Acute myeloid leukaemia (AML) is a haematological malignant disease characterized by an uncontrolled proliferation of immature hematopoietic cells. Over the last two decades, clinical trials have demonstrated an improved response rate in younger adult AML. Aggressive induction plus more potent intensification programs with chemotherapy alone or chemotherapy plus stem cell transplantation (SCT) has improved treatment results. Advances in understanding disease biology, improvements in induction and consolidation program, and better supportive care have also all contributed. A number of clinical and laboratory characteristics influence the response to treatment and, thus, the survival of patients with AML. Among them, cytogenetic at diagnosis represents the most important prognostic variable. However, other factors may have a prognostic value and may influence patient's outcome.

Anaemia and thrombocytopenia are cardinal manifestations of AML. Over the last decades, it has become apparent that the frequency of allogeneic blood transfusions can modify host immunity and clinical outcomes. Anaemia has long been recognized as an adverse prognostic factor in myelodysplastic syndrome (MDS), which represents a pre-leukemic disease. Red blood cell (RBC) transfusion need was identified as a strong and independent risk factor for survival in MDS, for which the presence and severity of anaemia were attributed to a clonally advanced and biologically more aggressive disease.

Based on these data, we retrospectively assessed the prognostic value of RBC and platelet transfusions at the time of diagnosis and the frequency of transfusions during the first induction course of chemotherapy in a large unselected group of patients with previously untreated AML.

Inclusion Criteria:

• Patient > 15 years old
• Newly diagnosed AML or post myelodysplastic syndrome (MDS)

Exclusion Criteria:

• Patients with M3 AML of FAB classification (APL, Acute Promyelocytic Leukemia)
• World Health Organization (WHO) performance status >2;
• Left ventricular systolic ejection fraction below the normal range
• Patients who have any severe and/or uncontrolled medical conditions or other conditions that could affect their participation in the study
• Serum creatinine concentration > 2x ULN (Upper Limit of Normal laboratory ranges),
• AST or ALT levels > 2.0 x ULN, except if AML-related

An observational, longitudinal, and retrospective study was conducted in 1067 patients (>15 years of age) fulfilling the clinical and laboratory criteria for diagnosis of AML. Patients were admitted at the Lyon University Hospital over two decades (fromMarch 1985 to December 2006) and enrolled onto 12 successive therapeutic trials according to age and time of enrollment. Diagnosis of AML was based on smears of bone marrow aspirates. Leukemic cells were classified according to the French-American-British (FAB) criteria. FAB M3 AMLs were not included in the study. AML with 20 % or more myeloid marrow blasts, either de novo or evolving from a MDS, was eligible in the absence of (i)World Health Organization (WHO) performance status >2; (ii) left ventricular systolic ejection fraction below the normal range; (iii) a creatinine or hepatic enzyme levels >2 ULN, except ifAML-related; and (iv) uncontrolled severe infection. All participants gave their written informed consent