The absolute incidence and mortality of hepatocellular carcinoma (HCC) worldwide reported in 2018 was 841,080 and 781,631, respectively [1
]. It is the second most common cancer and the most common cause of cancer-related deaths in men in Taiwan. Hepatitis B virus (HBV)/hepatitis C virus (HCV) infection and alcohol abuse are among the major etiologies for liver diseases and cancer. The underlying etiology of HCC due to HBV and HCV infections in Taiwan has been reduced because of universal HBV vaccination and the recent development of highly effective anti-HCV agents [2
]. However, there is a parallel increase in HCC related to nonalcoholic fatty liver disease [4
Once HCC is diagnosed, the Barcelona Clinic Liver Cancer algorithm, which includes in classifying HCC patients, clinical variables associated with tumor burden, the degree of liver dysfunction, and the patient’s symptoms, is used for disease staging and treatment allocation [5
]. A number of strategies have been employed for the treatment of patients with HCC. Transplantation remains the best option, but the supply of donor organs is limited. Alternative curative treatments including resection, radiofrequency ablation (RFA) [7
], and, potentially, systemic and targeted therapy may delay recurrence [8
]. When patients experience recurrence following the initial local therapies or transplantation, another local therapy and/or systemic treatments are required to obtain the best possible outcome. Abdominal ultrasound, radiological imaging, and serum alpha-fetoprotein (AFP) levels are the most common methods for surveillance of the treatment response. Imaging by computed tomography (CT) scans exposes patients to radiation, while serum AFP levels are in the normal range in 15–30% of patients with HCC [10
]. For patients who are in clinical remission, all of the above methods cannot detect early recurrence.
During cancer progression, the number of circulating tumor cells (CTCs) increases and correlates with tumor mass [11
]. Effective therapies that reduce tumor burden usually correlates with a decrease in the number of CTCs. Monitoring the CTC count is valuable in scrutinizing the treatment response of cancer patients and should provide for better patient care and management [12
]. The presence of CTCs in patients with HCC has been reported in several studies [15
]. CTC counts have been analyzed in most of these studies at a specific time point without longitudinal follow-up concerning the number or the change in CTC counts. Whether or not serial CTC counts during the course of treatment provides an advantage for monitoring the treatment response and disease status of patients with HCC remains to be explored.
In this study, patients with locally advanced or distant metastatic HCC were recruited for a series of CTC enumeration using a negative selection platform PowerMag which allows efficient depletion of CD45+
leukocytes and isolation of viable and label-free CTCs even only one CTC is present in 4 mL of peripheral blood [18
]. The CD45+
leukocytes are depleted for 10–20 folds more efficient by the PowerMag platform than by the EasySep method (StemCell Technologies, Vancouver, BC, Canada). The recovery rates of CTCs are 77–82% and 46–62% when cancer cells were enriched from leukocyte suspension and whole blood, respectively. The number of cancer cells recovered by PowerMag was linearly correlated with the number of cancer cells in the blood [18
]. By the PowerMag platform, the number of CTCs and the change in the CTC counts were analyzed and correlated with the disease status and treatment response of the patients. Performing a series of CTC enumeration in parallel with the current standard medical methods provide benefits for the clinical care of patients with locally advanced or metastatic HCC, in particular for patients without elevated serum AFP levels.
Monitoring disease status during treatment is crucial for the clinical management of patients with HCC. In the current study, the number of CTCs and the change in the CTC count are clinically significant and can supplement the use of serum AFP levels for evaluating the disease status of patients with HCC, in particular for the AFP-low cases.
Although serum AFP levels are routinely used in the clinical setting to monitor the disease status of patients with HCC, the serum AFP levels are found in the normal range for up to 42% of patients with HCC [25
]. The use of serum AFP levels for HCC surveillance is not recommended by the American Association for the Study of Liver Diseases due to the lack of specificity [26
]. Image examinations such as CT or MRI are still the standards for monitoring the treatment response. These procedures either need exposure to radiation or are expensive, and are not performed routinely. Novel HCC biomarkers such as des-gamma-carboxyprothrombin and lectin-bound AFP have been considered for overcoming the drawback of specifically using serum AFP levels for monitoring the treatment response of patients [27
]. These markers exhibit poor sensitivity for HCC, particularly in the detection of lesions <3 cm [28
]. There is a need for additional markers for monitoring disease status and evaluating the treatment response in patients with HCC, in particular, those patients without elevated serum AFP levels.
With the heterogeneity and complexity of HCC tissue and cancer cells, real-time monitoring of cancer burden in response to therapy is important during the treatment course of patients with HCC. Although imaging study provides a direct assessment of cancer burden on the tissue, the intrinsic property of radiation makes it not suitable for routine longitudinal follow-up of patients. Serum biomarkers such as AFP offer an alternative avenue for real-time monitoring of cancer burden but with clear disadvantages in the clinical setting. Recent studies indicate that liquid biopsies are the suitable biological resources for longitudinal measurements of cancer burden and disease status. Liquid biopsies such as CTC and cell-free circulating tumor DNA offer the option of taking serial samples from a patient to detect changes during disease history and imposed by treatment [29
]. Although serial monitoring of cell-free DNA was common, almost all reported CTC studies did only one to two measurements of CTC. For example, CTC counts have been used in previous studies for determining the diagnosis and assessing the prognosis of patients with HCC [15
]. A correlation exists between the postoperative early recurrence of HCC and mesenchymal CTCs [32
]. An elevated number of CTCs is associated with unfavorable clinicopathologic characteristics responsible for poor prognosis in patients with HCC [34
]. An increase in the number of postoperative CTCs when compared to the number of preoperative CTCs is associated with lower survival and higher recurrence among patients with low AFP levels and cirrhosis [35
]. Unlike prior studies, in which blood samples were collected at either a single or two time points prior to and after surgery or chemotherapy, blood samples in the current study were collected longitudinally before and during the course of systemic therapies. An evaluation of whether or not the number of CTCs or the change of CTC count related to the change of disease status in HCC patients was based on the CTC counts of each blood collection. With the continuous evolution of intrinsic cellular properties of HCC and the possible changes of therapeutic regimens during the treatment course, longitudinal follow-up of CTCs offers an option for real-time measurement of disease status for patients with HCC.
Several observations were made in the current study. First, CTCs were present in most of the blood samples collected from patients prior to systemic therapies. This is consistent with a number of HCC, papillary thyroid carcinoma, and head and neck cancer studies [19
]. Second, patients at PD usually had higher CTC counts compared to patients at SD or PR. Although there is overlap for the two groups of patients, this may have minimal effect on the longitudinal measurements of CTC count and its correlation with patient disease status which is based on comparison to each patient’s baseline CTC count. Further analysis of 85 blood samples from 17 patients with at least three sequential blood collections revealed that a change of the CTC count correlated with a change of disease status in most of the patients. The number of CTC or a change of the CTC count is not in accord with the clinical courses in several patients with high serum AFP, while in other cases AFP provides more useful clinical information and serves as a predictive biomarker even the CTC count reliably follows the clinical course of the patients. According to our analysis of patients who were AFP-low (Figure 6
) but underwent disease progression, longitudinal measurements of CTC are informative and provide advantages for monitoring treatment response and disease status of patients without elevated serum AFP.
Simultaneous counting of CTCs and measurement of serum AFP levels are likely to provide more comprehensive information regarding the disease status and treatment response of patients with HCC. This is similar to our previous studies of papillary thyroid carcinoma [36
]. The presence of CTC compromises serum thyroglobulin (Tg) measurements when the anti-Tg antibody is present in patients with papillary thyroid carcinoma [39
]. Combined analyses of CTC counts and serum Tg levels are superior to serum Tg levels alone in establishing the disease status of patients [36
]. The combination of CTC counts with serum AFP levels and Tg should decrease false-negative rates and is recommended for disease monitoring of patients with the respective cancer types who are undergoing treatment. Because peripheral blood is easily accessible and provides an important biological resource for diagnosis and monitoring disease progression of cancer patients, combined analyses of CTCs and serum AFP levels with one blood draw can be performed more frequently than imaging and could pave the way for monitoring the disease status and treatment response of HCC patients.
The disease status of several patients was not correlated with the number of CTCs or the change of the CTC counts as discussed in the previous section. The reasons for these observations are not clear. The epithelial–mesenchymal transition (EMT) occurs during cancer progression. A loss or a decrease in EpCAM expression and an increase in mesenchymal markers such as N-cadherin, vimentin, and podoplanin in cancer cells may accompany the EMT [33
]. The number of EpCAM+
-CTCs is likely to decrease when the disease status of patients changes from SD to PD. This may result in a decrease in EpCAM+
-CTCs and an increase in EpCAM−
-CTCs counts when a patient’s cancer is progressing. In the cell filtrate after depletion of CD45+
leukocytes, there are EpCAM−
cells of which the number is variable in different blood samples from 103
cells. Usually, less than 5 cells in the cell filtrate were CD45+
]. Because EpCAM−
cells are also present in the healthy control group, it is likely that most of the EpCAM−
cells are CD45−
circulating endothelial cells or leukocytes such as apoptotic neutrophils [42
], while some of the cells are EpCAM−
-CTCs. Future characterization of these EpCAM−
cells should provide more insight into the clinical implication of monitoring this cell population. Consistent with this notion, different CTC subpopulations exist and elicit a distinct impact on the disease progression of patients. In this regard, the number of CD133+
circulating cancer stem cells is present in the blood and is a biomarker of treatment selection and liver metastasis in patients with CRC [44
]. An RNA signature enables high specificity detection of CTCs in HCC [45
]. Therefore, it is worthwhile to further characterize different CTC subsets and determine their respective impact on the disease status of HCC patients.
Sequential CTC counts during therapy can supplement serum AFP measurements and is able to provide timely information for monitoring treatment efficacy and clinical outcomes. The findings of the current study may have a significant impact on the clinical management of patients with locally advanced or metastatic HCC, in particular for patients without elevated serum AFP levels.