PB1730 A SIMPLE FLOW CYTOMETRY‐BASED SCORE MAY OPTIMIZE TESTING FOR BIALLELIC CEBPA MUTATIONS IN PATIENTS WITH ACUTE MYELOID LEUKEMIA PATIENTS

Abstract

Background:Acute myeloid leukemia with biallelic mutation of CEBPA (CEBPA‐dmAML) is a distinct entity recognized by WHO 2016 classification with good prognosis. However testing for CEBPA mutationsischallengingdue tothe lack of hot spots and the presence of single nucleotide mutations. Sanger sequencing is the appropriate technique to detect CEBPA mutations, although it is expensive, time consuming and should be performed only in experienced centers.Aims:The aim of this study was to test the predictive value of a simple 7‐antigens cytofluorimetric (IF) score in predicting the presence of CEBPA‐dm in a cohort de novo cytogenetically normal (NK) AML patients.Methods:One‐hundred consecutive younger (<60 yrs), de‐novo AML patients, treated between January 2006 and January 2016, with available IF, cytogenetic and molecular work up performed at diagnosis have been evaluated. All patients received fludarabine‐high dose cytarabine‐idarubicine intensive induction followed by a risk‐adapted consolidation strategy. Patients with cytogenetic abnormalities and/or with NPM1 mutation were excluded from further analysis. Basing on previous reports on IF features of CEBPAdm AML, a comprehensive flow‐cytometry‐based CEBPA dm score was generated, assigning one point each for expression of HLA DR, CD7, CD13, CD15, CD33, CD34 and one point for lack of expression of CD14.Continuous variables were compared using Student's T test or, where necessary, Wilcoxon's Rank test. Dichotomous variables were compared using the Chi‐square test or, where necessary, Fisher's exact test. Survival curves were built using the Kaplan Meier method. A Cox Proportional Hazard Model was built for multivariate survival analysis. All two‐tailed p‐values < 0.05 were considered statistically significant.Results:Fifty patients were included in the analysis. Median age at diagnosis was 51.5 years (range 19‐60 years), CEBPA‐dmwas found in 9 patients (18%) and high allelic burden FLT3‐ITD in 6 patients (12%), ELN 2017 risk group was intermediate in 44 patient (88%) and high in 6 (12%).CEBPA‐dmscore was 7 in 2 patients, (4%), 6 in 16 (32%), 5 in 22 (44%), 4 in 12 (24%), 3 in 2 (4%), and 2 in 2 cases (4%).A score ≥ 6 was significantly associated with the presence of CEBPA‐ dm (p < 0.05), whereas no CEBPA‐ dm was recorded in patients with a score < 6.The predictive positive value (PPV) for a score >5 was 9/18 (50%), whereas the predictive negative value (PNV) for a score < 6 was 100% (Tab I).Furthermore, 4 out of the 6 patients with aCEBPA score of 6 with a positive CD7 had CEBPA‐dm. The PPV for >5 score including CD7 expression was 75%.Median follow‐up was 62 months (IC 95%: 39.89‐84.10 months), 3‐year Overall Survival (OS) was 54.9% (median not reached). Patients with CEBPA‐dm had better outcome if compared to unmutated patients (3‐year OS was 74.1% and 51.5% in CEBPA dm and wild type, respectively, p < 0.02). Multivariate OS analysis showed that CEBPA‐dm (p < 0.02) and FLT3‐ITD(p < 0.01) mutation status risk were the strongest independent predictor of OS.Summary/Conclusion:Although lacking prognostic value by itself, our score can promptly identify a vast group of AML patients who are not likely to harbor CEBPA‐dm, thusoptimizing time expenditure and costs of molecular work up at diagnosis, in particular in centers where complex molecular analysis cannot be carried on in house.image