A phase II study of adoptive immunotherapy using dendritic cells pulsed with tumor lysate in patients with hepatocellular carcinoma


  • Daniel H. Palmer,

    Corresponding author
    1. Cancer Research UK Clinical Trials Unit, CR UK Institute for Cancer Studies, Clinical Research Block
    • CR UK Institute for Cancer Studies, University of Birmingham, Vincent Drive, Edgbaston, Birmingham, B15 2TT, United Kingdom
    Search for more papers by this author
    • These authors contributed equally to this work.

    • fax: (44)-121-4143295.

  • Rachel S. Midgley,

    1. Cancer Research UK Clinical Trials Unit, CR UK Institute for Cancer Studies, Clinical Research Block
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Noweeda Mirza,

    1. Liver Research Laboratories, MRC Centre for Immune Regulation, 5th Floor Institute for Biomedical Research, University of Birmingham, Birmingham, UK
    Search for more papers by this author
  • Elizabeth E. Torr,

    1. Liver Research Laboratories, MRC Centre for Immune Regulation, 5th Floor Institute for Biomedical Research, University of Birmingham, Birmingham, UK
    Search for more papers by this author
  • Forhad Ahmed,

    1. Liver Research Laboratories, MRC Centre for Immune Regulation, 5th Floor Institute for Biomedical Research, University of Birmingham, Birmingham, UK
    Search for more papers by this author
  • Jane C. Steele,

    1. Cancer Research UK Clinical Trials Unit, CR UK Institute for Cancer Studies, Clinical Research Block
    Search for more papers by this author
  • Neil M. Steven,

    1. Cancer Research UK Clinical Trials Unit, CR UK Institute for Cancer Studies, Clinical Research Block
    Search for more papers by this author
  • David J. Kerr,

    1. Cancer Research UK Clinical Trials Unit, CR UK Institute for Cancer Studies, Clinical Research Block
    Search for more papers by this author
  • Lawrence S. Young,

    1. Cancer Research UK Clinical Trials Unit, CR UK Institute for Cancer Studies, Clinical Research Block
    Search for more papers by this author
  • David H. Adams

    1. Liver Research Laboratories, MRC Centre for Immune Regulation, 5th Floor Institute for Biomedical Research, University of Birmingham, Birmingham, UK
    2. NIHR Biomedical Research Unit for Liver Disease, University Hospital Birmingham NHS Foundation Trust, Birmingham, UK
    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.


This is a phase II clinical trial investigating the safety and efficacy of intravenous vaccination with mature autologous dendritic cells (DCs) pulsed ex vivo with a liver tumor cell line lysate (HepG2) in patients with advanced hepatocellular carcinoma (HCC). HCC is an attractive target for immunotherapy as evidenced by an active recruitment of tumor-infiltrating lymphocytes that are capable of lysing autologous tumor cells in ex vivo studies. DCs are the most potent antigen-presenting cells, with the capacity to take up, process, and present tumor antigens to T cells and stimulate an immune response, thus providing a rational platform for vaccine development. Thirty-five patients with advanced HCC and not suitable for radical or loco-regional therapies received a maximum of six DC vaccinations each at 3-week intervals. In total, 134 DC infusions were administered with no significant toxicity and no evidence of autoimmunity. Twenty-five patients who received at least three vaccine infusions were assessed clinically for response. The radiologically determined disease control rate (combined partial response and stable disease ≥3 months) was 28%. In 17 patients the baseline serum α-fetoprotein (AFP) was ≥ 1,000 ng/mL; in four of these patients, it fell to <30% of baseline following vaccination. In one patient there was a radiological partial response associated with a fall in AFP to <10% of baseline. Immune responses were assessed using an ELIspot assay of interferon-γ (IFN-γ) release. In several cases there was induction of T cell responses to the vaccine and/or AFP following vaccination. Conclusion: Autologous DC vaccination in patients with HCC is safe and well tolerated with evidence of antitumor efficacy assessed radiologically and serologically, with generation of antigen-specific immune responses in some cases. (HEPATOLOGY 2009;49:124-132.)

With an estimated 500,000 new cases per year, hepatocellular carcinoma (HCC) represents the third leading cause of cancer death worldwide. Although most deaths occur in the Far East and sub-Saharan Africa, the incidence is rising in the West, largely due to an increasing incidence of hepatitis C virus infection.1 The majority of patients present with disease too advanced for curative treatment by resection, transplantation, or local ablation. A small percentage of patients are suitable for transarterial chemo-embolization, but although this may provide a modest prolongation in survival, disease progression is inevitable.2 Recently, a randomized phase III placebo-controlled trial has shown a survival advantage for the multitargeted kinase inhibitor sorafenib in patients with advanced HCC, although this, too, is modest.3 No other systemic therapy has shown any survival benefit. Thus, there remains an urgent need for novel therapies for this disease.

Immunotherapy has demonstrated some clinical activity in tumors that are associated with an inflammatory or immune response, such as melanoma and renal cell carcinoma. HCC also possesses characteristics that render it a potential target for immunotherapeutic manipulation. For example, there is an active recruitment of lymphocytes to HCC, which have specific mechanisms to recognize and bind to tumor endothelium and infiltrate tumor tissue, suggesting a potential for cytotoxic effector cell activation.4 In addition, tumor-infiltrating lymphocytes derived from HCC and then expanded ex vivo in the presence of interleukin (IL)-2 display antitumor cytolytic activity.5 Lymphocyte infiltration is also associated with persistent expression of major histocompatibility complex (MHC) class I antigens and costimulatory molecules on tumor cells. However, tumor-infiltrating lymphocytes in HCC are only partially activated, proliferate only at very low levels, and fail to kill tumor cells unless activated by IL-2 in vitro, suggesting that immunosuppressive mechanisms prevent T cell maturation into useful antitumor effectors. The activation of tumor-specific cytotoxic T cells requires three synergistic signals: the presentation of tumor antigen by antigen-presenting cells to specific T-helper cells; the interaction between costimulatory factors (such as B7.1 and CD28 ligand); and the secretion of immunostimulatory cytokines (such as IL-2 and IL-12) from activated T-helper cells. Dendritic cells (DCs) are the most potent professional antigen-presenting cells that can capture, process, and present antigens to naïve T cells, stimulating their proliferation and activation. They provide the optimum costimulatory environment, with high levels of MHC I and II, costimulatory molecules (CD40, B7), adhesion molecules (intercellular cell adhesion molecule 1, vascular adhesion protein 1) and stimulatory cytokines (IL-12, interferon-γ) to evoke an immunostimulatory signal against that antigen. DC-based immunotherapy has been tested in clinical trials in melanoma, prostate cancer, renal cell cancer, and HCC.6–11 Safety has been demonstrated with variable efficacy.12, 13

Although variable expression of a number of antigens, including α-fetoprotein (AFP), Mage 1 and −3, and NY-ESO1, has been demonstrated in HCC, no single antigen is omnipresent and detailed knowledge of specific T cell reactivity against many of the antigens is limited.12, 14, 15 To overcome this problem, the phase II trial described here used a lysate of the hepatoblastoma cell line, HepG2, as a source of multiple antigens, many of which, including AFP, are shared with HCC. This trial investigated the safety and efficacy of vaccination with mature autologous DCs pulsed ex vivo with this lysate in patients with advanced HCC.


AFP, α-fetoprotein; DC, dendritic cells; IFN-γ, interferon-γ; HCC, hepatocellular carcinoma; HLA, human leukocyte antigen; IL, interleukin; MHC, major histocompatibility complex; PBMC, peripheral blood mononuclear cells.

Patients and Methods

Study Design.

This was a single-center, open-label, phase II study of adoptive immunotherapy in patients with HCC using mature autologous DCs pulsed with lysate of the HepG2 tumor cell line. A two-step Gehan design was employed to detect a minimum response rate of ≥20%, with recruitment of 14 patients in the first phase, and plans for further recruitment dependent on the response rate in the first cohort.

The primary objectives were radiological response rate and toxicity. Secondary objectives were changes in serum AFP and analysis of immunological responses to vaccination.


Patients with HCC not amenable to curative resection, transplantation, local ablation, or chemoembolization were eligible for inclusion in this study. The diagnosis of HCC was confirmed either histologically or according to European Association for the Study of the Liver criteria of known predisposing chronic liver disease, AFP >400 ng/mL, and characteristic imaging. Patients were required to be at least 18 years of age, have a World Health Organization performance status ≤2, and have a life expectancy of greater than 12 weeks. Laboratory values necessary for inclusion were as follows: hemoglobin >10g/dL, white blood cell count >2 × 109/L, platelet count >75 × 109/L, serum creatinine <150 mmol, aspartate aminotransferase <5 times the upper limit of normal, and serum bilirubin <2.5 times the upper limit of normal. For the safety of laboratory workers culturing the DCs, patients were required to have negative hepatitis B virus, hepatitis C virus, and human immunodeficiency virus serology. Testing for virus serology was performed in all patients following appropriate counseling.

Exclusion criteria included the following: pregnant or breast-feeding women or those not practicing adequate contraception; psychiatric, addictive, or any other disorder that prevented informed consent; active uncontrolled infection; concurrent systemic corticosteroid treatment; systemic autoimmune disease; clinically significant ischemic heart disease or cardiac failure; and chemotherapy or radiotherapy within the preceding 4 weeks.

Written informed consent was obtained before enrollment into the study. The study protocol was approved by the local research ethics committee and conformed to the provisions of the Declaration of Helsinki.

Clinical Protocol.

Screening assessments were completed within 2 weeks before entry into the trial and included medical history, physical examination, full blood count, serum renal and liver function tests, pregnancy test for women of child-bearing potential, hepatitis B and C serology, and baseline symptomatic assessment of cancer-related morbidity.

All known areas of disease were assessed clinically, serologically (liver biochemistry and AFP), and radiologically within 4 weeks before the first vaccination.

Patients received up to a maximum of six three-weekly DC vaccinations. On day 1 of each cycle, 150 mL of peripheral blood was taken for DC preparation, and full blood count, biochemical profile, and AFP estimation were performed. Mature lysate-pulsed DCs were reinfused intravenously in 20 mL of phosphate-buffered saline over 10 minutes on day 12. Patients were observed for 2 hours after each vaccination to assess any immediate complications.

Interim disease assessments were performed on day 1 of each cycle with medical history, examination, liver biochemistry, and AFP measurement. After three vaccinations, repeat imaging was performed and compared with baseline. Measurable disease and response were defined according to International Union Against Cancer criteria.

Toxicity was graded according to World Health Organization common toxicity criteria. Safety was assessed via clinical assessment, serial full blood count and biochemistry (every cycle). Prevaccination 1 and postvaccination 3 and 6 assays to assess autoimmunity—in particular estimation of titres of antinuclear antibody, double-stranded DNA antibody, and rheumatoid factor—were performed.

Preparation of DCs from Peripheral Blood.

DCs were prepared according to the method of Romani16 in a Good Manufacturing Practice (GMP)-compliant facility. Peripheral blood mononuclear cells (PBMCs) were collected via density gradient centrifugation over Ficoll-Paque (Uppsala, Sweden). PBMCs (10–20 × 107 cells) were suspended in 18 mL (6 × 3 mL) of RPMI 1640 medium (Iwaki, Japan) containing streptomycin and penicillin (Gibco, NY), fortified with 1% autologous plasma and allowed to adhere on a plastic surface for 2 hours. Nonadherent cells were removed via gentle washing and were frozen at −140°C into RPMI/autologous plasma/DMSO for later use as effector cells in immunological end-point assays. The adherent cells were cultured at 37°C in RPMI 1640 containing 1% autologous plasma, antibiotics and supplemented with granulocyte monocyte colony-stimulating factor (800 U/mL; Genzyme, Cambridge, MA) and IL-4 (500 U/mL; Genzyme). Fifty percent of this DC medium was exchanged on days 2 and 4. Immature DCs were harvested on day 8 and pulsed with lysate preprepared from the HepG2 cell line for 2 hours, before undergoing a maturation step in culture medium containing tumor necrosis factor α (20 ng/mL, TCS CellGenix). Finally, the mature DCs were washed, resuspended in 20 mL of phosphate-buffered saline, and reinfused intravenously into the patient. At all stages of culture, samples were taken for bacteriological assessment. Small aliquots of DCs for each patient were retained and frozen for subsequent immune end-point assays.

DC Phenotyping.

Where possible, immature and mature DCs generated for reinfusion were phenotypically characterized for specific DC markers via flow cytometric analysis using an EPICS XL-MCL flow cytometer (Beckman Coulter). Cells (1 × 105) were incubated for 1 hour with mouse monoclonal antibodies against human MHC class II (Dako), human CD11c (Beckman Coulter), human CD86 (Dako), and human CD83 (Beckman Coulter), washed, and then incubated for 1 hour with the appropriate fluorescein isothiocyanate–conjugated second antibody. After washing, cells were fixed in phosphate-buffered saline containing 1% paraformaldehyde before analysis.

Immune Monitoring.

Immune monitoring was performed using cryopreserved PBMCs or nonadherent cell populations remaining after DC harvest as a source of responder cells in ELIspot assays of interferon-γ (IFN-γ) release. After thawing, cells were recovered overnight in RPMI 1640 medium containing 10% fetal bovine serum. IFN-γ release was measured using a commercially available ELIspot kit (ELIspot assay for human IFN-γ; Mabtech) using cytokine capture and detection reagents according to the manufacturer's instructions. IFN-γ release was measured in response to the vaccine (autologous DCs pulsed with HepG2 tumor cell lysate) and in response to autologous DCs transduced with recombinant adenoviruses (infected for 48 hours at a multiplicity of infection of 70) encoding either AFP (RAdAFP) or β-galactosidase (RAdβGAL) as a negative control. The Epstein-Barr virus immunodominant antigen, EBNA-3, was used as a positive control. 1 × 105 responder cells and 1 × 104 target DCs were added to each well. Background wells containing PBMCs and autologous DCs alone were also included. Spots were counted using an automated system.



Treatment was performed in the Wellcome Trust Clinical Research Facility, Queen Elizabeth Hospital, Birmingham, UK. Thirty-nine patients were enrolled. Baseline characteristics are summarized in Table 1. The basis of the diagnosis of HCC was histological for the majority of patients. The male/female ratio followed that described for HCC in the literature. The age range was 18–78 years, the distribution appearing to be bimodal with two distinct peaks at 22 years and 62 years, reflecting the younger age of five patients with fibrolamellar HCC. Known predisposing disease was present in approximately half of patients, with alcoholic liver disease accounting for the majority of these. Routine liver biopsy of noncancerous liver was not performed, so this may be an underestimate of patients with underlying chronic liver disease. Three patients with hepatitis B or C were enrolled prior to viral serology results being available, and two of these patients received at least one vaccination. A small number of patients were previously treated with surgical resection, chemo-embolization, or systemic chemotherapy. Twenty patients were performance status 2 prior to entry into the trial, and only five patients were performance status 0, reflecting the general debility within this patient population. In 23 patients, the baseline serum AFP was ≥1,000.

Table 1. Patient Demographics
Patient CharacteristicNumber
Total number39
Median age (range)22
 63 (18–78)
Predisposing condition 
 Alcoholic liver disease13
 Hepatitis B/C3
 α1-Antitrypsin deficiency1
 European Association for the Study of the Liver criteria12
Performance status 
Prior treatment 

Toxicity Assessment.

In total, 134 DC infusions were administered to 35 of the 39 recruited patients (range, 1–6; median 3 cycles per patient). Four patients did not receive any DCs. In three cases this was due to clinical deterioration prior to the first infusion, and in one case it was due to hepatitis B positivity.

Intravenous infusion of DC was safe and well tolerated. Toxicity, summarized in Table 2, was mild and self-limiting. Eight patients experienced grade 1 myalgia, occasionally associated with low-grade fever. Two patients experienced nausea and vomiting, thought to be related to disease rather than DC infusion. No grade 3 or 4 toxicity was observed in any patient. Importantly, no hepatic toxicity nor de novo autoantibody formation was observed.

Table 2. Toxicity
ToxicityCTC GradePatient Number
  1. Abbreviation: CTC, common toxicity criteria.


Clinical Response Assessment.

Twenty-five of 39 recruited patients received three or more DC infusions. We considered it likely that at least two infusions would be required to stimulate an immune response, and for this reason objective response assessments were confined to these patients (Table 3). One patient with histologically confirmed multifocal HCC on a background of alcoholic liver disease achieved a radiological partial response in one measurable lesion and disease stabilization with evidence of tumor necrosis in the other lesion (Fig. 1A). This was mirrored by a fall in serum AFP from a pretreatment baseline of 4,310 ng/mL to a minimum of 365 ng/mL after three vaccinations (Fig. 1B). The progression-free interval was 6 months. A further six patients achieved disease stabilization for between 6 and 16 months following entry into the study. Thus, the disease control rate was 28% (7 of 25 assessable patients).

Table 3. Response Assessment
  • *

    Fall in AFP to <30% baseline value.

  • AFP <1,000 at baseline or <3 cycles received.

Complete response (CR)00
Partial response (PR)14*
Stable disease (SD)64
Progressive disease (PD)189
Not assessable1422
Figure 1.

Response assessment. (A) Contrast-enhanced axial CT scans through the liver. Baseline scan demonstrates an 8 × 3 cm lesion in the caudate lobe. After three DC infusions, this lesion measured 3 × 2 cm at the same level. This response was maintained after the sixth infusion. A large complex lesion in the left lobe remained static throughout. (B) Serial serum AFP measurements in the same patient.

Seventeen of the 23 patients with serum AFP values ≥1,000 ng/mL at study entry were eligible for serological response assessment (Table 3). In 4 of these 17 patients there was a decline in AFP to less than 30% of the baseline value following vaccination (Fig. 2). As already discussed, in one patient a fall in AFP to less than 10% of the baseline value corresponded to a radiological partial response (Fig. 1B).

Figure 2.

AFP responses in patients with baseline serum AFP ≥1,000 ng/mL. Data plotted logarithmically as the relative change from prevaccination baseline.


The median survival of the 35 treated patients was 168 days (Fig. 3). Six-month and 1-year survival rates were 33% and 11%, respectively. Exclusion of the five patients with fibrolamellar HCC did not alter these data. Note that this patient population comprised those who had exhausted conventional treatment options, often with poor prognostic factors, including poor performance status, Child-Pugh B cirrhosis, large tumors, and macroscopic vascular invasion.

Figure 3.

Kaplan-Meier survival curve.

DC Phenotyping.

The percentage expression of cell surface markers MHC class II, CD11c, CD86, and CD83 was analyzed, where possible, on immature and mature DC preparations. In some cases this was not possible due to low cell numbers. Representative results are shown in Table 4. The level of expression of MHC class II and CD86, molecules involved with antigen presentation, was with few exceptions very high on both immature and mature populations. The cell surface molecule CD11c, a marker of myeloid DC populations was expressed at high levels, indicating that the monocyte-enriched adherent cells had differentiated into DCs in vitro; expression of high levels of CD83 on the majority of the mature cells confirmed that the maturation conditions used generated mature DCs.

Table 4. Percentage Expression of Cell Surface Phenotype Markers on Immature and Mature Dendritic Cell Populations Prepared for Reinfusion
PatientExpression (%)
MHC Class IICD11cCD86CD83
Patient 17Immature97.497.890.855.9
Harvest 5Mature95.788.888.968.5
Patient 18Immature96.397.493.58.8
Harvest 5Mature93.793.981.020.4
Patient 19Immature94.296.495.928.6
Harvest 1Mature77.683.775.552.7
Patient 20Immature83.227.413.71.9
Harvest 1Mature62.
Patient 21Immature7.
Harvest 3Mature77.294.575.761.5
Patient 22Immature6.993.395.931.9
Harvest 1Mature38.891.088.341.6

Immune Monitoring.

ELIspot assays of IFN-γ release were performed using responder cells from 10 patients saved at each harvest where possible. Three gave uninterpretable results because of low cell numbers and poorly reproducible controls; two patients showed no response to the antigenic targets tested. Figure 4 summarizes the results from the remaining five patients, expressed as the fold increase in IFN-γ release over background levels using the appropriate negative controls. Cells were not available for testing at every harvest for every patient, but responder cells from all five patients demonstrated increased specific IFN-γ release to an antigenic target at some point during their participation in the trial. Particularly good responses were observed from patient 17 to autologous DCs pulsed with HepG2 lysate (the vaccine) and DCs expressing AFP at harvest 4 corresponding to a radiological partial response and fall in AFP to less than 10% of baseline (Figs. 1 and 4). The vaccine also induced good responses from patients 23, 27, and 28 that increased steadily during the trial. In addition, responses were seen to autologous DCs transfected with RAdAFP in patients 23 and 26, although these reduced in the later harvests.

Figure 4.

Immune monitoring. ELIspot asssays of IFN-γ release were performed using PBMCs or nonadherent cell populations as responder cells (2 × 105 cells/well). Responder cells were tested against autologous DCs (1 × 104 cells/well) (A) pulsed with HepG2 lysate or (B) transduced with recombinant adenoviruses encoding either AFP (gray bars) or β-galactosidase as a negative control (white bars). Results are expressed as the fold increase of experimental wells over background wells containing responder cells and autologous DCs only.


HCC possesses several characteristics that render it an attractive target for immunotherapy.13, 17 There is active recruitment of tumor-infiltrating lymphocytes able to recognize tumor-specific antigens as evidenced by their ability to lyse autologous tumor cells ex vivo. However, the fact that HCC develops in immunocompetent hosts indicates a failure to mount an adequate antitumor immune response naturally. Mechanisms by which tumors evade immune detection may include down-regulation of MHC and, through regulatory T cells, secretion of immuno-inhibitory cytokines.18

DCs have a unique ability to take up, process, and present antigen, which in the right context can stimulate both CD4 and CD8 antigen-specific T cell responses. However, DC function in HCC is suppressed as a result of local factors, including IL-10, and this leads to a failure of mature DCs to induce antitumor immune responses.19, 20 There is thus a rationale for activating DCs in vitro to overcome any tumor associated immunosuppression and reinfusing them into patients to stimulate antitumor immunity. The development of DCs as a platform for therapeutic vaccines is enhanced by the ability to harvest them from peripheral blood in sufficient numbers, and to manipulate them ex vivo to present antigens of interest.21

The study reported here demonstrates clinical antitumor activity of a DC vaccine against HCC in conjunction with evidence for the generation of antigen-specific immune responses. It is particularly encouraging that we demonstrated clinical responses associated with a logarithmic fall in serum AFP and the generation of immune responses following vaccination, suggesting that the generation of immune responses against the tumor is responsible for the clinical responses seen.

This study illustrates several key points in the development of immunotherapy for HCC. First, we demonstrate the feasibility of generating autologous DCs from this group of patients despite advanced malignancy and often coexistent chronic liver disease. These DCs can be loaded ex vivo with multiple antigens from lysates of a liver cancer cell line and matured using the appropriate cytokine cocktail prior to reinfusion. Although hepatitis B and C patients were excluded from this protocol, this was for the safety of laboratory workers and, with appropriate containment facilities, such patients should be included in future studies. Second, we show that this vaccine is safe when administered intravenously, with no significant toxicity and despite loading with multiple antigens from a whole cell lysate, no evidence of autoimmunity. Third, we report evidence of clinical response, with one patient achieving a radiological partial response associated with a significant decline in the serum AFP tumor marker. Although a single radiological response should not be overinterpreted, evidence of clinical activity is supported by our finding that several other patients achieved disease stabilization, and in three patients this was associated with significant reductions in serum AFP measurements. Finally, we identified immune responses following vaccination using ELIspot as a functional measure of the frequency of T cells that produce the Th1 immunostimulatory cytokine, IFN-γ, when stimulated by a specific antigen. This technique has a potential advantage over tetramer analysis, which gives no information about the functionality of the T cells detected. Although analyses were not possible in all cases due to low yields and poor viability of effector cells after cryopreservation in some patients, in those we were able to study there was clear evidence of antigen-specific T cell responses to the vaccine and to AFP. In the patient who achieved both a radiological and AFP response, this correlated with induction of immune responses to AFP ex vivo.

There remain several unanswered questions in the further development of DC vaccines for HCC. These include the optimal maturation status of the DC, the choice of tumor antigens, and the route of administration of the vaccine. Evidence suggests that immature DCs may induce anergy rather than an antitumor immune response, perhaps via secretion of Th2 type cytokines, whereas mature DCs, through high levels of MHC and costimulatory molecules, tend to induce and sustain a specific immune response.22 For this reason, we chose to infuse tumor necrosis factor α–treated DCs, which resulted in a mature CD83+ DC population. This may have contributed to the encouraging clinical responses seen in our study that were lacking from a previous study using immature DCs.9

The optimal antigens for incorporation into a DC vaccine for HCC is also an unresolved question. AFP is a potential candidate antigen. One might expect that because AFP is a nonmutated self antigen highly expressed in fetal liver, responsive T cells would be largely deleted. However, studies have reported that AFP-specific T cells can be detected in healthy individuals and in cirrhotic and HCC patients, and our study demonstrates that these can be restimulated and expanded in patients with HCC even in the presence of high serum levels.9, 15, 23–25 Other studies have also demonstrated an ability to induce T cell responses against AFP. For example, a study using DCs pulsed with human leukocyte antigen (HLA)-A0201–restricted AFP peptides reported the induction of AFP-specific CD8+ T cell responses, though no clinical responses were seen.9 The immunodominant epitopes from AFP are not well characterized. Published studies suggest that both dominant and subdominant epitopes are recognized by the human T cell repertoire in patients with HCC resulting in expansion of IFN-γ producing effector cells. The response is complex because T cells specific for subdominant epitopes are of similar or higher avidity than those specific for immunodominant ones, and in vitro DCs stimulate broad responses.20

Because no single antigen is ubiquitously expressed by HCC, we used a whole cell lysate as a source of multiple antigens (including AFP). An advantage of this approach is that DCs pulsed with whole protein will be more effective at eliciting antigen-specific responses from both CD4 helper T cells and CD8+ cytotoxic T cells to multiple antigens, independent of HLA type. Activation of both CD4+ and CD8+ T cells appears to be a prerequisite for a robust antitumor immune response,26 and the generation of T cells against multiple antigens may reduce the risk of tumor escape through down-regulation of a single specific antigen. We decided to use the hepatoblastoma cell line HepG2. It expresses high levels of AFP and is readily available, making it a pragmatic choice for a clinical study. Other liver cancer cell lines could equally have been selected. Autologous tumor lysate was not used because of unpredictable availability and potential variability in antigen expression between individuals that might limit the investigation of T cell responses. Previous studies have shown that specific anti-HCC T cells can be generated in vitro by DCs pulsed with HepG2 total RNA.27 We found that a combination of freeze thawing and sonication resulted in a lysate that could be easily handled, quantified, and used to pulse DCs. Preliminary studies with labeled lysates showed that it was taken up and processed appropriately by the DCs. The use of whole cell lysate may explain the encouraging evidence of clinical responses in our study. Some previous studies using HLA-restricted peptides failed to induce clinical responses, and this could be a consequence of the lack of CD4 help.9 Conversely, whole cell lysate may increase the risk of autoimmunity through shared epitopes with normal tissues. In this sense, the safe use of our vaccine without evidence of autoimmunity is reassuring.

The optimal route of vaccination also remains to be determined. Because antigen presentation to T cells by DCs normally occurs in lymph nodes, it appears logical to deliver vaccine via the route that optimizes their access to regional nodes. However, it is not yet known whether this requires direct inoculation to the regional nodes (which for the liver would be technically challenging), or whether a more convenient route such as inguinal nodes, intradermal, or intravascular (either through the hepatic artery or intravenously) might allow DCs to home to nodes. This process may be influenced by the maturation status of the DCs. Previous studies have shown that immature DCs injected into the tumor artery survive for several weeks within the tumor and are capable of generating T cell responses against tumor antigens.11, 28, 29 A recent study inoculating lysate-loaded DCs either intratumorally, subcutaneously or intravenously in a murine subcutaneous HCC model suggested intratumoral to be the optimal route. However, these DC were also engineered to express CD40 ligand, which may directly induce apoptosis in HCC when administered intratumorally and might explain the apparent superiority of this route rather than more potent immune stimulation.30 Other studies suggest that DCs may be trapped inside the tumor and are unable to migrate to draining lymph nodes; thus, whether intratumoral therapy is more effective than systemic delivery is yet to be determined.28, 29

This study was not able to monitor the migration of DCs following intravenous administration. Our ongoing studies aim to address the question by tracking indium-labeled DCs given by different routes of administration.

In conclusion, autologous DCs provide a practical basis for a therapeutic tumor vaccine.31 We report evidence of safety and immune activity, with some evidence of clinical response for patients with HCC using an autologous DC vaccine pulsed with tumor cell lysate. The fact that we were able to detect immune activity and clinical responses in a proportion of patients with advanced tumors and chronic liver disease is encouraging and suggests that future trials in less advanced disease may be associated with even better clinical responses. In future, immunotherapy should probably be used as an adjuvant to radical therapy. In particular, local ablation with radiofrequency ablation not only kills tumor cells but also causes tissue damage and the release of tumor antigens that evoke antitumor immune responses.32–34

Immunotherapy as an adjuvant to surgical resection may require greater caution, because liver autoimmunity has been reported in response to AFP DNA vaccination in a mouse model of liver regeneration following partial hepatectomy, a setting known to be associated with increased AFP expression,35 although other studies suggest no impairment of liver regeneration in this context.36

Further improvements in efficacy may come from combinatorial approaches to generating antitumor immune responses in which DC vaccination is combined with strategies to overcome regulatory T cells or the inhibitory effect of Th2 cytokines, or with conditioning chemotherapy.21