Fragmentomics profiling and quantification of plasma Epstein-Barr virus DNA enhance prediction of future nasopharyngeal carcinoma
W.K. Lam, Guannan Kang, Charles Ming Lok Chan, Vicky C.T. Lee, L Mary-Jane, Qing Zhou, Peiyong Jiang, Irene O L Tse, Ann D. King, Kenneth C.W. Wong, Edwin P. Hui, Brigette Ma, Anthony T.�C. Chan, K.C. Chan, Yuk Ming Dennis Lo
Abstract
Fragmentomics analysis of plasma autosomal DNA has shown promise in cancer diagnostics. Here we evaluated the clinical utility of plasma Epstein-Barr virus (EBV) DNA fragmentomics analysis for nasopharyngeal carcinoma (NPC) screening. Among our prospective cohort of approximately 20,000 subjects that underwent two rounds of screening, we analyzed the first-round blood samples of subjects who tested positive for EBV DNA via polymerase chain reaction (PCR) ( n = 558). We found that those who subsequently developed NPC in the second round exhibited a distinctive mononucleosomal size pattern, an NPC-associated end motif (specifically, a depletion of CC-motif) and aberrations in methylation identified through fragmentomics-based methylation analysis (FRAGMA). Subjects with these aberrant fragmentomics features and higher quantity of EBV DNA had a relative risk of 87.1 times greater for developing NPC in the second round compared to subjects tested negative for EBV DNA on PCR. These results demonstrate plasma DNA fragmentomics could predict future cancer risk. • There exists an NPC-associated fragmentomics profile of plasma EBV DNA • Plasma EBV DNA fragmentomics and quantitative analysis enhance NPC risk prediction • The analyses select those having 87 times the risk of others with negative EBV DNA • Fragmentomics-based cancer risk prediction is worthwhile in liquid biopsy research The analysis of fragmentation patterns in plasma DNA (i.e., fragmentomics) is actively explored. Lam et al. reveal that integrating fragmentomics and quantitative analysis of plasma Epstein-Barr virus DNA significantly improves the prediction of future nasopharyngeal carcinoma risk. This suggests the potential of fragmentomics-based risk prediction for other types of cancer.