Inflammatory pathways and cholangiocarcinoma risk mechanisms and prevention

Abstract

Cholangiocarcinoma (CCA), a neoplasm burdened by a poor prognosis and currently lacking adequate therapeutic treatments, can originate at different levels of the biliary tree, in the intrahepatic, hilar, or extrahepatic area. The main risk factors for the development of CCA are the presence of chronic cholangiopathies of various etiology. To date, the most studied prodromal diseases of CCA are primary sclerosing cholangitis, Caroli’s disease and fluke infestations, but other conditions, such as metabolic syndrome, nonalcoholic fatty liver disease and obesity, are emerging as associated with an increased risk of CCA development. In this review, we focused on the analysis of the pro-inflammatory mechanisms that induce the development of CCA and on the role of cells of the immune response in cholangiocarcinogenesis. In very recent times, these cellular mechanisms have been the subject of emerging studies aimed at verifying how the modulation of the inflammatory and immunological responses can have a therapeutic significance and how these can be used as therapeutic targets.

1. Introduction

Cholangiocarcinoma (CCA) is a highly aggressive neoplasia that can arise from the neoplastic transformation of cholangiocytes in any portion of the biliary system, from the finest intrahepatic ramifications to the main extrahepatic bile ducts. CCA is the most common biliary malignancy and the second most common primary liver cancer after hepatocellular carcinoma (HCC) (Banales et al., 2016, 2020), CCA remains an unmet need as less than 5% of patients survive up to 5 years from the diagnosis, making CCA responsible for 13% of overall cancer-related deaths (Welzel, McGlynn, Hsing, O’Brien, & Pfeiffer, 2006). A number of features contribute to the dismal prognosis of CCA: its insidious presentation due to late onset of the symptoms, lack of reliable clinical, laboratory or instrumental markers, and its early metastatic dissemination. Many patients are diagnosed with metastatic disease at presentation, significantly limiting the possibility of curative treatment.

CCA may arise from any tract of the biliary system and can be anatomically classified as intrahepatic (iCCA), perihilar (pCCA), or distal CCA (dCCA) (Rizvi & Gores, 2013). These site-restricted variants show significant clinical and genomic differences. The pCCA subtype is the most common type, and in the USA, it accounts for approximately 50–60% of all CCAs, followed by dCCA (20–30%) and iCCA (10–20%) (Banales et al., 2020). Globally, CCA has an incidence rate of 0.3–6/100,000 inhabitants per year, with a mortality rate of 1–6/100,000 inhabitants per year (Sarcognato et al., 2021), but its incidence has wide geographical variations, reflecting the heterogeneous distribution of environmental risk factors as well as the genetic predisposition of the different populations (Clements et al., 2020; Banales et al., 2020).

A number of risk factors for CCA have been identified and are listed in Table 1. There are significant difference in terms of carcinogenetic mechanisms according to the type of risk factors involved. CCA often arises in the setting of prolonged biliary inflammation, cholestasis and activation of reparative regenerative hepatic mechanisms, involving cell proliferation (Brindley et al., 2021; Leone, Ali, Weber, Tschaharganeh, & Heikenwalder, 2021). Recent transcriptomic analysis studies distinguished two iCCA profiles, enriched with pro-inflammatory (inflammatory class) and oncogenic pathways (proliferation class), respectively. The “inflammatory” subclass accounts for about 38% of the cases and is characterized by increased expression of immune-related signaling pathways, while the “proliferation” subclass, accounting for about 62% of iCCA cases, is enriched with oncogenic pathways and shows a worse prognosis (Sia et al., 2013). Similarly to iCCA, eCCA could also be differentiated in four subclasses, one of which is characterized by the upregulation of inflammatory mediators (Montal et al., 2020). There are marked differences in the genomic features of CCA, depending on the anatomical location and risk factors. Their discussion is beyond the scope of the current work and the reader is referred to excellent recent reviews (Banales et al., 2020; Cadamuro et al., 2021; Rodrigues et al., 2021; Simbolo et al., 2021).

Table 1.

Risk factors for CCA.

Risk factors	Refs.
Primary sclerosing cholangitis	Bjoro, Brandsaeter, Foss, and Schrumpf (2006), Lazaridis and LaRusso (2016), and Aune, Sen, Norat, Riboli, and Folseraas (2021)
Fibrocystic liver diseases: • Congenital hepatic fibrosis • Caroli disease • ARPKD • Choledocal cysts	Khan, Toledano, and Taylor-Robinson (2008), Strazzabosco and Fabris (2012), and Cannito et al. (2018)
Fluke infestation: • Platyhelminthes: Opistorchis viverrini and Opistorchis felineus • Thrematoda: Clonorchis sinensis	Brindley et al. (2021)
Hepatolithiasis	Kim et al. (2015)
Viral chronic hepatitis: • HBV • HCV	Zhou et al. (2012), Matsumoto et al. (2014), Wang, Sheng, Dong, and Qin (2016), and Seo et al. (2020)
Metabolic diseases: • Metabolic syndrome • Obesity • NAFLD/NASH	Welzel et al. (2011), Palmer and Patel (2012), Wongjarupong et al. (2017), and Petrick et al. (2017)
Diabetes mellitus	Tyson and El-Serag (2011) and Clements, Eliahoo, Kim, Taylor-Robinson, and Khan (2020)
Toxins (e.g., alcohol, thorotrast, dioxin)	McGee et al. (2019)
Genetic polymorphisms (e.g., ABCC2, MTHFR, KLRK1)	Tyson and El-Serag (2011) and Clements et al. (2020)

Here we will focus on the pathways that can be related to the inflammatory changes. The majority of CCA risk factors are associated with chronic inflammatory states and changes in signaling pathways involved in cholangiocyte biology, leading among others to uncontrolled proliferation, evasion of apoptosis, and loss of genome integrity that could eventually carry to cell senescence. Cellular senescence, often the result of chronic inflammation, has been little studied and it is not clear what its significance is in the pathogenesis of CCA even if it is hypothesized that it may be a nodal point in the development of this tumor (Sasaki & Nakanuma, 2016).

A distinguishing feature of CCA is the presence of a rich tumor-reactive stroma that it is believed to dictate the different steps of cholangiocarcinogenesis through the exchange of a complex network of signaling among tumor cells and a number of infiltrating cells, ultimately promoting cell proliferation, survival and genetic and/or epigenetic alterations. In iCCAs and pCCAs, the reactive desmoplastic stroma contains cancer-associated fibroblasts (CAFs) that exhibit an extensive crosstalk with CCA.

Bile acids are retained in cholestasis and may promote cholangiocarcinogenesis through the activation of epidermal growth factor receptor (EGFR), induction of cyclooxygenase-2 (COX-2), myeloid cell leukemia 1 (Mcl-1) and interleukin (IL)-6, and downregulation of farnesoid X receptor (FXR), which favors NF-kB-mediate inflammation (Sirica, 2011; Affo, Yu, & Schwabe, 2017; Cadamuro, Brivio, et al., 2018; Cadamuro, Stecca, et al., 2018). FXR expression was reported to be decreased in human CCA tumors, while the levels of TGR5, another bile acid receptor, were found to be increased in CCA tumors. In turn, IL-6–STAT3 may contribute to mitogenesis by upregulating Mcl-1 or altering EGFR promoter methylation (Wehbe, Henson, Meng, Mize-Berge, & Patel, 2006).

Given their relationships with biliary and liver repair, it worth mentioning that developmental pathways involved in biliary development (including Notch, WNT/β-catenin, Hedgehog, Hippo-Yes-associated protein, or YAP, and transforming growth factor β, or TGF-β), play an important role in cholangiocarcinogenesis. The Notch pathway, for example, is known to play a major role in biliary repair, tubulogenesis, fibrosis and maintenance of the stem cell niche and increased Notch activity has been associated with primary liver tumors. Overexpression or aberrant Notch receptor expression has been reported in all types of CCA (Aoki et al., 2016; Wu et al., 2014). Notch signaling also mediates hepatocyte-cholangiocyte transdifferentiation during carcinogenesis in iCCA (Sekiya & Suzuki, 2012; Wang et al., 2018). Furthermore, the WNT–β-catenin signaling pathway is also known to be activated in most CCAs, in part as an effect of the release of Wnt ligands by inflammatory macrophages infiltrating the stroma, but also as a consequence of mutations encoding key components of the canonical WNT–β-catenin signaling pathway (Perugorria et al., 2019). Relationships between β-catenin/YAP/inflammation will be discussed later (see fibropolycystic diseases). YAP is a transcriptional co-activator that could be also modulated by Hippo-independent signals, such changes in extracellular matrix (ECM) composition, stiffness and inflammation (Sugihara, Isomoto, Gores, & Smoot, 2019; Affo et al., 2021; see also Chapter 10). YAP nuclear expression is increased in CCA, and YAP has been shown in vitro to be activated in CCA cell lines by IL-6, platelet-derived growth factors (PDGF) and fibroblast growth factor (Rizvi et al., 2016; Smoot et al., 2018; Sugiura et al., 2019).

These pathways are complex and cross talk with each other in diverse ways, depending on specific predisposing conditions and risk factors. To focus more specifically on this concept, we will first describe some known conditions having a strong association with CCA and then discussed the respectively involved inflammatory pathways in more detail.

2. Diseases prodromal to cholangiocarcinoma development

2.1. Primary sclerosing cholangitis

Primary sclerosing cholangitis (PSC) is a rare chronic liver disease in which persistent peribiliary inflammation and progressive fibrosis lead to strictures of the intrahepatic or/and extrahepatic bile ducts. The etiology of PSC is still unknown, but it is well established that inflammatory damage to the biliary epithelium play a causative role (Dyson, Beuers, Jones, Lohse, & Hudson, 2018). The pathogenesis of PSC is that still unidentified, but it is probably the end-results of the interaction between a number of factors acquired and genetic determinants that trigger an inflammatory response in a predisposed individual. Inflammation damages cholangiocytes, which then elicits the onset of persistent inflammatory status sustained by an immune cell infiltrate and fibrotic tissue deposition around the bile ducts, resulting in the typical peribiliary onion-like lesion characteristic of PSC (Lewis, 2017). At the histological level, PSC is further characterized by the presence reactive ductular cholangiocytes, activated fibroblasts, macrophages and neutrophils, and by the deposition of extracellular matrix components, mainly collagen I and fibronectin (Matsumoto et al., 1999). The increasing deposition of fibrotic tissue not only causes strictures in the biliary tree, but also marks disease progression towards biliary cirrhosis and portal hypertension (Dyson et al., 2018).

Evidences of the role of inflammation/immunity in the pathogenesis of PSC is provided by a strong association with the Human Leukocyte Antigen (HLA) complex on chromosome 6, along with the evidence that in 25% of cases PSC is associated with other non-gastrointestinal autoimmune diseases (Karlsen, Folseraas, Thorburn, & Vesterhus, 2017). A genome-wide association study (GWAS) of large cohorts of patients has shown a strong association with HLA (Karlsen et al., 2010; Liu et al., 2013), but more than 20 other genes (e.g., genes of the IL-2 pathway, CARD9, REL, FUT2 and MST1) have been linked with PSC (Liu et al., 2013; Mells, Kaser, & Karlsen, 2013; Melum et al., 2011). However, to date, the combined genetic risk only accounts for less than 10% of the susceptibility to PSC (Karlsen et al., 2017), and the autoimmune nature of the inflammation in PSC remains unproven.

The gut-liver axis is also a factor in the pathogenesis of PSC, not only because of the known association with inflammatory bowel diseases. The innate immune response could be activated by the leakage of proinflammatory bacterial products (e.g., lipopolysaccharides) from the intestinal microbiota (Özdirik, Müller, Wree, Tacke, & Sigal, 2021). Furthermore, gut-derived antigens can be presented to the T-cell Receptor (TCR), thus activating T cells and inducing their migration to liver and gut because of the overlapping adhesion molecule profiles of the endothelium in these two organs (e.g., mucosal vascular address in cell adhesion molecule 1 (MadCAM-1), vascular cell adhesion molecule 1 (VCAM-1) and Chemokine C-C motif ligand 25 (CCL25)) (Trivedi & Adams, 2013). The role of Thy17 cells and IL17 is still under investigation, but it could be linked to the translocation of certain bacterial species (Kunzmann et al., 2020).

The composition of gut microbiota in PSC patients is usually altered, with an overall reduction in bacterial diversity and an increase of specific bacterial phyla when compared with a healthy patient. Studies regarding other diseases with altered gut microbiota suggest that the reduced bacterial diversity occurs prior to and independent from clinical disease manifestations, but its role remains unclear (Karlsen, 2016). Recent studies (Bajer et al., 2017; Pontecorvi, Carbone, & Invernizzi, 2016) support the involvement of the gut microbiota in the initiating events of PSC pathogenesis. It is interesting to note that changes in gut microbiota have been linked also to the development of HCC in other liver conditions (Schwabe & Greten, 2020; Zheng, Ran, Zhang, Wang, & Zhou, 2021).

As the disease progresses, chronic injury converges on the common mechanisms of fibrosis, involving hepatic stellate cells, portal myofibroblasts and cholangiocytes accompanied by several other cells involved in immune response (Mederacke et al., 2013). These cells take part in a cross talk that is still largely to be dissected, but cholangiocytes have a key role in driving mesenchymal cells and fibrosis (like in the “onion skin” histological lesions) (Karlsen et al., 2017). As the disease progresses there is an increased risk of developing dominant strictures and of recurrent episodes of bacterial cholangitis. Notably, dominant strictures could be the presentation of CCA (Williamson & Chapman, 2015). Patients with PSC are at increased risk to develop neoplasias, such as colon, and pancreatic cancers, but hepatobiliary malignancies are the most common cause of cancer-related mortality and the most dreadful complication is CCA (Aune et al., 2021; Bjøro et al., 2006; Lazaridis & LaRusso, 2016).

It has been reported that CCA that develops in PSC is driven by KRAS and TP53 mutations in association with FGFR2 fusion and IDH 1 and 2 mutations, similar to other iCCAs (Banales et al., 2020). Deregulation of micro RNAs (miRNAs) involved in cell proliferation, stemness, and migration, may also play a relevant role; for example, IL-6 reduces the expression of the tumor suppressor NF2, by upregulating miR-let-7a, along with several other miRNAs targeting different transcription factors (Meng et al., 2007).

The persistence of the inflammatory insult in PSC is associated with the secretion of pro-inflammatory mediators, in particular of IL-6, interferon (IFN)γ, and tumor necrosis factor (TNF)α (Pinto, Giordano, Maroni, & Marzioni, 2018), which upregulate the expression in the biliary structures of inducible nitric oxide synthase (iNOS) (Spirlìet al., 2003). This phenomenon leads to the increase in local concentration of nitric oxide (NO), which in turn stimulates infiltrating inflammatory cells to oversecrete various cyto- and chemokines, PDGF, and vascular endothelial growth factor (VEGF) (Cannito et al., 2018; Sirica, 2005). The local accumulation of high concentration of ${\text{NO}}_2^{-}$ promotes the formation of reactive oxidative species (ROS), such as peroxynitrites (ONOO⁻) that, being responsible for the peroxidation of lipids, play a major role in the neoplastic transformation, and in nitrosylation and inactivation of several proteins, including enzymes involved in DNA proofreading and repair of single or double-strand breaks in DNA, such as 8-oxo-deoxyguanine DNA glycosylase 1. Furthermore, NO and ROS could generate two mutagenic compounds, 8-oxo-7,8-dihydro-20-deoxyguanosine and 8-nitroguanine (8-NG) (Jaiswal, LaRusso, Shapiro, Billiar, & Gores, 2001). This sequence of events is particularly relevant, as increased iNOS expression and nitrosylation of protein has been also described in bile ducts of patients with early stages of PSC.

2.2. Congenital hepatic fibrosis/Caroli’s disease

Congenital hepatic fibrosis (CHF) and Caroli’s disease (CD) are two cholangiopathies belonging to the family of the fibropolycystic liver diseases, a group of genetic diseases caused by mutations to the polycystic kidney and hepatic disease 1 (PKHD1) gene, which encodes for the protein fibrocystin (FPC) (Harris & Torres, 2009; Zerres, Rudnik-Schöneborn, Steinkamm, Becker, & Mücher, 1998). These conditions are all characterized by saccular cyst-like biliary structures in continuity with the biliary tree, accompanied by robust fibro-inflammatory infiltrate preeminently with peribiliary localization, and the development of liver fibrosis (Lasagni, Cadamuro, Morana, Fabris, & Strazzabosco, 2021). Liver disease progresses overtime and 10–40% of the patients presents esophageal varices bleeding or need portosystemic shunting. In adulthood, patients with Caroli’s disease suffer from recurrent cholangitis and in 80% of the cases from portal hypertension (PH). No pharmacological therapies are available to relief or block the worsening of this fibroinflammatory liver disease, that could also be causative for CCA in 7–14% of patients with Caroli’s disease. (Cannito et al., 2018; Khan et al., 2008; Strazzabosco & Fabris, 2012).

FPC is a transmembrane protein expressed in the primary cilium and in centromeres of epithelial cells of the bile, pancreatic and renal tubules. The role of fibrocystin is not completely known, but appears to be crucial in the differentiation of the ductal epithelium and, if mutated, it induces a maturational arrest of the nascent ductal plate (Veigel et al., 2009). The hepatic phenotype is characterized by a wide spectrum of dysgenetic alterations of the biliary epithelium, including the formation of biliary microhamartomas, which can give rise to segmental dilatations of the bile ducts, both intra and extrahepatic, rather than real cysts. Biliary dysgenesis is associated with an exuberant production of peribiliary fibrosis accompanied by an inflammatory infiltrate, mainly composed by macrophages. This is clinically significant as it is accompanied by the development of PH, and by a non-negligible risk of malignant transformation to CCA (Locatelli et al., 2016; Pech et al., 2016). The mechanisms regulating the development of fibroinflammation in these diseases remain largely unknown.

Biliary cystogenesis in CD/CHF is mediated by altered intracellular signaling, including cAMP and Ca²⁺ (Banales et al., 2009; Rohatgi et al., 2008). The interconnection between these intracellular signals leads to the activation of various pathways involved in the proliferation of biliary structures such as the phosphatidylinositol 3-kinases (PI3K)/AKT/mammalian target of rapamycin (mTOR) or protein kinase A (PKA)/RAF/extracellular-signal-regulated kinase (ERK1/2) pathways (Fischer et al., 2009). The activation of these signaling pathways leads to the transcription of the cell division cycle 25 homolog A (Cdc25A) gene, which is essential in mediating biliary cystogenesis (Masyuk et al., 2012). It should be noted that Cdc25A is also modulated by epigenetic anomalies; in fact PCK rats, a model of CD, show the downregulation of miR-15A, whose experimental overexpression induces the repression of Cdc25A and consequently the reduction of the cystic area (Lee et al., 2008). Another epigenetic modulator involved in the pathogenesis of CD/CHF is histone deacetylase (HDAC) 6, an enzyme that plays a key role in cell cycle regulation, and whose overexpression has been demonstrated in cysts in both PCK rats and human patients (Gradilone et al., 2014). The inhibition of its function by ACY-1215, has been shown to induce a significant reduction in the growth of the cysts (Gradilone et al., 2014). The enzymes belonging to the HDAC family are known to modulate the Hippo pathway (Basu, Reyes-Múgica, & Rebbaa, 2013). This is a highly conserved signaling pathway, which orchestrate organ size, cell proliferation, stem cell stability and several other biological functions (Piccolo, Dupont, & Cordenonsi, 2014). Recent studies have shown that biliary cysts of PCK rat and of Pkhd1^{LSL(−)/LSL(−)} mice, both rodent model orthologs to the human disease, display and increased nuclear expression YAP, the main transcription factor of the Hippo pathway, and that this evidence correlate with increased biliary expression of proliferating cell nuclear antigen (PCNA) and of cyclin D1 (Jiang et al., 2017).

One of the most characteristic and concerning traits of CD/CHF is the presence of progressive fibro-inflammation. Recent studies point to the accumulation of β-catenin in the nucleus as a key step in the pathogenesis of CHF/CD. Beta-catenin is a structural protein that is part of the so-called WNT/β-catenin pathway and which is usually retained in the cytoplasm in a non-phosphorylated form linked to the so-called “β-catenin destruction complex.” When expressed into the nucleus, β-catenin binds to the T cell factor/lymphoid enhancer factor (TCF/LEF) complex (Nusse & Clevers, 2017) and acts as a transcriptional factor. In FPC-deficient cholangiocytes, β-catenin seems to favor the acquisition of a profibrotic and proinflammatory phenotype and promote the secretion of cytokines such as C-X-C Motif Chemokine Ligand (CXCL) 1, CXCL10, CXCL12. The secretion of these mediators into the pericystic space elicits the recruitment of inflammatory cells into the portal space, and, in particular of M1 and M2 macrophages. Macrophages, by secreting TNFα and TGF-β, in turn, induce the expression of integrin αVβ6, an activator of latent TGF-β, by cholangiocytes, driving fibrosis mediated by myofibroblasts (Locatelli et al., 2016). Furthermore, a recent paper (Tsunoda et al., 2019) demonstrated, using induced pluripotent stem Cell (iPSC) technology, that FPC-deficiency induces the secretion of IL-8 by cholangiocytes, which in turn, through an autocrine loop, stimulates the secretion of connective tissue growth factor (CTGF), a fundamental profibrotic mediator, providing further mechanistic insight into the pathogenesis of this family of diseases. How this intense inflammatory signaling contributes to biliary carcinogenesis is the focus of ongoing investigations.

2.3. Fluke infestations

A major cause of biliary neoplastic transformation is the infestation of the biliary tract with flukes. Liver fluke infestations, in particular of Platyhelminthes, such as Opistorchis viverrini, Opistorchis felineus, and of Thrematoda, such as Clonorchis sinensis, are endemic in several Asian countries and Eastern Europe (Brindley et al., 2021). O. viverrini, is endemic in Thailand, Cambodia and Vietnam and infects from 8 to 10 million people (Sripa, Kaewkes, Intapan, Maleewong, & Brindley, 2010), while O. felineus is mainly diffuse in Eastern Europe countries, in particular in Siberia, Ukraine, Belarus, and Kazakhstan (Fedorova et al., 2020). Finally, C. sinensis, is diffuse in the rural areas of Korea, Vietnam and China where it is responsible for the infestation of more than 35 million people (Keiser & Utzinger, 2005). The infestation of O. viverrini is due to the consumption of undercooked or raw fish, in particular of different species of cyprinids, a fish largely used in all South Asian regions. The first host of O. viverrini larvae is represented by freshwater snails, which are eaten by the fish. The metacercariae of the parasites ingested by man through the raw fish proliferate inside the bile ducts of the hosts releasing eggs. Eggs are then dispersed in the water through the stools and develop in the water perpetuating the cycle of infection (Sripa et al., 2010). This infection mechanism is also common to other flukes that infest, in particular, rural populations that do not have healthy drinking water available. Once encysted inside the bile ducts, the parasites develop and induce a strong fibroinflammatory reaction that triggers the development of a chronic cholangiopathy characterized by biliary proliferation, inflammation of the liver, and deposition of peribiliary fibrosis induced by an overt immune response (Qian, Utzinger, Keiser, & Zhou, 2016).

O. viverrini secretes a very large pool of growth factors, cytokine mediators, and other molecules (collectively named O. viverrini excretory/secretory products (OvESP)), that promote a chronic cholangiopathy (Young et al., 2010). Among OvESP, O. viverrini glutathione S-transferase (OvGST) stimulates cholangiocyte proliferation though AKT and ERK mediated pathways (Daorueang et al., 2012). Furthermore, O. viverrini granulin (Ov-GRN-1) and thioredoxin (Ov-Trx-1) can stimulate the proliferation of human cholangiocytes (Smout et al., 2009; Smout, Mulvenna, Jones, & Loukas, 2011; Suttiprapa et al., 2012). The response of biliary tree to the O. viverrini infestation is similar to the biliary response to pathogen-associated molecular pattern. In fact, OvESP act by activating the signaling pathway mediated by Toll-like receptors (TLR) 4, which results in nuclearization and consequent MyD88-dependent activation of the transcriptional action of NF-kB (Ninlawan et al., 2010). This signaling cascade leads to the secretion of various mediators, among which the most studied are IL-6 and IL-8, which are involved in stimulating the biliary ductular reaction typical of this liver disease and in the recruitment of neutrophils, respectively (Ninlawan et al., 2010). The TLR4 and TLRT2 pathways are also activated by the infestation of C. sinensis. This escalates into the generation of other mediators, including TNFα, IFNγ and several interleukins (1β, 4, 6, 10, and 13), further supporting the secretion of a chronic fibroinflammatory cholangiopathy (Kim et al., 2017; Prueksapanich et al., 2018). During the earliest stages of disease, there is a preponderance of infiltrating M1 macrophages with consequent release of M1-related factors, such as iNOS, TNFα, and CXCL9, while in the later phases there is a shift towards a population of profibrotic and pro-tumorigenic, arginase-1-positive M2 macrophages characterized by hypersecretion of CCL2 (Kim et al., 2017). For these reasons, both O. viverrini and C. sinensis have been included among the group 1 agents/biological carcinogens by the International Agency for Research on Cancer (IARC) (Brindley et al., 2021).

2.4. Other risk factors

As shown in table 1, several other risk factors related to the development of CCA have been identified. Among them, infection with hepatitis B and C viruses, hepatolithiasis, alcohol consumption, metabolic syndrome, obesity and non-alcoholic fatty liver disease (NAFLD) and non-alcoholic steatohepatitis (NASH). These are all risk factors for chronic liver disease. To date, the mechanisms responsible for the transition from chronic liver damage to tumor formation are unknown (Clements et al., 2020; Tyson & El-Serag, 2011), but in most cases, inflammatory pathways play a central role.

Hepatolithiasis is the presence of intrahepatic biliary stones, and is a well-known risk factor for CCA development in Far Eastern countries (Kim et al., 2015). Notably, hepatolithiasis has been reported to be a prodromal factor for CCA in the 1.6–9.9% of cases (Kim et al., 2015). The presence of stones generates an inflammatory response that induces the dedifferentiation of the biliary epithelium with possible transition to an intraductal papillary neoplasm of the bile duct (IPNB) and biliary intraepithelial neoplasia (BilIN), both of which are considered to be preneoplastic lesions (Aishima, Kubo, Tanaka, & Oda, 2014). Chronic inflammation induces the overexpression of COX-2, which positively modulates prostaglandin E2 that, similar to the increase in NF-kB transcription activity, transforms chronic hepatolithiasis into a chronic proliferating cholangitis (Shoda et al., 2003). The accumulation of gene mutations typically linked to conditions of neoplasia, such as EGFR, p16, deleted in pancreatic cancer 4 (DPC4)/signaling effectors mothers against decapentaplegic protein 4 (Smad4), c-Met, and c-ERBb2 ultimately lead to the development of CCA (Kim et al., 2015).

Hepatotropic virus infection, like HBV and HCV, are also a known cause of liver cancer. While there have been numerous reports on how these viruses induce neoplastic transformation of hepatocytes, favoring the development of HCC, only more recently have HBV and HCV been identified as independent risk factors for the development of iCCA (Matsumoto et al., 2014; Seo et al., 2020; Wang et al., 2016; Zhou et al., 2012). The mechanisms hypothesized for the viral related development of CA are similar to those of the HCC. They rely on the formation of chronic inflammation that induces fibrosis with development of cirrhosis, and subsequently to the accumulation of genetic modifications within cholangiocytes that underlie the sequential appearance of precancerous dysplastic foci, and ultimately to CCA. See at the Introduction concerning the role of Notch in driving the transdifferentiation of tumor cells from hepatocyte-like into cholangiocyte-like in iCCA.

Several studies have shown a correlation between alcohol consumption and smoking with the development of CCA (HR = 2.35 and HR = 2.15, respectively) (McGee et al., 2019). Furthermore, the association between diabetes, obesity and metabolic syndrome and cholangiocarcinogenesis is becoming clearer (Palmer & Patel, 2012; Vale, Gouveia, Gärtner, and Brindley, 2020). NAFLD is currently the most common liver disease and its incidence and prevalence is expected to reach epidemic proportions with a prevalence of 25% of the general population (Bedogni et al., 2005). NASH, the inflammatory variant of NAFLD, is a progressive disease in which a sizable proportion of patients develop liver fibrosis, progressing to cirrhosis and primary liver cancer. While most research efforts have been dedicated to study how hepatocellular carcinoma develops from NAFLD/NASH (Michelotti et al., 2021), not much is known about the mechanisms of CCA development in NASH. Recent epidemiological studies have shown that NAFLD is associated with a 3-time increased risk to develop iCCA (Petrick et al., 2017; Wongjarupong et al., 2017). Obesity, a metabolic disorder commonly associated with NAFLD/NASH, is significantly associated with CCA development (OR 1.52, 95 % CI 1.13–1.89) (Li et al., 2014); up to 20% of patients with resectable iCCA having been diagnosed with NASH (Reddy et al., 2013).

3. Mechanisms of neoplastic transformation

Chronic injury induced by inflammation of the biliary epithelia is a premalignant condition leading to CCA. Chronic cholangiopathies share similar carcinogenetic processes. Major mechanisms that support and facilitate the neoplastic transformation in CCA are the proliferation of cholangiocytes and their concomitant acquisition of resistance to apoptosis, together with the accumulation of oncogenic DNA mutations. These mechanisms are closely linked and are self-sustaining. “Reactive” cholangiocytes produce inflammatory mediators that act in an autocrine manner on the cholangiocytes themselves and in a paracrine way on cells of the tumor microenvironment (TME) (Fig. 1) (Banales et al., 2019; Cannito et al., 2018; Fabris, Sato, Alpini, & Strazzabosco, 2021; Sirica et al., 2019).

Fig. 1.

Mediators released by biliary epithelial cells involved in cholangiocancerogenesis. Cholangiocytes respond to signals that mediate responses involved in CCA tumorigenesis, such as proliferative responses, cell survival, chemotactic messages or others involved in the depression of proofreading enzyme activity. Among these effectors, the best known and studied is IL-6 which stimulates biliary proliferation through the ERK1/2 pathway, cellular senescence mediated by p21 overexpression, via p38, and secretion of Mcl1 (red arrow), through the PI3K/Akt pathway. Furthermore, by inhibiting the expression of miR-148a, miR-152, and miR-301, it stimulates, via DNMT1, the expression of the protooncogenes Rassf1a and p16^INK4a. IL-1β is involved in the secretion of CXCL5 (red arrow), a potent chemotactic agent for neutrophils. Another mediator of fundamental importance is TGF-β, which, through the nuclearization of the SMAD2/3/4 complex, fuels various responses involved in tumor biology, such as proliferation by stimulating cyclin D1. TNFα is involved in tumorigenesis because it stimulates the secretion of Bcl2 and the expression of S100A4, a protein involved in the metastasis of CCA. Finally, the regulation of COX-2, an enzyme involved in cell proliferation and in protection against apoptosis, is affected by two pathways, on the one hand is stimulated by the hyperactivation of EGFR, on the other hand by the increase of local NO concentration due to the de novo expression of iNOS by extracellular proinflammatory signals. NO can also repress the activity of 8-oxoDG, thus favoring the accumulation of mutations in the DNA. Legend: P, phosphorylation; CCA, cholangiocarcinoma; ERK, extracellular-signal-regulated kinase; IL, interleukin; Mcl1, myeloid cell leukemia 1; PI3K, phosphatidylinositol 3-kinases; miR, microRNA; SMAD, signaling effectors mothers against decapentaplegic protein; TGF-β, transforming growth factor β; TNF, tumor necrosis factor; COX-2, cyclooxygenase-2; EGFR, epidermal growth factor receptor; iNOS, inducible nitric oxide synthase; NO, nitric oxide; 8-oxoDG, 8-Oxo-2′-deoxyguanosine. Black lines, intracellular pathways; red lines, secretion of soluble proteins.

Cholangiocyte-derived secretion of IL-6 and consequent activation of IL-6/STAT3 signaling stimulates the growth of malignant cells by activating the p44/42 mitogen-activated protein kinases (MAPK) pathways, leading to cell proliferation (Park, Tadlock, Gores, & Patel, 1999), and, at the same time, stimulates the p38 MAPK pathway, which decreases the production of p21^WAF1/CIP1, a cyclin dependent inhibitor controlling cell cycle (Tadlock & Patel, 2001). Furthermore, the secretion of IL-6 through the stimulation of STAT3 expression and Akt activation induces the overexpression of Mcl-1, a fundamental anti-apoptotic mediator (Isomoto et al., 2005; Kobayashi, Werneburg, Bronk, Kaufmann, & Gores, 2005; Yu et al., 2005). The role of IL-6 in cholangiocarcinogenesis is also related to its ability to induce epigenetic modifications. In fact, IL-6 is able to modulate miRNAs (Braconi, Huang, & Patel, 2010) and to demethylate EGFR promoters, thus favoring their transcription and consequently their proliferative function (Wehbe et al., 2006). IL-6 inhibit the expression of miR-148a, miR-152 and miR-301. DNA methyltransferase-1 (DNMT-1) is a specific target of 148a and miR-152, and due to their absence, increases its activity, thus inducing the methylation and the consequent repression of two tumor suppressors, Rassf1a and p16^INK4a, both of which are involved in repressing tumoral transformation (Braconi et al., 2010).

The demethylation of the EGFR promoters hyperactivate the downstream signal pathways of EGFR and induces the expression of COX-2 (Endo, Yoon, Pairojkul, Demetris, & Sirica, 2002; Yoon, Canbay, Werneburg, Lee, & Gores, 2004; Yoon, Gwak, et al., 2004; Yoon, Higuchi, Werneburg, Kaufmann, & Gores, 2002) that promotes cell proliferation and angiogenesis, as well as and inhibits apoptosis (Yoon et al., 2002). In CCA, other mechanisms could modulate COX-2, such as the presence of oxysterols, which can both stabilize COX-2 expression and activate the hedgehog signaling pathway linked to CCA development (El Khatib et al., 2013; Fingas et al., 2011; Nachtergaele et al., 2012; Yoon, Canbay, et al., 2004; Yoon, Gwak, et al., 2004). Another inducer of COX-2 is the excessive local concentration of NO, as a result of the activation of inducible iNOS. As discussed earlier, the excess of NO induces nitrotyrosine stress and inhibits 8-oxo-2′-deoxyguanosine (8-oxoDG) base excision in DNA repair processes, causing the accumulation of oxidative DNA lesions and promoting carcinogenesis (Jaiswal et al., 2001; Jaiswal, LaRusso, Burgart, & Gores, 2000; Spirlì et al., 2003). Fluke infestations exploit similar mechanisms to induce neoplastic transformation of cholangiocytes. The accumulation of metabolites and factors released by the parasites, such as oxysterol or catechol estrogen quinone, stimulate the iNOS-mediated production of reactive oxygen species (ROS) and reactive nitrogen species (RNS). Both compounds not only induce DNA damage, but are also capable of interfering with the activity of DNA proofreading enzymes, such as 8-oxoDG that leads to the accumulation of 8-NG, a product of RNS-induced damage of DNA (Correia da Costa et al., 2014; Ohshima, Sawa, & Akaike, 2006; Vale et al., 2020). Overexpression of iNOS is an indicator of poor prognosis in CCA patients (Suksawat et al., 2017) and its in vitro inhibition with N^ω-nitro-l-arginine methyl ester hydrochloride leads to a reduced migratory and invasive ability of CCA cell lines (Suksawat et al., 2018).

Another critically important inflammatory mediator released into the TME is IL-1β. This interleukin plays a dual role by stimulating CXCL5 expression on CCA cells (Okabe et al., 2012). The expression of the latter chemokine is in fact able to stimulate the chemotaxis of neutrophils in the tumor area through the PI3K-Akt signaling pathway and to activate cell proliferation via p44/p42 MAPK (Zhou et al., 2014).

TGF-β is a cytokine/growth factor that is also of primary importance in CCA development. This multifunctional cytokine, primarily known as a mediator of organ fibrosis, is heavily involved in anti- and pro-carcinogenetic functions. Paradoxically, in the initial stages of the tumor it acts as a tumor suppressor, whereas in the advanced stages it stimulates various biological functions that support tumor growth and metastasis (Maemura, Natsugoe, & Takao, 2014; Morrison, Parvani, & Schiemann, 2013). In particular, in the most advanced phases of CCA, mutations found in the TFGRβ receptor, upregulate the expression of cyclin D1 via SMAD4 phosphorylation (Zen et al., 2005). A recent work, supporting the importance of TGF-β in cholangiocarcinogenesis has shown that pharmacological inhibition of the TGF-β signaling pathway is able to decrease both the clonogenic potential of tumor cells and the metastasis of CCA (Puthdee et al., 2021).

Finally, another fundamental piece of the puzzle composed by cytokines secreted into the TME is TNFα. TNFα can support the key neoplastic characteristics, not only by stimulating cell survival by activating the Akt/NF-kB/Bcl2 pathway (Luo, Maeda, Hsu, Yagita, & Karin, 2004), but also by increasing metastatization via modulation of S100A4 (Techasen et al., 2014). Moreover, TNFα stimulates CD4 lymphocytes to secrete trophic interleukins such as IL-17 and IL-23 (He et al., 2012; Tang et al., 2013).

4. Role of inflammatory cells in modulating CCA malignant features

As discussed previously, the TME is particularly important for supporting and sustaining the different biological functions of CCA, such as proliferation, metastasis and chemoresistance. The TME is composed of the ECM and of the infiltrating cell types. With regards to the non-cellular component of the TME, desmoplastic tumors, such as CCA, pancreatic adenocarcinoma, and breast cancer are characterized by a marked modification of the normal matrix, which becomes dense and stiff due to the increased deposition of collagen I, III and IV, fibronectin, nidogen, perlecan, and other supporting proteins (Brivio, Cadamuro, Strazzabosco, & Fabris, 2017; Cadamuro, Brivio, et al., 2018; Cadamuro, Stecca, et al., 2018). The increased stiffness provides a mechanical stimulus to the tumor cells, leading to the activation of signaling pathways mediated by mechanoreceptors, such as the Hippo pathway, resulting in a more aggressive tumor phenotype characterized by a marked cell proliferation and increases in chemoresistance and metastasis (Marti et al., 2015; Zanconato, Cordenonsi, & Piccolo, 2019; Zhang et al., 2018).

The TME is infiltrated by a “multiethnic” collection of cells and includes vascular structures (blood and lymphatic), CAFs, neural-derived cells, and cells belonging to innate (i.e., macrophages, granulocytes, and adaptive (T cells) immune responses. In this review, we focus on the cells of the immune response. For the pathogenetic mechanisms mediated in the CCA by other cell types, we refer the reader to other reviews (Banales et al., 2016, 2020; Sirica, Strazzabosco, & Cadamuro, 2021; Vaquero, Aoudjehane, & Fouassier, 2020).

Several recent works of functional genomics confirm the importance of the inflammatory infiltrate. These studies categorized iCCA (Job et al., 2020; Sia et al., 2013; Wang et al., 2022), eCCA (Montal et al., 2020) and mixed HCC-CCA (Nguyen et al., 2021) based on their genomic and transcriptomic fingerprint, identifying initially two classes in iCCA (proliferation and inflammatory) (Sia et al., 2013) and then four subclasses based on the characteristics of TME (immune desert, immunogenic, myeloid and mesenchymal) (Job et al., 2020), four for eCCA (metabolic, proliferation, mesenchymal, and immune) and two classes, further divided into two subclasses, for the mixed HCC-CCA (immune high, IH, and immune low, IL). In iCCA, immunogenic classes are characterized by accumulation of CD4+CD8+ lymphocytes, CD8RO+ T memory cells, CD20+ B cells, and reduced presence of CD68+ macrophages, as well as express significantly higher levels of IL1β, 6 and 15, C-X-C Chemokine Receptor (CXCR)4, CXCL9 and 13, and IFNγ, with the desert immune class showing the worst prognosis (Job et al., 2020). Similarly to iCCA, in eCCA, the immune class is the one with the better survival outcome, being characterized by an higher representation of CD8+ T cells and CD20+ B cells (Montal et al., 2020). Further confirming the importance of activation of the adaptive response in repressing and containing CCA aggressiveness, IH mixed HCC-CCA, characterized by increased tumoral expression of T lymphocytes and overexpression of chemotactic mediators, such as IL1β, 2, 4, 10, and 33, CCL9, CXCL19, CXCR3, IFNγ, also showed a better survival outcomes as compared to IL neoplasm (Nguyen et al., 2021).

It is interesting to note that these studies collectively demonstrate that CCAs with a high representation of the immune component have a better prognosis. It is worth noting that immune response cells have a dual role in cancer biology. Among cells that tend to contain tumorigenesis are Neutral Killer Cells (NKs), dendritic cells (DCs), and tumor-infiltrating lymphocytes (TILs). Macrophages and neutrophils, due to their plasticity and their ability to change phenotype and biological characteristics, can be either pro- or anti- tumorigenic, while Myeloid-Derived Suppressor Cells (MDSCs) act as tumor-stimulating factors (Fig. 2). For these reasons, a line of research in liver oncology is oriented towards modulating the immune response and the signaling pathways involved in the recruitment of the inflammatory infiltrate. Unfortunately, in CCA these studies are still in their infancy.

Fig. 2.

Involvement of inflammatory cells in CCA progression and malignant behavior. The cells of the immune response are differently involved in tumor progression of the CCA. Dendritic cells (DC), Natural Killer cells (NK) and tumor-associated lymphocytes (TIL), have marked anticancer properties. Conversely, myeloid-derived cells (MDSC), thanks to the ability to inhibit the activation of NK and DC, stimulate tumor malignancy. Finally, cancer-associated macrophages (TAMs) and cancer-associated neutrophils (TANs) can exhibit anti- or pro-tumor behaviors depending on their phenotype.

As previously mentioned, the cell types infiltrating the TME and capable of containing tumor aggressiveness include NKs, DCs, and TILs. Briefly, NKs are CD3-CD56+ cells capable of eliminating tumor cells via the secretion of lytic enzymes such as perforins, granzyme and proteases, or through an indirect pathway by stimulating the activation on target cells of the proapoptotic pathway mediated by Fas Cell Surface Death Receptor Ligand/TNF-related apoptosis-inducing ligand (FasL/TRAIL) (Banales et al., 2020; Cardoso Alves, Corazza, Micheau, & Krebs, 2021). DCs, on the other hand, acting as professional Antigen Presenting Cells (APCs), are able to activate the TILs in order to stimulate their ability to attack and kill cancer cells (Junking, Grainok, Thepmalee, Wongkham, & Yenchitsomanus, 2017).

Also cells of the adaptive response, i.e. tumor-infiltrating lymphocytes (TILs), either T or B cells, can be endowed with anti-tumor function (Fabris et al., 2019; Kitano et al., 2018). Their interaction within the microenvironment, may oppose tumor growth. On the other hand, MDSCs are myeloid-derived cells with a strong immunosuppressive activity, capable of inhibiting the action of TILs and DCs, thus generating a TME favorable to tumor growth (Bergenfelz & Leandersson, 2020; Job et al., 2020).

As alluded to above, macrophages and neutrophils can act in a dual way. In fact, both can have an inflammatory phenotype (M1 for macrophages and N1 for neutrophils), characterized by anti-tumor properties. Within the TME, however, there are M2 type macrophages or tumor-associated macrophages (TAMs) and N2 neutrophils, or tumor-associated neutrophils (TAN). TAMs are mainly present in the tumor front of the CCA (Raggi et al., 2017; Zhou et al., 2021) and are able to modify the matrix, and stimulate tumor growth thanks to the secretion of different cytokines (such as IL-6, IL-13, IL-34, and TNFα), promote neo-angiogenesis, inhibit the recruitment and activation of T cells, and induce apoptosis of M1-type macrophages (Ge & Ding, 2020). TANs can also secrete a plethora of cyto-chemokines and growth factors, such as VEGF, CXCL1, CXCL2, CXCL6 and CXCL8, having a trophic effect on CCA cells, as well as enzymes capable of modifying the extracellular matrix, such as metalloproteases (MMP) 8 and MMP9 (Fabris et al., 2019; Masucci, Minopoli, & Carriero, 2019; Rimassa, Personeni, Aghemo, & Lleo, 2019).

These data allow a better understanding of the complex and multifaceted biology of the CCA, but also indicate that, at least for some categories of CCA, it is necessary to study more in depth the inflammation that accompanies the tumor and how this can be targeted therapeutically. Currently, the biology and pathogenetic significance of TAMs in CCA represent an important chapter of study because their presence is a hallmark of poorer patient survival outcomes (Sun et al., 2020), probably due to their ability at stimulating cell proliferation, neoangiogenesis, and metastasis through the secretion of various mediators, including VEGF (Zhou, Wang, Lu, et al., 2021). In addition to participating in mechanisms more strictly linked to tumor aggressiveness, VEGF is also able to favor tumor malignancies by stimulating TAM proliferation. TAMs, in fact, have the ability to impede tumor antigen presentation, diminishing the T cell-mediated immune response against the tumor (Rothlin, Ghosh, Zuniga, Oldstone, & Lemke, 2007). Moreover, several studies have shown that the presence and relative quantity of CD4+, CD8+ and CD20+ cells, belonging to the adaptive immune response, are indices of more favorable outcomes in patients with CCA who exhibit lower disease recurrence and an increased overall survival (Goeppert et al., 2013; Loeuillard, Conboy, Gores, & Rizvi, 2019) A clinical trial is actually using Pembrolizumab, a monoclonal antibody anti programmed death (PD)-1, in combination with the anti-VEGF agent Lenvatinib. The rational for this clinical trial is to inhibit the immune suppressive effect of VEGF, thus promoting T cell tumor infiltration and differentiation of CD8+ T cells secreting IFNγ (NCT04550624). A similar approach was used to design another phase Ib/II study, which involves the use of Atezolizumab, a monoclonal antibody inhibiting PD-L1, in combination with Tivozanib, an oral VEGFRs inhibitor to induce activation of CD4+ and CD8+ T cells and DCs (NCT05000294). The stimulation of T cells with antitumor function has also been used in other studies. The phase 1/2, open label, ABILITY study, exploits the use of MDNA11, a modification of IL2 that enhances its action, alone or in combination with checkpoint inhibitor. The use of MDNA11 could therefore lead to an increased activation of naïve CD8+ T-cells and NK cells (NCT05086692). Finally, another phase 2 recruiting study involves the use of Atezolizumab and/or Cobimetinib, a MEK inhibitor, with a monoclonal antibody (CDX-1127) that inhibits CD27, having the ability to activate CD8+ T cells (NCT04941287).

Another cell population potentially important for the containment of CCA is that of DCs, thanks to their known ability to stimulate the activity of T cells (Junking et al., 2017; Panya et al., 2018). An early phase 1 clinical trial involves the infusion of autologous dendritic cells in patients with advanced CCA in combination with Prevnar, a pneumonia vaccine, which aims to further stimulate the immune response (NCT03942328). A second phase 1 study involves the use of CDX-1140, an antibody that targets CD40 alone or in combination with other chemotherapy drugs. CD40 is a key activator of immune response, which is found on dendritic cells, macrophages and B cells, as well as on neoplastic cells (NCT03329950).

NK cells are a fundamental building block for protection from microorganisms and for tumor immune surveillance. Xenograft studies have shown that infusion of NK in athymic mice xenotransplanted with human iCCA cells (HuCCT1) significantly reduces tumor growth in mice (Jung et al., 2018). Given the promising results, two subsequent phase 1/2 studies (“SMT-NK”) have been designed involving the infusion of allogenic NK cells in advanced CCA, but the data are not yet available (NCT03358849, NCT03937895). To date, although it has been shown that the accumulation of TAN is an indicator of poorer survival outcomes for both iCCA and eCCA (Kitano et al., 2018; Zhou et al., 2021), and that MDSCs are heavily involved in immune response depression (Fabris et al., 2019; Rimassa et al., 2019), there are currently no human clinical trials that foresee their modulation.

5. Conclusions

Despite a large body of literature that suggest a prominent role of inflammation in CCA associated with certain liver conditions or risk factors, and that the balance of cells of the innate and adaptive immune responses infiltrating the TME are fundamental in modulating the cancer/host balance, little is currently known about the mechanisms that mediate neoplastic transformation in chronic biliary diseases. We have discussed a few examples, such as PSC CD/CHF and fluke infestation, but the association discussed are mostly correlative and relate more to the pathogenesis of the disease than to the specific carcinogenetic mechanism. This is due on one side to the rarity of these conditions and on the other side to the lack of reliable animal and cellular models (Cadamuro, Brivio, et al., 2018; Cadamuro, Stecca, et al., 2018; Mariotti et al., 2019). There is an urgent need to develop animal models of neoplastic transformation more akin to the human conditions to better study how the inflammatory microenvironment and the presence of cells of the immune response interact with the tumor in order to hypothesize new pharmacological approaches. Functional genomic studies will be able to identify druggable inflammatory signature. Analysis of the microenvironment using single cell transcriptomics will be key to understand cell dynamics and crosstalk and their adaptation under treatment. The future certainly looks to be very active for scholars interested in cholangiocarcinoma and inflammation. Meanwhile, efforts should be devoted to advancing the prevention and treatment of known CCA risk factors and predisposing conditions.

Abbreviations

BilIN

biliary intraepithelial neoplasia

CAFs

cancer-associated fibroblasts

CCA

Cholangiocarcinoma

CCL25

Chemokine C-C motif ligand

CD

Caroli’s disease

Cdc25A

cell division cycle 25 homolog A

CHF

congenital hepatic fibrosis

COX-2

cyclooxygenase-2

CTGF

connective tissue growth factor

CXCR

C-X-C chemokine receptor

CXCL

C-X-C motif chemokine ligand

dCCA

distal CCA

DC

dendritic cells

DPC4

deleted in pancreatic cancer 4

eCCA

extrahepatic CCA

ECM

extracellular matrix

EGFR

epidermal growth factor receptor

ERK

extracellular-signal-regulated kinase

FPC

fibrocystin

FXR

farnesoid X receptor

GWAS

genome-wide association study

HCC

hepatocellular carcinoma

HDAC

histone deacetylase

HLA

human leukocyte antigen

HR

hazard ratio

iCCA

intrahepatic CCA

iNOS

inducible nitric oxide synthase

iPSC

induced pluripotent stem cell

IFN

interferon

IL

interleukin

IPNB

intraductal papillary neoplasm of the bile duct

MadCAM-1

mucosal vascular address in cell adhesion molecule 1

MAPK

mitogen-activated protein kinases

Mcl1

myeloid cell leukemia 1

MDSC

Myeloid-Derived Suppressor Cells

MMP

metalloproteases

miRNA

micro RNA

mTOR

mammalian target of rapamycin

NAFLD

non-alcoholic fatty liver disease

NASH

non-alcoholic steatohepatitis

NK

Neutral Killer Cells

8-NG

8-nitroguanine

NO

nitric oxide

OR

odds ratio

8-oxoDG

8-oxo-2′-deoxyguanosine

OvESP

O. viverrini excretory/secretory products

Ov-GRN-1

O. viverrini granulin

Ov-Trx-1

O. viverrini thioredoxin

PDGF

platelet-derived growth factors

pCCA

perihilar CCA

PSC

primary sclerosing cholangitis

PCNA

proliferating cell nuclear antigen

PKHD1

polycystic kidney and hepatic disease 1

PI3K

phosphatidylinositol 3-kinases

PKA

protein kinase A

ROS

reactive oxidative species

Smad4

signaling effectors mothers against decapentaplegic protein 4

TLR

Toll-like receptors

TCF/LEF

T cell factor/lymphoid enhancer factor

TNF

tumor necrosis factor

TGF-β

transforming growth factor β

TME

tumor microenvironment

TIL

tumor-infiltrating lymphocytes

TAM

tumor-associated macrophages

TAN

tumor-associated neutrophils

VEGF

vascular endothelial growth factor

VCAM-1

vascular cell adhesion molecule 1

YAP

Yes-associated protein