Acute myelogenous leukemia

Prognosis and genetics of AML 

Remission rates with standard induction chemotherapy in patients with AML range from 50% to 85% [1]. However, the majority of patients will relapse and die of their disease within 2 years of achieving a remission. Remission rates and overall survival depend on a number of features, including age of the patient, cytogenetics, other molecular changes within the malignant leukemic clone, previous bone marrow disorders (e.g., myelodysplasia [MDS] or a myeloproliferative disorder), and comorbid illnesses as well as others.

Prognosis and genetics of AML are tightly linked. Risk stratification based on cytogenetics divides patients into three main groups, those with favorable, intermediate, and unfavorable cytogenetics. Although the specific chromosomal aberrations within each group are not entirely consistent among all studies, a general consensus exists (Table 1) [2].

Table 1. Cytogenetic-based risk stratification

		Favorable risk	Intermediate risk	Unfavorable risk
	Chromosomal	t(15;17)(q22;q12-21)	Normal karyotype	Complex karyotype
	aberration	t(8:21)(q22;q22)	t(9;11)(p22;q23)	inv(3)(q21q26)/t(3;3)(q21;q26)
		inv(16)(p13q22)/	del(7q)	t(6;9)(p23;q34)
		t(16;16)(p13;q22)	del(9q)	t(6;11)(q27;q23)
			del(11q)	t(11;19)(q23;p13.1)
			del(20q)	del(5q)
			−Y	−5
			+8	−7
			+11
			+13
			+21

Data from reference 2.

Age and cytogenetics have a close relationship. Adverse cytogenetic abnormalities increase with increasing age and, within each cytogenetic risk group, prognosis with standard treatment worsens with increasing age. One study found that the percentage of favorable cytogenetics dropped from 17% in patients aged younger than 56 years to 4% in those aged older than 75 years. Furthermore, the percentage of patients with unfavorable cytogenetics increased from 35% in those younger than 56 years to 51% in patients older than 75 years 1, 3.

Core-binding factor (CBF) AML is a frequent subtype of AML with a relatively favorable prognosis. CBFs are a group of heterodimeric transcriptional regulators containing a common β (CBFβ-PEPBP2β) and one of three α (AML1-RUNX1, PEBP2αB, CBFα2) components. CBF AML results from translocations involving either AML1 or CBFβ. In t(8;21) AML, ETO on chromosome 8 is fused with RUNX1 on chromosome 21 and is associated with the French-American-British M2 morphology. In inv(16) AML, CBFβ on chromosome 16 is fused to the MYH11 gene on chromosome 16 and is associated with French-American-British M4Eo morphology 4, 5. Each of these fusion products is thought to lead to leukemia through dominant negative inhibition of normal myeloid differentiation. Despite a similar and favorable response rate among these two cytogenetic subsets of AML, it has recently been recognized that they differ with respect to their pattern of secondary chromosomal aberrations, their gene-expression signatures, and how various KIT mutations affect prognosis 6, 7, 8.

Approximately 40% to 50% of patients with AML have a normal karyotype and represent the largest subset of AML [6]. Not all patients in this subset have the same response to treatment. This is likely a result of the large variability in gene mutations and gene expression in this population. Mutations in the nucleophosmin, member 1 gene and the CCAAT/enhancer-binding protein-α gene seem to confer a better prognosis, whereas internal tandem duplications of the FMS-like tyrosine kinase 3 (FLT3), partial tandem duplication of the myeloid/lymphoid or mixed lineage leukemia gene, overexpression of the brain and acute leukemia gene, and overexpression of the ETS-related gene are associated with a poorer prognosis [9]. Although further research is needed, eventually gene-expression profiling may be used to help prognosticate patients with cytogenetically normal AML. It may also aid in guiding treatment, as each of these genetic alterations can ultimately become targets of therapeutic interventions.

Standard treatment 

Induction chemotherapy with 7 days of continuous intravenous infusion of cytarabine 100mg/m2 and 3 days of daunorubicin 45 to 60mg/m2 (“7 and 3”) followed by 5 days of continuous intravenous infusion of cytarabine and 2 days of daunorubicin, if disease persists based on a day-14 bone marrow biopsy, has been the standard initial treatment for AML for the past 3 decades [10].

Other induction regimens, including altering the doses of cytarabine, substituting different anthracyclines, and the addition of other agents to enhance the efficacy of this standard treatment, have been studied. Despite theoretical benefits, none have convincingly shown a survival advantage over “7 and 3” (Table 2). Whether to use a daunorubicin dose of 60mg/m2 or 45mg/m2 in this standard regimen has recently been a subject of much debate. While some have advocated use of the higher dose, no randomized trial has evaluated outcomes between the two dose levels.

Table 2. Trials of alternative induction regimens

	Induction regimen	Complete response	Overall survival	Other
	HDAC vs standard Ara-C 11, 12	No improvement	No improvement	Increased toxicity
	Ara-C (200 vs 100mg/m2) [13]	No improvement	No improvement
	Addition of etoposide to standard therapy [14]		No improvement (possibly improved in age <55 years)	Improved remission duration
	GCSF+priming [15]	No improvement	No improvement	May improve event-free survival
	Idarubicin vs daunorubicin 15, 16, 17, 18	Improved in two trials	Improved in two trials No improvement in 2 trials	Conflicting results in four trials
	Mitoxantrone vs daunorubicin [19]	Improved	No improvement
	Substitution of fludarabine for anthracycline [20]	Decreased	Decreased
	Adding fludarabine to Ara-C and anthracycline [21]	Inconclusive	No improvement
	Addition of MDR-1 modulator 22, 23	No improvement	No improvement
	Addition of gemtuzumab ozogamicin (3mg/m2) [24]		No improvement	Improved disease-free survival

Ara-C = cytarabine; GCSF = granulocyte colony-stimulating factor; HDAC = high-dose cytarabine; MDR-1 = multidrug resistance 1.

Postremission therapy 

Several studies have evaluated the role of postremission therapy in AML. These studies have demonstrated a clear benefit in survival and cure rates for patients who receive some form of accepted consolidation treatment 25, 26, 27, 28.

Postremission therapy traditionally has included three standard acceptable modalities, i.e., more chemotherapy, autologous stem cell transplantation, or allogeneic bone marrow transplantation (BMT). Intensity of the postremission therapy is typically dependent on the age of patient, comorbidities, chance for recurrence based on cytogenetics, and whether or not the patient has a suitable bone marrow donor.

Consolidation chemotherapy 

Standard consolidation chemotherapy for patients younger than 60 years old, provided the patient has good organ function, is high-dose Ara-C (cytarabine) [29]. Standard treatment doses are 3g/m2 twice a day on days 1, 3, and 5, for a total of six doses. Typically, three to four cycles of high-dose Ara-C are given. Patients who fall in the favorable or intermediate-risk groups appear to have the most benefit from this intensive consolidation regimen. From the available studies, it is clear that those patients who have favorable cytogenetics do equally well or better with intensive postremission chemotherapy as initial treatment. Patients who fall in the unfavorable risk group may not benefit from this intensive consolidation chemotherapy and are typically considered candidates for allogenic transplantation 29, 30, 31.

Patients older than 60 years of age do not benefit from the 3g/m2 dose of HiDAC regimen over other less intense forms of consolidation. Incidence of toxicity, including cerebellar toxicity, is higher in this age group, therefore, making it an unacceptable option 29, 30, 31.

A variety of consolidation regimens have been used for patients older than 60 years. One has not been shown to be clearly better than another. A reasonable approach is for two cycles of less-intense chemotherapy, such as 5 days of cytarabine 100 mg/m2 continuous intravenous infusion for 24hours and 2 days of daunorubicin 45mg/m2.

Autologous BMT as postremission therapy 

Several trials have evaluated autologous stem cell transplantation in first remission. Typically, patients who have been studied are those who do not have a human leukocyte antigen (HLA)−matched donor. Large trials that have randomized patients to consolidation chemotherapy vs autologous stem cell transplantation during first remission have consistently shown an improvement in disease-free survival, however, no overall survival advantage has been demonstrated 29, 30, 31, 32, 33, 34. A large meta-analysis published by Nathan et al. [32] of six randomized trials with 1044 patients comparing autologous BMT to nonmyeloablative chemotherapy also showed similar findings. This lack of survival advantage, despite improved disease-free survival, has been attributed in part to the ability to effectively salvage patients with autologous or allogeneic BMT after relapse from chemotherapy. Mortality for autologous BMT is <6%. The higher mortality of autologous BMT compared with consolidation chemotherapy as well as the improved efficacy in consolidation chemotherapy with HiDAC has also diminished the ability for autologous BMT to show a survival advantage when compared to standard consolidation chemotherapy [32].

Allogeneic BMT as postremission therapy 

Patients with unfavorable cytogenetics have traditionally been candidates for allogeneic BMT, given the poor results with standard consolidation chemotherapy. However, the results of randomized trials have been conflicting for its benefit. There have been five cooperative group trials in which the role of allogeneic BMT has been assessed in patients with AML in first remission 29, 30, 31, 33, 35, 36, 37. No prospective trials have been specifically designed to address an unfavorable risk group, allowing for interpretation to come only from the subset analyses of these larger trials. Conclusions from the subset analyses have been hampered in these studies because of lack of compliance of patients with the assigned treatment option, small amounts of patients, and the changing definition of an unfavorable risk group. Nonetheless, three trials do not suggest a survival benefit for allo-BMT for patients in first remission with unfavorable risk cytogenetics (i.e., European Organisation for Research and Treatment of Cancer/Gruppo Italiano Malattie Ematologiche dell'Adulto, AML 8/GOELAM [Groupe Ouest-Est Leuremies Aigues Myeloblastique], and Medical Research Council AML 10), while two trials do show a statistically significant benefit in overall survival with this modality (e.g., European Organisation for Research and Treatment of Cancer/Gruppo Italiano Malattie Ematologiche dell'Adulto AML 10, US Intergroup). However, because of the usual dismal results with standard chemotherapy in high-risk groups of patients with AML and the fact that two of the five trials have shown a survival benefit of allo-BMT in first remission in subset analysis for high-risk AML patients, and given that the lack of survival benefit may be due in part to the small numbers of patients, allogeneic BMT is the standard offered to high-risk patients in first remission.

Data is less clear for allo-BMT in patients with intermediate risk cytogenetics who are in first remission. The UK Medical Research Council AML trial did show an overall survival benefit for patients with intermediate-risk cytogenetics when allo-BMT was used in first remission, but this was a subset analysis and other studies have been contradictory [38].

Relapsed disease 

Relapsed AML after initial chemotherapy in patients who cannot tolerate an autologous or allogeneic BMT is almost always fatal. However, patients who have a matched HLA-identical sibling donor, a matched unrelated donor, or those who do not but who can tolerate an autologous BMT, can still be salvaged. The standard approach is for patients to undergo reinduction chemotherapy in an attempt to obtain a second remission prior to proceeding to BMT. The single best predictor of success for obtaining a second remission is the length of initial remission duration. If the initial complete remission (CR) was <12 months, the chance for obtaining a second CR is 10% to 20%. If the initial CR was >12 months, success for obtaining a second CR increases to 40% to 50% 39, 40, 41. Reinduction with “7+3” is very reasonable if a durable remission was obtained upon this initial induction regimen. Alternative reinduction regimens have been used with similar success rates, with no one regimen proving superior to the others 42, 43, 44, 45.

Autologous BMT has been extensively evaluated for patients in second remission who cannot tolerate an allogeneic BMT or do not have a suitable donor. A recent review of follow-up data in the British Society of Blood and Marrow Transplantation registry database to establish long-term outcomes of autologous transplantation in patients with intermediate and good risk AML in second CR reported 10-year survival rate of 32% [46].

Allogenic BMT with an HLA-matched sibling after the first remission is lost has consistently shown 5-year survival rates of 20% to 35%. Patients do better if the duration of first remission is >6 months and if they are in a second remission prior to allo-BMT [47].

Matched unrelated allo-BMT is also an option for patients who can tolerate it, but mortality (approaching 30%) and severe graft-vs-host disease incidence are both high with this form of transplantation [48]. Although Caucasians have a 50% chance of obtaining a match, it is only a 10% chance for minorities, and frequently there is a delay to obtain donor cells, which can allow for relapse prior to the patients getting the allograft [49].

Nonmyeloablative and reduced-intensity allogenicBMT 

The goal of both reduced-intensity conditioning regimens and nonmyeloablative regimens is to reduce the nonrelapse mortality of the allogeneic BMT, thereby allowing this treatment option to be offered to patients who otherwise could not tolerate a conventional myeloablative transplant, either due to age or comorbid illnesses. Both of these strategies are employed routinely for older individuals and appear effective, particularly in patients who are in a complete remission prior to either strategy [50].

Gemtuzumab ozogamicin 

Gemtuzamab ozogamicin (GO) is an anti-CD33 immunoconjugate that binds to the surface of myeloid blasts and, after internalization, releases calicheamicin, a cytotoxic drug [51]. GO is approved by the US Food and Drug Administration for use in patients 60 years or older with CD33+ AML in first relapse who are not candidates for cytotoxic chemotherapy. Although this novel drug theoretically has advantages of being a “targeted” treatment, its CR rate as a single agent was only 13%. Another 13% of patients had a CR with incomplete platelet recovery with median recurrence-free survivals of 6.4 months for patients who received a CR and only 4.5 months for patients who received a CR with incomplete platelet recovery. Infusion-related toxicity, such as fever, chills, shortness of breath, and hypotension, is appreciable and grade 3 to 4 neutropenia and thrombocytopenia occurred in all patients. Grade 3 to 4 elevations in bilirubin were seen in 29%, and 0.9% of patients who did not undergo prior or subsequent hematopoietic transplantation developed hepatic veno-occlusive disease, therefore, GO should be avoided in patients with preexisting liver pathology [52].

Higher survival rates have been seen when patients are able to subsequently go on to an allogeneic or autologous BMT, however, these patients have a much higher incidence of hepatic veno-occlusive disease, particularly when the BMT is done within 4 months of GO administration [52].

AML in the elderly 

There is no agreed upon age in which induction chemotherapy cannot be offered to patients with AML. However, once a patient is deemed too frail because of age or comorbidities for induction chemotherapy, treatment is focused mainly on supportive measures, such as transfusions when needed and antibiotics for treatable infections. Low doses of oral chemotherapy, such as hydroxyurea or melphalan, have been used palliatively to reduce the leukemia burden in patients. Low doses of cytarabine given either intravenously or subcutaneously have also been used palliatively with moderate success 53, 54, 55.

However, ultimately, these patients will die of their disease, typically within a matter of weeks to months. This population of patients may stand to benefit the most from novel agents that can be less toxic and more effective than current agents.

Novel agents 

In recent years, we have begun to understand the molecular biology of AML. The identification of specific gene mutations and their protein products, as well as alterations in gene transcription that result in aberrant regulation of the cell cycle, have led the way for development of targeted therapies for AML. The remainder of this review will discuss the results of clinical trials with many of these novel agents in AML.

FLT3 tyrosine kinase inhibitors 

FLT3 is a receptor tyrosine kinase that is normally expressed by hematopoietic stem/progenitor cells. It plays a role in cell survival, proliferation, and differentiation. FLT3 mutations occur in approximately 30% of AML in adults and confer a poor prognosis. The two major types of mutations that occur are internal tandem duplication mutations of the juxtamembrane region and point mutations, which frequently involve aspartic acid 835 of the kinase domain. Both mutations result in constitutive activation of the receptor's tyrosine kinase activity in the absence of ligand [56].

There are currently four FLT3 inhibitors undergoing clinical investigation: PKC-412 (Novartis, Summit, New Jersey), CEP-701 (Cephalon, Frazer, PA), MLN518 (Millenium, Cambridge, MA), and SU11248 (SuGen, San Francisco, CA). A number of phase I and II studies have been completed using these agents as monotherapy in relapsed/refractory AML. From these studies, we have learned that FLT3 inhibitors have demonstrated a clinical response in relapsed AML with activating mutations of FLT3 57, 58, 59, 60, 61, 62, 63, 64, 65, 66. There also appears to be a response in some patients without activating mutations, which may be explained by mutations at alternative sites that activate FLT3. Most studies have demonstrated only a transient reduction in peripheral blood and bone marrow blasts. However, preclinical studies have shown synergistic killing of leukemic cells in vitro when FLT3 inhibitors are combined with conventional chemotherapy [67]. A number of clinical trials are now underway evaluating FLT3 inhibitor therapy with chemotherapy. A phase III study of CEP-701 with MEC (mitaxantrone, etoposice, cytarabine) or HiDAC in relapsed FLT3 mutant AML is being conducted [68], as well as a study of PKC-412 with Ara-C and daunorubicin induction, and consolidation with HiDAC in newly diagnosed AML [69]. A phase I/II study of MLN518 plus standard induction with Ara-C and daunorubicin in newly diagnosed AML is also being investigated [70].

Farnesyl transferase inhibitors 

The 21-kD proteins encoded by the K-Ras, N-Ras, and H-Ras proto-oncogenes regulate the growth and differentiation of many cell types. Ras proteins that have been recruited to the plasma membrane act by transducing extracellular signals to the nucleus via multiple potential metabolic pathways. Point mutations of Ras genes at codons 12, 13, and 61 result in constitutively active Ras pathways. The incidence of Ras mutations in myelodysplastic syndromes and AML has been reported between 3% and 40%, with N-Ras mutations being the most common, followed by H-Ras, which is rarely seen. Farnesyl transferase inhibitors work by inhibiting farnesylation of Ras, a step that is required for the transfer of Ras to the plasma membrane [71]. Preclinical data indicated that R115777 (Zarnestra) inhibited proliferation and induced apoptosis in leukemic blast cells in vitro 72, 73. A phase II study of tipifarnib in 158 poor-risk elderly patients with untreated AML showed an overall response rate of 23%, with 14% of patients achieving a complete response [74]. There is currently a Southwest Oncology Group phase II study evaluating tipifarnib in previously untreated AML patients age 70 or older, as well as an Eastern Cooperative Oncology Group study evaluating tipifarnib as maintenance therapy after consolidation in AML patients in second or higher CR. A recent trial evaluated tipifarnib in combination with idarubicin and cytarabine in 74 newly diagnosed AML or high-risk MDS patients. Overall response was 77% with 65% achieving a complete response [75].

Transcription modulators (DNA demethylating agents and histone deacetylase inhibitors) 

Genetic changes refer to changes in the DNA sequence, whereas epigenetic changes refer to changes in gene transcription. Two well-known epigenetic modifications include DNA methylation and various histone modifications, which alter gene expression via the architectural remodeling of chromatin. It is believed that these mechanisms play a role in carcinogenesis through the silencing of tumor-suppressor genes. DNA demethylating agents and histone deacetylase inhibitors (HDAC) induce reexpression of tumor suppressor and proapoptotic genes [76].

5-Azacytidine and 5-aza-2′-deoxycytidine (Decitabine) are the most studied DNA demethylating agents in AML. Phase I trials of these agents used as monotherapy in relapsed/refractory AML or in patients not eligible for intensive chemotherapy have shown complete and partial responses. One study showed complete responses in 2 of 17 (12%) patients and complete or partial remissions in 11 of 17 (65%) patients 77, 78. Decitabine is currently being evaluated in a phase II Cancer and Leukemia Group B study as maintenance therapy following standard induction and consolidation therapy.

HDAC inhibitors that have been evaluated in clinical trials of AML and MDS include MS-275, MG-0103, sodium phenylbutyrate, Vorinostat (suberoyanilide hydroxamic acid), valproic acid, and depsipeptide (FK-228). Vorinostat used as monotherapy in a phase I trial of relapsed/refractory AML or MDS demonstrated complete and partial responses in 21% of patients [79]. Valproic acid has modest activity as monotherapy in low-risk MDS, and in AML in combination with all-trans retinoic acid 80, 81. Depsipeptide (FK-228) was shown to inhibit HDAC in vivo in a phase I trial in chronic lymphocytic leukemia and AML, but its administration was limited by progressive constitutional symptoms [82].

Several phase I/II studies have evaluated the combination of DNA demethylating agents with HDAC inhibitors. Results have indicated that the combination of these agents is safe and can effectively induce complete and partial remissions 83, 84, 85. Based on these promising results, a large intergroup study was initiated evaluating azacytidine with or without the HDAC inhibitor MS-275 for MDS, chronic myeloid leukemia, and AML with multilineage dysplasia.

Multidrug resistance-1 modulators 

Multidrug resistance-1 gene encodes a 170-kDa P-glycoprotein (P-gp) that functions as an adenosine triphosphate-dependent export pump. This pump transports numerous antineoplastic agents out of the cell. P-gp expression in AML cells increases with advancing age and represents an independent poor prognostic variable. Numerous antineoplastic agents are substrates of P-gp, including anthracyclines, vinca alkaloids, taxanes, camptothecins, and epipodophyllotoxins. Most multidrug resistance-1 modulators function as competitive inhibitors by binding to P-gp within its substrate channel, interfering with the binding of antineoplastics [86].

The first-generation modulators included quinine and cyclosporine. Studies with these agents did not show a significant benefit, however, a Southwest Oncology Group trial of cyclosporine with cytarabine and daunorubicin demonstrated increased relapse-free and overall survival [87]. This led to trials with second-generation modulators, of which PSC-833 (Valspodar; Novartis) was the only one to complete phase III trials. Results of these studies were disappointing, showing no improvement in the rate of CR and excessive toxicity [88].

A number of third-generation modulators are currently under investigation. These agents include Tariquidar (XR9576; Xenova, Cambridge, MA), Zosuquidar (LY335979; Eli Lilly, Indianapolis, IN), Laniquidar (R101933), and ONT-093. These agents are more specific for P-gp, have less CYP-3A4 interaction, and have less pharmacokinetic interactions with chemotherapy than second-generation agents [86]. After a phase II study of Zosuquidar with cytarabine and daunorubicin that showed promising results [89], the Eastern Cooperative Oncology Group initiated a phase III, double-blind, randomized trial evaluating Zosuquidar with standard induction and consolidation chemotherapy.

BCL-2 antisense oligonucleotide (Genasense) 

Bcl-2 is an apoptosis-regulating oncogene that is frequently overexpressed in lymphoid and myeloid malignancies. Normally, apoptotic signals are transmitted from the cell membrane to the mitochondria, resulting in release of cytochrome C, which induces caspase-mediated apoptosis. The bcl-2 protein stabilizes the inner mitochondrial membrane, preventing release of cytochrome C, which inhibits apoptosis [90]. Overexpression of the bcl-2 protein leads to resistance of tumor cells to apoptosis, and is associated with a poor prognosis in AML [91]. Oblimersen (Genasense; bcl-2 antisense oligonucleotide) binds to bcl-2 messenger RNA in the cytoplasm, which results in messenger RNA degradation and decreased level of the bcl-2 protein.

In a phase I study of Oblimersen with fludarabine, cytarabine, and granulocyte colony-stimulating factor in relapsed or refractory AML and acute lymphocytic leukemia, 9/20 (45%) of patients responded with 6 (5 AML, 1 acute lymphocytic leukemia) achieving a complete response [92]. In a subsequent phase I trial, Oblimersen was evaluated with standard cytarabine and daunorubicin induction therapy, and high-dose Ara-C consolidation therapy in previously untreated AML patients older than 60 years of age. In this study, 14 of 29 (48%) patients achieved a CR [93]. Based on these promising results, a randomized phase III study was initiated by the Cancer and Leukemia Group B evaluating oblimersen with cytarabine and daunorubicin induction, and with high-dose Ara-C consolidation in previously untreated patients older than 60 years with AML.

Antiangiogenesis agents 

The role for antiangiogenesis therapy in AML was prompted by recognition that bone marrow biopsies from patients with AML demonstrated increased microvessel density and that increased levels of vascular endothelial growth factor correlate with a poor prognosis [94]. Clinical trials involving a variety of antiangiogenesis agents alone or in combination with chemotherapy have largely been disappointing. These trials have included the agents PTK787/ZK222584, AG-013736, SU11248, SU5416, bevacizumab, and thalidomide 95, 96, 97, 98, 99, 100, 101, 102. However, a trial of bevacizumab, given after chemotherapy with high-dose cytarabine and mitoxantrone in patients with relapsed and refractory AML, demonstrated an overall response rate of 48% (23 of 48 patients) and a complete response rate of 33% (16 of 48 patients) [99].

Clofarabine 

Clofarabine is a second-generation nucleoside analog. It is a potent inhibitor of ribonucleotide reductase and is resistant to deamination by adenosine deaminase 91, 103. Clinical trials have demonstrated promising results when clofarabine is used alone or in combination with cytarabine in patients with AML or high-risk MDS. In a phase II trial of clofarabine monotherapy in patients with relapsed or refractory acute leukemia, 17 of 31 patients (55%) with AML had complete or partial responses [104]. In a subsequent phase I/II study of clofarabine plus cytarabine in relapsed/refractory acute leukemia, 12 of 29 (41%) patients with AML/high-risk MDS responded, with 24% achieving a CR [103]. In a recent phase II study of clofarabine plus cytarabine in previously untreated AML patients aged 50 years and older, an overall response rate of 60% (52% CR, 8% CRp) was achieved [105]. Based on these promising results, a randomized phase III trial of single-agent clofarabine compared to standard chemotherapy with daunorubicin and cytarabine in previously untreated elderly patients has been initiated.