FLT3-ITD confers poor prognosis in patients with acute promyelocytic leukemia treated with AIDA protocols: long-term follow-up analysis

Internal tandem duplication (ITD) of the fms-like tyro-sine kinase 3 gene (FLT3/ITD) occurs in 20–30% of young adults with acute myeloid leukemia (AML) and is associated with poor prognosis.1–4FLT3/ITD mutation is detected in 35–40% of acute promyelocytic leukemia (APL) patients and is associated with a high WBC count on presentation, hypogranular variant (M3v) morphology and short (bcr3) PML-RARA isoform. However, the prognostic significance of this alteration remains controversial5 whereas few data have been reported on the activity of arsenic trioxide in patients with FLT3/ITD mutation.5 We previously investigated the clinical-biological correlations of FLT3/ITD in 90 APL patients receiving the AIDA protocol.6 While we confirmed the association between FLT3/ITD and the above-mentioned base-line features, for FLT3/ITD patients we found no difference in response to induction and only a trend towards an inferior outcome. In the present study, we report the FLT3/ITD prognostic impact on an expanded series of 147 APL patients with a considerably longer follow up.

Table 1.Differences between FLT3-positive and negative patients.

Figure 1.Cumulative incidence of relapse (CIR) according to FLT3 status: continuous line represents CIR of FLT3-ITD positive patients; split curve represents FLT3-ITD negative patients.

During the period April 1993-October 2010, 147 patients with newly diagnosed APL were observed and treated with the AIDA04937 and AIDA20008 protocols at the Sapienza University of Rome. The diagnosis was initially established morphologically and confirmed by RT-PCR identification of PML/RARA fusion gene as reported.7,8 Seventy-three patients received AIDA 0493 regimen,7 while 74 patients diagnosed after May 2000 were treated according to the risk-adapted regimen AIDA-2000,8 with distinct post-induction approaches based on the initial risk stratification (ATRA included in each consolidation course and reduced intensity chemotherapy for low and intermediate risk).9 All patients received maintenance according to the protocols.7,8 The following clinical characteristics at diagnosis were analyzed according to the FLT3 status: age, gender, FAB classification, peripheral WBC count, platelet count, hemoglobin level, karyotype, PML/RARA isoform and relapse risk category. Molecular tests were performed after the third consolidation and thereafter every three months for two years and every six months after the end of maintenance. Molecular relapse was defined as positive RT-PCR test detected in two successive marrow samples collected at any time after consolidation and in the absence of morphologically detectable blasts in both the marrow and peripheral blood. Differentiation syndrome (DS) was defined as “definitely present” when at least three major signs were recorded according to the criteria of Frankel et al.10 Screening for FLT3-ITD mutations was performed as previously reported.6 D835 mutations were also investigated, but not considered in this study due to low prevalence. Wilcoxon-Mann-Whitney test was performed for comparison of non-parametric series and Fisher’s exact test was used to compare categories. Multivariable Cox proportional hazard regression models were performed and expressed as hazard ratios with 95% confidence intervals (CI). P<0.05 was considered statistically significant. Overall survival (OS) was estimated using the Kaplan-Meier method. Relapse-free survival (RFS) was estimated counting either relapse or death in CR as events. Cumulative incidence of relapse (CIR) was defined as the time from end of induction to date of hematologic or molecular relapse, whichever occurred first, considering death in remission as a competitive event.

Thirty-three of the 147 patients were FLT3-ITD, whereas FLT3-D835 mutation was detected in 9.5% of patients. Of the FLT3-ITD patients, 19 were males (57%) and 14 females; 27% were diagnosed as having a M3v and 36% as high relapse risk. In the negative cohort of 114 patients, 46 (40%) were males, 12% were diagnosed as having a M3v and 15.7% as high relapse risk. Twenty-one FLT3-ITD patients (63%) had the bcr3 transcript type as compared to 37 of 114 (32%) patients with germline FLT3 (P=0.002). Eight FLT3-ITD patients (24%) experienced a documented DS compared to 14 (12%) in the FLT3-ITD cohort (P=0.02). After a median follow up of nine years (range 5–19), a significant difference was seen in terms of OS: 96% in the FLT3-ITD cohort compared to 39% in the FLT3-ITD cohort (P=0.0001). RFS was 90% and 30% in the FLT3-ITD-negative and -positive cohorts, respectively (P=0.017) (Table 1). The CIR at nine years was 4% (95%CI: 1,279–6,238) for the FLT3-negative and 60% (95%CI: 22,632–81,236) for the FLT3 subset (Figure 1; P=0.0001). Cox multivariable analysis was performed for hazard of relapse and OS: of all previously mentioned factors tested, only FLT3-ITD had an independent prognostic value (hazard ratio = 2.4; 95%CI: 1–4.5; P=0.01) and bcr3 that carried a poorer prognosis compared to bcr1 (hazard ratio of 2.2; 95%CI: 1–4.2; P=0.02). Following our initial study reported in 2002,6 we prospectively analyzed FLT3 for the presence of ITD in all newly diagnosed patients. In this final analysis, we found a weak correlation of FLT3-ITD mutation with male gender and we confirm a strong correlation with WBC count, M3v and bcr3 at baseline. Although we confirmed no difference in response to induction, we observed in a prolonged follow up a considerably more unfavorable outcome for FLT3-ITD patients in terms of RFS, DFS and OS. The strong association of mutation with specific clinical features and the poor outcome demonstrated in long-term analysis, allows an aggressive subset of APL at diagnosis to be identified, which probably deserves an intensification of treatment in order to prevent relapse.

Conflicting results have so far been reported with respect to the correlation between FLT3 status and OS. While Gale et al. reported no association with outcome, even though FLT3-ITD patients had a higher rate of induction deaths,11 several groups reported a less favorable outcome for FLT3-ITD patients. A Spanish group described a shorter 5-year RFS in patients with ITD mutant/wild-type ratio or longer ITD size.12 A German study found that patients with FLT3 mutation/wild-type ratio of 0.5 or over had better 2-year OS and EFS rates compared to patients with a ratio less than than 0.5.13 In the PETHEMA/HOVON experience,14 in univariate analysis FLT3-ITD mutations were associated to higher relapse and lower OS not retained in multivariate analysis, probably due to a shorter median follow up. A Korean group also showed a higher relapse rate in FLT3-ITD patients with respect to those FLT3-negative.15 Given its distinct mechanism of action and non-crossresistance with chemotherapy and ATRA, it will be interesting to explore the efficacy of arsenic trioxide in APL patients harboring FLT3-ITD.