Table 1.
The overview of phenotypic features of circulating tumor cells in hepatocellular carcinoma.
| Phenotypic Markers | Enrichment Method | Specimen | Key Findings | Ref |
|---|---|---|---|---|
| Is CTCs heterogeneity compatible with EMT (epithelial to mesenchymal transition) markers? | ||||
| EpCAM | qRT-PCR-based platform | 299 HCC patients and 120 control subjects | Compared with pre-operation, the population of EpCAM+ CTCs decreased significantly after operation, and all the patients with CTC reduction showed tumor remission. | [16] |
| EpCAM | CellSearch system | 59 HCC patients and 19 control patients | CTCs in the presence of EpCAM were strongly correlated with tumor aggressiveness, and this allowed adequate stratification of HCC patients for curative or systemic therapy. | [17] |
| Twist, GPC-3 | CanPatrol system | 80 HCC patients and 10 healthy volunteers | The ratio of twist+ CTCs was closely correlated with the rate of metastasis or recurrence and the mortality rate; the prognostic evaluation of twist+ CTCs was better than CTCs alone. | [18] |
| EpCAM, CK8/18/19, and vimentin, twist | CanPatrol system | 165 HCC patients | The presence of mesenchymal CTCs tended to occur in patients with advanced stage, and was associated with decreased relapse-free survival. | [19] |
| EpCAM, CK8/18/19, and vimentin, twist | CanPatrol system | 113 HCC patients | The use of total CTCs was more effective than AFP for the diagnosis of HCC, and the combination of total CTCs and AFP could enhance diagnostic effectiveness. | [20] |
| EpCAM, CK8/18/19, E-cadherin, vimentin, twist, AKT2, and snail | CanPatrol system | 195 HCC patients | Mesenchymal and hybrid CTCs had higher invasive and metastatic abilities than E type CTCs. | [21] |
| E-cadherin, vimentin, and twist | Flow cytometric analysis, and immunofluorescence staining | 46 HCC patients | Co-expression of twist and vimentin in CTCs was significantly correlated with portal vein tumor thrombus, TNM classification, and tumor size. | [22] |
| EpCAM, CK8/18/19, and vimentin, twist | CanPatrol system | 62 HCC patients | HCC patients with positive peripheral mesenchymal CTCs had a higher risk of early recurrence. | [23] |
| EpCAM, CK8/18/19, and vimentin, twist | CanPatrol system | 33 HCC patients and 10 healthy volunteers | Epithelial-mesenchymal-mixed CTCs played an important role in EMT transition of HCC. The mixed CTCs might be a vital factor for intrahepatic metastasis, and mesenchymal CTCs could have potential to be a predictor of extrahepatic metastasis. | [24] |
| EpCAM, CK8/18/19, and vimentin, twist | CanPatrol system | 40 HCC patients | The average ratio of mesenchymal CTCs in each sample was increased in the later stages of cancer compared with the earlier stages of cancer. | [25] |
| EpCAM, E-cadherin, CK8/18/19, vimentin, and twist, BCAT1 | CanPatrol system | 112 HCC patients | The percentage of BCAT1 was positively correlated with EMT process, suggesting a potential marker for CTCs in evaluating tumor metastasis or recurrence. | [26] |
| Hepatocyte-specific markers of CTCs in HCC | ||||
| GPC3, AFP | Enzyme-linked immunoassay | 68 HCC patients | The combination of GPC3 and AFP improved the overall sensitivity for HCC; the positive rate of GPC3 was significantly higher than that of AFP in HCC patients. | [27] |
| GPC3 | Density gradient centrifugation and immunomagnetic positive enrichment | 85 HCC patients | Pre-operative GPC3-positive CTCs was a risk factor of microscopic portal vein invasion and poor prognosis, and therefore it might be a useful biomarker for HCC patient outcomes. | [28] |
| ASGPR | Microfluidic chip | 36 HCC patients | CTCs were detected in all the examined patients with HCC. | [29] |
| ASGPR, CPS1, P-CK | Density gradient Ficoll-Paque PLUS, and magnetic labeling and separation | 27 HCC patients | All the 16 HCC tissues had ASGPR staining on the membranes of the HCC cells, and CTCs in the presence of CPS1 and P-CK were detected in the majority of patients with HCC. | [30] |
| ASGPR, GPC3 | Magnetically assisted surface-enhanced Raman scattering biosensor | Eight HCC patients, three breast cancer patients, and three healthy controls | Dual labelling of ASGPR and GPC3 was effective in detecting HCC CTCs with a small volume of blood samples in clinical settings. | [31] |
| ASGPR, GPC3, CK | Semiquantitative immunocytochemistry | 62 HCC patients, seven HBV-infected patients, and 15 healthy individuals | The cells obtained from the blood of HCC patients had significantly higher levels of ASGPR, GPC3, and CK than cells derived from chronic HBV-infected patients or healthy controls; ASGPR, GPC3, and CK might be valuable as HCC biomarkers for CTC detection; the expression of ASGPR and GPC3 might be helpful for understanding OS of the patients. | [32] |
| Hep Par 1, GPC3, GS | Label-free Labyrinth technology, and immunoaffinity-based CTC-Chip (Microfluidic chip) | 42 HCC patients, four non-HCC patients | The HCC CTC detection rate was improved by using three HCC markers compared to EpCAM-based identification method. | [33] |
| ASGPR, Hep Par 1 | Magnetic separation and immunoidentification | 85 HCC patients, 37 patients with benign liver diseases, 20 healthy volunteers, and 14 patients with other advanced cancers | No healthy, benign liver disease, or non-HCC cancer subjects were detected with CTCs. CTCs were identified in 69 of 85 HCC patients. | [34] |
| ASGPR, CPS1 | Density gradient Ficoll-Paque PLUS, magnetic labeling, and separation | 32 HCC patients, 17 patients with other types of cancer, 40 patients with other liver diseases, and 20 healthy volunteers | CTCs that tested positive for ASGPR and CPS1 were detected in 91% of patients with HCC, and there were no CTCs detected in healthy volunteers and in patients with any other kinds of cancers, including breast, lung, esophageal, gastric, and colorectal cancer. | [35] |
| CK, EpCAM, EMA, CK18, AFP, GPC-3, and Hep Par 1 | BenchMark XT Slide Preparation system | 23 HCC patients, six patients with non-HCC | 57.1% of patients tested positive for EpCAM, 42.9% for EMA, and 21.4% for AFP. | [36] |
| How do CTCs respond in tumor microenvironment? | ||||
| phosphorylated ERK (pERK) and pAkt CTC | Density gradient centrifugation, magnetic separation | 109 HCC patients | Phosphorylated ERK (pERK) and pAkt expressions in CTCs were correlated to sorafenib efficacy in HCC patients; pERK+/pAkt− CTCs were mostly responsive to sorafenib; the population of pERK+/pAkt− CTCs could be a potential predictive factor for HCC patients treated with sorafenib. | [37] |
| CD4+CD25+Foxp3+ Treg cells | PCR and fluorescence-activated cell sorting | 49 HCC patients | The early recurrence rate in the group with combined higher EpCAM+ CTCs and Treg/CD4+ population was significantly higher than in the combined lower CTCs and Treg group; the combined detection of EpCAM+ CTCs and Treg/CD4+ might provide a novel prognostic predictor for HCC patients. | [38] |
| IGFBP1 | Density gradient centrifugation, and immunomagnetic beads | 25 HCC patients | IGFBP1 was correlated with the responsiveness to selective internal radiation therapy. | [39] |
| Are CTCs equivalent to CSCs (Cancer Stem Cells)? | ||||
| EpCAM, CD133 | CellSearch system | 123 HCC patients | CSC biomarkers CD133 and ABCG2 were observed in the blood samples of HCC patients with positive EpCAM+ CTCs. | [40] |
| GPC3, GS, Hep Par 1, and CD44 | Label-free Labyrinth technology, and immunoaffinity-based CTC-Chip | 37 HCC patients | CTCs with the expression of CD44 were observed in all the stages of HCC; CTCs with these three markers, GPC3, GS, and Hep Par 1 had a cancer stemness phenotype. | [33] |
| EpCAM, CD133, CD90, CK19, ABCG2, CD44, ICAM1, CD24, and Nestin | qRT-PCR | 956 HCC patients and 50 healthy donors | Compared with EpCAM, the prognostic significance of CTC panel (EpCAM, CD90, CD133, and CK19) was still retained in the EpCAM− subgroup. | [41] |
| CD133, ANXA3 | Enzyme-linked immunosorbent assay | 368 HCC patients | Serum ANXA3 could stimulate and maintain the stem cell-like traits of CD133 CTCs to promote tumor recurrence and metastasis; combining ANXA3 with AFP significantly improved the outcome prediction. | [42] |