Hypereosinophilic syndromes (HES) are a rare heterogeneous group of disorders characterized by marked peripheral blood and tissue eosinophilia resulting in organ damage.1 In 10%—50% of HES cases,2–4 a constitutively activated fusion tyrosine kinase was identified on chromosome 4q12, derived from an interstitial deletion, that fuses the platelet-derived growth factor receptor-α gene (PDGFRα) to a newly described human gene FIP1-like-1 (FIP1L1).4FIP1L1-PDGFRα is a clonal marker for chronic eosinophilic leukemia (CEL).
Imatinib mesylate, a tyrosine-kinase inhibitor, is effective in the treatment of chronic myeloid leukemia and other disorders with activation of tyrosine kinase family genes.5 The FIP1L1-PDGFRα-associated CEL patients also respond efficiently to imatinib mesylate with rapid and complete hematological remissions that were observed at lower doses than used for treating chronic myelogenous leukemia.6,7
In this study, using nested reverse-transcriptase polymerase chain reaction (RT-PCR), the FIP1L1-PDGFRα fusion gene was identified in 8 of 24 Chinese eosinophilia patients who were also confirmed to have chromosome 4q12 deletion by fluorescence in situ hybridization (FISH) technique. This fusion gene significantly correlated with disease progression in CEL patients. Imatinib mesylate induced molecular remission of FIP1L1-PDGFRα and reversed bone marrow fibrosis, dramatically altering the approach to the treatment of CEL. Meanwhile, tyrosine kinase family genes like PDGFRα, PDGFRβ, C-KIT, FGFR1, ABL and FLT3, as well as gene mutation “hotspots” frequently involved in myeloproliferative diseases like MPL515 and JAK2V617F, were also evaluated.
Twenty-four newly diagnosed HES, 18 male and 6 female, aged 16 to 72 years (median 43 years), were enrolled from 1997 to 2006. Clinical and biological findings of these patients are listed in Table 1. The diagnosis was established according to the criteria of Chusid et al8 and World Health Organization classifications. Ten patients with reactive eosinophilia and 367 healthy volunteers were referred as controls. The study was approved by the University and Institutional Review Board and informed consent was obtained from all study participants.
For each patient a bone marrow sample was collected at diagnosis and mononuclear cells were separated by density-gradient centrifugation with Ficoll solution. Chromosomes were R-banded and G-banded on unstimulated bone marrow cells after a 24-hour culture. Karyotypes were classified according to International System for Human Cytogenetic Nomenclature.9 Inclusion in the study required the analysis of ≥20 metaphase cells per patient.
RNA extraction and RT-PCR analysis
Total RNA was extracted from bone marrow cells using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). First-strand cDNA was synthesized from 1 μg total RNA using Superscript II reverse transcriptase (Invitrogen) with random hexamers according to the manufacturer's instructions. The common structure of tyrosine-kinase domains of PDGFRα, PDGFRβ, C-KIT, FGFR1, ABL, FLT3 as well as MPL515 and JAK2V617F mutations, were amplified using primer sequences listed in Table 2. The FIP1L1-PDGFRα transcript was detected by nested RT-PCR as described by Klion et al.10
The resultant PCR products were purified on Qiagen columns (Qiagen GmbH, Hilden, Germany) and sequenced by corresponding primers on ABI PRISM® 3700 DNA Analyzer using BigDye® Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems). For further confirmation of the insertion and deletion, the purified PCR products were ligated into pGEM®-T Easy Vector (Promega Corporation, Madison, WI, USA) and used to transform DH5α E. coli cells. On agarose gel with ampicillin, individual colonies were screened by IPTG and X-Gal, quantified on electrophoresis, and sequenced by T7 and SP6 primers following the same protocol.
Fluorescencein situhybridization (FISH)
Detection of the 4q12 deletion was performed by a dual-color FISH assay on metaphase cells from bone marrow using BAC clone 120K16 (mapped centromeric of FIP1L1) and BAC clone 3H20 (mapped between FIP1L1 and PDGFRα), labeled with Texas Red and fluorescence (FITC), respectively, following the manufacturer's instructions (Roche Diagnostics Co., Indianapolis, IN, USA). The FISH signals were evaluated as follows: normal chromosome 4 was seen as the appearance of either a yellow color signal as the result of the overlapped green/red signals (3H20-green signal and 120K16-red signal) or a combined green/red signals, deletion was detected as a sole 120K16-red signal because the 3H20-green signal is lost in cases with the FIP1L1-PDGFRα fusion.
Cell transfection andin vivosplicing assays
The PDGFRβ DNA-WT (intron 18 to intron 19) and PDGFRβ DNA-MUT (containing C>T at 35515 within intron 18 and A>G at 35936 within intron 19) were obtained by RT-PCR from the patient samples using the primers PDGFRβ-I18-F (5′-TCGAGACATCATGCGGGAC-3′) and PDGFRβ-I19-R (5′-CCCGTTTGATGGCATTGTAG-3′). The PCR products were cloned into pGEM®-T Easy Vector (Promega Corporation), cut by EcoRI and further inserted into the pFLAG-CMV expression vector (Sigma-Aldrich, St. Louis, USA). The resulting constructs (2.0 μg), named pCMV4-PDGFRβ DNA-WT and pCMV4-PDGFRβ DNA-MUT, were transfected into COS7 or HEK293 cells using SuperFect® Transfection reagent (Qiagen GmbH) as previously reported.11 Total RNA was extracted 36 hours after transfection and reverse transcripted as mentioned above. Sequence analysis of PCR amplification of the plasmid-derived cDNA was performed using the primers, PDGFRβ-I18-F and pFLAG-CMV4-R (5′-TATTAGGACAAGGCTGGTGGGCAC-3′).
Patient characteristics were compared using χ2 test and the Fisher exact test for categorical variables and the Wilcoxon test for continuous variables. Overall survival was measured from the date of diagnosis to either the date of death or the end date of the study, December 31st, 2006. Survival functions were estimated using the Kaplan-Meier method and compared using the log-rank test. Statistical differences were considered significant when the 2-sided P <0.05. All statistical analyses were performed using SAS 8.2 (SAS Institute, Cary, NC, USA).
FIP1L1-PDGFRαfusion gene expressed in CEL of subset of HES patients
The normal karyotype was identified by conventional cytogenetic analysis in 24 patients. The FIP1L1-PDGFRα fusion gene, which is a clonal marker, occurred in 8 of them (33.3%, cases 3, 4, 10, 12, 18, 20, 23, 24, Figure 1A). These cases were then diagnosed with CEL. None of 10 patients with reactive eosinophilia displayed this abnormality. To determine the relative sensitivity of the analysis we performed a set of dilutions of FIP1L1-PDGFRα positive cDNA extracted from the bone marrow of case 23 (43.5% eosinophils) and analyzed each dilution by nested RT-PCR. The results were positive at a dilution of 1/10 000, confirming that it is a sensitive technique. Consistent with RT-PCR data, FISH analysis revealed a chromosome 4q12 deletion in all 8 cases with FIP1L1-PDGFRα fusion transcript, two representative cases (cases 4 and 23) are shown in Figure 1B.
As demonstrated in Figure 2, sequencing of three fusion transcripts revealed breakpoints scattered on FIP1L1 exon 8a-9 (cases 20 and 23) or exon 10–10a (case 18), whereas breakpoints in the PDGFRα gene were restricted to exon 12. Despite different breakpoints, all FIP1L1-PDGFRα deletions were in frame.
FIP1L1-PDGFRαfusion gene reflected disease progression in CEL
Patients with FIP1L1-PDGFRα-associated CEL were all male patients (P=0.046, Table 1). These patients presented with a high incidence of hepato-splenomegaly and cardiac or pulmonary involvement (P <0.001, P=0.032, and P=0.043, respectively). Among the biological parameters, FIP1L1-PDGFRα fusion was significantly related to decreased levels of hemoglobulin and platelets, as well as the presence of bone marrow fibrosis (P=0.023, P=0.001 and P=0.013, respectively). Of the 24 HES patients, the 3-year overall survival rate was (87.4±8.4)%, with median survival not reached at the time of last the follow-up. When treated with conventional therapy such as steroids and chemotherapeutic drugs a poor overall survival rate was correlated with CEL. The 3-year overall survival rate for CEL patients (n=5) was (50.0±25.0)%, significantly shorter than eosinophilia patients without the fusion (n=16, 100%, P=0.0016).
FIP1L1-PDGFRα-positive CEL patients responded efficiently to imatinib
Insensitive to conventional therapy, three patients positive for FIP1L1-PDGFRα received imatinib. As described in Figure 3A, case 20 was initially treated with imatinib 400 mg daily. Peripheral eosinophil levels normalized after two weeks and imatinib, 100 mg daily, was administered as maintenance therapy. Complete hematological remission was also achieved in CEL case 24 after one-month of treatment with imatinib at 200 mg daily. The same dose of imatinib was continued and the eosinophil count remained stable. For CEL case 23, although disease was first controlled by imatinib at 200 mg daily, the patient relapsed after stopping imatinib. In case 23 FIP1L1-PDGFRα was positive at diagnosis, undetectable after imatinib treatment and reappeared after interruption of imatinib in (Figure 3B). All the three CEL patients were alive at the follow-up of 10 to 33 months.
In 2/3 CEL cases, analysis of the bone marrow aspirate and biopsy specimen obtained at diagnosis showed eosinophilia with large hypolobated eosinophils and myelofibrosis, accompanied by increased eosinophils in the blood smear (Figure 3Ci–iii, representative results of case 24). After three weeks of imatinib therapy, there was an obvious reversal of bone marrow and blood pathology, including the restoration of normal hematopoiesis (Figure 3Di), reversal of myelofibrosis (Figure 3Dii) and reduction of peripheral eosinophils (Figure 3Diii).
Other gene abnormalities involved in CEL
Two additional mutations were found in two CEL patients which reflected the aggressive disease behavior. No mutations were detected in any of the 367 healthy volunteers.
One deletion (2654_2753del) of the tyrosine kinase domain of FLT3 gene was identified in CEL case 4, inducing a premature stop codon (G885fsX888).12 This patient presented with hepato-splenomegaly, anemia and thrombocytopenia, skin lesions, cardiac involvement and peripheral neuropathy. Blood and bone marrow examination revealed a high eosinophil percentage (57.0% and 50.5%, respectively) and myelofibrosis. Not responding to prednisone, interferon-α and chemotherapy the patient died 45 months after diagnosis.
Another deletion (3056_3167del) of the tyrosine kinase domain of the PDGFRβ gene was detected in CEL case 20, producing a premature stop codon (T863fsX872).12 Although presenting with multi-organ involvement and nonresponsive to conventional therapy, the patient was treated by imatinib and remained in complete remission.
Interestingly, two SNPs (2587-110C>T; 2698+200A>G) flanking exon 19 (2587–2698 bp) of PDGFRβ gene existed in this CEL patient.12 (2587-110C>T) was derived from the mother and (2698+200A>G) from the father of the patient. Both SNPs rates were rather low in healthy volunteers (11/367 (3.0%) and 1/302 (0.3%), respectively) and no co-existence of these two SNPs was found. To further investigate its role on selective splicing of the PDGFRβ gene, pCMV4-PDGFRβ DNA-WT and pCMV4-PDGFRβ DNA-MUT were constructed and transfected to COS7 and HEK293 cells. Two transcripts were presented in cells transfected with pCMV4-PDGFRβ DNA-MUT, while only one transcript was found in those transfected with pCMV4-PDGFRβ DNA-WT (Figure 4). So, we demonstrated that the two SNPs were associated with the selective splicing of exon 19 in PDGFRβ gene.
No mutation of the tyrosine-kinase domains of PDGFRα, C-KIT, FGFR1, ABL, MPL515 or JAK2V617F mutation was found in any patients examined.
However, after direct sequencing of the FGFR1 gene we found a new variation, described as 1014_1019del AACAGT for a nucleotide change, resulting in T339_V340del at the protein level. Thirteen of 18 (72%) eosinophilia patients were identified as 1014_1019del AACAGT heterozygotes, which was similar to healthy volunteers (19/23, 82%). And no difference was found between patients who were positive and negative for the FIP1L1-PDGFRα fusion (4/7 vs 9/11, P=0.083)
Constitutive activation of tyrosine kinases is a key element in the pathogenesis of myeloproliferative diseases, including CEL. In our series, by RT-PCR and FISH analysis, we demonstrated a FIP1L1-PDGFRα fusion gene, a clonal marker for CEL in eight out of 24 HES patients (33.3%), which is comparable to the investigations from Western countries.2–4,13,14 The breaking points were scattered between introns 7 and 10, as previously reported.4,7 These fusions were not observed in patients with reactive eosinophila, indicating that they are restricted to CEL.
The mechanism of tyrosine kinase activation in CEL is not currently well understood. However, it would be predicted, based on analysis of structure-function in all known fusion tyrosine kinases, that FIP1L1 would contribute a promoter region that serves to constitutively activate the PDGFRα kinase. On the other hand, PDGFRα breakpoints invariably involve exon 12, which encodes a portion of the juxtamembrane domain known to have an auto-inhibitory function. Therefore, disruption of PDGFRα exon 12, in combination with unregulated expression from the FIP1L1 promoter, is the pivotal events required for the increased tyrosine kinase activity. Similar to other fusion tyrosine kinases, FIP1L1-PDGFRα transforms hematopoietic cells both in vitro and in vivo. The expression of FIP1L1-PDGFRα induces the murine hematopoietic cell line Ba/F3 to interleukin-3-independent growth.4 Transplantation of FLP1L1-PDGFRα- expressing IL-5-transgenic hematopoietic stem cells into syngeneic recipients causes a murine model of tissue-infiltrating hypereosinophilia resembling human CEL.15 These results all support FIP1L1-PDGFRα fusion being involved in the pathogenesis of CEL.
The clinical presentations of HES are extremely varied, ranging from asymptomatic eosinophilia to life-threatening organ damage. Frequently occurring in CEL, FIP1L1-PDGFRα fusion translates into differences in clinical behavior and disease progression in CEL patients. CEL patients appear to have a more severe disease phenotype involving extensive end-organ pathology, including hepato-splenomegaly, cardiac and pulmonary involvement. Moreover, those patients suffered from anemia, thrombocytopenia, as well as myelofibrosis, suggesting that normal bone marrow function was more profoundly disturbed.
Although recognized as a more aggressive form of HES, CEL patients positive for FIP1L1-PDGFRα demonstrated a dramatic response to imatinib. In the present study, complete hematological remission under imatinib was observed in three CEL patients treated, confirming previous results. Of note, its disappearance after remission confirmed that FIP1L1-PDGFRα acquired within the malignant clones, thus molecular monitoring of residual disease is important. Collectively, detection of the FIP1L1-PDGFRα fusion gene in CEL can be used prospectively to identify patients who are expected to respond to imatinib treatment, to molecularly monitor their response and to monitor for evidence of relapse.
PDGF mediates strong mitogenic signals via the PDGF receptors on fibroblasts and is responsible for the development of bone marrow fibrosis.16 Recently, it has been shown that imatinib has an anti-fibrogenic effect on myelofibrosis in chronic myelogenous leukemia.17 This could also explain why targeting FIP1L1-PDGFRα fusion by imatinib also reverse the myelofibrosis in parallel to molecular remission in CEL.
Mutations of other tyrosine kinase family genes were also found in addition to the FIP1L1-PDGFRα fusion. These mutations were not detected in 367 healthy volunteers. The FLT3 mutation within the tyrosine kinase region (case 4) made it independent to ligand interaction, resulting in constitutive FLT3 activition.18 The C-terminal tail of the PDGFRβ gene contains a proline- and glutamic acid-rich motif, which is identified as a negative regulator of receptor kinase activity.19 The mutation we found in case 20 could possibly cause deletion of this region and subsequent activation of the PDGFRβ gene. Clinically, disease in these two patients with dual abnormalities showed aggressive behavior, indicating that other tyrosine kinase activations could act concomitantly with the FIP1L1-PDGFRα fusion on disease progression. However, these abnormalities may be overcome by imatinib, as in case 20.
In conclusion, either alone or co-existing with other mutations of tyrosine kinase family genes, the FIP1L1-PDGFRα fusion gene reflects disease progression of CEL but predicts a good response to imatinib. FIP1L1-PDGFRα fusion gene should be considered as a biological marker of disease diagnosis and prognosis, as well as a means to follow therapy in the CEL subgroup of patients.
1. Klion AD. Recent advances in the diagnosis and treatment of hypereosinophilic syndromes. Hematology Am Soc Hematol Educ Program 2005: 209–214.
2. Klion AD, Noel P, Akin C, Law MA, Gilliland DG, Cools J, et al. Elevated serum tryptase levels identify a subset of patients with a myeloproliferative variant of idiopathic hypereosinophilic syndrome associated with tissue fibrosis, poor prognosis, and imatinib responsiveness. Blood 2003; 101: 4660–4666.
3. Roche-Lestienne C, Lepers S, Soenen-Cornu V, Kahn JE, Lai JL, Hachulla E, et al. Molecular characterization of the idiopathic hypereosinophilic syndrome (HES) in 35 French patients with normal conventional cytogenetics. Leukemia 2005; 19: 792–798.
4. Cools J, DeAngelo DJ, Gotlib J, Stover EH, Legare RD, Cortes J, et al. A tyrosine kinase created by fusion of the PDGFRA and FIP1L1 genes as a therapeutic target of imatinib in idiopathic hypereosinophilic syndrome. N Engl J Med 2003; 348: 1201–1214.
5. Goldman JM, Melo JV. Chronic myeloid leukemia — advances in biology and new approaches to treatment. N Engl J Med 2003; 349: 1451–1464.
6. Cools J, Stover EH, Wlodarska I, Marynen P, Gilliland DG. The FIP1L1-PDGFRalpha kinase in hypereosinophilic syndrome and chronic eosinophilic leukemia. Curr Opin Hematol 2004; 11: 51–57.
7. Gotlib J, Cools J, Malone JM 3rd, Schrier SL, Gilliland DG, Coutre SE. The FIP1L1-PDGFRalpha fusion tyrosine kinase in hypereosinophilic syndrome and chronic eosinophilic leukemia: implications for diagnosis, classification, and management. Blood 2004; 103: 2879–2891.
8. Chusid MJ, Dale DC, West BC, Wolff SM. The hypereosinophilic syndrome: analysis of fourteen cases with review of the literature. Medicine (Baltimore) 1975; 54: 1–27.
9. Shaffer LG., Tommerup N. ISCN 2005: An international system for human cytogenetic nomenclature (Cytogenetic & Genome Research). Basel: S Karger Pub; 2005: 1–130.
10. Klion AD, Robyn J, Akin C, Noel P, Brown M, Law M, et al. Molecular remission and reversal of myelofibrosis in response to imatinib mesylate treatment in patients with the myeloproliferative variant of hypereosinophilic syndrome. Blood 2004; 103: 473–478.
11. Miyaso H, Okumura M, Kondo S, Higashide S, Miyajima H, Imaizumi K. An intronic splicing enhancer element in survival motor neuron (SMN) pre-mRNA. J Biol Chem 2003; 278: 15825–15831.
12. den Dunnen JT, Antonarakis SE. Nomenclature for the description of human sequence variations. Hum Genet 2001; 109: 121–124.
13. Pardanani A, Brockman SR, Paternoster SF, Flynn HC, Ketterling RP, Lasho TL, et al. FIP1L1-PDGFRA fusion: prevalence and clinicopathologic correlates in 89 consecutive patients with moderate to severe eosinophilia. Blood 2004; 104: 3038–3045.
14. Vandenberghe P, Wlodarska I, Michaux L, Zachee P, Boogaerts M, Vanstraelen D, et al. Clinical and molecular features of FIP1L1-PDFGRA (+) chronic eosinophilic leukemias. Leukemia 2004; 18: 734–742.
15. Yamada Y, Rothenberg ME, Lee AW, Akei HS, Brandt EB, Williams DA, et al. The FIP1L1-PDGFRA fusion gene cooperates with IL-5 to induce murine hypereosinophilic syndrome (HES)/chronic eosinophilic leukemia (CEL)-like disease. Blood 2006; 107: 4071–4079.
16. Bock O, Loch G, Busche G, von Wasielewski R, Schlue J, Kreipe H. Aberrant expression of platelet-derived growth factor (PDGF) and PDGF receptor-alpha is associated with advanced bone marrow fibrosis in idiopathic myelofibrosis. Haematologica 2005; 90: 133–134.
17. Hasselbalch HC, Bjerrum OW, Jensen BA, Clausen NT, Hansen PB, Birgens H, et al. Imatinib mesylate in idiopathic and postpolycythemic myelofibrosis. Am J Hematol 2003; 74: 238–242.
18. Gilliland DG, Griffin JD. The roles of FLT3 in hematopoiesis and leukemia. Blood 2002; 100: 1532–1542.
19. Chiara F, Bishayee S, Heldin CH, Demoulin JB. Autoinhibition of the platelet-derived growth factor beta-receptor tyrosine kinase by its C-terminal tail. J Biol Chem 2004; 279: 19732–19738.