novel targeted therapies.

Acknowledgments

we thank Giambattista Bertani, who collaborated in the literature review.

Footnotes

• Funding: Dompè Biotech s.p.a. provided the financial support for literature search, consensus conferences and personnel for this project. None of the study participants disclosed any financial interest in this company.
• Legal note: the enclosed guidelines are intended to be general recommendations for clinical practice: they do not necessarily apply to single patients. The authors do not have any legal responsibility for the consequences of their application.
• Authorship and Disclosures EM was a member of the Expert Panel and wrote the paper; MM contributed to literature search, literature review and to writing the paper; GB co-ordinated the consensus meetings; ST conceived the study, co-ordinated the guideline production process and was a member of the Expert Panel; MV, AB, GM, FL, FF, CM were members of the Expert Panel and revised the paper.
• We thank Giambattista Bertani, who collaborated in the literature review.

• Received April 2, 2008.
• Revision received August 14, 2008.
• Accepted September 5, 2008.