PS922 BENEFITS OF RNA SEQUENCING IN DETECTING RECURRENT AND NOVEL FUSION TRANSCRIPTS IN PATIENTS WITH ACUTE LYMPHOBLASTIC LEUKEMIA

Abstract

Background:Chimeric transcripts are frequent genetic abnormalities in hematological malignancies often contributing to leukemogenesis. So far, analysis of chromosomal translocations giving rise to gene fusions is based on cytogenetic techniques such as chromosome banding analysis (CBA) and fluorescence in situ hybridization (FISH). However, these techniques only detect known gene fusions, while cryptic or novel translocations remain elusive. The impact of gene fusions is increasing as they are not only entity defining and important for risk stratification, but also potentially useful biomarkers for MRD monitoring or targeted therapy. Recently, RNA sequencing (RNAseq) has become a powerful tool to identify disease associated transcriptome profiles and determine recurrent and novel genomic gene fusions underlying the disease.Aims:Comprehensive analysis of the fusion gene landscape in ALL (acute lymphoblastic leukemia) patients and evaluation of the benefits of RNAseq in detecting recurrent and novel fusion transcripts.Methods:We screened the whole transcriptome of 288 patients with ALL (B‐ and T‐lineage) for the detection of recurrent and novel fusion transcripts and applied a combination of whole genome sequencing (WGS) and cytogenetics (CBA, FISH, and genomic arrays) for validation. Gene expression profiling was performed to determine expression signatures of emerging ALL subtypes.Results:The analysis included 288 patients that were screened for chimeric transcripts by RNAseq. After stringent filtering and validation by WGS data and/or cytogenetic analysis a total number of 122 fusion transcripts, corresponding to 61 unique gene fusions in 102 ALL patients, was identified with high confidence. These included 93/122 (76%) gene fusions comprising breakpoints located on different chromosomes while the remaining 29/122 (24%) fusion genes were located on the same chromosome. Among detected fusion events, 90/122 (74%) gene fusions were confirmed by WGS and cytogenetic data, 25/122 (20%) cases were confirmed by WGS only and 7/122 (6%) cases by cytogenetics only. Besides recurrent entity‐defining fusion gene families such as BCR‐ABL1 (n = 30), KMT2A/MLL‐r (n = 22), ETV6‐RUNX1 (n = 4), and TCF3‐PBX1 (n = 4), RNAseq detected so far unreported gene fusions or novel fusion partners of known genes in 38 cases including novel partners for RUNX1, ABL1, and MLLT10. Among patients lacking canonical chromosomal translocations, rearrangements involving 3’ ZNF384 were identified in 8/92 patients (9%) with either EP300 (n = 7) or EWSR1 (n = 1) as 5’ fusion partner, constituting the largest subtype in this subgroup. All ZNF384‐rearranged cases were validated by RT‐PCR. Notably, in none of the cases involving EP300 as fusion partner, standard cytogenetic techniques were able to detect the respective rearrangement as all cases displayed a normal karyotype. ZNF384‐rearrangements were recently defined as novel ALL subtype, based on gene expression signature and clinical features. As they are furthermore potentially targetable by small molecule inhibitors, accurate identification of patients harboring these and other cryptic aberrations by RNAseq is pivotal.Summary/Conclusion:Our data determine the capability of RNAseq to detect known as well as cytogenetically cryptic and novel fusion transcripts in ALL leading to a better characterization of patients with unresolved rearrangements and highlight its immediate importance as a valuable tool in routine diagnostics.