Detection of Known and Novel FGFR Fusions in Non–Small Cell Lung Cancer by Comprehensive Genomic Profiling

Abstract Introduction Activation of the fibroblast growth factor receptor (FGFR) family through fusion with various partners has been described in multiple cancer types, including NSCLC. FGFR inhibitors are currently being evaluated clinically for patients whose tumors harbor these fusions. Methods Hybrid capture–based comprehensive genomic profiling was performed on 26,054 consecutive formalin-fixed, paraffin-embedded specimens of NSCLC. Results FGFR fusions retaining the kinase domain were identified in 0.2% of NSCLC cases; they included 37 fibroblast growth factor receptor gene 3 ( FGFR3 ) – transforming acidic coiled-coil containing protein 3 gene ( TACC3 ) fusion–positive cases, two fibroblast growth factor receptor 2 ( FGFR2 )–shootin 1 gene ( KIAA1598 [also known as SHTN1 ]) fusion–positive cases, one BCL2 associated athanogene 4 gene ( BAG4 ) – fibroblast growth factor receptor 1 gene ( FGFR1 ) fusion–positive case, and 12 novel FGFR1, FGFR2, FGFR3, and fibroblast growth factor receptor 4 gene ( FGFR4 ) fusion–positive cases. Co-occurring EGFR or MNNG HOS Transforming gene ( MET ) alterations were observed in 8% of cases (four of 52), KRAS mutation was observed in three additional cases, and FGFR1 or FGFR3 amplification was observed in 10% of cases. The two patients with co-occurring EGFR mutations were previously treated with EGFR inhibitors. One patient with a novel FGFR2– leucine zipper transcription factor like 1 gene ( LZTFL1 ) fusion had a partial response to the pan-FGFR inhibitor JNJ-42756493 and remained progression-free for 11 months. Conclusion FGFR fusions were detected by using comprehensive genomic profiling in 0.2% of NSCLCs; they occurred primarily in the absence of other known driver alterations, or in a subset of cases, as likely mechanisms of acquired resistance. One patient with a novel FGFR2 fusion had clinical benefit from an investigational FGFR inhibitor, suggesting that these alterations may predict response to targeted therapies.