ISAL 2019 | Stratification of genetic risk factors in acute myeloid leukemia

On 25 February 2019 at the Acute Leukemias XVII Biology and Treatment Strategies biennial symposium in Munich, Germany, Klaus Metzeler from the Ludwig Maximilian University of Munich, Munich, Germany, presented data from a study which aimed to validate the European LeukemiaNet (ELN) genetic risk stratification system in adult patients with acute myeloid leukemia (AML), who received cytarabine and anthracycline induction chemotherapy enrolled on acute myeloid leukemia cooperative group (AMLCG) trials.^1,2

Overall, patients included in this study (n = 1118; median age = 58 years; range, 18–86) were analyzed in two cohorts. The incidence cohort (n = 772) included patients with all cytogenetic categories from the AMLCG-1999 and AMLCG-2008 studies, in genetically unselected patients. The extension cohort (n = 346) included patients with AML who had normal cytogenetic characteristics from the AMLCG-1999 study, with patients only included in the outcome analyses. Patients underwent genetic analysis, including sequencing for mutations in the NPM1, FLT3, CEBPA, RUNX1, ASXL1, and TP53 genes.

Key findings

Data shown based on the ELN 2017 classification

• Compared with the ELN 2010 classification, 27% of patients (n = 772) were reassigned to a category associated with an adverse risk or a more favorable outcome
• In patients ≥ 60 years, only 27.7% (n = 95) were classified with favorable genetic risk status compared to 41.3% (n = 177) in patients aged < 60 years
• Complete remission (CR) rates in the total population (n = 1118, P < 0.0001) vs patients < 60 years vs patients ≥ 60 years, who received induction chemotherapy:
  • Favorable genetic risk (n = 428): 72% vs 75% vs 68%
  • Intermediate genetic risk (n = 290): 67% vs 66% vs 66%
  • Adverse genetic risk (n = 400): 40% vs 43% vs 39%
• Five-year relapse-free survival (RFS) in all age groups, P < 0.0001:
  • Favorable genetic risk (n = 306): 54%
  • Intermediate genetic risk (n = 193): 26%
  • Adverse genetic risk (n = 161): 12%
• Five-year overall survival (OS) in patients < 60 years (n = 601) vs ≥ 60 years (n = 517) , P < 0.0001:
  • Favorable genetic risk: 64% (n = 265) vs 36% (n = 163)
  • Intermediate genetic risk: 41% (n = 167) vs 15% (n = 123)
  • Adverse genetic risk: 19% (n = 169) vs 6% (n = 231)
• In patients aged ≥ 75 years who have received intensive therapy, the ELN 2017 risk classification is uncertain, with prognostic indicators having less relevance in this population
• Outcomes are similar between favorable risk patients in the ELN 2017 and ELN 2010 groups, however, ELN 2017 improves risk stratification through the identification of greater numbers of high risk and low risk patients with AML

Data shown from the refined ELN 2017 classification using established prognostic markers

• Five-year OS in patients possessing mutated biCEBPA, inv(16), or t(16;16), defined as very favorable genetic risk: 69%
• Five-year OS in patients harboring mutated TP53 and complex karyotype, defined as very adverse genetic risk: 0%

Data shown from the refined ELN 2017 classification using the incorporation of additional pretreatment markers

• Shorter OS and event-free survival (EFS) was shown to associate with a high mutant NPM1 allele burden
• Multivariate analysis illustrated factors associating with OS:
  • FLT3-ITD^high:WT ratio: HR = 2.2 (95% CI, 1.6–2.9), P < 0.0001
  • Mutant DNMT3A: HR = 1.4 (95% CI, 1.1–1.9), P = 0.008

In conclusion, the ELN 2017 classification allows validated genetic risk based stratification of patients with AML receiving cytarabine and anthracycline induction chemotherapy. Furthermore, the study concluded that more patients are classified with favorable or adverse risk AML, and less as intermediate risk AML compared to the ELN 2010 classification system. Moreover, this study illustrates the possibility for further refinement of the ELN 2017 genetic risk classification system. The relevance of the risk classification system is unclear in patients receiving alternative therapies, such as venetoclax, or the effect of molecularly targeted agents, as it has been established primarily in patients receiving induction chemotherapy. Klaus Metzeler stated the importance of integrating pre-treatment risk classification with the measures of response and residual disease post-treatment. This study highlights that the future challenge is the integration of a multidimensional model to effectively aid in the selection of the most promising sequence of therapies for patients with AML.