Wang Siyuan, Peng Fan, Dang Miao, Jiao Huanmin, Zhang Huanqin, Zhou Kaixiang, Guo Wenjie, Gong Zhiyun, Guo Lin, Lu Renquan, Li Deliang, Liu Bingrong, Guo Xu, Xing Jinliang, Liu Yang
State Key Laboratory of Holistic Integrative Management of Gastrointestinal Cancers and Department of Physiology and Pathophysiology, Fourth Military Medical University, Xi'an, Shaanxi, China.
Department of Clinical Laboratory, Fudan University Shanghai Cancer Center, Shanghai, China.
Gut. 2025 May 7;74(6):961-970. doi: 10.1136/gutjnl-2024-333533.
Early detection of colorectal cancer (CRC) is crucial for improving the survival rates of patients.
We aimed to develop a novel strategy for early CRC detection using the fragmentomic features of circulating cell-free mitochondrial DNA (ccf-mtDNA).
Here, a total of 1147 participants, including 478 healthy controls (HCs), 112 patients with advanced adenomas (AAs) and 557 patients with CRC, were enrolled from five hospitals and plasma samples were collected for capture-based ccf-mtDNA sequencing.
Our data analysis revealed significantly aberrant ccf-mtDNA fragmentomic features in patients with CRC and AA when compared with HCs. Then, a CRC detection (CD) model was constructed based on the fragmentomic features of ccf-mtDNA from 246 patients with CRC and 168 HC in the training cohort, showing area under the curve of 0.9863, sensitivity of 92.68% and specificity of 93.45%. Both internal and two external validation cohorts demonstrated the excellent capacity of CD model in distinguishing patients with early-stage CRC from HCs, greatly surpassing the performance of serum biomarkers. Furthermore, our CD model can also detect patients with AA with a sensitivity of 79.35% in AA cohort 1 and 85.00% in AA cohort 2.
In conclusion, based on aberrant ccf-mtDNA fragmentomic features, a novel and non-invasive approach was established for the detection of patients with early-stage CRC or AA, with high performance.
早期发现结直肠癌（CRC）对于提高患者生存率至关重要。
我们旨在开发一种利用循环游离线粒体DNA（ccf-mtDNA）的片段组学特征进行早期CRC检测的新策略。
本研究从五家医院招募了1147名参与者，包括478名健康对照（HCs）、112名晚期腺瘤（AAs）患者和557名CRC患者，并收集血浆样本用于基于捕获的ccf-mtDNA测序。
我们的数据分析显示，与HCs相比，CRC和AA患者的ccf-mtDNA片段组学特征存在显著异常。然后，基于训练队列中246名CRC患者和168名HC的ccf-mtDNA片段组学特征构建了CRC检测（CD）模型，曲线下面积为0.9863，灵敏度为92.68%，特异性为93.45%。内部验证队列和两个外部验证队列均证明了CD模型在区分早期CRC患者和HCs方面的卓越能力，大大超过了血清生物标志物的性能。此外，我们的CD模型还可以检测AA患者，在AA队列1中的灵敏度为79.35%，在AA队列2中的灵敏度为85.00%。
总之，基于异常的ccf-mtDNA片段组学特征，建立了一种用于检测早期CRC或AA患者的新型非侵入性方法，具有高性能。
