%0 Journal Article %A Vicini, Paolo %A Gastonguay, Marc R. %A Foster, David M. %D 2002 %I Begell House %N 4-6 %P 379-418 %R 10.1615/CritRevBiomedEng.v30.i456.60 %T Model-based Approaches to Biomarker Discovery and Evaluation: A Multidisciplinary Integrated Review %U https://www.dl.begellhouse.com/journals/4b27cbfc562e21b8,5b51f41866ab7f77,0c1a7250501db3cb.html %V 30 %X The most common use of any mathematical or statistical model in physiology, pathophysiology, or therapy evaluation is to organize all relevant components characterizing the system behavior into a rigorously testable framework. This approach is widely applied both to the study of complex homeostatic paradigms involving endogenous substances and to the evaluation of the kinetics and dynamics of xenobiotics such as toxicants and drugs. In either case, one seeks a quantitative framework of the system that is consistent with known physiology and pharmacology and is "compatible" (in some meaningful sense) with all available data. The models are then evaluated and subjected to identifiability and validity tests and can then be used to estimate unknown parameters of interest, to make predictions about system behavior, to simulate previously unobserved behavior in response to a putative perturbation, and to aid in further experimental design. In a broader context, however, the focus and ultimate goal of this set of methodologies (whether this is explicitly stated or not) lies in understanding the mechanisms of physiology and pathophysiology and measuring the effect of therapeutic interventions through the accurate quantification of biomarkers of interest. In this review, we attempt to bring together under this comprehensive framework more than four decades of investigation on modeling and simulation in the life sciences (in particular, we will concentrate on the areas of pharmacology, physiology, and bioengineering). We demonstrate that such modeling approaches, when appropriately designed and evaluated, have significant potential and can be used to understand the multiple factors of disease progression and response to therapeutic interventions, the most likely causes of variability in population and individual responses to therapy, and the most appropriate timing of treatment administration. Lastly, they may allow estimation and prediction of significant outcomes in feasibility and clinical studies. %8 2002-11-30