1. Introduction to Liver Cancers

1.1. The Liver, an Extraordinary Organ with Multiple Functions

The liver is an essential organ involved not only in digestion but also in many other functions, among them immunotolerance ^[1]. Indeed, it is continuously exposed to antigens from food intake, gut microbiome, and possibly from pathogens. Kupffer cells, dendritic cells, and lymphocytes T participate in this tolerance.

Any dysfunction of the liver can therefore lead to important metabolic or immunological disorders, eventually threatening the survival of the organism and/or leading to liver tumorigenesis. In 2020, liver cancer ranked 6th in the world in terms of incidence (around 906,000 new cases) and 3rd in terms of mortality (around 830,000 deaths), with disparities between men and women as incidence and mortality rates were 2–3 times higher in men ^[2]. Liver cancer is expected to increase by 50% in the next 20 years with major geographical disparities in prevalence worldwide ^[3].

1.2. Hepatocellular Carcinoma in Adults

Hepatocellular carcinoma (HCC), also called hepatocarcinoma, is an adult primary liver cancer that develops from hepatocytes in 75–85% of cases ^[4] and, most of the time, in a diseased (chronic liver inflammation, fibrosis) and/or cirrhotic organ ^[5]. Other rarer forms of primary liver cancer can develop from cells in the bile ducts (cholangiocarcinoma, 10–15% of cases ^[4]) or, much more rarely, from blood vessels (epithelioid hemangioendothelioma). All causes of chronic liver disease are therefore directly or indirectly responsible for HCC. Among them the most frequent are the hepatitis B and C viruses, representing 56% and 20% of liver cancer deaths worldwide in 2012, respectively ^[6], alcohol abuse, aflatoxin B exposure, smoking, and non-alcoholic steatohepatitis (NASH). Diabetes and obesity are also risk factors of HCC. Hepatocarcinoma’s incidence has been steadily increasing for several years, due to the increase in viral infections or environmental toxins ^[7][8]. The etiology of HCC also lies in alterations of TERT (telomere reverse transcriptase), oncogenes and tumor suppressor genes (P53, etc.), and genes that lead to aberrant cell signaling pathways (Wnt-β-catenin pathway, etc.), in parallel with the development of cirrhosis and liver fibrosis ^[9].

In most cases, symptoms are absent in the early stages of the disease. By the time symptoms such as jaundice and weight loss occur, it is usually too late and the available treatment options extend survival by only a few weeks or months. This is why regular surveillance is recommended for patients with cirrhosis or chronic Hepatitis B or C Virus infections. Usually, it consists in an ultrasound scan, which may be combined with an evaluation of serum markers such as Alpha-fetoprotein (AFP) or Glypican 3 (GPC3).

Staging is a requisite for proposing the best treatment strategy, and it must consider the underlying liver condition that will affect the evolution of the disease and dictate the applicability and effectiveness of the treatment. Around 18 staging classifications have been proposed, and are reviewed in ^[10]. The number and size of nodules and the presence of an underlying liver disease are the most common criteria but the presence of tumor and vascular invasions or the extent of liver involvement is also examined. Some of these staging systems are also of prognostic value. However, it appears that there is no real consensus when it comes to deciding which staging system offers the best classification. Though the main drivers of HCC are numerous, the molecular characterization was thus far of no help for stratification.

Surgical resection and liver transplantation are the best curative options, as they improve the survival chances, at least at an early stage of HCC. Only a quarter of the liver mass is necessary for a human organism to live normally. After resection, the liver has extraordinary regenerative capacities as new cells are rapidly generated to allow the remaining liver to grow back to its original size. Transcatheter arterial chemoembolization (TACE) or radioembolization (TARE) leading to tumor necrosis are proposed for intermediate stage patients. For the most advanced cases, multikinase inhibitors sorafenib or regorafenib can extend life by a few months ^[11][12]. Recently, FDA approved as second line-therapy for advanced HCC immune check-point inhibitors, alone or in combination ^[13][14], and anti-VEGF antibodies as reviewed in ^[14]. Attempts have been made to vaccinate with HCC specific antigens such as AFP, GPC3, or the multidrug resistance-associated protein MRP3. They were not associated with an increased overall survival ^[1]. The complexity of the role of the liver in self-defense and immunotolerance makes it difficult to envision immunotherapy treatment and underlines the urgent need for better understanding the immune environment in HCC to provide efficient treatments for longer survival and better quality of life.

Any treatment that may improve liver conditions, such as steatosis, fibrosis, NASH, could help reduce HCC risk. In this context, statins have been evaluated, as reviewed in ^[13][15]. Statins inhibit the rate-limiting step of cholesterol synthesis catalyzed by the 3-hydroxy-3-methylglutaryl CoA reductase (HMG-CoA reductase). Since their discovery in 1973, they are widely used as lipid lowering agents in the prevention of heart attack and stroke. They are now considered potentially chemopreventive thanks to several observational and experimental studies, although additional investigations are needed to meet these expectations .

1.3. Hepatoblastoma in Children

Hepatoblastoma (HB) is a rare primary liver tumor with a worldwide incidence of 1.5 cases/million children per year. HB usually affects children under 5 years of age ^[16] for whom it is the most common liver tumor with the majority of cases appearing before 18 months. HB derives from parenchymal liver cells or hepatoblasts. It can occur sporadically, in children prematurely born or with a low birth weight, or in children with a family history of genetic diseases such as adenomatous polyposis or Beckwith-Wiedemann syndrome. Children who had hepatitis B at an early age or biliary atresia are at greater risk of developing the disease. HB can metastasize, mostly in the lung and the abdomen.

The mainstay of curative therapy based on clinical parameters and risk assessment ^[17] is a complete surgical resection. When not possible at first line, cisplatin and doxorubicin-based chemotherapies can help reduce the size of the tumor before resection. Four international groups have established standards of risk and treatment strategies that have helped to achieve a five-year survival rate of nearly 80%, up from 27% in the early 1990s ^[18], but the prognosis remains poor in 20% of cases. Hence, the side effects of chemotherapy and the management of the refractory cases remain to be improved.

Cisplatin is known to induce cytotoxicity, neurotoxicity, and nephrotoxicity, and cardiotoxicity of doxorubicin limits its use to treat HB. Other platinum derivatives such as carboplatin and oxaliplatin induced less cytotoxicity but were also less effective than cisplatin in HB treatment. These drugs induce irreversible DNA damage leading to cell death. The mechanisms of HB resistance to platinum derivatives could arise from DNA-unrelated effects of platinum and still need to be clarified ^[19]. Improvements in the understanding of HB are therefore of utmost importance. This implies answering several fundamental questions, including that of the HB driver genes and the evaluation of their drug-generability.

HB is associated with a significant increase in serum alpha-fetoprotein, the level of which is used as a prognostic and diagnostic marker ^[20]. β-catenin and the Wnt pathway were identified as key drivers of sporadic HB pathogenesis ^[21]. The Wnt/β-catenin has a key role in the development, regeneration, and zonation of the liver ^[22]. However, the list is far from close as cancers emerge as complex pathologies with different outcomes even though deriving from the same organ. For instance, in 2008, a 16-gene signature was proposed to discriminate two distinct HB subgroups ^[23], C1 and C2, and was further adopted.

More recently, a transcriptomic study was performed on tumor and non-tumor tissues from 22 hepatoblastoma patients. A 4-gene molecular signature was thus proposed to classify HB in three different types, C1, C2A, and C2B ^[24] based on the previous classification and not overlapping with the 16-gene signature. The C1 patients are at low risk. In the C2 types, which are of poorer prognosis, the C2A group is associated with high proliferation while the C2B type exhibits epithelial-mesenchymal transition features. Though different classifications are now internationally recognized, they are questioned because of the low number of patient samples and seem difficult to be validated in larger cohorts of patients. This stresses more than ever the need for a deeper understanding of the molecular biology of HB to propose other therapeutic options with less toxicity, notably for the chemoresistant HB. Research interests should focus on the crosstalk between two emerging hallmarks of cancer, metabolism and inflammation ^[25], whose better understanding may unravel new leads for such therapeutic perspectives.

2. Therapeutic Targets within the TRP Pathways

In mammals, TRP is a rare and essential amino acid provided only by dietary intake. In addition to being a component of proteins, TRP, is metabolized through the kynurenine (KYN) or the serotonin (5-HT) pathway to lead to the production of physiologically active metabolites, such as kynurenic acid (KYNA) and NAD⁺ or serotonin (5-HT) and melatonin ^[26]. The first and limiting step of the synthesis of 5-HT is catalyzed by the tryptophan hydroxylase (TPH1). The first and limiting step of the KYN pathway is catalyzed by the tryptophan dioxygenase (TDO2) in the liver or the indoleamine dioxygenase (IDO1/2) in immune cells.

2.1. Ongoing Clinical Trials

Given the close relationship between KYN metabolites and inflammatory responses, several drugs targeting COX-2, IDO1, IDO2, or TDO2 are already being tested in clinical trials for the treatment of some cancers (^[27]and Table 1), some of which have already been completed. The results of the most advanced clinical trial (Phase 3) using an IDO1 activity inhibitor, epacadostat (trial ECHO-301) have been disappointing for the treatment of metastatic melanoma ^[28]. Indeed, the combination of epacadostat with pembrolizumab (a reference anti-PD1 treatment in immunotherapy) does not show any beneficial effect of this inhibitor and does not appear to act on the immunosuppressive microenvironment. This failure calls into question the involvement of IDO1 in immunosuppression as well as the usefulness of the IDO1 inhibition strategy. IDO1 is involved in TRP degradation but also in cell signaling and the latter should be targeted along with the catalytic activity ^[28]. Some researchers have suggested that the immuno-modulatory effect of the inhibitors in preclinical studies may be the consequence of an off-target and independent action of IDO1 inhibition ^[29]. However, using the IDO inhibitor 1-methyl-tryptophan (1-MT) in association with diverse chemotherapeutic agents, such as cisplatin, cyclophosphamide, and doxorubicin, could effectively promote the regression of breast tumors known to be refractory to chemotherapy ^[30].

Table 1. List of clinical trials with anti-cancer drugs targeting COX-2, IDO1, or TDO2 in various cancers. Taken from https://clinicaltrials.gov/, last accessed 31 August 2021.

2.2. Other IDO Inhibitors

Other drugs targeting the key enzyme of the kynurenine pathway IDO1 have been used in vitro and/or in vivo and are listed in Table 2. 1-MT exists as two stereoisomers, 1-D-MT and 1-l-MT and IDO1 is the preferential target of 1-l-MT, while 1-D-MT, which has been used in clinical trials, preferentially inhibits IDO2 ^[30][31]. Despite the bulk of evidence supporting a role for IDO1 in promoting tumor formation and tumor immune escape, there have been clinical studies showing an anti-tumor effect of IDO1 by the intermediate of IFN-γ, which was effective in the therapy of ovarian carcinoma and bladder cancer ^{[32][33][34][35]}. At the preclinical level, a study on pre-immunized mice with tumor antigen has shown that the tumor volume was reduced with 1-methyl-tryptophan (1-MT) and a regression of established breast cancer was observed when 1-MT was combined with chemotherapy [36]. However, other studies have shown that IDO1 expression in tumors positively correlated with progression-free survival and long-term survival ^[34][36].

Table 2. List of molecules or drugs tested as inhibitors of IDO1 and/or TDO2 in vitro, some of which are from the library of the National Institute of Cancer.

In term of mechanism, 1-D-MT inhibits p38 MAPK phosphorylation, preventing the increase of IDO1 mRNA and therefore the increase in KYN production. Additionally, it contributes to the inhibition of JNK signaling, which attenuates the expression of IDO1 mRNA and as a consequence the release of KYN ^[41]. 1-l-MT, the stereoisomer of 1-D-MT was recently reported to suppress the IFN-γ-induced expression of IDO1 in mouse rectal carcinoma cells ^[41]. The TRP-catabolizing enzyme IDO2 was not induced by 1-D-MT. This information sheds light on the different regulation of IDO1 and IDO2 at the transcriptional level. Uncertainty remains about the clinical significance of IDO1 expression in tumors. However, at some point, it is relevant to link the anti-tumor outcome of 1-D-MT to the induction of IDO1 by IFN-γ.

Methyl-thiohydantoin-tryptophan (MTH-TRP), another IDO inhibitor, has been discovered and is 20-fold more potent than 1-MT. Its sidechain is a mimetic of the amino acid backbone of tryptophan and it is more soluble in water than 1-MT but it is also more rapidly cleared from serum, both compounds being orally bioavailable ^[30].

In conclusion, IDO1’s expression is known to be immunosuppressive and may be involved in tumor immune escape, but it has also been implicated in direct anti-tumor effects. More studies are needed to better understand the role of IDO1 and IDO2, the implication of their inhibition and to determine whether a correlation exists between these enzymes and other important enzymes of the TRP metabolism such as TDO2.

2.3. Other TDO2 Inhibitors

Several studies have reported functional TDO2 expression in various human cancers including bladder, melanoma, and hepatocellular carcinoma ^[40]. In a preclinical mouse model, it has been shown that TDO2 expression by tumor cells prevented their rejection by immunized animals ^[39]. Interestingly, this type of immunosuppression prevents allograft rejection usually observed after liver transplantation [40]. TDO2 promotes tumor progression through the production of kynurenine, which is an endogenous ligand of the aryl hydrocarbon receptor (AHR), known to be involved in increased tumor cell survival and motility, and reduced anti-tumor immune responses. Altogether, these data suggest that pharmacological inhibition of TDO2 could reactivate the immune system and promote tumor destruction. So far, several compounds were reported as TDO2 inhibitors such as 680C91 ^[38], which has poor bioavailability and poor solubility, replaced by LM10, which has high bioavailability and solubility ^[42].

In another study, ~2800 compounds from the Library of National Cancer Institute USA were screened, among which seven were potent inhibitors of TDO2, with inhibition rates in the nanomolar or low micromolar ranges, and six of them inhibited both IDO1 and TDO2. All these inhibitors have antitumor characteristics on different cancer cell lines ^[40](Table 2). NSC 26326 or β-lapachone, which is a topoisomerase I inhibitor, was the strongest IDO1/TDO2 inhibitor, with Ki = 97 ± 14 nM for TDO2 and 30–70 nM for IDO1 ^[40].

This pharmacological approach may be further extended to a wider range of eligible tumors before entering into clinical studies.

In conclusion, according to some preclinical data pharmacological inhibition of TDO2 and IDO1 might represent a safe and efficient approach to treat cancer by promoting tumoral destruction by the immune system, and consequently potentiating cancer immunotherapies whether it is a single shot inhibition or a two-shot inhibition (IDO1 and TDO2).