| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Published online before print May 3, 2007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Consultations in Molecular Diagnostics |
From the Departments of Human Genetics
* and Medical Chemistry,
Groupe Interdisciplinaire de Genoproteomique Appliquee, Centre Hospitalier Universitaire de Liège, University of Liège, Liège; and the Department of Human Genetics,
Free University of Brussels, Brussels, Belgium
Abstract
The idiopathic hypereosinophilic syndrome (HES) has remained for a long time a diagnosis of exclusion. Differential diagnosis between the HES and the related chronic eosinophilic leukemia (CEL) relied on the identification of signs of clonality that allowed, when present, the reclassification of patients as CEL. Recently, a new acquired mutation was described in approximately 50% of the HES/CEL patients: a cryptic deletion on chromosome band 4q12 generating a FIP1L1-PDGFRA fusion gene. According to the World Health Organization classification, this clonal abnormality has been proposed as a new surrogate marker for chronic eosinophilic leukemia diagnosis. Fluorescence in situ hybridization and reverse transcriptase-polymerase chain reaction protocols were developed for an accurate del(4)(q12q12) and FIP1L1-PDGFRA fusion gene detection. Here, we report a patient with a rare FIP1L1 intron 16 breakpoint located outside of the reported FIP1L1 breakpoint region (ie, from FIP1L1 introns 9 to 13). This case illustrates the risk of false-negative results with diagnostic procedures that do not take into account the occurrence of rare FIP1L1 breakpoints. As targeted therapy with tyrosine kinase inhibitors has dramatically changed the prognosis of FIP1L1-PDGFRA (+) CEL, false-negative results could hamper accurate diagnosis and treatment.
The idiopathic hypereosinophilic syndrome (HES) represents a heterogeneous group of diseases characterized by sustained hypereosinophilia (over 1500 eosinophils/µl), lasting for more than 6 months, and associated with end organ damages in the absence of any identifiable reactive conditions.1
According to the World Health Organization guidelines, the identification of a clonal origin of the disease allows the reclassification of these cases as chronic eosinophilic leukemia (CEL).1
Recently, empirical trials of imatinib mesylate (Gleevec; Novartis, Basel, Switzerland) targeted therapy allowed the discovery of an 800-kb submicroscopic deletion on chromosome 4q12 in some responsive HES/CEL patients.2
The cysteine-rich hydrophobic domain 2 (CHIC2) deletion (delCHIC2), a surrogate marker for the del(4)(q12q12), and the FIP1-like-1 (FIP1L1)-platelet-derived growth factor receptor 
(PDGFRA) fusion gene, resulting from the subsequent fusion of the 5' part of FIP1L1 to the 3' part of PDGFRA, represent powerful clonality markers for the direct diagnosis of CEL.3, 4, 5
The FIP1L1-PDGFR transforming properties have been described in vitro as well as in vivo.2, 6
As a consequence, powerful tools are mandatory for an accurate detection of these abnormalities, which offer new opportunities in HES/CEL differential diagnosis and therapy.3, 5
The del(4)(q12q12) cannot be detected by conventional cytogenetics, and consequently, the majority of HES/CEL patients harbor a normal karyotype.2, 5
The diagnosis therefore relies on fluorescence in situ hybridization (FISH) and on reverse transcriptase-polymerase chain reaction (RT-PCR) for the detection of the del(4)(q12q12) and the FIP1L1-PDGFRA fusion gene, respectively.2, 4, 7, 8
Both tools are sensitive, but false-negative results can occur. For these reasons, a combination of both FISH and RT-PCR has been advised for appropriate diagnosis.9
Previously reported patients presented a breakpoint invariably located in exon 12 of PDGFRA, whereas FIP1L1 breakpoints scattered over a 40-kb region encompassing introns 7 to 10, as numbered in Cools et al2, 5 nomenclature, or introns 8 to 13, according to the LeukemiaNet exon-intron numbering.2, 5, 10, 11 Here, we report a CEL patient with a breakpoint located in intron 16 of FIP1L1, thus outside of the reported FIP1L1 breakpoint cluster region. This case illustrates the sensitivity of the previously published RT-PCR technique and highlights the risk of false-negative results with real-time quantitative (RQ) RT-PCR approaches that do not take into account rare FIP1L1 breakpoints downstream of FIP1L1 exon 13.10
Case Reports
A 41-year-old man was referred for primary hypereosinophilia differential diagnosis. Secondary etiologies associated with reactive hypereosinophilia were extensively excluded. A peripheral blood cell count disclosed hyperleukocytosis with an absolute eosinophil count of 27.3 x 109/L. Serum vitamin B12 was >2000 pg/ml, and tryptase was >75 µg/L (N <11.5 µg/L). A bone marrow aspiration disclosed 38% of eosinophilic granulocytes, and trephine biopsy showed hypercellularity with marked granulocytic hyperplasia and eosinophilia. Conventional cytogenetic analysis revealed a 46,XY karyotype. However, combining a FISH analysis as described by Pardanani et al4
and the RT-PCR procedure described by Cools et al,2
a del(4)(q12q12) and a FIP1L1-PDGFRA fusion gene were demonstrated, respectively (Figure 1)
. According to the World Health Organization guidelines, a diagnosis of CEL was therefore established.1
A specific treatment with the tyrosine kinase inhibitor imatinib mesylate was initiated and led to a dramatic improvement of the symptoms and a rapid normalization of the white blood cell count (see below).
|
FISH Analysis
FISH assay was performed on fresh peripheral blood and fixed bone marrow cells. Deletion of the 4q12 region was assessed by a two-color FISH strategy. The CHIC2 locus is located telomeric to FIP1L1 but centromeric to PDGFRA and thus represents a surrogate marker for the FIP1L1-PDGFRA fusion gene, resulting from the 800-kb del(4)(q12q12) involving the region between FIP1L1 and PDGFRA. The CHIC2 status was evaluated both on metaphase and interphase nuclei. We used two overlapped bacterial artificial chromosomes (BACs), clones 89B16 (accession no. AC105384) and 98G22 (accession no. AQ317591), labeled with Spectrum Green that mapped the CHIC2 locus and a Spectrum Orange (Vysis, Downers Grove, IL) labeled probe, BAC clone 345F18 (accession no. AC110611) that mapped a sequence between the PDGFRA and the KIT loci. Glycerol stocks of the BAC clones were obtained from P. Dejong’s library (http://bacpac.chori.org/). The map position of each BAC clone was determined according to the University of California Santa Cruz Human Genome Project (http://genome.ucsc.edu). Nick translation labeling was performed using the Vysis Nick Translation Kit. Hybridization signals were visualized using an Olympus BX51 fluorescence microscope equipped with a cooled charge-coupled camera and were digitally recorded (Cytovision; Applied Imaging Ltd., Newcastle upon Tyne, UK). Hybridization of the three labeled probes on each chromosome results in two red-yellow-green fusion signal on normal 4q12 regions, whereas a chromosome 4 with a CHIC2 deletion would have lost its green signal.
Three thousand interphase nuclei from six normal individuals were scored in a blinded manner by three investigators for the presence or absence of the green (BACs 89B16 and 98G22) and red (BAC 345F18) hybridization signals. The normal cut-off value was evaluated from six healthy controls and 3000 nuclei observations, which yielded two signals in 96.3% of observed nuclei. The threshold value was established at 91.8% with a 95% confidence interval. At least 100 interphase nuclei of the studied patients were scored for the presence of the red and the absence of the green signals.
RT-PCR and Sequencing
RT-PCR detection of the FIP1L1-PDGFRA fusion was performed as described by Cools et al.2
In brief, total RNA was purified from blood and bone marrow using the RNeasy columns (Qiagen, Santa Clarita, CA). One µg of RNA was reverse transcribed using random hexamers and the Transcriptor First Strand cDNA Synthesis Kit (Roche, Mannheim, Germany) in a total volume of 20 µl, according to the manufacturer’s instructions. A first round of PCR amplification was performed starting from 1/10th of the cDNA in a total volume of 50 µl containing 15 mmol/L Tris-HCl, pH 8.0, 50 mmol/L KCl, 2.5 mmol/L MgCl2, 200 µmol/L dNTP, 20 pmol of each primer, and 2.5 U of AmpliTaq Gold polymerase (Applied Biosystems, Foster City, CA). The primer pairs were FIP1L1-F4 (5'-ACCTGGTGCTGATCTTTCTGAT-3') and PDGFRA-R1 (5'-TGAGAGCTTGTTTTTCACTGGA-3') for the first PCR. One µl of the first amplification was subsequently subjected to a second round of PCR with primers FIP1L1-F5 (5'-AAAGAGGATACGAATGGGACTTGTG-3') and PDGFRA-R2 (5'-GGGACCGGCTTAATCCATAG-3'). PCR conditions were as follows: 94°C, 10 minutes, followed by 35 cycles of 94°C, 1 minute; 61°C, 30 seconds; 72°C, 45 seconds, and a final elongation at 72°C for 10 minutes. The EOL-1 cell line (DSMZ, Berlin, Germany) bearing a FIP1L1-PDGFRA fusion gene was used as a positive control. The RNA quality as well as the cDNA synthesis efficacy were evaluated with an RQ RT-PCR amplifying the housekeeping gene ß-glucuronidase (GUS) as described.12
Amplified PCR products were purified using QIAquick PCR purification columns (Qiagen) and sequenced on both strands using the BigDye Terminator Sequencing protocol on an automated ABI 3100 sequencer. Sequences were analyzed with Sequence Analysis software V.5.1.1 (Applera, Norwalk, CT). After comparison with the published sequences (Ensembl software, http://www.ensembl.org/index.html), specific patient’s forward (FIP1L1S15: 5'-TGGACATTCCTCTGGTTATGA-3') and reverse (PDGFRAAS13: 5'-TGTTCCTTCAACCACCTTCC-5') primers were designed with the Primer3 freeware (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi). RNA (50-ng equivalent) was amplified using the following PCR conditions: 95°C, 10 minutes, followed by 35 cycles of 94°C, 30 seconds; 60°C, 30 seconds; 72°C, 45 seconds, and a final elongation at 72°C for 5 minutes.
For RQ RT-PCR, Primer Express 2.0 software (Applied Biosystems) was used to design a FIP1L1 exon 16 forward primer, FIP1L1E16T (5'-CCATCTTCCTGGTTCTGCTC-3'), a carboxyfluorescein dye-labeled TaqMan probe, P-PDGF12 (5'-TGACTCAAGATGGGAGTTTCCAAGAGATGG-3'), and a reverse primer, AS-PDGF12 (5'-CAAGACCCGACCAAGCACTAG-3'), located downstream to the PDGFRA exon 12 breakpoint. The GUS gene amplification was used as normalization control.12 One microgram of total RNA was isolated from blood and reverse transcribed as described for the qualitative RT-PCR. One-twentieth of the cDNA was then used for the RQ RT-PCR reaction on an ABI Prism 7000 Sequence Detection system (Applied Biosystems). PCR conditions and cycling parameters were in accordance with the default parameters of the ABI SDS 7000 apparatus (40 cycles of annealing/extension for 1 minute at 65°C). PCR products were detected using the TaqMan chemistry (Applied Biosystems).
Results
We report a case of HES with biochemical parameters compatible with a primary hematological process.8
Clonality of the eosinophils was assessed by molecular assays. A FISH analysis was first performed, with the aim of identifying the 4q12 deletion. One hundred interphase nuclei were examined, and 83% of them showed the CHIC2 locus deletion, a surrogate marker for the FIP1L1-PDGFRA fusion gene. The majority of the 17% of normal cells were round and mononucleated and were most likely lymphocytes (Figure 1A
and 1B)
. A FIP1L1-PDGFRA RT-PCR confirmed the presence of the fusion gene. Of note, a nested PCR was necessary to clearly identify the patient’s fusion gene (Figure 1C)
. According to the World Health Organization rules, these results demonstrated the clonal origin of the disease, and a definitive diagnosis of a CEL was thus made.1, 5
The identification of the FIP1L1-PDGFRA fusion gene is known to assess clonality in around 50% of hypereosinophilic syndrome.1, 2
To initiate a molecular follow-up of our patient, we sequenced the FIP1L1-PDGFRA PCR product. Whereas all of the previously reported patients had FIP1L1 breakpoints scattered between introns 7 to 10 (or 9 to 13 following the recently proposed LeukemiaNet nomenclature) (Figure 2A)
,11
we surprisingly observed a FIP1L1 breakpoint in intron 16 (Figure 2B)
. This broader dispersion of the FIP1L1 breakpoints could probably explain some false-negative FIP1L1-PDGFRA RT-PCR. Moreover, direct sequencing of the RT-PCR products revealed several splicing variants that joined exons 13, 15, or 16 of FIP1L1 to PDGFRA exon 12, generating PCR products from 629 to 738 bp long. Sequence analysis revealed an open reading frame only for one transcript. The resulting predicted sequence of the fusion protein is illustrated in Figure 2C
.
|
. The FIP1L1-PDGFRA transcript level dropped below the sensitivity level of this assay at day 55 of treatment (Figure 3A)
. We further analyzed the fusion gene kinetics by RQ RT-PCR. Patient’s specific primers and probes were designed after direct sequencing of the PCR product. We built a common reverse primer, AS-PDGF12, and a carboxyfluorescein dye-labeled TaqMan probe, P-PDGF12, located downstream the PDGFRA exon 12 breakpoints reported in the literature. The forward primer, FIP1L1E16T, was located in exon 16 of FIP1L1 (Figure 2B)
. With this assay, the FIP1L1-PDGFRA transcript dropped below the level of quantification at day 43 and remained undetectable thereafter (Figure 3B)
.
|
An accurate CEL diagnosis is essential as these patients greatly benefit from a specific and well-tolerated treatment with imatinib mesylate.2, 5, 8
Our report supports the use of both the FISH and RT-PCR approaches to avoid false-negative results and subsequent inadequate treatments for some patients. Indeed, the FISH technique is robust but not as sensitive as the RT-PCR. For instance, a HES patient previously treated with interferon- was addressed to our lab for FIP1L1-PDGFRA detection. FISH failed to detect the delCHIC2, whereas the FIP1L1-PDGFRA fusion gene was easily demonstrated by RT-PCR (data not shown). On the other hand, the FIP1L1-PDGFRA RT-PCR could be tricky, because it often generates multiple bands (Figure 1C)
. In theory, a FIP1L1 breakpoint localized downstream of exon 13 should generate, with the primers designed by Cools et al,2
an RT-PCR product longer than 1000 bp. Such a long PCR product could be difficult to amplify or detect. However, in our case, the skipping of exon 11 or part of exon 13 resulted in amplicon sizes of 629, 709, and 738 bp shorter than theoretically expected, which were easily visualized on agarose gel electrophoresis after two rounds of PCR (Figure 1C)
. Such a rare breakpoint had already been reported in a single patient by Chung et al.13
Moreover, we observed the present case and another one with a FIP1L1 exon 14 breakpoint out of six HES/CEL patients (data not shown). Therefore, although such events are rare, they should be considered when a new assay is proposed. For instance, Jovanovic et al10
recently proposed a new diagnostic approach based on real-time PCR. These RQ-PCR assays were designed for FIP1L1 breakpoints leading to fusion of exons 9 to 137, 8, 9, 10
of FIP1L1 to exon 12 of PDGFRA. Although this recent strategy is not the "gold" standard for FIP1L1-PDGFRA fusion gene detection at diagnosis, its easy procedure could encourage laboratories to choose it as a tool for screening purposes in HES. In another type of leukemia, we have recently described three false-negative BCR-ABL1 detection using a RQ RT-PCR screening approach in the process of chronic myelogenous leukemia diagnosis, as a consequence of complex genetic rearrangements.14
Our observation thus indicates a risk of false-negative results associated with simplified RQ RT-PCR strategies, whereas the technique described by Cools et al appropriately identified the fusion transcript.
Moreover, FIP1L1-PDGFRA-"negative" patients who respond to imatinib have been described.15 Our results illustrate that breakpoints in FIP1L1 may vary and that infrequent FIP1L1-PDGFRA fusion gene variants should be thoroughly excluded as a possible source of false-negative results in such cases. In conclusion, because of the therapeutic repercussions of a FIP1L1-PDGFRA (+) CEL diagnosis, it is essential to use the most appropriate diagnostic procedure to avoid false-negative results linked to unusual breakpoint.
Acknowledgments
We thank Antonietta Schioppa, Sophie Castermans, Andrée Villers, and Miguel Lopez y Cadenas for their excellent technical assistance as well as Drs. Li Rong and Jean-Louis Dargent for providing us with patient samples and clinical data.
Footnotes
Address reprint requests to Dr. F. Lambert, Department of Medical Genetics, CHU B35, University of Liège, Sart-Tilman, 4000 Liège, Belgium. E-mail: flambert{at}chu.ulg.ac.be
Supported by the Braconier-Lamarche, Lempereur, and Bonjean Oleffe Foundations as well as by the "Fonds d’Investissements pour la Recherche Scientifique" [(FIRS), Centre Hospitalier Universitaire de Liège] and the "Centre AntiCancéreux" (University of Liège, Liège, Belgium). A.C. is research Associate at the National Funds for Research (FNRS) and is supported by a grant from the Belgian Federation Against Cancer.
Accepted for publication February 21, 2007.
References
CT method. Methods 2001, 25:402-408[CrossRef][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
