Systematic analysis of actionable gene fusions by NGS-based in pancancer for precision oncology in China.

Abstract

e15072 Background: Detection of gene fusions with NGS is widely used for clinical diagnosis and drug development. To deeply understand clinically actionable gene fusions for implementation of precision oncology, we systematically analyzed the distribution of tumor types of each fused driver gene, and other mutation types in the same sample which were accompanied. Methods: Samples were performed with NGS based 450 tumor genes panel assay. Genomic alterations, including single nucleotide variations (SNV), short and long insertions/deletions (InDels), copy number variations and gene fusions were analyzed using a published algorithm. Results: We collected 1740 patients with 1478 known fusions and 338 novel fusions. Of all the patients, 48.5% were male, 6.5% were TMB-H (≥ 10 muts/Mb) and 0.6% were MSI-H. We found that breast carcinoma (BRCA) accounts for 28.6% of the ERBB2 fusions, intrahepatic cholangiocarcinoma (ICC) accounts for 60.4% of FGFR2 fusions, non-small cell lung cancer (NSCLC) accounts for 45.5% of FGFR1, 41.5% of FGFR3, and 87.3% of ALK fusions, respectively, renal cell carcinoma (RCC) accounts for 65.9% of TFE3 fusions, the soft tissue sarcoma (STS) accounts for 30.3% of NTRK1 fusions, 70.6% of PDGFB fusions, and 75.6% of SS18 fusions, respectively. We identified EGFR fusions in specific tumor types such as colorectal carcinoma (CRC) and Glioma (GBM), MET fusions in gallbladder carcinoma (GBC), PDGFRA fusions in NSCLC, BRAF fusions in ICC and FGFR3 fusions in urothelial carcinoma (UC). About novel fusions, many of them were detected in multiple tumor types, for example, PHF20-NTRK1 was detected in CRC, NSCLC and ovarian carcinoma (OC), FGFR2-INA was found in ICC and NSCLC, FGFR2-PAWR was detected in BRCA and ICC. Some partner genes were shared by multiple driver genes. For example, TTC28 presented as partner of FGFR2 in ICC, of NRG1 in ICC, of NTRK3 in hepatocellular carcinoma (HCC). Some novel fusions were found more than twice in one tumor type, for example, CMAHP-ALK was detected twice in NSCLC and EHBP1-MET was detected three times in ICC. These novel fusions are helpful to drug development and molecular diagnosis. Some co-occurring variations with fusions often existed in same patient. For example, patients with ALK fusion were also detected SNV of ALK such as L1196, F1174, and G1202. In addition, amplification of ERBB2 and EGFR were both 1.7% in patients with fusions, and CDKN2A/B deletion occupied 5.1% often with EML4-ALK and FGFR3-TACC3. Conclusions: In our study, many new novel fusions and known fusion in new tumor type were revealed. The regular pattern of co-occurring variations of fusion helps to improve the fusion detection rate and clinical application. This simultaneous detection of multiple meaningful variants is the advantage of NGS methods over other traditional methods. All of the above will helpful for improving the clinical benefit rate and accelerating target drug development.