Acute Myeloid Leukemia Genomics: Impact on Care and Remaining Challenges

Acute Myeloid Leukemia Disease Classification and Survival Outcomes

Introduction

Acute myeloid leukemia (AML) remains the most common acute leukemia in adults, with approximately 20 000 adults in the United States being diagnosed each year (1). Despite significant advances in the understanding of AML biology, long-term survival for most patients remains poor, as current treatment strategies lead to long-term complete remissions (CRs) in only approximately 40% of AML patients younger than 60 years of age and in 5% to 15% of patients older than age 60. Even with adaptation of risk-stratified therapies, 20% to 30% of AML patients never achieve CR, and over 50% of patients who do achieve CR subsequently experience disease relapse (2). In both settings, disease-related outcomes are typically poor.

Thus, increasing our understanding of both disease-associated and patient-associated contributors to differential treatment response and survival is imperative to improve outcomes. A complex interplay of several leukemia-associated drivers including genetic (3), epigenetic, and genomic features as well as more recently described microenvironment-related contributors have been shown to impact disease biology and therapy response. These leukemia-associated variables can be modified by patient-associated factors, including age, sociodemographics (4), race/ethnicity (5), and co-morbidities, including a history of prior malignancy or of leukemia-associated precursor lesions such as clonal hematopoiesis of indeterminate potential (CHIP), clonal cytopenias of unclear significance (CCUS), or myelodysplastic neoplasms.

In this review we summarize different aspects of AML genomics and their impact on patient care and survival outcomes, including a review of different genomic approaches to obtain complementary information about the disease and potential treatment vulnerabilities. Further, we highlight emerging fields in AML genomics and remaining challenges towards the best care for this clinically and molecularly diverse patient population.

Disease classification: WHO 2022, highlights

While the first formal classification of myeloid neoplasms dates to William Dameshek's work published in 1951, the World Health Organization (WHO) classification of myeloid neoplasms serves as the clinical reference for understanding disease biology and associated clinical ramifications. The latest revision of the “Blue Book” in 2022, which is the 5th edition of the WHO Classification of Hematolymphoid Tumors, brought major changes to the categorization of all myeloid diseases, largely driven by increased emphasis on genomics-specific features (6). This trend, in turn, has led to a more confluent classification spectrum of myeloid neoplasms.

An example of these changes is the entity of myelodysplasia-related (MR) AML, which transitioned from a diagnostic characterization based on morphology alone to one of integrated clinical, molecular/genetic, and pathologic parameters. As a result, the diagnosis of MR-AML also can be based on pathogenic mutations detected in any 1 of 8 genes: SRSF2, SF3B1, U2AF1, ZRSR2, ASXL1, EZH2, BCOR, and STAG2. Based on this revised classification, MR-AML now accounts for 25% to 34% of newly diagnosed AML, with increasing frequency by age. In addition, the spectrum of myeloid disorders has become more confluent, again reflective of the increased emphasis on molecular features rather than the former use of blast counts in the bone marrow or peripheral blood. The presence of rearrangements such as core binding factor [t(8;21) or inv(16)] or the PML::RARα fusion, for example, are established disease-defining molecular lesions sufficient for subtype-specific diagnosis of AML (or Acute Promyelocytic Leukemia (APL)) in previous classifications. In this latest WHO revision, several additional AML-associated molecular features including the presence of NPM1 mutations now also permit specific AML diagnosis classification irrespective of bone marrow blast percentage. Importantly, this emphasis on molecular drivers provides new opportunities for treatment, based on the availability of frontline treatment options designed to inhibit specific molecular drivers. However, remaining questions with respect to the molecular diversity of AML and the hierarchies of co-mutational patterns relative to the phenotype-driving molecular lesions remain.

Risk classification: European LeukemiaNet (ELN) 2022 highlights

Pretreatment cytogenetic findings were the first molecular hallmarks used to diagnose and prognostically stratify patients with AML. Subsequently, select recurrent AML-associated gene mutations were demonstrated to provide additional prognostic information which, starting in 2010 as part of the first edition of the European LeukemiaNet (ELN) recommendations for diagnosis and management of AML, were included into the proposed genetic risk stratification (7). This approach resulted in a standardized system for reporting cytogenetic and recurrent gene variants, enabling meaningful comparisons among studies correlating genetic features with clinical outcomes, and providing an important tool to guide clinicians about prognosis and therapeutic response to cytotoxic chemotherapy. The 2022 ELN recommendations (8) include a revised genetic-risk classification that incorporates recent advances in our understanding about the prognostic significance of recurrent somatic alterations in AML. Following changes in the WHO classification, the 2022 ELN now includes 7 myelodysplasia-related mutations into the adverse-risk group prediction (in the absence of favorable-risk markers). It further provides suggestions on select molecular hierarchies, including the placement of NPM1-mutated patients with adverse-risk cytogenetic abnormalities into the adverse-risk group. Similarly in this group, consideration of only the presence, not the allelic ratio, of FLT3 Internal Tandem Duplication (ITD); and the substitution of biallelic CEBPA mutations with in-frame mutations affecting the basic leucine zipper (bZIP) region of the CEBPA gene (CEBPA^bZIP) constitute favorable-risk markers. In addition, changes were made to the cytogenetic risk assignment, such as the addition of t(3;v)(q26.2;v)/MECOM(EVI1)-rearranged AML to the adverse-risk group, and the removal of hyperdiploid karyotypes with ≥3 trisomies but without structural abnormalities from consideration as a complex karyotype AML. Further, the very rare t(8;16)(p11.2;p13.3)/KAT6A::CREBBP was added to this classification.

History of AML Discovery Genomics

Cytogenetics-based discovery

Leukemias were the first cancer genomes to be studied by advanced chromosomal condensation, staining, and microscopy-based methods. In particular, the BCR-ABL1 translocation defined by the Philadelphia chromosome was the first such discovery (9), followed by the translocation of chromosomes 15 and 17 that yields the PML::RARα fusion (10). In both settings, although the underlying gene drivers were not known at the time, these chromosomal translocations served as diagnostic hallmarks of the disease, ushering in an era where the long-held notion that changes in DNA could drive the onset of cancer was supported by microscopically discernable alterations that tracked with a specific diagnosis.

Microarray/ fluorescent in situ hybridization (FISH)-based discovery

As chromosome-based diagnoses became more commonplace, higher-resolution methods were devised for discovery and clinical diagnosis. Microarray-based methods that used chromosomally ordered large-insert clones from the Human Genome Project permitted the hybridization of differential fluorescent labeled tumor- and normal-derived DNA from individual patients to the clone-spotted microarray. Signal intensity-based analysis of the resulting fluorescent images identified novel insertion, deletion, and ploidy changes in the tumor genome which, when compared across multiple patient samples, identified recurrent events. This approach parsed into the clinical laboratory as a directed assay known as array comparative genome hybridization (array CGH). A second approach to focused identification of diagnostic chromosomal abnormalities is known as fluorescent in situ hybridization (FISH). Here, DNA clones containing the gene(s) at a locus or event of interest (e.g., translocation, amplification) were labeled with unique fluorescent moieties, hybridized to synchronized cells in metaphase, then evaluated by fluorescence-based microscopy to identify the suspected chromosomal juxtaposition with a known fusion partner, or the amplified signal intensity compared to a control locus.

Sequencing-based discovery

Completion of the Human Genome Project permitted PCR-based sequencing of genes of interest from bulk cancer DNA extracts, toward the identification of mutations potentially driving onset or progression of cancers, including AML. These efforts were scalable but required large amounts of DNA and only proved marginally successful at identifying new genes and/or pathogenic mutations (11). The advent of next-generation sequencing (NGS) technologies in the late 2000s was transformative, enabling unbiased discovery of genomic alterations in the comparison of matched tumor and normal DNA. The first cancer genome studied by these methods was from an AML patient (12), and the second such effort (3) identified a novel somatic IDH1 mutation in an AML patient that was shown to be recurrent across a larger AML cohort. As the cost of data generation with these methods decreased, large-scale discovery efforts were enabled (13–15) and new AML driver genes were identified. Such studies also were contributory to defining AML disease biology by identifying co-mutation and mutual exclusivity across genes, as well as identifying structural variants (translocations, inversions, etc.) that identified novel fusion drivers. By combining standard pathology assays (e.g., FISH, cytogenetics) with NGS testing and analytics, refinements to diagnosis and prognosis, subtyping, and the identification of therapeutic vulnerabilities were obtained. This integrated approach across multiple assay types was challenged recently by a new approach relying solely on NGS-based whole genome sequencing data analyzed by multiple algorithms to provide somatic profiles as well as surrogacy for cytogenetics and FISH assay results (16). Other efforts have demonstrated that the personalized somatic genotype of an AML patient can inform assays to evaluate minimal residual disease, often identifying emergent resistant disease during patient management (17).

As we describe below and outline in Table 1, these approaches and the resulting therapies have coupled with standard of care diagnosis and treatment to yield new outcomes for patients in many AML subtypes.

Connecting Genomics to Therapeutics

Myelodysplasia-related AML

MR-AML is a clinically and molecularly heterogeneous entity and patient outcomes differ based on the underlying molecular aberration. Among the disease group-defining gene variants, high variability exists with respect to observed frequencies, as well as known association with clinical characteristics and outcomes. For example, while pathogenic variants in RUNX1, SRSF2, and ASXL1 are the most common and together account for approximately 60% of MR-AML, ZRSR2 driver variants are found in <1% of MR-AML patients (15, 35). Further, while ASXL1 and RUNX1 pathogenic variants were previously established as driving adverse prognoses, the prognostic impact of other genes and variants is yet to be established. While multiple studies have demonstrated inferior survival of MR-AML compared to other AML diagnoses, it is unclear whether all MR-AML defining genetic features equally associate with poor treatment response and survival, nor is it clear whether all genetically defined MR-AML have the same outcomes in the setting of dysplastic bone marrow features. In addition to the molecular heterogeneity of disease-defining lesions in MR-AML, there is additional heterogeneity with respect to co-occurring mutations, which may further alter treatment response. Notably, the ELN 2022 genetic risk classification recognizes this potential impact, although it is unclear whether all favorable genetic features outweigh those with adverse-risk association, or whether accompanying genetic features may further alter treatment response (36). Importantly, while different frontline treatment options for MR-AML are approved, data-driven guidance for the choice of treatment currently does not exist. Based on a hallmark study that compared the liposomal formulation of cytarabine/daunorubicin (CPX-351, Vyxeos) to the classic “7 + 3” regimen in older patients, 60 to 75 years of age, with newly diagnosed high-risk AML or secondary AML (sAML) (20), CPX-351 is now regarded as an upfront standard of care option in patients with newly diagnosed sAML. An additional effective novel agent, the oral BCL2-inhibitor venetoclax, when combined with hypomethylating agents (HMA) showed high response rates in sAML patients with a complete remission/complete remission with incomplete hematologic recovery (CR/CRi) rate of 67% and a 16.4 months-long overall survival (OS), thus providing another frontline option for MR-AML patients (18, 19). However, despite the availability of several frontline regimens, MR-AML is thus far still a difficult to treat AML subtype. The recent changes in disease and risk classification of myeloid neoplasms that enable earlier AML-directed therapy based on myelodysplasia-related genetic changes even at lower blast count, provide an opportunity to evaluate whether early, intensive/combinatorial treatment approaches will improve MR-AML survival outcomes.

NPM1-mutated AML

NPM1 mutations are one of the longstanding hallmarks of favorable-risk AML, yet the impact of specific, co-existing altered genes on the prognostic significance of NPM1 mutations has been acknowledged (37). While the 2010 and 2017 ELN genetic risk classifications stratified the favorable prognostic impact of NPM1 mutations based on the co-occurrence of FLT3-ITD, the 2022 risk classification removed the “FLT3-ITD rule” and instead used other co-occurring genetic features such as adverse-risk cytogenetics, to downgrade the prognostic significance of NPM1 variants. By contrast, NPM1 variants overrule the co-presence of MR–associated gene variation. Notably, while genetic risk associations are always contextualized by the treatment type they relate to, patients harboring NPM1 mutations also are reported to have favorable outcomes with alternative regimens such as venetoclax combined with hypomethylating agents (HMA), while NPM1 variants confer favorable response rates in the setting of co-occurrence with IDH mutations when treated with targeted inhibitors. Furthermore, patients with NPM1 mutations (and/or those with KMT2A-rearrangements) are thought to respond well to Menin inhibitors, another type of novel targeted treatment option specific for this molecular subgroup (21). Conversely, patient-associated factors such as age and race/ethnicity appear to alter the prognostic impact of NPM1 mutations. For example, for patients >60 years of age treated with conventional chemotherapy, the favorable treatment response of mutations resulting in cytoplasmic NPM1 (NPM1c) is less pronounced. In AML patients of African-American ancestry, the prognostic association of this mutated gene is greatly diminished. Taken together, these observations suggest the need to clearly define additional clinical or molecular features that influence treatment response to approved agents, and additional consideration of age and ancestry to identify the frontline treatment with the highest likelihood of success for each individual patient.

FLT3-mutated AML

The presence of a FLT3-ITD has long been one “hallmark” adverse-risk prognosticator. These activating mutations confer a poor response to standard cytotoxic induction therapy and correspondingly higher relapse rates, resulting in shorter overall survival. However, the adverse prognostic impact of FLT3-ITD is dependent on co-existing molecular features, such as mutated NPM1c and the clonal prevalence of the ITD, as estimated from NGS data. More recently, FLT3-ITD targeted inhibitors that are used either in combination with cytotoxic chemotherapy for patients eligible for intensive induction or with hypomethylating agents for patients who are not have shown encouraging response rates and superior survival. This approach may overcome the negative prognostic association of FLT3 alterations (22–24, 26). Both the technical difficulties associated with accurate ITD detection by NGS (and thus accurate allelic ratio estimation) combined with the hope for risk improvement with the addition of targeted inhibitors have led to a re-classification of FLT3-ITD in the ELN 2022. Now, it is considered an intermediate-risk marker, irrespective of co-existing NPM1 alteration status. Exciting results demonstrating the effectiveness of targeted inhibitors of FLT3-ITD and intensified consolidation (allogeneic hematopoetic stem cell transplant [HCT]) to overcome the adverse prognosis of this variant were recently presented by the MORPHO clinical trial investigators. Here, FLT3-ITD + patients received intensive induction (plus midostaurin), followed by allogeneic HCT and randomization to either placebo or gilteritinib maintenance (38). The addition of gilteritinib improved relapse-free survival in the group of minimal residual disease positive (MRD +) patients. Notably, the 2-year OS of both placebo and gilteritinib-treated patients was approximately 80%, suggesting that the combination of upfront addition of FLT3-inhibitor and early allogeneic HCT is a promising treatment strategy for this traditionally poor-risk patient cohort, and that maintenance may further improve the outcomes of patients with MRD post HCT. However, it should be noted that primary and secondary resistance to FLT3-inhibitors is common with, for example, almost half of patients who achieved a CR in the RATIFY clinical trial study experiencing relapse. Clonal evolution via acquisition of mitogen-activated protein kinase (MAPK) signaling pathway mutations or selection of drug-resistant ITD clones was a common escape mechanism, consistent with those observed to multiple targeted agents (39).

Of note, while data with respect to prognostic significance, treatment response and survival of patients with missense mutations in the FLT3 tyrosine kinase domain (TKD) treated within the RATIFY clinical trial were assessed in a dedicated subanalysis (25), associations are less clear for patients with mutations in the juxtamembrane domain. Dedicated, prospective studies for these FLT3-altered patients will be required to delineate their associated risk and response to available therapeutic regimens, in context with co-existing somatic alterations in genes that might modulate treatment response.

Lastly, in addition to the well-known FLT3-ITD and TKD mutations, uniparental disomy (UPD) of the chromosome 13 locus that encompasses FLT3 is recurrent and can co-occur with a second FLT3 mutation or (rarely) is detected as the sole alteration (40). Data on the frequency and significance of this UPD are relatively sparse, likely due to the use of gene panels in clinical testing that limit the detection of copy number and loss of heterozygosity (LOH). Should targeted inhibitors continue to show efficacy, consideration of these less common aberrations may provide more patients with additional therapeutic options.

IDH1/2-mutated AML

The identification of mutations in IDH1 and IDH2 in approximately 20% of AML patients ushered in the notion of precision medicine in AML. Mechanistically, IDH1/2 pathogenic mutations are neomorphic to protein function leading to an oncometabolite, 2-hydroxyglutarate, which dysregulates DNA and histone hypermethylation and impairs hematopoietic cell differentiation (41–44). This mechanistic impact has been addressed by targeted therapy regimens that are applied in the frontline, especially for unfit patients who will not tolerate aggressive chemotherapy. Isocitrate dehydrogenase (IDH) inhibitor therapies now are among the most promising targeted agents in the treatment of AML harboring IDH1/2 mutations. With excellent response rates as therapy both in the frontline and the relapsed/refractory setting, IDH inhibitors (IDHi) (IDH1, ivosidenib, olutasenib; IDH2, enasidenib), have reasonable toxicity profiles and gained rapid (ivosidenib: breakthrough) FDA approval (30–34). New data support the combined use of IDHi with hypomethylating agents for improving CR rates and corresponding outcomes. While it seems intuitive to consider such a targeted agent as best treatment option for IDH1/2-mutated patients, response data of the oral B cell lymphoma (BCL-2) inhibitor venetoclax, when combined with HMA has overall response rates in AML with IDH1/2 mutations of >80% in the frontline setting (45). Furthermore, AML patients harboring both IDH2^R140 and NPM1 mutations have good long-term survival rates in the setting of conventional 7 + 3 therapy. Hence, while the availability of therapies to address different molecular features is suggestive of likely response, actual head-to-head comparisons of general responses in the setting of specific co-occurring molecular aberrations have not been conducted. Given this lack of direct comparison studies, no guidance exists as to which treatment choice will provide best response in IDH1/2-mutated patients, including the possible importance of the sequence in which these targeted inhibitors should be offered to provide the longest therapeutic benefit and outcome. Further consideration should be given to an accompanying gauge of performance status allowing for possible curative allogeneic HCT, in which case likelihood of complete remission would inform treatment choice. Important groundwork has been completed to identify and understand resistance mechanisms to IDH inhibitors, including clonal switches, cis/trans-occurring second-site mutations, and the acquisition of MAPK somatic activating mutations (33, 46–48). However, these only occur in a small percentage of patients, and other contributing genomic alterations remain poorly understood.

Complex cytogenetics and TP53-mutated AML

AML with a complex karyotype, defined as the detection of ≥3 chromosomal abnormalities absent any disease-defining cytogenetic abnormalities, remains the subgroup with the lowest complete remission rates and shortest overall survival (49). Typical complex karyotype AML genomes carry TP53 alterations in >90% of cases and are defined by 5q, 7q, and/or 17p abnormalities (while their absence denotes atypical complex karyotype AML) (50). These genomic hallmarks denote very adverse-risk disease for which the use of standard cytotoxic chemotherapy, even for fit patients, is being discouraged. Initial excitement of venetoclax-based regimens possibly showing improved response rates for TP53-mutated AML unfortunately did not hold up in larger studies with longer follow-up. In fact, monotherapy with HMA seems to be the frontline treatment with the lowest toxicity and has response rates comparable to more intensive regimens (51). Venetoclax remains a superior treatment approach for other poor-risk genetics absent TP53 alterations, highlighting the importance of refined molecular risk groups for treatment determination (52). Consequently, identification and testing of agents that show promise of efficacy in TP53-alterated myeloid diseases is of highest significance. Evidence of immune dysregulation and a high inflammatory state in TP53-associated myeloid neoplasms suggests that novel immune-based therapies might benefit this poor survival cohort (53, 54). One such agent, the monoclonal CD47 antibody magrolimab, appears promising when combined with azacitidine based on phase Ib data in patients with previously untreated higher-risk myelodysplastic syndromes (55). Here, 40% of TP53-mutated patients achieved a CR with median OS of 16.3 months. Given the demonstrated equally poor survival of TP53-mutated AML and myelodysplastic syndrome (MDS) with excess blasts, there is a consideration for regarding both as a single entity to ease clinical trial enrollment and patient counselling (56, 57). Another important aspect is that multi-hit events affecting TP53 confer the poorest survival, as inferred by more than one TP53 variant, del(17p) or high variant allele fraction for a variant, indicative of copy number neutral LOH (58).

Inflammation and immune response

Immunotherapy has revolutionized the treatment of cancers (59), but AML study results have been largely disappointing. Numerous factors may contribute to this ineffectiveness, including rapid progression, high tumor burden, low mutational burden, and immune system dysregulation (60, 61). Another reason is our rudimentary understanding of the crosstalk between inflammation and immune response in the bone marrow and other tissues harboring AML.

Recently, inflammation and immune response have emerged as important, previously largely unrecognized contributors to leukemogenesis. In AML, inflammation has been linked to progression from MDS to AML (62, 63). Several mutations in genes associated with myeloid malignancies have been shown to render hematopoietic stem cells more susceptible to inflammation. Here, presence of such mutations in precursor stages such as CHIP and CCUS can confer a competitive advantage by means of a selective pressure that impairs growth and fitness of normal hematopoietic stem cells (64, 65). Chronic inflammation, a shared characteristic of myeloid neoplasms, has been established as an equally powerful selective pressure. Specifically, the inflammatory profile is characterized by overproduction of inflammatory cytokines including tumor necrosis factor α, interleukin (IL)-6 and IL-1β providing a compelling rationale that targeting aberrant inflammation may be a promising strategy to modulate leukemogenesis, treatment response, or possibly malignant transformation. In AML, high inflammation was found to be a powerful prognosticator of poor treatment response and survival (66, 67). Interestingly, while high inflammation associates with certain gene mutations such as FLT3-ITD, TP53 or MR-AML mutations, it still provides independent prognostic information and in fact refines current genetic risk stratification, suggesting the need to integrate information about the inflammatory response into AML clinical risk assessment (67).

Remaining Challenges: Understudied Patient Populations

Ethnic diversity in the setting of AML

Molecular features that influence disease aggressiveness, therapy response, and outcomes are key features of AML routinely used to guide daily clinical decision making. However, large genomic discovery studies included only 1% to 6% patients of any African descent, reflective of other such studies. Previous reports of AML in self-reported African-American race/ethnicity (hereafter referred to as Black) demonstrated worse survival in these patients (68, 69). The reasons underlying this outcome disparity are multifactorial and interconnected, including the established impact of socioeconomic features on survival, inferior access to care, and lower percentage enrollment into clinical trials, all of which are compounded by the multifaceted contributions of structural racism (4, 69–72). Data obtained from patients enrolled onto AML clinical trials over the past 3 decades demonstrated inferior survival of Black patients even when considering confounding variables such as access to treatment and socioeconomic status. Only 25% of Black patients with AML were disease-free and 29% were alive 3 years after diagnosis, compared with 38% and 42% of White patients, respectively (69). As patients on the study received only cytotoxic chemotherapy for induction and consolidation, and no allogeneic HCT in first complete remission (per protocol), these data suggest an increased chemotherapy resistance to standard of care in Black patients. Notably, pilot studies suggest differences in the frequencies of known AML-associated recurrent somatic alterations, including lower frequencies of NPM1 mutations, and higher frequencies of IDH1/2- mutations, and of core-binding factor rearrangements in Black patients (69). Furthermore, established molecular survival prognosticators seem to carry different weights in patients from ancestry diverse populations (5, 69), including suboptimal performance of the 2022 ELN genetic risk stratification for Black and Hispanic patients (36). As our current knowledge about AML and subsequent care seems severely biased towards patients of European ancestry, we have an incomplete depiction of the molecular landscape, including possible targetable lesions and/or pathways with ancestry association that contextualize the disease biology. Hence the need for systematic, genomics-based studies of diverse AML patient populations to overcome this unacceptable disparity in knowledge, risk stratification-based care, and disease outcomes.

Concluding Remarks

This review makes clear that large-scale genomic discovery in AML patient genomes in the era of NGS and analysis has informed our understanding of the unique somatic attributes of this disease type. Subsequent to these discoveries, novel therapeutic approaches to driver alterations have emerged and are increasingly being tested in the frontline in different subtypes of AML according to molecular characteristics, leading to refined treatment paradigms and risk stratification. While these discoveries continue to transform care and outcomes in our patients, it is clear from preliminary genomics studies that AML as a broad class of hematologic malignancy manifests itself based on very different genomic contexts in the setting of diverse ancestries. In order to more precisely and effectively address the AML patient population, specific studies of the genomics of diverse patients in the setting of the standard of care may revise risk stratification and advise the use of molecularly targeted therapies and combinations, as well as immunotherapeutic approaches.

Nonstandard Abbreviations

AML, acute myeloid leukemia; CR, complete remission; MR, myelodysplasia-related; ELN, European LeukemiaNet; FISH, fluorescent in situ hybridization; NGS, next-generation sequencing; HCT, hematopoetic stem cell transplant.

Human Genes

SRSF2, serine and arginine rich splicing factor 2; SF3B1, splicing factor 3b subunit 1; U2AF1, U2 small nuclear RNA auxiliary factor 1; ZRSR2, zinc finger CCCH-type, RNA binding motif and serine/arginine rich 2; ASXL1, ASXL transcriptional regulator 1; EZH2, enhancer of zeste 2 polycomb repressive complex 2 subunit; BCOR, BCL6 corepressor; STAG2, STAG2 cohesin complex component; PML, PML nuclear body scaffold; RARa, retinoic acid receptor alpha; NPM1, nucleophosmin 1; FLT3, fms related receptor tyrosine kinase 3; CEBPA, CCAAT enhancer binding protein alpha; MECOM (EVI1), MDS1 and EVI1 complex locus; KAT6A, lysine acetyltransferase 6A; CREBBP, CREB binding protein; IDH1, isocitrate dehydrogenase (NADP(+)) 1; RUNX1, RUNX family transcription factor 1; KMT2A, lysine methyltransferase 2A; TP53, tumor protein p53; IDH2, isocitrate dehydrogenase (NADP(+)) 2.

Author Contributions

The corresponding author takes full responsibility that all authors on this publication have met the following required criteria of eligibility for authorship: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; (c) final approval of the published article; and (d) agreement to be accountable for all aspects of the article thus ensuring that questions related to the accuracy or integrity of any part of the article are appropriately investigated and resolved. Nobody who qualifies for authorship has been omitted from the list.

Ann-Kathrin Eisfeld (Conceptualization-Equal, Writing—original draft-Equal, Writing—review & editing-Equal), and Elaine Mardis (Conceptualization-Equal, Writing—original draft-Equal, Writing—review & editing-Equal).

Authors’ Disclosures or Potential Conflicts of Interest

Upon manuscript submission, all authors completed the author disclosure form.

Research Funding

None declared.

Disclosures

E.R. Mardis is a member of the Supervisory Board of Qiagen N.V. and member of the Board of Directors at Singular Genomics Systems, Inc. for which she receives stock grants and honoraria, and receives honoraria as a member of the Scientific Advisory Board of Scorpion Therapeutics, LLC. A.K. Eisfeld has received grants or contracts (payment to institution) from the National Cancer Institute (NCI), Leukemia & Lymphoma Society, and American Cancer Society; discloses honoraria paid for serving on an Astra Zeneca Diversity, Equity and Inclusion Advisory Board, and for lectures at Baylor College of Medicine, the NCI, and OncLive; has received meeting support from the American Society of Clinical Oncology, American Society of Hematology (ASH), and European Hematology Association; has a patent pending on the use of inflammation Score (iScore) for acute myeloid leukemia risk refinement; has a leadership role on the ASH Committee for Diversity, Equity and Inclusion and ASH Committee for Precision Medicine; and spouse is employed by Karyopharm Therapeutics.

References