Genome-Wide Single-Nucleotide Polymorphism-Array Can Improve Prognostic Stratification of Core Binding Factor Acute Myeloid Leukemia, Especially in the Subgroup with Inv(16)/t(16;16) or without D816 C-KIT Mutation,

Abstract 3515 Background: The core binding factor (CBF) AML can be achieved long-term remission with high dose cytarabine-based chemotherapy alone. However, those with C-KIT gene mutation (esp. D816 C-KIT mutation) showed worse treatment outcomes compared to those with wild type C-KIT gene. The remaining cases without D816 C-KIT mutation is around 75% of CBF AML, which implies requirement of more sophisticated dissection of the patients according to their prognosis. Single nucleotide polymorphism (SNP) array (SNP-A) could detect cryptic abnormal genomic lesions, not identified by metaphase cytogenetics(MC). In this study, we analyzed the prognostic value of SNP-A based karyotyping combined with MC and its association with C-KIT mutation to facilitate further stratification of CBF AML patients. Methods and Materials: A total of 98 CBF AML patients were included and of whom, 63 (64%) and 35 patients (36%) were t(8;21) and inv(16)/t(16;16), respectively. Genome-Wide Human SNP 6.0 Array (Affymetrix, CA, USA) was performed using DNAs from marrow samples taken at diagnosis. Results: A total of 40 abnormal genomic lesions in 25 patients (26%) were detected by SNP-A karyotyping analysis, with a mean of 1.6 lesions per affected case (median size 33.6 Mb; range 0.4–145.9 Mb), including 3 CN-LOH lesions, 17 gain lesions, and 20 loss lesions. Survival of the patients with abnormal lesion(s) detected by SNP-A or/and MC was worse than those without any lesions in terms of 2 years' overall survival (OS; 57.5% vs 76.4%, p=0.028), event-free (EFS; 45.7% vs 66.2%, p=0.072) and leukemia free survival (LFS; 49.0% vs 77.4%, p=0.015). In contrast, MC alone could not stratify patients according to their long-term prognosis. Especially, in the subgroup with inv(16)/t(16;16), survival of patients with abnormal SNP-A/MC lesion showed worse than that of those without lesion (40.9±12.7% vs 80.2±10.4% at 2 yrs, p=0.040), but not in the subgroup with t(8;21) (66.85±9.1% vs 74.4±7.8% at 2 yrs, p=0.240). As for the subgroup with D816 C-KIT mutation, there were no differences of OS (p=0.417), EFS (p=0.380) and LFS (p=0.218) according to the presence of abnormal lesions detected by either SNP-A or MC. However, in the subgroup without D816 C-KIT mutation, those with abnormal lesions detected by either SNP-A or MC showed worse survival compared to those without abnormal lesions with respect to OS (61.6±8.7% vs 82.7±5.6% at 2 yrs, p=0.038). Multivariate analysis confirmed prognostic impact of abnormal SNP/MC lesions on OS (HR 2.743, p=0.020), EFS (HR 2.434, p=0.025), and LFS (HR 3.350, p=0.012). Conclusion: This study suggests that combined use of SNP-A with MC in the initial evaluation of CBF AML can provide an important prognostic value, especially in the inv(16)/t(16;16) subgroup or in the patients without having D816 C-KIT mutation. Disclosures: No relevant conflicts of interest to declare.