Comprehensive Analysis of Mutational Landscapes in Acute Myeloid Leukemia, Myelodysplastic Syndromes, and Chronic Myeloproliferative Neoplasms: Insights from a Myeloid NGS Panel on DNA Variants, Gene Fusions, and Gene Expression

Background

Next-generation sequencing (NGS) multigene panels have revolutionized the detection of actionable driver genes in myeloid malignancies. Now essential to modern hematology practice, these panels provide extensive real-world data that enhance evidence collection. This study examined a database containing results from an NGS myeloid panel, exploring correlations between detected DNA variants, gene fusions, and changes in gene expression in cases of known/suspected myeloid malignancies.

Methods

A retrospective analysis was conducted on a de-identified dataset from myeloid NGS panel tests applied at the time of diagnosis for myeloid malignancies during 2023. Cases indicated for Acute Myeloid Leukemia (AML), Myelodysplastic Neoplasm (MDS), and Chronic Myeloproliferative Neoplasms (CMN) testing were identified, and relevant demographic, specimen, and test outcome data were retrieved from the test result database. The applied myeloid NGS panel was AmpliSeq Myeloid Panel, performed on the MiSeq system (both from Illumina). This panel targets 40 DNA genes (17 full-length and 23 hotspots), 29 RNA (629 potential fusions), and the expression levels of 5 genes: MECOM, BAALC, WT1, MYC, and SMC1A. The validated limits of detection were 5% allele frequency for DNA targets and 5% relative amount for RNA fusions. Demographic, specimen, and test outcome data were analyzed and compared using the Mann-Whitney U test, chi-square test, and multiple correlation analysis, with additional insights provided by a perceptual map, as appropriate.

Results

157 cases were evaluated, 57 AML (24 males), 44 CMN (25 males), and 56 MDS (29 males), with no significant gender distribution difference (p=0.31). The average ages were 56.68 ± 21.20 years for AML, 58.07 ± 17.05 years for MDS, and 65.43 ± 17.95 years for CMN, showing statistically significant age differences (p=0.036, ANOVA). Distribution of bone marrow samples was 26, 5, 22, and for whole blood, it was 31, 39, 34 for AML, CMN, and MDS, respectively (p=0.0008). Distribution of gene fusion events across AML, CMN, and MDS was 33% (19/57), 2% (1/44), and 0% (0/56), respectively (p<0.0001). The most common fusions in AML were CBFB::MYH11 (47%, 9/19) and RUNX1::RUNX1T1 (21%, 4/19). Perceptual maps further clustered AML closely to all fusion events, except from BCR::ABL, that clusters with CMN. The presence of at least one somatic DNA variant was observed in 87% (50/57) of AML cases, 65.9% (29/44) of CMN cases, and 62.5% (35/56) of MDS cases (p=0.005). A total of 350 actionable, prognostic or classificatory DNA variants were detected. Perceptual maps positioned AML, CMN, and MDS in different quadrants. AML clusters with FLT3, IDH2, KRAS, NPM1, NRAS, PTPN11, and WT1. CMN clusters with BCOR, JAK2, MPL, and PRPF8. MDS clusters with CALR, EZH2, SF3B1, SETBP1, SH2B3, and SRSF2. The following genes clustered at the center among the three conditions: ASXL1, CSF3R, DNMT3A, NF1, RUNX1, TET2, TP53, U2AF1, and ZRSR2. Differences in gene expression across AML, CMN, and MDS include low MECOM expression in AML at 41.3% (19/46), compared to 18.2% in CMN (6/33) and 11.4% in MDS (5/44) (p=0.0173). High BAALC expression was observed in AML at 67.4% (31/46), with none in CMN (0/33) and 9.1% in MDS (4/44) (p<0.0001). Similarly, high WT1 expression was noted in AML at 67.4% (31/46), compared to 4.5% in CMN (2/44) and 4.3% in MDS (2/46) (p<0.0001). No significant differences were observed for MYC (p=0.681) and SMC1A (p=0.081) across AML, CMN, and MDS.

Conclusion

Despite variations in age distribution, the demographic data of AML, CMN, and MDS groups are comparable, providing a robust foundation for molecular comparisons. AML is notably marked by unique gene fusions like CBFB::MYH11 and a higher prevalence of somatic DNA variants. In contrast, CMN and MDS exhibit distinct genetic signatures driven by mutations in specific genes such as JAK2 for CMN and SF3B1 for MDS. Shared molecular mechanisms, evidenced by common mutations in genes like ASXL1 and TET2 across all conditions, suggest potential pathways influencing disease progression. Altered expression of MECOM, WT1, and BAALC in AML underscores the diagnostic importance of gene expression profiles. These findings, corroborated by experimental data, highlight the critical value of comprehensive databases in advancing our understanding of myeloid malignancies, providing crucial insights for targeted therapies and diagnostic advancements.

No relevant conflicts of interest to declare.