Abstract 1337P
Background
cfDNA has the potential to be used as diagnostic, predictive, prognostic and pharmacodynamic biomarker for NSCLC. Next Generation Sequencing of cfDNA is moving from the research setting to routine clinical use. As tumour genomics and burden will differ in the clinic from trials it is important to determine performance in this setting.
Methods
17ml of blood was taken from 230 patients with known or suspected NSCLC treated at The Christie and Newcastle upon Tyne Hospitals NHS Trusts from June 2019. Samples were transported to Foundation Medicine and analysed on a commercial hybrid-capture NGS platform for 70 oncogenes. Clinical details and rationale for test were extracted from clinical notes.
Results
Most patients were either untreated or had received 1 line of therapy (54% and 25% respectively). Histology was adenocarcinoma in 68%; squamous cancer in 13%, other in 9% and no histology in 10%. In 35% the rationale for testing was to avoid a contemporaneous biopsy (either failure of initial biopsy or inability to safely perform 15%, or rebiopsy on progression of targeted therapy 20%). Median time from collection to report was 13 days (range 4 to 21) with 85% received within 14 days; the failure rate was 5%, mostly due to cfDNA levels <20ng/ml. The number of abnormalities reported was 0 in 13%, 1 in 22%, 2 in 28% and 3 or more in 36% of patients. The median variant allele fraction (VAF) was 1.6% (range 0.11-91); however, 25% of cases had less than 0.5% VAF. Potentially actionable findings using the ESCAT scale were found for Tier 1: 22% (20% already known) Tier 2:14% and Tier 3: 4% of patients. 5 known fusions were not detected (4 ALK; 1 ROS-1) although 3 patients were on therapy at sampling.
Conclusions
In community practice results of cfDNA sampling are similar to reported in previous research series. 18% of patients either had failed analysis or no detectable abnormality. Only patients with EGFR mutations were able to access standard of care treatments but some patients were able to avoid biopsies or be routed to appropriate trials.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Experimental Cancer Medicine Centre Network.
Disclosure
All authors have declared no conflicts of interest.