P1.02-071 Detection of Multiple Low-Frequency Mutations by Molecular-Barcode Sequencing

Abstract

Recent advantage of next-generation sequencing (NGS) in the field of cancer genome research has revealed extreme levels of genetic heterogeneity, suggesting the major roles of subclonal mutations in cancer relapse and in rapid emergence of acquired resistance. Unlike inherited mutations, somatic variants often occur at low allele frequencies that require sensitive methods for detection. Based on the results using another highly sensitive method, such as digital PCR, the study of cancer subclones requires to detect mutations that are present in <1% frequency, however, such a level of resolution cannot be obtained by conventional NGS approaches. In the current study, we performed molecular-barcode sequencing for primary lung adenocarcinoma cases in order to analyze mutations with <1% low frequency. Fresh frozen tumor samples from 58 primary lung adenocarcinoma patients whose tumors previously tested positive for EGFR by conventional method (the PNA-LNA PCR clamp method) were collected. All samples were obtained from surgically resected specimen. We used HaloPlex HS method, which is a high sensitivity amplicon-based targeted sequencing method incorporating molecular barcodes in the DNA library, allowing for the identification of duplicate reads to significantly improve the base calling accuracy even at low allelic frequencies compared to conventional NGS methods. We used a panel designed for 47 cancer-related genes including EGFR, and sequenced data was obtained by using Illumina Miseq Reagent v3 (600 Cycle). Normal tissue samples were also sequenced for threshold adjustment. Out of 58 samples, EGFR ‘major’ mutation (L858R, Exon19 deletion, G719A/S, T790M) profiles of 57 samples by molecular-barcode sequencing corresponded to those by clinical method (98.3%). The mean coverage of EGFR major mutation was 3078 (243-8285), and the minimal detectable frequency was 0.45%. Of note, we could detect the minor frequency of T790M mutation in addition to the major frequency of L858R mutation, at the allele frequency of 1.92% (20/1042) for T790M and 10.62% (155/1459) for L858R mutation, respectively. Minor genetic alterations, except major mutation, in EGFR were detected in 82.8% (48/58) cases, and the mean number was 2.3 (0-11). These results suggest the clinical applicability of this method. We could demonstrate good concordance rate between clinical examination and molecular barcode sequencing. Minor genetic alterations in EGFR were detected in 82.8% (48/58) cases. Further investigations are warranted to establish the confident detection of subclonal mutations with low frequency.