Potential conflict of interest: Nothing to report.
Supported in part by grants from the Ministry of Education, Culture, Sports, Science and Technology of Japan.
Radiofrequency ablation (RFA) is one of the treatments for hepatocellular carcinoma (HCC) and is known to enhance host immune response. However, the epitopes to which enhanced immune responses occur, the impact on patient prognosis, and the functions and phenotype of T cells induced are still unclear. To address these issues, we analyzed immune responses before and after RFA in 69 HCC patients using 11 tumor-associated antigen (TAA)-derived peptides that we identified to be appropriate to analyze HCC-specific immune responses. The immune responses were analyzed using enzyme-linked immunospot (ELISPOT) assay and tetramer assays using peripheral blood mononuclear cells. An increase in the number of TAA-specific T cells detected by interferon-γ ELISPOT assays occurred in 62.3% of patients after RFA. The antigens and their epitope to which enhanced T cell responses occur were diverse, and some of them were newly induced. The number of TAA-specific T cells after RFA was associated with the prevention of HCC recurrence, and it was clarified to be predictive of HCC recurrence after RFA by univariate and multivariate analyses. The number of TAA-specific T cells after RFA was inversely correlated with the frequency of CD14+HLA-DR−/low myeloid-derived suppressor cells (MDSCs). The modification of T cell phenotype was observed after RFA. The number of TAA-specific T cells at 24 weeks after RFA was decreased. Conclusion: Although RFA can enhance various TAA-specific T cell responses and the T cells induced contribute to the HCC recurrence-free survival of patients, besides immunosuppression by MDSCs, the memory phenotype and lifetime of TAA-specific T cells are not sufficient to prevent HCC recurrence completely. Additional treatments by vaccine or immunomodulatory drugs might be useful to improve the immunological effect of RFA. (Hepatology 2013;1448–1457)
Hepatocellular carcinoma (HCC) is the sixth most frequent type of cancer worldwide, and it is becoming an important public health concern due to its increased incidence in Western and Asian countries.1, 2 Although there are many types of treatments for HCC, the posttreatment recurrence rate is very high.3 To inhibit HCC recurrence and improve prognosis, an immunotherapeutic approach is considered an attractive strategy.
Radiofrequency ablation (RFA) is one of the treatments for HCC and is now widely used for curative strategies.4 In recent studies, it has been reported that RFA creates a tumor antigen source for the generation of antitumor immunity and enhances host immune responses.5 Our previous mouse study also showed that RFA induced antitumor immune responses with massive T cell infiltration into a tumor, and the effect was enhanced by an active variant of CC chemokine ligand 3.6 These studies suggest that additional immunological approaches to RFA may reduce HCC recurrence after treatment. However, in human studies, important data needed to develop a new immunotherapeutic approach have been lacking. First, the types of tumor-associated antigens (TAAs) and the epitopes to which these enhanced immune responses occur have not been fully identified. Second, the proportion of patients with enhanced antitumor immune responses and the effect of antitumor immunity for a patient's prognosis after RFA are still unclear. Third, the factors that affect TAA-specific immune responses and the functions and phenotype of T cells induced by RFA have not been identified.
In the present study, we analyzed immune responses in peripheral blood mononuclear cells (PBMCs) before and after RFA in 69 HCC patients using 11 TAA-derived peptides that we identified previously to be appropriate for analyzing HCC-specific immune responses. This approach offers useful information to develop a new strategy for HCC immunotherapy and improve the prognosis of patients treated by RFA.
In this study, we examined 69 human leukocyte antigen (HLA)-A24–positive HCC patients with RFA. The diagnosis of HCC was histologically confirmed in 11 patients. For the remaining 58 patients, the diagnosis was based on typical hypervascular tumor staining on angiography in addition to typical findings, which showed hyperattenuated areas in the early phase and hypoattenuation in the late phase on dynamic CT.7
RFA was performed with a cool-tip RFA system consisting of an 18-gauge, cooled-tip electrode with a 2- or 3-cm exposed tip (Radionics, Burlington, MA) and radiofrequency generator (CC-1 Cosman Coagulator, Radionics). After local anesthesia, the electrode was inserted through a guide needle under ultrasound guidance. Radiofrequency energy was delivered for 6 to 12 minutes for each session. The energy was increased from 40 watts to 120 watts in a stepwise fashion. During ablation, the electrode was cooled by circulating ice-cooled saline in the electrode lumen to maintain the tip temperature below 20°C. During each treatment, the electrode tip was inserted into the tumor 1-3 times until the target tumor was surrounded by a high-echoic area. Complete necrosis after RFA was confirmed by dynamic computed tomography (CT) or magnetic resonance imaging (MRI). RFA was repeated in some cases until complete necrosis was confirmed. Thirty-nine and 30 patients received RFA 1 and 2-4 times, respectively. After treatments, HCC recurrence was evaluated with dynamic CT or MRI every 3-4 months.
All patients gave written informed consent to participate in the study in accordance with the Helsinki declaration, and this study was approved by the regional ethics committee (Medical Ethics Committee of Kanazawa University).
Blood samples were tested for hepatitis B surface antigen and hepatitis C virus (HCV) antibody using commercial immunoassays (Fuji Rebio, Tokyo, Japan). The patients with HCV antibody were tested for serum HCV RNA by real-time PCR (Roche, Tokyo, Japan), and 49 of 52 patients with HCV antibody were HCV RNA–positive. HLA-based typing of PBMCs from patients and normal blood donors was performed using reverse sequence-specific oligonucleotide analysis with polymerase chain reaction (PCR-RSSO). The serum alpha-fetoprotein (AFP) level was measured via enzyme immunoassay (AxSYM AFP, Abbott Japan, Tokyo, Japan), and the pathological grading of tumor cell differentiation was assessed according to the general rules for the clinical and pathological study of primary liver cancer.8 The severity of liver disease was evaluated according to the criteria of Desmet et al. using biopsy specimens of liver tissue, where F4 was defined as cirrhosis.9 Fifty-five patients who participated in the present study received liver biopsy with RFA. Another 14 patients received liver biopsy 1-3 years before RFA.
Peptides and Preparation of PBMCs.
Eleven peptides that we previously identified as being useful for analysis of immune response in HLA-A24–positive HCC patients were selected.10-13 Human immunodeficiency virus (HIV) envelope-derived peptide (HIVenv584)14 and cytomegalovirus (CMV) pp65-derived peptide (CMVpp65328)15 were also selected as control peptides. Peptides were synthesized at Sumitomo Pharmaceuticals (Osaka, Japan). They were identified using mass spectrometry, and their purities were determined to be >90% by analytical high-performance liquid chromatography. PBMCs were isolated before and 2-4 weeks after HCC treatments as described.11 In the patients who received RFA 2-4 times, PBMCs were obtained 2-4 weeks after the final treatment. In some patients, PBMCs were also obtained 24 weeks after RFA. PBMCs were resuspended in Roswell Park Memorial Institute 1640 medium containing 80% fetal calf serum and 10% dimethyl sulfoxide and cryopreserved until use.
Interferon-γ ELISPOT Assay.
Interferon-γ (IFN-γ) ELISPOT assays were performed as described.11 Negative controls consisted of an HIV envelope–derived peptide (HIVenv584).14 Positive controls consisted of 10 ng/mL phorbol 12-myristate 13-acetate (PMA, Sigma) or a CMVpp65-derived peptide (CMVpp65328).15 The colored spots were counted with a KS ELISpot Reader (Zeiss, Tokyo, Japan). The number of specific spots was determined by subtracting the number of spots in the absence of an antigen from the number in its presence. Responses to peptides were considered positive if more than 10 specific spots were detected, which is greater than the mean + 3 SD of the baseline response detected in 11 HLA-A24–positive normal blood donors against TAA-derived peptides, and the number of spots in the presence of an antigen was at least twofold that in its absence. The results of an ELISPOT assay with more than 25 spots in the wells without peptides (control wells) were excluded from the analysis.
IFN-γ ELISPOT assays were also performed using PBMC-depleted CD4+ or CD8+ cells to determine what kind of T cell is responsive to the peptides. In the assay using PBMC-depleted CD4+ or CD8+ cells, the number of cells was adjusted to 3 − 105 cells/well after the depletion. Depletion of CD4+ or CD8+ cells was performed using the MACS separation system with CD4 or CD8 MicroBeads (Miltenyi Biotec, Auburn, CA) in accordance with the manufacturer's instructions.
Detection of Myeloid-Derived Suppressor Cells.
For the detection of myeloid-derived suppressor cells (MDSCs), PBMCs were isolated from 20 randomly selected patients 2-4 weeks after HCC treatment. To determine the frequency of CD14+HLA-DR−/low MDSCs, two-color fluorescence-activated cell sorting analysis was performed using the following antibodies: anti-CD14 and anti–HLA-DR (Becton Dickinson). Flow cytometry was performed using the FACSAria II system (Becton Dickinson). The frequency of CD14+HLA-DR−/low MDSCs was calculated as a percentage of HLA-DR−/low cells in CD14+ cells.
Tetramer Staining and Flow Cytometry.
Peptide MRP3765-, AFP357-, AFP403-, and hTERT461-specific tetramers were purchased from Medical Biological Laboratories Co., Ltd. (Nagoya, Japan). PBMCs were stained with anti–CD8-APCAb (Becton Dickinson, Tokyo, Japan), anti–CCR7-FITCAb (eBioscience, Tokyo, Japan), anti–CD45RA-PerCP-Cy5.5Ab (eBioscience, Tokyo, Japan), and tetramer-PE for 30 minutes at room temperature. Cells were washed, fixed with 0.5% paraformaldehyde/phosphate-buffered saline, and analyzed using the FACSAria II system.
Data are expressed as the mean ± SD. The estimated probability of tumor recurrence-free survival was determined using the Kaplan-Meier method. The Mantel-Cox log-rank test was used to compare curves between groups. The prognostic factors for tumor recurrence-free survival were analyzed for statistical significance using the Kaplan-Meier method (univariate) and the Cox proportional hazard model (multivariate). Linear regression lines for the relationship between the frequency of CD14+HLA-DR−/low MDSCs and the number of TAA-specific T cells were calculated using Pearson's correlation coefficient. A level of P < 0.05 was considered significant.
The clinical profiles of the 69 patients analyzed in the present study are shown in Table 1. HCC was histologically classified as well, moderately, and poorly differentiated in 7, 3, and 1 cases, respectively. In the other cases, HCC was diagnosed on the basis of typical CT findings and elevated AFP levels. In terms of size and number, the tumor was classified as large (>2 cm) in 28 cases, small (≤2 cm) in 41 cases, multiple in 29 cases, and solitary in 40 cases. Vascular invasion was noted in one patient. Using tumor-node-metastasis staging of the Union Internationale Contre Le Cancer (UICC) system (6th edition),16 patients were classified as having stage I (n = 39), II (n = 29), IIIA (n = 0), IIIB (n = 0), IIIC (n = 1), or IV (n = 0) tumors.
Table 1. Patient Characteristics (n = 69)
Data are expressed as the mean ± SD (median) or as the number of patients.
Abbreviations: ALT, alanine aminotransferase; HBV, hepatitis B virus; ND, not determined, TNM, tumor-node-metastasis.
67.3 ± 9.4 (69.0)
Platelet count, ×104/μL
15.9 ± 26.5 (10.9)
Platelet count, >15 × 104/≤15 × 104/μL
46.7 ± 33.8 (38.0)
ALT, >30/≤30 IU/L
Prothrombin time, %
78.4 ± 14.6 (77.0)
Prothrombin time, >70%/≤70%
3.6 ± 0.5 (3.6)
Albumin, >3.5/≤3.5 g/dL
Total bilirubin, mg/dL
1.1 ± 0.6 (0.9)
Total bilirubin, >2.0/≤2.0 mg/dL
134.7 ± 468.3 (11.0)
AFP, >100/≤100 ng/mL
HCC differentiation, well/moderate/poor/ND
Tumor diameter, >2/≤2 cm
Tumor multiplicity, multiple/solitary
Vascular invasion, +/−
TNM stage, I/II/IIIA/IIIB/IIIC/IV
Histology of nontumor liver, liver cirrhosis/chronic hepatitis
Detection of TAA-Specific T Cells Before and After RFA.
Detection of TAA-specific T cells was performed by direct ex vivo analysis (IFN-γ ELISPOT assay). Positive T cell responses against each TAA-derived peptide were observed in 0 to 11 (0.0%-17.2%) patients before RFA (Table 2). The same responses against HIV- and CMV-derived peptides were observed in 1 (1.6%) and 43 (62.3%) patients, respectively. After HCC treatments with RFA, positive T cell responses against TAA-, HIV- and CMV-derived peptide were observed in 8-24 (11.6%-35.3%), 2 (2.9%), and 39 (58.2%) patients, respectively. The increase of the frequency of TAA-specific T cells after RFA observed in 7 of 11 peptides (SART2899, SART3109, MRP3503, MRP3765, AFP357, AFP403, and hTERT461) was statistically significant (Table 2).
Table 2. Peptides and Response Frequency
Amino Acid Sequence
Number of Specific Spots in Normal Donors (mean ± SD)
The magnitude of TAA-specific T cell responses determined by the frequency of T cells and the proportion of the patients who showed a significant increase of TAA-specific T cells are shown in Fig. 1. When the T cell responses against a single peptide with more than or equal to 10 specific spots and two-fold increase were defined as significant, a significant increase was observed in 4-16 (6.5%-24.6%) patients for each TAA-derived peptide and in 24 (39.3%) patients for total of TAA-derived peptides. On the other hand, the numbers of patients who showed a significant increase against HIV- and CMV-derived peptide were 1 (1.6%) and 8 (11.9%), respectively. The number of patients who showed a significant increase against at least one TAA-derived peptide after RFA was 43 (62.3%).
To determine what kind of T cell is responsive to the peptides, TAA-derived peptide-specific IFN-γ–producing T cells were also analyzed by ELISPOT assay using PBMC-depleted CD4+ or CD8+ cells. The assay showed that IFN-γ–producing T cells against the peptides (SART2899, SART3109, MRP3503, MRP3692, MRP3765, AFP357, AFP403, AFP434, hTERT167, hTERT324, and hTERT461) mainly consisted of CD8+ cells (Supporting Fig. 1).
Effect of Increase of TAA-Specific T Cells After RFA for the Prognosis of Patients.
To examine the effect of increase of TAA-specific T cells after RFA for the prognosis of patients, we analyzed the relationship between the number of TAA-specific T cells and HCC recurrence-free survival after RFA. First, we divided the patients into two groups with high (above median) and low (below median) specific spots detected via ELISPOT assay. In the analysis, we found that a high number of TAA-specific T cells after HCC treatment correlated significantly with the length of HCC recurrence-free survival (P = 0.044) (Fig. 2A). The difference between the groups was emphasized when 50 spots were defined as highly specific spots (P = 0.006) (Fig. 2B). On the other hand, there was no correlation between the number of TAA-specific T cells before HCC treatment and the length of HCC recurrence-free survival (P = 0.758) (Fig. 2C). Furthermore, the magnitude of enhancement of TAA-specific immune responses did not correlate significantly with the length of HCC recurrence-free survival (P = 0.267) (Fig. 2D).
When univariate analysis of prognostic factors for HCC recurrence-free survival was performed, γ-glutamyltransferase (<30), AFP (<400), Okuda stage,1 and number of TAA-specific T cells after RFA (≥50) were detected as factors that decrease HCC recurrence rate after RFA (Table 3). When multivariate analysis including these three factors was performed, only the number of TAA-specific T cells after RFA (≥50) was found to be a factor that decreases HCC recurrence rate after RFA.
Table 3. Univariate and Multivariate Analysis of Prognostic Factors for Tumor-Free Survival
HR (95% CI)
Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BCLC, Barcelona Clinic Liver Cancer; CLIP, Cancer of the Liver Italian Program; DCP, des- γ-carboxy prothrombin γ-GTP, γ-guanosine triphosphate.
Number of TAA-specific T cells before and after RFA was calculated per 3 × 105 PBMCs.
Number of TAA-specific T cells after RFA, <50/≥50*
Analysis of the Factors that Affect the Number of TAA-Specific T Cells After RFA.
To identify the factors that affect the number of TAA-specific T cells after RFA, we analyzed clinical parameters of patients and the frequency of CD14+HLA-DR−/low MDSCs after HCC treatment. We could not find any clinical parameters correlated with the number of TAA-specific T cells after RFA.
The frequency of CD14+HLA-DR−/low MDSCs after RFA showed various levels and depended on the patient (Fig. 3A,B). The frequency decreased significantly after RFA (P = 0.022) except in three patients (Fig. 3B) and correlated inversely with the number of TAA-specific T cells after RFA, but not with that of CMV-specific T cells (Fig. 3C).
Phenotypic Analysis of TAA-Specific T Cells Before and After RFA.
Next, we examined the naïve/effector/memory phenotype of increased TAA-specific T cells after RFA using a tetramer assay. The memory phenotype was investigated by the criterion of CD45RA/CCR7 expression.17 In tetramer analysis, the frequency of TAA-derived peptide-specific CD8+ T cells before RFA was 0.00%-0.03% of CD8+ cells (Fig. 4A). On the other hand, the frequency was increased after RFA in 10/12 (83.3%) patients, and the range was 0.00%-0.10% of CD8+ cells. The frequency of CD45RA−/CCR7+ (central memory), CD45RA−/CCR7− (effector memory), and CD45RA+/CCR7− (effector) T cells in tetramer-positive cells depended on the patients, and the ratio of these cells changed after RFA (Fig. 4B). The frequency of tetramer-positive cells with CD45RA−/CCR7+ and CD45RA−/CCR7− in CD8+ cells was increased in 6/7 (85.7%) and 6/7 (85.7%) patients, respectively, whose samples were available for the assay before and after RFA. Interestingly, the tetramer-positive cells with CD45RA−/CCR7+ were newly induced after RFA in 5/7 (71.4%) patients.
Kinetics of TAA-Specific T Cells Induced by RFA.
Although the number of TAA-specific T cells was a predictive factor of a decrease of HCC recurrence rate after RFA (as shown in Fig. 2A), more than 50% of the patients with a high number of TAA-specific T cells showed HCC recurrence for 25 months after treatment. To identify the relationship between TAA-specific T cell responses and HCC recurrence more precisely, we examined the kinetics of TAA-specific T cells in 16 patients whose PBMCs were available for analysis at 24 weeks after RFA. The frequencies of TAA-derived peptide-specific T cells decreased in most of the peptides and patients at 24 weeks after RFA (Fig. 5). In the analysis of the total of each type of TAA-derived peptide-specific T cells, the frequency decreased in 14/16 (87.5%) patients analyzed, and most of them showed fewer than 50 specific spots per 3 × 105 PBMCs, with the exception of one patient. In contrast, the frequencies of CMV-derived peptide-specific T cells were maintained in most of the patients.
In recent years, HCC-specific TAAs and their T cell epitopes have been identified, which has made analysis of immunological status in HCC patients possible and shown that TAA-specific T cell responses can be detected in peripheral blood.11, 18-20 The immunological analysis of HCC patients with RFA using 11 TAA-derived peptides in this study showed that the enhancement of TAA-specific T cell responses occurred in 62.3% of patients, the antigens and their epitope to which enhanced T cell responses occurred were diverse, and some of them were newly induced. The mechanism of enhancement of tumor-specific immune response by RFA is still unclear. den Brok et al.5 showed that RFA created an antigen source for antitumor immunity by destruction of tumor cells using a mouse tumor model. The antigens used in this study have been reported to be located in the cell membrane (MRP3), cytoplasm (SART2 and AFP), and nucleus (hTERT and SART3).21-24 The diversity of the target proteins of enhanced T cells suggests that the central mechanism of enhancement of tumor-specific immune response by RFA is due to tumor cell destruction, which supports the results mentioned previously.5
In the present study, we also showed that the number of TAA-specific T cells after RFA was associated with the HCC recurrence-free survival of patients. The univariate and multivariate analyses clearly showed it was a predictive factor for HCC recurrence after RFA. These results suggest that TAA-specific T cells induced by RFA contribute to protection from HCC recurrence, and additional immunological approaches should be applied to enhance the protective effect after treatment.
To understand the precise mechanism that RFA enhances TAA-specific T cell responses, we analyzed the factors that affected the number of TAA-specific T cells after RFA. Among the factors analyzed, the frequency of CD14+HLA-DR−/low MDSCs after RFA was inversely correlated with the number of TAA-specific T cells, suggesting these MDSCs may have a negative effect on TAA-specific immune responses. Regarding the function of MDSCs in cancer patients, it has been reported that they inhibit T lymphocyte responses.25 In HCC patients, it is reported that the frequency of CD14+HLA-DR−/low MDSCs in PBMCs is significantly increased in comparison with healthy controls and they exert immunosuppressive function via induction of regulatory T cells.26 Taken together with our results, these reports suggest that an additional immunological approach to inhibit the function of MDSCs after RFA may enhance TAA-specific immune responses.
On the other hand, patients with a high number of TAA-specific T cells were not completely protected from HCC recurrence. To examine the mechanisms behind the failure to control HCC recurrence completely by RFA-induced TAA-specific immune responses, we performed phenotypic and kinetic analysis of T cells enhanced by RFA. The results showed that the frequency of T cells with each memory phenotype depended on the patient, and the ratio of these cells changed after RFA. The memory phenotype of T cells that showed a more than two-fold increase was the CD45RA−/CCR7+ (central memory) phenotype, which required secondary stimulation by antigen to exert stronger antitumor effects.17 Interestingly, they were newly induced, suggesting that RFA may modify not only the frequency but also the phenotype of TAA-specific T cells.
The frequencies of TAA-derived peptide-specific T cells decreased in most of the patients at 24 weeks after RFA, suggesting that RFA could not induce long-lived T cells. In a previous study, it was reported that tumor-specific immune responses induced by RFA could not protect from HCC recurrence completely because of tumor immune escape.27 In addition to this mechanism, our results suggest that one of the reasons that RFA-induced tumor-specific immune response is insufficient for controlling HCC recurrence is the weak induction of long-lived T cells.
Taken together with these results, the present study suggests that the antitumor effect of TAA-specific T cells induced by RFA should be enhanced by an additional immunological approach. In recent studies of cancer immunology, cancer vaccines consisting of TAA-derived protein or peptide, recombinant virus, and engineered tumor cells have been considered as candidates to enhance host immune responses.28 Alternatively, immunomodulating antibodies such as anti-cytotoxic T-lymphocyte antigen 4 (CTLA-4) and anti-programmed cell death 1 (PD-1) have been considered to reactivate T cell function.28, 29 These approaches may also be effective to enhance the antitumor effect induced by RFA.
In conclusion, the results of this study show that RFA can enhance various TAA-specific T cell responses and the number of T cells induced is associated with HCC recurrence-free survival. To maintain the TAA-specific T cell responses induced by RFA and to improve the immunological effect for HCC, additional treatment by vaccine or immunomodulatory drugs might be useful.