Gene fusions are frequently associated with tumorigenesis malignant tumors. Highly sensitive and reproducible detection of gene fusions in clinical samples, especially formalin-fixed paraffin-embedded (FFPE) specimens, is important for diagnosis, prognosis, and treatment guidance. Although many clinical tests utilize next-generation-sequencing (NGS) for detection of genetic alterations, reliable detection of gene fusions by NGS in FFPE samples remains a challenge.
Five FFPE non-small cell lung cancer samples confirmed by FISH to contain EML4-ALK or KIF5B-RET fusions were selected for testing along with the Seraseq FFPE Tumor Fusion RNA Reference Material, v1, which evaluates 12 known fusions. Four operators prepared NGS libraries for each sample (89–250 ng input) using the commercially available QIAseq RNAscan Oncology Panel, which targets 223 fusion gene pairs or 576 unique breakpoints. The resulting libraries were sequenced on an Illumina NextSeq 500 sequencer. Demultiplexed FASTQ files were analyzed using cloud-based QIAseq RNAscan Analysis pipeline v188.8.131.52. Detected fusions that are on the panel’s target list were reported as curated gene fusions and assigned a score based on unique molecules supporting the transcript.
For the Seraseq FFPE Tumor Fusion RNA Reference v1 sample, all 12 expected known fusions were detected in each of the libraries generated by four different operators. The analysis pipeline correctly called the expected fusion in all four replicates of the five clinical FFPE samples tested, which varied in both input quantity amount and quality. Reproducibility was evaluated by comparing the score values across replicate libraries for each sample; expected fusion detection was 100% consistent between users, and the average correlation coefficient for the fusion score was 0.995.
The commercially available QIAseq RNAscan Oncology panel is a highly sensitive and robust product for detection of gene fusions in FFPE samples. QIAseq RNAscan methodology provides a useful tool for studying gene fusions in tumor samples and with optimization, customization, and validation can further fit the specific needs of clinical laboratories.
Jeffrey Falk1, Parth Sitlani1, Claire Orosc1, Maya Panjikaran 1, Dana Weiner1, Song Tian2, Raed Samara2, John DiCarlo2, Yexun Wang 2, Eric Lader2, Frank Reinecke3. 1 Genoptix, Inc., Carlsbad, CA; 2 Qiagen Sciences, Inc., Fredrick, MD; 3 Qiagen GmbH, Hilden, Germany