Meta-analysis was employed by a recently published study to estimate the correlation between hepatitis virus infection and the risk of ICC and extrahepatic cholangiocarcinoma (ECC) [23
]. Because different anatomic location of CC has distinct epidemiologic features indicating different sets of risk factors [11
], here we used meta-analysis to examine the relationship between HBV/HCV infection and the risks of ICC, and restricted our analysis to patients only with ICC.
In this meta-analysis, we found positive relationship between HBV/HCV infection and the development of ICC. This conclusion is further supported by the evidences from a series of experimental studies. In a study from the United States, HBV and/or HCV DNA was present in 3 (27%) out of 11 ICC tissue samples obtained at the time of surgical resection [24
]. Another study from China also showed that HBV DNA was detected in 34.8% (8/23) of ICC cases [25
]. Recently, Torbenson et al.
] found that dysplasia of the intrahepatic bile ducts, the histologic precursor lesion of ICC, was present in HCV related cirrhotic livers.
The mechanism of the development of ICC in patients with HBV or HCV infection is still uncertain. HBV and HCV infection can cause chronic inflammation of the liver. Indeed, chronic inflammation and cancer are closely associated. In this context, long-term expression of several viral oncoproteins, mostly the HCV core protein and HBx protein, might participate in the tumourigenic process. Battaglia et al.
] found that HCC derived-HCV core variants alleviate TGF-beta cytostatic responses and increase TGF-beta-mediated epithelial to mesenchymal transition (EMT) in mouse or human primary hepatocytes. Interestingly, Li et al.
] reported that HCV core protein expression can induce EMT in cholangiocarcinoma cell line. On the other hand, HBx acting as a transactivator on various cellular genes that are involved in the control of the cell cycle, proliferation or apoptosis has been shown to be present in some HBV-infected ICC specimens [29
]. There are reports providing evidence that HBx gene transfection can upregulate the transcriptional expression of human telomerase reverse transcriptase mRNA both in HCC and cholangiocarcinoma cell lines [30
]. Clinically, Peng et al.
] reported that ICC patients with HBV infection were more common in male and younger patients as compared to ICC patients with seronegative HBsAg, similar to HCC patients. Furthermore, the profile of age distribution of ICC patients with HBV infection was roughly close to that of HCC patients. Similarly, Lee et al.
] also found that both viral hepatitis-associated ICC and HCC had similar age profiles and difference in age distributions between patients with chronic hepatitis B and patients with chronic hepatitis C. In addition, the surgical outcomes of ICC patients with hepatitis, preoperatively diagnosed as HCC, is favorable, similar to that of typical HCC patients [31
]. Taking these clinical and experimental findings into account, it seems to be reasonable that hepatitis virus–associated ICC and HCC share common disease processes for carcinogenesis.
Some researchers suggest that hepatocytes and cholangiocytes might originate from hepatic progenitor cells (HPCs) [32
]. Therefore it is possible that both ICC and HCC in patients infected with HBV and HCV may be derived from HPCs. In fact, cumulative studies have provided strong evidence to this hypothesis. Cumulative evidence suggests that HBV- and HCV-associated HCC and ICC are derived from HPCs. (i) It has been demonstrated that HPCs can be infected with HBV and HCV, and that proliferation of large numbers of HPCs was seen in HBV- and HCV-associated cirrhosis which is a risk factor for liver cancer [33
]. (ii) Based on gross morphological features, ICC are classifiable into three representative types of growth patterns: the mass-forming (MF) type, the periductal infiltrating (PI) type, and the intraductal growth (IG) type [35
]. It is likely that PI and IG type tumors arise from malignant transformation of epithelial cells lining the larger bile ducts, whereas the MF type arises from smaller bile ducts or liver progenitor cells within portal areas [36
]. Yamamoto et al.
] speculated that HCV-infected proliferating cholangioles in patients with chronic hepatitis C might be associated with the development of MF type ICC. Yu et al.
] found that viral hepatitis-associated ICC was more likely to be of the MF type rather than the PI and IG type. (iii) Alpha-fetoprotein (AFP) is a marker of HPCs compartments [38
]. Viral hepatitis-associated ICC patients were found had elevated serum alpha-fetoprotein levels as compared with seronegative ICC patients [19
]. One possible mechanism for the development of ICC is that the neoplastic transformation of HPCs is involved in the genesis of ICC and that the HPCs retain their ability to produce AFP through the process of malignant transformation [38
]. Indeed, Ishii et al.
] demonstrated that AFP-producing cells in cholangiocarcinomas possessed cancer stem cell-like properties. More importantly, one most recent study from China found that HBx induces intrinsic cellular transformation, promoting the expansion and tumorigenicity of HPCs in 3,5-diethoxycarbonyl- 1, 4-dihydrocollidine treated mice [40
ICC is characterized by wide variability in incidence and risk factors. HCV seems to be associated with ICC in regions with relatively low prevalence of HBV infection such as United States, Italy and Japan. In contrast, several studies from China [9
], a highly endemic area for HBV infection, found that not HCV but HBV infection was significantly associated with ICC. On the other hand, a recent study from Taiwan, where both HBV and ICC are endemic, found that both HBV and HCV are significantly associated with ICC [17
]. These studies indicate that the different endemic hepatitis virus types in different regions may be a determinant to the variability of risk factors for ICC.
Our study also has some weaknesses which should be considered when interpreting results. First, seropositivity for HBsAg and anti-HCV was used as the sole indicator of HBV and HCV infection. It seems that occult HBV and HCV infection may also play a role in the development of HCC [41
]. Therefore, the definition in the present report might result in underestimation of the effect of hepatitis viruses infection. Second, there were significant heterogeneities between different studies, which would lower the reliability of the summary odds ratio. The presence of heterogeneity is most likely to have been introduced by diversity in study design, population demographics, location of studies conducted and adjustments for confounders. Finally, the possibility of publication bias is of concern. The fact that studies with positive results are more likely to be published than negative and non-English studies were excluded in the current meta-analysis might be sources of potential publication bias.