Extreme Variability of FIP1L1-PDGFRalpha Transcripts Do Not Influence to Imatinib Mesylate Response in CEL: Clinical Follow-up and Molecular Analysis of the Italian Multicenter Prospective Study

Background. The presence of an interstitial deletion in the long arm of chromosome 4 leads to the formation of the fusion gene FIP1L1-PDGFRalpha (F/P) coding for a constitutively activated form of PDGFRalpha. F/P has become the molecular marker of clonal hypereosinophilic syndrome (CEL) and it predicts a dramatic response to imatinib mesylate (IM). Different F/P transcripts have been described, but the correlation with kinetic of molecular response to IM and with the clinical presentation is unknown.

Aims. The aim of this study was to evaluate the duration of IM response in FIP1L1-PDGFRalpha positive CEL patients and the correlation between their clinical behaviour and molecular characteristics of the transcripts.

Methods. A prospective multicenter study of the HES was established in 2001. The 33 FIP1L1-PDGFRalpha positive patients (F/P+) were monitored every three months for two years then every six months using a nested retrospective reverse transcriptase polymerase chain reaction (RT-PCR). The RNA was subsequently reverse transcribed in cDNA and then amplified by reverse transcription-PCR. The samples were purified and sequenced using the Big Dye Terminator Cycle Sequencing Kit (Applied Biosystems). Patients were systematically screened for organ damage with instrumental evaluation and for the presence of symptoms. Patients were treated with IM 100 to 400 mg daily. The observation period ranges between 15 and 72 months (median 32 months).

Results. There were 32 males and one female patients. Organ involvement was recorded in 42% of F/P+. After imatinib therapy all patients achieved a complete hematologic response (CHR) in less than one month, and PCR negativity in a median time of 3 months (range 1–9). They became negative for organ localizations and free of symptoms. All 29 patients who continue imatinib therapy remain in CHR and RT-PCR negative, with a dose of 100 to 400 mg daily. In four patients IM treatment was discontinued and the fusion transcript became rapidly detectable. CHR was maintained. The transcript was again undetectable upon treatment resumption. 28 samples of the totality were valuable for molecular analysis. All deletions of 4q12 generates in-frame chimeric fusion gene. FIP1L1 breakpoints scattered between exon 9 to 18, with several splicing variants. All breakpoints in PDGFRA are located within exon 12. Fusion gene sequencing demonstrate an extreme variability. Region of FIP1L1 varies in length from 109 to more than 500 nucleotides. The more conserved regions are exon 10 and exon 11 of FIP1L1, that are repeated together in 13 out of 28 analyzed transcripts. Transcript of the only female patient is the same of one of the males. No evidence of correlation was noted with kinetic of molecular response or with the presence at diagnosis of peculiar organ involvement. More complexity of transcript is noted in patients with longer history of disease prior to imatinib therapy. Interpretation and conclusion. with this large series of patients we can confirm more complexity and variability in FIP1L1-PDGFRalpha transcripts than data reported in other series, but no evidence of clinical correlation between this heterogeneity and phenotype of disease. Moreover, the response to the imatinib therapy is not influenced by the biologic features, but it depends on continuous therapy.