Title: Aflatoxin B1, Hepatitis B, and IFNA17 on the Risk of Liver Cancer: An example of the application of exposure markers in cancer epidemiology
1Aflatoxin B1, Hepatitis B, and IFNA17 on the Risk
of Liver Cancer An example of the application
of exposure markers in cancer epidemiology
- Binh Y. Goldstein, PhD
- Epidemiology 243
2Introduction
3BackgroundEpidemiology of liver cancer
- Worldwide
- Sixth most common new cancer
- Third most common cause of cancer death
- Rates among men 2-3x higher than women
- Over 80 of new cases occur in developing
countries - Low 5-year relative survival rates (lt15)
- Source GLOBOCAN 2002
4Age standardized incidence of liver cancer in
world among men
5BackgroundEpidemiology of liver cancer
- US
- 19,200 new cases and 16,800 deaths
- Five year survival rate is 10
- 8th leading cause of deaths from cancers in both
sexes - 6th for men
- 10th for women
- Source American Cancer Society, 2007
- China
- 345,844 new cases and 321,851 deaths
- accounts for over 53 of all liver cancer cases
and deaths worldwide - 3rd most common cancer among men
- Most common cause of death from cancer among men
- Source GLOBOCAN 2002
6BackgroundRisk factors for liver cancer
- Hepatitis B virus
- 350 million people chronically infected worldwide
- About two-thirds of liver cancers in China and
Southeastern Asia are attributed to HBV infection
- Chronic carriers have a 20-fold increase in risk
compared with non-carriers - Major pathways by which HBV infection increases
risk for liver cancer are - (1) viral DNA integration
- (2) oncogenic proteins
- (3) inflammation
7BackgroundRisk factors for liver cancer
- Aflatoxin B1
- Toxin found in mildewed grains and nuts
- Bioactivated intermediate AFB1-exo-8,9-epoxide
has carcinogenic effect - Adducts associated with increased risk of liver
cancer - Associated with a specific mutation in codon 249
of p53 tumor suppressor gene - Potential multiplicative interaction with
Hepatitis B viral infection (HBV)
8BackgroundMetabolism of aflatoxin B1
9Background
- Other risk factors
- Hepatitis C virus
- Alcohol consumption
- Tobacco smoking
- Contaminated drinking water (microcystins)
- Potential protective factors
- Antioxidants
- Dietary nutrients
- Selenium
10Background IFNA17
- Located at position 9p22
- Encodes interferon, alpha 17
- Has viral inhibitory and viral anti-proliferative
effects - Interferon, alpha investigated as a treatment for
HBV and HCV infection and prevention of liver
cancer among HBV infected individuals - IFNA17 has a polymorphic site that results in
either an arginine or isoleucine amino acid in
codon 184 - The 184Arg allele has a higher frequency in
Chinese populations (50) than in other
populations (lt35) - Studied in cancers that have a viral component,
including cervical cancer (papillomavirus) and
nasopharyngeal cancer (Epstein-Barr virus)
11Objectives
- Overall objective
- Gain insight into the mechanism of interaction
observed between aflatoxin and HBV infection - Hypothesis
- AFB1 may differentially suppress IFNA17 protein
activity, thereby increasing a persons
susceptibility to the sequelae of HBV if
chronically infected and increasing the risk for
liver cancer - Specific Aims
- Assess the independent effects of IFNA17
Ile184Arg polymorphic site on liver cancer risk - Explore the joint effects and three-way
interaction among HBsAg-positivity, AFB1-albumin
adduct level, and IFNA17 polymorphisms on the
development of liver cancer
12Research Design and Methods
13Study Design
- Population-based case-control study with 204
incident liver cancer cases (57 of all cases)
and 415 randomly selected healthy population
controls (89 response rate) in Taixing City,
China - Collected epidemiologic data and blood specimens
- IFNA17 genotyped using PCR-RFLP
- Markers used to assess HBV chronic infection and
aflatoxin B1 exposure (and HCV infection)
14Taixing City, China
- Located on the east bank of the Yangtze River in
middle of Jiangsu province - Consists of 24 small villages, with estimated
population of 1.28 million (660,000 males and
627,000 females), most of whom are farmers - Has a high rate of alimentary cancer, among the
highest in the world - Has a tumor registry
- After esophageal cancer, liver cancer is the
second largest cause of deaths from cancers - In 2000, crude incidence rates for top three
cancers - 65.2 (esophageal), 55.6 (liver), 54.8
(stomach) per 100,000 - Incidence rate of liver cancer for males
(84.6/100,000) is over three times the rate for
females (25.0/100,000)
15Taixing City, China
Taixing City
16Exposure Marker HBV
- Different markers used to assess extent of
infection with HBV - Detection of HBV surface antigen (HBsAg) used to
assess new and chronic infections - Enzyme-linked immunosorbant assay (ELISA) used to
detect HBsAg in serum
17(No Transcript)
18Exposure Marker Aflatoxin B1
- AFB1-exo-8,9-epoxide intermediate binds to DNA
and proteins - Aflatoxin-albumin adduct detection to assess
aflatoxin exposure - Competitive ELISA used to detect
aflatoxin-albumin adducts in plasma
19Statistical Analysis Model
- Unconditional logistic regression model
- Complete analysis was used and missing data for
independent variables were not imputed - For potential confounders that were missing a
large amount of data (gt10), like BMI, we imputed
the median of controls by sex - When aflatoxin B1 (10 missing) was included as
a potential confounder in a model, missing data
was imputed using the median of controls
20Results
21Demographic data Average (SD) age, BMI, and
smoking pack-years of cases and controls
22Demographic data Gender, education, and alcohol
consumption
23Main Effects of HBsAg, AFB1 levels, and IFNA17 on
liver cancer development
Variables Variables Case Case Control Control Crude Age Sex Adjusted Fully Adjusted
N () N () N () N () OR (95CI) OR (95CI) OR (95CI)
HBsAg - 72 (35.3) 312 (75.4) 1 1 1
132 (64.7) 102 (24.6) 5.61 (3.90-8.07) 5.21 (3.60-7.53) 5.68 (3.80-8.51)
AFB1 Mean (SD) 508.1 (328.7) 426.2 (250.4)
lt247 33 (18.1) 94 (24.9) 1 1 1
247.1-388.8 46 (25.3) 94 (24.9) 1.39 (0.82-2.37) 1.38 (0.81-2.37) 1.15 (0.61-2.14)
388.9-545 42 (23.1) 95 (25.2) 1.26 (0.74-2.16) 1.27 (0.74-2.20) 1.19 (0.64-2.21)
gt545.1 61 (33.5) 94 (24.9) 1.85 (1.11-3.08) 1.75 (1.04-2.94) 1.63 (0.90-2.96)
p(trend)0.031 p(trend)0.055 p(trend)0.109
IFNA17 II 33 (17.4) 94 (24.5) 1 1 1
RI 104 (54.7) 193 (50.4) 1.54 (0.97-2.44) 1.49 (0.93-2.38) 1.67 (0.95-2.93)
RR 53 (27.9) 96 (25.1) 1.57 (0.94-2.64) 1.58 (0.93-2.68) 1.99 (1.06-3.73)
p(HW)0.878 p(HW)0.878 p(trend)0.104 p(trend)0.102 p(trend)0.037
RIRR 157 (82.6) 289 (75.5) 1.55 (1.00-2.41) 1.52 (0.97-2.38) 1.77 (1.04-3.03)
Model includes age, sex, BMI, education,
alcohol consumption, tobacco smoking, HBsAg,
imputed AFB1 levels, anti-HCV
24Interaction between HBV and AFB1 and IFNA17
HBsAg HBsAg HBsAg Case Case Control Control Crude Age Sex Adjusted Fully Adjusted
N () N () N () N () OR (95CI) OR (95CI) OR (95CI)
AFB1 AFB1 AFB1
lt247 - - 12 (6.6) 69 (18.4) 1 1 1
247.1-388.8 - - 19 (10.4) 67 (17.8) 1.63 (0.74-3.62) 1.64 (0.73-3.65) 1.72 (0.73-4.08)
388.9-545 - - 15 (8.2) 71 (18.9) 1.22 (0.53-2.78) 1.22 (0.53-2.80) 1.34 (0.55-3.27)
gt545.1 - - 17 (9.3) 77 (20.5) 1.27 (0.57-2.85) 1.26 (0.56-2.82) 1.15 (0.48-2.74)
lt247 21 (11.5) 25 (6.6) 4.83 (2.08-11.23) 4.61 (1.97-10.80) 6.43 (2.56-16.16)
247.1-388.8 27 (14.8) 27 (7.2) 5.75 (2.55-12.96) 5.30 (2.34-12.02) 4.68 (1.92-11.38)
388.9-545 27 (14.8) 24 (6.4) 6.47 (2.84-14.74) 6.20 (2.70-14.21) 6.65 (2.72-16.25)
gt545.1 44 (24.2) 16 (4.3) 15.82 (6.84-36.57) 13.75 (5.90-32.06) 16.72 (6.60-42.38)
1ORint (95CI) 1ORint (95CI) 1ORint (95CI) 0.73 (0.24-2.24) 0.70 (0.23-2.18) 0.42 (0.12-1.45)
2ORint (95CI) 2ORint (95CI) 2ORint (95CI) 1.10 (0.35-3.49) 1.10 (.35-3.52) 0.77 (0.22-2.70)
3ORint (95CI) 3ORint (95CI) 3ORint (95CI) 2.58 (0.82-8.12) 2.38 (0.75-7.55) 2.27 (0.65-7.92)
IFNA17 IFNA17 IFNA17 IFNA17 IFNA17
II II II - 13 (6.8) 66 (17.3) 1 1 1
RIRR RIRR RIRR - 50 (26.3) 220 (57.6) 1.15 (0.59-2.25) 1.14 (0.58-2.23) 1.34 (0.64-2.82)
II II II 20 (10.5) 27 (7.1) 3.76 (1.64-8.62) 3.49 (1.51-8.04) 3.99 (1.54-10.32)
RIRR RIRR RIRR 107 (56.3) 69 (18.1) 7.87 (4.04-15.34) 7.17 (3.66-14.06) 9.18 (4.34-19.43)
ORint (95CI) ORint (95CI) ORint (95CI) 1.81 (0.71-4.62) 1.81 (0.71-4.63) 1.71 (0.60-4.92)
Model includes age, sex, BMI, education,
alcohol consumption, tobacco smoking, imputed
AFB1 levels, anti-HCV 1ORint for AFB1
(247.1-388.8 fmol/mg) and HBsAg 2ORint for AFB1
(388.9-545 fmol/mg) and HBsAg 3ORint for AFB1
gt545.1 fmol/mg) and HBsAg
25Effects of IFNA17 stratified by HBsAg
HBsAg Crude Age Sex Adjusted Fully Adjusted
OR (95CI) OR (95CI) OR (95CI)
IFNA17 - 1.15 (0.59-2.25) 1.11 (0.57-2.18) 1.35 (0.63-2.85)
(RIRR vs. II) 2.09 (1.09-4.02) 2.08 (1.06-4.08) 2.19 (1.01-4.76)
Model includes age, sex, BMI, education,
alcohol consumption, tobacco smoking, imputed
AFB1 levels, anti-HCV
Effects of HBV and IFNA17 stratified by AFB1
AFB1 Crude Age Sex Adjusted Fully Adjusted
OR (95CI) OR (95CI) OR (95CI)
HBsAg lt247 4.83 (2.08-11.23) 4.72 (2.02-11.05) 7.65 (2.82-20.77)
(Pos. vs. Neg.) 247.1-388.8 3.53 (1.69-7.37) 3.14 (1.42-6.96) 2.77 (1.16-6.66)
388.9-545 5.33 (2.44-11.65) 5.27 (2.38-11.67) 5.89 (2.38-14.60)
gt545.1 12.46 (5.73-27.08) 12.24 (5.42-27.63) 18.34 (7.02-47.92)
IFNA17 lt247 0.52 (0.20-1.34) 0.55 (0.21-1.44) 0.26 (0.07-0.92)
(RIRR vs. II) 247.1-388.8 1.62 (0.66-3.99) 1.21 (0.45-3.21) 2.85 (0.82-9.96)
388.9-545 3.27 (1.05-10.19) 3.11 (0.99-9.84) 5.89 (1.16-29.87)
gt545.1 1.19 (0.53-2.67) 1.19 (0.53-2.70) 1.42 (0.43-4.71)
Model includes age, sex, BMI, education,
alcohol consumption, tobacco smoking, HBsAg,
anti-HCV
26Interaction between HBsAg and IFNA17 stratified
by AFB1
AFB1 HBsAg IFNA17 Case Control Crude Age Sex Adjusted Fully Adjusted
N N OR (95CI) OR (95CI) OR (95CI)
lt388.9 - II 8 26 1 1 1
- RIRR 20 99 0.66 (0.26-1.66) 0.63 (0.24-1.62) 0.70 (0.24
II 9 13 2.25 (0.70-7.19) 2.04 (0.62-6.74) 2.07 (0.52-8.18)
RIRR 37 37 3.25 (1.30-8.11) 2.81 (1.10-7.19) 3.45 (1.21-9.83)
ORint (95CI) ORint (95CI) ORint (95CI) 2.20 (0.58-8.38) 2.20 (0.56-8.70) 2.39 (0.50-11.45)
gt388.9 - II 5 34 1 1 1
- RIRR 25 104 1.63 (0.58-4.60) 1.62 (0.58-4.59) 2.09 (0.64-6.86)
II 11 9 8.31 (2.29-30.10) 8.07 (2.21-29.42) 9.22 (2.08-40.86)
RIRR 57 27 14.35 (5.05-40.77) 13.88 (4.80-40.09) 21.80 (6.36-74.75)
ORint (95CI) ORint (95CI) ORint (95CI) 1.06 (0.25-4.44) 1.06 (0.25-4.45) 1.13 (0.22-5.81)
Model includes age, sex, BMI, education,
alcohol consumption, tobacco smoking, HCV
27Discussion
28Findings summarized
- Main Effects
- Strong association between liver cancer and HBsAg
- Moderate association between liver cancer and
aflatoxin B1 and IFNA17 R allele - Possible Interactive Effects
- HBV-AFB1
- HBV-IFNA17
- AFB1-IFNA17
- HBV-AFB1-IFNA17
29IFNA17
- Genotype frequencies are similar to previous
studies in Chinese populations - No previous studies have evaluated association
between IFNA17 and liver cancer - 184Ile allele, the lower-risk allele for our
study, was previously found to be positively
associated with cervical and nasopharyngeal
cancers - Difference in risk alleles may be due to
differences in their specific activities, i.e.
Ile protein product may have more antiviral
activity against HBV/HCV, whereas Arg protein
product may have more against human
papillomavirus and Epstein-Barr virus
30IFNA17
- Positive results may be due to
- 1. False positive
- 2. Direct functional involvement with HCC
development - 3. Linkage disequilibrium with a nearby risk gene
(like IFNA10 or p16)
31Interpretation of HBV, AFB1, and IFNA17 Joint
Effects
- AFB1 may negatively interact with IFNA17, leading
to a differential decrease in protein function,
resulting in a decreased resistance against HBV
and increasing risk for the development of liver
cancer
32Limitations
- Recall bias
- Subjects awareness of disease status may alter
recall of past exposures - Selection bias
- Selection of patients with less advanced and
aggressive cancers - Reporting bias
- Behaviors or habits carry social stigma, like
smoking and alcohol drinking - Confounding by indication
- Since blood samples were collected after
diagnosis, cases may have altered their diet to
contain less AFB1
33Strengths
- Population-based study design
- controls were randomly selected from base
population from which cases arose - Relatively large sample size
- Detailed and extensive questionnaire
- dietary habits, smoking, alcohol
- Racially homogeneous population
- race would not be a potential confounder or
effect modifier
34Significance Public Health Applications
- The associations and joint effects for IFNA17
have never been previously studied in liver
cancer - Early detection of liver cancer and
identification of high-risk individuals for
intervention - Prevention strategies
- HBV vaccine
- Control intake of foods that typically have
higher levels of AFB1 and modify storage
condition of food to prevent mold growth - Chemoprevention of liver cancer
- Interferon mixtures are currently under study to
prevent liver cancer among HBV chronic carriers - Oltipraz protects against AFB1-induced liver
cancers by inhibiting phase I enzymes and
increasing phase II enzymes
35Acknowledgements
- Laboratories at UCLA
- Dr. Steven Dubinett
- Dr. Robert Lehrer
- Dr. John Timmerman
- Dr. Gang Zeng
- Collaborators
- Dr. Zuo-Feng Zhang
- Dr. Regina Santella
- Dr. Li-Na Mu
- Dr. Shun-Zhang Yu
- Dr. Qing-Wu Jiang
- Dr. Wei Cao
- Dr. Xue-Fu Zhou
- Dr. Bao-Guo Ding
- Dr. Ru-Hong Wang
- Dr. Jinkou Zhou
- Dr. Lin Cai
- Mr. John Garcia