| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
Division of Cancer Epidemiology and Genetics, National Cancer Institute, Bethesda, Maryland 20892 [R. S., C. A. S., M. C. R. A.]; Division of Biostatistics, Department of Community Medicine and Health Care, University of Connecticut School of Medicine, Farmington, Connecticut 06030 [M. K.]; Information Management Services Inc., Silver Spring, Maryland 20904 [J. C.]; and Department of Community Health and Prevention Research Center, School of Public Health, Saint Louis University, St. Louis, Missouri 63103 [R. C. B.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We investigated the role of three HCAs in the etiology of lung cancer. We estimated daily ingestion of three HCA compounds, DiMeIQx, MeIQx, and PhIP, among Missouri women participating in a population-based case-control study of lung cancer. Exposure levels were estimated using a newly created HCA database (2–4) in combination with a FFQ. We also investigated the effect of HCAs by smoking category and histological cell type of lung cancer.
|
PATIENTS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Controls between ages 30 and 64 years were randomly selected from driver’s license files identified through the Missouri Driver’s License Registry. Controls between the ages of 65–84 years were randomly selected from lists provided by the Health Care Finance Administration. A two-stage randomized recruitment was used to avoid the expected imbalance in smoking among cases as compared with controls (8, 9) . From the 3386 controls who were found to be eligible, 730 subjects were selected, 700 completed an interview, and 628 (86% of the 730 eligible controls) provided information on diet, meat cooking practices, and relevant potential confounders. Five control subjects were excluded due to implausible dietary information or missing dietary information, resulting in 623 controls (85% of 730 eligible controls).
FFQ, Cooking Methods, and HCA Database.
We used a modified version of the 100-item Health Habits and History
Questionnaire to obtain information on usual diet (frequency of
consumption and portion size) approximately 2–3 years before diagnosis
(1, 7)
. The first of two sections obtained information on
consumption frequency and portion size for 22 meat and fish items and
other foods including fruits, vegetables, cereals, and grains. The
second portion of the questionnaire focused on methods of cooking
(i.e., pan-fried, oven-broiled, grilled/barbecued,
microwaved, or other) and doneness levels (rare, medium-rare,
medium-well, well-done, and very well-done).
We have developed a HCA database (2–4, 10) for three HCA compounds: (a) MeIQx; (b) DiMeIQx; and (c) PhIP. Using the responses from the FFQ, we estimated gram consumption of the meat groups using frequency and portion size by cooking technique and doneness level. HCA consumption was then derived by multiplying grams of the meat type (doneness level and cooking method) by the HCA concentration measured in that meat. The HCA concentration was summed across all meat items in the diet to estimate total intake for individual study subjects. For subjects with missing data on cooking practices of an individual meat item, the doneness level was imputed using the median value among controls, whereas the cooking method was imputed using the most common choice among controls. No imputations were made for 86% of the subjects, 11% had one imputation, 2% had two imputations, and less than 1% had more than two imputations.
The tumor slides were examined simultaneously using multiheaded microscopes by three pathologists who were blinded regarding the referring pathologists’ diagnoses. Consensus diagnoses were obtained for surgical specimens with the criteria outlined in the WHO classification scheme (11) . When only cytological material was available, a consensus was obtained with cytological criteria (12) .
Analyses.
ORs were computed separately for dietary DiMeIQx, MeIQx, and PhIP using
both categorical and continuous data. For the latter, the ORs were
calculated using the fitted logistic regression (13)
taking the ratio of the estimated odds at two different consumption
levels, the medians of the fifth and first quintiles (i.e.,
the 90th and 10th
percentiles) based on the controls. We checked for nonlinearity by
adding a quadratic term to the continuous model. In no case was the
quadratic term statistically significant; thus, it was left out of the
final logistic regression model. This does not necessarily mean that
the true relation is linear, but any existing nonlinearity is not
strong enough to be determined with the current sample size. The 95%
CIs are given for the ORs. To test for trend, we determined whether the
fitted logistic regression parameter was significantly different from
0. The trend is significant at the 0.05 level, if the 95% CI did not
contain 1. The magnitude of the ORs based on quintiles or percentiles
cannot be directly compared to one another because the levels of
ingestion in the upper and lower quintiles are different for various
subsets of HCAs (1)
. Therefore, we present a second set of
ORs reflecting the relative risk associated with an increased
consumption of 10 ng HCA/day.
These models included age, pack-years of smoking, smoking status (never/ever/current), years-quit-smoking, BMI, calories, fat, fruit/fruit juices, and vegetables using continuous variables and education using a categorical variable (<12 years, 12 years, and >12 years). In addition to these variables, subsequent analysis for lung cancer risk was further controlled for the effect of cooking practice and doneness level. This analysis was performed to assess whether the excess risk associated with a specific HCA could be explained by meat cooking method and doneness level. We also examined the role of HCAs on the risk stratified by smoking status and for specific histological types (squamous cell carcinomas, small cell carcinomas, adenocarcinomas, and others types) of lung cancer.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Considerable variability in dietary HCA concentrations was observed
for all compounds, as seen in the HCA concentrations at the
10th versus 90th
percentile.
|
, age-adjusted ORs, multivariate-adjusted ORs, and multivariate analysis
accounting for all three HCAs simultaneously are shown both for 10
ng/day increments of HCAs in the diet and for comparisons of lung
cancer risk at the 90th and
10th percentiles of HCA intake. After
multivariate adjustments were made for the ORs comparing the
90th and 10th percentiles
of ingested HCAs, risk was significantly elevated for MeIQx (OR, 1.5;
CI, 1.1–2.0), but not for DiMeIQx (OR, 1.2; CI, 0.9–1.6) or PhIP (OR,
0.9; CI, 0.8–1.0). Across quintiles of MeIQx consumption, the ORs were
0.9 (CI, 0.6–1.3), 1.1 (CI, 0.7–1.7), 1.2 (CI, 0.8–1.8), and 1.4
(CI, 0.9–2.1) for the second, third, fourth, and fifth
quintiles, respectively. The test for trend was P = 0.006 (Table 2)
.
|
. A 23% excess risk was associated with a 50
ng/day increase in MeIQx. A 50% excess lung cancer risk was observed
for a 100 ng/day increase in MeIQx. Moreover, the OR for MeIQx remained
elevated when further adjusted with DiMeIQx and PhIP. Neither dietary
DiMeIQx nor PhIP was associated with significant excess lung cancer
risk in the adjusted models. We examined the association of lung cancer risk with MeIQx ingestion, controlling for the effects of the baseline covariates as well as meat groups (i.e., total meats, red meat, red meat doneness, fried and nonfried red meat, grilled meat, and smoked meat) to investigate whether MeIQx is associated with lung cancer above and beyond what can be explained by the different meat variables. Comparing the 90th to 10th percentile of MeIQx ingestion, statistically significant elevations in risk continue to be observed when controlling for total meat (OR, 1.4; CI, 1.0–1.9; P = 0.03), red meat (OR, 1.4; CI, 1.0–1.9; P = 0.05), barbecued red meat (OR, 1.5; CI, 1.1–2.0; P = 0.007), and smoked meat (OR, 1.5; 1.1–2.0; P = 0.009). The association of lung cancer and MeIQx was still elevated but was not statistically significant when controlled for intake of well-done red meat (OR, 1.3; CI, 0.9–1.9; P = 0.13) or fried red meat (OR, 1.3; CI, 0.9–1.8; P = 0.19). In contrast, the risk associated with well-done meat and fried meat disappeared when adjusted for MeIQx but not when adjusted for DiMeIQx and PhIP.
The correlation of cigarette smoking and MeIQx and PhIP intake were very small. The nonparametric Spearman correlation coefficients are 0.10 for packyears and MeIQx; 0.02 for packyears and the PhIP; -0.13 for years-quit-smoking and MeIQx; and 0.01 for years-quit-smoking and PhIP. Furthermore, there was no association of cigarette smoking and various meat cooking techniques (7) .
When stratifying the subjects based on pack-years of smoking (0,
0.1–39.9, or 40+ pack-years), higher MeIQx consumption was associated
with increased risk of lung cancer for nonsmokers (OR, 3.6; CI,
1.3–10.3; P = 0.015) and light/moderate
smokers (OR, 2.1; CI, 1.3–3.3; P = 0.003),
but not for heavy smokers [OR, 1.0; CI, 0.7–1.5;
P = 0.97 (Table 3
)]. Looking at the interaction, the difference in MeIQx ORs for the
different strata was statistically significant for the nonsmokers
versus heavy smokers (P = 0.04),
but not for the light smokers versus the heavy smokers
(P = 0.09).
|
. There was
elevated risk with MeIQx intake in subjects with squamous cell
carcinomas (OR, 1.9; CI, 1.2–3.1) and "other" histological cell
type (OR, 1.6; CI, 1.1–2.5). There was no statistically significant
risk associated with small cell carcinomas and adenocarcinomas. Neither
DiMeIQx nor PhIP showed associations with any histological type of lung
cancer. |
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In an earlier analysis of dietary data from this study (1) , consumption of well-done red meat and consumption of fried red meat were observed to be risk factors for lung cancer. In the analysis presented here, we investigated whether specific HCAs found in well-done red meats could explain this association. Our data support the hypothesis that MeIQx, but not DiMeIQx or PhIP, may be part of the reason for the elevated risk of lung cancer observed from meat consumption. There is also evidence from animal carcinogenesis studies (5, 6) that MeIQx may be a lung carcinogen.
The correlation between HCA intake and the amount of meat consumed makes it difficult to separate an effect due to a specific HCA versus that of meat or meat cooking practices in general without a very large sample size. However, they do not measure the same thing. For example, well-done pan-fried steak contains 4.1 ng of MeIQx per gram, whereas well-done oven-broiled hamburger patty does not contain any measurable amount of MeIQx. When we controlled for various correlates, such as total meat, red meat, barbecued meat, or smoked meats, the statistically significant lung cancer risk associated with MeIQx remained. The MeIQx effect was still present but was no longer statistically significantly associated with lung cancer risk when the model included well-done red meat or fried red meat. The opposite was also true; the previously significant effects of well-done and fried red meat were no longer significant after adjusting for MeIQx.
It is interesting that we observe increased risk due to MeIQx intake only in nonsmokers and light/moderate smokers and not in heavy smokers. Cigarette smoking is a major risk factor for lung cancer that it is likely to overwhelm the risk associated with MeIQx.
This is the first effort to assess the role that specific HCAs play in the etiology of human lung cancer. However, the limitations of our study should be recognized. Because the current HCA database is limited to a fraction of the compounds associated with high-temperature cooking of meats, the excess risk observed here may be explained by an unidentified meat pyrolysis byproduct that is strongly associated with MeIQx. We also recognize that the evidence from the literature for the relationship between meat consumption and lung cancer is equivocal. Both cohort and case-control studies have examined the role of meat intake in lung cancer (1, 14–25) . A few studies have examined the relationship between meat cooking methods and the risk of lung cancer (26, 27) . In a Swedish study (26) , lung cancer risk was not associated with consumption of meat or fish cooked and preserved by various methods. In contrast, a study from Uruguay (27) found results compatible with ours, showing an increase in lung cancer risk with higher consumption of fried meat.
The biological relevance of our observation that MeIQx is associated with a significant excess risk of squamous cell carcinoma and "other" cell types but not of small cell or adenocarcinoma is unclear because the link between exposure and histological type of lung cancer remains uncertain. Some studies have observed an increase in risk of squamous cancer with red meat, fried meat, and smoked meat and poultry consumption (22, 27) . Evidence suggests for other exposures, such as smoking, there is a stronger risk factor for squamous and small cell carcinomas (28) and a weaker risk factor for adenocarcinomas, whereas adenocarcinoma of the lung has been associated with asbestos and pesticide exposure (15, 29) .
In conclusion, even after adjusting for smoking, we found evidence of increased lung cancer risk among the high consumers of MeIQx, but not among consumers of DiMeIQx and PhIP. The increased risk of lung cancer with MeIQx consumption may be associated with smoking status and certain histological types of lung cancer. However, these preliminary findings need to be replicated.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 To whom requests for reprints should be
addressed, at Division of Cancer Epidemiology and Genetics, National
Cancer Institute, 6120 Executive Boulevard, EPS Room 7028, Bethesda, MD
20892. Phone: (301) 496-6426; Fax: (301) 496-6829; E-mail: sinhar{at}nih.gov 
2 The abbreviations used are: HCA, heterocyclic
amine; MeIQx,
2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline;
DiMeIQx,
2-amino-3,4,8-trimethylimidazo[4,5-f]quinoxaline;
PhIP, 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine;
FFQ, food frequency questionnaire; OR, odds ratio; CI, confidence
interval.
Received 11/15/99. Accepted 5/17/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Bendaly, S. Zhao, J. R. Neale, K. J. Metry, M. A. Doll, J. C. States, W. M. Pierce Jr., and D. W. Hein 2-Amino-3,8-Dimethylimidazo-[4,5-f]Quinoxaline-Induced DNA Adduct Formation and Mutagenesis in DNA Repair-Deficient Chinese Hamster Ovary Cells Expressing Human Cytochrome P4501A1 and Rapid or Slow Acetylator N-Acetyltransferase 2 Cancer Epidemiol. Biomarkers Prev., July 1, 2007; 16(7): 1503 - 1509. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Tang, J. J. Liu, A. Rundle, C. Neslund-Dudas, A. T. Savera, C. H. Bock, N. L. Nock, J. J. Yang, and B. A. Rybicki Grilled Meat Consumption and PhIP-DNA Adducts in Prostate Carcinogenesis Cancer Epidemiol. Biomarkers Prev., April 1, 2007; 16(4): 803 - 808. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. J. Cross, U. Peters, V. A. Kirsh, G. L. Andriole, D. Reding, R. B. Hayes, and R. Sinha A Prospective Study of Meat and Meat Mutagens and Prostate Cancer Risk Cancer Res., December 15, 2005; 65(24): 11779 - 11784. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Kikugawa Prevention of mutagen formation in heated meats and model systems Mutagenesis, November 1, 2004; 19(6): 431 - 439. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Lippman and M. R. Spitz Lung Cancer Chemoprevention: An Integrated Approach J. Clin. Oncol., September 15, 2001; 19(90001): 74s - 82. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Brown, B. E. Hingerty, E. A. Guenther, V. V. Krishnan, S. Broyde, K. W. Turteltaub, and M. Cosman Solution structure of the 2-amino-1- methyl-6-phenylimidazo[4,5-b]pyridine C8-deoxyguanosine adduct in duplex DNA PNAS, June 28, 2001; (2001) 151251898. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Bardelli, D. P. Cahill, G. Lederer, M. R. Speicher, K. W. Kinzler, B. Vogelstein, and C. Lengauer Carcinogen-specific induction of genetic instability PNAS, April 5, 2001; (2001) 81082898. [Abstract] [Full Text]
|
|
|
|
|
|
A. Bardelli, D. P. Cahill, G. Lederer, M. R. Speicher, K. W. Kinzler, B. Vogelstein, and C. Lengauer From the Cover: Carcinogen-specific induction of genetic instability PNAS, May 8, 2001; 98(10): 5770 - 5775. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Brown, B. E. Hingerty, E. A. Guenther, V. V. Krishnan, S. Broyde, K. W. Turteltaub, and M. Cosman Solution structure of the 2-amino-1- methyl-6-phenylimidazo[4,5-b]pyridine C8-deoxyguanosine adduct in duplex DNA PNAS, July 17, 2001; 98(15): 8507 - 8512. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
