Use of NGS for stratification of patients with advanced NSCLC within the NHS using FFPE-extracted DNA from diagnostic biopsies

Abstract

Background: The Cancer Research UK Stratified Medicine Programme 2 (SMP2) screens NSCLC cancer patients by NGS for clinically-actionable gene aberrations that are likely drug targets. Eligible patients can be recruited to a targeted therapy or ‘Non Actionable (immunotherapy) arm of the National Lung Matrix Trial (NLMT). SMP2 aims to establish genomic screening at a national level within existing NHS pathways and ensure equity of access to targeted therapies to eligible patients in the UK. Methods: The CRUK SMP2 pre-screening programme recruits patients with late stage NSCLC and utilises DNA from excess diagnostic biopsy tissue. Regions of interest are enriched using a custom 28-gene Illumina Nextera Rapid Capture panel before sequencing on MiSeq using V3 chemistry. Validation of the assay involved cell lines, blood and FFPE samples with known aberrations. This comprehensive validation evaluated a number of different factors, including mean target coverage, detection limits, concordance, tumour content cut-off, number of reads required for variant calling, and number of base pairs on either side of the region of interest to be analysed. This validation confirmed that the panel and analytics are able to identify SNV (5% VAF threshold), indels, translocations, amplifications (>5 copies) and deletions (<0.5 fold). The panel confirms wildtype for genes where no aberration is present. CNVs and SVs have been validated against FISH. The analytical pipeline was developed in-house at West Midlands Regional Genetics Service to accommodate a busy diagnostic service, and includes a reporting algorithm for determining confidence in calling wildtype, based on tumour content of the sample, and read depth and coverage achieved. Results: Greater than 3000 patients have been consented and greater than 2000 patients successfully screened through SMP2 over the past two years. QC fail rates have significantly reduced, as optimum sample selection improved throughout the network. CNVs and SVs called by NGS showed good concordance with FISH, although low tumour content and poor quality samples could result in low level CNVs being missed. Preliminary analysis indicates prevalence and distribution of SNVs consistent with published reports, including 33.5% KRAS and 16.1% EGFR mutations. The SMP2 testing panel has enabled the recruitment of > 150 patients to NLMT, with 31% of samples sent for screening eligible for one of 20 biomarker-specific NLMT cohorts, which rises to 51% when the immunotherapy arm is included. Conclusions: Delivering a nationwide screening programme within the NHS has raised a number of challenges across the entire patient and sample pathway: from patient recruitment, to sample preparation, right the way through to analysis of the results. Lessons learned will now enable informed updates to the NGS panel and analysis pipeline. Together with improvements to pre-analytical sample processes, this will ultimately increase the number of molecularly eligible patients that can benefit from novel targeted therapies. Clinical trial identification: The National Lung Matrix Trial Legal entity responsible for the study: NA Funding: Cancer Research UK with support from partners AstraZeneca and Pfizer, and in collaboration with Illumina. Disclosure: All authors have declared no conflicts of interest.