Targeted RNA Sequencing Is Capable of Identifying Thus Far Unknown Partner Genes in Leukemias with Rare Translocations and Provides Important Clinical and Therapeutical Information

Background: The genomic landscape of hematological malignancies has been resolved mainly based on whole exome and whole genome sequencing, primarily targeting gene mutations. Beside mutations also gene fusions function as therapeutic targets, impressively shown for e.g. BCR-ABL1 and ETV6-PDGFRB. Hence, the need for a comprehensive genetic analysis is increasing, as it is the basis for precision medicine, selecting treatment based on genotype and providing markers for disease monitoring.

Aim: To test the value of targeted RNA sequencing in a routine diagnostic work up.

Patients and Methods: 38 cases were selected in which rearrangements involving KMT2A (n=8), RUNX1 (n=19), ETV6 (n=9), RARA (n=1) and JAK2 (n=1) had been identified by chromosome banding analysis (CBA) complemented by FISH analysis. In all cases the partner gene could not be identified using standard methods. Targeted RNA sequencing was performed using the TruSight RNA Fusion panel (Illumina, San Diego, CA) consisting of 7690 probes covering 507 genes known to be involved in gene fusions. Library was prepared according to manufacturer's protocol with ~50ng DNA extracted from fresh/frozen samples. This assay allows the capture of all targeted transcripts. Sequencing was performed on two NextSeq runs (Illumina, San Diego, CA) with 20 multiplexed samples including two samples with known fusions as positive control samples. Analysis was performed with the RNA-Seq Alignment App (BaseSpace Sequence Hub) using Star for Alignment and Manta for gene fusion calling with default parameters (Illumina, San Diego, CA).

Results: In 22/38 cases with rearrangements involving KMT2A (n=8), RUNX1 (n=8), ETV6 (n=4), RARA (n=1) or JAK2 (n=1) this approach led to important new information: The following partner genes for KMT2A were identified: MLLT10 (n=2), MLLT1 (n=2), ITPR2, FLNC, ASXL2 and ARHGEF12. MLLT10 and MLLT1 are two of the most frequent partner genes of KMT2A, while KMT2A-ARHGEF12 fusions are rare. Fusion of KMT2A to ITPR2, FLNC, or ASXL2 have not been reported yet.

Seven different partner genes were identified in RUNX1 translocated cases. These were PLAG1 (n=2), PRDM16, MECOM, ZFPM2, MAN1A2, N6AMT2, and KIAA1549L. PRDM1, MECOM and ZFPM2 have previously been described in the literature as RUNX1 partner genes but were not suspected in our cases as partner genes due to complex cytogenetic rearrangements in CBA. The other identified partner genes have not been described so far. Interestingly, PRDM1, MECOM, ZFPM2 and the newly identified PLAG1 are all members of the C2H2-type zinc finger gene family.

Four different partner genes were identified in ETV6 rearranged cases: ABL1, CCDC126, CLPTM1L, and CFLAR-AS1. Most strikingly was the identification of the ETV6-ABL1 fusion, which could not be suspected by cytogenetics as the 5' ETV6 FISH signal was located on chromosome 7. This ETV6-ABL1 fusion was confirmed by conventional RT-PCR.

In an ALL patient a JAK2-PPFIBP1 fusion was identified leading to classification as a BCR-ABL1-like ALL. In an APL patient showing an ins(17;11)(q12;q14q23) in chromosome banding analysis a ZBTB16-RARA fusion was identified and thus resistance to all-trans retinoic acid, arsenic trioxide, and anthracyclines can be predicted.

All these fusions were not detectable by our routine RT-PCR analyses as these assays cover only the most frequently occurring breakpoints in fusions with known partner genes, but might miss very rare variants. For all yet undescribed fusion partners routine assays are not available. Based on the results of targeted RNA sequencing quantitative PCR assays for MRD monitoring can now be established.

In 11 cases with a RUNX1 rearrangement and 5 cases with an ETV6 rearrangement no fusion transcript was identified. Further analyses will have to clarify whether in these cases no transcript was derived from the genomic rearrangement.

Conclusions: 1) Targeted RNA sequencing was able to identify and characterize rare gene fusions and thus provided the basis for the design of RT-PCR based assays for monitoring MRD. 2) Targetable genetic aberrations were identified, which were not identifiable by chromosome banding analysis but would now lead to more individualized treatment. 3) Thus, targeted RNA sequencing may be a valuable tool in routine diagnostics for patients with rearrangements unresolved by standard techniques, also paving the way to precision medicine in a considerable number of patients.