Co-mutational landscape of Indian core binding AML: An answer to the unfavorable outcome in a favorable AML?

Abstract

6534 Background: Core binding factor AML, which encompasses RUNX1/RUNX1T1 and CBFB/MYH11, comprise 10-15% of all AML, currently classified as favorable risk in the 2022 European LeukemiaNet Risk Stratification. The Indian CBF AML cohort as reported in earlier studies tends to be younger, with extramedullary disease and with higher number of co-mutations. Also due to unavailability of the anti CD33 gemtuzumab ozogamicin in India, the outcomes to standard therapies tend to be dismal when compared to the West. This raises a dilemma for the ELN risk stratification of this entity as a favorable risk. Methods: 165 patients of CBF AML were accrued, from two different centers. Non-disclosure consents from the participating centers were obtained. Patients with documented CBF AML who have undergone NGS based testing were included. Descriptive statistical analysis was done for continuous and ordinal data. The analysis was done using SPSS version 24. A two tailed p value of <5% was considered significant. The data from TCGA studies was compared which was accessed through www.cbioportal.org and all studies of AML were included. Results: A total of 165 patients of CBF AML were included in the study. Median age was 33.4 years (range 16-45) with a clear male predilection (male to female ratio of 3.1:1).RUNX1/RUNX1T1 rearrangement was detected in 70% (116) cases and 49 (30%) cases harbored the CBFB/MYH11 gene rearrangement. At least one additional mutation was detected in 70% cases with RUNX1/RUNX1T1 rearrangement and 80% cases of CBFB/MYH11.The most frequent co-occurring alteration was NRAS (32% cases across both subtypes, more common in CBFB/MYH11). Another statistically significant finding was patients with inv(16) manifested a significantly higher frequency of Spliceosome mutations (p = 0.028) and WT1 mutations (p = 0.021) compared to t(8;21). Two patients harbored concomitant TP53 mutations predicted to be pathogenic based on ClinVar database as well as IARC TP53 database. Both these patients also had a complex karyotype along with RUNX1/RUNX1T1 fusion. Mutation in DNA methylation genes occurred more in CBFB/MYH11 vs RUNX1/RUNX1T1 cases (p<0.06). The patients with CBFB/MYH11 also depicted a higher prevalence of mutations in the polycomb repressor genes ASXL1 and ASXL2 when compared to RUNX1/RUNX1T1(p<0.06).Another feature was higher prevalence of cohesin complex gene mutations involving SMC1A and SMC3 which were seen in 20% cases overall. The inv(16) group showed a lower predilection compared to the t(8;21) group (p<0.061). Conclusions: The presence of higher number of additional mutations in the Indian cohort underscores the need of larger studies/controlled trials of this subset, which otherwise has been stratified as favorable. The access to standard of care like GO is a necessity and there is a need to formulate national/regional guidelines for the Indian subset.