Associations of components of PTEN/AKT/mTOR pathway with cancer stem cell markers and prognostic value of these biomarkers in hepatocellular carcinoma

Aim: We aimed to investigate the associations between components of the phosphatase and tensin homolog deleted on chromosome 10/protein kinase B/mammalian target of rapamycin (PTEN/AKT/mTOR) pathway and liver cancer stem cell (LCSC) markers, including CD133, CD90, CD44, and epithelial cell adhesion molecule (EpCAM), and to further evaluate the predictive values of these biomarkers for recurrence and survival in hepatocellular carcinoma (HCC). Method: Protein expressions and mRNA levels of PTEN and LCSC markers were determined in 110 HCC tissues and 98 adjacent non-tumor tissues. Protein expressions of phosphorylated AKT (p-AKT) and phosphorylated mTOR (p-mTOR) were detected to evaluate the activation of the PTEN/AKT/mTOR pathway by using immunohistochemistry. Prognostic significance was analyzed by univariate and multivariate analysis. Results: Loss of PTEN expression was negatively correlated with positive expression of CD133, CD90, and EpCAM (P< 0.05). Positive expression of p-AKT and p-mTOR were positively associated with positive expression for CD133, CD90, and EpCAM (P< 0.05). By univariate and multivariate analysis, a higher level of a-fetoprotein, loss of PTEN expression, and CD133-positive, p-AKT-positive, p-mTOR-positive, and EpCAM-positive signals were predictors for HCC recurrence, whereas advanced TNM stage, loss of PTEN expression, and positive expression of p-AKT, p-mTOR, and CD133 were predictors for survival. Patients with PTEN-/CD133(+) or PTEN-/EpCAM(+) HCC had shorter recurrence-free survival and overall survival times. Conclusion: The PTEN/AKT/mTOR pathway might play a crucial role in driving recurrence and influencing prognosis in HCC. There could be a potential repressive relationship between components of the PTEN/AKT/mTOR pathway and LCSCs. The combination of PTEN with CD133 or EpCAM expression may serve as a screening tool to monitor recurrence and predict prognosis.