FMS-like tyrosine kinase 3 (FLT3) is a transmembrane tyrosine kinase receptor that is involved in cell proliferation upon activation. Mutations in FLT3 are reported in 10% to 30% of cytogenetically normal AML. Two types of FLT3 mutations have been recognized. The most common has activating internal tandem duplications (FLT3-ITD), which along with point mutations in the activating loop, result in ligand-independent activation of FLT3. The presence of FLT3-ITD confers a worse prognosis and inferior outcomes, secondary to higher relapse rates. Current data suggests that internal tandem duplication mutations of FLT3 are stronger predictors of poor prognosis than point mutations.^3,4,5 Clinical trials with FLT3 tyrosine kinase inhibitors, sorafenib, and sunitinib have shown evidence of activity, albeit moderate and transient.^6,7 Other FLT3 inhibitors are currently in development.

Through its association with nucleolar ribonucleoprotein and ribosome function, NPM1 is involved in the regulation of cell proliferation and differentiation. Mutations in NPM1 are seen in up to 50% of cases of cytogenetically normal AML. Current studies indicate that presence of NPM1 mutations in cytogentically normal AML confer a favorable prognosis comparable to that in the good-risk cytogenetics AML. The combined presence of NPM1 mutation and FLT3-ITD results in significantly poorer outcome compared with patients with NPM1 and wildtype FLT3.^8,9,10,11,12

KIT is a class of receptor tyrosine kinase that is involved in several pathways of cell differentiation and proliferation. KIT mutations, seen in 20% to 30% of AML, result in ligandindependent activation of downstream pathways, which translates clinically into higher relapse rates and shorter survival.^13,14 Studies attempting to incorporate tyrosine kinase inhibitors into treatment regimens have not shown significant activity. A study evaluating the benefit of adding imatinib to low-dose cytarabine as induction regimen in elderly c-KIT positive AML revealed no added benefit to the combination.¹⁵

The CEBPA gene product is involved in myeloid lineage differentiation, and mutations in the CEBPA gene occur in 10% to 20% of cytogenetically normal AML. Studies have shown that the presence of CEBPA mutations confers a favorable outcome.^16,17,18 Data from recent studies, however, suggests that the favorable effect conferred by CEBPA mutations is limited to cases with two copies of the mutant allele, whereas single allele mutations have outcomes comparable with wild-type CEBPA genes.^19,20,21 Hypermethylation of the CEBPA gene also appears to negate the benefit conferred by CEBPA mutations.²²

Isocitrate dehydrogenase (IDH) gene 1 and 2 products are involved in NADPH production through the Krebs cycle. Mutations of IDH 1 and 2 genes are found in 20% to 30% of AML with normal cytogenetics and appear to confer a negative outcome. In a study evaluating the prognostic significance of IDH1/IDH2 mutations, the presence of IDH1 mutations in otherwise low-risk, cytogenetically normal young AML subjects, with NPM1 mutations and negative FLT3-ITD, was associated with shorter 5-year disease-free survival rates of 42% versus 59% and a trend toward shorter 5-year overall survival of 50% versus 63% when compared with wild-type IDH1. A similar picture was seen with IDH2 mutations, where a lower
complete remission (CR) rate of 38% was seen compared with 75% in patients with wild-type IDH2.²³

Despite early studies showing Wilms tumor 1 (WT1) gene to exhibit a tumor-suppressor function, more recent data suggests that the WT1 gene exhibits a poorly understood dual function of an oncogene or a tumor suppressor, depending on the tissue of origin.²⁴ Mutations in WT1 gene are reported in 10% to 20% of cytogenetically normal AML cases.²⁵ The bulk of currently available data suggests that mutations in WT1 gene in cytogenetically normal AML constitute a negative prognostic factor.^26,27,28

The meningioma gene 1 (MN1) product is involved in transcriptional regulation, and mutations involving the MN1 gene have been reported to confer negative outcomes.^29,30

The most common induction regimen for AML is a combination of a 7-day continuous infusion of cytarabine with three daily infusions of an anthracycline (7 + 3 regimen). The most common anthracycline used is daunorubicin. Studies attempting to substitute or incorporate other agents to induction or to increase the cytarabine dose have resulted in worsening toxicity and no added benefit.^31,32 In a study evaluating anthracycline dose intensification for young adults, two different daunorubicin doses were compared; 45 versus 90 mg per m². Higher rates of CR and overall survival were noted in the higher daunorubicin group with no increase in adverse effects.³³ This data supports the use of more intense induction regimens in young adults. There is currently no standard anthracycline agent for induction. In a study comparing idarubicin with daunorubicin for induction, the idarubicin-allocated arm exhibited higher rates of CRs and overall survival compared with the daunorubicin arm. However, the use of nonequitoxic dose of daunorubicin might have resulted in the favorable outcome in the idarubicin-treated arm.³⁴

Response assessment with bone marrow aspiration and biopsy, cytogenetics and FISH or PCR-based studies for known leukemic mutations, is commonly performed between 21 and 28 days after induction with a “classical” 7 + 3 regimen. Earlier response assessment at 7 to 14 days after induction is indicated in the case of new or investigational agent use.¹

Postremission therapy includes consolidation chemotherapy with or without hematopoietic stem cell transplantation (HSCT). Selection of the appropriate treatment regimen largely depends on the risk of relapse, which is mainly dictated by karyotype, although increasingly other risk factors are being incorporated into medical decision making as described previously¹

The classical induction regimen (7 + 3) continues to be a viable option in fit elderly patients. In a study comparing two different daunorubicin doses as part of 7 + 3 induction regimen in elderly AML patients, the group assigned to daunorubicin at 90 mg per m² compared with 45 mg per m² exhibited higher CR rates with no significant increase in toxicity. Subgroup analysis revealed that patients 60 to 65 years of age and with core-binding factor leukemias; t(8;21), t(16;16), and inv(16) exhibited longer event-free and overall survivals.⁵⁰ Elderly AML patients harboring favorable translocations, t(8;21) and inv(16), were shown to exhibit better response compared with historical controls.⁵¹ Currently, there is no clear evidence to recommend one anthracycline agent over another if used at equitoxic doses.⁵² Hypomethylating agents (azacitidine and decitabine) are well tolerated and constitute a viable alternative in frail elderly patients unable to tolerate intensive standard induction. Available data suggests that patients with chromosome 5 and 7 abnormalities might derive the most benefit from hypomethylating agents.^53,54,55,56 In a study comparing cytarabine at a low daily dose (LDAC) compared with hydroxyurea in elderly AML patients unfit for intensive treatment, LDAC was associated with higher CR rates of 18% versus 1% for hydroxyurea, and hence is an option for frail elderly patients.⁵⁷

No standard postremission regimen exists for elderly patients. Studies have shown that HDAC consolidation results in extremely poor outcomes and increased toxicity in patients over the age of 60 years.³⁵ In a study randomizing elderly AML patients after attainment of first CR, patients who were randomized to receive six monthly courses of outpatient chemotherapy, with 1 day of anthracycline and five daily subcutaneous cytarabine injections, exhibited higher rates of disease-free and overall survival, and lower rate of death compared with a second intensive induction course.⁵⁸

Allogeneic HSCT is preferred for the treatment of high-risk younger adults with unfavorable cytogenetics secondary to the associated GVL effect. To gain the benefits of GVL effect while avoiding high transplant-related morbidity and mortality associated with allogeneic HSCT, reduced-intensity conditioning (RIC) allogeneic HSCT were devised. Several studies
employing RIC allogeneic HSCT showed that development of graft-versus-host-disease (GVHD) was associated with lower rates of relapse.^59,60 A large study evaluating the effect of age on the outcomes of RIC allogeneic HSCT in elderly AML patients in first CR revealed no significant effect of age on nontransplant mortality, relapse rates, and disease-free and overall survival. This study supports the notion that available modalities of treatment should be considered thoroughly in elderly patients. Chronologic age alone should not be used as the sole discriminant to deny potentially effective therapy.⁶¹ Patients’ preferences, performance status, and comorbidities should be integrated into the medical decision making regarding further postinduction therapy in elderly adults. Figure 44.3 depicts the treatment algorithm for AML patients older than 60 years.

Multiple studies have evaluated different combinations of cytotoxic agents for relapsed AML. Remissions are not durable and are often achieved at a cost of excessive toxicity. Intermediate and HDAC either as single agent, or in combination with mitoxantrone, etoposide in combination with mitoxantrone or cyclophosphamide, have been tried with minimal encouraging results.^{70,71,72,73,74} A combination of fludarabine, cytarabine, and granulocyte colony-stimulating factor (G-CSF) resulted in CR rates of 30% to 50% in primary refractory or relapsed AML. The combination was well tolerated and was not associated with significant toxicity. Median survival was considerably shorter with early relapsed AML compared with late relapses.^75,76

Autologous HSCT is likely to be of no benefit as first-line therapy in low-risk AML patients; however, it could be a viable option at relapse. Relapses are common secondary to loss of GVL effect, and there is no available data to show any survival benefit to autologous HSCT in AML.⁷⁷

Myeloablative allogeneic HSCT are associated with significant transplant-related morbidity and mortality, especially in patients with relapsed AML. Thus, multiple studies have evaluated the role of RIC regimens followed by allogeneic HSCT. Overall survival rates of 30% to 40% have been achieved with RIC allogeneic HSCT. It was well tolerated by elderly patients with no excessive mortality.^61,78,79

For patients who relapse after allogeneic HSCT few options are available. A second RIC allogeneic HSCT can be attempted in medically fit patients.^80,81 Reduced immunosuppression and donor lymphocyte infusion are potentially viable options, however, this intervention is associated with worsening GVHD. CR rates and 2-year overall survival rates of <20% have been reported.^82,83 Figure 44.4 depicts the treatment algorithm for relapsed AML.

Monotherapy with ATRA results in high rates of CR, however, remissions are not durable. Studies have shown that a combination of ATRA and chemotherapy administered simultaneously for induction is superior to either ATRA or chemotherapy alone in terms of CR and relapse rates, and longer disease-free survival.^88,89,90,91 ATRA has shown beneficial effects in decreasing the markers and duration of coagulopathy associated with APL.^88,92 Currently, there is no evidence supporting the use of ATRA in non-APL variants of AML.⁹³