Background

Hypereosinophilia is a feature of a variety of uncommon hematologic disorders like hyperseosinophilic syndrome (HES), systemic mastocytosis (SM) and chronic eosinophilic leukemia (CEL). Approximately 4% of patients with HES or SM have interstitial deletion of chromosome 4q12 leading to juxtaposition of FIP1L1 and PDGFRA[1]. The fusion product is exquisitely sensitive to therapy with imatinib mesylate, and hence its identification has important therapeutic ramifications particularly in hematologic disorders presenting with hypereosinophilia [2–5]. We herein present the case of a patient in whom FIP1L1-PDGFRA was discovered at the time of evolution from chronic myelomonocytic leukemia (CMML) to refractory acute myeloid leukemia and how therapy with imatinib resulted in durable complete remission.

Case presentation

A 64-year-old Caucasian man presented to our institution with a 6-month history of progressive leukocytosis. Per the patient’s outside medical records, a bone marrow biopsy at initial presentation had shown a 100% cellular marrow with marked myeloid hyperplasia. Conventional cytogenetics demonstrated a diploid male karyotype. Fluorescence in situ hybridization studies (FISH) were negative for myelodysplasia-associated abnormalities. Molecular studies were negative for BCR-ABL rearrangement, and JAK2V617F and MPLW515L mutations. Based on these features, he was diagnosed with myelodysplastic/myeloproliferative neoplasm, unclassifiable, and was started on hydroxyurea, 1.5 grams daily. At presentation to our institution, his white blood cell (WBC) count was 44.6 ×109/L with 76% neutrophils, 6% metamyelocytes, 10% monocytes, 6% lymphocytes, 1% eosinophils and 1% blasts, with absolute monocytosis (4.46 × 109/L) and eosinophilia (0.45 × 109/L). He was anemic (hemoglobin 9.0 g/dL) and mildly thrombocytopenic (platelet count 119 × 109/L). He did not have splenomegaly on physical examination. Bone marrow evaluation performed at presentation to our institution revealed a 100% cellular bone marrow with myeloid hyperplasia. Megakaryocytes were decreased in number and included rare dysplastic forms. Wright-Giemsa stained smears prepared from the bone marrow aspirate were remarkable for increased myeloid cells and trilineage dysplasia; a 500-cell differential count showed mildly increased monocytes (6%) and myeloid blasts (7%). The constellation of findings was diagnostic of chronic myelomonocytic leukemia (CMML-1). Conventional cytogenetics showed trisomy 8 in two of twenty analyzed metaphases; this was confirmed by fluorescence in situ hybridization (FISH) using an alpha-satellite (D8Z2) CEP8 probe (positive in 4% of the cells studied). Reverse transcriptase polymerase chain reaction (RT-PCR) performed on the bone marrow aspirate was negative for BCR-ABL fusion. Targeted next-generation sequencing mutation analysis was negative for 53 “hotspot” mutations analyzed as described previously [6].

The patient was enrolled on the SGI-110 clinical trial, a phase 1–2 dose escalation, multicenter study of SGI-110, a DNA hypomethylating agent, in subjects with intermediate or high-risk myelodysplastic syndromes (MDS) or AML. He received two courses of SGI-110, after which he experienced rapidly progressive leukocytosis with a peak WBC count of 126 × 109/L and new onset of peripheral eosinophilia (11%). Bone marrow evaluation performed at this time demonstrated nearly 100% cellularity with myeloid hyperplasia; in contrast to the previous biopsy, prominent eosinophilia and moderate myelofibrosis were noted in this sample (Figure 1a-b). Wright-Giemsa stained smears showed increased granulocytes with left-shifted maturation and prominent eosinophilia (16%) in a background of trilineage dysplasia. There was no significant increase in bone marrow monocytes (5%). Myeloid blasts comprised 12% of total nucleated cells. The unexpected and abrupt presence of prominent eosinophilia in the peripheral blood and bone marrow at this point in time prompted us to evaluate for PDGFRA rearrangement. FISH analysis performed using a LSI-4q12 tricolor rearrangement probe that hybridizes to the chromosome 4q12 region containing the FIP1L1, CHIC2 and PDGFRA genes revealed deletion of the CHIC2 gene in 86.5% of the cells analyzed indicating the presence of the FIP1L1-PDGFRA rearrangement. The FIP1L1-PDGFRA fusion transcript was further confirmed by RT-PCR. Low-level trisomy 8 was also detected by FISH in this sample. Based on these findings a diagnosis of myelodysplastic/myeloproliferative neoplasm with eosinophilia and PDGFRA rearrangement was rendered. A follow-up bone marrow biopsy after one month showed acute myeloid leukemia with 26% blasts. In addition to persistent low-level trisomy 8, conventional cytogenetics and FISH demonstrated a new clone with TP53 gene deletion. FISH was positive for deletion of the CHIC2 gene, TP53 deletion and trisomy 8 in 90%, 10% and 9% of the analyzed cells, respectively.

Figure 1
figure 1

Bone marrow core biopsy at the time of initial FIP1L1- PDGFRA rearrangement discovery. The bone marrow is hypercellular (100%), with prominent eosinophilia and myeloid hyperplasia, mild increase in immature cells, and features of myelofibrosis manifesting primarily as cellular streaming. (a: 10× objective; b: 20× objective; hematoxylin and eosin stain).

Full size image

The patient was then started on imatinib mesylate 400 mg daily along with a short course of idarubucin and subcutaneous cytarabine for cytoreduction. He achieved complete hematologic and morphologic remission and went on to receive a matched unrelated allogeneic stem cell transplant (SCT). To date, the patient remains in complete hematologic, morphologic and molecular remission with successful engraftment as demonstrated by chimerism studies. The sequence of events is provided in Table 1.

Table 1 Sequence of clinical events
Full size table

Discussion

The PDGFRA and PDGFRB proteins are members of the class III receptor kinase family that also includes c-KIT, and FLT3 [7]. PDGFRA is located on chromosome 4q12 [8]. A small interstitial deletion of 4q12 leads to juxtaposition of FIP1L1 and PDGFRA resulting in a gain of function fusion protein with signal independent kinase activity and therefore increased cell proliferation and survival [9]. This interstitial deletion is generally cryptic and not detectable using standard cytogenetic banding techniques.

FIP1L1- PDGFRA rearrangements are often associated with chronic eosinophilic leukemia and hypereosinophilic syndromes [9, 10] as well as systemic mastocytosis [11]. Pardanani et al. reported a prevalence of approximately 4% for FIP1L1- PDGFRA fusion gene in a large series of patients with suspected or established HES or systemic mastocysosis [1]. We recently described a case of chronic neutrophilic leukemia associated with FIP1L1- PDGFRA rearrangement [4]. The basis for the apparent lineage predilection of FIP1L1- PDGFRA for eosinophils is not well understood. The hypothesis is that it is present in all myeloid lineages, but that eosinophils are particularly sensitive to the FIP1L1- PDGFRA proliferative signal [12]. In contrast to rearrangements involving PDGFRB, the FIP1L1- PDGFRA rearrangement is exceedingly rare in the setting of chronic myelomonocytic leukemia [13–16]. To our knowledge, only one other case has been reported in the literature by Zota et al. [17] In contrast to our patient, the patient reported by Zota et al. did not evolve in to acute myeloid leukemia, albeit the acquisition of FIP1L1- PDGFRA fusion was considered a feature of disease evolution. Initiation of imatinib resolved eosinophilia but did not effectively improve other counts, and the patient subsequently progressed to CMML-2 and developed extramedullary disease in abdominal lymph nodes; she succumbed in 10 months. A case series from Germany described five patients with FIP1L1- PDGFRA who presented with AML and eosinophilia, but no history of antecedent myeloid malignancy was reported for any of the patients [18]. A summary of case reports describing PDGFRA rearrangements arising in patients with myeloid neoplasms commonly not associated with such rearrangements is provided in Table 2. All patients received imatinib therapy and achieved at least a hematologic response. One patient got sorafenib and another got dasatinib after acquiring resistance to imatinib. Two of eight patients maintained molecular response, while two maintained hematologic response at last reported follow up. Notably, clonal acquisition of FIP1L1-PDGFRA has not been reported in the setting of acute myeloid leukemia evolving from CMML.

Table 2 PDGFRA rearrangement in unusual adult myeloid neoplasms
Full size table

Our case highlights the importance of assessing for PDGFRA rearrangement in myeloid neoplasms with de novo or subsequently acquired eosinophilia. The identification of the FIP1L1- PDGFRA fusion gene is significant since imatinib has excellent efficacy at low doses (100-400 mg daily) in FIP1L1-PDGFRA-positive neoplasms [2, 5]. Of note, due to the 250-fold lower IC50 as compared to BCR-ABL, reports suggest that even once weekly doses of imatinib are adequate in the setting of FIP1L1-PDGFRA[3]. However, these responses are eventually lost due to emergence of an imatinib-resistant T614I mutation in the ATP-binding site of PDGFRA[9]. In our patient, the ability to induce a complete remission using imatinib at a time when the patient was unresponsive to chemotherapy induction permitted subsequent allogeneic SCT and an ensuing durable remission as of last follow up.

Conclusion

We describe a case report of a patient who transformed from CMML to AML which was refractory to standard chemotherapy. Emergence of peripheral and bone marrow hypereosinophilia during this transformation led to suspicion of presence of FIP1L1-PDGFRA rearrangement, which was confirmed by FISH and RT-PCR. Treatment with imatinib led to a complete remission and permitted allogeneic SCT therapy.

Endnote

The identification of new onset of eosinophilia in acute myeloid leukemia arising in a patient with chronic myelomonocytic leukemia might indicate acquisition of imatinib-responsive FIP1L1-PDGFRA rearrangement.

Consent

Granted under protocol approved by the Institutional Review Board of The University of Texas M.D. Anderson Cancer Center.