A New Molecular and Clinical Prognostic Score for Risk Stratification in CN-AML.

Abstract 2635 Poster Board II-611 Background: Cytogenetically normal acute myeloid leukemia (CN-AML) is associated with an intermediate outcome. A number of clinical and molecular risk factors have been characterized pointing to the heterogeneity of this group. The purpose of the study was to define a prognostic model based on pre-treatment patient characteristics to facilitate choice of therapy by definition of patient groups with different prognoses. Patients and methods: We evaluated four molecular markers (mutations of NPM1, CEBPA, MLL-PTD; FLT3-ITD mutant level; interaction term NPM1 and FLT3-ITD mutant level) and nine clinical parameters (white blood count (WBC), platelet count, hemoglobin level, lactase dehydrogenase (LDH) level, bone marrow blasts, de novo AML vs. non de novo AML, performance status, sex and age) at initial diagnosis in 648 patients with CN-AML treated in the AMLCG (German AML Cooperative Group) 1999 trial. The outcome parameter overall survival (OS) was calculated from randomization to death from any cause or to the latest follow-up date. Event-free survival (EFS) was defined as the period from the start of therapy until lack of a complete remission (CR), relapse of AML after CR or death without relapse. Relapse-free survival (RFS) was determined for responders from the first day of a CR until relapse or death without relapse. Univariate and multivariate Cox regression analyses for OS were performed. All parameters with p'0.05 in multivariate analyses after backward elimination and their regression coefficients were applied in the prognostic score. The minimal p-value approach was used to identify the risk groups with the greatest differences in OS. Results: In our patient cohort 84% had de novo AML. Median age was 60 years (17–85 years) and 70% had an ECOG score ≤1. Median platelet count was 57 G/l (5–643 G/l), median WBC was 18 G/l (0.1–798 G/l) and median hemoglobin level was 9.2 g/dl (4.2–16.4 g/dl). Mutations of NPM1, FLT3-ITD, MLL-PTD and CEBPA were present in 51%, 27%, 8% and 10% of patients, respectively. Median FLT3-ITD mutant level in FLT3-ITD mutated patients was 0.42 (0.02–1.00). Of 648 patients 377 had died. Median OS was 20 months with a median follow-up of 45 months. In the multivariate analyses for OS, the following parameters were significant: age (+10, years, HR: 1.3, p<0.001), WBC (10 fold, ×109/l, HR: 1.4, p<0.001), NPM1 (mutation vs. wild-type, HR: 0.35, p<0.001), CEBPA (mutation vs. wild-type, HR: 0.47, p=0.001), interaction term NPM1/FLT3-ITD mutant level (+1, HR: 4.5, p=0.006), performance status (ECOG 0,1 vs. ECOG 2-4, HR: 1.4, p=0.006) and platelet count (10 fold, ×109/l, HR: 0.70, p=0.016). After calculation of the prognostic score for each patient and definition of two cutpoints, we could identify three risk groups (median OS (N=590): not reached (n=169) vs. 22.7 months (n=220) vs. 8.4 months (n=201), p<0.001; median EFS (N=583): 42.3 months (n=168) vs. 7.6 months (n=216) vs. 3.2 months (n=199), p<0.001; median RFS (N=383): not reached (n=136) vs. 15.3 months (n=143) vs. 7.6 months (n=104), p<0.001). Furthermore this model was valid in both age subgroups (<60 years / ≥60 years). Interestingly, a subset of 31% of patients within the molecular favorable NPM1+/FLT3-ITD- risk group were assigned to the intermediate group according to our prognostic score and 31% of the low risk group were not NPM1+/FLT3-ITD-. Conclusions: We propose a new prognostic score based on pre-therapeutic clinical and well-established molecular markers that could be easily applied in the routine patient care setting for risk stratification and risk-adapted therapy. Further prospective validation is required to confirm the clinical relevance of this score. Disclosures: Unterhalt: Roche: travel support. Hoster:Roche: travel support.