INTRODUCTION — Management of acute myeloid leukemia (AML) is informed by the individual's medical fitness for intensive antileukemic therapy. Importantly, medical fitness is based on performance status and physiologic function, but not on age per se; we do not apply age limits when judging medical fitness.

This topic will discuss treatment and outcomes of AML in medically-unfit adults.

The following topics are presented separately:

●Overview of AML (see "Overview of acute myeloid leukemia in adults")

●Pretreatment evaluation and prognosis of AML in older adults (see "Pretreatment evaluation and prognosis of acute myeloid leukemia in older adults")

●Treatment of medically-fit adults with AML (see "Induction therapy for acute myeloid leukemia in medically-fit adults")

SPECIAL CONSIDERATIONS DURING THE COVID-19 PANDEMIC — The coronavirus disease 2019 (COVID-19) pandemic has increased the complexity of cancer care. Important issues include balancing the risk from treatment delay versus harm from COVID-19, ways to minimize negative impacts of social distancing during care delivery, and appropriately and fairly allocating limited health care resources. These issues and recommendations for cancer care during the COVID-19 pandemic are discussed separately.

●(See "COVID-19: Considerations in patients with cancer".)

DETERMINING MEDICAL FITNESS — Medical fitness is a key determinant of management of AML and informs the goals of care and choice of therapy. Medical fitness should be determined before initiating therapy for AML.

Clinical and laboratory evaluation — Pretreatment evaluation of the adult with AML should identify findings related to complications of AML and comorbid conditions that may influence the choice of therapy.

●Clinical – The clinical evaluation should assess effects of cytopenias associated with AML. The history should evaluate clinical consequences of anemia (eg, dyspnea, weakness, palpitations) that may impair the quality of life and document the history of transfusions. The patient should be examined for evidence of infections or excessive bleeding/bruising and clinical manifestations of comorbid illnesses, including heart disease, pulmonary processes, nutritional deficits, and renal impairment should be documented.

●Laboratory studies – Laboratory studies should include:

•Complete blood count and differential

•Prothrombin time (PT) and partial thromboplastin time (PTT)

•Serum electrolytes, renal and liver function tests

•Electrocardiogram (EKG)

If cytogenetic and molecular features of the leukemia were not determined at the time of diagnosis, they should be defined prior to selecting therapy using either peripheral blood or bone marrow. (See "Clinical manifestations, pathologic features, and diagnosis of acute myeloid leukemia", section on 'Diagnosis' and "Classification of acute myeloid leukemia (AML)".)

Assessment instruments — Assessment of medical fitness includes evaluation of:

●Eastern Cooperative Oncology Group (ECOG) performance scale (table 1).

●Physiologic fitness (eg, comorbid conditions, activities of daily living, physical performance tests, cognition), as measured by the Charlson comorbidity index (CCI) (table 2). Other measures of medical fitness, such as the Short Physical Performance Battery (SPPB; https://geriatrictoolkit.missouri.edu/SPPB-Score-Tool.pdf) are also acceptable [1].

For some older patients with AML, a comprehensive geriatric assessment and geriatric consultation may aid assessment of medical fitness [2,3]. For older patients with AML, a detailed assessment of medical fitness and its impact on clinical outcomes are discussed separately. (See "Pretreatment evaluation and prognosis of acute myeloid leukemia in older adults".)

Classification of fitness — We classify patients with AML according to their medical fitness for treatment. Determination of medical fitness for patients with hematologic malignancies is based on assessment of performance status (PS) and medical fitness. (See 'Assessment instruments' above.)

The following should be considered in judging medical fitness:

●Age, per se, does not determine medical fitness. However, because of age-related comorbidities, caution should be used when considering intensive therapy for patients ≥75 years old, as most outcomes data with intensive antileukemic therapy came from studies in younger, fit adults.

●Chronic comorbid conditions should be weighted more heavily than transient medical complications of the leukemia itself (eg, infection, heart failure exacerbated by anemia). The burden of AML can contribute to a lack of fitness and, in some cases, treatment may alleviate consequences/complications of AML and enhance the patient's ability to tolerate and benefit from subsequent treatment.

It should be recognized that there are no clear distinctions between various categories of medical fitness that are described below. Furthermore, some measures of PS or physiologic fitness can apply to different fitness categories. In selecting a category of fitness, we seek to protect frail patients from treatment that they are unlikely to tolerate, while not depriving a medically-unfit, but not frail patient of the opportunity to achieve remission and prolonged survival. Note that our use of the term frail for managing patients with hematologic malignancies may differ from other definitions of frailty. (See "Frailty", section on 'Concepts and definitions'.)

●Medically-fit – Medically-fit patients are those who are judged to be able to tolerate intensive treatment for AML, based on both of the following (see 'Assessment instruments' above):

•ECOG PS (table 1): 0 to 2

•CCI (table 2): 0 to 2

For medically-fit patients who choose to receive intensive induction therapy for AML, management is discussed separately. (See "Induction therapy for acute myeloid leukemia in medically-fit adults".)

Some medically-fit patients may, for reasons of age or personal preference, choose to receive less intensive therapy. For such patients, management is described below. (See 'Medically-unfit, but not frail' below.)

●Medically-unfit, but not frail – Patients who are medically-unfit, but not frail are unlikely to tolerate intensive antileukemic therapy because of impaired PS or comorbid conditions. This fitness category includes a broad range of physical function. Some patients have only modest, recent, or transient impairment of functional status, while others have substantial comorbid illnesses, cognitive impairment, or other conditions that may affect their ability to tolerate treatment.

We judge patients to be medically-unfit, but not frail if either of the following applies (see 'Assessment instruments' above):

•ECOG PS (table 1): 3

•CCI (table 2): 3

Treatment for medically-unfit patients is described below. (See 'Medically-unfit, but not frail' below.)

●Frail – Frail patients are those whose debility or comorbid conditions would not permit treatment aimed at modifying the disease course, as reflected by both the following:

•ECOG PS (table 1): ≥3

•CCI (table 2): ≥3

Management of frail patients with AML should focus primarily on providing supportive care to improve the quality of life by lessening symptoms caused by AML and comorbid conditions. It is important to carefully discuss the goals of care, likely outcomes, and available resources for personal and medical assistance. (See 'Frail patients' below and 'Goals, benefits, and timing of treatment' below.)

EMERGENCIES — Medical emergencies are encountered regularly in patients who present with AML. (See "Overview of the complications of acute myeloid leukemia".)

●Acute promyelocytic leukemia (APL) – APL with t(15;17) is a unique category of AML that constitutes a medical emergency and requires urgent and specific treatment because of its distinctive clinical manifestations and complications.

APL is more likely in patients who present with bleeding or bruising, when leukemic blasts have coarse or dense cytoplasmic granules, and when there are few circulating leukemic cells or a low white blood cell count. The diagnosis is confirmed by detection of the PML-RARA fusion gene and/or an associated chromosomal translocation, as described separately. (See "Clinical manifestations, pathologic features, and diagnosis of acute promyelocytic leukemia in adults".)

The mainstay of therapy is prompt initiation of all-trans retinoic acid (ATRA), with or without arsenic trioxide. Medical fitness is generally not a barrier to such treatment and APL responds equally well across the age spectrum. Management of APL is presented separately. (See "Initial treatment of acute promyelocytic leukemia in adults".)

●Hyperleukocytosis/leukostasis, which may be manifest as respiratory and/or neurologic distress in a patient with AML and myeloblasts >50,000/microL, or any patient with AML and myeloblasts >100,000/microL. (See "Hyperleukocytosis and leukostasis in hematologic malignancies".)

●Metabolic abnormalities, including tumor lysis syndrome (eg, hyperkalemia, hyperphosphatemia, hyperuricemia, and/or renal insufficiency), can be diagnosed and managed as described separately. (See "Tumor lysis syndrome: Definition, pathogenesis, clinical manifestations, etiology and risk factors" and "Tumor lysis syndrome: Prevention and treatment".)

●Bleeding/coagulation abnormalities may be caused by severe thrombocytopenia, disseminated intravascular coagulation, and other coagulopathies. (See "Hematopoietic support after hematopoietic cell transplantation", section on 'Platelet transfusion' and "Evaluation and management of disseminated intravascular coagulation (DIC) in adults".)

●Unexplained neurologic or visual findings may be manifestations of leukostasis (described above), thrombotic or hemorrhagic cerebrovascular accident, or leukemic involvement of the central nervous system. (See "Involvement of the central nervous system with acute myeloid leukemia".)

GOALS, BENEFITS, AND TIMING OF TREATMENT

●Goals of care – Goals of care are influenced by medical fitness and individual values. Age, per se, should not define or limit the goals of care. Determination of medical fitness is described above. (See 'Determining medical fitness' above.)

•Medically-fit – For most medically-fit patients, the goal of care is to achieve long-term survival with the possibility of cure, as discussed separately. (See "Induction therapy for acute myeloid leukemia in medically-fit adults".)

•Medically-unfit, but not frail – For most medically-unfit patients who are not frail, the goal is to achieve a remission as the most effective way to alleviate symptoms, improve the quality of life, and/or extend survival. It should be recognized that current therapies for this population have not been proven to achieve long-term survival or the possibility of cure. (See 'Medically-unfit, but not frail' below.)

•Frail – For frail patients, the primary goals are relief of symptoms and improved quality of life through supportive care. The potential for harm resulting from more intensive therapy is likely to outweigh the benefits. (See 'Frail patients' below.)

Goals of care should be established in discussions between the patient and clinicians at the time of diagnosis and periodically through the course of the disease. Discussions should acknowledge that for most patients, AML will be a life-ending disease. Many patients overestimate their long-term prognosis and chance for cure, and don't recall being offered more than one treatment [4,5]. Mutual understanding between the patient, family members, and practitioners facilitates conversations regarding goals of care, selection of therapy, health care proxies, do-not-resuscitate directives, and end-of-life decisions.

●Benefits of treatment – Patients of any age who are medically-unfit, but not frail can benefit from treatment of AML. Compared with supportive care alone, treatment of AML can relieve symptoms, improve the quality of life, and potentially extend survival.

Some clinicians and patients fear that treatment will cause more harm than good. However, the major causes of death from AML are infection and hemorrhage related to disease-associated cytopenias; treatment of the underlying AML is the most effective and efficient way to control those complications [6]. While some population-based studies have suggested that treatment only benefits patients ≤80 years, other studies have documented that treatment of selected patients in their 80s and 90s can achieve high rates of response and improved survival [7-10].

Compared with supportive care alone, treatment of AML is associated with improved quality of life, including fewer hospitalizations and less need for transfusions and antibiotics [8,11]. Improved outcomes are reported in prospective, retrospective, and population-based studies, even while recognizing that retrospective analyses may exhibit a bias that favors treatment for more medically fit and/or younger patients [7-9,11,12]. United States Surveillance, Epidemiology, and End Results (SEER)-Medicare data of 5480 patients ≥65 years (median age 78 years) reported improved survival in treated patients, compared with untreated patients (median six versus two months, respectively) [7]. Outcomes with treatment of medically-unfit patients with AML are discussed below. (See 'Medically-unfit, but not frail' below.)

●Timing of therapy – For most patients with AML, it is not harmful to temporarily delay treatment to permit an adequate diagnostic evaluation, assessment of fitness, medical stabilization, and management of complications.

An important exception is patients who are suspected of having acute promyelocytic leukemia (APL), for whom treatment should begin immediately, even before cytogenetic and/or molecular confirmation is available. Features that suggest the presence of APL are described above. (See 'Emergencies' above.)

Aside from APL, it is generally not harmful to delay treatment of AML by 96 hours or longer to stabilize the patient's condition and/or characterize the cytogenetic and molecular features of the AML cells. Infections, bleeding, hyperuricemia, renal dysfunction, anemia, and thrombocytopenia should be managed and comorbid conditions optimized to improve baseline physical function prior to initiating antileukemic therapy. Oral hydroxyurea can be used for cytoreduction if the peripheral blast count must be reduced during this time.

A retrospective study of 2263 patients with AML reported no differences in rates of remission, early death, or overall survival based on the time between diagnosis and initiation of treatment [13] and other studies have reported similar findings [14,15].

MEDICALLY-UNFIT, BUT NOT FRAIL — Classification of medical fitness is described above. (See 'Classification of fitness' above.)

The use of prophylactic antibiotics, antifungal agents, and myeloid cytokines for medically-unfit patients receiving induction therapy for AML varies between institutions. Supportive care for patients who are undergoing induction therapy for AML are discussed separately. (See "Induction therapy for acute myeloid leukemia in medically-fit adults", section on 'Supportive care'.)

Eligible for a hypomethylating agent (HMA) — For most patients who are medically-unfit, but not frail, our preferred approach for initial treatment of AML is a regimen that includes an HMA (eg, azacitidine, decitabine) (algorithm 1). The choice of a specific regimen is informed by the availability of venetoclax (BCL2 inhibitor), patient preference, and other factors, as described in the following sections.

Possible exceptions to an HMA-based regimen include:

●Ineligible  

•Severe liver or kidney disease (excessive toxicity)

•Prior HMA therapy (eg, for a preceding myelodysplastic syndrome)

●Patient/clinician preference – Some patients favor an alternative approach because of convenience, toxicity profile, or other reasons. In addition, some experts consider certain non-HMA-containing regimens acceptable in this setting, as described below. (See 'Alternative treatments' below.)

HMA plus venetoclax — In settings where venetoclax is available, we suggest treatment with an HMA (eg, either azacitidine or decitabine) plus venetoclax, rather than treatment with an HMA alone or other approaches. This suggestion is based on prospective and retrospective studies in which venetoclax plus either azacitidine or decitabine was associated with improved response rates and comparable toxicity; in addition, a randomized trial reported that venetoclax plus azacitidine improved survival compared with azacitidine alone [16-18]. Outcomes with venetoclax plus an HMA are superior to outcomes with targeted agents, low-dose cytarabine (LoDAC), and other approaches, but they have not been directly compared in prospective studies.

●Setting of care – We generally administer the first treatment of an HMA plus venetoclax in an inpatient setting, to monitor and manage the possibility of tumor lysis syndrome (TLS). Some patients with nearby housing and dependable transportation can be treated safely with daily outpatient visits. Although TLS is not well-documented with venetoclax for AML, it can occur with lymphoid malignancies.

●Dosing and administration – Venetoclax plus either azacitidine or decitabine is administered in 28-day cycles [17,19]. Patients should be monitored clinically and biochemically for TLS and should receive adequate hydration and an anti-hyperuricemic agent. (See "Tumor lysis syndrome: Prevention and treatment".)

•Either azacitidine or decitabine:

-Azacitidine 75 mg/m² daily intravenously (IV) or subcutaneously for the first 7 days of each 28-day cycle. Other protocols for azacitidine administration are described below, but they have not been evaluated in combination with venetoclax. (See 'Azacitidine' below.)

or

-Decitabine 20 mg/m² IV on days 1 to 5 of each 28-day cycle. (See 'Decitabine' below.)

Plus

•Venetoclax – In cycle 1 only, venetoclax should be administered according to the following dose escalation schedule, to lessen the risk of TLS. Venetoclax should be taken orally with food. The dose of venetoclax should be reduced if it is administered with a moderate or strong CYP3A inhibitor (eg, antifungal azoles) [20].

-Cycle 1, day 1: 100 mg

-Cycle 1, day 2: 200 mg

-Cycle 1, days 3 to 28: 400 mg

We attempt to adhere to this protocol of 28 days of venetoclax administration to maximize the rate of complete responses (CR).

Subsequent cycles can begin with 400 mg venetoclax daily.

In the event of substantial hematologic toxicity, venetoclax administration can be shortened to the first 21 days or 14 days of subsequent 28-day treatment cycles, although the efficacy of such schedules is not yet well-defined.

●Monitoring – During the first treatment cycle, serum potassium, renal function tests, calcium, phosphate, and lactate dehydrogenase (LDH) should be obtained daily, until the threat of TLS has passed.

Prior to each treatment cycle, a complete blood count (CBC), liver chemistries, and renal and liver function tests should be obtained. Interval visits and CBCs may be required, guided by the degree of cytopenias and/or need for transfusion support. Response assessment is discussed below. (See 'Response assessment' below.)

Responses to venetoclax plus an HMA are relatively rapid, and we generally discontinue treatment if there is no meaningful response after three or four treatment cycles. We continue treatment for as long as the patient is responding and tolerating treatment, although adjustments to the dose or schedule of treatment may be needed, as described below. (See 'Maintenance therapy' below.)

●Adverse effects – The primary adverse effects (AEs) of venetoclax plus an HMA are prolonged cytopenias, febrile neutropenia, and mild or moderate gastrointestinal effects (eg, nausea, vomiting, diarrhea), fatigue, and edema. Venetoclax plus an HMA should be used with caution in patients who have liver disease, renal dysfunction, or are receiving CYP3A inhibitors.

●Efficacy – Compared with azacitidine alone, venetoclax plus azacitidine is associated with superior survival and venetoclax plus decitabine is associated with superior rates of response (survival data are not currently available). Toxicity of either combination is only modestly increased relative to HMA monotherapy.

•Venetoclax plus azacitidine – A double-blind, phase 3 trial (VIALE-A) of 431 patients with previously untreated AML reported that, compared with azacitidine plus placebo, azacitidine plus venetoclax achieved superior overall survival (OS), a higher rate of remission, more rapid and durable responses, and comparable quality of life, but was associated with a higher incidence of febrile neutropenia [16]. Patients who were ≥75 years of age or had comorbid conditions that precluded intensive remission induction therapy were randomly assigned (2:1) to azacitidine plus venetoclax versus azacitidine plus placebo. With median follow-up of 21 months, azacitidine-venetoclax achieved superior OS (15 versus 10 months; hazard ratio [HR] 0.66; 95% CI 0.52-0.85), more CR (37 versus 18 percent), a higher composite score of CR plus CR with incomplete hematologic recovery (CRi; 65 versus 23 percent), and more transfusion-independence from red blood cells (60 versus 35 percent) and platelets (69 versus 50 percent). Grade ≥3 AEs that were more common with azacitidine-venetoclax included febrile neutropenia (42 versus 19 percent), neutropenia (42 versus 28 percent), and thrombocytopenia (45 versus 38 percent). AEs (of all grades) associated with azacitidine-venetoclax included nausea or vomiting (44 and 30 percent, respectively), constipation or diarrhea (43 and 41 percent), hypokalemia (29 percent), peripheral edema (24 percent), and infections (84 percent), but the incidence of AEs were only modestly increased compared with azacitidine plus placebo.

Other studies of azacitidine plus venetoclax reported similar rates of remission and toxicity [17,19].  

•Venetoclax plus decitabine – Venetoclax plus decitabine is associated with a toxicity profile and response rate that is similar to venetoclax plus azacitidine, but survival data are not yet available. Treatment of 145 patients with venetoclax plus either decitabine or azacitidine was associated with 67 percent CR plus CRi, 11 month median duration of response, and 18 month median OS; toxicity profiles and responses were similar for both regimens [17]. In a study of 31 patients treated with venetoclax plus decitabine, rates of CR and CRi were 54 and 8 percent, respectively; median duration of remission was five months [18].

•Single-agent venetoclax – Venetoclax as a single agent has limited activity against AML. In a study of 32 patients with relapsed or refractory AML, venetoclax was associated with 19 percent response rate and 5 month median OS [21].

Venetoclax is approved by the US Food and Drug Administration (FDA) in combination with an HMA in medically-unfit patients or those ≥75 years with newly diagnosed AML; venetoclax is not approved as a single agent for treatment of AML [18]. The European Medicines Agency (EMA) approved decitabine for treatment of AML and approved azacitidine for treatment of AML that arose from a myelodysplastic syndrome (MDS); neither HMA is approved by the US FDA for treatment of AML.

If venetoclax is not available — Where venetoclax is not available, we suggest other HMA-based therapy, rather than other treatments, based on a more favorable balance of efficacy and toxicity. Specifically, if there is a mutation of isocitrate dehydrogenase 1 (IDH1) and ivosidenib is available, we treat with an HMA plus ivosidenib; otherwise, we treat with HMA monotherapy. (See 'HMA plus ivosidenib' below.)

Some experts consider certain non-HMA-containing regimens acceptable in this setting, as described below. (See 'Alternative treatments' below.)

HMA plus ivosidenib — Azacitidine plus the IDH1 inhibitor, ivosidenib, is a well-tolerated combination that can achieve deep and durable responses in patients with newly diagnosed AML with mutant IDH1.

Azacitidine (75 mg/m² on days 1 to 7 subcutaneously) plus ivosidenib (500 mg once daily orally) in 28-day cycles was associated with 78 percent overall response and 61 percent CR among 23 patients with newly diagnosed AML with mutant IDH1 who were ineligible for intensive induction chemotherapy [22]. OS at 12 months was 82 percent and median OS was not reached after 16 months. Among patients who achieved CR, 71 percent had undetectable measurable residual disease (MRD) by polymerase chain reaction (PCR). Grade ≥3 AEs included neutropenia, anemia, and thrombocytopenia (22, 13, and 13 percent, respectively) and QT prolongation in 13 percent. Differentiation syndrome (DS) was reported in 17 percent, but no cases of DS were grade ≥3.

Ivosidenib plus azacitidine achieved superior outcomes compared with azacitidine plus placebo in patients with newly diagnosed AML with mutant IDH1 who were ineligible for intensive induction chemotherapy, according to preliminary results from a phase 3 trial (AGILE; NCT03173248; presented in abstract form) [23]. Compared with azacitidine plus placebo, ivosidenib plus azacitidine achieved superior median OS (24 versus 8 months), CR (47 versus 15 percent), ORR (63 versus 19 percent), and EFS (hazard ratio [HR] 0.33 [95% CI 0.16-0.69]). In both trial arms, >90 percent of patients had grade ≥3 AEs and 4 percent had grade ≥3 DS, but there was less febrile neutropenia with ivosidenib plus azacitidine (28 versus 34 percent).

HMA plus other agents — Compared with HMA monotherapy, no other agents besides venetoclax or ivosidenib have proven benefit when administered along with an HMA. Treatment with an HMA plus either venetoclax or ivosidenib are described above. (See 'HMA plus venetoclax' above and 'HMA plus ivosidenib' above.)

Other HMA-based combinations include:

●Decitabine plus sapacitabine – Alternating cycles of decitabine with sapacitabine (an orally administered deoxycytidine analogue) did not improve outcomes compared with decitabine alone in a phase 3 trial [24]. Patients (≥70 years) with newly diagnosed AML who were unsuitable for intensive induction therapy were randomly assigned (241 patients in each arm) to single-agent decitabine (20 mg/m² infusion on days 1 to 5 every 28 days) versus the same dose of decitabine every 8 weeks, alternating with sapacitabine (300 mg by mouth twice daily on days 1 to 3 per week for 2 weeks every 8 weeks). Compared with decitabine alone, alternating decitabine/sapacitabine was associated with similar median OS (six months for each arm), one-year OS (34 to 35 percent), CR (17 percent for combination therapy versus 11 percent for decitabine alone), transfusions, and hospitalizations. Grade ≥3 hematologic AEs and grade ≥3 pneumonia, sepsis, and disease progression were similar in both arms.

●Azacitidine plus gilteritinib – Treatment with azacitidine plus gilteritinib was associated with similar OS (9 to 10 months) and CR (14 to 16 percent), but higher ORR (58 versus 27 percent) compared with azacitidine plus placebo, according to an abstract report of preliminary results of a phase 3 trial that included 123 patients with newly diagnosed AML who were ineligible for intensive treatment [25]. We await further data before using this combination outside of a clinical trial.

●Decitabine plus bortezomib – In a randomized phase 2 study of 163 patients ≥60 years, compared with decitabine (20 mg/m², days 1 to 10) alone, addition of bortezomib did not improve the rate of response or survival [26]. Outcomes did not differ between the study arms and, overall, the rate of CR plus CRi was 39 percent and median OS was 9 months. The most common AE was febrile neutropenia.

HMA monotherapy: choice of drug — For patients who are eligible for an HMA, but neither venetoclax nor ivosidenib is available or suitable, we generally treat with HMA monotherapy (algorithm 1), based on prospective trials in which HMA monotherapy was associated with superior responses and similar toxicity compared with LoDAC and was associated with improved symptom relief and possible prolongation of survival compared with supportive care alone [27-29].

Both azacitidine and decitabine are well-tolerated and they have comparable efficacy for newly diagnosed AML. Both HMAs are given parenterally, but a patient may favor one or the other based on schedule and/or method of administration. Some patients may favor decitabine because it is given intravenously (convenient for patients who have venous access devices), while others favor azacitidine because it requires less time in an infusion center and can be give subcutaneously (which can cause pain or bruising in thrombocytopenic patients). Orally available HMAs are in development or approved for other indications, but their efficacy has not yet been reported for induction therapy of AML.

Azacitidine and decitabine have not been directly compared in randomized trials, but prospective and retrospective analyses indicate similar efficacy and toxicity profiles [27-40]. A population-based survey of 2263 adults >65 years old with AML reported that patients who were treated with azacitidine or decitabine had similar median OS (seven to eight months) and both achieved red blood cell transfusion-independence in approximately one-third of patients [35]. Details of these studies, including some that compared HMAs with other approaches are discussed below. (See 'Azacitidine' below and 'Decitabine' below.)

The US FDA has not approved either azacitidine or decitabine for AML, but both are commonly used for off-label treatment of AML. The EMA has approved decitabine for treatment of AML and has approved azacitidine for treatment of AML that arose from MDS.

Azacitidine

●Administration – Azacitidine (75 mg/m² subcutaneous) can be administered in an outpatient setting for 7 consecutive days in 28-day cycles [27]. Other azacitidine regimens (eg, 100 mg/m² per day for 5 consecutive days in 28-day cycles) are more convenient and are also acceptable for treatment of AML [36]. Dose adjustment may be needed for patients with liver or kidney disease [41].

●Monitoring – CBC, liver chemistries, and renal function should be evaluated prior to each treatment cycle and as clinically indicated. We continue treatment for as long as the patient is responding and tolerating treatment, although it may be necessary to insert a brief interval between cycles or adjust doses to enable recovery of blood counts. Response assessment is discussed below. (See 'Response assessment' below.)  

●Toxicity – Grade ≥3 AEs include cytopenias and febrile neutropenia in one-quarter to one-third of patients and most will have at least one hospitalization for neutropenic fever during treatment. Growth factor support may be considered.

●Efficacy – Azacitidine is associated with a hematologic response in up to half of patients with AML. Compared with supportive care alone, azacitidine treatment is associated with longer survival:

•A phase 3 trial of 488 patients ≥65 years old with newly diagnosed AML reported that azacitidine achieved superior survival compared with best supportive care (BSC) [28]. Each enrolled patient was preselected for a conventional care regimen (CCR; either LoDAC, intensive induction chemotherapy, or BSC); within each CCR subgroup, patients were randomly assigned to azacitidine versus the preselected CCR. Compared with the overall CCR cohort, azacitidine achieved marginally superior median OS (10 versus 7 months; HR 0.85; 95% CI 0.69-1.03) and one-year OS (47 versus 34 percent). Among those preselected for BSC, patients randomized to azacitidine achieved superior OS (6 versus 4 months; HR 0.60; 95% CI 0.38-0.95); OS did not differ between patients who were randomly assigned to azacitidine versus either LoDAC or intensive induction therapy. For patients treated with azacitidine, the rate of CR plus CRi was 28 percent and grade ≥3 AEs included febrile neutropenia (28 percent), thrombocytopenia (24 percent), pneumonia (19 percent), and anemia (16 percent).

•In another phase 3 trial, azacitidine achieved superior OS compared with CCR; this trial included 113 patients with AML among 358 patients with either AML or high-risk MDS [27]. Compared with CCR (preselected as either LoDAC, intensive induction chemotherapy, or BSC), azacitidine was associated with superior median OS (25 versus 16 months, respectively; HR 0.47; 95% CI 0.28-0.79), two-year OS (50 versus 16 percent), and hematologic response (29 versus 12 percent). Grade ≥3 cytopenias were the most common serious AEs for all treatments.

•Among 35 patients treated with azacitidine for newly diagnosed AML, the rate of CR/CRi was 31 percent and median response duration was 13 months [30].

The EMA approved azacitidine for treatment of AML that arose from MDS; the US FDA has not approved azacitidine for treatment of AML.

An oral preparation of azacitidine (CC-486) has been approved by the US FDA for post-remission maintenance therapy of AML, but it is not well-studied for induction therapy of AML; use of CC-486 for maintenance therapy is discussed below. (See 'Post-remission management' below.)

Treatment with azacitidine plus either venetoclax or ivosidenib is presented above. (See 'HMA plus venetoclax' above and 'HMA plus ivosidenib' above.)

Decitabine

●Administration – Decitabine (20 mg/m² IV daily over one hour) can be administered in an outpatient setting for 5 or 10 consecutive days, in 28-day cycles [29,31-34]. Decitabine can also be administered in the inpatient setting (eg, 15 mg/m² IV over four hours every eight hours for three consecutive days, every six weeks) [37]. Dose adjustment may be needed for severe liver or kidney disease [42].

●Monitoring – CBC, liver chemistries, and renal function should be evaluated prior to each cycle of therapy and as clinically indicated. Response to decitabine typically requires at least two cycles of therapy, and treatment should continue for as long as a benefit persists with tolerable side effects. Growth factor support and dose adjustment may be necessary. Response assessment is discussed below. (See 'Response assessment' below.)  

●Toxicity – Grade ≥3 AEs include cytopenias and febrile neutropenia in one-quarter to one-third of patients, and most patients will have at least one hospitalization for neutropenic fever during treatment. Growth factor support may be considered.

●Efficacy – Decitabine monotherapy is associated with a hematologic response in up to half of patients with AML and five to eight month median OS [29,31-34,37-40]. Compared with treatment for 5 consecutive days, treatment for 10 consecutive days may achieve higher response rates [33,34,38]. Decitabine appears to be effective in patients with AML with mutations in TP53 and/or unfavorable cytogenetic features [32]. Studies of decitabine for AML include:

•A phase 3 trial randomly assigned 485 patients (median age 74 years) with AML and poor- or intermediate-risk cytogenetics to decitabine versus treatment choice (preselected BSC or LoDAC) [29]. The rate of CR plus CR without platelet recovery (CRp) was superior with decitabine (18 versus 8 percent), but OS was not significantly different (8 versus 5 months; HR 0.85; 95% CI 0.69-1.04). AEs of thrombocytopenia and neutropenia (27 and 24 percent, respectively) with decitabine were similar to those of LoDAC.

•A multicenter study reported that decitabine achieved clearance of bone marrow blasts in nearly half of 54 patients ≥60 years old [32]. Response rates were higher in patients with unfavorable-risk cytogenetics (67 percent of 43 patients) and TP53 mutation (100 percent of 21 patients), compared with intermediate-risk or favorable-risk cytogenetics (34 percent) and wild-type TP53 (41 percent). Maximum clinical response required at least two cycles of treatment. Other studies that used a similar 10-day decitabine regimen reported CR rates of 30 to 47 percent [33,34,38].

•In a multicenter study of 227 older adults (median age 72 years) in which decitabine (135 mg/m² total dose infused over 72 hours every six weeks) was administered to patients who were ineligible for intensive induction chemotherapy, the CR rate was 13 percent, median OS was 6 months, and one-year and two-year OS rates were 28 and 13 percent, respectively [37].

•A randomized phase 2 study reported that addition of bortezomib to decitabine did not improve outcomes [26], as described separately. (See 'HMA plus other agents' above.)

Decitabine is approved for treatment of AML by the EMA, but not by the US FDA. A fixed-dose oral agent (35 mg decitabine plus 100 mg cedazuridine, an inhibitor of cytidine deaminase in the gut and liver) is approved by the US FDA for treatment of MDS and chronic myelomonocytic leukemia, but not for AML. (See "Chronic myelomonocytic leukemia: Management and prognosis", section on 'Hypomethylating agents'.)

Alternative treatments — Alternative, non-HMA-based treatment may be appropriate for selected patients:

●Patient is ineligible for HMA:

•Prior HMA-based treatment

•Severe liver or kidney dysfunction

●Patient declines an HMA:

•Higher value assigned to convenience or toxicity than outcomes

•Reduce daily clinic visits or hospitalization for treatment

Some experts consider alternative approaches to be acceptable for a broader population of medically-unfit, but not frail patients.

IDH inhibitors — For patients with AML with an IDH mutation, IDH inhibitors are generally well-tolerated and associated with substantial hematologic responses [43,44].

Treatment with an IDH inhibitor plus azacitidine is discussed above. (See 'HMA plus ivosidenib' above.)

●IDH1 mutation – For IDH1-mutated AML, ivosidenib monotherapy is well-tolerated and can induce durable remissions and transfusion-independence. Ivosidenib is not available in all medical settings.

Ivosidenib is administered 500 mg once daily by mouth, with or without food; high-fat meals should be avoided. Grade ≥3 adverse reactions that were reported in ≥5 percent of patients include differentiation syndrome (DS; which can be fatal if not treated promptly), prolonged QTc interval, and fatigue; rare cases of posterior reversible encephalopathy syndrome (PRES) and Guillain-Barré syndrome have been reported. (See "Treatment of relapsed or refractory acute myeloid leukemia", section on 'Remission re-induction' and "Differentiation syndrome associated with treatment of acute leukemia".)

We generally continue treatment with an IDH inhibitor for as long as the patient continues to respond and the agent is tolerated. (See 'Maintenance therapy' below.)

A study that included 34 patients with previously untreated IDH1-mutant AML who were ineligible for intensive induction therapy (median age 76 years) reported that ivosidenib monotherapy achieved 30 percent CR plus 12 percent CRi [43]. Among 21 transfusion-dependent patients, 9 became transfusion-independent. Median OS was 13 months after median follow-up of nearly two years. Treatment was well-tolerated; grade ≥3 DS occurred in 9 percent, but did not require treatment interruption.

Ivosidenib is approved by the US FDA for adults ≥75 years old with AML that has a susceptible IDH1 mutation detected by an FDA-approved diagnostic test [45].

●IDH2 mutation – Enasidenib is an inhibitor of IDH2 and is generally well-tolerated, but its efficacy for newly diagnosed AML is not well-defined and it is not available in all medical settings.

Enasidenib is administered 100 mg once daily by mouth for a minimum of six months (to allow time for clinical response) and should be continued until disease progression or unacceptable toxicity. The dose should be reduced for elevated bilirubin. Grade ≥3 adverse reactions in ≥5 percent of patients include diarrhea, nausea, and DS (which can be fatal if not treated promptly). (See "Treatment of relapsed or refractory acute myeloid leukemia", section on 'Remission re-induction' and "Differentiation syndrome associated with treatment of acute leukemia".)

The efficacy of enasidenib for newly diagnosed AML with an IDH2 mutation is not well-defined, but treatment of relapsed or refractory AML in 23 evaluable patients ≥60 years achieved CR/CRi in 43 percent; azacitidine was added for some patients who did not achieve CR/CRi by cycle 5 [44].

Enasidenib is approved by the US FDA for adults with relapsed or refractory AML with an IDH2 mutation as detected by an FDA-approved test; it is not currently approved for treatment of newly diagnosed AML [46].

Low-dose cytarabine — Low-dose cytarabine (LoDAC) has limited therapeutic efficacy in AML, but we consider it acceptable for patients who are not candidates for HMA-based treatment (eg, patients who are medically-unsuitable or previously received HMA therapy) and do not have an actionable mutation (eg, IDH mutation); however, LoDAC may be ineffective for patients with higher-risk AML (ie, adverse cytogenetic and/or molecular features). Some patients may favor LoDAC because it can be administered subcutaneously in the home (ie, to reduce clinic visits or hospitalization) or because they value the convenience or toxicity profile more than its modest efficacy.

We favor treatment with LoDAC combined with either venetoclax or glasdegib, based on superior outcomes compared with LoDAC monotherapy [47-50]. (See 'LoDAC combination therapy' below.)

LoDAC combination therapy

●LoDAC plus venetoclax – Venetoclax is an orally available inhibitor of BCL2 that is effective for initial treatment of AML in combination with LoDAC. Note that the following dose-escalation schedule for venetoclax differs from that used for venetoclax plus azacitidine. (See 'HMA plus venetoclax' above.)

To lessen the risk for TLS, when venetoclax is administered with LoDAC (20 mg/m² once daily for 10 days in 28-day cycles), the following dose-escalation schedule should be used:

•Day 1: Venetoclax 100 mg

•Day 2: Venetoclax 200 mg

•Day 3: Venetoclax 400 mg

•Day 4 and beyond: Venetoclax 600 mg

Compared with LoDAC monotherapy, venetoclax plus LoDAC is associated with improved rates of remission and OS and an acceptable safety profile:

•In an international phase 3 trial of 211 patients with untreated AML, LoDAC plus venetoclax was associated with superior outcomes compared with LoDAC plus placebo [47]. Patients with medical comorbidities that precluded intensive induction chemotherapy and/or age ≥75 years were randomly assigned (2:1) to LoDAC plus venetoclax versus LoDAC plus placebo. With 18 month follow-up, median OS was superior in the venetoclax arm (8 versus 4 months; HR 0.70; 95% CI 0.50-0.98). Rates of CR plus CRi were 48 versus 13 percent, respectively, for venetoclax versus placebo, respectively. Key grade ≥3 AEs were febrile neutropenia (32 versus 29 percent), neutropenia (47 versus 16 percent), and thrombocytopenia (45 versus 37 percent).

•LoDAC plus venetoclax was associated with 54 percent CR plus CRi, 8 month median duration of remission, and 10 month median OS in a phase 1-2 study in 82 patients with newly diagnosed AML [50]. Response rates were more favorable in patients with intermediate-risk cytogenetics, de novo AML, and no prior HMA treatment.

Venetoclax is approved by the US FDA for use in combination with LoDAC in medically-unfit patients or those ≥75 years with newly diagnosed AML [18].

●LoDAC plus glasdegib – Glasdegib (an oral inhibitor of the hedgehog pathway) combined with LoDAC is effective for treatment of newly diagnosed AML [49]. Glasdegib does not have a demonstrated role as a single agent for de novo AML. Glasdegib can prolong the QT interval and may be associated with ventricular fibrillation and/or tachycardia, inhibitors and inducers of CYP3A4 should be avoided, and it is potentially teratogenic [51].

Glasdegib is given orally 100 mg daily along with LoDAC. An electrocardiogram (EKG) should be obtained prior to treatment with glasdegib, one week after initiation of treatment, and then once monthly for the next two months to assess QTc prolongation [49]. Strong inhibitors and inducers of CYP3A4 should be avoided for patients treated with glasdegib; if a moderate CYP3A4 inducer cannot be avoided, the glasdegib dose should be increased [49]. The FDA label warns of potential embryo-fetal death or severe birth defects. Male partners of fertile women should use a condom for ≥30 days after the last dose and patients receiving glasdegib should not donate blood or blood products for ≥30 days after the last dose to avoid risk of teratogenicity in fertile blood recipients.

The most common serious adverse reactions in patients receiving glasdegib with LoDAC were febrile neutropenia, pneumonia, hemorrhage, anemia, and sepsis (29, 23, 12, 7, and 7 percent, respectively) [49]. Studies of LoDAC plus glasdegib include:

•The BRIGHT AML 1003 study (NCT01546038) included 115 patients with newly diagnosed AML who were ≥75 years old or who had severe cardiac disease, renal impairment, or impaired performance status [49]. Patients were randomized 2:1 to receive glasdegib (100 mg daily) plus LoDAC (20 mg subcutaneously twice daily on days 1 to 10 of a 28-day cycle; 77 patients) versus LoDAC alone (38 patients). With median follow-up of 20 months, outcomes with glasdegib plus LoDAC were superior to LoDAC alone, with median OS 8 versus 4 months, respectively (HR: 0.46; 95% CI 0.30-0.71), and rates of CR were 18 versus 3 percent.

•LoDAC plus glasdegib achieved superior OS (9 versus 5 months) and CR (17 versus 2 percent), when compared with LoDAC alone, in a randomized, open label phase 2 study of 132 medically-unfit adults with AML or high-risk MDS [48]. Addition of glasdegib to LoDAC was more effective in patients with favorable or intermediate-risk cytogenetics than in those with poor-risk cytogenetics.

Glasdegib was approved by the US FDA, in combination with LoDAC, for treatment of newly diagnosed AML in medically-unfit adults or those ≥75 years old [49].

●LoDAC plus clofarabine – Treatment with LoDAC plus clofarabine (20 mg/m² daily for five days) was associated with 58 percent CR, 14 month median relapse-free survival, 13 month median OS, and 7 percent induction mortality rate at eight weeks in 60 patients ≥60 years [52].

Treatment with single-agent clofarabine is discussed below. (See 'Clofarabine' below.)

LoDAC monotherapy — LoDAC combined with another agent is preferred over monotherapy, but other agents (eg, venetoclax, glasdegib) are not available in all settings, are associated with increased risk for some patients (eg, glasdegib for patients with cardiac dysrhythmias), and some patients may choose monotherapy for ease of administration or other reasons. (See 'LoDAC combination therapy' above.)

Single-agent LoDAC is generally administered subcutaneously 20 mg twice daily for 10 to 14 days every 28 days [53]. The dose may require adjustment for severe liver impairment. Moderate cytopenias and mild nausea and vomiting are the predominant AEs.

We generally continue treatment with LoDAC for as long as the patient is responding and tolerating treatment. (See 'Maintenance therapy' below.)

Outcomes with LoDAC monotherapy include:

●In a multicenter trial, 217 patients with AML or high-risk MDS were randomly assigned to receive LoDAC (20 mg twice daily for 10 days) versus hydroxyurea; patients were also randomly assigned to receive all trans retinoic acid (ATRA) or no ATRA [53]. Compared with hydroxyurea, LoDAC resulted in a higher remission rate (18 versus 1 percent, respectively) and improved OS (HR 0.60; 95% CI 0.44-0.81); addition of ATRA had no effect. The median disease-free survival in patients who achieved a CR with LoDAC was eight months. Toxicity was generally mild, with the most common grade ≥3 toxicities being cardiac (10 percent), nausea/vomiting (6 percent), and diarrhea (4 percent). Induction death occurred in 26 percent of patients.

●A multicenter trial randomly assigned 87 patients >65 years to LoDAC (10 mg/m² every 12 hours for 21 days) versus intensive chemotherapy and reported that LoDAC achieved 32 percent CR and 22 percent partial response [54]. There was no difference in OS or remission duration between the two arms of the study, but intensive therapy was associated with a higher rate of CR (52 versus 32 percent) but more early deaths.

Gemtuzumab ozogamicin — Gemtuzumab ozogamicin (GO) is an immunoconjugate that comprises an anti-CD33 monoclonal antibody linked to calicheamicin.

●GO administration – When used as a single agent for remission induction of AML, we treat with GO 6 mg/m² on day 1 and 3 mg/m² on day 8, and continue treatment for up to eight monthly cycles [55,56]. We restrict use of GO to patients with CD33-expressing leukemic blasts (although the threshold expression level has not been defined). The GO label includes a boxed warning about the risk of hepatotoxicity, including severe or fatal hepatic sinusoidal obstruction syndrome (hepatic veno-occlusive disease) [56].

For patients who achieve a CR, maintenance therapy with GO is described below. (See 'Maintenance therapy' below.)

●Outcomes – Induction therapy with single-agent GO achieved one-year survival in about one-quarter of older adults with AML [56]. A trial that randomly assigned 237 patients ≥61 years to single-agent GO versus BSC reported that GO achieved superior median OS (5 versus 4 months; HR 0.69, 95% CI 0.53-0.90) and one-year OS (24 versus 10 percent) [55]. Rates of serious AEs were similar in both arms of the study.

GO is approved by the US FDA as a single agent for newly diagnosed AML [56]. Treatment with GO for de novo or relapsed/refractory AML in medically-fit adults is discussed separately. (See "Induction therapy for acute myeloid leukemia in medically-fit adults", section on 'Other agents'.)

Clofarabine — Single-agent clofarabine can achieve responses in medically-unfit patients with AML [57,58].

In a phase 3 trial, 406 medically-unfit adults with AML/high-risk MDS were randomly assigned to single-agent clofarabine versus LoDAC [59]. Compared with LoDAC, clofarabine achieved superior rates of overall response (38 versus 19 percent, respectively; HR 0.41 95% CI 0.3-0.6) and CR (22 versus 12 percent). Two-year OS with clofarabine was 13 percent, but this was not superior to LoDAC. Clofarabine caused greater myelosuppression and gastrointestinal and hepatic toxicity.

Clofarabine is not approved for treatment of AML by the US FDA or the EMA.

Treatment with clofarabine plus LoDAC is discussed above. (See 'Low-dose cytarabine' above.)

Gilteritinib — Gilteritinib is an oral inhibitor of mutant FLT3 that is used for treatment of relapsed or refractory AML, but its role for treatment of de novo AML is not defined at present. The FDA label for gilteritinib carries a warning about differentiation syndrome; other significant toxicities include posterior reversible encephalopathy syndrome (PRES), prolonged QT interval, and pancreatitis.

Gilteritinib monotherapy was superior to salvage chemotherapy in 247 adults with relapsed or refractory AML with a FLT3 mutation, as described separately [60]. (See "Treatment of relapsed or refractory acute myeloid leukemia", section on 'Targeted agents'.)

FRAIL PATIENTS — We consider frail patients to be those whose debility or comorbid conditions would not permit treatment aimed at modifying the disease course of AML (see 'Classification of fitness' above). Note that our use of the term frail for managing patients with hematologic malignancies may differ from other definitions of frailty. (See "Frailty", section on 'Concepts and definitions'.)

It is important to carefully discuss the goals of care, outcomes, and available resources (eg, assistance for personal and medical needs). (See 'Goals, benefits, and timing of treatment' above.)

Supportive care — Management of the frail patient should focus primarily on supportive care to improve the quality of life by lessening symptoms caused by AML. Supportive care for patients with AML includes transfusions with red blood cells and platelets, antibiotics to treat infections, growth factor support, and correction of coagulopathies. Prospective and retrospective studies report that, compared with supportive care alone, various treatments for AML are associated with improved survival, better relief of symptoms, and comparable days spent in the hospital, as described above. (See 'Goals, benefits, and timing of treatment' above.)

Other treatments

●Hydroxyurea may be useful to avoid symptomatic leukostasis in patients with hyperleukocytosis blast counts (eg, >50,000 blasts/microL) or to manage discomfort from massive organomegaly. Management of leukostasis is discussed separately. (See "Hyperleukocytosis and leukostasis in hematologic malignancies".)

●Targeted agents – Some frail patients with certain molecular features may choose treatment that would be offered to the medically-unfit, non-frail patient. As an example, some patients might accept treatment with an IDH inhibitor for AML with an IDH mutation, because they place greater weight on possible benefits than on the adverse effects. (See 'IDH inhibitors' above.)

RESPONSE ASSESSMENT — We assess response to treatment based on history, physical examination, and laboratory evaluation. We generally do not create a predefined schedule of routine bone marrow examinations because if blood counts return to near normal, we consider that there is little to learn from repeating a bone marrow biopsy. However, for patients with prolonged pancytopenia, a bone marrow biopsy may be helpful to distinguish residual or recurrent AML from treatment-related hypoplasia.

●Response definitions – For medically-unfit patients, we judge response to treatment based on blood counts, transfusion needs, and relief of symptoms, as follows:

•Hematologic complete remission (CR) – Normalization of the complete blood count (CBC), including absolute neutrophil count (ANC) ≥1000/microL and platelets ≥100,000/microL, transfusion-independence, and relief of AML-related symptoms.

•CR with incomplete hematologic recovery (CRi) – Transfusion-independent and meets above criteria for CR, but without complete platelet recovery (usually) or ANC recovery.

•Hematologic improvement – Improvements in CBC, but not to levels that define CR/CRi, with ongoing transfusion needs and/or inadequate symptom relief. Hematologic improvement is a reasonable goal for frail patients, in whom the goals are to reduce transfusion needs and/or provide symptomatic relief.

•Refractory disease – No meaningful improvement in CBC, reduction of ongoing transfusion needs, and/or inadequate symptom relief.

For patients who achieve CR or CRi and whose clinical condition has improved to the point that they now meet criteria for medically-fit, we perform a bone marrow examination to document the response morphologically. Classification of medical fitness and definitions of morphologic response are described above and separately. (See 'Determining medical fitness' above and "Remission criteria in acute myeloid leukemia and monitoring for residual disease", section on 'Definition of complete remission (CR)'.)

For patients with partial response or refractory AML, management is discussed below. (See 'Relapsed or resistant AML' below.)

POST-REMISSION MANAGEMENT — Medical fitness and the patient's preferences should inform post-remission management. We reassess fitness after completing initial therapy because, for some patients, treatment of AML is associated with an improvement in medical fitness, as the effects of AML and its complications respond to therapy. (See 'Determining medical fitness' above.)

Consolidation therapy — For patients who are medically-fit after achieving a hematologic complete remission (CR), we perform a bone marrow examination. For those who have a morphologic CR, intensive consolidation therapy is an acceptable option. Consolidation therapy and its benefits and risks are discussed separately. (See "Post-remission therapy for acute myeloid leukemia in younger adults".)

For patients who remain medically-unfit or frail after achieving CR, we advise not treating with consolidation therapy because the adverse effects generally outweigh the potential benefits. Medically-unfit patients may be candidates for maintenance therapy, as described below.

Maintenance therapy — For medically-unfit, but not frail patients who are responding and tolerating treatment, we generally continue the prior treatment indefinitely as maintenance therapy, to prolong the response and/or extend survival. While some studies have arbitrarily stopped maintenance therapy after one or two years, the optimal duration of maintenance therapy is undefined and should be individualized. Some patients may favor observation rather than maintenance therapy, because they place greater emphasis on reducing the burdens or toxicity of continued treatment.

Over the course of maintenance therapy, recovery of blood counts may require dose reduction, increased intervals between cycles, and/or brief treatment interruptions. Examples of maintenance therapy in this setting include:

●Hypomethylating agents (HMA) – Patients who were treated with an HMA can continue to receive that drug as maintenance therapy for as long as they are responding and tolerating treatment.

Importantly, the oral formulation of azacitidine is pharmacokinetically and pharmacodynamically distinct from injectable azacitidine; these distinct drugs should not be considered interchangeable [61,62].

•Oral azacitidine (CC-486) – Maintenance therapy with an oral formulation of azacitidine (CC-486) was associated with superior overall survival (OS) and relapse-free survival (RFS) when compared with placebo in newly diagnosed patients ≥55 years in first CR after intensive chemotherapy who were not eligible for allogeneic hematopoietic cell transplantation (HCT) [63]. In the multicenter QUAZAR trial, 472 patients were randomly assigned to CC-486 (300 mg orally once daily on days 1 to 14 of 28-day cycles) versus placebo. Compared with placebo, CC-486 achieved superior median OS (25 versus 15 months; hazard ratio [HR] 0.69; 95% CI 0.55-0.86) and RFS (10 versus 5 months). Maintenance therapy with CC-486 appears to delay relapse rather than prevent it, as the survival curves for CC-486 and placebo treatment in this study ultimately converged. Neutropenia was the most frequent grade ≥3 adverse event (41 versus 24 percent for CC-486 versus placebo, respectively); most other adverse effects were mild gastrointestinal symptoms (diarrhea, nausea, vomiting). Overall health-related quality of life was preserved during CC-486 treatment.

Oral azacitidine is approved by the US Food and Drug Administration (FDA) for continued treatment of AML in first CR or CR with incomplete hematologic recovery (CRi) following intensive induction chemotherapy in patients who are not able to complete intensive curative therapy [64]. No age limit is specified in the FDA label, but it is unclear if the benefits of CC-486 maintenance will extend to other patients (eg, <55 years, eligible for HCT) than those in the trial above.

•Parenteral azacitidine – In a study of parenteral azacitidine, 116 patients ≥60 years old who achieved CR or CRi were randomly assigned to observation versus azacitidine maintenance (50 mg/m², subcutaneously, days 1 to 5, every four weeks) until relapse, for a maximum of 12 cycles [65]. Compared with observation, maintenance azacitidine was associated with superior 12-month disease-free survival (64 versus 42 percent), but OS did not differ between the groups.

●Isocitrate dehydrogenase (IDH) inhibitors – Treatment with an IDH inhibitor can continue for at least two years, if the patient has a continued response and is tolerating the agent [43].

●Gemtuzumab ozogamicin – For patients who received gemtuzumab ozogamicin (GO), maintenance therapy with single-agent GO is an acceptable option. For maintenance therapy, single-agent GO is administered at 2 mg/m² on day 1 every 4 weeks for up to 8 cycles [55,56].

FOLLOW-UP — We individualize follow-up according to clinical needs (eg, transfusion support), severity of symptoms, and informed by concerns on the part of the patient and clinicians. For the first two years, we generally schedule visits every one to three months, which include interval history and physical examination, complete blood count, and serum chemistries. We do not perform routine surveillance bone marrow examinations, unless warranted by a change in blood counts or other worrisome features.

RELAPSED OR RESISTANT AML — AML that recurs after an initial response (relapsed AML) or that fails to respond to induction therapy (resistant AML) has a grim prognosis in patients who are medically-unfit. Participation in a clinical trial is the preferred option.

Outside of a trial, there is no consensus regarding a preferred approach to management of relapsed or refractory AML. There is no evidence of benefit to treatment with a different hypomethylating agent (eg, azacitidine in a patient previously treated with decitabine, or vice versa) for patients who did not have an adequate response to the initial hypomethylating agent (HMA). We consider treatment with a targeted agent (if an actionable mutation is present), gemtuzumab ozogamicin (if available), or palliative management to be acceptable options.

CLINICAL TRIALS — Often there is no better therapy to offer a patient than enrollment onto a well-designed, scientifically valid, peer-reviewed clinical trial.

Additional information and instructions for referring a patient to an appropriate research center can be obtained from the United States National Institutes of Health (www.clinicaltrials.gov).

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Acute myeloid leukemia".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient education" and the keyword(s) of interest.)

●Basics topics (see "Patient education: Acute myeloid leukemia (AML) (The Basics)" and "Patient education: Leukemia in adults (The Basics)")

●Beyond the Basics topics (see "Patient education: Acute myeloid leukemia (AML) treatment in adults (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●Management of acute myeloid leukemia (AML) is informed by the individual's medical fitness for intensive antileukemic therapy. Medical fitness is based on performance status (PS) and physiologic function, not age; we do not apply age limits when judging medical fitness. (See 'Determining medical fitness' above.)

•Fitness assessment – We categorize medical fitness based on clinical evaluation, laboratory studies, and:

-Eastern Cooperative Oncology Group (ECOG) PS (table 1).

and

-Physiologic fitness (eg, comorbid conditions, activities of daily living, physical performance tests, cognition), as measured by the Charlson comorbidity index (CCI) (table 2).

•Classification of fitness – There are no clear distinctions between categories of fitness, but following are suggestions for classifying medical fitness of a patient with AML (see 'Classification of fitness' above):

-Medically-fit – Patients who can tolerate intensive treatment for AML (eg, ECOG PS 0 to 2 and CCI 0 to 2); management of medically-fit patients is discussed separately. (See "Induction therapy for acute myeloid leukemia in medically-fit adults", section on 'Medically-fit patients'.)

-Medically-unfit, but not frail – Patients are unlikely to tolerate intensive therapy due to impaired PS or comorbid conditions (eg, ECOG PS 3 or CCI 3) but are judged to be able to tolerate lower-intensity treatments.

-Frail – Patients whose debility or comorbid conditions (eg, ECOG PS ≥3 and CCI ≥3) would not permit treatment aimed at modifying the disease course. Note that our use of the term frail for managing patients with hematologic malignancies may differ from other definitions of frailty. (See "Frailty", section on 'Concepts and definitions'.)

●Medically-unfit, but not frail – For patients who are medically-unfit, but not frail, selection of therapy depends on whether the patient is eligible and accepts treatment with a hypomethylating agent (HMA; eg, azacitidine, decitabine); availability of venetoclax (BCL2 inhibitor); and the presence of an actionable mutation (eg, mutated IDH1 or IDH2) (algorithm 1).

•Eligible for HMA therapy – For patients who are eligible/accepting of HMA-based therapy, our approach follows:

-If venetoclax is available – Where venetoclax is available, we suggest treatment with venetoclax plus an HMA, rather than an HMA alone, venetoclax alone, low-dose cytarabine (LoDAC), or supportive care alone (Grade 2C). Compared with an HMA alone, addition of venetoclax to an HMA is associated with higher response rates and little additional toxicity, and a randomized trial demonstrated that venetoclax plus azacitidine improved survival compared with azacitidine alone. Selection of a regimen of venetoclax plus an HMA, administration, toxicity, and details of outcomes are discussed above. (See 'HMA plus venetoclax' above.)

-If venetoclax is not available – When venetoclax is not available, we suggest HMA-based therapy or a targeted agent (when applicable and available), rather than LoDAC or supportive care alone (Grade 2C). This suggestion is based on prospective trials in which, compared with LoDAC, HMAs were associated with comparable toxicity but more favorable responses; compared with supportive care alone, HMAs were associated with improved relief of symptoms and may prolong survival. Both azacitidine and decitabine are well-tolerated and they have comparable efficacy for newly diagnosed AML; factors that might influence the choice of azacitidine versus decitabine, specific regimens, toxicity, and outcomes are discussed above. (See 'HMA monotherapy: choice of drug' above.)

For patients with a mutation of IDH1 or IDH2, treatment with an IDH inhibitor (where available) is an acceptable alternative to HMA-based therapy, based on limited data that these oral agents have low toxicity and activity against AML, which is comparable to HMAs. Administration, toxicity, and outcomes with IDH inhibitors are discussed above. (See 'IDH inhibitors' above.)

•Ineligible for HMA therapy – For patients who are not candidates for an HMA (eg, prior treatment with an HMA, significant liver or kidney dysfunction) and those who decline HMA-based treatment (eg, to reduce clinic visits and/or hospitalization), alternative treatment options include targeted agents and LoDAC. Some experts consider such treatment acceptable for a broader population of medically-unfit patients. Patients may choose an alternative approach because they place a high value on convenience and low toxicity and/or are willing to accept an uncertain, but possibly lower likelihood of achieving remission or longer-term benefit.

-Actionable mutation – For patients who have an actionable mutation, we suggest treatment with a targeted agent, such as ivosidenib for an IDH1 mutation or enasidenib for an IDH2 mutation, rather than LoDAC (Grade 2C). IDH inhibitors have not been directly compared with other approaches, but they are associated with acceptable response rates and modest toxicity, but they are not universally available and long-term outcomes are not well-defined. Administration, toxicity, and outcomes with IDH inhibitors are discussed above. (See 'IDH inhibitors' above.)

-No actionable mutation – For patients who are ineligible for an HMA and do not have an actionable mutation/available targeted agent, we suggest a LoDAC-based regimen, rather than other treatments (Grade 2C). Our preference for LoDAC in this setting is based on its well-established track record, modest toxicity, convenience, widespread availability, tolerability in patients with kidney and liver dysfunction, and moderate efficacy. Selection of LoDAC monotherapy versus a LoDAC-based combination, toxicity, and outcomes are discussed above. (See 'Low-dose cytarabine' above.)

●Frail patients – For medically frail patients, the mainstay of management is supportive care to relieve symptoms and to improve the quality of life. Some frail patients may, instead, choose to receive lower-intensity treatments, as discussed above. (See 'Frail patients' above.)

●Post-remission management – For patients who achieve a complete hematologic response, management should be individualized according to medical fitness and patient preferences. We reassess fitness upon completion of initial therapy because, for some patients, treatment of AML is associated with an improvement in medical fitness, as the effects of AML and its complications respond to therapy.

Options for maintenance therapy for medically-unfit patients are described above. (See 'Maintenance therapy' above.)