How Genetics Can Drive Initial Therapy Choices for Older Patients with Acute Myeloid Leukemia

Introduction

Acute myeloid leukemia (AML) is the most common acute leukemia in adults with approximately 80% of new cases occurring in those aged ≥60 years [1]. Despite its frequency in older adults, care of this population remains clinically challenging. Multiple upfront treatment options exist, which clinicians must choose between while considering aging-related conditions (e.g., comorbidities, functional impairment), therapy benefits and risks, patient preferences, and unique disease characteristics compared to younger adults. This has led to efforts to create specific recommendations for older patients in consensus treatment algorithms published by the National Cancer Center Network (NCCN) which defines older adults as those aged ≥60 years, and the European Leukemia Network (ELN) which uses a cutoff of ≥65 years [2, 3]. Nonetheless, outcomes of older adults with AML remain poor; five-year overall survival (OS) for adults aged ≥60 years is approximately 3–10% compared to 60% in those aged ≤50 years [1, 4, 5].

Beyond aging-related conditions, older patients also have higher rates of disease characteristics associated with unfavorable survival outcomes [6–9]. This includes higher rates of secondary and therapy-related AML, unfavorable cytogenetic and molecular features, and increased risk of resistance to chemotherapy [7]. Older adults are also underrepresented in clinical trials, therefore limiting access to new therapies and the generalizability of efficacy and safety data to their clinical care [10]. As a result, older adults are frequently undertreated, with as many as 60% of this group not offered any anti-leukemic therapy following diagnosis due to concerns about treatment-related toxicities, despite their adverse disease characteristics [4, 11]. Yet recent studies have revealed that older patients even up to age 80 may benefit from leukemia-directed therapy, including intensive induction, and that survival rates improve when these treatments are offered [4, 11]. Additionally, newer targeted agents and lower-intensity regimens have extended the arsenal of treatments available to older patients less fit for intensive induction.

In addition to assessment of aging-related conditions, cytogenetic and molecular testing should guide initial treatment [12, 13]. In this article, we briefly review the genomic characteristics of older adult AML and available upfront treatment options as guided by our current understanding of clinically relevant cytogenetic and molecular changes in AML.

The genomic landscape of older adults with AML

The biology of older adult AML has patterns distinct from that of AML affecting younger adults. Specifically, higher frequencies of mutations associated with poor outcomes occur in older populations, including those in transcription factors RUNX1 and BCOR, tumor suppressor gene TP53, epigenetic regulators ASXL1, DNMT3A, and TET2, spliceosome factor SRSF2, and drug-targetable metabolic genes IDH1 and IDH2 [13–19]. Unfavorable-risk cytogenetics such as a complex karyotype, del(5), del(7), and abnormalities in 17p are also more frequent in the older AML population [17, 18]. Meanwhile favorable-risk mutations and cytogenetics including t(8;21), inv(16), and biallelic CEBPA decrease with age [14, 17–19]. The presence of drug-targetable FLT3-ITD and FLT3-TKD mutations do not significantly increase with age and per some reports, FLT3-ITD mutations may even decrease with age [18, 20, 21]. Common genetic features of older adult AML, and their frequencies (for those ≥ 5%), are summarized in Table 1.

Table 1.

Common molecular and cytogenetic features in older adult AML

Genetic Mutation	Risk Stratification by ELN [2, 15, 80]	Frequency in adults < 60 [16, 18, 19]	Frequency in adults ≥ 60 [14, 16–19]
TET2		8–10%	22–42%
DNMT3A		12–19%	21–35%
NPM1*	Favorable to intermediate risk†	19–34%	16–32%
FLT3-ITD*	Intermediate to adverse risk#	23–32%	18–27%
RUNX1*	Adverse risk	9–10%	17–23%
SRSF2		6%	14–25%
ASXL1*	Adverse risk	7–8%	15–22%
IDH2		10–11%	11–18%
IDH1		5–7%	7–17%
TP53*	Very adverse risk	4–6%	10–14%
NRAS		12–24%	4–19%
BCOR		5%	10%
FLT3-TKD*		6%	7–12%
CEPBA*	Very favorable risk§	9–17%	5–10%
Cytogenetic Abnormality
−7 or 7q−	Adverse risk	3–9%	8–19%
−5 or 5q−	Adverse risk	2–7%	8–17%
+8	Intermediate risk	6%	9%
17p abnormality	Adverse risk	2%	8%
Cytogenetic Risk Group
Unfavorable¶	Adverse risk	11–33%	21–42%
Intermediate‡	Intermediate risk	46–70%	53–75%
Favorable††	Favorable risk	14–19%	4–5%

^*Recommended as part of genetic screening at diagnosis [3].

^†Favorable risk for mutated NPM1 without FLT3-ITD or with FLT3-ITD allelic ratio 0.5. Intermediate risk for mutated NPM1 with FLT3-ITD allelic ratio 0.5.

^#Intermediate risk for wild-type NPM1 without FLT3-ITD or with FLT3-ITD allelic ratio 0.5. Adverse risk for wild-type NPM1 with FLT3-ITD allelic ratio 0.5. See NPM1 for risk stratification of AML with mutated NPM1 and concurrent FLT3-ITD mutations.

^§Very favorable risk for biallelic mutation of CEBPA.

^¶Unfavorable cytogenetic risk group includes patients with t(6;9)(p23;q34.1), t(v;11q23.3), t(9;22)(q34.1;q11.2), inv(3)(q21.3q26.2), t(3;3)(q21.3;q26.2), −5, del(5q), −7, −17, abn(17p), complex karyotypes (3 or more unrelated chromosomal abnormalities), and monosomal karyotypes

^‡Intermediate cytogenetic risk group includes patients with t(9;11)(p21.3;q23.3) and other cytogenetic abnormalities not characterized as favorable or adverse.

^††Favorable cytogenetic risk groups includes patients with t(8;21)(q22;q22.1), inv(16)(p13.1q22), and t(16;16)(p13.1;q22) unless complex cytogenetic abnormalities were also present.

Epigenetic differences such as altered methylation patterns have also been noted in older vs. younger patients, though the prognostic and therapeutic implications of these changes are less clear [14]. Genetic changes facilitating drug resistance via increased expression of the multidrug resistance glycoprotein (MRD1) are more common in older patients and contribute to worse outcomes [7]. Additionally, age-associated del(7) and SRSF2 mutations have been associated with multidrug resistance to tyrosine kinase inhibitors (TKIs), cytarabine, and daunorubicin [17, 18, 22, 23].

Initial assessment of older adults with AML

The increased prevalence of unfavorable-risk genetics and drug resistance with rising age should be used to motivate discovery of new therapeutic agents and mechanisms to bypass drug resistance to improve outcomes. Age alone should not be used as a sole justification for reduction of therapy intensity [4, 11]. When evaluating newly diagnosed older patients with AML, we suggest the following framework to help guide decisions regarding treatment intensity and specific regimen selection.

Clarify genomic risk

Karyotyping, fluorescence-in-situ hybridization, and next-generation sequencing should be used to investigate the genomic features of each new patient’s AML. ELN genetic risk stratification can then be employed to assess prognosis and has been adopted by multiple society consensus guidelines including the NCCN and the European Society of Medical Oncology (ESMO) [3, 24]. Refinement of these categories is ongoing; for instance, there is evidence that mutations in TP53 with complex karyotype, GATA2, MECOM/EVI1, or the presence of inv(3)(q21.3q26.2) or t(3;3)(q21.3q26.2) are associated with very adverse prognosis [15, 16].

While historically there has been concern that treatment delays to wait for genomic tests could harm patient outcomes, there is data to suggest this is not the case. In a study of 599 French patients (40% were aged ≥60 years), a delay in treatment with a median time-to-treatment of 8 days was not associated with early death, complete response (CR) rate, or OS in multivariable analysis [25]. More recently, a preliminary analysis of the Beat AML Master trial of 395 patients aged ≥60 years of age demonstrated the feasibility of obtaining cytogenetic and NGS testing to target induction therapy [26]. Approximately 95% of patients had these tests completed within 7 days and median OS for patients assigned to three investigational sub-studies on the basis of genetic features was significantly longer than those who received standard of care treatment (12.8 vs 3.9 months) [26].

Clarify patient fitness

A geriatric assessment (GA) should be used to assess aging-related conditions that are predictive of mortality and morbidity [6, 10, 27, 28]. Feasibility and tools have previously been described [10, 28].

Treatment

Candidates for intensive therapy

Favorable to intermediate ELN risk disease

For older patients deemed fit to undergo intensive chemotherapy who have favorable to intermediate-risk genomic features per ELN, intensive induction is recommended as the benefits generally outweigh the risks (Figure 1). Combination “7+3” regimens containing 100–200 mg/m2 cytarabine plus anthracycline (daunorubicin or idarubicin) are most commonly used. The HOVON43 trial of adults aged ≥60 receiving either 45 mg/m2 or 90 mg/m2 of daunorubicin demonstrated a CR rate of 82% and 2-year OS of 60% for those with favorable-risk disease, and a CR rate of 60–65% with a 2-year OS of 31–34% for those with intermediate-risk disease [2, 29]. Thirty-day mortality was approximately 12%, and 52% of patients experienced grade 3–4 adverse events [29]. Subgroup analysis in HOVON-43 demonstrated that only patients aged 60–65 years, and patients with core-binding factor mutations (CBF) had a survival benefit with 90 mg/m2 daunorubicin, though no significantly increased toxicities were seen with higher dosing [29]. NCCN guidelines currently recommend use of 60–90 mg/m2 daunorubicin for older adults [3, 29]. Idarubicin is also an NCCN recommended option and has demonstrated similar outcomes and toxicities in older adults compared to mid-dose intensity daunorubicin 60–80 mg/m2 [3, 30].

Figure 1.

A proposed approach for older adults with acute myeloid leukemia. Abbreviations: AML = acute myeloid leukemia. AML-MRC = acute myeloid leukemia with myelodysplasia-related changes. IDH = isocitrate dehydrogenase. FLT3 = FMS-like tyrosine kinase 3. ITD = internal tandem duplication. TKD = tyrosine kinase domain. 7+3 = cytarabine (7 days) and anthracycline (3 days). GO = gentuzumab ozogamicin. HMA = hypomethylating agents. CPX-351 = liposomal cytarabine and daunorubicin. AZA = 5-azacytidine. LDAC = low-dose cytarabine.

Older patients who are fit with favorable to intermediate-risk disease and who are CD33 positive appear to benefit from the addition of anti-CD33 drug-antibody conjugate gemtuzumab ozogamicin (GO) to 7+3 [3]. A study of patients with a median age of 67 demonstrated no increased toxicities and a statistically significant 3-year OS improvement in patients who received 7+3 with a single dose of 3 mg/m2 GO given day 1, vs. those who received 7+3 alone (25% vs. 20%) [31]. ALFA-0701 examined patients aged 50–70 years without secondary AML receiving 3 mg/m2 GO on days 1, 4, and 7 in conjunction with 7+3 and demonstrated improved event free survival (EFS) and OS at a median follow up of 14.8 months (hazard ratio [HR] 0.58, 95% confidence interval [CI] 0.43–0.78, and HR 0.69, 95% CI 0.49–0.98, respectively) [32]. The benefit was especially pronounced in patients with favorable to intermediate-risk cytogenetics and preserved in patients ≥60 in subgroup analyses [32, 33]. Higher hematologic toxicity comprised mostly of prolonged thrombocytopenia was observed without an increase in mortality [32, 33].

FLT3 TKD or ITD mutations

Older patients who are candidates for intensive induction and have FLT3 mutations should be considered for treatment with midostaurin (a multitargeted kinase inhibitor) plus standard 7+3 [3]. In a Phase 2 single-arm study including patients up to age 70, the addition of midostaurin on day 8 following 7+3 resulted in a 2-year EFS of 34% and 2-year OS of 46% [34]. Propensity score-weighted analysis showed significant improvement in EFS compared to historical induction controls [34]. A phase 3 study of patients aged <60 years demonstrated improved OS with the addition of midostaurin following 7+3 vs. 7+3 alone (74.7 vs. 25.6 months) with benefit retained in both TKD and ITD subgroups [35]. Adverse events and subsequent treatment discontinuation between the treatment arms were similar, except an increased risk of rash was seen in patients receiving midostaurin [35]. In the single-arm study of older adults, 61% of patients terminated post-induction midostaurin early, predominantly due to GI side effects and infections. [34].

Adverse-risk genomics by ELN

Those with adverse-risk disease treated with 7+3 had lower CR (56%) and 2-year OS (19%) than those with favorable-risk disease treated per HOVON43 [29]. However, attempts have been made to refine prognostication for older adults receiving 7+3 beyond ELN risk stratifications given growing recognition of the different genetic patterns seen in this population. For example, the ALFA decision tool validated a 7-gene model in patients ≥60, suggesting that some patients with adverse-risk cytogenetics by ELN still derive a reasonable 2-year OS of 39% with intensive induction [36]. Patients who did not benefit per this model were patients with poor-risk cytogenetics plus either a TP53 or KRAS mutation (2-year OS 3%) [36]. This highlights that for fit patients with ELN adverse-risk disease, 7+3 may still be considered as a treatment option, and it remains listed as such within NCCN guidelines [3]. Optimal prioritization of lower vs. higher intensity therapy options for many adverse-risk patients remains unknown due to a lack of head-to-head studies.

Notable adverse-risk subgroups

Secondary AML and AML with myelodysplasia-related changes

Secondary AML (sAML) is a challenging disease subset associated with increased resistance to conventional cytotoxic therapy and increased frequencies of adverse-risk genomic features [37–39]. Patients with AML with myelodysplasia-related changes (AML-MRC) also have high frequencies of unfavorable-risk mutations, particularly ASXL1 and TP53 [40].

Standard 7+3 for fit patients with sAML has demonstrated a trend towards lower CR rates compared with patients with de novo AML (63% vs. 67%) and significantly decreased overall survival (4-year OS 25.5% vs. 39.5%) [41, 42]. However, in recent years, CPX-351 (Vyxeos) has emerged as a new therapeutic option. A phase III trial of CPX-351 vs. 7+3 in older patients (60–75 years) with newly diagnosed sAML or AML-MRC reported significantly improved composite CR rates, defined as CR or complete remission with incomplete hematologic recovery (CRi), of 47.7% vs. 33.3%, respectively [43]. OS was also significantly improved with CPX-351 vs. 7+3 therapy, with a median OS of 9.56 vs 5.95 months, and 1-year OS of 41.5% vs. 27.6% [43]. Five-year follow up continued to show the survival benefit of CPX-351 with a median OS of 9.33 vs. 5.95 months, and 5-year OS of 18% vs. 8%, respectively [44].

Following induction, fit patients should be considered for consolidation with allogeneic hematopoietic stem cell transplant (HSCT). Patients who underwent HSCT and were initially treated with CPX-351 had better OS than patients who received standard 7+3 (median OS was not reached vs. 10.25 months) [44]. At 3 years, OS was 56% for patients initially treated with CPX-351 vs. 23% for patients treated with 7+3 [44].

TP53-Mutated AML

TP53 gene mutations lead to loss of tumor suppressor functions and are associated with particularly adverse prognosis in AML [45–48]. For fit patients eligible for allogeneic HSCT, induction therapy with 7+3 followed by transplant in first CR is recommended. Unfortunately, CR may not be achieved in the majority of cases. A retrospective study of patients with newly diagnosed TP53-mutated AML revealed poor CR rates of only 28% vs. 50% in the unmutated group [15].

Hypomethylating agents (HMA) with or without the B-cell lymphoma 2 (BCL-2) inhibitor, venetoclax, can be a reasonable therapeutic strategy in patients with TP53-mutated AML who have not responded to standard cytotoxic therapy, or in those ineligible for intensive induction chemotherapy. A prospective trial of decitabine at a dose of 20mg/m2/day for 10 days demonstrated favorable outcomes in older patients with TP53-mutated AML [49]. The 21 patients with mutated TP53-mutated AML had a 100% rate of blast clearance (bone marrow with <5% blasts) vs. 41% with wild-type (WT) TP53. No difference in OS was observed between TP53-mutated and WT TP53 cases [49]. Unfortunately, leukemia-specific mutations were detected even during morphological remission in all patients which indicated that decitabine likely leads to incomplete clearance of disease and relapse may happen in the short-term after therapy is discontinued [49]. A subsequent randomized phase II trial compared 5-day vs. 10-day schedules of decitabine in older patients with AML who were deemed ineligible for intensive therapy [50]. No significant differences in response rates were seen between patients with TP53 mutations in both groups (29% in the 5-day decitabine group vs. 47% in the 10-day decitabine group) [50]. Additionally, no difference in OS was seen [50]. The addition of venetoclax to hypomethylating agents (HMA) has also shown modest efficacy in this patient population, as discussed in subsequent sections.

Given continued poor outcomes in TP53-mutated AML, emerging novel therapies are of particular interest. Eprenetapopt (APR-246) is a novel small molecule that selectively induces apoptosis in TP53 mutant cancer cells by restoring the WT conformation of mutant p53. Based on promising preclinical results, a phase I/II study evaluated the combination of azacitidine and eprenetapopt in older patients (median age=66 years) with TP53-mutated MDS or AML [51]. An overall response rate (ORR) of 64% and CR rate of 36% were reported in the subgroup of patients with TP53-mutated AML and the combination was well-tolerated. Thirty-five percent of patients in this study were able to proceed to HSCT with a median OS of 14.7 months [51].

Magrolimab, a CD47-targeted IG4 monoclonal antibody targeted to CD47 which activates antibody-dependent cellular phagocytosis, is another promising novel therapy for patients with TP53 mutations. In combination with azacitidine, it has shown encouraging outcomes in a phase 1b trial of patients at a median age of 73, with an ORR of 71%, CR of 48%, and median OS of 12.9 months in TP53-mutated AML [52]. Relevant recruiting clinical trials of magrolimab for older adults with AML are outlined in Table 2.

Table 2.

Selected actively recruiting older adult AML trials

Eligible Genomic/Prognostic Group	Trial Identifier	Abbreviated Description	Phase
All	NCT03013998	Biomarker-Based Treatment of AML	Ib/II
All	NCT03226418	Integrating Geriatric Assessment and Genetic Profiling to Personalize Therapy Selection in Older Adults With AML	II
Adverse-risk (not including FLT3)	NCT04817241	Oral Decitabine and Cedazuridine (ASTX727) in Combination With Venetoclax for Higher-Risk AML Patients	Ib/II
Adverse-risk, complex karyotype, secondary AML, TP53	NCT04435691	Magrolimab, Azacitidine, and Venetoclax for the Treatment of AML	II
FLT3	NCT05010122	ASTX727, Venetoclax, and Gilteritinib for the Treatment of Newly Diagnosed, Relapsed or Refractory FLT3-Mutated AML or High-Risk MDS	I/II
FLT3	NCT03836209	Gilteritinib vs Midostaurin in FLT3 Mutated AML	II
FLT3-ITD	NCT04518345	TP-0903 for the Treatment of FLT3 Mutated AML	Ib/II
IDH1	NCT02074839	Orally Administered AG-120 in Subjects With Advanced Hematologic Malignancies With an IDH1 Mutation	I
IDH1	NCT03471260	Ivosidenib and Venetoclax With or Without Azacitidine in Treating Patients With IDH1 Mutated Hematologic Malignancies	Ib/II
IDH1	NCT04493164	CPX-351 and Ivosidenib for the Treatment of IDH1 Mutated AML or High-Risk MDS	II
IDH2	NCT03728335	Enasidenib as Maintenance Therapy in Treating Patients With AML With IDH2 Mutation After Donor Stem Cell Transplant	I
IDH2	NCT04092179	IDH2 Inhibitor Enasidenib in Combination With BCL2 Inhibitor Venetoclax in Patients With IDH2-Mutated Myeloid Malignancies	Ib/II
IDH1 and IDH2	NCT04774393	Decitabine/Cedazuridine and Venetoclax in Combination With Ivosidenib or Enasidenib for the Treatment of Relapsed or Refractory AML	Ib/II
IDH1 and IDH2	NCT03839771	Ivosidenib or Enasidenib in Combination With Induction Therapy and Consolidation Therapy in Patients With Newly Diagnosed AML or MDS With Excess Blasts and an IDH1 or IDH2 Mutation	III
Secondary AML	NCT04848974	Uproleselan, Cladribine, and Low Dose Cytarabine for the Treatment of Patients With Treated Secondary AML	Ib/II
Secondary AML	NCT04763928	Treatment Combination including Decitabine and Venetoclax in Patients With AML Secondary to MPNs Unfit for Intensive Chemotherapy	II

Patients not suitable for intensive therapy

For older patients who are not candidates for or choose not to receive intensive therapy, an increasing number of treatment options have been approved since the 1990s. It is prudent to note that newer trials report responses based on genomic features in subgroup analyses with increasing frequency; this can further guide treatment selection.

Hypomethylating agents and venetoclax

All older patients who are not candidates for intensive therapy should be strongly considered for treatment with a HMA plus venetoclax on the basis of recent practice-changing trials in older adults [53–55]. Efficacy and safety were demonstrated in a phase 1b study of adults aged ≥65 years without favorable-risk features by NCCN and not fit for intensive therapy, who received any HMA (either 7 day azacitidine or 5 day decitabine) plus 400 mg/day venetoclax. CR+CRi rate was 74% [54, 55]. Median OS was 16.3 months [54, 55]. Patients with adverse-risk disease retained benefit with this regimen, with a CR+CRi of 67%, though their duration of remission (DOR) remained shorter compared to those with intermediate-risk disease (7.8 vs. 26.5 months) [54, 55]. In the subsequent phase 3 VIALE-A study of azacitidine plus venetoclax vs. azacitidine alone, outcomes were significantly better in the venetoclax-containing arm [53]. CR+CRi was 66% vs. 28.3%, and median OS was 14.1 vs. 9.6 months [53]. Notably, the azacitidine plus venetoclax arm achieved a higher CR+CRi and median OS even in secondary AML (median OS 16.4 vs. 10.6 months in the control group) [53]. Patients with TP53 or NPM1 mutations were also more likely to achieve a response with the addition of venetoclax. In the TP53 subgroup CR+CRi was 55.3% vs. 0% with azacitidine+venetoclax vs. azacitidine alone, with a trend toward OS benefit (HR 0.76, 95% CI 0.40–1.45) [53]. The NPM1 subgroup similarly had an improved CR+CRi of 66.7% vs. 23.5%, again with a trend toward OS benefit (HR 0.73, 95% CI 0.36–1.51) in the azacitidine+venetoclax arm [53]. Smaller sample sizes may have limited the detection of statistically significant OS improvements in these subgroups [53]. For all patients, grade 3–4 adverse events occurred at similar rates between the two arms, however more infections, cytopenias, and febrile neutropenia occurred in the venetoclax-containing arm [53].

Notably, patients with targeted therapy options also benefited from combination HMA plus venetoclax in VIALE-A, and this is a recommended first-choice for lower-intensity therapy in those with IDH1/2 or FLT3 mutations [53, 55]. Patients with IDH1/2 mutations had a significantly improved CR+CRi rate (75.4% vs. 10.7%) and 1-year OS (66.8% vs 35.7%) with azacitidine+venetoclax vs azacitidine alone [53]. Those with FLT3 mutations had significantly improved CR+CRi rates (72.4% vs. 36.4%) with azacitidine+venetoclax, with a trend toward an OS benefit (HR 0.66, 95% CI 0.35–1.26) though statistical significance was not reached in this smaller sample [53].

Prior studies of HMA plus venetoclax specifically excluded older adults with favorable-risk disease and those eligible for intensive therapy. However, promising outcomes in older patients with intermediate to poor-risk AML raise the question: is HMA plus venetoclax superior to intensive therapy in older adults? While we lack head-to-head studies comparing these regimens, two retrospective propensity-score matched analyses comparing intensive chemotherapy to either azacitidine plus venetoclax, or decitabine plus venetoclax suggest improved outcomes with HMA and venetoclax for older adults [56, 57]. Patients aged＞60 years receiving 10-day decitabine plus venetoclax had significantly higher CR+CRi rates (81% vs. 52%) and median OS (12.4 vs. 4.5 months) compared to a matched cohort of patients who received intensive induction regimens including at least moderate doses of cytarabine ≥1g/m2/day [57]. Similarly, comparison of matched cohorts of adult patients receiving azacitidine plus venetoclax vs. patients receiving 7+3 demonstrated significantly improved OS for patients aged ≥65 years with azacitidine and venetoclax (HR 0.27, 95% CI 0.1–0.6) [56]. Azacitidine and venetoclax also appeared to improve OS in patients of any age with adverse-risk disease (HR 0.424, 95% CI 0.2–0.9), with particular OS benefit noted in those with RUNX1 mutations (HR 0.081, 95% CI 0.01–0.6) [56].

Targeted therapy options

IDH1 targeted therapy

Mutations in the isocitrate dehydrogenase (IDH) genes cause DNA hypermethylation and lead to impaired hematopoietic differentiation which can be reversed by the inhibition of IDH with small molecules [58–60]. The prognosis of patients with AML and IDH mutations is still controversial [58, 59]. Ongoing trials of IDH-inhibitors in AML are outlined in Table 2.

Ivosidenib is a first-in-class targeted inhibitor of mutant IDH1 and is given orally. An analysis of patients (median age=76.5 years) with newly diagnosed IDH1-mutated AML ineligible for standard therapy demonstrated efficacy of single-agent ivosidenib [61]. The ORR was 54.5% and composite CR rate, defined as CR + CR with incomplete hematologic recovery (CRh), was 42.4%, and CR rate was 30.3% [61]. For those patients who achieved CR or CRh, 64% had IDH1 mutation clearance. Median durations of ORR, CR+CRh, and and CR were not reached, but the estimated proportions of patients remaining in remission/response at 12 months were promising at 63%, 61.5%, and 77.8%, respectively [61]. Median OS of the overall population was 12.6 months with a 1-year OS rate of 51.1% [61]. Differentiation syndrome, a known toxicity of IDH-inhibitors, occurred in 18% of patients, none of whom required therapy cessation [61]. Currently, ivosidenib is FDA-approved as monotherapy for patients with newly diagnosed, chemotherapy-ineligible IDH1-mutated AML, and relapsed/refractory disease.

Additional investigations of ivosidenib in combination with other therapies are ongoing. A phase 1b clinical trial examined ivosidenib plus azacitidine in older patients with newly diagnosed IDH1-mutated AML who were ineligible for intensive induction therapy [62]. ORR was 78.3% and CR rate was 60.9%; clearance of IDH1 was seen in 71.4% patients who achieved CR [62]. One-year OS was 82% [62]. The combination was determined to be well tolerated without dose-limiting toxicities [62]. This study led to the development of a phase III study of azacitidine with or without ivosidenib in patients with newly diagnosed AML ineligible for intensive therapy (AGILE) which further demonstrated the benefit of this combination [63]. CR rates were significantly improved to 38% with ivosidinib combination therapy vs. 11% with azacitidine+placebo. The primary outcome of EFS favored the ivosidinib-containing arm (HR 0.33, 95% CI 0.16–0.69). Median OS was also significantly prolonged to 24 months with ivosidinib+azacitidine vs. 7.9 months with azacitidine alone [63].

IDH2 targeted therapy

Enasidenib is an oral inhibitor of mutant IDH2. A phase I/II study of patients with newly-diagnosed IDH2-mutated AML ineligible for intensive chemotherapy (median age=77 years) who were treated with enasidenib monotherapy showed reasonable benefit [64]. ORR was 30.8%, CR rate was 18%, and median OS for all patients was 11.3 months, with responder median OS not reached at data cutoff. Differentiation syndrome occurred in 10% of patients [64].

Enasidenib has also been evaluated in combination with HMAs with promising results. One phase II study of enasidenib plus azacitidine vs. azacitidine alone in patients (median age=75 years) with newly diagnosed IDH2-mutated AML showed that ORR and CR rates were significantly improved in the combination group (71% vs. 42%, and 53% vs. 12%, respectively), and also demonstrated longer duration of response (24.1 vs. 12.1 months) [65]. Median OS was not significantly different at 22 months in both arms [65]. Enasidenib is currently approved in the US for the treatment of adult patients with relapsed/refractory IDH2-mutated AML and is recognized in NCCN guidelines as an off-label option for the treatment of newly diagnosed patients with IDH2 mutations on the basis of this phase II data [3].

FLT3 targeted therapy

The first generation FLT3-inhibitor, sorafenib, in combination with HMAs is an additional option for treatment of patients with FLT3-mutated AML on the basis of a phase II trial and a small case-series of older adults utilizing either azacitidine or decitabine plus sorafenib [66, 67]. Patients treated with azacitidine/sorafenib achieved an ORR of 78% and CR of 44%, with a median duration of response of 14.5 months [67]. Median OS was 8.3 months [67]. A six-patient case-series which included both patients with newly-diagnosed and relapsed/refractory FLT3-positive AML reported an ORR of 83% with a median OS of 5.2 months [66]. Treatment was well tolerated in both reports and is currently recognized in the NCCN guidelines as an alternative for patients ineligible for intensive therapy [68].

Investigational treatment approaches include triplet combinations of HMAs plus venetoclax and a FLT3-inhibitor [69]. Promising outcomes of this approach were presented at ASH 2020 with use of venetoclax, decitabine, and gilteritinib or sorafenib in 16 patients with newly diagnosed FLT3-postitive AML [70]. Composite CR (comprised of CR, CRi, and CR with incomplete platelet recovery) was 88% with FLT3-negative PCR achieved in 100% of responders [70]. Additional trials evaluating varying combinations of FLT3-inhibitor therapy with venetoclax, HMA, and other novel agents are shown in Table 2.

Patients without targetable mutations

Additional lower-intensity combination therapies

Prolonged cytopenias, which can lead to neutropenic fever and severe infections, may limit the use of HMA/venetoclax in some patients. Low-dose cytarabine (LDAC) can be used in combination with venetoclax 600 mg daily on the basis of a phase III trial which compared the combination to LDAC alone in older patients ineligible for intensive therapy (median age=76 years) [71]. CR+CRi was improved in the venetoclax-containing arm (48% vs. 13%), however there was no significant median OS benefit at the planned 1-year analysis (7.2 vs. 4.1 months, HR 0.75, 95% CI 0.52–1.07) [71]. This may have been related to imbalanced rates of adverse-risk genomics and secondary AML between the groups [71].

Glasdegib, a hedgehog pathway inhibitor, has also been used in combination with LDAC with modest benefits [72]. A phase II study of patients with a median age of 76 demonstrated improved CR (19.2% vs. 2.6%) and median OS (8.8 vs. 4.1 months) with glasdegib plus LDAC vs. LDAC alone [72]. In subgroup analyses, patients with secondary AML showed particularly favorable survival responses with use of glasdegib combination therapy, with a median OS of 9.1 vs. 6.6 months for those with sAML vs. de novo AML, respectively [72].

Single agent therapy options

In patients whom the risks of combination therapies outweigh the benefits, options include single agent HMAs, which have shown efficacy compared to best supportive care and LDAC [73–75]. Two phase III studies in older adults comparing azacitidine to conventional care (predominantly best supportive care or LDAC) demonstrated improved OS benefit depending on marrow blast burden [73, 74]. Patients with blasts <30% had a significantly improved median OS (24.5 vs. 16 months) and decreased hospitalizations compared to conventional care [74]. Patients with blasts >30% only demonstrated a trend toward improved OS which did not meet statistical significance (median OS 10.4 vs. 6.5 months, P=0.1) [73]. Subgroup analyses demonstrated improved survival in patients with adverse-risk cytogenetics (HR 0.68; 95% CI 0.5, 0.94) and AML-MRC (HR 0.69; 95% CI 0.48, 0.98) [73]. Decitabine similarly led to improved median OS compared to BSC or LDAC (8.5 vs. 5.3 months) though this only met significance after extended unplanned follow up (HR 0.82, 95% CI 0.68–0.99) [75]. Subgroup analyses indicated improved survival outcomes for those with de novo AML (HR 0.71, 95% CI 0.56–0.91) [75].

Single-agent gemtumab ozogamicin (GO) therapy in CD33 positive patients led to a median OS benefit of 4.9 months vs. 3.6 months in a phase III trial (median age=61 years) compared to BSC [76]. Significant improvement in survival was particularly observed in patients with CD33 expression >80% (HR 0.49, 95% CI 0.32–0.76) [76]. LDAC alone may also be used and has demonstrated minimally improved outcomes compared to BSC (CR 18% vs. 1%), though no benefit was observed in patients with adverse-risk cytogenetics [77].

Conclusion

Major strides have been made in recognizing the presence of higher-risk leukemia biology – including genetic alterations – in older adults. Meanwhile, the field of geriatric hematology/oncology continues to build the framework for assessment of aging-related conditions which can help decrease overtreatment and undertreatment [78]. Further investigation into the optimal sequencing of treatment for older AML patients based on specific genomic profiles and fitness levels is needed [79]. The former is being undertaken at the level of new drug trials (such as Beat AML and the upcoming MyeloMATCH SWOG/National Cancer Institute trial; Table 2) which increasingly target and report responses on the basis of genetic characteristics. The latter will be improved by continued representation of older adults in therapeutic clinical trials, with inclusion of geriatric assessments and quality-of-life-based outcomes.

OPINION STATEMENT.

Treatment of older adults with acute myeloid leukemia (AML) is challenging. Therapy decisions must be guided by multiple factors including aging-related conditions (e.g., comorbidities, functional impairment), therapy benefits and risks, patient preferences, and disease characteristics. Balancing these factors requires understanding the unique, and frequently higher-risk cytogenetic and molecular characteristics of AML in older adult populations, which should caution providers not to reduce therapy intensity on the basis of age alone. Instead, geriatric assessments should be employed to determine fitness for therapy. Treatment options in AML are increasingly targeted to specific mutations or recognized to have differential benefits on the basis of genomics, and representation of older adults and geriatric outcome reporting in clinical trials is improving. Additionally, newer studies have begun to explore personalized therapy strategies on the basis of initial genetic testing. Review and refinement of practice guidelines for older patients on the basis of these advances is needed and is anticipated to remain an important topic in ongoing hematology/oncology clinical education.