A blood-based three-gene signature for the non-invasive detection of early human hepatocellular carcinoma.

BACKGROUND:Identifying early stages of disease in high-risk individuals for the development of hepatocellular carcinoma (HCC) would greatly improve the clinical outcomes of these individuals. The aim of this study was to develop a blood-based gene set that could identify early-stage HCC. METHODS:Comprehensive gene expression profiling of purified RNA of peripheral blood mononuclear cells (PBMC) was performed using microarrays. Gene signatures were developed through bioinformatics-driven approaches and their diagnostic value was evaluated by custom-designed, quantitative, multiplex polymerase chain reaction (PCR) assays. RESULTS:Bioinformatics-driven analysis of microarray data derived from PBMC RNA samples of patients with HCC (N=10), pancreatic cancer (N=3), gastric cancer (N=3) and 10 normal individuals identified six genes that were differentially expressed in HCC. Subsequent multiplex-PCR validation and univariate analyses performed with an independent cohort of 114 HCC patients, 48 normal individuals and 14 patients with chronic hepatitis B (CHB) validated that three genes, namely Chemokine (C-X-C motif) receptor 2 (CXCR2), C-C chemokine receptor type 2 (CCR2) and E1A-Binding Protein P400 (EP400), were able to identify HCC individually with accuracies of 82.4%, 78.4% and 65%, respectively. In combination, these three genes gave an area under the curve (AUC) of 0.96 (95% confidence interval (CI), 0.93-0.99) using multivariate logistic regression and yielded a sensitivity of 93% and a specificity of 89%. When these three genes were used in combination with alpha-fetoprotein (AFP) to predict HCC, the accuracy of predicting HCC improved slightly with an AUC of 0.99 (95% CI, 0.98-1.0), sensitivity of 93% and specificity of 95%. CONCLUSIONS:CXCR2, CCR2 and EP400 can provide a promising non-invasive multiplex PCR diagnostic assay to monitor high-risk individuals for the development of HCC.

KEYWORDS: {{ getKeywords(articleDetailText.words) }}