Diagnostic Utility of Gene Fusion Panel to Detect Gene Fusions in Fresh and Formalin-Fixed, Paraffin-Embedded Cancer Specimens

Abstract BACKGROUND Anaplastic lymphoma receptor tyrosine kinase (ALK), ROS proto-oncogene 1, receptor tyrosine kinase (ROS1), and ret proto-oncogene (RET) fusions are present in 5%–7% of patients with advanced non–small-cell lung cancer (NSCLC); their accurate identification is critical to guide targeted therapies. FISH and immunohistochemistry (IHC) are considered the gold standards to determine gene fusions, but they have limitations. The nCounter platform is a potentially useful genomic tool for multiplexed detection of gene fusions, but has not been validated in the clinical setting. METHODS Formalin-fixed, paraffin embedded (FFPE) samples from 108 patients with advanced NSCLC were analyzed with an nCounter-based assay and the results compared with FISH, IHC, and reverse transcription PCR (RT-PCR). Data on response to fusion kinase inhibitors was retrospectively collected in a subset of 29 patients. RESULTS Of 108 FFPE samples, 98 were successfully analyzed by nCounter (91%), which identified 55 fusion-positive cases (32 ALK, 21 ROS1, and 2 RET). nCounter results were highly concordant with IHC for ALK (98.5%, CI = 91.8–99.7), while 11 discrepancies were found compared with FISH (87.5% concordance, CI = 79.0–92.9). For ROS1, nCounter showed similar agreement with IHC and FISH (87.2% and 85.9%), but a substantial number of samples were positive only by 1 or 2 techniques. Of the 25 patients deriving clinical benefit from fusion kinase inhibitors, 24 were positive by nCounter and 22 by FISH. CONCLUSIONS nCounter compares favorably with IHC and FISH and can be used for identifying patients with advanced NSCLC positive for ALK/ROS1/RET fusion genes.

Download Full-text

Fusion transcript discovery in formalin-fixed paraffin-embedded human breast cancer tissues and its relation to tumor progression.

Journal of Clinical Oncology ◽

10.1200/jco.2013.31.15_suppl.11018 ◽

2013 ◽

Vol 31 (15_suppl) ◽

pp. 11018-11018

Author(s):

Yan Ma ◽

Ranjana Ambannavar ◽

James Stephans ◽

Jennie Jeong ◽

Andrew Dei Rossi ◽

...

Keyword(s):

Breast Cancer ◽

Tumor Progression ◽

Clinical Outcomes ◽

Fusion Transcript ◽

Gene Fusions ◽

Fusion Transcripts ◽

Formalin Fixed Paraffin ◽

Formalin Fixed Paraffin Embedded ◽

Multiple Patients ◽

Formalin Fixed

11018 Background: While several recently discovered gene fusions already play an important role in personalized cancer treatment, many cancer gene fusions remain to be discovered. Next generation sequencing has enabled identification of many rare gene fusion events in fresh or frozen solid tumors. There is a need to detect gene fusions in transcriptomes of formalin-fixed paraffin-embedded (FFPE) tumor tissue, for which there is long-term clinical outcome data. We therefore sought to develop bioinformatics methods to detect fusion transcripts in FFPE tissue and to characterize their association with clinical outcomes. Methods: RNA sequencing libraries were created and sequenced from tumor biopsy tissues (Plos One 2012 7(7): e40092) of two ER+ breast cancer cohorts consisting of 136 and 77 patients, for which clinical outcomes were available. The fusion junctions were nominated by the RNA-seq aligner GSNAP and further filtered to consider discontinuous expression patterns at exon/intron levels. Results: A total of 108 candidate fusion transcripts were detected and RT-PCR assays confirmed 89% of the top ranking fusion transcript candidates. The majority (82%) of identified fusion gene partners are listed in the COSMIC database of known cancer sequence variations. Of note, several patients expressed multiple fusion transcripts that are significantly associated with tumor progression (P<0.001), including genes associated with cell proliferation and cellular metabolism. Furthermore, these patients also harbored inter-chromosomal gene fusions. It is noteworthy that several gene fusions were present in multiple patients. In one of these recurrent fusions the estrogen receptor gene acts as the fusion pair donor. Conclusions: Novel bioinformatics approaches developed here demonstrate the ability to detect fusion transcripts as biomarkers from archival FFPE tissues that associate with breast cancer progression. Some gene fusions were common in multiple patients and deserved further study.

Download Full-text