Acute Erythroleukemia is a rare form of acute myeloid leukemia recognized by its distinct phenotypic attribute of erythroblast proliferation. After a century of its descriptive history, many diagnostic, prognostic, and therapeutic implications relating to this unique leukemia subset remain uncertain. The rarity of the disease and the simultaneous involvement of its associated myeloid compartment have complicated in vitro studies of human erythroleukemia cell lines. Although murine and cell line erythroleukemia models have provided valuable insights into pathophysiology, translation of these concepts into treatment are not forthcoming. The integration of knowledge gained through a careful study of these models with more recent data emerging from molecular characterization will help elucidate key mechanistic pathways and provide a much-needed framework that accounts for erythroid lineage-specific attributes. In this article, we discuss the evolving diagnostic concept of erythroleukemia, translational aspects of its pathophysiology, and promising therapeutic targets through an appraisal of the current literature.
Types of Acute Erythroleukemia
Acute erythroid leukemias can be classified as follows:
- M6a (Erythroleukemia) – 50% or more of all nucleated bone marrow cells are erythroblasts, dyserythropoiesis is prominent and 20% or more of the remaining cells (non- erythroid) are myeloblasts.[rx][rx]
- M6b (Pure erythroid leukemia) – In rare cases the erythroid lineage is the only obvious component of an acute leukemia; a myeloblast component is not apparent. The erythroid component consists predominantly or exclusively of proerythroblasts and early basophilic erythroblasts. These cells may constitute 90% or more of the marrow elements. Despite this lack of myeloblasts, these cases should be considered acute leukemias. In a WHO proposal the blastic leukemias that are limited to the erythroid series are designated pure erythroid malignancies.[rx]
- M6c (Erythroleukemia and Pure erythroid leukemia) – Myeloblast- and proerythroblast-rich mixed variant.[rx]
Models of erythroleukemia
It is conceivable that the molecular oncogenesis of PEL, characterized by the features of both early differentiation block and proliferation in its blast populations, would share the paradigm of the proposed double-hit model involved in AML evolution [rx]. While the transformative processes in human erythroid progenitors are incompletely understood, described murine and avian models have proven valuable in this regard. The conceptual model of multistage carcinogenesis is exemplified by the Friend disease, a model system of erythroleukemia first described in 1957 [rx]. Friend erythroleukemia can be induced in susceptible strains of mice by infection with the ‘Friend’ retrovirus complex, constituted by the replication-defective spleen focus-forming virus (SFFV) and a replication-competent Friend murine leukemia virus [rx, rx]. The disease evolves through two stages; the first stage being preleukemic and involving the polyclonal expansion of erythroblasts. Glycoprotein 55 (Gp55), encoded by the envelope (Env) gene of SFFV, activates the erythropoietin receptor (EpoR) to cause erythropoietin-independent erythroblastosis, the effects of which are modulated by the expression of SF-Stk (a truncated form of STK receptor tyrosine kinase) [rx, rx]. The second stage is transformative and is initiated by pro-viral integration upstream to the promoter of the Sfpi-1 gene that encodes PU.1, a transcription factor of the ETS (E26 transformation-specific) oncogene family. These tumorogenic erythroblasts are clonal and have shown to be consistently altered with respect to two genes, Sfpi-1 and the TP53 gene [rx, rx]. Deregulated expression of PU.1 is believed to induce blockade in the erythroid differentiation program through upregulation of various potential targets including Fli1, another pro-oncogenic transcription factor, to complete the causal chain of leukemogenesis [rx–rx]. Later developed murine models, such as the Spi-1 transgenic model system, share more similarities with the human erythroleukemic disease than the Friend model. At the onset of disease in the Spi-1 transgenic model, hematopoietic tissues were massively invaded with non-tumorigenic proerythroblasts that expressed a high level of Spi-1 protein causing profound anemia. The transgenic proerythroblasts are dependent on erythropoietin for their proliferation during this stage. In the second stage, proerythroblasts became tumorigenic and achieved growth factor autonomy through the acquisition of additional genetic events including TP53 alterations [rx]. Unlike in the Friend disease, polycythemia does not precede the occurrence of erythroleukemia despite autocrine stimulation by Epo. This clinical feature is shared by human erythroleukemic disease in which EPO autocrine stimulation is similarly observed and may represent a secondary event related to the acquisition of malignant state [rx] [rx, rx. With the currently available evidence, it is not yet known if the two-hit model of mouse leukemogenesis may be an oversimplification of human erythroleukemic disease. Unfortunately, the rarity of this leukemia variant and associated involvement of the myeloid compartment complicates in vitro studies of primary erythroleukemia.
Erythroleukemic cell lines
Other in vitro models for study include an extensive repertoire of human erythroleukemic cell lines, the oldest and most extensively studied of which is the K-562 cell line. Cell line characterization studies, while confirming the shared similarities of karyotypic complexity marking the disease, have identified them to be heterogeneous in their cytogenetic, molecular and cytokine-related profiles [rx]. A list of leukemic cell lines with erythroid features is outlined. These available erythroleukemia cell lines exhibit a wide range of functional characteristics, such as variable cytokine requirements and differing differentiation potential when grown in culture with or without differentiation agents. Still, significant information has been acquired from studying these cell lines, especially related to transcriptional and epigenetic programs in AEL. Some of these cell lines (K562, HEL, OCIMI, OCIM2, LAMA-84) express markers of multiple cell lineages, a characteristic found only infrequently in other primary leukemias [rx]. Few of the erythroleukemia lines are bipotential precursors, with an ability to differentiate, in vitro, along megakaryocytic or erythroid lineage pathways in the presence of appropriate inducers [rx].
The role of transcription factors, molecular mutations and epigenetic alterations
GATA1 and PU.1
GATA1 and PU.1 function as two opposing lineage-specific transcription factors that regulate erythroid and myeloid programs [rx]. GATA1 plays a critical role in erythroid survival and terminal erythroid differentiation [rx, rx]. The pivotal functional significance of these interactions in determining lineage fate and lineage-specific differentiation may underlie the leukemogenic mechanisms of maturation arrest consequent to dysregulated expression of these transcription factors [rx].
In vitro studies have shown that although the upregulation of PU.1 contributed to erythroleukemia by causing differentiation arrest, it induced growth inhibition and apoptosis in the presence of differentiating agents such as dimethyl sulfoxide (DMSO) [rx]. The growth-inhibitory and apoptotic effects of PU.1 on erythroleukemic cells are associated with the downregulation of pro-survival oncogenes such as c-MYC and BCL2, and reduced DNA binding activity of the GATA1 transcription factor [rx, rx]. PU.1 interacts directly with GATA1 and represses its erythroid differentiating action; introducing exogenous GATA1 was able to induce differentiation in PU.1 blocked murine erythroleukemia (MEL) cells [rx]. GATA1 represses PU.1 expression by binding to the PU.1 promotor and its other regulatory upstream response elements. Available evidence suggests that the antagonistic influence of these transcription factors’ function in the erythroid maturation-differentiation program is finely modulated by their relative expression. While the complete loss is lethal, graded reductions in PU.1 gene expression associated with an increasingly aggressive AML phenotype [rx]. Unlike in normally differentiating erythroid precursors where GATA1 completely shuts down PU.1 to repress the myeloid program, expression of PU.1 in human AEL is maintained, albeit at reduced levels, from its incomplete repression by GATA1 [rx]. It appears that complete transcriptional repression of PU.1 leads to the loss of PU.1-dependent repression of GATA1 targets, facilitating erythroid differentiation. On the other hand, blockade of GATA1 mediated repression on PU.1 would increase PU.1 expression leading to differentiation and growth arrest [rx]. One may speculate that an attenuated expression of PU.1 may paradoxically enhance leukemogenesis through changes in the nuclear environment brought about by interactions with GATA1.
GATA-1 mediated repression mechanisms exhibit distinct inter-species differences in human and murine AEL cell lines. Unlike in murine AML-EL, PU.1 repression by GATA1 additionally involves DNA binding along with H3K9 and H3K27 trimethylation at regulatory upstream enhancer and promoter regions. Along these lines, the upper response enhancer elements in human AEL cells contain a DNA methylation mark. Reversal of the DNA methylation mark via inhibition of DNMT3A, a co-occupant of the GATA1 repression complex, by hypomethylating agents has been demonstrated to inhibit leukemic growth and induce differentiation [rx]. This deregulated mechanism of upper response element DNA methylation is also shared by MDS, and may even predict for clinical response to hypomethylation therapy [rx].
Rose et al reported on molecular mutation data in a cohort of 166 AEL (M6) patients and showed that TP53 was the only gene occurring at a higher frequency within M6 as compared with the remaining overall AML cohort (36% vs 11%, respectively). Relatively lower mutational frequencies were observed for other genes including ASXL1, DNMT3A, FLT3-ITD, IDH2, NPM1, NRAS, RUNX1, and TET2 [rx]. Recent data reported by Montalban-Bravo and Benton et al. revealing an especially high prevalence of at least two TP53 abnormalities (including both mutations and aberrant or deleted chromosome 17p) in >90% of PEL patients [rx]. The high frequency of TP53 alterations suggest a crucial role of TP53 in the leukemic transformation to AEL. It remains to be determined whether TP53 alterations are a proximate effect of excess cytoplasmic iron sequestered within the dysfunctional pronormoblasts. In this context, there appears to be a link between iron/heme homeostasis and p53 signaling, with p53 downregulated during iron excess, via mechanisms operating at various levels influencing nuclear export of the p53, p53 stabilization and p53-DNA interactions [rx]. Iron induced DNA damage, via free radical oxygen species formation, may well contribute to the high rate of TP53 mutations and complex karyotype observed in this malignancy.
GATA1 and p53 interactions: a role in erythroleukemia?
GATA1 directly influences p53 by interacting with the p53 transactivation domain and inhibiting its transactivation in erythroid precursor cells [rx]. This interaction is an an erythroid cell-specific event with inhibition of p53 by GATA1, not observed in non-erythroid cells. GATA1 is crucial in mediating erythroid differentiation with GATA1 knockdown shown to induce erythroid leukemia in mice [rx]. In this context, hyperproliferative GATA1 null erythroid cells which escape cell death may accumulate secondary mutations leading to transformation [rx]. In the absence of GATA1 mediated p53 inhibition in GATA1 deficient cells, functional p53 pathway activation may be crucial in inducing cell cycle arrest with TP53 mutations may allow for the abnormal expansion of leukemic cells predisposing to leukemic transformation [rx, rx].
c-MYC and Bromodomain inhibition
The bromodomain (BRD) family are an epigenetic class of histone modification proteins with an ability to ‘read’ the genome and modulate gene expression through transcriptional regulator recruitment to specific genome locations [rx]. The protein family comprises four homologous proteins: BRD2, BRD3, BRD4, and BRDT, with widely varying roles on cell cycle growth and regulation. It has been demonstrated that BRDs of these reader proteins promote aberrant gene expression and sustain leukemic maintenance, at least in part to sustained MYC expression, thus paving a rationale for developing inhibitors against this class [rx]. In vitro experiments with BRD inhibitors, such as JQ1, have demonstrated these agents to carry anti-leukemic activity [rx]. Consistent with this, JQ1 treatment of UT7, a human erythroleukemia cell line, was able to rescue erythropoietin differentiation within a matter of two days [rx]. This also highlights the importance of a cellular erythroid cycle break mediated by c-MYC inhibition before initiation of the erythropoiesis program. The therapeutic potential of BRD inhibition merits further exploration within this subtype of leukemia.
GFI-1B and LSD-1 interactions
Another important transcription factor that has been implicated in erythroid leukemia is the growth factor-independent 1B protein (GFI-1B). GFI-1B plays a crucial role in erythroid progenitor cell growth and differentiation induction [rx]. Of interest, its overexpression is restricted to the AML-M6 and AML-M7 subtypes and is associated with the increased proliferative capacity of progenitor cell lines [rx]. Silencing its expression through siRNA was shown to decrease the proliferative capacity of HEL cell lines. GFI-1B interacts with histone demethylase LSD1 thereby repressing GBI-1B target genes and consequent differentiation of lineage-specific cells. Treatment with an LSD-1 inhibitor, T-3775440, was able to disrupt the GFI-1B LSD1 interaction, leading to transdifferentiation and cell growth arrest suggesting a novel mechanism of action specifically against AEL.
KIT receptor-ligand system
A focused network of lineage-specific transcription factors plays a decisive role in determining lineage fate at the crossroads of erythroid and megakaryocytic differentiation. The erythroid and megakaryocytic lines share early lineage similarities in regulatory transcription factors and cell surface marker expression; evidence suggests that the lineages diversify from a common erythroid–megakaryocytic progenitor [rx]. Flow cytometric analyses have revealed HEL cell lines to exist in two distinct sub-clones: CKIT-positive, CD41b-negative (erythroid lineage markers), and CKIT-negative, CD41b-positive (megakaryocytic lineage markers). KIT receptor expression relates to the expression of lineage-specific antigens and also determines phenotypic fate towards erythroid differentiation [rx]. MiR-221 and miR-222, miRNAs downregulated during erythroid differentiation, down-modulate CKIT protein production through translational repression. MiR-221 and -222 gene transfer impairs proliferation and accelerates differentiation of the CKIT-positive TF-1 erythroleukemic cell lines [rx].
RUNX1 and KLF1
Another key transcription factor in AEL is RUNX1 which, along with multiple micro-RNAs, negatively affects erythroid differentiation by repressing an erythroid master regulator Knueppel-like factor 1 (KLF1). The repression of KLF1 turns off the erythroid gene expression program and facilitates megakaryocytic lineage specification [rx]. Recent studies have reported RUNX1 mutations in erythroleukemia, albeit at significantly lower frequencies compared with overall AML [rx]. There have not yet been studies directly implicating KLF1 in human erythroleukemia.
Other transcription factors
Other relevant transcription factors involved in the erythroid differentiation program such as GATA2, SCL/TAL, NF-E2, nuclear factor-B, forkhead transcription factors (FOXO), EKLF have also been studied in various in vitro models [rx]. Although these factors seem to have a pathogenic role in murine tumor models and other malignancies, mutations in genes encoding these transcription factors have not been directly implicated in human erythroleukemia.
Signaling proteins and pathways
The proliferative and differentiating effects of erythropoietin (EPO) commences with ligand binding to its cognate EPO receptor. Subsequent subunit dimerization and JAK2 recruitment results in phosphorylation of several tyrosine residues on the receptor. These phosphorylated residues serve as docking sites for signal transducer and activator of transcription (STAT) transcription factors, most prominently STAT5, and phosphorylate them. Phosphorylated STAT transcription factors dimerize and enter the nucleus to activate the transcription of specific genes [rx].
In vitro studies on freshly isolated human primary erythroleukemic cells demonstrated the constitutive activation of STAT1 and STAT3 and their role in promoting cell growth through c-MYC activation [rx]. STAT proteins also influence various aspects of erythroid differentiation. In the Friend disease model, activated STAT3 upregulates PU.1 thus promoting the progression of erythroleukemia by inhibiting erythroid differentiation [rx]. Conversely, activation of transcription factors such as STAT5 correlates with erythropoietin mediated erythroid differentiation and its conditional inactivation in erythroleukemic cell lines has been demonstrated to prevent terminal differentiation [rx]. Furthermore, erythroleukemic cell lines exhibit activation of multiple signaling pathways apart from JAK-STAT, including mTOR, PI3K/Akt pathways, thus serving multiple potential targets for targeted inhibitors [rx]. In the Friend disease model, activating mutations in receptor tyrosine kinases including CKIT results in clonal expansion through activation of multiple signaling pathways such as ERK & MAP kinases, PI3Kinase, and Src kinases [rx].
MicroRNAs are non-coding RNAs that play a crucial role in cell growth and differentiation through the regulation of gene expression [rx]. A miRNA profiling study in MEL cell lines found more than a hundred miRNAs to be dynamically expressed, with wide variations in expression levels, during the process of terminal differentiation induction after DMSO treatment [rx]. MiR-451 and miR-144 are upregulated, whereas miR-221, miR-222, miR-24, and miR-223 are downregulated during erythroid differentiation. Among the many identified miRNAs, miR-451, an erythroid differentiation promoting miRNA, in particular, was found to increase significantly with erythroid differentiation. The investigators were further able to demonstrate, through transfection of synthetic anti-sense miR-451 oligonucleotides into MEL cells thereby inducing miR-451 knockdown, that miR-451 positively regulates erythroid differentiation. Bruchova-Votavova et al demonstrated that enforced expression of miR-451 induced erythroid differentiation in K562 cells, an erythroleukemic cell line [rx]. Comprehensive miRNA profiling data in human PEL while lacking, perhaps due to the rarity of the neoplasm, could be performed to help provide deeper insights into disease pathogenesis and potentially inform miRNA targeted cancer therapies.
Participating member proteins of the JAK-STAT pathway are tightly regulated by a network of regulatory transcription factors and microRNAs. Su et al demonstrated that the miR-23a, -27a, and -24 miRNA cluster in particular, was dramatically downregulated in AEL patients and that restoration of their expression was able to induce apoptosis through inhibition of JAK-STAT3 pathway cascade [rx]. The investigators were also able to show that the inhibition of the pathway by the miRNA cluster simultaneously involved an upregulated expression of GATA1, which through PU.1 inhibition, was able to induce erythroid differentiation. This attests to the important role of the dysregulated GATA1-JAK-STAT pathway in AEL pathogenesis.
Causes of Acute Erythroleukemia
The causes of AEL are unknown.[rx] Prior to a 2008 reclassification by the World Health Organization, cases that evolved from myelodysplastic syndromes, myeloproliferative neoplasms, chemotherapy for other cancers or exposure to toxins were defined as secondary AEL.[rx] These cases are now likely to instead be classified as acute myeloid leukemia with myelodysplasia-related changes or therapy-related AML.[rx]
Symptoms of Acute Erythroleukemia
The most common symptoms of AEL are related to pancytopenia (a shortage of all types of blood cells), including fatigue, infections, and mucocutaneous bleeding.[rx] Almost half of the people with AEL exhibit weight loss, fever, and night sweat at the time of diagnosis.[rx] Almost all people with AEL are anemic, and 77% have a hemoglobin level under 10.0 g/dl.[rx] Signs of thrombocytopenia are found in about half of people with AEL.[rx]
Acute myeloid leukemia often begins with flu-like symptoms. You might have:
- anemia due to a lack of red cells, causing persistent tiredness, dizziness, paleness, or shortness of breath when physically active
- frequent or repeated infections and slow healing due to a lack of normal white cells, especially neutrophils
- increased or unexplained bleeding or bruising, due to a very low platelet count
- bone pain, swollen lymph nodes (glands), swollen gums, chest pain, and abdominal discomfort due to a swollen spleen or liver.
- Weight loss or loss of appetite
- Unusual bleeding or bruising
- Tiny red spots on your skin (petechiae)
- Swollen gums
- Swollen liver or spleen
- More infections than usual
- Achy muscles
If you have fewer platelets than usual, your blood may not clot as well as it should. You might have symptoms like these
- Easy bruising
- Bleeding that can be hard to stop
- Bleeding gums
- Small red spots under your skin caused by bleeding
- Sores that don’t heal
Diagnosis of Acute Erythroleukemia
Your doctor will ask about your medical history for acute granulocytic leukemia. They’ll do a physical exam to look for signs of bleeding, bruising, or infection. You might have tests including:
CBC tests – A complete blood count (CBC) shows how many of each type of blood cell you have. A peripheral blood smear checks for blast cells. Specific types include tests for
- RBC – the numbers, size, and types of RBC in the blood
- WBC – the numbers and types of WBC in the blood
- Platelets – the numbers and size of the platelets
- Hemoglobin – an iron-rich protein in red blood cells that carries oxygen
- Hematocrit – how much space red blood cells take up in your blood
- Reticulocyte count – how many young red blood cells are in your blood
- Mean corpuscular volume (MCV) – the average size of your red blood cells
The complete blood count (CBC) includes most or all of these. CBC is one of the most common blood tests.
- Blood tests – Most people with acute myelogenous leukemia have too many white blood cells, not enough red blood cells, and not enough platelets. The presence of blast cells — immature cells normally found in bone marrow but not circulating in the blood — is another indicator of acute myelogenous leukemia.
- Imaging tests – X-rays, CT scans, MRIs, and ultrasounds give a clearer picture of what’s going on inside you. They can help find infections or show when cancer has spread to other parts of your body.
- Bone marrow tests – Your doctor uses a needle to take a sample of marrow, blood, and bone from your hip or breastbone. A specialist looks at it under a microscope for signs of leukemia.
- Spinal tap – This is also called a lumbar puncture. Your doctor uses a needle to take some cerebrospinal fluid from around your spinal cord. A specialist checks it for leukemia cells.
- Genetic tests – A laboratory can look at your leukemia cells for gene or chromosome changes. The results will tell your doctor more about your AML so they can help you decide on the best treatment.
- Bone marrow biopsy – sections are usually hypercellular with clusters or sheets of immature cells and a marked reduction in the normal hematopoietic components. The erythroid series are dysplastic and left shifted, but are usually found in all stages of maturation. Some of the myeloblasts may show cytoplasmic granules. Occasionally, Auer rods may be present. Bone marrow iron stores are often increased, and ring sideroblasts may be present.
Blood smears show anisopoikilocytosis with the presence of schistocytes, tear-drops, and macrocytes. Basophilic stippling is present. Granulocytic series may show hypogranulation and hyposegmentation. Giant and/or hypogranular platelets are often present. Various numbers of nucleated red blood cells and blasts are often present.
Treatment of Acute Erythroleukemia
Santos et al reported on clinical outcomes of AML M6 in 91 patients treated at a single institution [rx]. The study investigators found no statistically significant difference in survival between M6 and other AML subtypes (p = 0.60). This was confirmed on a multivariate analysis for overall survival where AML-M6 was not an independent risk factor. The median overall survival of the overall M6 cohort was approximately 9 months and significantly lower in the M6b cohort (Pure Erythroid) compared with M6a (15 weeks vs 39 weeks, p = 0.007). Interestingly, the subtype of AML-M6 (6a and 6b) was not an independent prognostic factor for disease-free and overall survival. The authors concluded that AML-M6 by itself did not carry additional prognostic import.
Recent emphasis has been the evaluation of the efficacy of hypomethylating agents in the treating TP53 mutated leukemia due their ability to function through p53 independent mechanisms to affect responses. A recent clinical study supporting these mechanisms of action reported high response rates with a 10-day regimen of decitabine in TP53 mutated AML and myelodysplastic syndrome (MDS) [rx]. In a study of 36 AEL patients (81% classifiable as MDS, per 2016 WHO), decitabine-10 day regimens showed comparable overall survival and a non-significant trend towards improved event-free survival when compared with cytarabine-based regimens [rx].
PEL is associated with complex and high-risk karyotypes including chromosomes 5q and 7q abnormalities [rx]. In a study evaluating patients with erythroid-predominant myeloid neoplasms, morphologic features of PEL, adverse risk cytogenetics, and other features such as hypoalbuminemia and high serum lactate dehydrogenase emerged as independent prognostic factors of death [rx]. Whether prognosis in erythroleukemia links solely to its association with unfavorable karyotype or relates also to additional disease-specific characteristics, is not well-understood. Median survival among PEL patients is 1–3 months, with no survival differences observed in patients treated with intensive chemotherapy versus hypomethylating agents (HMA) [rx, rx, rx]. A more recent multinational study evaluating clinical outcomes of 217 patients with acute erythroleukemia demonstrated that intensive chemotherapy was superior to hypomethylating agents in affecting overall response but not associated with superior progression-free or overall survival. Importantly, however, patients with high-risk cytogenetics treated with HMA lived longer compared with intensive chemotherapy. The excellent therapeutic sensitivity to hypomethylating agents appears selective to the decitabine-10 day regimen and given the dismal outcomes with PEL and its association with high-risk cytogenetics, prospective evaluation of 10-day decitabine is needed. It must be noted that while this may represent a good treatment option for patients not ultimately going to transplant, the standard of care for a transplant-eligible patient would be induction chemotherapy based on the current body of evidence Allogeneic hematopoietic cell transplantation improves outcomes of AEL and should be considered in all AEL patients with high-risk cytogenetic features eligible for transplantation [rx]. Nevertheless, studies are yet to reassess the role of transplant in PEL as currently defined in 2016 WHO classification.
Novel therapies based on a more detailed understanding of dysregulated TP53-related molecular pathways may predictably improve current outcomes in PEL patients.