Multi-feature next-generation liquid biopsy for diagnosis and prognosis in HPV-associated head and neck cancer.

作者：Ling Aye, Michael Bryan, Dipon Das, Shun Hirayama, Yana Al-Inaya, Julia Mendel, Saskia Naegele, Adam S. Fisch, William C. Faquin, Peter M. Sadow, Annie W. Chan, Michael S. Lawrence, Lisa Mirabello, John Iafrate, Tim Waterboer, Lori J. Wirth, Jeremy D. Richmon, Daniel L. Faden · 发表于：Journal of Clinical Oncology · 年份：2024 · DOI：10.1200/jco.2024.42.16_suppl.6064 · 被引用次数：13 · 研究领域：Molecular Biology Techniques and Applications、Cancer Genomics and Diagnostics

6064 Background: Human papillomavirus-associated head and neck squamous cell carcinoma (HPV+HNSCC) releases circulating tumor HPV DNA (ctHPVDNA) into the blood which is a highly accurate biomarker of disease status. Existing approaches based on droplet digital PCR (ddPCR) have limitations in sensitivity, genotype coverage, and do not annotate additional prognostic genomic features. In this study, we tested the diagnostic and prognostic performance of a custom HPV whole-genome next-generation sequencing (NGS) liquid biopsy, termed HPV-DeepSeek, at the time of diagnosis with HPV+HNSCC, compared to ctHPVDNA ddPCR, HPV serology, and clinical work-up. Methods: 304 participants (152 untreated HPV+HNSCC patients, 152 population-level controls) were prospectively enrolled in this single-center prospective cohort study. Plasma samples were analyzed using HPV-DeepSeek, a multiplexed serology assay, single-plex ddPCR, and a 5-genotype multiplexed ddPCR assay. We tested the hypothesis that HPV-DeepSeek would have the highest diagnostic accuracy. We secondarily examined the accuracy of HPV-DeepSeek for prognostic feature annotation including high-risk HPV16 single nucleotide polymorphisms (SNPs), integration events, and PIK3CA mutations. Results: Of 152 cases, 114 (75%) were AJCC stage 1 and 138/152 (91%) were oropharynx primary site cancers. The sensitivity (98.7%), specificity (98.7%), and diagnostic accuracy (0.974) of HPV-DeepSeek was superior (P<0.001) to singleplex ddPCR (94.2%, ...