Role of Chromosomal Microarray and RNA Fusion Analysis in Detecting KMT2A Partial Tandem Duplication with High Mortality in Myeloid Neoplasms: A Retrospective Study

KMT2A partial tandem duplication (PTD) is a recurrent genetic alteration in myeloid neoplasms, conferring poor prognosis. KMT2A -PTDs are complex gene rearrangements that cannot be fully ascertained using a single genomic platform. There is no universally accepted gold standard for detecting KMT2A -PTDs due to the technical limitations of individual methods and the genomic complexity of PTDs. A retrospective analysis was performed at Fox Chase Cancer Center to compare the utility of chromosomal microarray (CMA; Affymetrix CytoScan HD, 2,696,168 markers) and RNA fusion analysis (Illumina TruSight 523-gene RNA Fusion Panel) in 97 specimens (79 bone marrow, 18 peripheral blood) from 17 patients (12 males, 5 females; median age 68 years) with KMT2A -PTD-positive myeloid neoplasms (11 AML, 4 MDS, 2 MPN), diagnosed between January 2011 and February 2025. A total of 31 specimens were confirmed as KMT2A -PTD-positive by chromosomal microarray (CMA; n = 24) or RNA fusion panel (n = 21). The 73.3% overall concordance (κ = 0.467) between CMA and RNA fusion, with 45.2% KMT2A -PTD-specific concordance, underscores their complementary roles. RNA fusion demonstrated superiority in discordant cases (p = 0.035), reflecting its sensitivity for cryptic PTDs, while CMA identified non-canonical PTDs (e.g., Patient 16) and secondary genomic abnormalities (e.g., trisomy 8, CN-LOH in patient 14). The KMT2A-PTD-positive group exhibited greater genomic complexity, with 17 secondary chromosomal abnormality types compared to 6 in the PTD-negative group, reflecting increased clonal heterogeneity. RNA fusion analysis revealed significantly higher median split reads (p = 0.032) and split read-to-coverage (SR/C) ratios (p = 0.046), suggesting elevated KMT2A -PTD transcript expression in CMA KMT2A -PTD(+) patients lacking additional chromosomal abnormalities (vs. CMA KMT2A -PTD (−) patients with abnormalities, p = 0.014-0.020). This may reflect purer PTD clones or fewer competing genomic events. Chromosomal microarray analysis (CMA) identified genomic coverage of KMT2A exons 2-20 (except in patient 16). RNA fusion analysis showed consistent breakpoints between exon 10 and exon 2 (16,200 bp) and between exon 11 and exon 2 (13,720 bp). These disruptions affect critical KMT2A oncogenic domains: CXXC, PHD, and Bromodomain. Targeted 275-gene DNA-sequencing NGS panel analysis (n=60/97 specimens) revealed specific mutational profiles: DNMT3A (51.67%), TET2 (30.00%), ASXL1 (21.67%), RUNX1 (41.67%), GATA2 (15.00%), JAK2 (25.00%), FLT3 (13.33%), NRAS (20.00%), PIK3CA (11.67%), EZH2 (18.33%), and KMT2A (13.33%). RNA fusion analysis revealed significantly higher median split reads and SR/C ratio of KMT2A -PTD in specimens with DNMT3A (p ≈ 0.0423, 0.0387), RUNX1 (p ≈ 0.0198, 0.0234), and FLT3 (p ≈ 0.0087, 0.0123). Similarly, significant correlations with RUNX1T1 gain (p ≈ 0.0345, 0.0412) and cnLOH (p ≈ 0.0214, 0.0178) indicate higher PTD expression with RUNX1T1 gain and lower expression with cnLOH.

Kaplan-Meier analysis estimated 41.2% 6-month overall survival (OS). Hematopoietic stem cell transplantation (HSCT) improved outcomes (68.6% vs. 0% OS at 17 months, p = 0.028). Non-transplanted patients (10/17, 8 deceased) had a median survival of 6 months. Abnormal karyotypes and FISH-detected KMT2A gains were associated with high-risk profiles. Concurrent FLT3 mutations correlated with 100% mortality (p=0.0456). CMA and RNA fusion analysis are complementary, with RNA fusion excelling for cryptic PTDs and CMA for genomic alterations.

Optimal management of myeloid neoplasms harboring KMT2A -PTD is achieved through the integration of combined testing methodologies, next-generation sequencing, and hematopoietic stem cell transplantation.