Hepatitis B and C virus infection and the risk of biliary tract cancer: A population-based study in China



Emerging data suggest that chronic hepatitis B virus (HBV) and hepatitis C virus (HCV) infections may also play a role in extrahepatic bile duct cancers. To test the HBV hypothesis, we examined the relationship of HBV/HCV infection with risks of biliary tract cancer and biliary stones in a population-based case-control study conducted in Shanghai, China. Standard assays were used to detect HBV surface antigen (HBsAg) and antibodies against HBV core antigen (anti-HBc) and hepatitis C virus (anti-HCV) in sera from 417 patients with biliary tract cancers, 517 with biliary stones, and 762 healthy controls randomly selected from the population. Unconditional logistic regression was used to calculate the odds ratios (ORs) and 95% confidence intervals (CIs) for each disease type. HBsAg seroprevalence was 7.3% among population controls and 14.2% among patients with extrahepatic bile duct cancer, resulting in a 2.4-fold risk of extrahepatic bile duct cancer (95% CI 1.2–4.5). No association was found for cancers of the gallbladder (prevalence 8.2%) or the ampulla of Vater (6.1%), or for stones in the gallbladder (10.1%) or bile duct (9.3%). Further adjustment for education, smoking, body mass index, diabetes and gallstones did not materially change the results. Prevalence of HCV infection in this population was low (2%), limiting our ability to detect an association with biliary diseases. In Shanghai, an HBV endemic area, chronic HBV infection was associated with a 2.4-fold risk of extrahepatic bile duct cancer. These results should be confirmed in other populations with varying risks of HBV and HCV infection. © 2007 Wiley-Liss, Inc.

Cancers of the biliary tract encompass tumors arising from the gallbladder, extrahepatic bile ducts and ampulla of Vater. Biliary tract cancer is relatively rare in most parts of the world, although high rates are reported for populations in various geographic locations, such as Japan, Korea, India, Eastern Europe and Latin America.1 In the United States, elevated rates occur in the Native American and Hispanic populations.1 From 1972 to 1994, biliary tract cancer was the most rapidly increasing malignancy in Shanghai, China, with an increase in incidence of 119% in men and 124% in women.2

Other than a close link between gallstones and gallbladder cancer and between sclerosing cholangitis and extrahepatic bile duct cancer, the etiology of biliary tract cancer is poorly understood. Obesity, parity, inflammatory conditions of the biliary tract, ulcerative colitis, abnormal choledochopancreatic junction, and choledochal cysts have been suggested as risk factors.1, 3 In addition, chronic infections with Salmonella typhi, Helicobacter species and liver flukes, Clonorchis sinensis and Opisthorchis viverrini, have been implicated.1, 4–6

We reported previously that self-reports of a history of chronic hepatitis or liver cirrhosis were associated with an increased risk of extrahepatic bile duct cancer.7 Because chronic HBV infection in China is the most common cause of hepatitis, cirrhosis and liver cancer, we hypothesized that HBV infection may also be involved in the pathogenesis of biliary tract cancer. Recent studies from China have detected HBV DNA and HCV RNA in tissue specimens of extrahepatic bile duct cancers,8–10 although one recent U.S. study of subject with a low prevalence of HBV chronic infection (0.4%) did not find an association between HBsAg positivity and extrahepatic bile duct cancer.11 As part of a population-based case-control study conducted in Shanghai to investigate the etiology of biliary tract cancer, we examined the association of HBV and HCV infections with biliary tract cancers as well as biliary stones.

Material and methods

Study subjects

Details of the study have been reported elsewhere.7, 12, 13 Between June 1997 and May 2001, using a rapid reporting system established between the Shanghai Cancer Institute (SCI) and 42 collaborating hospitals in 10 urban districts of Shanghai (henceforth referred to as urban Shanghai), we recruited and enrolled a total of 627 patients newly diagnosed with primary biliary tract cancer (ICD9 156). This reporting system captured over 95% of biliary tract cancer patients in Shanghai. A total of 1,037 biliary stone case patients (774 gallbladder and 263 bile duct) were selected by frequency matching to cancer case patients on age (5-year intervals), gender, and hospital. Population controls were 959 healthy subjects without a history of cancer who were randomly selected from all permanent residents listed in the Shanghai Resident Registry, frequency matched to cancer cases on age (5-year intervals) and gender. Of the eligible cancer patients and control subjects, 95% and 82% respectively agreed to participate in the study. For all study subjects, informed consent was obtained and the study protocol was approved by the Institutional Review Boards of the U.S. National Cancer Institute (NCI) and the SCI. All subjects provided written informed consent.

Pathology and clinical review

The diagnoses of biliary tract cancer and stones were confirmed, whenever possible, by review of the pathology slides of cancer and stone cases independently by two pathologists. Clinical data and imaging studies, such as ultrasonography, magnetic resonance imaging (MRI), computed tomography (CT) scans and endoscopic retrograde cholangiopancreatography (ERCP), were used to confirm the diagnosis of cancer cases without histologic materials and gallstone controls without surgery. The primary site of cancer was obtained from surgical/histopathology reports from local hospitals as well as imaging studies.


Information on potential risk factors, including smoking, use of alcohol, body size, diet, reproductive history, family history of cancer and medical history, was elicited from all study subjects through an in-person interview by trained interviewers using a structured questionnaire. The response rates for in-person interviews were over 90% for both cases and controls. To ensure interviews were conducted uniformly among subjects and high quality data were collected, all interviews were tape-recorded and reviewed. In addition, 5% of the study subjects were re-interviewed within three months of the original interview, and the concordance of responses to key questions between interviews was above 90%.

Gallstone assessment

Gallstone status, in either the gallbladder or extrahepatic bile ducts, was determined for cancer and biliary stone patients, using data from clinical imaging studies (abdominal ultrasound, CT scan, ERCP and MRI), medical record review, and self-reported history of gallstones at interview. Individuals who reported having prior cholecystectomy or history of gallstones were classified as having had biliary stones. For control subjects, status of biliary stones was determined based on self-reported gallstone history at interview and abdominal ultrasound data. About 85% of the controls agreed to ultrasound screening for the detection of stones.

Biological samples

At interview, subjects were asked to donate an overnight fasting blood sample, which was obtained from over 90% of the cases and 80% of the controls. Collected blood was processed within four hours at the SCI central laboratory. Serum, plasma, buffy coat samples and red blood cells were stored at −70°C at SCI until shipped on dry ice via express mail to the NCI repository and kept frozen until use. Gallstones, bile and tissue samples were collected from patients who underwent surgical resection of the gallbladder or bile duct.

Serologic assays

Serology testing was carried out on subjects with blood specimens, including all of the biliary tract cancer cases from the three subsites, all subjects with bile duct stones, and a random sample of 40% of subjects with gallstones in view of the large sample size of this group. For comparison, we included all population controls, since in the general population in Shanghai, the prevalence of chronic HBV infection is modest (<10%) and HCV infection relatively low (<3%). Both HBV and HCV assays were conducted at the Science Applications International Corporation (SAIC)-Frederick Viral Epidemiology Section (VES) laboratory located at NCI-Frederick, MD.

HBV assays

To assess HBV status among study subjects, we measured HBV surface antigen (HBsAg) and IgG antibody to HBV core antigen (anti-HBc) using enzyme immunoassays, following the directions provided by the manufacturers. The presence of HBsAg was determined using the Bio-Rad Genetic Systems™ HBsAg EIA 3.0 kit (Bio-Rad Laboratories, Hercules, CA) and anti-HBc was tested using the Hepatitis B Virus Core Antigen (Recombinant) ORTHO® HBc ELISA Test System (Ortho-Clinical Diagnostics, Raritan, NJ). HBV-uninfected subjects were defined as those negative for both anti-HBc and HBsAg. Subjects who were positive for anti-HBc but negative for HBsAg were classified as having had prior, but not current, infection with HBV (also called resolved infection). Chronic HBV infection was defined as the presence of both HBsAg and anti-HBc. Fourteen subjects (0.8%), including 3 gallbladder cancer cases, 4 gallstone cases, 1 bile duct stone case and 6 population controls, who tested positive for HBsAg but negative for anti-HBc were excluded from the analysis as it was not possible to determine whether they had a chronic HBV infection, an acute HBV infection, or a false-positive HBsAg result.

HCV assays

HCV infection status among study subjects was determined by measuring IgG antibodies to HCV (anti-HCV) using an ELISA assay (ORTHO® HCV Version 3.0 ELISA Test System, Ortho-Clinical Diagnostics). Specimens that were positive for anti-HCV EIA were further tested with a recombinant immunoblot assay (RIBA) for viral proteins (Ortho-Clinical Diagnostics, version 3) for confirmation. Subjects that were positive by both anti-HCV ELISA and HCV RIBA were considered to be HCV positive.

Quality control

For the HBsAg assay, each subject was tested at both the VES SAIC-Frederick and the SCI laboratories. Those with discrepant results (<1%) were retested by the VES laboratory for confirmation. Ten percent of the submitted samples were duplicate quality control samples included for the assessment of assay reproducibility. In addition, in each batch, both positive and negative control samples were included to monitor the quality of the assay.

Statistical analysis

Fisher's exact test was used to detect differences between cases and controls for selected characteristics. In order to make the appropriate case-control comparisons, gallbladder cancer cases were compared with controls without a history of cholecystectomy; bile duct cancer cases and ampulla of Vater cancer cases were compared with all controls; and biliary stone cases were compared with population controls without biliary stones. All variables, including education, body mass index, gallstones and diabetes, that differed significantly between cases and controls (p < 0.05) and were related to prevalence of HBV or HCV were considered potential confounders of risk. Unconditional logistic regression analyses were used to compute the odds ratios (ORs) and 95% confidence intervals (CIs) for biliary tract cancer in relation to serologic markers of HBV and HCV.14 Odds ratios were estimated for biliary tract cancer and each anatomic subsite. In all the analyses, ORs were adjusted for age and p values were two-sided.


A total of 417 patients with biliary tract cancer (234 gallbladder, 134 bile duct and 49 ampulla of Vater), 301 with gallbladder stones, 216 with bile duct stones and 762 population controls were included in the analysis (Table I). Age at diagnosis ranged from 35 to 75 years (median 67) for cancer cases. Subjects with gallstones were slightly younger than cancer cases; gallbladder cancer and gallstone subjects were more likely to be female. Compared with controls, cancer cases for all 3 subsites were more likely to have gallstones, while gallbladder cancer cases were more likely to have diabetes and bile duct cancer cases were more likely to be smokers, use alcohol or have a history of liver cirrhosis.

Table I. Selected Characteristics of Cases and Controls, Shanghai, China
Characteristics  Biliary tract cancerBiliary stones
Controls1GallbladderExtrahepatic bile ductAmpulla of vaterGallbladder stonesBile duct stones
  • *

    Fisher's exact test, p < 0.05.

  • 1

    Current smokers defined as those who were smoking cigarettes regularly at the time of interview.

  • 2

    Current drinkers defined as those drinking alcohol regularly at the time of interview.

 Jr. Middle18624.45523.53627.11428.69230.65726.4
 Sr. Middle14719.32812.02216.51020.47825.94621.3
 Some College11415.0229.41914.3510.25116.93717.1
Marital status            
 Never Married121.620.921.500.020.720.9
Smoking (%)122229.16427.55843.3*2142.9*7825.98037.0*
Alcohol use (%)215920.93916.74332.1*1122.54916.34320.0
Diabetes (%)618.02912.5*197.536.13010.02411.2
Hypertension (%)31441.28536.34029.9*1326.5*9029.9*5826.9*
Gallstones (%)17923.519784.2*8966.4*2449.0*301100.0*216100.0*
Liver cirrhosis (%)101.331.375.2*24.110.352.3
Body mass index            

Among population controls, 73.8% were anti-HBc positive, indicating a history of HBV infection, including 7.3% with chronic HBV infection (HBsAg+, anti-HBc+) (Table II). The prevalence of chronic HBV infection in controls was similar to that in patients with gallbladder cancer (8.2%), ampulla of Vater cancer (6.1%), gallstones (10.1%) and bile duct stones (9.3%). In contrast, patients with extrahepatic bile duct cancer were more likely to be positive for HBsAg (14.2%), resulting in a 2.4-fold risk (95% CI 1.2–4.5) after adjustment for age. Further adjustment for education, smoking, drinking, obesity and diabetes did not materially change the results. No association was seen between resolved infection (anti-HBc +, HBsAg -) and biliary disease. Although extrahepatic bile duct cancer was slightly more common in men, there was no gender difference in the prevalence of HBsAg (men 6.8%, women 7.6%), and similar risks were seen for men (OR = 2.2, 95% CI 0.9–5.4) and women (OR = 2.5, 95% CI 1.0–6.7). As expected, the prevalence of HCV infection was low, 1.3% in men and 2.4% in women. No association between HCV and biliary tract cancer or biliary stones was evident. Of the 1,682 subjects, only 1 person (0.06%) had concurrent infection with both HBV and HCV.

Table II. Odds Ratios (ORs) and 95% Confidence Intervals (CIs) for Biliary Tract Cancer and Stones in Relation to Hepatitis B Virus (HBV) and Hepatitis C Virus (HCV) Infection, Shanghai, China
HBV/HCV  Biliary tract cancer
ControlsGallbladderExtrahepatic bile ductAmpulla of vaterGallstonesBile duct stones
N%N%OR195% CIN%OR195% CIN%OR195% CIN%OR195% CIN%OR195% CI
  • 1

    Adjusted for age.

  • 2

    Excluded 16 subjects with anti-HBc negative and HBsAg positive.

Total762100.0234100.0  134100.0  49100.0  301100.0  216100.0  
Anti-HBc, HBsAg2                      
 Negative, negative19826.25222.51.02921.61.01020.41.08026.91.06429.81.0
 Positive, negative50366.516069.31.20.8–1.78664.21.20.7–1.83673.51.40.7–2.818763.01.00.7–1.413160.90.80.6–1.2
 Positive, positive557.3198.21.30.7–2.41914.22.41.2–4.536.11.20.3–4.33010.11.10.7–1.9209.31.10.6–2.0

The risks of bile duct cancer associated with chronic HBV infection in combination with biliary stones and liver cirrhosis are shown in Table III. Among subjects without biliary stones, chronic HBV infection was associated with a 2.2-fold risk (95% CI 0.9–5.2). Among subjects with biliary stones, the risk was 7.2-fold with stones alone and 9.4-fold in conjunction with HBsAg positivity; no significant interaction was seen between HBsAg and gallstones (p interaction =0.39). Chronic HBV infection in conjunction with liver cirrhosis (self-reported) appeared to heighten the risk of bile duct cancer (OR = 16.6, 95% CI 1.0–266.5), although the estimate was based on small numbers.

Table III. Odds Ratios (ORs) and 95% Confidence Intervals (CIs) for Bile Duct Cancer in Relation to HBsAg, Gallstones and Liver Cirrhosis, Shanghai, China
 ControlsExtrahepatic bile duct cancerp-interaction
N%N%OR195% CI
  • 1

    Adjusted for age.

Gallstones, HBsAg       
 No gallstones, HBsAg (−)54070.93828.41.0 
 No gallstones, HBsAg (+)435.675.22.20.9–5.2 
 Gallstones, HBsAg (−)16121.17757.47.24.6–11.10.39
 Gallstones, HBsAg (+)182.4129.09.44.2–21.1 
Liver cirrhosis, HBsAg       
 No cirrhosis, HBsAg (−)69290.811082.11.0 
 No cirrhosis, HBsAg (+)607.91712.71.50.8–2.8 
 Cirrhosis, HBsAg (−)–8.40.30
 Cirrhosis, HBsAg (+)10.121.516.61.0–266.5 


In this population-based case-control study in Shanghai, chronic HBV infection was associated with a 2.4-fold risk of extrahepatic bile duct cancer. In contrast, HBV infection was not associated with the risk of cancers of the gallbladder or ampulla of Vater. Although our study did not examine cancers of the intrahepatic bile ducts, evidence from other studies suggests that these tumors may also be linked to HBV infection.15, 16 It is noteworthy that of the 19 bile duct cancer cases that were positive for HBsAg in our study, over 70% of the tumors were in the upper third of the extrahepatic bile ducts (vs. 62% in all bile duct tumors), where the intrahepatic and extrahepatic bile ducts meet. These results suggest that HBV, the main causative agent for hepatocellular carcinoma (HCC) worldwide, also increases the risk of bile duct tumors.

The link between chronic HBV infection and bile duct cancer appeared to be independent of gallstones, with the risk being slightly more pronounced among subjects without gallstones (OR = 2.2). Among those with gallstones, HBV was associated with a 1.3-fold risk of bile duct cancer. No interaction was seen between HBV infection and biliary stones. Although gallstones are a major risk factor (OR = 7) for bile duct cancer in our study population, only 50% of the bile duct cancer cases in our study had evidence of gallstones, despite the rigorous review of clinical and pathological data, suggesting that factors other than gallstones may be involved in the etiology of bile duct cancer. On the basis of the prevalence of gallstones and HBsAg in our study, 57% (95% CI 46–68%) of the bile duct cancers in the population could be explained by gallstones alone, 21% (12–36%) by the combination of gallstones and HBsAg, and only 8% (0.2–20%) by HBsAg alone.

Consistent with our finding, recent clinical surveys in China have detected HBV DNA and HCV RNA in tissue specimens from both intrahepatic and extrahepatic bile duct cancers, with extrahepatic tumors arising mainly from the hilum, where the intrahepatic and extrahepatic bile ducts meet,8–10, 17, 18 although the mechanism relating HBV infection to extrahepatic bile duct cancer is unknown. HBV induces HCC mainly by causing a chronic necroinflammatory process with regeneration of hepatocytes.19 A similar mechanism may be involved in intrahepatic and extrahepatic bile duct cancers, particularly since common progenitor stem cells differentiate into hepatocytes and cholangiocytes. The molecular pathway underlying HBV-related bile duct cancer, as for HCC, may involve activation or suppression of cellular genes by direct integration of viral DNA in the cellular genome or by expression of viral protein, such as the HBV X protein.17 HBx mRNA expression has been reported in a high fraction of extrahepatic bile duct carcinomas from China,18 with 61% of the tumor tissue having detectable HBx mRNA levels, compared with only 18% of the non-neoplastic bile ducts. Similarly, HBV X gene expression was shown in 31 of 40 patients with intrahepatic cholangiocarcinoma by in situ hybridization.20

Strengths and limitations of this study should be noted. Selection bias appeared to be minimal, due to relatively complete case ascertainment and high response rates in study subjects. Misclassification of the cancer subsite was unlikely due to the rigorous clinical and pathological review procedures implemented in the study. Misclassification of gallstone status, the major risk factor for biliary tract cancer, should also be minimal, since assessment of gallstone status was nearly complete, as a result of careful review of medical records for cases and ultrasound examination of population controls. Similarly, misclassification of HBV status was unlikely, since assays of high sensitivity and specificity were conducted independently in both NCI and SCI laboratories, with highly concordant results. However, without further molecular work, we are not able to establish HBV viral load and genotype, major determinants of HCC.19 Because Shanghai is an endemic area for HBV infection, we cannot rule out the possibility that HBV infection among some bile duct cancer cases was acquired during ERCP examination through contaminated equipment. However, such an effect should be minimal, since only about 10% of subjects infected with HBV as adults become chronic carriers. In addition, the HBV effect observed in the study was specific for bile duct cancer primarily in the upper third of the duct system adjacent to the liver, although all biliary tract cancer cases had similar diagnostic procedures. Despite being the largest study to date, a major weakness of the study was the limited statistical power to examine the risks associated with ampulla of Vater cancer or the interactive effects of HBV with gallstones and other risk factors. Because of the low prevalence of HCV infection (2%) in our study population, we lacked sufficient power to evaluate its role in bile duct cancer, as previously suggested by surveys relating HCV infection to intrahepatic bile duct cancer,15, 21 and we did not test for HCV RNA to estimate the extent of chronic HCV infection in the Shanghai population.

In summary, this case-control study of biliary tract cancer revealed a 2.4-fold risk of extrahepatic bile duct cancer associated with chronic HBV infection. Worldwide, there are 350 million chronic carriers of HBV and 50 million new cases of HBV infection occur annually,22 with 5–10% of adults and 90% of infants becoming chronic carriers. Although biliary tract cancer has a low incidence in most populations, its high fatality rate intensifies the need for further studies to clarify the role of HBV and HCV in extrahepatic as well as intrahepatic bile duct cancer.


The authors thank Dr. Jie Deng, Ms. Jiarong Cheng, Ms. Lu Sun, Mr. Kai Wu and the staff at the Shanghai Cancer Institute for data collection, specimen collection, and processing, surgeons at the collaborating hospitals for data collection, and local pathologists for pathology review; Ms. Hope Cohen-Webb, Ms. Lisa McFarland, Ms. Kathy Hurt-Mullen, Sara Glashofer, Nancy Odaka, John Heinrich, and Linda Lannom of Westat for data preparation and management; Ms. Georgina Mbisa of the SAIC-Frederick VES laboratory for performing HBV and HCV serological testing, and Ms. Janis Koci of the Scientific Applications International Corporation for management of the biological samples. The content of this publication does not necessarily reflect the reviews or policies of the Department of Health and Human Services, nor does mention of trade name, commercial products, or organizations imply endorsement by the U.S. Government.