FIP1L1/PDGFRA is a molecular marker of chronic eosinophilic leukaemia but not for systemic mastocytosis

Abstract

Sir,During the past decade, a number of molecular markersspecific for haematological malignancies have been identi-fied. For chronic eosinophilic leukaemia (CEL) [1], onerecurrent molecular marker is the FIP1L1/PDGFRA fusiongene product that exhibits tyrosine kinase activity and rep-resents a target of imatinib unless the gene is specificallymutated [2,3]. Over the past few years, an attendant dis-cussion about the relationship of FIP1L1/PDGFRA to sys-temic mastocytosis (SM) has ensued [3]. In this letter, wewish to provide the following comments to clarify this issue.In SM, the frequency of associated haematological non-mast cell disorders (AHNMD) is relatively high [4–6]. UsingKIT D816V as a marker of SM, it has been noted that theAHNMD usually arises from the same clone [7,8]. Even ifKIT D816V is not detected in the AHNMD, both malig-nancies may be of monoclonal origin. Because of therapeuticimplications, the consensus has been to incorporate a specificcategory of SM in the World Health Organization (WHO)classification of mastocytosis, termed SM-AHNMD [9,10].Now, a number of reports have commented that patientswith CEL often present with an increase in bone marrowmast cells [2,3]. These mast cells usually are spindle-shapedand display aberrant expression of CD25 [3], features thatotherwise are found only with mast cells in SM [9]. In mostCEL patients, however, criteria for SM [9,10] are notfulfilled. This is because the mast cells are diffusely dis-tributed, do not form aggregates and do not display KITD816V. Nevertheless, in some CEL patients, mast cells doform multifocal clusters consistent with the diagnosis SM[3]. These patients are appropriately classified as SM-CEL.However, in these patients, the molecular defect, FIP1L1/PDGFRA, is not indicative of SM, but is indicative of theAHNMD, i.e. CEL. It is noteworthy that in these patients,disease symptoms are due to increased eosinophils, and donot overlap with symptoms of SM. Imatinib therefore is pre-scribed in these patients to bring the eosinophil numberunder control, but not to treat the SM-component of thedisease, which is indolent. Interestingly, most patients withSM-CEL do not have a detectable KIT mutation, which isof importance as KIT D816V confers resistance againstimatinib. Thus, in most patients with typical (KIT D816V-positive) SM, therapy with imatinib is not recommended.Regardless of the presence or absence of FIP1L1/PDGFRA,patients with SM may present with eosinophilia [4,6,9,10].Thus, in SM, eosinophilia must be considered as a ‘pre-diagnostic’ checkpoint (SM-eo) at which molecular markersneed to be determined, but is not a final diagnosis or deci-sion point at which treatment with imatinib or other drugscan be recommended. Rather, only a complete staging andgrading of symptoms and findings [9,10] and a consecutivesearch for targets will yield the information necessary fortreatment decisions and drug-selection [3,9,11]. This isimportant as patients with indolent SM usually have a normallife expectancy and do not require cytoreductive or targeteddrugs, even when presenting with eosinophilia.Therefore, in order to establish the correct diagnosis andselect the appropriate approach to management, we recom-mend the use of WHO criteria [9,10] in all patients withSM, including those with eosinophilia (SM-eo).