Introduction

Acute myeloid leukemia (AML) is a disease that primarily affects older adults, and the median age at diagnosis is over 65 years [1–3]. The cut-off age for differentiating younger from older AML is arbitrary; age 65 or older has traditionally been used as the eligibility criterion for previous studies of elderly AML by the Japan Adult Leukemia Study Group, whereas other study groups have chosen, among others, 60 or older, 70 or older, or 50–70 years. For practical purposes, however, age 60 or over is generally used to define elderly AML [4, 5]. Elderly AML is a biologically and clinically distinct disease with a diminished response to chemotherapy. Previous clinical trials of intensive chemotherapy showed rates of complete remission (CR) around 50 % and of long-term survival at less than 10 %, which are much worse than for younger patients [1–3]. Furthermore, such data likely overestimate the true outcome for elderly AML, as patients entered into clinical trials are screened using criteria such that they often do not represent the general patient population. In contrast with the progress made for younger adults, the outcome of treatment of elderly AML has improved little, if at all, in recent decades [6, 7]. The adverse prognostic impact of older age is attributable to differences both in disease-related factors (i.e., cytogenetics, secondary AML, and expression of the multidrug resistance phenotype) and patient-related factors (i.e., general condition, organ dysfunctions, and comorbidities). In addition to these therapeutic drawbacks, the net incidence of elderly AML is expected to increase as the population continues to age, making the management of elderly AML an even more critical issue.

Characteristics of AML in older adults

The unfavorable biologic characteristics of AML amplified in older adults, such as a higher proportion of unfavorable cytogenetics, higher frequency of antecedent hematologic disorders or previous treatment for one or more other malignancies, and more frequent expression of the multidrug resistance phenotype. Cytogenetic findings at diagnosis have important prognostic implications for both younger and older patients [8–10]. Favorable cytogenetic characteristics, e.g., core binding factor (CBF) abnormalities as defined by t(8;21) or inv(16)/t(16;16), are relatively uncommon in older adults, and are seen in less than 5 % of patients aged over 60 years [8–10]. In contrast, unfavorable cytogenetics represented by complex karyotype is predominant in older patients. Secondary AML arising from myelodysplastic syndrome (MDS) or myeloproliferative neoplasm (MPN), or AML related to prior chemotherapy for previous malignancies, both of which are known as subtypes with increased resistance to chemotherapy, is also common in this age group [1–3]. Response to chemotherapy is affected by the expression of genes that confer drug resistance, such as the multidrug resistance 1 (MDR1) gene that encodes the chemotherapy efflux pump P-glycoprotein. Overexpression of P-glycoprotein reportedly occurs in 71 % of older patients, compared to only 35 % of younger patients [11].

Patient-related factors, such as poor general condition, significant comorbidities, and diminished functional reserves, also contribute to the poorer outcomes for older patients. Moreover, because of their reduced performance status (PS) and increased prevalence of significant comorbidities, older patients are less tolerant of complications associated with chemotherapy. These conditions often make physicians reluctant to administer intensive chemotherapy to older patients. Indeed, according to a survey conducted in the United States, chemotherapy was administered to only 30 % of patients over the age of 65 years [12]. Although selected patients can tolerate and benefit from intensive chemotherapy, older adults as a whole are more likely to experience treatment-related mortality (TRM) and less likely to benefit from standard induction and post-remission therapies.

Taken together, these findings indicate that the effects of age in terms of both patient- and disease-related factors result in a higher incidence of early death during induction therapy, a lower rate of CR, and a reduced chance of long-term survival.

Prognosis and prognostic factors

The prognosis of AML worsens with age. When treated with intensive chemotherapy, CR rates for older adults range between 40 and 60 % [1–3], which is much lower than those for younger adults. Even if CR is achieved, older adults are more likely to experience relapse, leading to an extremely low expectation for long-term survival. Such a poor chance of treatment success is combined with a high risk of TRM, which ranges from 10 to 40 % even for selected older adults [1–3].

Because of intrinsic resistance to chemotherapy and excessive toxicity, the benefits associated with standard chemotherapy for older adults remain debatable. Since these patients represent a heterogeneous population, it is clinically important to identify those who are and are not likely to benefit from standard chemotherapy. This medical requirement has prompted the generation of several prognostic models incorporating multiple covariates (Table 1) [13–17]. These systems may help determine whether a patient should receive standard chemotherapy or alternative treatment. Among the most important prognostic factors are age, cytogenetics, and PS. In addition, some recent studies have shown genetic mutations involving the fms-like tyrosine kinase 3 (FLT3) gene, the nucleophosmin 1 (NPM1) gene, and others are as prognostically relevant for older as they are for younger patients [18–21], while other studies have shown that gene expression profiles can identify prognostically distinct subgroups [22, 23]. These recent findings nonetheless need to be validated before they can be adopted for therapeutic decision-making.

Table 1 Covariates constituting predictive models for overall survival in older adults with AML
Full size table

The PS classification is commonly used for assessing whether or not a patient is fit for intensive chemotherapy [4, 5]. However, this classification may be suboptimal, as it does not differentiate between functional impairments resulting from leukemia and those resulting from unrelated comorbidities. This distinction is important, as the former can be reversible with treatment, while the latter are not. In this respect, comorbidity scoring rather than PS assessment may be a potentially more accurate basis for therapeutic decisions [24].

Treatment

Induction therapy

Since a randomized study reported by Lowenberg et al. [25] in 1989 demonstrated that intensive chemotherapy could change the natural history of AML in elderly patients, induction therapy has been considered a viable option in the management of elderly AML. A recent population-based study in Sweden indicates that this may also be the case with unselected patients [26]. Another retrospective study using a propensity score matching analysis suggests that intensive chemotherapy has a beneficial effect on OS even for patients aged 70 years or older [27].

The standard induction therapy consists of an anthracycline combined with cytarabine (AraC), known as “3 + 7”, a regimen that has been the mainstay for more than three decades. For selected patients, standard induction therapy yields CR rates of around 50 % [1–3]. As shown in Table 2, a multitude of attempts have been made in the hope of improving the outcome, such as escalating doses, replacing or adding drugs, and using growth factors [25, 28–47], but most if not all failed to demonstrate survival benefit. However, a number of studies have reported positive results. Schlenk et al. [34] evaluated the effect of all-trans retinoic acid (ATRA) administered in combination with induction and consolidation therapy to 242 patients more than 60 years old. They showed that addition of ATRA significantly improved CR rate, event-free survival (EFS) and overall survival (OS). Lowenberg et al. [41] compared the effect of a doubled dose of daunorubicin (DNR) of 90 mg/m2 with that of a conventional dose of 45 mg/m2 in the context of the “3 + 7 regimen” administered to 813 patients aged 60 or older. Although no overall difference in outcome was observed, patients who were between 60 and 65 years old significantly benefited from the doubled dose of DNR. Very recently, Castaigne et al. [46] reported results from a phase 3 study that examined the effect of adding gemtuzumab ozogamicin (GO). In their study, a total of 280 patients aged 50–70 were randomly assigned to receive GO or a placebo in combination with chemotherapy during induction and consolidation courses. While CR rates for the two arms were not different, GO significantly improved relapse-free survival (RFS), EFS, and OS. Since none of these positive results have as yet been reproduced elsewhere, definitive conclusions regarding the effectiveness of these treatments await the results of further validation studies.

Table 2 Selected randomized studies using induction therapy for AML in older adults
Full size table

For patients unfit for intensive chemotherapy, treatment options include less-intensive chemotherapy, best supportive care, or enrolment in clinical trials. For less-intensive chemotherapy, low-dose AraC has been utilized since the 1960s. Burnett et al. [37] recently demonstrated in a phase 3 study that low-dose AraC significantly prolonged OS when compared to best supportive care for older patients who were considered unfit for intensive chemotherapy. Stratification by cytogenetic risk made it clear that low-dose AraC was beneficial for patients with favorable and intermediate cytogenetic risks. Despite the significant superiority of low-dose AraC over best supportive care, outcome with low-dose AraC remained far from satisfactory, as the probabilities of CR and 1-year OS were only 18 and 25 %, respectively [37].

Post-remission therapy

Once CR has been reached after induction therapy, however, there is no standard post-remission strategy for elderly AML. Table 3 summarizes selected randomized studies using post-remission therapy [30, 31, 36, 38, 40, 43, 47–51]. None of these studies have shown improved survival resulting from treatment intensification: increasing the dose of AraC did not result in better outcome when compared to the standard dose [48], and results for four cycles of consolidation therapy were not superior to those for three cycles [40], or even for a single cycle [31]. However, prolonged therapy with lower doses of chemotherapy may be beneficial for elderly patients. Gardin et al. [36] compared consolidation therapy consisting of six less-intensive courses administered monthly on an outpatient basis with one course of intensive chemotherapy. After achieving CR, 164 patients aged 65 years or older were randomized to either the outpatient or intensive arm, with the former showing significantly superior DFS and OS. In addition to this finding, several studies showed the beneficial effect of prolonged maintenance therapy for the prevention of relapse, although they did not show significant survival advantage [30, 49]. On the other hand, Schlenk et al. [50] reported that one course of intensive consolidation chemotherapy was superior to one year of oral maintenance therapy in terms of relapse prevention and OS. The reason for this discrepancy is unclear, but may be partly explained by differences in treatment regimens used in these studies. Older adults are highly likely to be unable to complete or even start planned post-remission therapy because of comorbidities or residual toxicities from previous chemotherapy. Hence, special consideration needs to be given to the feasibility of treatment when considering post-remission therapy for older adults.

Table 3 Selected randomized studies using post-remission therapy for AML in older adults
Full size table

Hematopoietic cell transplantation

While allogeneic hematopoietic cell transplantation (HCT) offers the best chance of long-term survival for patients with poor-risk hematologic malignancies such as seen in elderly AML, it does so at the cost of higher TRM. Due to this high risk, allogeneic HCT has traditionally been limited to younger patients. However, advances in supportive care and the use of reduced-intensity conditioning (RIC) have extended the application of this procedure to older adults who would previously not have been candidates for myeloablative allogeneic HCT. Indeed, recent studies have demonstrated the feasibility of allogeneic HCT for selected older patients, with reported post-transplant survival rates of around 50 % [2, 3]. A retrospective analysis of a large number of registry data showed that older age was not associated with inferior post-transplant OS [52]. In the absence of prospective studies, the efficacy of RIC-HCT for older patients remains to be verified, although accumulated evidence suggests that RIC-HCT during CR represents a reasonable therapeutic option for selected older patients. The most critical factor that interferes with the administration of allogeneic HCT could be the low CR rate attained with induction chemotherapy. Availability of a suitable donor is another obstacle for successful HCT for older patients, because their siblings are usually also elderly. The major issues for further incorporation of allogeneic HCT into treatment strategies for elderly AML that need to be addressed are improvement of the CR rate for chemotherapy, identification of a suitable donor in a timely manner, and development of transplant procedures that lead to reduced risk of post-transplant relapse and complications.

Novel therapies

Given the limitations of the intensive chemotherapy approach for AML in older adults, current clinical trials have focused on less-intensive therapies with the potential to be efficacious without impairing patients’ quality of life. A growing understanding of the underlying molecular mechanisms of AML has led to development of novel agents, most of which selectively target leukemic cells, such as monoclonal antibodies, tyrosine kinase inhibitors, farnesyl transferase inhibitors, hypomethylating agents, histone deacetylase inhibitors, and others. Details about individual agents are reviewed extensively elsewhere in this issue of the journal. Development of less toxic and more targeted agents may provide treatment alternatives for a majority of older adults with AML. Given the dismal prognosis with currently available therapies and the limited tolerance for toxicity, such patients are promising candidates for such targeted therapies. The National Comprehensive Cancer Network (NCCN) panel has therefore recommended the use of investigational therapy in clinical trials for all patients over 60 years [5].

Conclusions

AML in older adults poses a significant therapeutic challenge because of the refractoriness of the disease and the frailty of the patients. These patients are biologically and clinically heterogeneous, which means that intensive chemotherapy may be beneficial for some, whereas it may be harmful for others. There is thus an urgent need to accurately identify patients likely to benefit from intensive chemotherapy. Such decisions should be individualized after systemic assessment of disease biology and patient characteristics. Currently, a number of novel agents are under investigation, most of which are selective for certain molecular targets and these involve less toxicity than do existing chemotherapeutic agents. The overall dismal prognosis for the treatment of elderly AML patients with currently available treatments constitutes a strong motivation for their participation in prospective clinical trials that aim to develop more effective and less toxic therapies.