Prognostic impact of C-KIT mutations in Core Binding Factor-Leukemia

C-KIT gene mutations are not a rare event in Core Binding Factor Leukemia (CBFL). To investigate their prognostic impact in this setting , we attempted at correlating the KIT mutational status with the WBC count, the level of the WBC-index, the presence of extramedullary disease (EMD) and the outcome, in t(8;21)(n=30) and inv(16) (n=20) AML patients. On the basis of genomic DNA analysis of c-KIT gene exons 2, 8, 10, 11, 17 for mutation detection, we categorized 21 patients (42%) in the "KIT mutated group" (KIT+) and 29 patients (58%) in the "KIT unmutated group" (KIT–). The median age at diagnosis was 46.4 and 45.6 years for the KIT(+) and KIT(–) groups, respectively (P=0.872). Among the KIT(+) patients, we found mutations in exon 8 (n = 4), exon 10 (n =1), exon 11 (n = 1) and exon 17 (n = 15). We recorded mean WBC counts of 28.6 x 109/L vs 34.5 x 109/L (P=0.051), and a mean WBC-index, expressed as WBC x (% Bone Marrow blasts/100), of 37.2 vs 15.5 (P=0.0395), for KIT(+) and KIT(–), respectively. Seven out of 50 patients experienced an EMD: in 6 cases (28.57%) this event occurred in the KIT(+) group. 40 patients (age < 60 years) were evaluable for clinical response; 18 out of them were KIT mutated. At a median follow-up of 24 months (range 10–107), we recorded for KIT(+) and KIT(–), respectively: CR incidence of 88.8 % (16/18) vs 100% (22/22) (P=0.38); Relapse Incidence 81.3% (13/16) vs 31.81% (7/22), P=0.0072; OS 27.7% (5/18) vs 77.3% (17/22), P=0.0049; DFS 16.7% (3/18) vs 77.3% (17/22), P=0.0005. Using a log-it linear model for univariate and multivariate regression analysis, we found a significant contribution to death (P=0.048) and relapse (P= 0.045) exerted by mutation in the whole group of patients; this effect was more evident when the analysis was restricted to patients <60 years (P=0.003 and P=0.014, respectively). In conclusion, this study confirms the correlation between c-KIT mutational status and high WBC-index in CBFL patients. Furthermore, our data indicate a statistical correlation between the presence of c-KIT mutations and both the OS and the DFS.