Value of comprehensive genomic profiling in pre-screening patients for NTRK fusion in STARTRK2 trial: Single centre experience

Abstract

Background: Comprehensive genomic profiling (CGP) by next-generation sequencing (NGS) is increasingly used as a pre-screening tool for clinical trials. The aim of this project was to retrospectively determine the scope of alterations identified by CGP that could render patients suitable for alternative early phase clinical trials of genomically-matched (GM) / immunotherapy (IO) or ‘off-label’ drug use. Methods: Patients were pre-screened for the STARTRK2 study (Roche sponsored study of Entrectinib, NCT02568267) at The Christie NHS foundation Trust using FoundatioCDx. Testing is validated for NTRK, ROS1 and ALK fusion testing but all pathogenic alterations are reported on a clinical trial specific Foundation Medicine (FM) report. Results were scrutinised for actionable alterations that could direct patients to alternative clinical trials or off label drug use. Results: A total of 269 patients with were consented since FM testing was introduced in the trial in Jan’2019. FM yielded results in 229 patients (85.2%), mean age was 54, 58.4% were male and 45.8% had 1-2 prior systemic lines. Most prevalent tumour subtypes were colorectal (26.4%), head and neck (21.6%) and sarcomas (7.1%). Most prevalent alterations occurred in: TP53 (12.6%), APC (8.4%) and KRAS (4.6%). MSI High was 1.5%. No patients had NTRK/ROS1 fusions, 1 non-small cell lung cancer patient had ALK fusion. 104 (45.4%) patients were potentially eligible for matched clinical trials (101 for GM and 3 for IO) and 61 (26.6%) patients could have been considered for off-label drug use. The most prevalent actionable alterations found across common and rare disease types were PI3KCA (10%), ERBB2 (6.1%), PTEN (3.1%), tumour mutation burden 10 mut/Mb (2.6%) and HRAS (1.7%). The following alterations occurred in <1%: AKT1, KRAS G12C, BRAF V600E, BRCA, FGFR3 and IDH1. Conclusions: Our results highlight the relevance of CGP in identifying patients for GM or IO within clinical trials or off-label drug use. The retrospective nature of this work and the fact that FM results provided within STARTRK2 are not intended for clinical use precluded implementing these recommendations. NTRK fusions were not detected in our cohort which highlights the rarity of this event in our population