2.50
Hdl Handle:
http://hdl.handle.net/10146/17399
Title:
Correlations among biomarkers.
Authors:
Gyorffy, Erika; Anna, Livia; Rudnai, Peter; Kovacs, Katalin; Schocket, Bernadette
Abstract:
An extensive literature survey on correlation of biomarkers resulted in the following preliminary conclusions: • No significant correlation between individual pairs of DNA adducts was found in a number of the 32P-postlabelling and immunoassay studies, indicating limited overlapping of the substrate specificity of the different DNA adduct methods. • Attempts to show correlations between a group of related DNA adduct structures and a chemically specific single DNA adduct structure by using the same type of methodology gave controversial results. These suggest the co-existence of both closely linked and independent metabolic activation pathways for complex mixtures of xenobiotics, and may also reflect differences in the kinetics of DNA adduct formation and elimination. • A larger number of studies revealed a positive correlation between DNA adduct levels in target and surrogate tissues than did not find a correlation. Correlation may depend on exposure dose and the metabolic capacity of the corresponding tissues. • In the majority of the studies there was a positive correlation between DNA adduct levels in tumour and normal tissues, suggesting similarities in the xenobiotic activation/elimination processes of the tumour and normal tissues. However, the levels of DNA adducts found suggested organ specificity. • There was a positive correlation for different urinary metabolites and urinary mutagenicity in most of the studies. • No correlation between DNA adducts and urinary polycyclic aromatic hydrocarbon metabolites was generally found, but stratification by genotype suggested that correlation might be present. • Correlation was more probable between structurally specified protein adducts and urinary polycyclic aromatic hydrocarbon metabolite 1-hydroxypyrene than with less specific xenobiotic-protein structures. • Stratification of the study population for confounding factors, such as smoking status, may reveal hidden correlations. • Cytogenetic biomarker studies also gave complex results. Examples of both positive correlation and lack of correlation with exposure markers were found. Molecular epidemiological studies of cancer that include exposure biomarkers have greatly increased in number in recent years, the rationale being to measure the biologically relevant aspect(s) of exposure. However, the use of biomarkers to measure exposure is not a panacea. Most biomarker-based studies, of both prospective and retrospective design, rely on a single biological sample. Most exposures in cancer epidemiology are time-related variables, and both carcinogenesis models and empirical evidence strongly point towards the importance of induction and latency periods in cancer onset, the need to separate the role of duration and intensity of exposure, and the decrease in effect after cessation of exposure. While most biomarkers measure recent exposure, it is possible to apply them in the measurement of temporal changes, e.g. by collecting repeated samples from subjects enrolled in prospective studies or from a sample of the original cohort. Another important parameter is the in vivo lifetime of the biomarker after it has been generated by a carcinogen exposure. Biomarker-based epidemiological results are subject to the methodological problems affecting all types of observational studies, namely bias and confounding. A further problem in the measurement of exposure biomarkers in retrospective studies is their possible dependence on the disease process. With the increasing use of biomarkers of dose and effect of carcinogens and the possible input of these data into the regulatory area, it is essential that the methodology be standardised and wherever possible internationally accepted protocols be established for this. The need to validate exposure biomarkers before their application in population-based studies arises from the variability in biomarker-based measurements of exposure, which can be due to interindividual sources (e.g. differences between exposed and unexposed individuals, usually the component of variability of primary interest), intraindividual sources (e.g. variability in hormonal levels) and observer sources, including measurement error. This is the domain of so-called transitional studies, which aim to characterise the biomarker itself rather than the underlying biological phenomenon. This is the area in which most work is needed in the near future: efforts such as the ECNIS Network of Excellence will be instrumental in providing a logical framework for the development and validation of biomarkers of exposure with the ultimate goal of their application in molecular epidemiological studies. • There was a positive correlation for different urinary metabolites and urinary mutagenicity in most of the studies. • No correlation between DNA adducts and urinary polycyclic aromatic hydrocarbon metabolites was generally found, but stratification by genotype suggested that correlation might be present. • Correlation was more probable between structurally specified protein adducts and urinary polycyclic aromatic hydrocarbon metabolite 1-hydroxypyrene than with less specific xenobiotic-protein structures. • Stratification of the study population for confounding factors, such as smoking status, may reveal hidden correlations. • Cytogenetic biomarker studies also gave complex results. Examples of both positive correlation and lack of correlation with exposure markers were found. Molecular epidemiological studies of cancer that include exposure biomarkers have greatly increased in number in recent years, the rationale being to measure the biologically relevant aspect(s) of exposure. However, the use of biomarkers to measure exposure is not a panacea. Most biomarker-based studies, of both prospective and retrospective design, rely on a single biological sample. Most exposures in cancer epidemiology are time-related variables, and both carcinogenesis models and empirical evidence strongly point towards the importance of induction and latency periods in cancer onset, the need to separate the role of duration and intensity of exposure, and the decrease in effect after cessation of exposure. While most biomarkers measure recent exposure, it is possible to apply them in the measurement of temporal changes, e.g. by collecting repeated samples from subjects enrolled in prospective studies or from a sample of the original cohort. Another important parameter is the in vivo lifetime of the biomarker after it has been generated by a carcinogen exposure. Biomarker-based epidemiological results are subject to the methodological problems affecting all types of observational studies, namely bias and confounding. A further problem in the measurement of exposure biomarkers in retrospective studies is their possible dependence on the disease process. With the increasing use of biomarkers of dose and effect of carcinogens and the possible input of these data into the regulatory area, it is essential that the methodology be standardised and wherever possible internationally accepted protocols be established for this. The need to validate exposure biomarkers before their application in population-based studies arises from the variability in biomarker-based measurements of exposure, which can be due to interindividual sources (e.g. differences between exposed and unexposed individuals, usually the component of variability of primary interest), intraindividual sources (e.g. variability in hormonal levels) and observer sources, including measurement error. This is the domain of so-called transitional studies, which aim to characterise the biomarker itself rather than the underlying biological phenomenon. This is the area in which most work is needed in the near future: efforts such as the ECNIS Network of Excellence will be instrumental in providing a logical framework for the development and validation of biomarkers of exposure with the ultimate goal of their application in molecular epidemiological studies.
Citation:
In: Biomarkers of carcinogen exposure and early effects. Ed. Peter B. Farmer, Jean M. Emeny. Lodz 2006, p. 143-159.
Publisher:
The Nofer Institute of Occupational Medicine
Issue Date:
2006
URI:
http://hdl.handle.net/10146/17399
Additional Links:
http://www.ecnis.org/images/stories/ecnis/documents/reports/Biomarkers/6.correlations...%20143-159.pdf
Type:
Book chapter
Language:
en
Description:
6.1. Introduction 6.2. Study populations and types of exposure 6.3. Most frequent biomarkers used in the studies evaluated here 6.4. Correlation between levels of DNA adducts in human samples from a methodological point of view 6.5. Correlation between levels of the same DNA adduct in various tissues 6.6. Correlation between different biomarkers of human genotoxic exposure 6.7. Correlations among multiple biomarkers monitoring genotoxic exposure and effect 6.8. Preliminary conclusions of the literature review
Series/Report no.:
ECNIS Report; 1
ISBN:
83-88261-78-9; 978-83-88261-78-1
Sponsors:
ECNIS Network of Excellence
Appears in Collections:
Chapters

Full metadata record

DC FieldValue Language
dc.contributor.authorGyorffy, Erika-
dc.contributor.authorAnna, Livia-
dc.contributor.authorRudnai, Peter-
dc.contributor.authorKovacs, Katalin-
dc.contributor.authorSchocket, Bernadette-
dc.date.accessioned2008-02-04T10:14:15Z-
dc.date.available2008-02-04T10:14:15Z-
dc.date.issued2006-
dc.identifier.citationIn: Biomarkers of carcinogen exposure and early effects. Ed. Peter B. Farmer, Jean M. Emeny. Lodz 2006, p. 143-159.en
dc.identifier.isbn83-88261-78-9-
dc.identifier.isbn978-83-88261-78-1-
dc.identifier.urihttp://hdl.handle.net/10146/17399-
dc.description6.1. Introduction 6.2. Study populations and types of exposure 6.3. Most frequent biomarkers used in the studies evaluated here 6.4. Correlation between levels of DNA adducts in human samples from a methodological point of view 6.5. Correlation between levels of the same DNA adduct in various tissues 6.6. Correlation between different biomarkers of human genotoxic exposure 6.7. Correlations among multiple biomarkers monitoring genotoxic exposure and effect 6.8. Preliminary conclusions of the literature reviewen
dc.description.abstractAn extensive literature survey on correlation of biomarkers resulted in the following preliminary conclusions: • No significant correlation between individual pairs of DNA adducts was found in a number of the 32P-postlabelling and immunoassay studies, indicating limited overlapping of the substrate specificity of the different DNA adduct methods. • Attempts to show correlations between a group of related DNA adduct structures and a chemically specific single DNA adduct structure by using the same type of methodology gave controversial results. These suggest the co-existence of both closely linked and independent metabolic activation pathways for complex mixtures of xenobiotics, and may also reflect differences in the kinetics of DNA adduct formation and elimination. • A larger number of studies revealed a positive correlation between DNA adduct levels in target and surrogate tissues than did not find a correlation. Correlation may depend on exposure dose and the metabolic capacity of the corresponding tissues. • In the majority of the studies there was a positive correlation between DNA adduct levels in tumour and normal tissues, suggesting similarities in the xenobiotic activation/elimination processes of the tumour and normal tissues. However, the levels of DNA adducts found suggested organ specificity. • There was a positive correlation for different urinary metabolites and urinary mutagenicity in most of the studies. • No correlation between DNA adducts and urinary polycyclic aromatic hydrocarbon metabolites was generally found, but stratification by genotype suggested that correlation might be present. • Correlation was more probable between structurally specified protein adducts and urinary polycyclic aromatic hydrocarbon metabolite 1-hydroxypyrene than with less specific xenobiotic-protein structures. • Stratification of the study population for confounding factors, such as smoking status, may reveal hidden correlations. • Cytogenetic biomarker studies also gave complex results. Examples of both positive correlation and lack of correlation with exposure markers were found. Molecular epidemiological studies of cancer that include exposure biomarkers have greatly increased in number in recent years, the rationale being to measure the biologically relevant aspect(s) of exposure. However, the use of biomarkers to measure exposure is not a panacea. Most biomarker-based studies, of both prospective and retrospective design, rely on a single biological sample. Most exposures in cancer epidemiology are time-related variables, and both carcinogenesis models and empirical evidence strongly point towards the importance of induction and latency periods in cancer onset, the need to separate the role of duration and intensity of exposure, and the decrease in effect after cessation of exposure. While most biomarkers measure recent exposure, it is possible to apply them in the measurement of temporal changes, e.g. by collecting repeated samples from subjects enrolled in prospective studies or from a sample of the original cohort. Another important parameter is the in vivo lifetime of the biomarker after it has been generated by a carcinogen exposure. Biomarker-based epidemiological results are subject to the methodological problems affecting all types of observational studies, namely bias and confounding. A further problem in the measurement of exposure biomarkers in retrospective studies is their possible dependence on the disease process. With the increasing use of biomarkers of dose and effect of carcinogens and the possible input of these data into the regulatory area, it is essential that the methodology be standardised and wherever possible internationally accepted protocols be established for this. The need to validate exposure biomarkers before their application in population-based studies arises from the variability in biomarker-based measurements of exposure, which can be due to interindividual sources (e.g. differences between exposed and unexposed individuals, usually the component of variability of primary interest), intraindividual sources (e.g. variability in hormonal levels) and observer sources, including measurement error. This is the domain of so-called transitional studies, which aim to characterise the biomarker itself rather than the underlying biological phenomenon. This is the area in which most work is needed in the near future: efforts such as the ECNIS Network of Excellence will be instrumental in providing a logical framework for the development and validation of biomarkers of exposure with the ultimate goal of their application in molecular epidemiological studies. • There was a positive correlation for different urinary metabolites and urinary mutagenicity in most of the studies. • No correlation between DNA adducts and urinary polycyclic aromatic hydrocarbon metabolites was generally found, but stratification by genotype suggested that correlation might be present. • Correlation was more probable between structurally specified protein adducts and urinary polycyclic aromatic hydrocarbon metabolite 1-hydroxypyrene than with less specific xenobiotic-protein structures. • Stratification of the study population for confounding factors, such as smoking status, may reveal hidden correlations. • Cytogenetic biomarker studies also gave complex results. Examples of both positive correlation and lack of correlation with exposure markers were found. Molecular epidemiological studies of cancer that include exposure biomarkers have greatly increased in number in recent years, the rationale being to measure the biologically relevant aspect(s) of exposure. However, the use of biomarkers to measure exposure is not a panacea. Most biomarker-based studies, of both prospective and retrospective design, rely on a single biological sample. Most exposures in cancer epidemiology are time-related variables, and both carcinogenesis models and empirical evidence strongly point towards the importance of induction and latency periods in cancer onset, the need to separate the role of duration and intensity of exposure, and the decrease in effect after cessation of exposure. While most biomarkers measure recent exposure, it is possible to apply them in the measurement of temporal changes, e.g. by collecting repeated samples from subjects enrolled in prospective studies or from a sample of the original cohort. Another important parameter is the in vivo lifetime of the biomarker after it has been generated by a carcinogen exposure. Biomarker-based epidemiological results are subject to the methodological problems affecting all types of observational studies, namely bias and confounding. A further problem in the measurement of exposure biomarkers in retrospective studies is their possible dependence on the disease process. With the increasing use of biomarkers of dose and effect of carcinogens and the possible input of these data into the regulatory area, it is essential that the methodology be standardised and wherever possible internationally accepted protocols be established for this. The need to validate exposure biomarkers before their application in population-based studies arises from the variability in biomarker-based measurements of exposure, which can be due to interindividual sources (e.g. differences between exposed and unexposed individuals, usually the component of variability of primary interest), intraindividual sources (e.g. variability in hormonal levels) and observer sources, including measurement error. This is the domain of so-called transitional studies, which aim to characterise the biomarker itself rather than the underlying biological phenomenon. This is the area in which most work is needed in the near future: efforts such as the ECNIS Network of Excellence will be instrumental in providing a logical framework for the development and validation of biomarkers of exposure with the ultimate goal of their application in molecular epidemiological studies.en
dc.description.sponsorshipECNIS Network of Excellenceen
dc.language.isoenen
dc.publisherThe Nofer Institute of Occupational Medicineen
dc.relation.ispartofseriesECNIS Reporten
dc.relation.ispartofseries1en
dc.relation.urlhttp://www.ecnis.org/images/stories/ecnis/documents/reports/Biomarkers/6.correlations...%20143-159.pdfen
dc.subjectbiomarkersen
dc.subjectDNAen
dc.subjectpopulationen
dc.titleCorrelations among biomarkers.en
dc.typeBook chapteren
All Items in ECNIS-NIOM are protected by copyright, with all rights reserved, unless otherwise indicated.