In vitro size-selection of short circulating tumor DNA fragments from late-stage lung cancer patients enhance the detection of mutations and aneuploidies

Introduction: Recent studies have demonstrated differences between the fragment length profiles of cell-free DNA (cfDNA) from cancer patients and healthy individuals. This has led to the development of in vitro size-selection procedures which can isolate the short fragments that are enriched with mutated circulating tumor DNA (ctDNA). This has yet to be investigated in a large cohort of lung cancer patients. Materials and methods: We used plasma samples from 35 stage III and IV lung cancer patients and performed targeted next-generation sequencing (NGS) and variant calling from cfDNA with and without size-selection of short fragments. We identified clonal hematopoiesis (CH) and germline mutations using targeted NGS on paired buffy coat (BC) samples. In addition, we performed a genome-wide copy-number alteration analysis on the cfDNA samples with and without size-selection. Results: ctDNA containing tumor mutations had a different fragment length profile compared to cfDNA fragments with CH or germline mutations. In vitro size-selection resulted in a median 1.36-fold (interquartile range (IQR): 0.63 to 2.48) mutational allele fraction (MAF) enrichment of tumor mutations whereas CH/germline mutations had a median 0.95-fold (IQR: 0.62 to 1.05) MAF enrichment. Key oncogenic drivers, including KRAS and EGFR were more likely to have a MAF increase with size-selection. Size-selection also increased the number plasma aneuploidy positive samples from 8 of 35 to 20 of 35. Conclusion: This study expands the knowledge regarding ctDNA fragmentation in lung cancer patients and we demonstrate that in vitro size-selection can increase MAF of tumor mutations and plasma aneuploidy calls. Size-selection could lead to increased sensitivity of ctDNA detection, which is crucial for clinical implementation of liquid biopsies. This study is the largest of its kind studying cfDNA samples from 35 lung cancer patients containing 109 mutations in total.