In order to design a simple and reproducible classifier predicting the overall survival (OS) of patients with acute myeloid leukaemia (AML) of at least 60 years of age treated with intensive ‘7+3’ chemotherapy, 37 genes in 471 patients from the ALFA1200 study were sequenced. Overall, mutations in 7 genes independently predicted overall survival in distinct cytogenetic risk groups of patients.
“Current treatment options in AML have recently been widened and the need for intensive chemotherapy is a recurring question. In patients above the age of 60, a less intensive regimen might be beneficial in bringing them towards allogeneic transplantation when needed. The mutational profile of the disease can guide us in the choice of treatment. Rafaël Itzykson describes the different risk groups and treatment attitude in a clear overview.”
Despite the fact that acute myeloid leukaemia (AML) is most commonly diagnosed in patients above the age of 60, recent improvements in survival have been confined to younger adults. To design future randomised studies of intensive and less intensive therapies in fit older patients with AML, specific decision tools must be developed to identity the minority of patients in whom 7+3-based induction chemotherapy is unequivocally beneficial or futile among most older, fit patients with AML. The ALFA1200 study aimed to leverage a 37-gene panel in AML patients ≥60 years who received intensive chemotherapy, in order to identify risk groups that could be used for future prognostic and investigative purposes.
Between September 2012 and June 2016, this study enrolled 509 newly diagnosed AML patients aged ≥60 years with an ECOG performance status ≤3 eligible for intensive chemotherapy. Patients received a ‘7+3’ chemotherapy induction regimen that included idarubicin (12 mg/m2 per day, on days 1-3) and cytarabine (200 mg/m2 per day, days 1-7) as continuous infusions. Patients who did not achieve a complete remission (CR), or a CR with incomplete platelet recovery (CRp), would be eligible to receive a salvage regimen of intermediate-dose cytarabine boluses. Patients who did achieve a CR/CRp received two cytarabine bolus courses as a consolidation regimen. Molecular genetic analysis was carried out via targeted sequencing of a 37-gene panel, which used bone marrow or peripheral blood samples taken at inclusion. Of the 509 enrolled patients, 92% provided samples at baseline. The median age was 68 years and 82.8% had clinically defined de novo AML. The median white blood cell (WBC) count was 5.3 x 109/L and bone marrow blasts were ≥30% in 80.7% of patients. After 1 or 2 courses of treatment, 72.4% had achieved either a CR or CRp. Additionally, 131 patients received a haematopoietic stem cell transplant, which included 87 CR responders. At a median follow-up of 44.8 months, there were 207 relapses and 318 deaths, which translated to a median RFS and OS of 14.8 and 21.2 months, respectively. Cytogenetic risk was good, intermediate, poor and not available in 2.8%, 72.0%, 17.8% and 7.4% patients, respectively. Among the 37 sequenced genes, 19 lesions in 17 genes were found in at least 5% of cases, with the most frequent being DNTM3A (28.7%), NPM1 (27.0%), TET2 (21.0%) and FLT3-ITD (18.7%).
Overall, 76.1% of patients with good or intermediate cytogenetic risk achieved a CR/CRp, whilst this was only 56.0% in patients with a poor risk (p= 0.0004). This translated into a median OS and RFS of 25.0 and 16.8 months, in good/intermediate-risk patients, and 9.5 months and 7.1 months in poor-risk patients, respectively (p–value for OS ≤ 10-4; p–value for RFS= 0.0005). Mutations in seven genes independently predicted OS in distinct cytogenetic groups of patients with AML. TP53 (hazards ratio [HR], 2.49; p= 0.0003) and KRAS (HR, 3.60; p= 0.001) mutations independently worsened the OS of patients with poor-risk cytogenetics. In those without poor-risk cytogenetics, NPM1 (HR, 0.57; p= 0.0004), FLT3 internal tandem duplications with low (HR, 1.85; p= 0.0005) or high (HR, 3.51; P< 10-4) allelic ratio, DNMT3A (HR, 1.86; p< 10-4), NRAS (HR, 1.54; p= 0.019), and ASXL1 (HR, 1.89; p= 0.0003) mutations independently predicted OS.
In order to develop the ALFA molecular decision tool, cytogenetic risk and mutations in these 7 genes were combined to assign patients to different groups. Patients with non-poor cytogenetics and either an NPM1 mutation with up to 1 mutation in FLT3-ITD low allelic ratio, DNMT3A, ASXL1 or NRAS, or those wild-type NPM1, FLT3-ITD, DNMT3A, ASXL1 or NRAS (N= 184) were assigned to the very favourable go-go group. Patients with poor cytogenetics and either a mutation in TP53 or KRAS were put into the ‘no-go’ group (N= 36), and the remaining patients were put into the ‘slow-go’ group (N= 251). In these groups, the 2-year survival estimates were 66.1%, 2.8% and 39.1%, respectively (overall p ≤ 10-5). 2-year RFS estimates mirrored those observed in OS at 49.7%, 5.3% and 30.2%, respectively. External validation of the ALFA decision tool was performed in three distinct cohorts of 141 to 466 older patients. Statistically significant differences in OS were seen between decision tiers in all three cohorts and to a lesser extent on RFS.
This study was able to identify 7 genes predictive of OS in distinct cytogenetic groups of patients with AML aged ≥60 years and treated intensively. Using these genes to create a decision model, the study was also able to create and validate a simple genetic model to identify older patients with AML with very good, intermediate or poor outcome with 7+3 chemotherapy.
Itzykson R, Fournier E, Berthon C, et al. Genetic identification of patients with AML older than 60 years achieving long-term survival with intensive chemotherapy. Blood. 2021;138(7):507-19.