P1.14-56 The Dynamic Range of Mutant Allele Fraction Detected in ctDNA from Patients with NSCLC: Clinical Experience Data and Clinical Implications

Abstract

The detection of both activating and resistance mutations in the circulating tumor DNA (ctDNA) fraction of total cell free DNA (cfDNA) isolated from plasma in patients with NSCLC has gained widespread adoption and clinical practice guideline recommendations. However, the vast majority of cfDNA in blood samples is derived from non-cancerous tissue or white blood cells (WBCs). In patients with NSCLC, ctDNA must be distinguished from cfDNA. Studies have suggested that the contribution of non-tumor genetic material in circulation is greater than 90% whilst the DNA derived from tumor or circulating tumor cells (CTCs) is < 1%. Hence the ability to accurately identify mutations in patients with NSCLC, who would potentially benefit from targeted therapy as well as for monitoring for both emergence of resistance, recurrence and tumor burden, require assays to consistently and reliably detect low level mutations. Additionally, blood collection tubes should maximally stabilize cells to prevent excess breakdown of WBCs that would significantly increase the amount of cell free non-tumor DNA in the plasma fraction leading to potential false negative results for mutations analyzed in patients with NSCLC. Patient samples were collected in CEE-Sure™ blood collection tubes to minimize the non-tumor cfDNA content. Plasma was removed from the blood samples and circulating nucleic acid was extracted. The plasma fraction was then used in the highly sensitive EGFR, BRAF, and KRAS Target-Selector™ assays. The Target-Selector™ assays utilize forward and reverse primers and a Target-Selector™ probe to specifically block wild-type amplification of cfDNA, and selectively enrich for mutant sequences (ctDNA). Sanger sequencing of the amplified Target-Selector™ product is used to confirm presence of the mutation. A total of 1410 NSCLC patient samples were analyzed for EGFR, 877 for BRAF and 300 for KRAS. Overall mutations were detected in 28% of cases. The dynamic range of the mutant allele frequency (MAF) was 0.05% to 77%. For EGFR mutations, 23% of total mutations per sample were detected at <1% and for individual EGFR alterations (del19, L858R, T790M) within a sample, 74% were detected at <1% MAF. 37% of cases were positive for BRAF and 41% positive for KRAS at <1% MAF. 3% of EGFR mutant NSCLC, 10% of individual EGFR alterations and 11% of BRAF mutations were detected at <0.1% MAF. The consistent detection of mutations at a lower MAF of <1% corresponds to previously published biological thresholds of ctDNA in patients with cancer against an excess of non-tumor cfDNA. The ability of an assay to detect mutations consistently at <1% and <0.1% as demonstrated, is an important performance characteristic, as 26%, 47% and 41% of patients with NSCLC harbor mutations at <1% for EGFR, BRAF and KRAS respectively. This has potential clinical implications regarding patient qualification for targeted therapies and monitoring. The blocking of amplification of non-tumor cfDNA component allows for a more sensitive analysis of ctDNA.