Development and Validation of an Rnaseq NGS Assay to Comprehensively Detect Clinically Relevant ALL Gene Fusions

Background

Recent advances in molecular technologies have identified a subtype of B-progenitor acute lymphoblastic leukemia (B-ALL) with a Philadelphia chromosome-like (Ph-like or BCR-ABL1-like) gene expression profile without the BCR-ABL1 fusion. The subtypes containing Ph-like fusions are considered the more aggressive subtype. Although FISH and RT-PCR assays can be used to detect these fusions, these technologies are labor intensive and low-throughput in nature. An optimal solution is to utilize RNA-sequencing for high-throughput detection of characterized and novel fusions with high accuracy and sensitivity in a clinical setting. Here we report the development and characterization of a fusion panel targeting 115 genes of which 23 are specifically designed to detect fusions of clinical relevance in ALL. The genes targeted in the ALL panel include: ABL1, ABL2, BCR, CRLF2, CSF1R, ETV6, IL2RB, IL3, JAK2, KMT2A, MEF2D, MLLT109, NUP98, PAX5, PDGFRB, PTK2B, RUNX1, TAL1, TCF3, TLX1, TLX3, TYK2, and ZNF284.

Materials & Methods

Samples (peripheral blood, bone marrow, or fixed cell pellets) were selected based on FISH or molecular analysis that contain gene fusions in the ALL panel. RNA was extracted from these samples using the Qiagen AllPrep or the MagNA Pure isolation platform. A custom ArcherDX Anchored Multiplex PCR (AMP) assay was designed to enrich for a set of oncology-related genes including ALL relevant genes using 250ng input RNA. The enriched libraries were pooled and sequenced on an Illumina NextSeq500. Analysis was performed using Archer Analysis software, version 5.1. Additional filter settings were employed to reduce the false-positive rate and increase confidence in true-positive calls. The passing fusions were parsed through Cartagenia and N-of-One clinical interpretation for annotation and patient reporting. The accuracy, specificity, analytical sensitivity, precision, and reproducibility of the RNASeq assay were evaluated.

Results

Studies utilizing clinical patient samples as well as synthetic G-block constructs containing specific clinically relevant fusion break points were performed using the custom RNA FusionPlex panel to detect the fusions. The criteria for successful fusion detection were established at 10 unique molecules supporting the fusion with 3 unique start sites. These studies clearly demonstrated that the custom RNA FusionPlex test was able to successfully detect fusions in the ALL genes contained in the fusion panel. In addition, the fusions detected included most of the relevant fusions that have been reported previously by either FISH or RT-PCR testing, demonstrating that the custom RNA FusionPlex test is effective in detecting clinically relevant fusions. The accuracy of the assay was >90%, with precision and reproducibility at >95%. The false positive rate of the assay using these criteria was observed to less than 2%. Additional metrics and criteria will also be presented.

Conclusions

The studies presented clearly demonstrate that the Genoptix custom RNA Fusionplex test is capable of detecting most clinically relevant ALL gene fusions in patient samples in a reproducible and effective manner. The Genoptix ALL panel is a highly sensitive and robust product for detection of gene fusions in a clinical setting and presents a viable alternative to traditional FISH and RT-PCR testing, capable of delivering comprehensive patient gene fusion information.