Abstract
To explore new molecular diagnosis approaches for early detection and differential diagnosis of hepatocellular carcinoma (HCC), we analyzed genomic copy number variations (CNV) using plasma cell-free DNA from patients with HCC by next generation DNA sequencing. Plasma samples from 31 patients with HCC and 8 patients with chronic hepatitis or cirrhosis were analyzed. In HCC group, most samples with large tumor size (tumor dimension greater than 50 mm) showed CNVs that are visually recognizable at chromosome CNV plots, few samples with small tumor and none samples with chronic liver diseases showed CNVs recognizable at CNV plots. CNV Z score analysis showed significant CNVs in samples with HCC and chronic liver diseases although more significant changes were found in HCC group, some are differentially valuable (such as gain in 1q, 7q, and 19q in HCC), while others are less differentially valuable (such as loss in 4q, 13q, gain in 17q, 22q). We proposed a CNV scoring method that generated positive result in 26 of the 31 HCC patients (83.9%) or 11 of the 16 HCC with tumor dimension 50 mm or less (68.8%) or 4 of the 7 HCC with tumor dimension 30 mm or less (57.1%), while all the 8 samples with chronic hepatitis or cirrhosis scored negative. Ten HCC patients had normal or low serum AFP levels, among them, 7 were scored positive by CNV analysis, including 4 with tumor dimension 50 mm or less. Our study suggested that non-invasive genomic CNV analysis using plasma samples could be a valuable tool for early detection and differential diagnosis of HCC. Although CNV analysis itself cannot establish the diagnosis, it can help identify patients at high risk for HCC among patients with chronic liver diseases, which would prompt closer and more frequent surveillance for early tumor detection and intervention.