Background: Acute myelogenous leukemia (AML) is characterized by recurrent variants that have implications in diagnosis, prognosis and management. We analyzed all variants that were detected in our patients since the inauguration of our Next-Generation-Sequencing (NGS) facility.
Methods: This is a retrospective study that analyzed all variants that were reported at our institution from Jan 2018 to Dec 2019 in patients with de-novo AML. NGS was performed using illumina miseq platform utilizing an amplicon panel that includes 50 genes known to be altered in AML. Variant Call Format (VCF) files were analyzed in a weekly meeting. For the purpose of this study, all VCFs were concatenated into a single MAF file. The Bioconductor package maftools was utilized to analyze this MAF file.
Results: 78 patients (24 children, 48 males, median-age=31 yrs) were studied. 193 variants were detected in 63 samples (81% of patients). The median number of variants-per-sample were 2 (range, 1 to 6). Most commonly mutated genes wereTET2(26%),NPM1(21%), FLT3(17%), DNMT3A(15%),NRAS(13%)CBLC, CEBPA, KIT(8%). The most commonly found nucleotide alteration was C>T transition which represented 50% of all mutations. Enrichment analysis showed the following significant associations (p<0.05):FLT3rearrangement associated withNPM1andDNMT3Aalterations, andRUNX1T1/RUNX1rearrangement associated withKITalterations. Utilizing the drug gene interaction database, 72 mutations were categorized as druggable, the most frequently listed categories were clinically actionable variants (N=17), kinase-dependent (N=8) and druggable genome (N=7). Using log-rank test,NRASandDNMT3Avariants were found to be associated with significantly worse EFS (p=0.017 and 0.038, respectively), whileNRASvariants were associated with significantly worse OS (p=0.015). The pediatric age group had some noticeable difference; NPM1andTET2were significantly more altered in adults whileKITwas significantly more altered in children. Finally, we compared our results to The Cancer Genome Atlas (TCGA) de-novo AML data for 192 patients that was published previously. The TCGA raw data is available publicly and were used run through our pipeline for comparison. A striking similarity in the frequency of alterations (81% vs. 82%) and the top mostly altered genes. The following genes were significantly more altered in our cohort:TET2,GATA2, andCSF3R;whileIDH2was more altered in the TCGA cohort. Interestingly,CBLC(n=7) andZRSR2(n=5) variants were identified in KHCC patients but none in the TCGA cohort.
Conclusion: We reported variants in 81% of our patients. Prognosis is significantly altered in patients withDNMT3AandNRASvariants. A wealth of actionable mutations were detected. Our results carry many similarities to the TCGA cohort, but significant difference in alterations involving 6 genes were reported. Our study lay the foundation for future studies that can evaluate big-data results obtained from our growing NGS facility.
No relevant conflicts of interest to declare.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal