Imatinib-Induced Clinical Response in Myeloid Neoplasm with Eosinophilic Pneumonitis: A Case Report

Introduction

The 5th edition of the World Health Organization (WHO) and International Consensus Classification recently revised the classification schemes of primary eosinophilic disorders to the major category of “myeloid/lymphoid neoplasms with eosinophilia and tyrosine kinase gene fusions” (MLN-TK) and the myeloproliferative neoplasm (MPN) subtype of “chronic eosinophilic leukemia” (CEL), that lack all known recurrent tyrosine kinase fusion genes [1,2]. Most MLN-TK cases are associated with PDGFRA rearrangements, frequently resulting from inter-stitial deletion of 4q12 resulting in FIP1L1::PDGFRA, or PDGFRB rearrangements, most commonly t(5;12)(q32;p13.2) leading to ETV6::PDGFRB. The ETV6::ACSL6 fusion is an extremely rare abnormality with ~ 20 cases reported worldwide, although the incidence might be underestimated as several studies reported the recurrent translocation of t(5;12)(q31;p13) without performing molecular analysis to verify whether it contained ETV6::ACSL6 fusion [3].

Patients with ETV6::ACSL6 exhibit eosinophilic and/or basophilic leukocytosis with a presentation that is similar to that of patients with the ETV6::PDGFRB rearrangement [4,5]. Notably, while various studies have shown remarkable efficacy with the tyrosine kinase inhibitor (TKI) imatinib in patients with PDGFRA/ PDGFRB rearrangements, responses to TKIs have not been seen in patients with ETV6::ACSL6, which is presumed to be due to its lack of involvement of a tyrosine kinase activity [5–9]. Here, we report a rare case of a male patient diagnosed with myelodysplastic syndrome/neoplasms (MDS) with an acquired t(5;12)(q31;p13) leading to an ETV6::ACSL6 fusion presenting with eosinophilia who demonstrated a clinical response and normalization of the eosinophil blood count to imatinib treatment. We aimed to show the potential benefit of imatinib in patients with an ETV6::ACSL6 fusion associated myeloid neoplasms who are refractory tol1 corticosteroids.

Case Report

A 71-year-old man with a past medical history significant for coronary artery disease presented with symptoms of fatigue and weight loss, without other systemic B-symptoms. No lymphadenopathy, organomegaly, or petechiae were noted. His complete blood cell count revealed a low hemoglobin (Hgb) of 9.3 g/dL (normal range, 12.4–17.3 g/dL), a normal white blood cell (WBC) count of 10.0×109/L (normal range, 4.0–11.0×109/L), and a mildly low platelet count of 139×109/L (normal range, 150–400×109/L). Bone marrow biopsy demonstrated a markedly hypercellular bone marrow with multilineage dysplasia and ring sideroblasts (50%), and the absolute number of monocytes was only minimally increased (0.9×109/L, normal range, 0.2–0.8×109/L), with a normal blast percentage of 2%.

Flow cytometry did not reveal any increase in aberrant myeloblasts. Cytogenetic analysis revealed a normal male karyotype, without abnormalities identified by fluorescence in situ hybridization (FISH) for MDS-associated abnormalities, and targeted DNA-sequencing identified mutations in DNMT3A (R882H, VAF 38.4%), RUNX1 (S293fs, VAF 37.2%), and SF3B1 (K700E, VAF 37.8%). He was diagnosed with MDS with mutated SF3B1 and multilineage dysplasia, formerly known as MDS with ring sideroblasts, with a moderate-low risk score based on the molecular international prognostic scoring system. He was treated with luspatercept for several months with no response, and a repeat bone marrow biopsy 6 months later showed persistent disease with erythroid and megakaryocytic dysplasia, an increase in reticulin fibrosis (Grade 2), and a stable blast percentage of 2%. Treatment was switched to azacitidine subcutaneously for 5 cycles of 28 days, but he remained transfusion dependent requiring weekly red blood cells and platelet transfusions. To reevaluate his disease status, a bone marrow biopsy was performed, demonstrating persistent trilineage dysplasia with 1% blasts. Cytogenetic analysis, however, identified an acquired t(5;12)(q31;p13)[6]/46,XY[14] abnormality (Figure 1A), consistent with clonal evolution of the MDS. FISH was negative for FIP1L1/PGFFRA or PDGFRB gene rearrangements, but an ETV6::ACSL6 fusion was identified by RNA-sequencing (Figure 1B, 1C). His complete blood count revealed an absolute eosinophil count of 3.0×109/L (44% of the WBC differential) (normal range, 0.0–0.7×109/L). A diagnosis of CEL was considered, but in the absence of an increased bone marrow blast count (2%), he did not meet the diagnostic criteria. He was diagnosed with a myeloid neoplasm associated with t(5;12)(q31-q33;p13) based on the history of an underlying MDS [1,2,4]. He then received allogeneic hematopoietic cell transplantation (HCT) from a 10/10 HLA-matched unrelated donor with fludarabine (30 mg/m2) and melphalan (100 mg/m2) conditioning, and post-transplant graft-versus-host-disease prophylaxis included cyclophosphamide (50 mg/ m2 on D+3 and D+4), tacrolimus, and mycophenolate mofetil. Post-transplant chimerism at Day +100 was >98% donor (recipient DNA was below the threshold of detection) and post-transplant bone marrow chromosome analysis revealed a normal male karyotype. Complete blood count showed a normal absolute eosinophil count of 0.11×109/L.

However, at Day +154 after his allogeneic HCT, he developed disease relapse with loss of donor chimerism, without overt dysplasia. Chromosome analysis again revealed t(5;12)(q31;p13) [6]/46,XY[5], with mutations in DNMT3A (R882H, VAF 25.7%), RUNX1 (S293fs, VAF 11.3%), and SF3B1 (K700E, VAF 26.6%) identified by targeted DNA-sequencing, and his complete blood cell count showed an absolute peripheral eosinophil count of 0.7 x 109/L (13% of the WBC differential). Peripheral smear and bone marrow morphology showed an increase in eosinophils and a hypercellular bone marrow, with flow cytometric studies showing a high side-scatter with bright expression of CD45, CD11b, and CD15 and dim CD33, and negative for CD10, HLA-DR, CD16 and CD64 (Figure 2A–2C). There was no increase in the blast percentage.

He rapidly clinically deteriorated with refractory hypoxic respiratory failure and was diagnosed with acute eosinophilic pneumonitis after ruling out other inflammatory and infectious processes. He received high-dose methylprednisone 250 mg every 6 hours, with modest but incomplete improvement in his eosinophilia, and he persistently required high levels of supplemental oxygen. In the absence of other good therapeutic options and based on data supporting the use of imatinib in MLN-TK which clinically has a similar presentation and FDA approval for the use of imatinib in patients with CEL, a decision was made to empirically start imatinib 400 mg given a rising eosinophil count and steroid refractoriness. After initiation of imatinib, the patient’s absolute eosinophil counts rapidly normalized over the course of 5–10 days (Figure 3). Concomitantly, he clinically improved and we were able to gradually wean him off supplemental oxygen. He was discharged from the hospital on Day 40 of imatinib.

Two weeks later, he presented again with symptoms of acute-on-chronic heart failure and a gastrointestinal bleeding in the setting of persistent severe thrombocytopenia. His eosinophil count (0.4×109/L) remained well controlled within the normal limits on imatinib and a stable respiratory status, but he developed septic shock and died due to complications.

Discussion

Myeloid neoplasms associated with t(5;12)(q31-q33;p13) and eosinophilia often result from the fusion between PDGFRB and ETV6 (formerly known at TEL) [10]. The ETV6::ACSL6 fusion is an exceptionally rare variant that clinically mimics ETV6::PDGFRB but has not been shown to be responsive to imatinib due to the lack of involvement of a tyrosine kinase and the treatment implications are not well understood [3,4,7,11]. Based on the promising response to imatinib in our case, we suggest that imatinib may in fact show some activity in a subset of patients with corticosteroid-refractory disease and should be considered in patients who are refractory to treatment with corticosteroids.

ETV6 is a transcription factor belonging to the ETS family that is frequently mutated in leukemia and MPN/MDS and is critical for (early) hematopoiesis and megakaryocyte development. It has been described to be fused to at least 30 partner genes in myeloid neoplasms [11,12]. While the mechanism through which ETV6 translocations t(5;12) cause eosinophilia is unknown, it has been hypothesized that it is driven by a super-enhancer that located within the ETV6 gene locus that interacts with genes adjacent and distal to breakpoints on chromosomes 5 and 12 coding for pro-inflammatory factors such as IL-3 and IL-5, triggering the eosinophilia [14,15]. ACSL6 codes for a long-chain acyl-CoA synthetase that plays a role in lipid metabolism and ATP generation. It has been shown to correlate with progression in AML, but its mechanism in tumori-genesis remains unclear [16].

The prognosis of ETV6::ACSL6-associated disease is usually poor, and essentially all organ systems may be susceptible to the effects of sustained eosinophilia, posing life-threatening risks as seen in our case. While fusion genes with rearrangement of PDGFRA or PDGFRB are known to be associated with rapid and durable complete hematological and clinical remission on treatment with imatinib, it is unclear how to treat ETV6::ACSL6-associated myeloid neoplasms. Imatinib is an oral tyrosine kinase inhibitor that was initially found to target PDGFR but has also been found to target tyrosine kinases such as c-KIT and the BCR-ABL fusion protein. It first received FDA approval for chronic myelogenous leukemia in 2001 and has since received FDA approval for gastrointestinal stromal tumors, and various other hematological conditions, including Philadelphia-positive acute lymphoblastic leukemia, systemic mastocytosis, hypereosinophilic syndrome and/or CEL in patients with or without the FIP1L1/PDGFRA fusion kinase [17]. Although imatinib is approved for CEL in the absence of a tyrosine kinase gene fusion, ETV6::ACSL6-associated myeloid neoplasms are thought to be primarily resistant to currently TKI due to the lack of involvement of a tyrosine kinase [4,6]. This assumption, however, might be preliminary, as only 6 cases to date have received a TKI; 4 patients received imatinib and 2 patients received sorafenib due to its multitargeted activity towards signal transduction molecules (Table 1) [5–7]. Our patient, who was refractory to high-dose steroids, exhibited a rapid and sustained normalization of the eosinophil count after treatment with imatinib, leading to substantial clinical improvement.

Conclusions

We report an uncommon case of MDS with an acquired t(5;12) (q31;p13) resulting in a ETV6::ACSL6 fusion. While ETV6::ACSL6 has clinical features that overlap with ETV6::PDGFRB but generally lacks the exquisite responsiveness to imatinib due to its lack of tyrosine kinase activity, confirmation of the underlying fusion genes is crucial. Based on the promising clinical response with normalization of the eosinophil count to imatinib in our case, we suggest that imatinib may in fact show activity in a subset of patients lacking the presence of a tyrosine kinase gene fusion with corticosteroid-refractory disease. Further molecular investigations are needed to elucidate the underlying mechanism of imatinib sensitivity in ETV6::ASCL6-associated disease, given the absence of genetic involvement of a tyrosine kinase. Given that the kinases inhibited by imatinib (eg, ABL, KIT) may overlap with IL-3 and IL-5 in the downstream signals that they activate, we speculate that inhibition of such kinase pathways may blunt the effect of the increased IL-3 and IL-5 engendered by ETV6::ASCL6 [15]. Thus, imatinib combined with drugs targeting IL-3 or its associated pathways should also be considered in future studies.