Prognostic factors of clinical use in acute myeloid leukemia

Date 20 October 2018
Event ESMO 2018 Congress
Session Poster display session: Biomarkers, Gynaecological cancers, Haematological malignancies, Immunotherapy of cancer, New diagnostic tools, NSCLC - early stage, locally advanced & metastatic, SCLC, Thoracic malignancies, Translational research
Topics Leukaemia
Presenter Pedro Chorão
Citation Annals of Oncology (2018) 29 (suppl_8): viii359-viii371. 10.1093/annonc/mdy286
Authors P.A. Chorão1, N.T. Tavares2, E. Aguiar1, M.L. Amorim1, P. Gomes1, F. Príncipe1, J.E. Guimarães1
  • 1Clinical Hematology Department, Centro Hospitalar São João, EPE, 4200-319 - Porto/PT
  • 2Medical Oncology Department, Centro Hospitalar São João, EPE, 4200-319 - Porto/PT



The heterogeneity in acute myeloid leukemia (AML) is influenced by disease and host-specific factors. In recent years, independent molecular factors and host characteristics that affect prognosis have been described. Nevertheless, data is needed to warrant their clinical use.


We retrospectively reviewed data on 356 AML patients from October 2008 to June 2017, including age, Charlson Comorbity Index (CCI), ECOG performance status (PS), laboratory parameters (complete blood count, bone marrow blasts, lactic dehydrogenase) and disease factors (de novo or secondary AML (sAML), genetic alterations). The 2017 European Leukemia Net genetics risk stratification (GRS) was used. Univariate and multivariate Cox regression analysis was performed.


The intensive chemotherapy (IC) group comprised of 241 patients, median age 58 years (y), 95% having PS ≤ 1, 67% CCI ≤2 and 22% were sAML. GRS was 23% favorable, 44% intermediate, 27% adverse and 6% unknown. Median overall survival (OS) was 15 months (mo), with 28% 3y-OS. In multivariate analysis (Table), age older than 60y (median 6 versus (vs) 22mo), PS ≥ 2 (median 7 vs 15mo) and higher risk GRS (median 35mo favorable vs 13mo intermediate and 11mo adverse) impacted on survival. The non-IC group included 112 patients, median age 67y, 76% having PS ≤ 1, 6% CCI ≤2 and 37% were sAML. GRS was 12% favorable, 49% intermediate, 24% adverse and 15% unknown. Median OS was 3mo, with 10% 3y-OS. In multivariate analysis (Table), age older than 60y (median 2 vs 17mo), PS ≥ 2 (median 1 vs 5mo) and higher risk GRS (median not reached in favorable vs 4mo intermediate and 1mo adverse) impacted on prognosis.Table: 1028P

Intensive chemotherapy HR [95% CI]Non-intensive therapy HR [95% CI]
Age≤60 yearsRefRef
>60 years2.41 [1.76-3.31]3.09 [1.70-5.65]
ECOG PS≤ 1RefRef
> 12.36 [1.23-4.53]2.25 [1.27-3.98]
ELN2017 geneticsFavorableRefRef
Intermediate2.04 [1.33-3.14]3.26 [1.30-8.15]
Adverse2.26 [1.42-3.59]5.55 [2.13-14.44]


In spite of the high number of recognized risk factors, in real-life only GRS, age and PS were of clinical use to predict survival in both IC and non-IC sets. In our sample, there was no significant impact of sAML, CCI and laboratory parameters. Efforts are needed to identify more factors that aid clinical decision in the treatment of AML.

Clinical trial identification

Legal entity responsible for the study

Clinical Hematology Department, Centro Hospitalar São João, EPE.


Has not received any funding.

Editorial Acknowledgement


All authors have declared no conflicts of interest.