The Genetic And Epigenetic Abnormalities Of Plasma Cfdna As Liquid Biopsy Biomarkers To Diagnose Hepatocellular Carcinoma

Liver cancer is the second leading cause of cancer-related death in China and worldwide, where hepatocellular carcinoma (HCC) represents the major histological type. Previous studies have demonstrated that surveillance program combining serum marker AFP and ultrasound could greatly reduce liver cancer mortality. However, it is widely recognized that AFP has lower sensitivity for early stage of the disease and the specificity is also sub-optimal, limiting its application for early detection and timely intervention. Tumor-originated circulating cell-free DNA (ctDNA), harboring cancer-related genetic and epigenetic changes, provides new opportunity for non-invasive detection of liver cancer. In this study, we have performed parallel genetic and epigenetic profiling of cfDNA from Chinese hepatocellular carcinoma patients as well as healthy individuals by targeted bisulfite sequencing and by targeted ultra-deep sequencing. For methylome profiling, we first identified HCC-specific differentially methylated regions (DMRs) and then employed machine learning approaches to build diagnostic models to classify the plasma of HCC patients from that of healthy individuals. The training cohort consists of 148 hepatocellular carcinoma cases (median age of 63) and 84 healthy individuals (median age of 60). A random forest classifier achieved an AUC of 0.94±0.04 with 10-fold cross-validation using a panel of 21 DMR markers. Meanwhile, cfDNA mutation profiling achieved a sensitivity of 50.8% and a specificity of 95.3% in the training cohort, providing an inferior diagnostic performance compared to the methylation assay. Further analyses were performed in an independent validation cohort, including HCC patients (n=112) as well as healthy control (n=96). The cfDNA methylation model achieved a sensitivity of 82.9%, and a specificity of 93.8%; all stage sensitivity excluding BCLC stage 0 was 92.8%. On the other hand, the diagnostic model based on mutation profile achieved a sensitivity of 43.8% and a specificity of 97.9% in this cohort. Furthermore, we found that the methylation model had a sensitivity of 76.6% for early HCC (BCLC stage 0 + A), while serum AFP level (\u003e20ng/ml) had a sensitivity of 27.3%. In conclusion, our results suggest that cancer-derived abnormal methylation pattern of cfDNA provides promising biomarkers for the diagnosis of HCC with high sensitivity and specificity. Citation Format: Yuying Wang, Yupeng Wang, Ao Huang, Ruijingfang Jiang, Jianchao Zheng, Zhilong Li, Jiaxi Peng, Jianlong Sun, Chichuan Liu, Guanghui Yang, Jie Yuan, Xinrong Yang, Jian Zhou, Jia Fan. The genetic and epigenetic abnormalities of plasma cfDNA as liquid biopsy biomarkers to diagnose hepatocellular carcinoma [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 782.