We read with interest the report by Klion et al1 that describes a myeloproliferative variant of hypereosinophilic syndrome (HES) that is associated with elevated serum tryptase and hyperplasia of dysplastic mast cells (MCs) in the bone marrow (HES-tryptase). It was argued that HES-tryptase is distinct from systemic mast cell disease associated with eosinophilia (SMCD-eos). In support of this notion, the authors considered several features that were absent in HES-tryptase but expected in SMCD-eos: focal mast cell aggregates, MC coexpression of CD2 and CD25, and the presence of the D816V c-kit mutation. However, we question the accuracy of such a classification.
Clinical presentation in adults with SMCD is markedly heterogeneous,2 and making or refuting the diagnosis requires a careful morphologic analysis of the bone marrow. In general, SMCD is characterized by focal, dense aggregates of dysplastic MCs. However, the bone marrow MC infiltration pattern in aggressive SMCD, including SMCD-eos, can be diffuse, and whether one appreciates a “dense” or “loose” scattering of MCs in this setting is open to subjective bias.3 In our experience, either dense or loose aggregates of dysplastic MCs are seen in both FIP1L1-PDGFRA+ and c-kit D816V+ SMCD-eos. This was illustrated in a recent report of 5 patients with SMCD-eos in which all 3 patients who carried the FIP1L1-PDGFRA fusion had pathognomonic MC aggregates in the bone marrow in a pattern that was not different from 1 of the patients with the c-kit D816V mutation.4
In regard to immunophenotypic characteristics of neoplastic MCs, we have recently reported that aberrant MC expression of CD2 is not a uniform disease feature in SMCD.5 While CD25 was aberrantly expressed in all 22 patients studied in that report, the prevalence of CD2 coexpression was much lower (41%), and CD2 expression was only occasionally seen in SMCD that was associated with another clonal hematologic disorder. In our recent report of the 5 patients with SMCD-eos, MC CD25+CD2– expression profile was seen in all 5 patients, including the 3 with the FIP1L1-PDGFRA fusion and the 2 with the c-kit D816V mutation.4
Not considering a diagnosis of SMCD on the basis of absence of c-kit D816V mutations is not accurate given the wide variation in the reported prevalence of such mutations in sporadic SMCD, which may be as low as 20%, depending upon the source of patient sample analyzed and the patient population being studied.6 The occurrence of the FIP1L1-PDGFRA fusion gene has only recently been described for patients with HES7 and has also been demonstrated in clonal eosinophilia, including chronic eosinophilic leukemia and chronic myeloproliferative disorder associated with eosinophilia.8 Therefore, the suggestion that patients carrying this fusion gene have HES as opposed to SMCD-eos may not be accurate. In fact, we had reported on the efficacy of imatinib in SMCD-eos well before the discovery of the drug target, the FIP1L1-PDGFRA fusion protein.9 We have subsequently shown that SMCD-eos patients who responded to imatinib carried the FIP1L1-PDGFRA fusion.4
These observations suggest that the cases described by Klion et al may actually represent SMCD-eos rather than HES associated with reactive mast cell proliferation. Further clarification awaits the performance of either interphase cytogenetics or other molecular approaches for FIP1L1-PDGFRA detection in purified primary mast cells. Until then, our data indicate, as has been demonstrated for gastrointestinal stromal tumors,10 that activating mutations in c-kit and PDGFRA may operate as alternative and mutually exclusive oncogenic events in SMCD-eos.
Hypereosinophilic syndrome with elevated serum tryptase is a syndrome that differs from systemic mast cell disease with eosinophilia
We thank Dr Tefferi et al for their provocative discussion concerning the distinction between a primary mast cell disorder with eosinophilia (SMCD-eos) and hypereosinophilic syndrome with elevated tryptase (HES-tryptase). Although we agree with Dr Tefferi et al that our data do not definitively prove that HES-tryptase is an entity distinct from SMCD-eos, we feel strongly that HES-tryptase is a distinct clinical syndrome in which the primary pathology is due to the presence of excessive numbers of activated eosinophils. As such, it differs from pure systemic mast cell disease in a number of aspects. First, the clinical features of HES-tryptase, including endomyocardial fibrosis and mucosal ulcerations, are hallmarks of HES and are not a feature of systemic mast cell disease.1 Second, as has been shown previously, biopsies of affected tissues (other than bone marrow) in patients with HES did not show evidence of mast cells histologically or upon staining with antitryptase antibody, but did, in the 2 instances in which staining was performed, show eosinophil granule protein deposition. Third, in our experience, all of the patients with clinical signs and symptoms of SCMD accompanied by peripheral eosinophilia have bone marrow findings characteristic of SCMD with evidence of the codon 816 c-kit mutation in bone marrow aspirate samples enriched for mast cells.
It is certainly possible that the FIP1L1-PDGFRA fusion could cause a different phenotype (eg, SMCD-eos vs HES-tryptase) depending on the cell type of origin or the breakpoints of the fusion. An example of this is the TEL-TRKC fusion gene that has been associated with both acute myelogenous leukemia and infantile fibrosarcoma.2 In this regard, we note with interest the description of a single patient with the FIP1L1-PGDFRA mutation and the presence of mast cell aggregates in the bone marrow. Whereas we agree that molecular analysis of purified mast cells for the FIP1L1-PDGFRA (F/P) fusion is of interest, the fusion has already been found in purified eosinophils by us (A.D.K., unpublished results, 2003) and others.3 Consequently, only the absence of the mutation in mast cells would shed light on the issue of the cellular pathogenesis of HES-tryptase. However, in view of the rapid disappearance of spindle-shaped mast cells from the bone marrow in response to imatinib mesylate therapy, the specific absence of the mutation in the mast cells seems unlikely.
If both eosinophils and mast cells derive from the same neoplastic clone in patients with HES-tryptase, as has been described in patients with SMCD-eos and the codon 816 mutation,4 the distinction between HES-tryptase and SMCD-eos on the basis of cell lineage involvement becomes irrelevant. More important is the identification of clinical features, such as those described in our study, that can be used to distinguish patients with imatinib-responsive F/P fusion-mediated disease from patients with the imatinib-unresponsive codon 816 c-kit mutation. The danger in using the same terminology (SMCD-eos) for the 2 disorders is that clinicians who do not have access to the techniques necessary to make a molecular diagnosis will treat all patients with spindle-shaped bone marrow mast cells and peripheral eosinophilia, including those with the codon 816 c-kit mutation, a condition in which treatment with imatinib may have potentially serious consequences.
Correspondence: Amy D. Klion, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bldg 4, Rm 126, 9000 Rockville Pike, Bethesda, MD 20892; e-mail: aklion@nih.gov
