Detecting NTRK, ROS1, and ALK gene fusions in gastrointestinal tumor patients.

Abstract

619 Background: The STARTRK-2 trial is a potentially registration-enabling Phase 2 global basket trial of the tyrosine kinase inhibitor entrectinib in patients with solid tumors harboring NTRK1, NTRK2, NTRK3, ROS1, or ALK gene fusions. Phase 1 studies of entrectinib reported a 79% ORR across multiple histologies in patients with gene fusions of these targets who were naïve to inhibitors of these targets, received an effective dose, and had extracranial disease. Patients harboring these fusions are typically rare in the cancer population ( < 3%); however, they have been seen across a wide variety of GI tumors. In this presentation we report on the occurrence of these fusions in GI tumors, based on Ignyta’s internal and partner testing. We discuss the development of an assay for GI samples and immunohistochemistry (IHC) screening efficiency rates in a clinical cohort of 157 GI samples. Methods: The occurrence of NTRK, ROS1, and ALK gene fusions in GI tumors was studied from Ignyta internal and partner collaborations. We developed a 2-step diagnostic test to identify fusions in FFPE clinical specimens. IHC screening using a pan-receptor tyrosine kinase cocktail of antibodies precedes an RNA-based anchored multiplex PCR next generation sequencing (NGS) assay. Results: IHC positivity rates ranged from 0 to 24 percent depending on histology and averaging 8% positivity rate across 15 tumor locations. The screen is particularly effective in cholangiocarcinoma, colon, esophagus, gastric, peritoneum, rectum and small bowel. It was highest in pancreas (24%). Out of 157 GI samples, no instances were detected where the IHC was screened negative and gene fusions were observed by NGS. The IHC cocktail enriches detection of the patient population for gene fusions over 11-fold and has a 100% negative predictive value. Conclusions: This 2-step assay is an efficient method for detection of gene rearrangements in both high-volume clinical testing and studies of archival formalin-fixed paraffin-embedded specimens. Depending on how it is implemented in clinical trials, the staged approach can substantially decrease the costs and tissue expended for NGS. Screening efficiency may be improved even further as additional diagnostic antibodies become available. Clinical trial information: NCT02568267.