Myeloid/Lymphoid Neoplasms With (or Without?) Eosinophilia and Tyrosine Kinase Fusion Genes: Search and Ye Might Find

A 41-year-old woman was referred to hematology for isolated mild leukocytosis on routine laboratory testing. She was experiencing mild fatigue, which she attributed to a busy lifestyle and being a mother to small children; she was otherwise having no symptoms. A physical examination was unremarkable, though she had abdominal imaging remarkable for mild splenomegaly (14.9 cm). A complete blood count showed a white blood cell count of 12.9 THO/µL (61.4% segmented neutrophils, 16.8% lymphocytes, 11.9% monocytes, 2.9% eosinophils, 1% basophils, 1% metamyelocytes, 1% myelocytes, and 1% blasts); hemoglobin, 13.3 g/dL; hematocrit, 41 percent; mean corpuscular volume, 96 fL; and platelets, 176 THO/µL. A review of a peripheral blood smear showed teardrop poikilocytes and nucleated red blood cells, and confirmed that she had circulating blast cells. Next-generation sequencing panel for myeloproliferative neoplasms (MPNs) from a reference laboratory identified no mutations in JAK2, MPL exon 10, and calreticulin (CALR) exon 9. Fluorescence in-situ hybridization (FISH) for BCR-ABL rearrangements was normal. Vitamin B12 level was elevated (1,499). A bone marrow biopsy was recommended. The aspirate was hemodilute, erythroid and myeloid precursors showed orderly maturation without atypia, and megakaryocytes were not well represented; there were 0 percent blasts and 8 percent eosinophils. Ring sideroblasts were not observed.

The core biopsy was hypercellular (95%), and the pathologist commented that the erythroid precursors were left shifted with full maturation; myeloid precursors were relatively decreased, though with full maturation; and megakaryocytes were decreased and showed occasional cytologic atypia (hypo- and abnormally lobated hyperchromatic forms). A reticulin stain showed grade 1 fibrosis. Cytogenetics showed no metaphases. FISH studies performed on the bone marrow aspirate were also negative for a BCR-ABL1 rearrangement. Molecular studies showed a CALR exon 9 variant (c.1142_1144delAGG, p. E381fs, estimated variant allele frequency 50%). Molecular studies were also negative for JAK2 V617F, JAK2 exon 12, JAK2 exon 13 (G5712S), and MPL W515L or MPL W515K mutations. Based on the combined molecular and morphologic findings, the initial interpretation was favoring early primary myelofibrosis associated with a CALR mutation. Next-generation sequencing panel of peripheral blood showed no disease-associated mutations including CALR, which prompted a reconciliation of the previously known outside studies. Upon review, these study results showed a three base-pair in-frame deletion without known disease associations rather than a pathogenic frameshift. The overall interpretation continued to be that of an MPN. However, in the absence of an identified pathogenic cytogenetic or molecular alteration, additional studies were recommended, including repeat cytogenetics and, given modest eosinophilia noted on the aspirate smear, FISH studies to exclude a myeloid neoplasm with eosinophilia harboring a rearrangement of PDGFRA, PDGFRB, FGFR1, or PCM1-JAK2. Peripheral blood FISH studies for these were sent and results showed detection of a PCM1-JAK2 fusion involving (chromosome 9p24) rearrangement in 58 percent of cells.

A 67-year-old woman was referred after routine blood work found mild anemia and left shift at a post-retirement physical. She had previously been found to have elevated platelet count as part of an occupational health screen on several occasions dating back more than 10 years. Now her white blood cell count was 3.8 (differential notable for 64% neutrophils, 9% lymphocytes, 5% monocytes, 1% eosinophils, 2% basophils, 6% metamyelocytes, 3% myelocytes, and 3% blasts; 8% nucleated red blood cells were also observed); hemoglobin, 10.1 g/dL; mean corpuscular volume, 81 fL; and platelet count was normal (169 THO/µL). Her history was otherwise unremarkable, though examination identified an enlarged spleen. Peripheral blood studies included FISH for a BCR-ABL1 translocation (not detected), and molecular testing for JAK2 V617F and MPL mutations (not detected). A deletion mutation in CALR was identified and was considered to be the common type 1 frameshift mutation. A bone marrow biopsy was done next, showing hypercellularity with trilineage hematopoiesis and atypical megakaryocytes occurring in frequent dense clusters. There was no increase in blasts, and reticulin stain highlighted grade 3 fibrosis that was interpreted as myelofibrosis in fibrotic phase, associated with a mutation in CALR.

A karyotype reported subsequently was notable for the following abnormal female karyotype: 46,XX,t(4;12) (q12,p13)[2]/46,XX[18], likely involving the PDGFRA gene at 4q12 with the ETV6 gene at 12p13; confirmatory FISH testing showed an extra signal for PDGFRA likely resulted from a rearrangement of the PDGFRA gene.

As is typical in oncology and hematology, all patient management needs to begin with specific and accurate diagnostics. These cases were notable for several reasons; both patients were relatively asymptomatic, and both lacked concerning eosinophilia on peripheral smear; both appeared initially to have a classical MPN. However, our experience with MLNE is that it always feels like a bit of a surprise. What should our thresholds be for considering this rare group? How hard should we look? Or worse yet, how often do we miss them?

The NCCN provides some helpful guidelines for suspecting MLNE (Table). Informally, “triggers” other than eosinophilia for considering MLNE tend to be by association or a “Spidey sense” that the data for another provisional diagnosis are not quite holding together. An example of this is an apparent MPN/myelodysplastic syndrome (MDS) overlap with no clonal abnormalities identified, or an elevated tryptase without obvious systemic mastocytosis. Another example is the hopefully rare event of an apparent myeloproliferation arising soon after a T- or B-cell lymphoma or acute lymphoblastic leukemia. However, we are not superheroes and in reality, patients are not reassured by reliance on superpowers. Instead, we cast a wide net and rationalize the generous use of FISH (pun intended) by the idea that an actionable mutation might be identified by this modality alone.

One cautionary observation is that neither of these cases had eosinophilia. Nevertheless, there were some breadcrumbs — an elevated B12 level, enlarged spleen, and leukoerythroblastosis in the first case. However, all these features can be present in classical MPNs. Before the identification of the rearrangement involving JAK2, the case might have been interpreted as the entity unofficially referred to as “triple-negative” myelofibrosis or myelofibrosis without one of the three common driver mutations (estimated to be <5% of myelofibrosis). A careful study of a series of such patients has shown that they are commonly misclassified, often readjudicated as MDS with myelofibrosis or MDS/MPN ²  This patient highlights a variation on that misclassification.

The identification of a translocation involving a tyrosine kinase has an initial “Aha!” moment: For all, there is a small molecule inhibitor to try. For some, such as the chronic phase of MLNE associated with FIP1L1-PDGFRA, we anticipate a quick and sustained response with low doses of imatinib. But for others, such as PCM1-JAK2 rearranged disorders, identifying them in chronic phase is unusual, and response to small molecule inhibitors appears to be brief.³  Early referral for allogeneic stem cell transplantation is likely to be important.

The second case is a woman with a disease associated with a PDGFRA rearrangement. But her disease seemed to “miss the memo” on what we expect to see for this rare disorder. In addition to durable response to treatment, two memorable features of PDGFRA rearranged eosinophilic disorders include a striking male predominance and challenges in disease detection.⁴ FIP1L1-PDGFRA is generally a cryptic rearrangement, hence the recommendation for a generous application of FISH to identify this group of disorders writ large. This patient is the exception that proves both rules, though easily resolved by a different fusion partner (ETV6) in a non-cryptic rearrangement. These deviations from the “text book case” are also likely also to have implications for response to therapy and future treatment recommendations. The patient also has a canonical CALR mutation and a medical history and bone marrow findings suggestive of an underlying classical MPN, with the PDGFRA rearrangement as a possible secondary event. Additional testing found disease-associated mutations in ASXL1 and DNMT3A, again pointing to a possibly more pernicious disease course and consideration of allogeneic stem cell transplantation. But not without imatinib first!

The path to a diagnosis of a rare disease is often winding and inelegant. But it is not a completely unworn path. Guidelines for rare disorders, like colleagues, can be particularly helpful. A toast to the NCCN MLNE guidelines!

Dr. Hexner has received research support from Blueprint Medicines, Novartis Pharmaceuticals, and Tmunity Therapeutics and serves on the hematology exam committee for the American Board of Internal Medicine.