Chronic myelomonocytic leukemia with double-mutations in DNMT3A and FLT3-ITD treated with decitabine and sorafenib

ABSTRACT

Chronic myelomonocytic leukemia (CMML) is a heterogeneous neoplastic hematologic disorder with worse overall survival. Half of CMML have mutations, but case with concomitant mutations of DNA methyltransferase 3A (DNMT3A) and Internal tandem duplications of the juxtamembrane domain of FLT3 (FLT3-ITD) in CMML was not reported before. We reported a 51-year-old man who had CMML with concomitant mutations in DNMT3A and FLT3-ITD.The patient received decitabine and sorafenib combined treatment. In this report, we reviewed DNMT3A mutation and FLT3 mutation, and we reviewed treatment of decitabine and sorafenib. This report is significant. First: This is the first report on CMML with double-mutations of DNMT3A and FLT3-ITD. Second: It shows the importance of targeted drug in combined treatment of CMML.

KEYWORDS:

• CMML
• DNMT3A
• double-mutations
• FLT3-ITD
• sorafenib

Introduction

Chronic myelomonocytic leukemia (CMML) is a heterogeneous neoplastic hematologic disorder characterized by peripheral blood monocytosis and is a tendency for acute myeloid leukemia (AML) transformation.¹ CMML is further subdivided into 2 subsets, CMML-1 and CMML-2, depending on the number of blasts plus promonocytes in the peripheral blood and bone marrow.² The majority of CMML patients are classified as World Health Organization (WHO) CMML-1.³ Cytogenetic abnormalities are found in about 30% of patients with CMML, but none were specific.² The frequent affected genes in CMML are tet methylcytosine dioxygenase 2 (TET2) (50–60%), Serine And Arginine Rich Splicing Factor 2 (SFSR2) (36–46%), additional sex combs like 1 (ASXL1) (43–44%),⁴ the frequency of affected DNMT3A is 5–10%, and the frequency of affected FLT3 is 1–3%.⁴ There was no report about concomitant mutations of DNMT3A and FLT3-ITD in CMML. We reported a case of CMML-1 with concomitant mutations in DNMT3A and FLT3-ITD. The patient was treated with decitabine and sorafenib. We reviewed DNMT3A mutation and FLT3 mutation, and we reviewed treatment of decitabine and sorafenib in CMML. This report displayed the significant role of targeted drug in combined treatment of CMML.

Case report

A previously healthy 51-year-old Chinese man presented to our hematology department in our hospital with dizzy for 20 d since October 10, 2014. The result of physical examination showed that there was no superficial lymph node enlargement except for mild splenomegaly. Ultrasound showed the thickness of spleen was 4.2cm (normal: 2–3.5cm).The laboratory data showed an increased level of white blood cell (WBC) [115 × 10⁹/L (normal: 4–10 × 10⁹/L)], increased percentages of monocytes (82 × 10⁹/L) on peripheral blood analysis associated with decreased level of hemoglobin(HB) [112 g/L (normal: 120–160)] and normal level of platelet(PLT) [134 × 10⁹/L (normal: 100–300 × 10⁹/L)]. Level of lactate dehydrogenase (LDH) was increased [LDH 420U/L, (normal: 135–225)]. The characteristic of monocytes (folded nuclei and delicate nuclear chromatin) can be appreciated among the granulocytes in bone marrow smear, and bone marrow aspiration revealed obvious granulocytic component (Fig. 1A, B).The blast percentage (including promocytes) is 8% in bone marrow and 4.2% in peripheral blood. The immunohistochemical staining in bone marrow biopsy specimen showed CD68+ CD15+CD34-MPO-(Fig. 1C, D, E, F). Flow cytometric analysis of bone marrow revealed that 78% nucleated cells were monocytes, and there were 56% monocytes with phenotypic abnormalities (CD14-CD64+CD33+) (Fig. 2).Conventional cytogenetic analysis of bone marrow showed normal karyotype. The results from RT-PCR showed there were no fusion genes including BCR-ABL1, PML-RARα, AML1-ETO, CBFβ-MYH11, MLL-AF9, DEK-CAN, NPM1-MLF, FIP1L1-PDGFRA. Molecular genetics analysis revealed the R882H DNMT3A mutation and FLT3-ITD mutation (Fig. 3) without mutation of C-kit exon8 and exon17, CEBPA, ASXL1 exon12, SF3B1 exon15, and SRSF2 exon1.The patient received decitabine(15mg/m², d1–6, CHIA TAI TIANQING, China) and sorafenib (200mg,BID, NATCO, India) treatment after hydroxyureaoral administration decreased the leucocyte count. One month later, bone marrow aspiration revealed complete hematological remission, but mutational analysis was not performed. The patient rejected hospitalization again, so he received long-term treatment of oral sorafenib (200mg, BID). He took blood routine examination once every month and the results were no obvious abnormality. He had no discomfortableness, rash, abnormality of liver and renal function. The peripheral monocytosis did no increase significantly (Fig. 4). On August 14, 2015, the patient was again admitted to our department for fever and petechiae. The physical examination showed that there was an abscess on dorsum of left foot, and there were scattered petechiae on lower limbs without superficial lymph node enlargement and without hepatosplenomegaly. The laboratory data showed normal level of WBC [10 × 10⁹/L (normal: 4–10 × 10⁹/L)], normal level of HB [133 g/L (normal: 120–160)] and increased percentages of mature monocytes(49%) on peripheral blood analysis associated with decreased level of PLT [9 × 10⁹/L (normal: 100–300 × 10⁹/L)]. Bone marrow aspiration revealed obvious increased monocytes (mature monocyte 63%).The ratio of blasts (including promocytes) is 7% in bone marrow. Molecular genetics analysis revealed there was still the R882H DNMT3A mutation, but PCR did not detectFLT3-ITD mutation again. Although the patient was treated with meropenem and teicoplanin, he had persistent hyperpyrexia and died of disseminated intravascular coagulation at 4th day after the patient admitted to hospital. We guessed the uncontrolled infection should be fundamental cause.

Figure 1. Histopathology of bone marrow. (A) the Wright-Giemsa stain in bone marrow smear, the characteristic of monocytes (folded nuclei and delicate nuclear chromatin) is obvious. (B) a bone marrow biopsy specimen, the granulocytic component is most obvious in the biopsy specimen. (C,D) positive staining for CD15 and CD68 E, F: negativity for MPO and CD34.

Figure 2. Phenotypic abnormalities of monocytes detected by a multicolor flow cytometric approach at initial diagnosis.

Figure 3. Molecular genetics analysis of R882H DNMT3A mutation and FLT3-ITD mutation. (A) sequencing of R882H DNMT3A mutation. (B) PCR of FLT3-ITD mutation.

Figure 4. Change trend of the peripheral monocytosis.

Discussion

CMML is described by WHO as a clonal disease of the haematopoietic stem cell represented by persistent monocytosis. In the light of the WHO 2008 classification, diagnosis of CMML needs sustained monocytosis for greater than 1.0 × 10⁹/L for 3 months. In patients with obviously increased leukocyte counts, the monocytes should account for greater than 10% of the WBC.⁵ Cases are divided into CMML-1 (<5% peripheral blasts and promonocytes, <10% bone marrow blasts and promonocytes) and CMML-2 (5–19% peripheral and 10–19% marrow blasts and promonocytes).⁵ Our case met WHO criterion of CMML-1; and results of fusion genes such as BCR-ABL1 ruled out CML. The immunophenotyping analysis of this case showed immature abnormal monocytes without increased blasts. The immunophenotyping analysis did not play the critical role in diagnosis of CMML, but flow cytometric analysis is valuable as a tool for distinguishing CMML from reactive monocytosis.⁶

With a median survival of only 20∼30 months and leukemic transformation rates(5 years) of 45%,the prognosis of CMML is poor overall.⁷ Prognostic parameter recognized by WHO classification did not satisfy demand of treatment, especially for cases with normal karyotype. With whole-exome sequencing application, more and more mutated genes were found in CMML. Overall, about 50% of CMML has mutations in genes. The genes covered many cellular pathways, such as signaling, RNA splicing, transcription, cohesin complex DNA methylation, histone modifications.⁴ In the future, these mutations may play important roles in the prognosis of CMML. ASXL1 mutations in CMML have predictable value; the median survival in cases of ASXL1 mutation (13.2 months) is significantly shorter than that of ASXL1 wild type (30.0 months).⁸ SRSF2 mutations were linked with a statistically significant reduction in All survival (OS).⁹ The genes such as DNMT3A and FLT3 were not studies further in CMML, but have been researched deeply in AML.

DNA methylation is an epigenetic modification that is important in cell differentiation (e.g. haematopoietic stem cell differentiation) and gene regulation. DNA methylation refers to the addition of a methyl group to the C5 position of the pyrimidine ring of cytosines to form 5–methylcytosine.Disrupted DNA methylation patterns in malignancies suggest a role in the development of cancer, but there was seldom evidence that genetic aberrations directly linked the DNA methylation machinery to malignancies. DNMT3A(encoding DNA methyltransferase 3A) is a critically important new tumor-suppressing gene. DNMT3A is a member of the DNA methyltransferase proteins that enzymatically add a methyl group to 5′cytosine in the CpG dinucleotides. The DNMT3A mutations resulted in loss-of-function with a dominant negative effect.¹⁰ Present studies shows that DNMT3A mutation is related to tumorigenesis and prognosis of hematologic malignancies.¹¹ FLT3 is a receptor tyrosine kinase which is expressed on immature haematopoietic cells and takes part in the development of stem cells. The wild type FLT3 receptor is transmembrane protein, and this protein contains 5 extracellular immunoglobulin-like domains, a transmembrane domain, a JM domain and a split intracellular kinase domain. FLT3 is activated and is dimerized when FLT3 ligand binds to the extracellular domain of FLT3.²⁷ Subsequent phosphorylation of FLT3 activates downstream signaling pathways. Mutations of DNMT3A and FLT3 influenced the pathophysiology and clinical features of AML at a different angle.

DNMT3A mutation was intensively studied in AML (Table 1). Recent studies have found that frequency of DNMT3A mutations in AML is 17%∼30% and most of DNMT3A mutations are heterozygous^12,15,16,18 Besides Common DNMT3A R882 mutation, mutations of V897D, G543C and R478W are also found.^12,13 DNMT3A mutations in adult AML are associated with clinical features such as gender, age, white blood cell count¹⁴ andtype of AML.²⁰ Although there were many studies, there was no definite conclusion about effect of DNMT3A mutations on AML prognosis (Table 1).^15-21,26

Table 1. DNMT3A and FLT3 in leukemia.

		DNMT3A	FLT3
Gene function		DNA methyltransferase	receptor tyrosine kinase
Aftermath of mutation		loss-of-function	constitutive kinase activation
Type of mutation	AML	R882, V897, G543, R478	FLT3-ITD,FLT3-TKD
	CMML	R882	FLT3-ITD,FLT3-TKD
Frequency of mutation	AML	21%[12],17.8%[18],30%[15],17.7%[16]	28–34% [29]
	CMML	2/52[22],0/15[23],9%[26]	1/68[30],3/29[31],4.3%[32],3/51[33]
prognosis	AML	no definite conclusion shorter OS in total population [15] [16] [26] [18] [19][20]	worse prognosis [29]
		no influence on OS in total population[17][21] and in the favorable genotype subgroup[26].
		shorter RFS in total population[15] [16][17] [26] [19][20]
		no influence on RFS in total population [18][21]
		no influence on DFS in total population [19]
		no influence on RFS in the favorable genotype subgroup [18]
	CMML	Shorter OS[22][26][9]	no influence on OS[32]

OS: All Survival, RFS: relapse-free survival, DFS: disease-free survival

FLT3 is the most frequently mutated gene in AML (Table 1), with an approximately 28%∼34%of AML cases harboring FLT3 mutations^28,29 FLT3 have 2 kinds of mutations: an internal tandem duplication (ITD) in the juxtamembrane domain and a point mutation of the tyrosine kinase domain (TKD). Estimated 20–25% of AML cases harbor FLT3-ITD mutations,²⁷ and FLT3-ITDmutations in adult AML are associated with clinical features such as having higher white blood cell count and higher peripheral blood blast cell count.⁴⁷ These mutations lead to constitutive kinase activation. Many studies showed that FLT3-ITD was associated with worse prognosis.²⁹

Although there were many studies about DNMT3A and FLT3 in AML, there were fewer reports on DNMT3A in CMML (Table 1). Abdel-Wahab did not detect DNMT3A mutations in 15 CMML.²³ Anna M. Jankowska analyzed DNMT3A mutations in CMML and CMML-derived AML. DNMT3A mutations were identified in 2 of 52 cases with CMML (4%), all with CMML-2 (2 of 16; 12%). 5 mutant cases were detected in cases with CMML-derived AML (6 of 20; 25%). The DNMT3A mutations were mostly heterozygous and they affected the R882 position.²² The result from analysis of 75 CMML (8/75 DNMT3A mutations) showed OS was associated with mutations in DNMT3A(P = 0.015).²⁶ Another report from analysis of 312 CMML (5/227 DNMT3A mutations) showed ASXL1 mutations, not DNMT3A mutations, predicted inferior OS.⁸ From the current research results, the frequency of DNMT3A mutation in CMML tends to increase in progression to AML^{22, 26} So, more clinical studies are needed to define the role of DNMT3A mutations in prognosis of CMML.

Reports of FLT3-ITD mutations were also less in CMML (Table 1). Wang analyzed 68 cases of CMML, and detected only one case with FLT3-ITD mutation.³⁰ Chen W analyzed 29 cases of CMML, and detected 3 cases with FLT3-ITD mutation.³¹ At MD Anderson Cancer Center, analysis of FLT3 mutation was performed on 302 cases of CMML. 13 (4.3%) CMML patients had FLT3 mutations: 8 patients had FLT3-ITD mutations and 5 patients had FLT3-TKD mutations. There was no significant difference in disease characteristics between CMML with and without FLT3 mutations, and there was no significant difference in median OS between CMML with and without FLT3 mutation.³² Shih LY evaluated FLT3-ITD mutation in 198 cases of MDS, and detected 3 cases with FLT3-ITD mutation from 51 CMML patients. There was no significant difference in clinicohematologic characteristics, cytogenetic characteristics, or international prognostic scoring system score among MDS patients with and without FLT3 mutation. But FLT3-ITD was identified as an independent predictor of reduced time to development of AML and reduced overall survival on multivariate analysis. The role of FLT3-ITD in CMML was not estimated because of fewer cases.³³ Although frequency of FLT3-ITD mutation was less in CMML, the result from murine model showed Flt3-ITD “knockin” mice developed myeloproliferative disease resembling CMML.³⁴

Concomitant mutations in DNMT3A and FLT3 were frequent in AML,²¹ and ‘double-mutated' (FLT3/ITD+, DNMT3A+) patients were considered as having very poor prognosis.³⁵ Although there were no reports on CMML with concomitant mutations in DNMT3A and FLT3, we think that concomitant mutations in DNMT3A and FLT3-ITD also might bring poor prognosis in this patient.

Aberrant epigenome changes are now recognized to be vital in driving the development of leukemia. Acquisition of the mutation of DNMT3A in haematopoietic stem cells can lead to clonal amplification resulting in a pre-leukemic stem cell (pre-LSC) population. Pre-LSCs can become fully transformed into fully transformed leukemic stem cells with the acquisition of additional driver mutations (e.g., FLT3-ITD).⁴⁵ CMML with DNMT3A mutation may quickly transform into AML after the mutation of FLT3-ITD happened. One case reported before gave the direct evidence that a pre-leukemic clone can evolve to AML.¹¹ This allogenic bone marrow transplantation case revealed definitely the process of transplanted bone marrow cells with DNMT3A mutation (for one lymphoma patient) transforming into AML after the acquisition of additional driver mutation of FLT3-ITD. After the patient acquired complete remission by inductive treatment, only DNMT3A mutation could be detected without mutation of FLT3-ITD. After the patient relapsed, mutations of DNMT3A and FLT3-ITD could be detected again. Another research showed that DNMT3A deletion in oncogenic Ras models not only significantly promoted CMML progression but also led to transformation to AML.²⁵ We might give a reasonable speculation that DNMT3A might be trunk somatic mutation in CMML^11,24; and DNMT3A mutant brought out CMML through disturbing gene expression/DNA methylation in mice.²⁴ FLT3-ITD mutation might be a key subsequent mutation which transforms CMML to acute leukemia. So the ‘double-mutated' (FLT3/ITD+, DNMT3A+) CMML transformed very quickly to acute leukemia and had a very poor prognosis. We hope there will be zooperal study and clinical study to approve the guess.

The case of ‘double-mutated' (FLT3/ITD+, DNMT3A+) CMML in our report received one cycle treatment of decitabine and sorafenib long-term oral, he got approximately 10 months' remission. This case gave us a good example of combined targeted drug treatment.

Decitabine, as hypomethylating agent, was approved for treatment of patients with MDS. Three studies have now been completed using decitabine specifically in patients with CMML. The results showed complete remission ranged from 10% to 58%^{36, 37} Low-dose, dose-intensity schedule of decitabine optimized epigenetic modulation and clinical responses in MDS and CMML.³⁸ For CMML with normal karyotype, response to decitabine is highly variable. Specific molecular signatures were evaluated to predict decitabine response in CMML.³⁹ A case report showed combination regimen of low-dose (15mg/m² intravenously daily for 5 days) decitabine, low-dose cytarabine, low-dose aclarubicin and granulocyte colony-stimulating factor successfully managed one case of AML transformed from CMML in the elderly.⁴⁰ Until now, the role of DNMT3A mutation was undefined in CMML treated with decitabine.

Sorafenib is a new oral tyrosine kinase inhibitor, and main targets of sorafenib are deregulated receptor tyrosine kinase pathways in cancer including the Ras/Raf/MEK pathway, vascular endothelial growth factor (VEGF) pathway, and FLT3. Sorafenib was used for advanced renal cell cancer and hepatocellular cancer and shows good clinical activity. Sorafenib had superior performance in AML with FLT3-ITD mutation; though there were FLT3 mutations resistant to sorafenib.⁴¹ Tschan-Plessl A studied 17 patients with FLT3-ITD positive AML. These patients were treated with sorafenib (pre-transplant and post-transplant) combined with allogeneic HSCT. The result showed sorafenib with synergy of cGvHD can achieve high rates of sustained remission in high-risk patients.⁴² There were 2 case reports about successful treatment of CMML using sorafenib^43,44 Mutations of DNMT3A and FLT3 influenced the pathophysiology of leukemia at a different angle, the combination of FLT3 and DNA methyltransferase inhibition is synergistically cytotoxic to FLT3/ITD AML cells.⁴⁸

The case in our report received decitabine and sorafenib combined treatment and got approximately 10 months' remission. When he admitted to hospital again, he did not transform to AML. We could still detect DNMT3A mutation, but could not detect FLT3 mutation. In the course of CMML transforming into AML, there was clone evolution. DNMT3A mutation should be trunk somatic mutation in CMML^{11, 45} FLT3-ITD mutation is pivotal subsequent cooperating mutations in development of AML.⁴⁵ The patient did not transform to AML after sorafenib treatment, we surmise loss of FLT3-ITD played an important role. DNMT3A mutations are persistent at long-term remission in adult patients with AML.⁴⁶ The results from decitabine treatment showed CMML complete remission ranged from 10% to 58%^36,37 and there were no reports on the effect of decitabine in CMML with DNMT3A mutation. This patient received only one course of decitabine treatment. After the patient got complete hematological remission induced by treatment of decitabine and sorafenib, the patient rejected treatment of decitabine; he only received treatment of oral sorafenib. Persistence of DNMT3A mutation implied existence of trunk tumor clone and led the relapse of CMML. Because the mutation of FLT3-ITD was controlled by sorafenib, the patient did not transform into AML for 10months. So sorafenib was very important for ‘double-mutated' (FLT3/ITD+, DNMT3A+) CMML. Treatment of sorafenib may effectively prolong ‘double-mutated' (FLT3/ITD+, DNMT3A+) CMML transforming into AML.

In conclusion, we reported one case of CMML with double-mutations of DNMT3A and FLT3-ITD, and this case showed the importance of targeted drug in combined treatment of CMML with FLT3 mutation.

Disclosure of potential conflicts of interest

The authors declare that they have no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.

Funding

This work was supported by a Grant from the National Natural Science Foundation of China (No 81370658).