Discovery of recurrent KIF5B-RET fusions and other targetable alterations from clinical NSCLC specimens.

Abstract

7510 Background: Many NSCLCs have driving oncogenic alterations including in EGFR, KRAS, ERBB2, BRAF, ALK and ROS1. Clinically effective drugs are approved for EGFR and ALK and clinical trials are underway for other genomic targets. Thus having a means of identifying genomic alterations in routine formalin fixed paraffin embedded (FFPE) clinical specimens is critical. Methods: We sequenced 24 FFPE NSCLC specimens with a next generation sequencing (NGS) assay that captures and sequences 2574 coding exons of 145 cancer relevant genes plus 37 introns from 14 genes often rearranged in cancer. Tumors from 643 additional patients were genotyped for KIF5B-RET. Results: We identified 50 alterations in 21 genes with at least one in 83% (20/24) of tumors (range 1-7). In 72% (36/50) of NSCLCs, at least one alteration was associated with a current clinical treatment or targeted therapy trial, including mutations in KRAS, BRAF, EGFR, MDM2, CDKN2A, CCNE1, CDK4, NF1 and PIK3CA. We also found an 11,294,741 bp pericentric inversion on chromosome 10 generating a novel gene fusion joining exons 1-15 of KIF5B to exons 12-20 of RET (K15:R12) in a Caucasian never smoker. This fusion gene contains the kinesin motor and coiled-coil domains of KIF5B and the entire RET tyrosine kinase domain. In 643 additional tumors we identified 11 fusion positive patients who were all wild type for known oncogenes (frequency of 6.3% (10/159)). Four unique KIF5B-RET variants were found: 8 K15:R12, 3 K16:R12, 1 K22:R12 and 1 K15:R11. We introduced K15:R12 into Ba/F3 cells and observed IL-3 independent growth consistent with oncogenic transformation. KIF5B-RET Ba/F3 cells were sensitive to sunitinib, sorafenib and vandetinib, multi-targeted kinase inhibitors that inhibit RET, but not gefitinib, an EGFR kinase inhibitor. Sunitinib, but not gefitinib, inhibited RET phosphorylation in these cells. Conclusions: We identified both known and novel genomic alterations from NSCLC FFPE specimens using a single test. Our findings suggest that RET inhibitors should be tested in prospective clinical trials in NSCLC patients bearing KIF5B-RET rearrangements and that NGS is a feasible approach to stratifying patients for treatment based on their genomic profiles.