Key points are not available for this paper at this time.
BACKGROUND AND AIMS: Early detection of primary liver cancer (PLC), including HCC, intrahepatic cholangiocarcinoma (ICC), and combined HCC-ICC (cHCC-ICC), is essential for patients' survival. This study aims to develop an accurate and affordable method for PLC early detection and differentiating ICC from HCC using plasma cell-free DNA (cfDNA) fragmentomic profiles. APPROACH AND RESULTS: Whole-genome sequencings (WGS) were performed using plasma cfDNA samples from 192 patients with PLC (159 HCC, 26 ICC, 7 cHCC-ICC) and 170 noncancer controls (including 53 liver cirrhosis LC or HBV-positive) enrolled in the training cohort. An ensembled stacked model for PLC detection was constructed using the training cohort. The model performance was assessed in an independent test cohort (189 patients with PLC 157 HCC, 26 ICC, 6 cHCC-ICC, 164 noncancer controls including 51 LC/HBV). Our model showed excellent performance for cancer detection in the test cohort (AUC: 0.995, 96.8% sensitivity at 98.8% specificity). It showed excellent sensitivities in detecting early-stage PLC (I: 95.9%, II: 97.9%), small tumors (≤3 cm: 98.2%), and HCC (96.2%) or ICC (100%). The AUC for distinguishing PLC from LC/HBV reached 0.985 (96.8% specificity at 96.1% specificity). Promisingly, our model maintained consistent performances during the downsampling process, even using 1X coverage data (AUC: 0.994, 93.7% sensitivity at 98.8% specificity). A separate model showed potential for distinguishing ICC from HCC (AUC: 0.776). CONCLUSIONS: Our model, outperforming previous reports at a lower cost by solely using low-coverage WGS data, exhibits excellent clinical potential for ultrasensitive and affordable detection of PLC and its subtypes.
Zhang et al. (Sun,) studied this question.