Progress in the treatment of acute myeloid leukemia



Significant progress in understanding the mechanisms leading to the development of acute myeloid leukemia (AML) has led to the identification of numerous molecular abnormalities that may be responsible for leukemogenesis. Over the same period, large trials have established standard regimens combining cytotoxic agents for the treatment of patients with AML. Current research is attempting to better stratify patients by identifying risk factors responsible for resistance, and to discern ways for incorporating newer agents with specific and targeted activity into our standard regimens, Herein the recent developments in the diagnosis and treatment of AML are reviewed. Cancer 2007. © 2007 American Cancer Society.

The traditional chemotherapy regimens used to treat adults with acute myeloid leukemia (AML) have not changed significantly. Large randomized trials failed to demonstrate the superiority of adding new cytotoxic agents to the traditional cytarabine and anthracycline-based regimens. With the exception of high-dose cytarabine in younger patients with favorable cytogenetics, variations in the dose have not improved survival. Understanding the molecular events of leukemogenesis has led to the identification of molecular alterations with prognostic significance, which may also be targets for therapeutic intervention. This has already led to improvements in outcome in patients with specific subsets such as acute promyelocytic leukemia (APL). Agents with significant in vitro activity targeted at leukemia-specific aberrations have been investigated in early clinical trials with promising activity. Herein we review the progress in the therapy of AML.

Risk Stratification in Adult AML

Predictors of response and survival in AML include: 1) patient-related factors such as age, performance status, and organ function; 2) disease-related factors recognizing the inherent resistance of some subsets of AML to existing therapies; and 3) treatment-related factors (recognizing that identification of highly effective therapies can overcome the other 2 determinants).

Previous studies have identified a number of consistent prognostic factors. These include age, karyotype, performance status, organ dysfunction, white blood cell count at presentation, antecedent hematologic disorders, and molecular abnormalities (eg, MDR, FLT-3 ITD, C/EBPα, BAALC, NPM). Age is an important determinant of outcome in AML, with investigators defining age cutoffs of 60 or 65 years beyond which the outcome is worse. Poor prognosis in older patients is a result of an intrinsically more resistant disease (higher incidence of poor-risk cytogenetics, MDR, and other as-yet undefined risk factors), and limited tolerability of traditional cytotoxic agents. In a study of 998 patients aged 65 years or older with AML or high-risk myelodysplastic syndrome (MDS), we identified predictive factors for complete remission (CR), 8-week mortality, and overall survival (OS).1 In all 3 categories, patients aged 75 years or older fared worse than younger patients. Other studies have confirmed the influence of age on outcome of AML.2

Risk Stratification by Cytogenetics

The prognostic effect of karyotype is well established (Table 1). A Medical Research Council (MRC) study in 1612 patients confirmed the importance of cytogenetics in risk stratification for AML.3 Patients in this study were younger (age <55 years) and received uniform treatment. They could be divided into 3 cytogenetic categories: favorable, intermediate, and unfavorable karyotypes. Patients with adverse cytogenetic abnormalities had 5-year survival rates of 10% to 21%. Patients with structural (eg, del[9q], abn[11q23], del[7q]) or numeric (eg, +8, +21, +22) cytogenetic abnormalities (as well as diploid cytogenetics) had 5-year survival rates ranging from 23% to 60% and were identified as having intermediate risk disease. Patients with favorable abnormalities, including t(8;21), inv(16), and t(15;17), had 5-year survival rates of 60% to 70%. The presence of secondary abnormalities did not influence outcome.4 In a follow-up study including an older population (ages 44–91 years), all subgroups had worse outcomes, further emphasizing the importance of age4 (Table 1).

Table 1. Cytogenetics and Prognosis in Patients With AML
StudyNo. of patientsAge, years5-Year survival by cytogenetics, %
  1. AML indicates acute myeloid leukemia; MRC, Medical Research Council; CALGB, Cancer and Leukemia Group B; SWOG, Southwest Oncology Group; ECOG, Eastern Cooperative Oncology Group.


The Cancer and Leukemia Group B (CALGB) determined the prognostic impact of cytogenetic abnormalities on CR rate, 5-year cumulative incidence of recurrence (CIR), and 5-year survival.5 In this study of 1213 patients with AML, aged 15 to 86 years, the characteristics of the 3 identified subgroups were slightly different from the MRC study. They defined complex as the presence of more than 3 abnormalities and identified del(9q), as favorable for survival but intermediate for CR and CIR. In addition, among translocations involving 11q23, t(6;11), and t(11;19) were unfavorable for survival, whereas t(9;11) was intermediate. The 3 categories had different expectations for 5-year survival (Table 1). In a more recent study, the CALGB investigators further confirmed the importance of pretreatment cytogenetics in predicting outcome in patients aged 60 years and older.6 The Southwest Oncology Group (SWOG) and Eastern Cooperative Oncology Group (ECOG) studied 609 adults with AML (ages 16–55 years) who were treated with allogeneic transplantation if a donor was available and identified cytogenetic risk groups similar to those observed in the previous studies (Table1).7

Cytogenetic risk assessment has led to specific treatment strategies. Patients with APL may be treated effectively with ATRA and arsenic trioxide.8–10 Patients with favorable-risk AML other than APL are highly sensitive to cytarabine, and multiple courses of high-dose cytarabine (range, 3–4 courses) significantly improved their survival.11

Molecular Abnormalities as Predictors of Outcome

Normal or diploid karyotype constitutes the largest single cytogenetic group in AML.11 Intermediate-dose or high-dose cytarabine improved outcome in younger patients (age ≤60 years) with diploid karyotype.12 The need to define risk factors within cytogenetic subgroups (particularly with diploid AML) has led to the identification of several prognostic molecular abnormalities (Table 2).13 The role of these molecular subgroups in identifying subsets within the favorable and poor-risk cytogenetic categories is less clear.

Table 2. Molecular Abnormalities in AML as Predictors of Outcome
ReferencesMolecular abnormalityAML subtypeDisease-free survivalOverall survival
  1. AML indicates acute myeloid leukemia; MLL PTD, mixed lineage leukemia partial tandem duplication; FLT-3 ITD, FMS-like tyrosine kinase-3 internal tandem duplication; BAALC, brain and acute leukemia cytoplasmic; CEBPα:,CCAAT/enhancer binding protein-α; NPM, nucleophosmin; WT-1, Wilms tumor-1; ERG, ETS-related gene; APL, acute promyelocytic leukemia; NR, not reported.

Caligiuri et al., 199814; Schnittger et al., 200015; and Dohner et al., 200216MLL PTDDiploid, trisomy 11ShortenedNone
Kottaridis et al., 200118 and Schnittger et al., 200292FLT-3 mutationsDiploidShortenedShortened
Gale et al., 200593FLT-3 mutationsAPLNoneNone
Baldus et al., 200620 and Baldus et al., 200394↑ BAALC mRNADiploidShortenedShortened
Preudhomme et al., 200221 and Frohling et al., 200422CEBPα mutationDiploidProlongedProlonged
Falini et al., 200523; Dohner et al., 200524; Schnittger et al., 200525; and Verhaak et al., 200595NPM mutationDiploidProlongedProlonged
Bergmann et al., 199796WT-1All subtypesNRShortened
Bacher et al., 200697NRAS mutationAll subtypesNoneNone
Paschka et al., 200628KIT mutationinv(16), t(8;21)ShortenedShortened
Marcucci et al., 200598↑ ERG mRNADiploidShortenedShortened
van Waalwijk et al., 200399↑ EVI1 mRNA3q26, unfavorableShortenedShortened

The FMS-like tyrosine kinase (TK) 3, or FLT-3 is a member of class 3 TK receptors. FLT-3 protein is expressed in early hematopoietic progenitors and plays a major role in myeloid differentiation.17 An important mutation in FLT3 is the internal tandem duplication (ITD), which results in constitutive activation of the TK activity of FLT-3, leading to activation of downstream signaling pathways mediating survival and proliferation.17 Another mutation, occurring in 7% of patients, is the TK domain point mutation in codon 835. FLT3 ITD mutations occur in 20% to 30% of patients with AML and are common in patients with diploid cytogenetics or APL. FLT3 mutations are associated with a poor prognosis, particularly the presence of an ITD in conjunction with loss of the second wild-type allele.18, 19

The brain and acute leukemia, cytoplasmic gene (BAALC) encodes a protein of unknown function. High BAALC expression is an independent predictor of resistance, high CIR, and shorter survival in patients with AML and normal karyotype.20 Based on BAALC mRNA expression and FLT3 ITD mutations, the investigators distinguished 4 subgroups of patients with diploid AML. Patients with both high levels of BAALC expression and FLT3 mutations had the worst survival, whereas patients with neither mutation did best.20

The CCAAT/enhancer binding protein α (C/EBPα) is a transcription factor involved in the regulation of myelopoiesis. Mutation of the C/EBPα gene occurs in 15% of patients with AML and normal karyotype and is associated with longer remission and survival.21, 22 The effect of the coexpression of FLT3 ITD mutations and mutated C/EBPα is unclear.21, 22

Partial tandem duplication (PTD) of the MLL gene (located at 11q23) was the first molecular abnormality described to occur in approximately 10% of patients with AML and normal karyotype.14MLL PTD occurs in conjunction with other cytogenetic abnormalities (such as complex abnormalities) but less commonly with favorable cytogenetics (such as inv(16) or t(8;21)).15 The duration of remission was shorter if MLL PTD was present.16

Nucleophosmin (NPM) is a nucleo-cytoplasmic protein that shuttles molecules between cellular compartments. NPM gene mutations have been described in a rare form of APL with a translocation involving NPM and RARα and in anaplastic large cell lymphoma. NPM gene disruptions are seen in approximately 35% to 50% of patients with AML, particularly those with normal karyotype, and have been associated with FLT3 mutations.23 The presence of NPM1+ AML is associated with a higher CR rate, longer survival, and improved event-free survival (EFS).24 A German cooperative group study of 401 patients with AML and diploid karyotype found that NPM mutations were frequently associated with FLT3 mutations but rarely with other mutations.25 In this study, the negative influence of FLT3 mutations overcomes the positive influence of NPM mutations on survival.25

More recently, 872 patients with normal karyotype, ages 16 to 60 years (median age, 48 years), were evaluated for the presence of molecular alterations including NPM, FLT3 ITD, FLT3 TKD, C/EBPα, MLL PTD, and NRAS.26 The overall CR rate was 76%. In a logistic regression model, the NPM1+FLT3-ITD (P < .0001) and C/EBPα (P = .05) genotypes were associated with induction success and were prognostic for recurrence-free and overall survival. Allogeneic stem cell transplantation (SCT) in first CR was beneficial only in patients with normal karyotype who did not have the NPM1+FLT3-ITD phenotype.26 However, further validation of the impact of FLT3 status on the outcome of transplants is needed. The MRC investigators have suggested that neither an autograft nor an allograft can overcome the adverse prognostic effect of mutated FLT3 and that the presence of FLT3 mutation should not be an indication for either procedure in first remission.27

In summary, several prognostic molecular abnormalities have been identified that are particularly useful in AML with diploid karyotype. Similar abnormalities with potential impact on outcome are being reported for other cytogenetic subgroups. For example, The presence of c-kit mutations predicts a worse outcome in patients with favorable risk disease with inv(16) and t(8;21).28 More recently, polymorphisms in the genes encoding drug-metabolizing enzymes and DNA repair mechanisms (XPA and SULTIC2) have also been found to have prognostic significance.29 New techniques including gene expression profiling are employed to identify different subgroups of patients.30, 31 Whether these subset profiles will assist in risk stratification or in identification of new molecular targets for therapy remains to be determined. Increased availability of these markers and their prospective incorporation into large multicenter trials will further validate their influence on the outcome of patients.

Treatment Options in Adult AML

Despite multiple clinical trials examining combinations of cytotoxic agents, the standard regimen for AML induction is 3 days of an anthracycline and 7 days of standard dose cytarabine (3 + 7). Variations in this regimen have included the incorporation of high-dose cytarabine,32 double-dose induction, and the addition of other agents such as etoposide,33 chlorodeoxyadenosine,34 and fludarabine.35

Controversies concern the optimal dose of daunorubicin, whether high-dose cytarabine during induction would be more effective, if idarubicin is better than daunorubicin, and which newer agents may improve outcome. Large trials have compared idarubicin with daunorubicin in newly diagnosed AML.36, 37 In a study by Wiernik et al.37 there was an improvement in CR duration for patients who received idarubicin. In 2 studies, there was a survival advantage for idarubicin.36, 37 However, the difference in the efficacy of the 2 anthracyclines may not exist if a higher dose of daunorubicin is used (eg, 60 mg/m2 or higher, daily for 3 days).

Vignetti et al.38 reported the results of a European study randomizing 2157 patients with AML to receive induction with daunorubicin (50 mg/m2 daily for 3 days), mitoxantrone (12 mg/m2 daily for 3 days), or idarubicin (10 mg/m2 daily for 3 days) in addition to cytarabine and etoposide. In patients achieving CR, 1 course of consolidation with the same agents was followed by allogeneic or autologous SCT. The CR rates were similar for the 3 arms and during induction the recovery times and toxicity were comparable. However, during consolidation the incidence of grade 3 and 4 infections was 14% with daunorubicin, 23% with mitoxantrone, and 24% with idarubicin. Recovery times were also significantly longer with mitoxantrone and idarubicin, suggesting a more pronounced hematopoietic toxicity with these agents.38 Among patients who did not undergo SCT, the disease-free survival was significantly superior in the idarubicin and mitoxantrone arms.

High-dose cytarabine (HIDAC) (generally considered as ≥1 g/m2/dose), as part of induction therapy, was associated with longer remission duration and EFS in younger patients (age <60 years).39, 40 Kern and Estey41 performed a pooled analysis of data from randomized clinical trials that compared HIDAC with standard-dose cytarabine (SDAC, ≤0.2 g/m2/dose) during induction in patients with AML. No differences were noted with regard to CR rates. However, the use of HIDAC was associated with better long-term disease control and survival in adults younger than 60 years of age.41

Age, concomitant cardiotoxicity, congestive heart failure, or other comorbid conditions may prohibit the use of anthracyclines. Alternative regimens such as the combination of fludarabine plus cytarabine or topotecan plus cytarabine have been used.42, 43 These regimens have produced equivalent CR rates and outcome to standard anthracycline/cytarabine regimens, without any associated cardiotoxicity.44

Prognostic factors are currently used to guide therapy in specific subsets of patients such as those with APL or favorable cytogenetics leukemias. Strategies to develop more effective therapies for patients with other subsets of AML are needed. These include: novel nucleoside analogs (eg, clofarabine), strategies to reverse drug resistance (eg, multidrug resistance, cell cycle inhibition), monoclonal antibodies (eg, gemtuzumab ozogamicin), and molecularly targeted therapies (eg, FLT-3 inhibitors, hypomethylating agents, and farnesyl transferase inhibitors).45 A detailed discussion of new agents in AML is beyond the scope of this article and has been reviewed elsewhere.45 Here we focus on selected agents with promising early clinical data.

Several inhibitors of the FLT-3 kinase are undergoing evaluation.46 Lestaurtinib (CEP701) is being evaluated in the salvage setting. Patients with FLT-3-activating mutations are randomized to treatment with chemotherapy with or without lestaurtinib.47 Preliminary results indicate that the addition of the FLT-3 inhibitor may improve the response rate. This agent has also produced clinical responses as first-line treatment in older patients with AML.48

PKC412 is another FLT-3 kinase inhibitor with single-agent activity in AML.49 Preliminary results showed a favorable CR rate for the combination of daunorubicin plus cytarabine plus PKC412 in patients age 60 years or younger with newly diagnosed AML.50 In the latest follow-up of the study, 38 patients (26 with wild-type FLT-3 [FLT-3wt] blasts and 12 with mutated FLT-3 [FLT-3mut] blasts) were evaluable. CR occurred in 27 of 38 (71%); the CR rate in patients with FLT-3wt was 18 of 26 (69%) and in patients with FLT-3mut 11 of 12 (92%).50 Another agent, tandutinib (MLN518), has shown antileukemic activity in early trials.51

Tipifarnib (R115777) is a nonpeptidomimetic farnesyl transferase inhibitor (FTI) that has been combined with chemotherapy for first-line treatment of AML or high-risk myelodysplastic syndrome (MDS) (blasts >10%).52 Induction with idarubicin, cytarabine, and tipifarnib was followed by 5 courses of attenuated dose consolidation and by maintenance with single-agent tipifarnib for 6 months. There was no improvement in CR rate or remission duration and an increase in toxicity, particularly diarrhea and liver dysfunction.52 Karp et al.53 are testing the role of tipifarnib in maintenance in a phase 2 trial for patients with poor-risk AML in CR after induction and consolidation.

Aberrant DNA methylation, and other epigenetic events, are important in the progression of a number of human neoplasms.54 Methylation of promoter-associated CpG-rich regions (CpG islands) can lead to silencing and inactivation of tumor suppressor genes, acting as an alternative mechanism to deletions and mutations.55 Methylation of promoters of genes such as p15 have been associated with disease progression and with worse outcome in myeloid malignancies.56 Decitabine (5-aza-2′-deoxycitidine) irreversibly inhibits DNA methyltransferases (DNAMT), enzymes that methylate newly synthesized DNA, leading to hypomethylation of the promoters of tumor suppressor genes and their activation.54 Decitabine and 5-azacytidine, either alone or in combination with histone deacetylase inhibitors, are currently under evaluation for the treatment of patients with AML.57 A recent reanalysis of data from 3 sequential CALGB trials of 5-azacytidine demonstrated that among 103 patients with AML at baseline (using World Health Organization [WHO] criteria), 48% had hematologic improvement or better responses.58

Role of Maintenance Therapy

Maintenance therapy is a well-established strategy in acute lymphoblastic leukemia (ALL)59 but has not become part of routine treatment of AML due to conflicting results from prior clinical trials. An early study by ECOG randomized patients age younger than 60 years in remission to no further therapy, short-term high dose consolidation therapy, or prolonged maintenance.60 The observation arm of the study was closed early due to a significantly poorer outcome compared with maintenance (P = .002). In a follow-up report, the same investigators compared the long-term outcome of patients age younger than 41 years with no available donor for a transplant who were randomized to either 2 years of continuous outpatient maintenance therapy with cytarabine and 6-thioguanine or a single course of in-patient high-dose cytarabine-based consolidation therapy.61 The EFS and survival were significantly better for patients who received consolidation therapy.

A randomized study by the CALGB demonstrated that protracted maintenance for 36 months was not superior to 8 months of maintenance.62 Sauter et al.63 reported that patients in remission after a course of consolidation who were assigned to maintenance chemotherapy every 8 weeks for 2 years did not do better than those randomized to observation only. In studies by the German AML cooperative group, patients achieving CR were randomized to protracted maintenance or no further therapy.64–66 Monthly maintenance after consolidation was superior to consolidation alone in prolonging the remission duration.64–66 The benefit of this strategy was particularly evident in patients with poor-risk features, defined by unfavorable karyotype, age ≥60 years, high lactate dehydrogenous levels, or residual Day 16 bone marrow blasts >40% (P = .006).65 In another European study, patients in CR received 1 cycle of consolidation therapy and were randomized to no further therapy or low-dose cytarabine at 42-day intervals for a total of 8 cycles or until recurrence.67 Low-dose cytarabine prolonged recurrence-free survival but not overall survival.67 In a study conducted by SWOG, patients ineligible or unable to undergo an allogeneic SCT were randomized to receive intensification alone or intensification followed by monthly maintenance. In multivariate analysis, maintenance was associated with a prolonged disease-free survival (DFS) but not survival.68

The available data suggest that maintenance therapy with cytotoxic drugs similar to those used in induction and consolidation leads to improved DFS in AML. A potential reason for the lack of success of maintenance therapy in improving survival in AML is the use of drugs with mechanisms of action similar to those used in the induction-consolidation therapy. These drugs are not discriminatory toward malignant myeloid cells, are toxic to the normal hematopoietic myeloid progenitors, and the residual leukemia cells are likely to be resistant to their effect. This argues for the use of alternative agents with different mechanisms of action. The use of immune-modulating agents like interleukin-2 (IL-2) in an effort to stimulate the host antileukemic immune effect has been evaluated in the postremission setting.69 In a recent international randomized trial, 320 patients with AML in CR were stratified by CR1 or subsequent CR (CR >1) and randomized to treatment with IL-2 plus histamine dihydrochloride or no treatment.70 Patients in the experimental arm had significantly longer leukemia-free survival compared with those in the control arm (40% vs 26% at 3 years; P = .01). Other agents such as hypomethylating agents, FTIs, and vaccines may have a potential role in this setting.

Management of AML in the Elderly

Treatment of elderly patients with AML is challenging and deserves a separate mention. The U.K. AML trial experience since the 1970s showed that the proportion of patients who achieve 5-year or 10-year survival remains very low.4 Results from the AML11 trial showed no difference in survival with different types of regimens or different lengths of treatment.71 The design of the AML14 trial included selection of intensive or nonintensive chemotherapy for patients older than age 60 years with AML. This study, which included 1237 patients, demonstrated no difference in CR rate or overall survival with high-dose or low-dose daunorubicin, low-dose or high-dose cytarabine induction, or 3 or 4 total courses of therapy.72 The5-year survival for intensive chemotherapy was 13%.72

Results from the nonintensive chemotherapy arm of the AML14 trial showed that median survival significantly improved with low-dose cytarabine (20 mg twice daily for 10 days, every 4 to 6 weeks) compared with hydroxyurea.73 CR occurred in 1 of 92 (1%) patients in the hydroxyurea treatment arm and in 15 of 92 (17%) patients in the low-dose cytarabine arm. Benefits were limited to patients who had intermediate-risk karyotype.

Appelbaum et al.2 analyzed data from 5 SWOG clinical trials that included 968 patients. Increasing age was associated with less favorable cytogenetics, poorer performance status scores, lower white blood cell counts, and a lower percentage of marrow blasts. Increasing age was also associated with a higher incidence of early death after induction therapy, a lower CR rate, and shorter survival.

The poor outcome of elderly patients with AML has led to the search for less toxic and more effective agents in this setting. Gemtuzumab ozogamycin has been approved for treating older patients (age ≥60 years) with AML in first recurrence with CR1 duration of >3 months.74 Gemtuzumab is not used widely to treat these patients due to its possible liver toxicity. Among the patients treated on the U.K. AML15 trial, which combined gemtuzumab with chemotherapy in predominantly younger patients, approximately 140 patients aged 60 years and older were treated. Preliminary data indicate that these patients had a high CR rate of 82%.75 This study is likely to generate further interest in combining low doses of gemtuzumab with chemotherapy in both the adult and elderly populations.75

The adenosine nucleoside analog clofarabine was recently approved for the treatment of pediatric patients with refractory or recurring ALL. Clofarabine is being evaluated as monotherapy or in combination for the treatment of older patients with AML. Faderl et al.76 reported the results of a phase 2 study of clofarabine (40 mg/m2 given as a 1-hour infusion on Days 2–6) plus intermediate dose cytarabine (1 g/m2 daily given as a 2-hour infusion for 5 days) in patients aged 50 years or older with newly diagnosed AML. Among the 60 enrolled patients, 48% had secondary AML and 50% had abnormal karyotype. The overall response rate was 60% (52% CR rate and 8% CR without platelet recovery [CRp]); myelosuppression was frequent but other toxicities were mostly grade 2 or lower. Four patients (7%) died during induction.

In another study, patients aged 60 years or older with untreated AML were randomized to a lower dose of clofarabine monotherapy (30 mg/m2, Days 1 through 5) or the same dose of clofarabine plus low-dose cytarabine (20 mg/m2 daily, Days 1 through 14).77 The addition of low-dose cytarabine to this lower dose of clofarabine improved the CR rate compared with clofarabine monotherapy.78 Unfortunately, due to the Bayesian randomization strategy, very few patients were treated with single-agent clofarabine. A phase 2 trial of single-agent clofarabine (30 mg/m2/day given intravenously every day for 5 days, every 28 days) in older patients (age >65 years) with previously untreated AML with a nonfavorable cytogenetic profile for whom standard intensive treatment is deemed unsuitable has been conducted in Europe.79 Among the 66 enrolled patients, 69% had intermediate-risk cytogenetics and 31% had adverse-risk cytogenetics. Twenty-five percent had secondary AML and 74% presented with 1 or more comorbidities. Preliminary efficacy data demonstrated an encouraging 48% overall CR and CRp rate and little difference in response rates among patients with intermediate versus unfavorable karyotypes.79 Based on these results, a phase 2 study in patients aged 60 years old or older with a high mortality risk not fit for intensive therapy currently is under way in the U.S.

Tipifarnib has been used in patients age older than 65 years with previously untreated AML and myelodysplastic syndrome (MDS). In a recent study, 14% of 158 elderly patients with poor-risk AML achieved a CR.80 Partial remission or hematologic improvement occurred in 15 patients for an overall response rate of 23%. The median duration of CR was 7.3 months and the median survival of complete responders was 18 months. Adverse karyotype, age ≥75 years, and poor performance status were found to be negatively correlated with survival. Early death in the absence of progressive disease was rare; drug-related nonhematologic serious adverse events were observed in 74 patients (47%).80

Cloretazine, a sulfonylhydrazine alkylating agent, has demonstated activity as monotherapy in first recurrence of AML after an initial CR lasting less than 12 months.81 An ongoing phase 2 trial is evaluating cloretazine in patients aged 60 years or older with chemotherapy-naive AML or high-risk MDS. A total of 107 patients were treated.82 The CR rate was 47% in de novo AML and 11% in secondary AML. Induction mortality was 18%. An ongoing phase 3 trial is examining the activity of cytarabine with or without cloretazine in AML in first recurrence.

Stem Cell Transplantation

Allogeneic SCT for early consolidation therapy in patients with AML in first remission has been evaluated since early studies of transplantation. Several studies have attempted to define the optimal transplant strategy. In prospective trials, patients age younger than 55 years, in first CR, and with an human leukocyte antigen (HLA) identical sibling were assigned to allogeneic transplantation (Table 3). Those with no donor were randomized to receive an autologous transplantation or further chemotherapy.83–86 Overall, these trials suggested a possible advantage for allogeneic transplantation in patients with poor-risk cytogenetics, although long-term results remain poor. To our knowledge, there is no clear advantage with any of these modalities in patients with other cytogenetic risk groups with the exception of those with inv(16) and t(8;21), who fare better with chemotherapy. A recent study of 872 patients with diploid AML treated in 4 German trials evaluated the prognostic impact of NPM1, FLT3, C/EBPα, MLL, and NRAS gene mutations on survival after different postremission therapies.26 They reported a strong benefit for allogeneic SCT in first CR only in the subgroup of patients without the NPM1+FLT3-ITD marker constellation.26 Therefore, in patients with diploid cytogenetics, allogeneic SCT should be considered in first CR only for those without NPM1+FLT3-ITD.

Table 3. Selected Large Randomized Trials of Transplantation in First CR in Patients With AML
ReferencesNo. of patients% Achieving CR% Receiving planned therapyEFS at 4 years, %
  1. AML indicates acute myeloid leukemia; CR, complete remission; EFS, event-free survival; Allo, allogeneic stem cell transplantation; Auto, autologous stem cell transplantation; Chemo, chemotherapy.

Zittoun et al., 1995839416655554830
Burnett et al., 19988518578138 5440
Harousseau et al., 19971005177160444440
Suciu et al., 200310111986969, 565242 
Cassileth et al., 1998847407081, 54, 91433535

Newer transplantation strategies such as nonmyeloablative transplants, use of intravenous busulfan, and better management of graft-versus-host disease have permitted extension of the transplant-eligible age, and may lead to a better efficacy/toxicity ratio for this procedure in the future.87–89 To our knowledge to date, there are no randomized trials confirming the benefit of nonmyeloablative SCT in older patients with AML. However, retrospective comparative studies have demonstrated that, in older patients, this strategy may be superior to the traditional transplants.90 Furthermore, the reduced toxicity associated with these less intensive regimens may allow their use in high-risk groups otherwise not suited for transplants.91 Although autologous transplant is used commonly for consolidation of remission, no randomized data to date have demonstrated a survival benefit compared with consolidation chemotherapy.


Recent developments in understanding the biology of AML have provided multiple targets for developing agents with potential activity against the disease. It is likely that a number of these agents, singly or in combination, will be effective against subgroups of patients with AML. It is also likely that these agents will complement traditional and newly developed cytotoxic agents used for treating AML. New strategies, such as maintenance therapy to suppress minimal residual disease using 1 or more of these agents, may prove effective in improving survival.