Acute myeloid leukemia (AML) is a heterogeneous disease. In recent years, the knowledge and management of AML has changed significantly. The classification of AML has shifted from purely focusing on morphological features, cytogenetics, and phenotypic characteristics, towards a greater emphasis on molecular genetic abnormalities. During the 1st NCRI AML academy meeting, Professor Brian Huntly from the University of Cambridge, Cambridge, UK delivered a presentation detailing how genetic studies can facilitate clinical decision making in terms of diagnosis, patient prognosis and also predicting which patients may be sensitive to the novel agents.1
Professor Huntly discussed the molecular classification and risk stratification schemas for AML: the 2016 World Health Organization (WHO) classification system, which distinguishes leukemias by their characteristic cytogenetic or genetic abnormalities2 and the 2017 ELN genetic risk stratification system, which uses a 3-group classification (favorable, intermediate, and adverse)3. Risk stratification informs prognosis and therapeutic approach, but as Professor Huntly noted, there is notable heterogeneity within the three ELN groups in terms of clinical outcome, indicating that some patients are currently being treated inappropriately.
Collaborative studies in AML genetics
Professor Huntly described the ongoing collaboration between the National Cancer Research Institute (NCRI) working group and the Wellcome Trust Sanger Institute which is the largest study to-date of leukemia genes in approximately 2100 patients from eight NCRI AML clinical trials. Professor Huntly presented the preliminary results to the NCRI audience and described the ongoing work to create a simplified, practical combined risk stratification system for AML. The data have not yet been published and, therefore, are not summarized here.
Professor Huntly highlighted the HARMONY project – a public-private European Network established in January 2017 – which so far has captured data on 45,000 patients with blood cancers from multiple sources such as clinical trials and registries. For AML, the HARMONY project will create a dataset of approximately 4000–5000 patients with the aim of identifying molecular profiles that can predict clinical course and drug response.