|
|||||||||
|
|
|
|
|
|
|
|
|
|
|
Blood, 15 April 2004, Vol. 103, No. 8, pp. 2879-2891. Prepublished online as a Blood First Edition Paper on November 20, 2003; DOI 10.1182/blood-2003-06-1824.
REVIEW ARTICLES
The FIP1L1-PDGFR
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
Abstract |
|---|
gene (PDGFRA) to an uncharacterized human gene FIP1-like-1 (FIP1L1). However, not all HES and CEL patients respond to imatinib, suggesting disease heterogeneity. Furthermore, approximately 40% of responding patients lack the FIP1L1-PDGFRA fusion, suggesting genetic heterogeneity. This review examines the current state of knowledge of HES and CEL and the implications of the FIP1L1-PDGFRA discovery on their diagnosis, classification, and management. (Blood. 2004;103:2879-2891) |
Introduction |
|---|
(FIP1L1-PDGFRA) in cases of HES/CEL adds to a growing list of activated fusion tyrosine kinases linked to the pathogenesis of chronic myeloproliferative disorders.3 It is unique, however, because it is the first description of a gain-of-function fusion protein resulting from a cryptic interstitial deletion between genes rather than a reciprocal chromosomal translocation. The FIP1L1-PDGFR fusion protein transforms hematopoietic cells, and its kinase activity is inhibited by imatinib at a cellular 50% inhibitory concentration (IC50) 100-fold lower than BCR-ABL.3 Acquisition of an imatinib resistance mutation in the adenosine triphosphate (ATP)–binding domain of PDGFRA in a relapsed patient previously responsive to imatinib supports a critical role for FIP1L1-PDGFR in the pathogenesis of disease and demonstrates that FIP1L1-PDGFR is the therapeutic target of imatinib.3 The identification of this novel molecular target in HES and CEL patients will help refine genotype-phenotype correlations in these diseases and should aid basic research of the biologic pathways involved in eosinophil proliferation, differentiation, and signaling. |
Epidemiology |
|---|
|
Current classification |
|---|
|
One method for establishing a diagnosis of CEL is demonstration of eosinophil clonality; however, this is frequently not assessed or is difficult to confirm. Methods for demonstrating clonality include fluorescence in situ hybridization11 or cytogenetic analysis of purified eosinophils12 and, also, X chromosome inactivation analysis in women.13,14 X inactivation–based assessment of clonality is of limited value in HES because most patients are male. To avoid the emphasis placed on demonstrating clonality of eosinophils, Brito-Babapulle has advocated that, in cases of clonal eosinophilia, it is sufficient to demonstrate only that eosinophils are part of a clonal bone marrow disorder and not necessarily part of the malignant clone, with treatment tailored to the underlying disease.15 In this scheme, blood eosinophilia is divided into 3 categories: reactive (nonclonal eosinophilia), clonal disorders of the bone marrow associated with eosinophilia, and HES, which remains a diagnosis of exclusion.15
Prior reviews have discussed the difficulty in using abnormal eosinophil morphology (eg, cytoplasmic hypogranularity or vacuolization, abnormal lobation, ring nuclei) to reliably distinguish reactive from clonal eosinophilia because these cytologic changes may be present in both conditions.15-17 Roufosse et al have proposed that disease presentations with CML-like features be segregated into a "myeloproliferative variant" of HES.18 Clinical and laboratory features associated with this variant include hepatomegaly, splenomegaly, anemia, thrombocytopenia, bone marrow dysplasia or fibrosis, and elevated levels of cobalamin. This is in contrast to a "lymphocytic variant" or T-cell–mediated HES (discussed in "T-cell–mediated HES"), which typically follows a more benign course and is usually manifested primarily by skin disease.18 Some patients exhibit overlapping characteristics of both variants. Although such groupings may correspond to biologic subsets of HES, evaluation of a large cohort of patients is required to validate their prognostic relevance.
|
Clinicopathologic manifestations |
|---|
|
Hematologic findings
Although persistent eosinophilia without a clinically identifiable cause is the sine qua non of HES, the hematologic picture can vary. Relatively modest elevations in the leukocyte count (eg, 20-30 x 109/L [20 000-30 000/mm3]) with peripheral eosinophilia in the range of 30% to 70% are commonly observed,9 but significantly higher leukocyte counts have also been reported.19,20 Neutrophilia, basophilia, myeloid immaturity, and both mature and immature eosinophils with varying degrees of dysplasia may be found in the peripheral blood or bone marrow.19,22 In one series, anemia was present in 53% of patients, thrombocytopenia was more common than thrombocytosis (31% versus 16%), and bone marrow eosinophilia ranged from 7% to 57% (mean, 33%).22 Charcot-Leyden crystals were frequent marrow findings, whereas increased blasts and myelofibrosis were less often observed.22
Cardiac disease
The multistep process of cardiac injury illustrates some of the pathophysiologic mechanisms contributing to organ damage in HES (previously reviewed by Fauci et al4 and Weller and Bubley21). In the initial necrotic stage, cardiac disease may be initiated by eosinophil damage to the endocardium, with local platelet thrombus subsequently leading to formation of mural thrombi that have the potential to embolize (thrombotic stage). The contents of eosinophil granules, including major basic protein and eosinophilic cationic protein, may promote endothelial damage and hypercoagulablity, enhancing the thromboembolic risk.23,24 In the later fibrotic stage, organization of thrombus can lead to fibrous thickening of the endocardial lining and, ultimately, restrictive cardiomyopathy.4,21 Valvular insufficiency in HES is commonly related to mural endocardial thrombosis and fibrosis involving leaflets of the mitral or tricuspid valves.25-27 Table 2 lists manifestations of HES that have been reported in hematologic, cardiac, and other organ systems.
|
T-cell–mediated HES |
|---|
|
Cytogenetic and molecular features of hematologic malignancies with eosinophilia |
|---|
Over the last 3 decades, the list of chromosomal abnormalities in cases reported as HES or CEL has grown (reviewed by Bain102). However, a unique clonal karyotype has not been associated with these diseases, and most patients exhibit a normal karyotype by conventional cytogenetics. Although trisomy 8 is frequently detected in these eosinophilic disorders,103-106 it is also observed in other hematologic malignancies. Three HES case reports have described balanced reciprocal translocations within or near the chromosome 4q12 locus of the PDGFRA and KIT tyrosine kinases: t(3;4)(p13;q12),107 t(4;7)(q11;q32),108 and t(4;7)(q11;p13).109 The genes involved in these translocations were not identified. In a recent report, a 6-year-old girl with hypereosinophilia presented with a t(5;9)(q11;q34) constitutional translocation, involving genes for the ABL tyrosine kinase and possibly granzyme A on chromosome 5.7 In this case, no mention was made of the use of imatinib.
Eosinophils have been found to be part of the malignant clone in systemic mastocytosis,110 CML and other chronic myeloproliferative disorders (MPDs), and in specific subtypes of acute myeloid leukemia (AML). The best-characterized examples in the French-American-British classification of AML are M4Eo inv(16)(p13q22) or t(16;16)(p13;q22),111 resulting in chimeric fusion of the CBF
and MYH11 genes, and M2 t(8;21)(q22;q22),112 which links the AML-1 and ETO genes. Other abnormal karyotypes reported in AML with eosinophilia include monosomy 7,113 trisomy 1,114 t(10;11)(p14;q21),115 t(5;16)(q33;q22),116 and t(16;21)(p11;q22).117 The latter 2 may represent chromosome 16 variants with an underlying cryptic fusion gene. Eosinophil clonality has been demonstrated in cases of eosinophilic myelodysplastic syndrome (MDS) with t(1;7) or dic(1;7) karyotypes.11,118 Eosinophilia is also a feature of acute and chronic hematologic malignancies with rearrangements involving transcription factor ETV6 (ETS translocation variant 6, TEL) on chromosome 12p13. Examples include the ETV6-ABL fusion in t(9;12)(q34;p13) AML119 and, also, the small subset of chronic myelomonocytic leukemia patients with t(5;12)(q33;p13), which fuses platelet-derived growth factor receptor-beta (PDGFR) on chromosome 5q33 to ETV6.120 In this latter disease, imatinib produces clinical remissions by inhibiting the deregulated activity of the fusion tyrosine kinase.121 Proliferation of eosinophils in some AML, MPD, and MDS cases is associated with rearrangements involving the long arm of chromosome 5 (eg, 5q31-33) where several genes encoding eosinophilic cytokines reside.122-124 In a study of 9 patients with MPD or mixed MDS/MPD and a translocation involving 5q31-33, fluorescence in situ hybridization (FISH) unmasked disruption of the PDGFRB gene in 6 cases.125 The translocations included t(1;5)(q21;q33), t(1;5)(q22;q31), t(1;3;5)(p36;p21;q33), t(2;12;5)(q37;q22;q33), t(3;5)(p21;q31), and t(5;14)(q33;q24). Eosinophilia was noted in 3 of these patients and noted in an additional case at the time of transformation to AML.125 Recent cloning of the t(1;5)(q23;q33) breakpoint revealed that PDGFRB is fused to the novel partner protein myomegalin.126 In a subset of B-cell acute lymphoblastic leukemias, translocation of the IL-3 gene on chromosome 5q31 to the immunoglobulin heavy chain gene on chromosome 14q32 is typically associated with eosinophilia.127 In the "stem cell" myeloproliferative disorders, mutation in a pluripotent hematopoietic progenitor results in a spectrum of diseases including T- or B-cell lymphoblastic lymphoma, bone marrow myeloid hyperplasia, and eosinophilia. These poor-prognosis disorders are related to recurrent breakpoints on chromosome 8p11 that involve translocation of the fibroblast growth factor receptor 1 (FGFR1) gene to 5 currently identified partner loci: FOP at 6q27,128 CEP110 at 9q33,129 FIM/ZNF198 at 13q12,130 BCR at 22q11,131 and the human endogenous retrovirus gene (HERV-K) at 19q13.132 Before attributing eosinophilia to HES or CEL, these various hematologic malignancies should be given diagnostic consideration.
|
Prognosis |
|---|
|
Current treatment |
|---|
(IFN-) can elicit long-term hematologic and cytogenetic responses in HES and CEL patients resistant to other therapies, including prednisone and hydroxyurea.109,140-145 Some have advocated its use as initial therapy for these diseases.144 Remissions have been associated with improvement in clinical symptoms and organ disease, including hepatosplenomegaly,140,144 cardiac and thromboembolic complications,109,141 mucosal ulcers,143 and skin involvement.145 IFN- exerts pleiotropic effects including inhibition of eosinophil proliferation and differentiation.146 Inhibition of IL-5 synthesis from CD4+ helper T cells may be relevant to its mode of action in T-cell–mediated forms of HES.147 IFN- may also act more directly via IFN- receptors on eosinophils, suppressing release of mediators including cationic protein, neurotoxin, and interleukin-5.148 Responses to cyclosporin A149,150 and 2-chlorodeoxyadenosine have also been reported in HES.151 Bone marrow/peripheral blood stem cell allogeneic transplantation has been attempted in patients with aggressive disease. Disease-free survival ranging from 8 months to 5 years has been reported152-156 with one patient relapsing at 40 months.157 Allogeneic transplantation using nonmyeloablative conditioning regimens has been reported in 3 patients, with remission duration of 3 to 12 months at the time of last reported follow-up.158,159 Despite success in selected cases, the role of transplantation in HES is not well established. Transplantation-related complications including acute and chronic graft versus host disease as well as serious infections have been frequently observed.160,161
Advances in cardiac surgery have extended the life of patients with late-stage heart disease manifested by endomyocardial fibrosis, mural thrombosis, and valvular insufficiency.4,21 Mitral and/or tricuspid valve repair or replacement26,35-37,162 and endomyocardectomy for late-stage fibrotic heart disease37,163 can improve cardiac function. Bioprosthetic devices are preferred over their mechanical counterparts because of the reduced frequency of valve thrombosis.
Leukapheresis can elicit transient reductions in high eosinophil counts but is not an effective maintenance therapy.164-166 Similar to other myeloproliferative disorders, splenectomy has been performed for hypersplenism-related abdominal pain and splenic infarction but is not considered a mainstay of treatment.19,167 Anticoagulants and antiplatelet agents have shown variable success in preventing recurrent thromboembolism.19,47,50,168
|
Imatinib in HES |
|---|
. He was treated with imatinib based on the drug's efficacy in CML, with the hypothesis that the 2 diseases may share a common pathogenetic mechanism. The patient achieved a rapid and complete hematologic remission after taking 100 mg imatinib daily for 4 days. Complete disappearance of peripheral eosinophils occurred by day 35. Imatinib was decreased to 75 mg daily for headaches, which still effectively controlled eosinophil levels.
|
A subsequent report included 5 HES patients treated with 100 mg imatinib daily.170 Four male patients with normal serum levels of interleukin-5 achieved complete hematologic remissions. One female patient with high levels of serum interleukin-5 did not respond to imatinib. All patients who responded were able to discontinue other treatments.
A third report described a 54-year-old man with HES and organ involvement including splenomegaly, skin, cardiac, and central nervous system disease.167 He was resistant to steroids and chemotherapy, including 2-chlorodeoxyadenosine and cytosine arabinoside. Before treatment, the WBC count was 9.7 x 109/L (9700/mm3) with 68% eosinophils. After 18 days of imatinib (100 mg daily), the patient achieved a complete hematologic remission with a WBC count of 3.9 x 109/L (3900/mm3) and 0% eosinophils. His hematologic response was accompanied by marked symptomatic improvement.
Another study reported the efficacy of 100 to 400 mg imatinib daily in 5 HES patients and 2 patients with a diagnosis reported as eosinophilia-associated chronic myeloproliferative disorder (eos-CMD).171 At a median follow-up of 17 weeks, 1 eos-CMD and 2 HES patients achieved complete clinical remissions, and an additional HES patient achieved a partial remission. Screening for known targets of imatinib, including BCR-ABL, or mutations in the coding exons of KIT and PDGFRB was negative. In contrast to the earlier report where complete remitters had normal serum interleukin-5 levels, the current group of responders had high serum levels. These disparate findings demonstrate that levels of this eosinophil-stimulating cytokine are not necessarily predictive of imatinib responsiveness in HES patients. Although imatinib was generally well tolerated, one responding HES patient experienced cardiogenic shock within the first week of treatment with a marked decrease in the left ventricular (LV) ejection fraction. Endomyocardial biopsy revealed eosinophilic myocarditis with evidence of eosinophil infiltration, degranulation, and myocyte damage. The patient was successfully treated with high-dose corticosteroids. LV function recovered, and the patient was restarted on imatinib and achieved a hematologic remission. More data are needed to evaluate the role of prophylactic steroids in HES patients with cardiac disease who receive imatinib treatment.
An additional cohort of 9 symptomatic HES patients (6 male; median age, 50) was treated with imatinib starting at 100 mg daily.172 They had received an average of 3 prior therapies. With median follow-up of 13 weeks, 4 male patients achieved a complete remission. Three exhibited response within the first 2 weeks of therapy, while the fourth required a dose increase to 400 mg daily at day 28 to attain a normal eosinophil count. Overall, treatment was well tolerated, with primarily grade 1 toxicities previously associated with imatinib reported.
The largest study described 16 patients, including 11 treated with imatinib.3 At presentation, the median eosinophil count was 14.5 x 109/L (14500/mm3) (range, 4.96 x 109/L to 53 x 109/L [4960/mm3 to 53000/mm3]). Nine of the treated patients had normal karyotypes. One CEL patient had a clonal cytogenetic abnormality t(1;4)(q44;q12), and one patient with AML arising from CEL had a complex karyotype, including trisomies 8 and 19, add2q, and del6q. Hematologic responses were observed in 10 of 11 HES patients treated with imatinib at doses of 100 to 400 mg daily. The median time to response was 4 weeks (range, 1 to 12 weeks). Nine of the 10 patients demonstrated sustained hematologic responses (lasting at least 3 months), with a median duration of 7 months at the time of publication (range, 3 to 15 months). One patient had a transient response lasting several weeks and failed to derive benefit from an increase in the imatinib dose. Figure 1 shows a bone marrow biopsy from the CEL/AML patient with the complex karyotype before imatinib treatment and at the time of hematologic remission following 3 months of therapy.
|
The molecular basis for response in most patients was inhibition of a novel fusion tyrosine kinase, FIP1L1-PDGFR, in which a newly described human gene, FIP1-like-1 (FIP1L1), is fused to the gene encoding platelet-derived growth factor receptor- (PDGFRA).3 The FIP1L1 gene encodes a protein that is homologous to a previously characterized Saccharomyces cerevisiae protein, Fip1, a synthetic lethal component of the mRNA polyadenylation apparatus.173 The fusion gene is generated by an interstitial deletion on chromosome 4q12 rather than a reciprocal translocation.3 FIP1L1-PDGFRA was present in 9 (56%) of 16 HES patients. In patients for whom FIP1L1-PDGFRA testing was performed, the fusion was detected with a similar frequency in treated (5 of 10) and untreated (4 of 6) patients.3 All 5 patients with the FIP1L1-PDGFRA fusion responded to imatinib.3 However, an additional 4 patients with durable responses to imatinib lacked FIP1L1-PDGFRA, indicating that an as yet unidentified target of imatinib is responsible for HES in these cases.3 To date, no primary treatment failures to imatinib have been reported in patients with the FIP1L1-PDGFRA fusion.
The FIP1L1-PDGFRA genotype may cosegregate with a clinical phenotype including myeloproliferative-like HES (HES-MPD), tissue fibrosis, and increased serum tryptase levels.174,175 The FIP1L1-PDGFRA fusion was identified in all 7 HES-MPD patients with elevated serum tryptase levels (all were treated with imatinib and responded)175; in an earlier companion study, the fusion was not detected in 4 HES patients with normal serum tryptase or 2 patients with familial eosinophilia. FIP1L1-PDGFRA may also be related to the pathogenesis of eosinophilic subsets of systemic mastocytosis (SM). Deletion of the CHIC2 locus, a surrogate for the FIP1L1-PDGFRA fusion, was detected in imatinib-responsive patients diagnosed with systemic mastocytosis (SM) and eosinophilia but not in 2 other patients with SM and the KIT Asp816Val (D816V) mutation who exhibited no response to imatinib.176
Currently, limited data are available regarding imatinib's ability to reverse eosinophil-related organ damage. Among 3 HES-MPD patients with endomyocardial fibrosis and congestive heart failure, there was no improvement in cardiac disease despite complete hematologic responses to imatinib.175 However, significant improvement of respiratory symptoms associated with clearing of interstitial infiltrates on chest computed tomography (CT)171,175 and normalization of pulmonary function testing have been reported.175 We and others have demonstrated reversal of myelofibrosis (Figure 1).175
|
Molecular biology of FIP1L1-PDGFRA |
|---|
|
|
FIP1L1-PDGFR has also been discovered in the cell line EOL-1, derived from a patient with acute eosinophilic leukemia following hypereosinophilic syndrome. Imatinib and 2 other inhibitors of PDGFR, vatalanib and THRX-165724, reduced the viability of EOL-1 cells and a prominent 110-kDa phosphoprotein, ultimately identified as FIP1L1-PDGFR.177
FIP1L1 is a 520–amino acid protein that contains a region of homology to Fip1, a yeast protein with synthetic lethal function that is involved in polyadenylation.174,178 Similar proteins are found in plants, worm, fly, rat, and mouse. All share the well-conserved 42–amino acid "Fip1" motif (pfam domain no. PF05182; http://pfam.wustl.edu/cgi-bin/getdesc?acc=PF05182), which is also present in the FIP1L1-PDGFR fusion protein. The exact function of the human or mouse FIP1L1 protein is not known. Based on the abundance of FIP1L1 expressed sequence tags (ESTs) in the databases that are derived from different tissues and cell types, FIP1L1 is predicted to be under the control of a ubiquitous promoter.
Similar to other fusion tyrosine kinases, FIP1L1-PDGFR is a constitutively active tyrosine kinase that transforms hematopoietic cells in vitro and in vivo.3,179 FIP1L1-PDGFR phosphorylates itself
fusion tyrosine kinase in hypereosinophilic syndrome and chronic eosinophilic leukemia: implications for diagnosis, classification, and management
Top
Abstract
Introduction
