The Cell Expression of PML-RARa at Diagnosis Has a Marginal Correlation with Prognosis in Patients with Acute Promyelocytic Leukemia (APL).

Acute promyelocytic leukaemia (APL) is characterised by the t(15;17) leading to the fusion of the PML gene with the retinoic acid receptor a (RARa) gene. APL is associated with favourable prognosis, however, approximately 15–20% of patients ultimately relapse. With regards to pre-treatment characteristics, is now clear that long-established prognostic factors such as age and leukocyte count have a major impact on outcome. However, none of these parameters is robust enough, in order to adjust treatments strategies according to the risk of relapse. The quantification of PML-RARa fusion gene transcripts (leukemia-specific chimeric mRNA) can add important prognostic information useful for risk group stratification. However, few clinical studies have been carried out and there are discrepancies concerning the prognostic significance of the quantification of fusion gene transcripts of newly diagnosed patients. In this study we tested the PML- RARa fusion gene transcripts level in 126 newly diagnosed patients with t (15;17) in order to determine their correlation with disease characteristics at presentation and to identify a subset of patients at high risk of relapse. The presence of the PML- RARa fusion gene transcripts was analysed from 1mg of RNA by real-time quantitative RT-PCR (RQ-PCR) based on TaqMan probes and the 7700 Sequence Detector, performed according to the “Europe Against Cancer Program” (

Results: In 121 out of 126 patients, evaluable data were obtained: 73 (60,3%) cases had bcr-1, 4 (3,3%) cases had bcr-2 and 44 (36,4%) cases had bcr-3 fusion types. The expression of the fusion gene was reported as the PML- RARa normalized quantities (NQ, number of PML- RARa copies per 1x10⁴ ABL copies as gene control). The median of NQ was 3030 and the expression was highly variable (range of 826 to 9605). The NQ was not significantly associated with disease characteristic at diagnosis, such as age, percentage of blasts cells, fusion types and platelet counts. Nevertheless, patients with NQ above 3000 had lower white cell counts (7,4±12,1 vs 24,2±39,3x10⁹/L, p=0,03). Therefore, the fusion transcript copy number per leukemic cell was higher in patients with lower leukocyte count. Moreover, in survival analysis, the NQ did not correlate with disease free survival or overall survival.

Conclusions: Our data confirms the wide range of differences in expression activity between individual patients. The NQ did not correlate with the clinical and biological disease characteristics except for white cell counts. Neither did it correlate with principal prognostic parameters (response and survival).