Correlation between biomarkers of human exposure to genotoxins with focus on carcinogen-DNA adducts.

Correlations among biomarkers, an important issue in biomarker research, provide enhanced insight and understanding of the complexity of molecular mechanisms initiated by environmental genotoxic agents in the human organism. Occupational and environmental exposures mostly represent mixtures of genotoxic agents, whereas the specificity of biomarker measurements varies widely. Here, we give an overview of the correlation studies with particular emphasis on DNA adduct biomarker analysis of exposure to polycyclic aromatic hydrocarbons (PAHs) and/or tobacco smoke. We have collected data on correlations between different DNA adduct detection methods, DNA adduct structures and DNA adduct levels in human tissues. Data are also presented on the correlation between DNA adducts and other biomarkers of exposure and of early biological effects, including protein adducts, urinary metabolites and cytogenetic end points. In numerous studies, 32P-postlabelling and immunoassay measurements of DNA adducts recognized the difference between exposure groups similarly; however, at the individual level, there was, in general, not a statistically significant correlation between the two determinations. Inconsistency was found regarding the correlation between the levels of total bulky adducts and specific single DNA adduct structures. A number of studies found a positive correlation between DNA adduct levels in target and surrogate tissues, although stratification for exposure level may have influenced the results. Characteristically, there was a positive correlation between DNA adduct levels in tumour and normal tissue pairs. In general, there was a lack of correlation between DNA adducts and urinary PAH metabolites, but after stratification for particular genetic polymorphisms correlation may have emerged between the two biomarkers of exposure. The correlations with cytogenetic biomarkers were very complex, with examples of both positive correlation and lack of correlation. Exploration of correlations among biomarkers contributes to the further progress of molecular cancer epidemiology and to the selection of the optimal biomarkers for the investigation of human exposure to carcinogens.