Introduction

Hepatocellular carcinoma (HCC) is a malignant neoplasm of hepatocytes. It is a frequently occurring tumor, ranking fifth in worldwide prevalence and third in mortality [1, 2]. HCC is a worldwide health problem, but it is especially prevalent in South-East Asia and sub-Saharan Africa. Globally, the incidence of HCC is increasing. This is especially due to a rise in the incidence of infections with the hepatitis C virus (HCV) [3]. Chronic infections with HCV and the hepatitis B virus (HBV) induce a chronic inflammation which can lead to liver fibrosis and subsequently cirrhosis, i.e., a replacement of liver tissue by connective tissue and regenerative nodules. Of note, chronic hepatic inflammation and liver cirrhosis can also be caused by several other factors such as alcohol abuse or hereditary liver diseases [4]. This inflamed and cirrhotic environment can ultimately give rise to HCC lesions.

The diagnosis of HCC is particularly difficult since it usually remains clinically asymptomatic until late stages of disease. However, even at early stages, therapeutic options for HCC are limited [5, 6]. Treatment options include liver transplantation, partial liver resection, or less invasive procedures such as radio-frequency thermal ablation (RFTA) and trans-arterial chemo-embolization [6]. Several new molecularly targeted therapies are currently under investigation. So far, sorafenib, a multi-kinase inhibitor, has been approved for treatment of patients with advanced HCC but has only shown modest clinical benefits [7]. Thus, there is an urgent need for new therapeutic options for HCC to treat end-stage patients or to limit the rate of recurrence of HCC after liver transplantation or liver resection.

Immunotherapy is a relatively new approach to tumor therapy. The aim of immunotherapy is to boost the immune system in order to destroy pathogens or malignantly transformed cells [8, 9]. For the development of such a complex therapeutic regimen, a detailed understanding of the natural role of the immune system in the development and control of HCC lesions is required.

The role of the immune system in hepatocarcinogenesis

Tumors develop from single cells that accumulate mutations over time. These mutations alter the expression and function of genes responsible for cellular proliferation among others. This process ultimately leads to cells with uncontrolled proliferative and metastatic potential [10]. The exact molecular nature of these changes in HCC has been reviewed elsewhere and is not in the scope of this review [11, 12]. In brief, it could be demonstrated that HCC lesions accumulated diverse genomic alterations over time, whose number correlated with the pathological appearance of the lesions [13]. Especially signaling pathways such as Hedgehog and Wnt/β-catenin pathways were altered during hepatocarcinogenesis [14, 15]. Some of these changes can also be induced by chronic viral hepatitis, the leading cause for HCC [16]. For example, chronic HCV infection induces changes at the level of lipid metabolism and gene expression in the liver, whereas the HB antigen derived from HBV is believed to be capable of altering the transcription of several cellular genes which might also contribute to HCC development [17, 18]. Apart from these direct changes, chronic viral hepatitis also induces a chronic inflammation of the liver. Indeed, in the HBV transgenic mouse model, it could be shown that this chronic inflammation alone, i.e., in the absence of a productive viral infection, was sufficient to induce hepatocarcinogenesis [19]. Inflammation is associated with an abundance of different cytokines and growth factors in order to repair the tissue by cellular proliferation. However, in chronic inflammation, this can lead to permanently increased cellular turnover and finally the development of various malignancies [20]. This finding also elucidates why different causes of chronic hepatitis ultimately give rise to HCC. In HCC, especially cytokines of the lymphotoxin (LT) family have been associated with tumor development [21]. LTs are produced by a variety of lymphocytes such as T cells or natural killer (NK) cells, thereby demonstrating that the adaptive as well as the innate immune response are involved in the development of HCC. Interestingly, a recent study was able to demonstrate that the rate of DNA double strand breaks is increased in a mouse model of chronic hepatitis [22]. This directly contributes to mutagenesis and thus carcinogenesis. In addition, Toll-like receptors (TLRs) have been suggested to play a role in the development of HCC [23]. TLRs are pattern-recognition receptors that normally recognize conserved danger signals found on pathogens or also on necrotic cells, so called pathogen-associated molecular patterns. They are capable of inducing inflammatory processes, again supporting a role for chronic inflammation in hepatocarcinogenesis [24]. Indeed, the activation of TLR9 by unmethylated CpG DNA induced the proliferation as well as the expression of several anti-apoptotic proteins in hepatoma cells [25].

A direct role in HCC progression has also been shown for other cells of the immune system. For example, the detection of intratumoral neutrophils has been shown to predict a reduced recurrence-free survival time of HCC patients after liver resection [26, 27]. Of note, neutrophils the tumor edge induced angiogenesis and thus directly enhanced tumor growth [28]. This observation becomes especially intriguing in light of the finding that neutrophils can be recruited by interleukin-17 (IL-17) [29]. IL-17 is produced by T cells, termed TC17 or TH17, in the CD8+ or CD4+ T cell compartment, respectively. Indeed, IL-17 producing cells have been found to be enriched in HCC lesions and to correlate negatively with overall survival of HCC patients after liver resection [30]. Furthermore, TC17 and TH17 cells were also found to be enriched in the peritumoral stroma, where their numbers correlated with those of tumor-associated monocytes (TAMs) that have been shown to readily induce expansion of IL-17 producing cells [31, 32].

This data supports a model in which the immune system is creating an environment of increased cellular turnover which prone to mutagenesis and thus ultimately the development of neoplastic lesions. These data also support the concept that tumors of a certain size gain the ability to manipulate immune cells for their own benefit.

Immune responses against HCC

In HCC, the immune system certainly can induce hepatocarcinogenesis. However, several lines of evidence also suggest a protective role, e.g., by controlling tumor growth. Indeed, this assumption is based on several findings. First, HCC patients with an intratumoral accumulation of lymphocytes had a superior 5-year survival rate and a prolonged recurrence-free survival after liver resection [33, 34]. A certain level of protection was especially conferred by cytotoxic CD8+ T cells [35]. Of note, these tumor infiltrating lymphocytes (TILs) were associated with an inflammatory microenvironment that was a predictor of overall patient survival, indicating a protective role of TILs in HCC [36, 37]. Furthermore, a strong CD8+ T cell response against several tumor-associated antigens (TAAs) was found to coincide with improved recurrence-free survival after liver resection [38]. The important role of CD8+ T cells in HCC control is further supported by a study in mice. Here, interferon γ (IFNγ) produced by CD8+ TILs could be shown to be one effector mechanism for apoptosis of hepatoma cells [39].

Of note, the antigen specificities of TILs in HCC patients are currently still under investigation. To date, several TAAs of HCC have been identified. This has been reviewed recently by Breous and Thimme [8]. Briefly, several shared tumor-specific antigens could also be identified as antigens targeted by CD8+ T cells in HCC, e.g., human telomerase-reverse transcriptase (hTERT) or NY-ESO-1. Expression of these antigens has been reported for other malignancies as well. They comprise genes required for sustained proliferation such as hTERT. Other antigens are expressed specifically in HCC and are also recognized by cells of the immune system, e.g., α-fetoprotein (AFP) or Glypican-3. The latter two antigens belong to the family of oncofetal antigens. Oncofetal antigens are expressed during ontogenesis but not in the adult. However, re-expression of such antigens is frequently observed in tumors. The exact reasons for the specific reexpression of these antigens in HCC are still not completely clear. Further research is also required to determine the frequency, immunodominance and strength of the immune responses induced against different TAAs.

Next to HCC-specific cytotoxic T cells, other cell types have also been implicated in a successful immune response against HCC. For example, NK cells play an important role in anti-tumor immune responses, e.g., by direct lysis of malignant cells [40]. Indeed, in HCC patients an increased preoperative NK cell activity correlated with prolonged recurrence-free survival [41]. Furthermore, a decreased functionality of NK cells was shown to predict the occurrence of HCC [42]. Other studies also showed that the stimulation of natural killer T cells (NKT cells) leads to an activation of NK cells and clearance of hepatoma cells in mice [43].

The role of B cells in HCC control is unclear thus far. Auto-antibodies against several antigens have been described in mouse models of HCC as well as in HCC patients [44, 45]. Additional studies showed that monoclonal antibodies against Glypican-3 were able to induce antibody-dependent cellular cytotoxicity (ADCC) and thus lysis of human hepatoma cells in vitro and in mice [46]. However, the importance of ADCC and antibodies in general in the natural immune response against HCC has not been investigated so far.

Factors contributing to a successful immune response against HCC are summarized in Fig. 1.

Fig. 1
figure 1

Components of a successful immune response against HCC. Macrophages (M θ ), or other APCs such as DCs, take up material from necrotic tumor cells. Next, they become activated and present epitopes derived from TAAs on their respective MHC class I and class II molecules. Subsequently tumor-specific CD4+ helper T cells (T H ) are activated and support the activation of CD8+ cytotoxic T cells (T C ). Activated TC are able to lyse tumor cells that present the respective epitope on their MHC class I molecules. NK cells can lyse tumor cells directly or upon encounter of antibodies specific for antigens on the surface of tumor cells. The latter process is termed antibody-dependent cellular cytotoxicity (ADCC). Antibodies specific for TAAs are produced by B cells. Cytokines secreted by NKT cells and TH modulate the activity of TC and NK cells as well as the class switch of B cells

Full size image

Mechanisms of failure of immune responses against HCC

In most patients, HCC-specific immune responses fail to control the tumor. This failure is partly due to the fact that HCC evades the immune response. The growing tumor is exposed to various effector mechanisms by which especially cytotoxic CD8+ T cells and NK cells lyse tumor cells [47, 48]. Thus, tumor cells that have acquired mutations allowing them to escape are selected by the pressure exerted by the immune system. This selection of tumor cells that are able to subvert the anti-tumor immune response is termed cancer immunoediting [49]. Indeed, HCC lesions were frequently found to have lost expression of Fas (CD95) and the IFNγ receptor. Both are important effector mechanisms usually engaged by cytotoxic cells of the immune system [50, 51].

On the other hand, a dysfunctionality of tumor-specific CD8+ T cells in HCC patients has also been reported. These cells were found to express inhibitory receptors such as programmed death-1 (PD-1). Additionally, they displayed downregulated costimulatory molecules such as CD28 or components of the T cell receptor complex, e.g., CD3ζ [52–54]. Accordingly, these cells were functionally impaired and prone to apoptosis. A recent study could also show that this impaired CD8+ T cell phenotype predicted an earlier recurrence of HCC after liver resection [55]. This supports the hypothesis that the failure of the immune response to HCC is mainly a failure of tumor-specific CD8+ T cells.

There are probably several reasons for this functional impairment. First, PD-1 on tumor-specific CD8+ T cells is engaged by its ligand programmed death-ligand 1 (PD-L1). PD-L1 is expressed on tumor cells and intratumoral Kupffer cells [56]. Second, DCs were described to have a distinct phenotype in HCC [57]. These cells were found to secrete less IL-12 than DCs from controls. IL-12 is required for proper activation and differentiation of T cells. Thus, this particular phenotype of DCs might lead to an improper activation of CD8+ T cells and thus help to explain their dysfunctional phenotype. Third, regulatory T cells (Treg) that are highly capable of suppressing CD8+ T cells were readily detectable in HCC. Indeed, the fraction of Treg among TILs correlated negatively with clinical outcome [35]. In addition, an increase in the number of Treg could be shown for HCC patients in the blood as well as in the liver [58, 59]. These immunosuppressive Treg were also found to directly infiltrate HCC lesions [60]. Indeed, a direct functional impairment of CD8+ T cell function by increased Treg numbers could be demonstrated for HCC patients [61]. Of note, an increase in Treg numbers was also shown to coincide with progressing hepatocarcinogenesis, i.e., the more malignant the tumor, the higher the Treg count observed [62]. Further studies revealed that Treg are induced by supernatants derived from HCC, specifically by factors secreted by TAMs [63, 64]. Thus, HCC actively recruits Treg in order to keep CD8+ T cell responses suppressed. Fourth, another immunosuppressive cell type termed myeloid-derived suppressor cells (MDSCs) has recently gained increasing interest in the field of tumor immunology. MDSCs are a mixture of diverse immature cell types of myeloid origin that are able to dampen immune responses. In HCC, MDSCs were found to infiltrate the tumor and to induce the Treg phenotype in naïve T cells [65].

As described, NK cells are also potent effectors against tumor cells. Interestingly, NK cell number and function were found to be generally reduced in HCC patients, which might point to sequestration or deletion of these cells [66, 67]. The reasons for this failure of NK cell responses in HCC patients are still under investigation. There is evidence that hepatoma cells can secrete a soluble form of MHC class I-related chain A (MICA) [68]. MICA is a ligand for the activating NK cell receptor NKG2D and a soluble form of this ligand might thus prevent proper NK cell activation at the tumor site. Other studies also revealed a role for MDSCs in suppression of NK cell function in HCC [69]. Thus, the functional impairment of immune cells is not restricted to CD8+ T cells but also comprises other cytotoxic cell types.

The role of CD4+ helper T cells in HCC remains elusive. A recent study in our group found that there is a pronounced lack of helper T cells specific for AFP in blood, liver, and tumors of HCC patients especially when compared to CD8+ T cell responses against the same antigen [70, 71]. Another recent study proposed that AFP specific CD4+ and CD8+ T cells expanded at different stages of disease [72]. However, the exact mechanism of this apparent insufficiency of helper T cells in HCC is still elusive and further research, especially for other TAAs, is still urgently required.

The mechanisms leading to an impairment of the immune response against HCC are summarized in Table 1. Taken together, many pathways have been shown to contribute to the apparent failure of the immune system in HCC. Further research is needed to better understand this failure and eventually be able to overcome it.

Table 1 Mechanisms of failure of tumor-specific immune responses
Full size table

Perspectives for treatment

There is urgent need for additional therapeutic options for HCC. Liver transplantation is a curative option, but due to the shortage of donor organs, it is only applicable for a few patients. Instead, surgical removal of the tumor by liver resection is a frequently applied treatment. RFTA is another potentially curative treatment but only suitable for early stages of disease [6]. However, the frequent occurrence of recurrences limits the overall success rate of these approaches [73]. As shown earlier, the rate of recurrence is largely influenced by the lymphocyte infiltrate found within the tumor. Presumably, tumor-specific CD8+ T cells and NK cells are able to control the tumor cells remaining after resection and thus limit the rate of recurrence. A therapeutic approach that effectively enhances this tumor-specific immune response might thus be a very promising adjuvant approach to conventional surgical therapy. Indeed, the efficacy of autologous lymphocytes activated in vitro using anti-CD3 antibodies and recombinant human IL-2 in lowering recurrence rates and improving post-surgical survival could be demonstrated [74]. Clearly, this is one of the current aims of immunotherapy, although further research is still required.

Interestingly, recent studies also showed a role for some currently available therapeutic approaches in boosting the natural immune response against HCC. For example, it could be shown that RFTA increases the costimulatory capacity of antigen-presenting cells (APCs) and boosts NK cell activity [75, 76]. This is probably mediated by the release of tumor material by necrosis and thus in a pro-inflammatory context. This observation leads to the hypothesis that RFTA or similar ablative therapies could be applied prior to partial liver resection in an attempt to boost the immune response and thus limit the rate of recurrence after surgery. By this approach, the apparent power of the immune system could be used without the laborious and costly requirements of adoptive transfer of in vitro-activated cells, as described above. However, due to the novelty of some of these observations, this procedure has not been tested prospectively in the clinical setting.

With a better understanding of the underlying mechanisms of failure of the natural tumor-specific immune response, it might even become possible to apply immunotherapy as a first line of treatment [9]. However, thus far only a few trials testing immunotherapy of HCC have been published. Nevertheless, in mice, the general feasibility of such an approach could already be demonstrated using a DNA-based vaccine containing AFP [77]. In this trial, a partial tumor regression and improved survival coincided with the induction of AFP-specific immune responses. Furthermore, a phase I/II clinical trial using DCs pulsed with AFP peptides showed that immune responses to AFP were also inducible in humans although no clinical response was seen in this study [78]. However, a recent phase II clinical trial was able to show that autologous DCs primed with tumor lysate are well tolerated by HCC patients [79]. Additionally, this immunotherapeutic approach showed some anti-tumor effect as evidenced by stabilization of disease and a drop in serum AFP levels in some HCC patients.

For other malignancies, especially melanoma, immunotherapy is already several steps further into development. Importantly, a vaccine containing recombinant human NY-ESO-1 protein mixed with ISCOMATRIX adjuvant was able to achieve stabilization of disease in some patients [80]. As mentioned above, the immunogenicity of this antigen was also proven for HCC, raising hopes that vaccines developed for other malignancies might also be applicable for HCC patients. Other approaches in HCC immunotherapy focused on the immunomodulatory role of NKT cells in order to orchestrate a comprehensive immune response involving multiple effector arms [81, 82]. However, there is little data so far and this approach thus requires further investigation.

In conclusion, immunotherapy holds promise to become a useful adjuvant treatment to support current surgical regimens in the near future. The possible use of immunotherapy as a stand-alone therapy, however, still requires and warrants further investigation.

Summary

HCC is a malignancy of increasing importance due to a rising incidence accompanied by limited treatment options. The immune system plays a pivotal role in the development of HCC as chronic inflammation is one of the most important causes of HCC. Limiting such inflammations might thus be a promising approach to reduce the occurrence of this disease. However, once the tumor is established, the immune system is usually able to confer a certain level of protection against the tumor, at least against recurrence after surgical removal. Enhancing this natural anti-tumor immunity and breaking the immunosuppressive barriers established by HCC are central aims of immunotherapy and might in the future be of benefit for patients’ survival.