Advances in molecular genetics and treatment of core-binding factor acute myeloid leukemia

Abstract

Purpose of review Core-binding factor (CBF) acute myeloid leukemia (AML) is among the most common cytogenetic subtypes of AML, being detected in approximately 13% of adults with primary disease. Although CBF-AML is associated with a relatively favorable prognosis, only one-half of the patients are cured. Herein we review recent discoveries of genetic and epigenetic alterations in CBF-AML that may represent novel prognostic markers and therapeutic targets and lead to improvement of the still disappointing clinical outcome of these patients.

Recent findings Several acquired gene mutations and gene-expression and microRNA-expression changes that occur in addition to t(8;21)(q22;q22) and inv(16)(p13q22)/t(16;16)(p13;q22), the cytogenetic hallmarks of CBF-AML, have been recently reported. Alterations that may represent cooperative events in CBF-AML leukemogenesis include mutations in the KIT, FLT3, JAK2 and RAS genes, haploinsufficiency of the putative tumor suppressor genes TLE1 and TLE4 in t(8;21)-positive patients with del(9q), MN1 overexpression in inv(16) patients, and epigenetic and posttranscriptional silencing of CEBPA. Genome-wide gene-expression and microRNA-expression profiling identifying subgroups of CBF-AML patients with distinct molecular signatures, different clinical outcomes, or both, have also been reported.

Summary Progress has been made in delineating the genetic basis of CBF-AML that will likely result in improved prognostication and development of novel, risk-adapted therapeutic approaches.

Introduction

Core-binding factor (CBF) acute myeloid leukemia (AML) is characterized by the presence of t(8;21)(q22;q22) or inv(16)(p13q22)/t(16;16)(p13;q22). These chromosome aberrations create, respectively, RUNX1-RUNX1T1 and CBFB-MYH11 fusion genes, that disrupt subunits α and β of CBF, a heterodimeric transcription factor involved in hematopoietic differentiation [1]. RUNX1-RUNX1T1 can be generated not only by t(8;21) but also by its infrequent variants, including complex translocations involving one or two additional chromosomes [2–5], and insertions [e.g., ins(21;8)(q22q22;q22) and ins(8;21)(q22;q22q22)] [2,5,6]. Likewise, variants of inv(16)/t(16;16) and cryptic rearrangements leading to CBFB-MYH11 have been reported [7^•,8,9].

CBF-AML is fairly common, with t(8;21) detected in 6% and inv(16)/t(16;16) in 7% of adult de novo AML patients [10]. CBF-AML patients have a relatively favorable prognosis, especially if intensive postremission treatment with multiple cycles of high-dose cytarabine is administered [11–13]. However, only 50–60% of CBF-AML patients are cured using contemporary treatment [14–16]. Thus, genetic, clinical, or both, markers are needed that identify CBF-AML patients unlikely to respond to current treatment who could potentially benefit from more intensive, novel, or both, therapies.

Genetic alterations cooperating with core binding factor-acute myeloid leukemia-associated chimeric proteins

Expression of the RUNX1 and CBFB genes appears essential for the development of normal hematopoiesis. In murine knockout models, the homozygous loss of RUNX1 or CBFB caused lack of definitive hematopoiesis and embryonic death [17]. Both mouse and in–vitro models have shown that the RUNX1-RUNX1T1 and CBFB-MYH11 fusion proteins are critical for CBF-AML development. However, neither fusion protein alone is sufficient for leukemogenesis [18^•,19^•]. Thus, for overt leukemia to occur additional genetic, epigenetic alterations, or both are necessary. This is consistent with a multistep model of leukemogenesis, in which two classes of mutations cooperate: one affecting genes encoding transcription factors capable of impairing hematopoietic differentiation, such as RUNX1-RUNX1T1 and CBFB-MYH11 (class II mutations), and another involving genes such as KIT, FLT3, NRAS or KRAS, whose mutations lead to increased cell proliferation, survival (class I mutations), or both [20].

Secondary chromosome aberrations

Secondary chromosome aberrations are detected at diagnosis in approximately 40% of patients with inv(16)/t(16;16), and 70% of t(8;21) patients [14–16]. The patterns of secondary aberrations are nonrandom, and differ between inv(16)/t(16;16) and t(8;21) patients (Table 1 Table 1). One recent study suggested that secondary aberrations other than +8, +21 and +22 are more common in inv(16)-positive patients with rare, nontype A CBFB-MYH11 transcripts [21^•]. Secondary aberrations, such as +22 in inv(16)/t(16;16) patients, and −Y, −X and del(9q) in t(8;21) patients are less frequent in non-CBF-AML signifying their importance for CBF-AML leukemogenesis [22]. However, molecular alterations associated with +22, −Y, and −X have not yet been elucidated.

Table 1.

Secondary chromosome aberrations in core binding factor-acute myeloid leukemia: frequency and prognostic significance

Secondary chromosome aberration	inv(16)	t(8;21)
	% (No. of patients with aberration/total no. of patients studied)a, b	% (No. of patients with aberration/total no. of patients studied) a, b	Prognostic significance [reference no.]
−Y	2.0% (4/205)c	55.3% (157/284)c	OS and duration of first CR significantly shorter for male t(8;21) patients with −Y [14]; No prognostic impact [15].
−X	1.3% (2/159)d	37.0% (81/219)d	No impact on OS of female patients with t(8;21) [14]
del(9q)	0 (0/364)	16.1% (81/503)	Among nonwhite t(8;21) patients, OS of those with del(9q) was better than OS of patients with sole t(8;21) and those with secondary abnormalities other than del(9q) [15]. No impact on OS in white t(8;21) patients [15]. No impact on CR rates, RFS and OS [16].
+22	18.0% (99/551)	0.3% (1/318)	CIR significantly lower for inv(16) patients with +22 compared with CIR of patients with isolated inv(16) [15]. RFS longer for inv(16) patients with +22 than for inv(16) patients without +22 [14].
+8	10.7% (59/551)	6.2% (31/503)	No impact on RFS and OS [14].
+21	4.9% (27/551)	0.6% (2/318)	Not established.
None	60.7% (335/551)	30.2% (152/503)

CIR, cumulative incidence of relapse; CR, complete remission; OS, overall survival; RFS, relapse-free survival.

Data used to establish proportion of cases with a given secondary chromosome aberration.

^aAdapted with permission from [14,15,16].

^bPercentages for particular secondary aberrations calculated using only those studies that provided relevant data.

^cPercentage of male patients.

^dPercentage of female patients.

In contrast, the TLE1 and TLE4 genes have recently been suggested as candidate tumor suppressors residing within the 2.4Mb region at 9q21.32–9q21.33 invariably lost in t(8;21) patients with del(9q) [23^••]. The authors knocked down expression of several genes mapped to 9q21.32–9q21.33, and demonstrated that decreased expression of only TLE1 and TLE4 affected myeloid cell survival, proliferation and apoptosis. Additionally, TLE1 and TLE4 expression was lower in t(8;21) patients with del(9q) compared with normal CD34+ cells or cells from t(8;21) patients without del(9q). Moreover, lower expression of TLE1 and TLE4 increased the cell division rate of the t(8;21)-positive Kasumi-1 cell line without del(9q), whereas forced TLE1 and TLE4 expression led to apoptosis and death of the Kasumi-1 cells. These data suggest that haploinsufficiency of TLE1 and TLE4 may cooperate with RUNX1-RUNX1T1 in the leukemogenesis of t(8;21) patients with del(9q) [23^••].

Gene mutations

Genetic events cooperating with CBF-AML-associated gene fusions may also include activating mutations in genes encoding receptor tyrosine kinases (RTKs) or small GTPases.

KIT

The most common RTK mutations are those in the KIT gene, which encodes a member of the type III RTK family. They are found in 12–47% of patients with t(8;21), and in 22–38% of those with inv(16)/t(16;16) [24,25,26^•]. In t(8;21) AML, KIT mutations occur mostly in exon 17 and confer adverse prognosis, whereas the prognostic significance of KIT mutations (both in exon 8 and 17) in inv(16) AML is less well established (Table 2 Table 2) [27–33, 34^•]. However, in two recent large studies, the presence of KIT mutations (predominantly in exon 17) was associated with a higher relapse risk [27,29]. Although in one study overall survival was not impacted [29], in another, KIT mutations affected unfavorably overall survival after adjustment for sex [27]. Notably, KIT mutations can be targeted by tyrosine kinase inhibitors, which are selectively active against specific KIT mutations. For instance, imatinib is active against various exon 8 mutations and the exon 17 mutation involving codon N822, but not mutations involving codon D816, which can be successfully targeted with dasatinib and midostaurin [35]. Thus, determination of the exact type of KIT mutation is necessary for selecting the appropriate tyrosine kinase inhibitor in individual patients.

Table 2.

Prognostic significance of mutations in the KIT, RAS and JAK2 genes in core binding factor-acute myeloid leukemia patients

Mutated gene	% mutated (No. of patients with mutation/total no. of patients studied)	Age range, years (median)	Prognostic relevance	Reference
			[0,1-5]inv(16)/t(16;16)
KIT	30% (18/61)	19–57 (40)	CIR significantly higher for both patients with any KIT mutation and those with exon 17 KIT mutation than for patients with wild-type KIT. OS worse for patients with any KIT mutation in multivariable analysis, after adjustment for sex	Paschka et al. [27]
KIT	32% (20/63)	15–74 (44)a	RR significantly higher for patients with KIT mutations in exon 8 compared with patients without exon 8 mutations. No impact on OS.	Care et al. [28]
KIT	31% (13/42)b	17–<60 (NR)b	RI significantly higher for patients with KIT mutation than for patients with wild-type KIT. No impact on OS.	Cairoli et al. [29]
KIT	22% (10/46)	1–75 (33)c	No significant difference in EFS and OS between patients with and those without KIT mutations.	Boissel et al. [24]
KIT	38% (5/13)	<17 (NR)	No significant difference in EFS and OS between patients with and those without KIT mutations.	Shih et al. [26•]
RAS d	36% (17/47)	1–75 (33)c	No significant difference in EFS and OS between patients with and without RAS mutations.	Boissel et al. [24]
NRAS	36% (41/113)	18–92 (63)e	No significant difference in EFS and OS between patients with and without NRAS mutations.	Bacher et al. [30]
			[0,1-5]t(8;21)
KIT	22% (11/49)	18–71 (37)	CIR significantly higher for patients with KIT mutations compared with those without KIT mutations. OS not significantly different between patients with and without KIT mutations.	Paschka et al. [27]
KIT	13% (8/64)	15–90 (NA)	EFS and OS significantly shorter for patients with exon 17 KIT mutations.	Schnittger et al. [31]
KIT	47% (17/36)	16–76 (40.5)	RI significantly higher and OS shorter for patients with any KIT mutation and for those with exon 17 KIT mutation compared with patients with wild-type KIT.	Cairoli et al. [25]
KIT	12% (6/50)	1–75 (33)c	EFS, RFS and OS significantly shorter for patients with KIT mutations.	Boissel et al. [24]
KIT	17% (8/46)	2–15 (7.5)	RR significantly higher, and DFS and OS shorter for patients with KIT mutations compared with those without KIT mutations.	Shimada et al. [32]
KIT	43% (12/28)	<17 (NR)	No significant difference in RR, EFS and OS between patients with and those without KIT mutations.	Shih et al. [26•]
RAS f	8% (4/50)	1–75 (33)c	No significant difference in EFS and OS between patients with and without RAS mutations	Boissel et al. [24]
NRAS	11% (13/116)	18–92 (63)e	No significant difference in EFS and OS between patients with and without NRAS mutations.	Bacher et al. [30]
		[0,1-5]inv(16)/t(16;16) and t(8;21) combined into one CBF AML group
KIT	41% (11/27)g	0–18 (NR)h	No significant difference in EFS between patients with and those without KIT mutations.	Goemans et al. [33]
KIT	43% (17/41)	<17 (NR)	No significant difference in EFS and OS between patients with and those without KIT mutations.	Shih et al. [26•]
RAS	41% (8/27)i	0–18 (NR)h	No significant difference in EFS between patients with and those without KIT mutations.	Goemans et al. [33]
JAK2	4% (5/138)	NR	DFS significantly shorter for patients with JAK2 mutation compared with those without JAK2 mutations. No impact on OS.	Illmer et al. [34•]

CIR, cumulative incidence of relapse; DFS, disease-free survival; EFS, event-free survival; NR, not reported; OS, overall survival; RFS, relapse-free survival; RI, relapse incidence; RR, risk of relapse.

^aMean age.

^bNumbers and age of patients who received intensive chemotherapy and were analyzed for clinical outcome. The total number of patients studied was 50, including 17 with KIT mutations, and their age range was 17–88 years (median 47).

^cAge range and median provided for all patients with CBF AML, including 56 with t(8;21) and 47 with inv(16)/t(16;16).

^dIncluding 15 patients with NRAS and 2 with KRAS mutations.

^eAge range and median provided for all 2502 patients included in this study.

^fIncluding 2 patients with NRAS and 2 with KRAS mutations.

^gIncluding 11 patients with inv(16)/t(16;16), 6 of whom harbored KIT mutation, and 16 with t(8;21), 5 of whom harbored KIT mutation.

^hAge range provided for all 150 patients included in this study.

ⁱIncluding 11 patients with inv(16)/t(16;16), 2 of whom harbored RAS mutation, and 16 with t(8;21), 6 of whom harbored RAS mutation.

FLT3

An internal tandem duplication (ITD) within the juxtamembrane domain of the FLT3 gene, that encodes an RTK protein, is relatively infrequent in CBF-AML, being detected in 2–9% of t(8;21) patients [24,28,36–39] and 0–7% of inv(16) patients [24,28,36–38]. Nevertheless, mouse models have demonstrated that when FLT3-ITD coexists with RUNX1-RUNX1T1 [39] and CBFB-MYH11 [40^•] in leukemic blasts, it collaborates with the gene fusions in inducing acute leukemia.

The second type of FLT3 mutations, point mutations within the activation loop of the tyrosine kinase domain (TKD), are more frequent in inv(16) patients (6–24%) [24,38,41], but occur in only 2–7% of t(8;21) patients [24,38,39,41]. To our knowledge, the prognostic significance of both FLT3-TKD and FLT3-ITD in CBF-AML remains to be established. Given their relatively high incidence in inv(16) patients, large studies analyzing prognostic impact of FLT3 mutations, which constitute potential therapeutic targets for tyrosine kinase inhibitors, are warranted.

JAK2

The most recent addition to the list of potentially leukemogenic mutations in CBF-AML is the activating, gain-of-function JAK2V617F mutation, previously reported in most patients with chronic myeloproliferative disorders [42]. JAK2 mutations are infrequent in CBF-AML, being identified in 4 of 64 (6%) t(8;21) patients and none of 99 inv(16)/t(16;16) patients in one study [43], and in 5 of 138 (3.6%) CBF-AML patients [2 had t(8;21) and 3 inv(16)/t(16;16)] in another [34^•]. Additionally, Schnittger et al. [44] found JAK2 mutations in two of three t(8;21)-positive patients with therapy-related AML, but in none of 20 t(8;21)-positive de novo AML patients, and suggested JAK2 mutations might be associated with therapy-related CBF-AML after treatment with anthracyclines and topoisomerase inhibitors. Occasionally, JAK2V617F mutation may precede acquisition of t(8;21) during progression from myeloproliferative syndrome to CBF-AML [45^•]. However, JAK2V617F mutations are also detected in de novo CBF-AML [34^•,43,46]. Importantly, disease-free survival (DFS) of CBF-AML patients with JAK2 mutations was significantly worse than DFS of patients without these mutations in one study [34^•]. These results require corroboration. JAK2 mutations are potential targets for small-molecule inhibitors of JAK2 kinases [42].

RAS

Mutations in the RAS genes, NRAS and, less often, KRAS, both of which encode small GTPases, are common among inv(16)/t(16;16) patients, being detected in over one-third of them; RAS mutations are present in 8–11% of t(8;21) patients [24,30]. Thus far, RAS mutations have not been correlated with clinical outcome (Table 2). Mutated NRAS may become a target for molecularly targeted therapy.

Gene overexpression

Gene overexpression has been recently identified as another type of cooperating genetic event in CBF-AML.

MN1

High expression of the MN1 gene has been consistently detected in inv(16) patients at both the RNA [47,48,49^••] and protein levels [49^••]. Carella et al. [49^••] transplanted bone marrow cells overexpressing MN1 into lethally irradiated mice, which rapidly developed myeloproliferative disease, but not AML. Similarly, mice transplanted with MN1-expressing cells in another study developed a hematopoietic disorder resembling AML and rapidly died [50], suggesting that overexpressed MN1 is a highly efficient oncogene. As Heuser et al. [50] could not detect genetic alterations collaborating with MN1 overexpression, Carella et al. [49^••] demonstrated that mice transplanted with cells containing both overexpressed MN1 and CBFB-MYH11 quickly developed AML. Thus, MN1 overexpression seems to constitute a cooperating event in inv(16)-associated leukemogenesis. The mechanism through which MN1 becomes overexpressed is currently unknown. It is unclear whether MN1 expression has prognostic significance in inv(16) patients, as it does in cytogenetically normal AML patients [51,52].

Microarray gene-expression studies

Initial microarray gene-expression profiling (GEP) studies identified signatures allowing reliable separation of patients with inv(16)/t(16;16) from those with t(8;21) and from other cytogenetic, molecular genetic AML subgroups [48,53–55], or both. More recent studies have focused on CBF-AML, and investigated whether GEP could improve outcome prediction. Paschka et al. [56] analyzed GEP in t(8;21) patients lacking KIT mutations and, separately, in inv(16)/t(16;16) patients without secondary aberrations or KIT mutations. Gene-expression-based outcome predictors for event-free survival (EFS) dichotomized each cytogenetic group into two subsets with markedly different EFS [56].

Bullinger et al. [57^••] performed GEP analyses on 93 CBF-AML patients, including both inv(16) and t(8;21) patients. Unsupervised 2-way hierarchical cluster analysis stratified the patients into two groups with different GEPs and significantly different survival. Group I, which had worse overall survival, had more cases with inv(16), higher white blood counts (WBC), and higher incidence of FLT3-ITD-positive cases, although the latter were infrequent (4/28 in group I versus 0/52 in group II). Review of the genes expressed differentially between the groups revealed that the prognostically adverse group I was characterized by overexpression of the proto-oncogene JUN and constitutive activation of the JNK and MAPK signaling pathways. Conversely, deregulated mTOR signaling and antiapoptotic mechanisms were suggested to play a role in the prognostically more favorable group II. Overexpression of such antiapoptotic genes as BIRC3 and BIRC6 is of interest because they encode inhibitors of apoptosis proteins (IAPs), which are potential therapeutic targets for antisense and chemical IAP inhibitors [57^••].

Micrornas in core binding factor-acute myeloid leukemia

MicroRNAs are small (19-25 nucleotides), noncoding RNA molecules that regulate expression of genes at the posttranscriptional level by hybridizing to complementary mRNA targets and inhibiting their translation or causing degradation [58]. As aberrant expression of many microRNAs has been demonstrated in several types of cancer [59], including AML [60,61,62^••], the role of microRNAs in CBF-AML has only begun to be investigated. Fazi et al. [63^•] reported that RUNX1-RUNX1T1 chimeric protein caused epigenetic silencing of the microRNA-223, which is involved in regulation of myelopoiesis. Demethylating treatment with 5-azacytidine in vitro or forced microRNA-223 expression enhanced microRNA-223 levels and restored cell differentiation, suggesting that microRNA-223 might become a therapeutic target.

Another study assessed expression of 260 microRNAs (using quantitative RT-PCR) in 215 AML patients belonging to several cytogenetic and molecular subsets [62^••]. Unsupervised analysis identified distinctive microRNA signatures for most patients with t(8;21) and for most with inv(16), and these distinguished CBF-AML from other subtypes of AML. However, some inv(16) patients clustered among t(8;21) cases and vice versa. Most (83%) t(8;21) cases were grouped in a cluster characterized by mainly downregulated microRNAs, which included members of a known tumor suppressor microRNA family, let-7b and let-7c. Both these microRNAs and microRNA-127, reported to function as a tumor suppressor, were also downregulated in inv(16) AML. The authors speculated that downregulation of microRNA-let-7, which regulates the expression of RAS and HMGA2 oncogenes, and microRNA-127, that regulates BCL6 expression, might contribute to CBF-AML leukemogenesis [62^••].

Epigenetics in core binding factor-acute myeloid leukemia

Epigenetic alterations silencing gene function without changing the DNA coding sequence are intrinsic to CBF-AML leukemogenesis. Under physiologic conditions, the N-terminus of RUNX1 interacts with the CBFB subunit and binds the consensus sequence TGT/cGGT at the promoter regions of its target genes [64]. This allows the RUNX1 C-terminus to interact with a p300-containing co-activator complex resulting in transcriptional activation of RUNX1 target genes via histone acetylation. In t(8;21)-AML, the p300-binding site is replaced by the fusion partner RUNX1T1, which recruits histone deacetylases (HDACs) and induces transcriptional repression of the target genes. Their expression can be restored by treatment with HDAC inhibitors [65,66]. In addition to recruiting HDACs, RUNX1-RUNX1T1 also recruits DNA-methyltransferase 1 (DNMT) [66,67]. This leads to the silencing of RUNX1-target genes by the RUNX1-RUNX1T1 fusion protein through both histone deacetylation and promoter hypermethylation. Combinations of DNMT and HDAC inhibitors induce re-expression of the silenced genes in RUNX1-RUNX1T1-positive leukemic cells [68]. Therapeutic targeting of epigenetic alterations in CBF-AML is now being pursued in the clinic.

Epigenetic alterations may also represent secondary ‘hits’ contributing to CBF-AML leukemogenesis [69,70]. They include methylation of the promoter region of the CEBPA gene, a transcription factor involved in myeloid lineage differentiation, reported recently in inv(16)-AML [71^•]. Downregulation of CEBPA in inv(16)-AML may also be achieved through posttranscriptional interaction of CEBPA mRNA with the RNA-binding protein calreticulin [72] or by microRNA-124a, which is also regulated by promoter methylation [71^•]. In t(8;21)-AML, CEBPA downregulation may result from inhibition of positive autoregulation of the CEBPA promoter by RUNX1/RUNX11T1 protein [73,74^•].

Hematopoietic stem-cell transplantation in core binding factor-acute myeloid leukemia

Two recent retrospective studies have addressed the role of hematopoietic stem-cell transplantation (HSCT) in treatment of CBF-AML patients. The first, restricted to adults less than 60 years harboring t(8;21), compared outcomes of 118 patients who received human leukocyte antigen (HLA)-matched sibling HSCT with 132 patients receiving cytarabine-based chemotherapy postremission [75^••]. The transplant and chemotherapy groups had similar relapse-free survival (RFS, 5-year rates, 55 versus 64%). Although the transplanted patients had lower relapse rates, this did not translate into improved overall survival because HSCT was associated with a higher treatment-related mortality. With regard to overall survival, patients without secondary −Y or − X had a higher risk of death after HSCT than those receiving chemotherapy, with no difference between treatment groups among patients harboring −Y or −X. In both treatment groups, WBC more than 25 × 10⁹/l conferred a higher relapse risk and worse RFS and overall survival. Schlenk et al. [75^••] concluded that cytarabine-based chemotherapy offers results similar or better than HLA-matched sibling HSCT in first remission.

Another study compared results of autologous and allogeneic HSCT in CBF-AML. It found that both transplantation types resulted in similar outcomes; treatment-related mortality was lower and relapse incidence higher in patients receiving autologous HSCT [76^•]. Lower WBC was associated with lower treatment-related mortality and better DFS in t(8;21), but not inv(16)/t(16;16), patients. Among inv(16) patients, women had a lower relapse incidence and a better DFS than men.

Monitoring of minimal residual disease in core binding factor-acute myeloid leukemia

Both CBF-associated fusion transcripts constitute molecular markers useful for assessing MRD. Qualitative PCR assays are of limited value because patients in early or long-term remission may continue to test positive for the fusion transcripts [77]. Assessment of RUNX1-RUNX1T1 by quantitative RT-PCR (Q-RT-PCR) seems to be more promising. The reduction of initial RUNX1-RUNX1T1 levels measured by Q-RT-PCR after induction and consolidation therapies was shown recently to influence prognosis. Patients with RUNX1-RUNX1T1 expression levels below the median after induction and consolidation therapies had a lower cumulative incidence of relapse, and longer EFS and overall survival than patients with transcript levels above the median [78^•]. Q-RT-PCR and flow cytometry at the end of treatment have been successfully used to identify t(8;21) patients with higher relapse risk in another study [79]. However, other studies using Q-RT-PCR have not corroborated these results [80^•,81].

As in t(8;21)-AML, Perea et al. [79] reported that higher MRD levels at the end of treatment measured by flow cytometry or RT-PCR were associated with a higher relapse risk in inv(16) AML. In another study, more than 1-log increase in the CBFB-MYH11 or RUNX1-RUNX1T1 levels in bone marrow during remission caused more than eight times increase in the relapse risk [80^•]. Although promising, the use of MRD analysis for clinical decision making in CBF-AML still requires validation in larger trials.

Conclusion

Recent discoveries of gene mutations, changes in gene-expression and microRNA-expression levels and epigenetic events have increased our understanding of CBF-AML leukemogenesis. Many genetic and epigenetic alterations acquired in addition to the t(8;21)/RUNX1-RUNX1T1 and inv(16)/t(16;16)/CBFB-MYH11 may serve as prognostic factors, therapeutic targets, or both. It is hoped that ongoing studies will ultimately lead to the design of more effective, molecularly targeted therapies that will increase the cure rate of patients with CBF-AML.