CSF3R T618I mutated chronic myelomonocytic leukemia: A proliferative subtype with a distinct mutational profile.

Chronic myelomonocytic leukemia (CMML) is a clonal myeloid neoplasm characterized by sustained monocytosis and mutations in TET2, ASXL1, SRSF2, SETBP1, NRAS, and KRAS. We describe a rare case of CSF3R T618I mutated CMML that has a proliferative phenotype, myelodysplasia, and additional mutations in ASXL1, SETBP1, KRAS, and PTPN11. Comparing the clinicopathologic features of this case to previously reported cases of CSF3R T618I mutated CMML and CSF3R non-T618I mutated CMML, CSF3R T618I seems to define a unique proliferative subtype of CMML with a distinct mutational profile. The diagnostic challenges and molecular pathogenesis associated with this case are also briefly discussed.

Introduction

Chronic myelomonocytic leukemia (CMML) is a clonal myeloid neoplasm characterized by persistent absolute (>1 × 109/L) and relative (>10% of leukocytes) monocytosis in peripheral blood (PB), the presence of myelodysplasia or acquired clonal genetic abnormalities, and no morphologic or cytogenetic/genetic features suggestive of another myeloid malignancy [1, 2]. The neoplasm is subdivided into a dysplastic subtype with white blood cell count (WBC) <13 × 109/L and a proliferative subtype with WBC >13 × 109/L. Patients with a dysplastic subtype are often cytopenic and thus more likely to present with symptoms such as bleeding, infection, and fatigue, whereas patients with a proliferative subtype more often present with hepatosplenomegaly, fever, and weight loss. The typical mutational profile for CMML involves TET2, ASXL1, SRSF2, SETBP1, and genes in the RAS signaling pathway [1, 2].

Colony-stimulating factor 3 receptor gene (CSF3R) mutation is rare in CMML and is more often associated with chronic neutrophilic leukemia (CNL), a neoplasm characterized by sustained neutrophilia and hypercellular bone marrow with a predominance of neutrophilic granulocyte forms [3]. CSF3R mutation is detected in ∼80–90% of CNL cases (90% of mutated cases with CSF3R T618I mutation) and is highly sensitive and characteristic of CNL [4, 5]. It is also detected in ∼30% of atypical chronic myeloid leukemia cases (aCML, ∼60% of mutated cases with CSF3R T618I mutation) [6], [7], [8].

We describe a rare case of CSF3R T618I mutated CMML and compare the clinicopathologic features of this case with previously reported CSF3R T618I mutated CMML and CSF3R non-T618I mutated CMML. The diagnostic challenges and molecular pathogenesis associated with this case are also briefly discussed.

Case report

A 27-year-old woman was incidentally noted to have leukocytosis and macrocytic anemia during her pregnancy. She later presented for sustained leukocytosis, monocytosis for about one year, macrocytic anemia, and hepatosplenomegaly. There was no family history of cancer or personal evidence of a syndrome. PB, bone marrow (BM) aspirate, and core biopsy were submitted for morphology, flow cytometric immunophenotyping, cytogenetics, and molecular studies.

The PB smear (Fig. 1A) demonstrated granulocytic leukocytosis (WBC 35 × 109/L) with left-shifted neutrophils, including myelocytes and metamyelocytes [5%, green arrow; mature neutrophils (segmented/banded) 62%] and occasional dysplastic neutrophils, markedly relative and absolute monocytosis (29%, 7 × 109/L, blue arrow), severe anemia (hemoglobin 4.3 g/dL) with occasional coarse basophilic stippling (orange arrow), and mild thrombocytopenia (platelets 146 × 109/L). BM aspirate smear demonstrated left-shifted granulopoiesis with 3% myeloblasts (Fig. 1B, red arrow), scattered atypical/dysplastic neutrophils, increased monocytic cells (10%, blue arrow), dysplastic erythroids with ring sideroblasts (10% of erythroids, Fig. 1C, orange arrow) and nuclear irregularity (Fig. 1D, orange arrow), increased myeloid: erythroid ratio (7.9:1), and dysplastic hypolobated and small megakaryocytes (Fig. 1E, violet arrow). Core biopsy and clot sections (Fig. 1F) showed a hypercellular BM (cellularity ∼100%) with increased left-shifted granulocytic and monocytic cells and occasional hypolobated megakaryocytes.

Fig. 1

Morphologic findings in peripheral blood and bone marrow. Peripheral blood smear (A) shows left-shifted neutrophils (green arrow), monocytosis (blue arrow), and coarse basophilic stippling (orange arrow). Bone marrow aspirate shows scattered myeloblasts (red arrow, B), monocytic cells (blue arrow, B), ring sideroblasts (C), nuclear irregularity in erythroids (D), and dysplastic megakaryocytes (E). Clot section shows hypercellularity, atypical megakaryocytes, and left-shifted granulocytic and monocytic cells (F). [Wright-Giemsa stains in images A, B, D, and E (100x objective), iron stain in C (100x objective), and hematoxylin and eosin stain in F (20x objective)].

Flow cytometric analysis on the BM aspirate (Fig. 2) was performed on 10-color BD FACSCanto flow cytometer (Becton Dickinson, San Jose, CA) and analyzed using cluster analysis with Cytopaint Classic Software (Leukobyte, CA). Analysis revealed a 0.88% population of CD34(+)/CD117(+) myeloblasts (in red) with immunophenotypic aberration [CD56(+), CD5(dim+), CD33(dim+), CD11b (few +)], abnormal maturation pattern in neutrophils (in green) on the CD11b/CD13 plot, and increased immunophenotypically aberrant monocytes (in blue, 11% with a nearly uniform expression of CD56) and a high fraction of CD14(+)/CD16(-) classical monocytes (estimated at 96% of total monocytes).

Fig. 2

Immunophenotypic findings in bone marrow aspirate by 10-color flow cytometry: There are immunophenotypic aberrancy on CD34(+)/CD117(+) myeloblasts (in red, 0.88% population, aberrantly expressing CD5, CD56, and dimly expressing CD33), increased monocytes [(in blue, 11%, with a nearly uniform expression of CD56 and a high fraction of CD14(+)/CD16(-) classical monocytes (estimated at 96% of total monocytes)], and neutrophils (in green) with abnormal maturation pattern on the CD11b/CD13 plot. The fractions of CD14(+)/CD16(+) intermediate and CD14(-)/CD16(+) non-classical monocytes could not be estimated due to the panel design.

The cytogenetic study revealed a normal female karyotype. Next Generation Sequencing (NGS) by Foundation Heme panel revealed mutations in CSF3R T618I [VAF (variant allele frequency) 11.2%)], KRAS G12D (5.8%), ASXL1 S1014fs*10 (48.2%), BCORL1 R1149W (51.5%), PTPN11 E76Q (2.0%), SETBP1 D868N (3.8%), and SETBP1 G870S (5.1%); there was no evidence of mutations in JAK2, MPL, CARL, TET2, or SRSF2, or fusion of BCR-ABL1. PTPN11 E76Q has been reported in hematologic malignancies [9] and is likely a somatic mutation given its low VAF (2.0%). Overall, there are no-known risk factors for inherited predisposition syndromes.

The constellation of morphologic and immunophenotypic findings in conjunction with molecular abnormalities led to a diagnosis of CMML-1 with an unusual CSF3R T618I mutation. The follow-up clinical information regarding treatment and prognosis is not available.

Discussion

The presence of the CSF3R T618I mutation along with leukocytosis, sustained monocytosis, and myelodysplasia in our case raises differential diagnoses that include CNL, aCML, and CMML. Key diagnostic criteria for CNL include peripheral leukocytosis (WBC count >25 × 109/L) with a neutrophilic predominance (>80% segmented/banded neutrophils), no dysgranulopoiesis or monocytosis (<1 × 109/L monocytes in PB), and either presence of an activating CSF3R mutation (most commonly T618I) or persistent (>3 months) neutrophilia and splenomegaly with no other identifiable causes. Key diagnostic criteria for aCML include peripheral leukocytosis (WBC count >13 × 109/L) with neutrophilic precursors accounting for >10% of the WBC, dysplastic neutrophils, hypercellular BM with granulocytic/neutrophilic proliferation and dysplasia, and no or minimal PB basophilia and monocytosis (basophils <2% of WBC, monocytes <10% of WBC).

In our case, the findings of sustained absolute and relative monocytosis, multilineage dysplasia, and lack of neutrophilic predominance in PB (<80% segmented/banded neutrophils) argue against CNL despite the CSF3R T618I mutation. Additionally, while both aCML and CMML are classified as myelodysplastic/myeloproliferative neoplasms and share many overlapping morphologic and genetic features, including myelodysplasia and recurrent mutations in KRAS, ASXL1, and SETBP1 (as seen in our case) [10], significant relative and absolute monocytosis with immunophenotypically aberrant monocytes (uniform expression of CD56) and a high fraction of classical monocytes (96% of total monocytes, higher than the cut-off of 94%) argues for CMML with an unusual CSF3R T618I mutation over aCML. Specifically, these immunophenotypic findings support the neoplastic nature of monocyte proliferation and help to distinguish CMML from myelodysplasic/myeloproliferative neoplasms with associated monocytosis [11, 12].

One unusual finding in our case is the presence of ring sideroblasts without mutation in spliceosome-related genes. Ring sideroblast status has not yet been reported in CSF3R mutated CMML. While the mutation in spliceosome-related genes (such as SF3B1 and SRSF2) plays a role in the formation of ring sideroblasts, the mechanisms responsible for mitochondrial iron accumulation are not completely understood. It is possible that ring sideroblasts in this case resulted from alteration(s) in the genes involved in iron (heme and iron-sulfur cluster biosynthesis) and mitochondrial protein metabolism [13, 14].

Notably, the CSF3R mutation is rare in CMML, with only ∼40 cases being reported (∼30 cases with CSF3R non-T618I and ∼10 cases with CSF3R T618I, accounting for ∼4% and ∼1% of CMML, respectively) [4, 8, [15], [16], [17]]; our case is thus unusual. Although the mechanism by which the CSF3R mutation contributes to pathogenesis in CMML is largely unknown, it seems that the CSF3R T618I mutation, a membrane-proximal mutation, results in ligand-independent receptor activation and constitutive downstream signaling through JAK2, whereas the other membrane-distal CSF3R mutations may lead to ligand hypersensitivity and increased cell surface CSF3R expression [6, 18]. This proposed mechanism may explain why leukocytosis (proliferative subtype) is commonly associated with CSF3R T618I mutated CMML (a higher median WBC of 38 × 109/L in 6 cases [16] and WBC of 35 × 109/L in our case), whereas the non-proliferative subtype is commonly associated with CSF3R non-T618I mutated CMML (a median WBC of 11.3 × 109/L in 6 cases [15]).

Moreover, the mutational profile in CSF3R T618I mutated CMML also appears to be distinct from CSF3R non-T618I mutated CMML and CSF3R unmutated CMML. CMML is characterized by recurrent mutations in TET2 (∼60%), ASXL1 (∼40%), SRSF2 (∼50%), SETBP1 (∼15%), and genes in the RAS signaling pathway (∼30%) [1, 2], characteristic of an aging-associated disease with mutations affecting epigenetic (DNA methylation and histone modification), splicing, and signaling pathways. Compared to CSF3R unmutated CMML, CSF3R T618I mutated CMML (including our case) seems to have more frequent mutations in ASXL1 (∼80%) and infrequent mutations in TET2 (0/8, 0%) or SRSF2 (1/8, ∼14%), a feature suggestive of minimal disturbance of DNA methylation and splicing pathways [16, 17]. This contrasts with the mutational profile in CSF3R non-T618I mutated CMML, with the frequency of TET2 and SRSF2 mutations (∼30% for each) similar to CSF3R unmutated CMML [4, 8, 15]. Our case additionally has mutations in genes in the RAS signaling pathway (KRAS and PTPN11) and SETBP1; the presence of multiple mutated genes in the signaling pathways (KRAS and CSF3R) may contribute to the proliferative phenotype of CMML. The mutation status in KRAS, PTPN1, and SETBP1 in CSF3R mutated CMML was not systematically reported. It would be interesting to evaluate any changes in the mutational patterns over the course of the disease.

Similarly, CSF3R T618I mutated CNL seems to have a mutational profile different from CSF3R non-T618I mutated CNL regarding the SETBP1 mutation; this mutation is more frequent in CSF3R T618I mutated CNL (∼50%) than in CSF3R non-T618I mutated CNL (∼5–10%). The mutational frequency of ASXL1 is similar between these two subtypes of CNL [4, 5, [19], [20], [21]].

Finally, the proposed mechanism of CSF3R T618I pathogenesis may provide a rationale for molecularly-directed targeting of CSF3R T618I mutated CMML with JAK2 kinase inhibitors such as ruxolitinib [6, 21]. Encouragingly, the clinical benefit of ruxolitinib was recently demonstrated in a subset of CNL patients (with lower-risk features) harboring CSF3R T618I mutation with hematologic responses and allele burden reduction [22]. However, given the complex mutational profile in our CMML patient, including ASXL1 and SETBP1, therapies including hypomethylating agents, hydroxyurea in addition to supportive care with packed red cell transfusions to alleviate severe anemia were considered. Additionally, JAK2 inhibitors are increasingly recognized as potential therapies for CSF3R T618I mutated myeloproliferative neoplasms, however the effects may be transient and are not well established [23]. Given the patient's overall fitness, young age, and desire for curative-intent, allogenic hematopoietic cell transplantation with a matched unrelated donor was ultimately successfully pursued.

Conclusion

CSF3R mutations in a myeloid neoplasm with monocytosis may pose a diagnostic challenge. A comprehensive morphologic, immunophenotypic, and molecular study is essential to arrive at the correct diagnosis. CSF3R T618I mutation is rare in CMML and seems to define a unique proliferative subtype of CMML with a mutational profile distinct from both CSF3R non-T618I mutated CMML and CSF3R unmutated CMML. However, additional cases of CMML with this unusual mutation need to be studied to further characterize genotype-phenotype associations and better understand clinical and therapeutic outcomes.

Funding

None

Informed consent

Not Applicable (The samples were collected only for diagnosis and a general IRB rule was applied in this study.)

Declaration of Conflict of Interests

None