ETV6-FLT3–positive myeloid/lymphoid neoplasm with eosinophilia presenting in an infant: an entity distinct from JMML

We present the first report of an infant with an ETV6-FLT3 fusion-driven myeloid/lymphoid neoplasm with eosinophilia, T-lymphoid, and myeloid differentiation (MLN-Eo), which exhibited unique sensitivity to type I FMS-like tyrosine kinase 3 (FLT3) inhibitors.

The study was conducted in accordance with the Declaration of Helsinki. An 8-month-old male infant presented to a local emergency department with a 2-week history of fever, irritability, and decreased oral intake. On examination, the patient was found to be febrile, with massive hepatosplenomegaly, lymphadenopathy, and a diffuse rash (Figure 1A). A complete blood count (CBC) showed a white blood count (WBC) of 250 × 10³/μL, with eosinophilia (23%), circulating promonocytes (7%), and blasts (14%). Bone marrow aspirate was performed, and a preliminary diagnosis of juvenile myelomonocytic leukemia (JMML) was made. After leukapheresis and initiation of hydroxyurea, the patient was transferred to Memorial Sloan Kettering Cancer Center for further workup and management.

On transfer to our center, the aforementioned physical findings were observed; the patient was also noted to have significant edema and associated respiratory distress. A CBC on arrival revealed a WBC of 68 × 10³/μL. Flow cytometric analysis of peripheral blood revealed the following: an abnormal immature T-cell population (12%), an abnormal myeloid blast population (0.1%), monocytosis (17%), and eosinophilia (21%). Review of submitted bone marrow aspirate smears confirmed increased blasts/blast equivalent (23% in total) but no significant dysplasia (Figure 1B). Cytogenetics (Figure 1C) revealed a clonal t(12;13)(p13;q12) by karyotype, confirmed by fluorescence in situ hybridization (FISH) as a rearrangement involving ETV6 in 83% of cells. The fusion was identified in multiple flow-sorted cell lineages, including CD34⁺ myeloid blast, immature T-cell, and monocyte populations, at high levels. Targeted DNA- and RNA-sequencing revealed an ETV6-FLT3 fusion, fusing ETV6 exons 1-6 to FLT3 exons 14-24 (Figure 1D) and no evidence of RAS pathway mutations. Based on these findings, a diagnosis of MLN-Eo was made. Results of the rash skin biopsy confirmed involvement by the same leukemic process, containing the same ETV6-FLT3 fusion as in the blood and marrow.

Given the critical status of the patient, sorafenib (150 mg/m² per dose, twice daily), targeting the FLT3 fusion, was started to provide clinical stabilization. After 12 days of sorafenib and hydroxyurea therapy, there was marked improvement of hepatosplenomegaly and leukemia cutis, resolution of respiratory distress, and improvement of leukocytosis to a WBC of 12 × 10³/μL. However, evaluation of the bone marrow revealed residual leukemia with 13% blasts morphologically and 57% ETV6-rearranged cells by FISH. The patient went on to receive 3 cycles of intensive chemotherapy combined with continuous sorafenib (Figure 2A). This regimen led to reduction in disease burden as shown by a decrease in FISH positivity for the ETV6 rearrangement after each cycle of chemotherapy. However, after completion of these 3 cycles of chemotherapy and continuous sorafenib, bone marrow evaluation revealed persistent low-level minimal residual disease (MRD) by FISH and targeted RNA-sequencing.

During treatment with combination chemotherapy and sorafenib, patient-derived leukemia cells from diagnosis were treated with a panel of FLT3 inhibitors for 96 hours and 50% inhibitory concentration values calculated following colorimetric (alamarBlue, MilliporeSigma) assessment of viability (Figure 2B). Relatively lower 50% inhibitory concentration values were observed across type I FLT3 inhibitors compared with type II inhibitors, suggesting preferential sensitivity to type I FLT3 inhibition. Based on these results, gilteritinib (a type I inhibitor) was started with a goal to achieve MRD-negative remission. After 2 months of single-agent gilteritinib, the patient had no detectable rearrangement of ETV6 by FISH. He then underwent a conventional bone marrow transplant (BMT) from his HLA-identical brother. He developed stage III skin graft-versus-host disease, which responded rapidly to treatment. The patient restarted gilteritinib on day +45 posttransplant. All marrow evaluations after BMT have been negative for ETV6 rearrangement by FISH. Furthermore, his most recent marrow evaluation was negative for ETV6-FLT3 by ddPCR of bone marrow aspirate–derived DNA (Figure 2B). The patient completed 1 year of posttransplant therapy with gilteritinib and remains free of disease.

MLN-Eo and rearrangement of PDGFRA, PDGFRB, or FGFR1, or with PCM1-JAK2, is a World Health Organization–established category of hematologic malignancies primarily arising in adults.¹  Rare case reports of patients with similar clinical features harboring ETV6-FLT3^2-9  and other FLT3 rearrangements^9,10  have been reported, supporting the inclusion of FLT3-rearranged disorders into this World Health Organization–defined category of MLN-Eo. To our knowledge, this report is the first describing an infant with MLN-Eo with ETV6-FLT3 rearrangement.

In adults, the clinical presentation of MLN-Eo with ETV6-FLT3 rearrangements primarily includes features of chronic myeloproliferative neoplasm (MPN), such as lymphadenopathy and splenomegaly,^2,3,6,7  although progression from chronic to acute/proliferative phases has been described.^3,4  Our patient presented at a young age with an aggressive phenotype marked by high-grade leukocytosis with >20% blasts and massive hepatosplenomegaly. The clinical features of this patient overlapped with the common presenting signs of children with JMML, a more common MPN in this age group, characterized by mutations leading to RAS pathway activation.¹¹  However, the high percentage of blast/blast equivalents (>20%), presence of an expanded immature T-cell population, and pronounced eosinophilia strongly pointed to an alternative diagnosis. Recently, a case of a child with JMML characterized by a novel CCDC88C-FLT3 fusion was reported,¹²  and previous reports have also described rare MPN variants lacking RAS pathway mutations and characterized by tyrosine kinase fusions of PDGFRB, ALK, and ROS.^13,14  Despite clinical and laboratory similarities with JMML, tyrosine kinase fusion–driven MLN-Eo may be a distinct entity¹⁵  and present a therapeutic opportunity unavailable to patients with classic RAS pathway–driven JMML. In fact, our patient, as well as others described,^2,3,6,16  have benefited from tyrosine kinase inhibitor therapy to attain complete remission.

The oncogenic kinase fusion in this case was defined by the joining of exon 6 of ETV6 to exon 14 of FLT3. The FLT3 product retains the juxtamembrane and both tyrosine kinase domains. In addition, FLT3-fusion MPN has shown in vitro and in vivo sensitivity to FLT3 inhibitors.^2,3,6,16,17  Ex vivo drug screening of patient-derived neoplastic cells revealed significantly increased sensitivity to type I FLT3 inhibitors, which bind to the gatekeeper domain of FLT3 regardless of receptor conformation.¹⁸  Type II FLT3 inhibitors, such as sorafenib, bind to a region adjacent to the adenosine triphosphate (ATP)–binding domain when the protein is in its inactive conformation.¹⁸  This suggests that the ETV6-FLT3 fusion leads directly to FLT3 activation, as this protein conformation results in increased sensitivity to type I, compared with type II, FLT3 inhibitors. We started our patient on the type II FLT3 inhibitor sorafenib based on safety and efficacy data from a Children’s Oncology Group study evaluating its efficacy in children with FLT3-internal tandem duplication–positive acute myeloid leukemia.¹⁹  Although we were able to reduce disease burden down to an MRD-positive level with sorafenib and chemotherapy, it was only after treatment with the type I inhibitor gilteritinib that we were able to achieve MRD negativity according to FISH.

In conclusion, this infant presented with clinical features suggesting JMML but was ultimately identified to have a tyrosine kinase fusion–driven MLN-Eo. Our clinical experience, as well as others reported in the literature,^2,3,6,20,21  suggests that a strategy incorporating tyrosine kinase inhibitors with hematopoietic stem cell transplantation is an effective treatment paradigm for this rare group of kinase-driven MPN. Ex vivopatient-specific data suggested preferential sensitivity to type I FLT3 inhibitors, which aided our selection of a more sensitive inhibitor and ultimately a cytogenetic complete response for our patient. Further studies are needed to better compare the genetic and biologic features of these entities compared with classic RAS pathway–driven JMML.

Funding support was provided in part by a National Institutes of Health, National Cancer Institute Cancer Center Support Grant (P30 CA008748).

Contribution: B.S. and N.S. provided direct patient care, performed analyses, and wrote the paper; F.S.D.C. and G.D.I.S. designed and performed the experiment; Y.Z., W.X., M.R., and R.B. performed analyses and diagnostic evaluation and edited the paper; A.M. provided direct patient care and diagnosis; M.I.R.-S. and N.B. obtained clinical data and helped write the paper; and A.L.K. helped design the experiments and edited the paper.

Conflict-of-interest disclosure: W.X. has received research support from Stemline Therapeutics. R.B. has received a grant and travel credit from ArcherDx; honoraria for advisory board participation from Loxo Oncology; and speaking fees from Illumina. A.M. is on the advisory board of AstraZeneca; and receives research funding from Incyte. M.R. has equity in Auron Therapeutics; and has been involved with provision of services from Celgene and Physicians’ Education Resource. The remaining authors declare no competing financial interests.

Correspondence: Neerav Shukla, MSK Kids, 1275 York Ave, New York, NY 10065; e-mail: shuklan@mskcc.org.