Impact of FLT3 Mutation Status and Other Genetic Parameters In Acute Promyelocytic Leukemia (APL) with t(15;17)(q22;q12)/PML-RARA.

Abstract 1685

Acute promyelocytic leukemia (APL) with t(15;17)(q22;q12)/PML-RARA displays the most favourable entity among the different subtypes of acute myeloid leukemia with a five year overall survival (OS) of more than 80%. However, around 10–20% still experience relapse. In order to find pretreatment parameters that may be predictive for outcome in APL, we have comprehensively analyzed 147 PML-RARA-positive patients (pts) at diagnosis. All pts were treated with ATRA in addition to standard chemotherapy. Patients were selected according to availability of cytogenetics, FLT3-ITD mutational status and FLT3-TKD mutational status. The cohort was comprized of 85 males and 62 females. Median white blood cell count (WBC) was 1,8×10⁹/L (range: 0.2–183.4 × 10⁹/L), median platelet count was 30.0 (range: 1.0–228.0 × 10⁹/L) and median hemoglobin level (Hb) was 9.7 (range: 3.6–16.4 g/dL). In 115 pts bone marrow smears were available: 68 were classified as M3 and 47 as M3v (FAB criteria). According to PML-RARA fusion type 89 pts had a bcr1 (long variant), 52 a bcr3 (short variant) and 6 a bcr2 (exon6) breakpoint. Diagnostic %PML-RARA/ABL1 transcript levels were heterogeneously distributed ranging from 0.6 to 96.7 (median: 18.5). In 57/147 pts (38.8%) additional cytogenetic aberrations (ACA) were detected (+8/+8q: n=26, i(17q): n=11, 9q-: n=3, complex: n=2, all others: n=15). A FLT3-ITD was detected in 47 pts (32.0%) and a FLT3-TKD mutation in 19 pts (12.9%). Thus, a total of 65 pts (44.2%) had a mutated FLT3 status (one case revealed both ITD and TKD mutations). FLT3-ITD was highly associated with bcr3 breakpoints (32 FLT3-ITD vs. 20 FLT3wt compared to 14 FLT3-ITD vs 75 FLT3wt in bcr1 and 1 FLT3-ITD vs. 5 FLT3wt in bcr2; p<0.001). Furthermore, FLT3-ITD was associated with higher WBC (mean: 33,153 compared to 5,170 × 10⁹/L in FLT3wt pts, p<0.001) and a lower platelet count (mean: 30,351 vs. 63,324 × 10⁹/L in FLT3wt pts, p=0.001). All parameters mentioned above were analyzed for a possible impact on OS and EFS. Median follow up time of this cohort was 16 months. OS was significantly better in males (2 year OS: 94.2% vs. 78.5% in females; p=0.038). Age as a continuous variable was found significantly related to both OS and EFS (p=0.002, each). Overall, the presence of ACAs had no impact on OS or EFS. In a next step the different ACAs as defined above were evaluated separately. The only group with significantly shorter OS and EFS was the one including the non recurrent “other” ACAs (2 years OS/EFS: 63.6% each, compared to 87.5%/81.4% in the remaining 5 cytogenetic groups, p=0.014/p=0.040). Importantly, these differences exclusively are due to four early deaths in this “other non recurrent” group. No significant effect on OS or EFS was found for WBC, Hb, platelet count, M3/M3v, PML-RARA breakpoint, diagnostic %PML-RARA/ABL1 transcript levels as well as FLT3-ITD, FLT3-TKD or combined FLT3-ITD/TKD status. However, when the FLT3-ITD/wildtype ratio was taken into account a significantly worse EFS was found for those with a FLT3-ITD/wildtype ratio >0.5 (n=21; 2 years EFS: 61.2% vs. 83.5% in the combined group with FLT3wt or FLT3-ITD/wildtype ratio < 0.5, p=0.009). Parameters with significant impact in univariate analysis were included into the multivariate analyses. For OS this was performed for gender, age and “other non recurrent ACA”. All three parameters were proven independent prognostic factors (p=0.026, RR: 0.24; p=0.004, RR: 1.44/decade; and p=0.013, RR: 4.32, respectively). For EFS age, FLT3-ITD/wildtype ratio >0.5, and “other non recurrent ACA” were analyzed (p=0.003, RR: 1.41/decade; p=0.077, RR: 2.73, and p=0.049, RR: 2.46, respectively). In conclusion, specific treatment in APL is extremely efficient what results in minor prognostic impact of otherwise established pretreatment parameters like WBC count, additional cytogenetic aberrations and mutated FLT3 status. Age is the strongest prognostic factor for OS and EFS. Non-recurrent ACAs are associated with an inferior OS. Most importantly, FLT3-ITD mutations with high allelic burden of more than 0.5 are associated with a shorter EFS. This data should be confirmed in controlled prospective studies to draw final conclusions for clinical decision making.