HBV & Risk for Cancers Outside the Liver
Report by Jules Levin
It is well known that hepatitis C & B can both lead to hepatocellular carcinoma. There were two studies at DDW (May 2005) finding an association between HBV & cancers outside the liver; both are reported below. One study found an association between hepatitis B and pancreatic cancer (increased risk for pancreatic cancer of 2-7 times); a second study found an association between HBV & non Hodgkins Lymphoma (NHL); hazard ratios of 2 to 3 fold increased risk, after adjusting for HIV & HCV).
A third study at DDW evaluated viral hepatitis (HVB, HBV) and liver cirrhosis as risk factors for intrahepatic (ICC) and extrahepatic cholangiocarcinoma (ECC). The study authors concluded that chronic HCV and possibly HBV are risk factors for ICC but not ECC. Liver cirrhosis is a common and strong risk factor for ICC and less so for ECC. A cholangiocarcinoma may arise anywhere along the liver secretion (biliary) ducts. These tumors produce symptoms by blocking the bile ducts. They affect both sexes, and a majority of cases are found in patients above the age of 65.
"CHRONIC HEPATITIS B INFECTION in ACTIVE REPLICATION is a RISK FACTOR for PANCREATIC CANCER - a Long-Term Cohort Study"
HI Yang1, J Su2, CL Jen1, UH Iloeje2, E Kuo3, CJ Chen1
1National Taiwan University, Taiwan; 2Bristol-Myers Squibb Pharmaceutical Research Institute, Wallingford, CT, USA; 3Bristol-Myers Squibb, Taipei, Taiwan
Introduction: The relationship between hepatitis B virus (HBV) and pancreatic cancer remains unclear. Hepatitis B surface antigen (HBsAg) has been found to localize within the pancreatic cells as well as to be associated with pancreatitis. It has been hypothesized that HBV can infect and replicate in human pancreatic cells and play some role in the oncogenesis of pancreatic carcinoma.
Methods: A cohort of 23,819 subjects without evidence of pancreatic cancer was recruited from seven townships in Taiwan between 1991 and 1992. Serum samples obtained at enrollment were tested for HBsAg and hepatitis B e-antigen (HBeAg) by radioimmunoassay. The diagnosis of pancreatic cancer was ascertained through data linkage with the computerized profiles of the National Cancer Registry and Death Certification System in Taiwan. Cox’s proportional hazard models were used to calculate the relative risk of pancreatic cancer among subjects who were positive for HBsAg alone, and for HBsAg and HBeAg together, in comparison with those who were negative for both HBV markers.
Results: During 254,726 person-years of follow-up, 23 patients with pancreatic cancer were newly diagnosed. After adjustment for gender, age, residential area, educational level, betel quid chewing habits, cigarette smoking, and alcohol consumption, the relative risk of developing pancreatic cancer was found to be 2.4 (95% confidence interval, 0.9-6.7) in subjects who were positive for HBsAg alone, and 7.3 (95% confidence interval, 1.6-33.9) in those who were positive for both HBsAg and HBeAg, as compared with subjects who were negative for both HBV markers.
Number of cohort members: 19,665 in HBsAg & HBeAg negative (229,223 person-year of follow-up); 3.461 in HBsAg+ & HBeAg-neg (39,999 person-year of followup); 615 in HBsAg & HBeAg positive (6,799 person-year of followup).
Conclusion: Chronic hepatitis B infection and active replication of HBV is considered to be associated with an increased risk of pancreatic cancer.
AUTHOR CONCLUSIONS & SUMMARY:
--In this large sample population-based prospective observational cohort study, infection
with HBV was associated with an increased incidence of pancreatic cancer
--The highest incidence of pancreatic cancer was seen in persons sero-positive for both
HBsAg and HBeAg indicative of active replication
--These data strongly suggest that HBV infection may play a causative role in the development of pancreatic cancer even when adjusted for other risk factors
--To our knowledge, these data represent the largest and most robust study to date evaluating the relationship between chronic HBV infection and any nonhepatic malignancy
Chronic infection with HBV should be considered a risk factor for the development of pancreatic cancer. In addition to the well described hepatic complications of HBV infection, pancreatic cancer should be considered an additional cause of morbidity associated with HBV infection.
--Globally, 232,306 new cases of pancreatic cancer and 227,023 deaths from
pancreatic cancer were reported in the year 20021
--Pancreatic cancer is one of the most lethal malignant diseases with almost all patients dying within one year of diagnosis2
--Environmental risk factors for pancreatic cancer are not completely understood and cigarette smoking and alcohol consumption are two well described risk factors
--Hepatitis B virus has been isolated and can replicate in human pancreatic cells, and it has been postulated that infection with HBV may be a risk factor for pancreatic cancer3-6
--Large-scale epidemiologic studies are useful for examining risk factors for diseases such as pancreatic cancer
--We report on a prospective observational population-based cohort study of the
relationship between HBV in active replication and pancreatic cancer in 23,819
This is a population-based prospective observational cohort study
--47,079 male and 42,214 female Taiwanese subjects living in seven townships in Taiwan were invited to participate in this study between 1991 and 1992
--Permanent cohort size of 23,819 subjects (11,972 males and 11,847 females) was established between 1991 and 1992
Subject screening and entry:
--Eligible subjects were aged 30 to 65 years at study entry
- Subjects completed a structured questionnaire administered by trained public health nurses, had a health examination and provided a blood sample
Risk factor assessment:
- Serologic testing was conducted on serum samples at cohort entry as follows:
HBsAg and HBeAg (Abbott Laboratories, North Chicago, IL)
HBsAg and HBeAg testing was completed in 23,741 (99.7%) subjects
- Assessment of demographics and environmental risk factors was done at cohort entry and included the following:
Betel quid chewing
Personal medical and surgical history
Family history of malignancies
Gynecological history and use of hormonal replacements for females
Pancreatic Cancer Ascertainment
--All subjects were free of detectable pancreatic cancer at cohort entry
--Cases were ascertained through linkage to the Taiwan National Cancer Registry. To ensure complete ascertainment, data linkage was also conducted with the Taiwan National Death Certificate Profiles and the Taiwan Bureau of National Health Insurance profiles
There were 24 cases of confirmed pancreatic cancer:
- 10 confirmed histopathologically as pancreatic adenocarcinoma
- 14 diagnosed by computed tomography scan and/or endoscopic retrograde
1. Ferlay J, Bray F, Pisani P, Parkin DM. GLOBOCAN 2002: Cancer Incidence, Mortality and Prevalence Worldwide. IARC CancerBase No. 5. version 2.0, IARCPress, Lyon, 2004 http://www-dep.iarc.fr (accessed November 8th 2004).
2. Gloeckler Ries LA, Hankey BF, Miller BA, et al., eds. Cancer Statistics Review. 1973-1988. Bethesda, Md.: Department of Health and Human Services, 1991.
3. Hoefs JC, Renner IG, Ashcavai M, Redeker AG. Hepatitis B Surface Antigen in Pancreatic and Biliary secretions. Gastroenterology. 1980;79:191-194.
4. Shimoda T, Shikata T, Karasawa T, Tsukagoshi S, et al. Light Microscopic Localization of Hepatitis B Virus Antigens in the Human Pancreas. Gastroenterology. 1981;81:998-1005.
5. Yoshimura M, Sakurai I, Shimoda T, et al. Detection of HBsAg in the Pancreas. Acta Pathol Jpn. 1981;31:711-717.
6. Hohenberger P. Detection of HBs-Ag in the Pancreas in Cases of Pancreatic Carcinoma. Hepato-Gastroenterol. 1984;31:239-241.
"INCIDENCE of NON-HODGKINS LYMPHOMA AMONG PATIENTS EXPOSED TO HEPATITIS B VIRUS"
Authors: CA Caldwell. C Quesenberry, Jr, K wells, D Guo, J Shan, G Kuerra, M Skovron, MU Yood
Background: Some research suggests that HCV may be associated with Non Hodgkins Lymphoma (NHL), however little is known regarding the incidence of NHL among patients exposed to HBV.
Objectives: The purpose of this analysis is to quantify the incidence of NHL among patients diagnosed with HBV.
Methods: Using automated encounter & lab data (1/1/95-12/31/2001) from two USA health systems, we categorized 3 distinct cohorts of patients based on available HBV antigen or diagnosis of HBV: 1) Laboratory Confirmed CHB (CHB): at least two positive HBV surface antigen tests recorded >= 6 months apart or confirmation through subsequent lab testing/record review; 2) Surface Antigen Positive (SA): at least 1 positive HBV surface antigen test with no subsequent testing/confirmation; 3) ICD-9 Coded (ICD-9): at least one ICD-9 coded encounter recorded with HBV. Three different cohorts were studied, due to inherent limitations of automated data case identification. This approach allowed for assessment of potential misclassification of cases & the effect on results. Three mutually exclusive comparison cohorts were randomly sampled (50 comparison patients for each HBV case) matched on age (+/- 5 yrs), sex, index date, etc. Cohorts were merged with tumor registries to obtain cancer diagnosis data. NHL was diagnosed after the determination of HBV. Patients were followed through date of NHL diagnosis or disenrollment from the plan or 12/31/2002.
We identified 4,190 CHB patients and 209, 358 comparison patients; 7,449 SA patients and 372,425 comparison patients.
3,963 ICD-9 patients and 198,453 comparison patients.
The rate of NHL was as follows:
--CHB: 1.03 per 100,000 person-years (95% confidence interval (CI) 0.62-1.61)
--SA: 0.43 per 100,000 person-years (CI 0.20-0.80)
--ICD-9: 1,10 per 100,000 person-years (CI 0.55-1.92)
Comparing these rates to respective comparison cohorts produced the following
crude rate ratios (RR):
--CHB: crude RR=6.74 (CI 3.91, 10.86)
--SA: RR=3.08 (CI 1.39-5.86)
--ICD-9: RR=4.33 (CI 2.13-7.81)
Adjusting for age, race, sex, income, HIV & HCV infection yielded the following
adjusted hazard ratios (HR):
--CHB: adjusted HR=2.23 (CI 1.18-4.23)
--SA: HR=2.97 (1.41-6.27)
--ICD-9: HR=2.83 (1.34-5.96)
Patients with HBV are at increased risk for NHL, even after adjusting for HIV & HCV, age, race, sex & income. This increased risk is consistent, regardless of case-definition.
"HEPATITIS C Is A RISK FACTOR FOR CHOLANGIOCARCINOMA"
Authors: YH Shaib, HB El-Serag, M HASSAn
This is a hospital-based prospective case-control study. Cases were patients with histologically confirmed cholangiocarcinoma referred to MD Anderson Cancer Center between 1992-2002 and were matched (1:2 for ICC and 1:1 for ECC) by age and gender to controls obtained from the companions of patients with a diagnosis of cancer, other than hepatobiliary cancers. Data were collected using a structured questionnaire about liver disease, family historu, diabetes, smoking, and alcohol. Blood samples from all participants were tested for HBsAg, HBc and anti-HCV antibodies. Comparisons were conducted in bivariate and multivariate logistic regression analyses.
They identified 246 cases (83 cases with ICC and 163 with ECC) and 156 controls. There were no significant differences between cases and controls in age, gender, and race.
Compared with controls, ICC cases had higher prevalence of anti-HCV antibodies (6%) vs 1.3%, p=0.04) and Anti-HBc antibodies (10.8% vs 0.65%, p=0.0002) but not HBsAg (1.2% vs 0.7%, p=0.07). The race-adjusted odds ration for HCV was 4.5 (1.1-34.5) and for Anti-HBc antibodies was 18.3 (3.3-340.0).
The prevalence of diabetes (14.5% vs 9.2%), p=0.2), smoking (44.6% vs 47.4%, p=0.7) and alcohol (55.4% vs 57.9%, p=0.7) were not significantly different between ICC cases and controls.
Most of the ICC patients with HCV (80%) had liver cirrhosis. Only 2 ICC cases (2.4%) had ulcerative colitis or sclerosing cholangitis. The prevalence of most of the risk factors was not different between ECC cases and controls: anti HCV (3.7% vs 1.3%, p=0.2), HBsAg (0 vs 0.6%, p=0.3), diabetes (11.6% vs 9.2%, p=0.5), and smoking (47.2% vs 47.4%, p=0.9).
The prevalence of cirrhosis was significantly higher among patients with ICC than those with ECC (24.1% vs 4.9%, p<0.0001).
Conclusion: Chronic HCV and possibly HBV are risk factors for ICC but not ECC. Liver cirrhosis is a common and strong risk factor for ICC and less do for ECC.