A Method for Biomarker Validation and Biomarker-Based Dose Response: A Case Study with a Bayesian Network Model for Benzene

To facilitate the use of biomarkers in human health risk assessments, a biomarker decision support system is presented that systematically identifies, documents, validates, and incorporates biomarkers into occupational risk assessments. A biomarker database structure was developed and decision rules were identified to organize the diverse types of data to support an occupational risk assessment for benzene. A suite of biomarker validation approaches was applied to evaluate potential biomarkers of exposure and effects of exposure for benzene-induced acute myeloid leukemia (AML). Traditional biomarker evaluation approaches based on the Hill criteria and regression analysis were coupled with a Bayesian network approach to validate (or discount) individual biomarkers and ultimately link the validated biomarkers along the exposuredisease continuum. Dose-response analyses using validated biomarkers were conducted to contrast various biomarker-based dose-response approaches. Although multiple analysis and validation approaches are described in this benzene case study, it is envisioned that the system will be most useful as a set of options allowing the user to choose the analytical approach(es) appropriate to the data set and needs of the analysis. Ultimately, this work aims to identify appropriate tools for critically evaluating biomarkers, as well as to provide a quantitative approach for linking changes in biomarkers of effect both to exposure information and to changes in disease response. Such linkage can provide a scientifically valid point of departure that incorporates precursor dose-response information without being dependent on the difficult issue of a definition of adversity for precursors.