Page 27 - 《南京医科大学学报》2026年第1期

第46卷第1期 南京医科大学学报（自然科学版）
2026年1月 Journal of Nanjing Medical University（Natural Sciences） · 21 ·

·专题研究：肿瘤·

基于 cfDNA 多组学模型进行头颈部鳞状细胞癌早期诊断的研

究进展

杨明哲，陈仁杰 *

南京医科大学第二附属医院耳鼻咽喉头颈外科，江苏 南京 210003

［摘 要］ 头颈部鳞状细胞癌（head and neck squamous cell carcinomas，HNSCC）是全球第六大常见癌症，其早期诊断因症状隐
匿和特异性标志物缺乏而存在困难。随着液体活检技术的发展，循环游离DNA（circulating free DNA，cfDNA）作为其核心检测
对象，具有非侵入性、动态监测和反映肿瘤特征的优势，在HNSCC早期筛查、预后评估和疗效监测中展现出应用前景。但现有
的单组学研究仍受限于低灵敏度和肿瘤异质性的影响。通过结合基因组、转录组、表观组、蛋白质组信息的多组学整合策略，
可显著提高诊断准确率，并揭示疾病分子机制。近年来，随着深度学习和机器学习技术被广泛运用于cfDNA的多组学数据分
析，推动了相关标志物筛选和模型构建。现有研究表明，基于cfDNA的多组学模型有望提高HNSCC的早期诊断水平，并为精
准医学的发展提供新方向。
［关键词］ 头颈部鳞状细胞癌；循环游离DNA；液体活检；多组学整合；深度学习；早期诊断；生物标志物；精准医学
［中图分类号］ R739.91 ［文献标志码］ A ［文章编号］ 1007⁃4368（2026）01⁃21⁃10
doi：10.7655/NYDXBNSN251049

Research progress on early diagnosis of head and neck squamous cell carcinoma based
on cfDNA multi⁃omics models

*
YANG Mingzhe，CHEN Renjie
Department of Otolaryngology and Head and Neck Surgery，the Second Affiliated Hospital of Nanjing Medical
University，Nanjing 210003，China

［Abstract］ Head and neck squamous cell carcinomas（HNSCC）rank as the sixth most common cancer globally，with early diagnosis
challenging due to subtle symptoms and the absence of specific biomarkers. With the development of liquid biopsy technology，
circulating free DNA（cfDNA），as its core test object，has the advantages of non⁃invasive，dynamic monitoring and reflecting tumor
characteristics，and shows promising applications in early screening，prognostic assessment and efficacy monitoring of HNSCC.
However，existing single⁃omics studies are still limited by low sensitivity and tumor heterogeneity. Multi⁃omics integration strategies
combining genomic，transcriptomic，epigenomic，and proteomic information can significantly enhance diagnostic accuracy and reveal
disease molecular mechanisms. In recent years ，as deep learning and machine learning technologies have been widely applied to
analyze multi⁃omics data of cfDNA，promoting relevant marker screening and model construction. Existing research indicates that multi⁃
omics models based on cfDNA hold promise for improving early diagnosis of HNSCC and offer new directions for precision medicine
development.
［Key words］ head and neck squamous cell carcinomas；cell⁃free DNA；liquid biopsy；multi⁃omics integration；deep learning；early
diagnosis；biomarkers；precision medicine
［J Nanjing Med Univ，2026，46（01）：21⁃30］

［基金项目］ 江苏省科技计划项目（BL2013038）
通信作者（Corresponding author），E⁃mail：crjent@njmu.edu.cn（ORCID：0000⁃0002⁃4768⁃7650）
∗