B Academy of Molecular Imaging 2005 Mol Imaging Biol (2005) 7:5Y13 DOI: 10.1007/s11307-005-0954-7 Published Online: 30 December 2004 REVIEW ARTICLE Molecular Imaging in Drug Discovery and Development: Potential and Limitations of Nonnuclear Methods Markus Rudin, PhD, Martin Rausch, Markus Stoeckli DTC/Analytical and Imaging Sciences, Novartis Institute for Biomedical Research, WSJ-386.2.02, CH-4002, Basel, Switzerland Abstract Noninvasive conventional imaging methods are established technologies in modern drug discovery and development providing valuable morphological, physiological, and metabolic information to characterize disease phenotypes, to evaluate the efficacy of therapy and to identify and develop potential biomarkers for clinical drug evaluation. The development of target-specific or molecular imaging has added a new dimension: molecular events such as the target expression, the drugYtarget interaction, or the activation of signal transduction pathways can be studied in the intact organism with high spatial and temporal resolution. Molecular imaging is inherently a multimodality approach. In this article, we review the role of molecular imaging for drug discovery and development focusing on nonnuclear imaging methods, i.e., magnetic resonance imaging (MRI) and optical imaging techniques based on fluorescence and bioluminescence readouts. Examples discussed are direct visualization of target expression using target-specific ligands or reporter genes, pathway imaging, and cell-trafficking studies. Key words: Molecular imaging, Drug discovery, Nonnuclear methods Introduction T he primary role of biomedical imaging is in diagnos- tics, i.e., the characterization of a disease phenotype based on morphological or physiological readouts. When combining a contrast-enhancing principle (reporter group) with a target-specific carrier moiety (receptor ligand, anti- body, cell), highly specific information on targeted molec- ular interactions can be obtained. Such target-specific or molecular imaging approaches have raised considerable interest both from a diagnostic and therapeutic point of view [1], and will become important tools for target validation providing temporo-spatial information on target expression levels and function in relation to a disease process. It is not surprising that target-specific imaging methods designed to study such molecular events in an intact organism are currently being developed at a rapid pace [1]. In later phases of drug development, before entering a clinical develop- ment programVin particular when targeting chronic dis- eases with late clinical endpointsVit is important to identify reliable biomarkers to allow for the validation of the mechanism of action of man and for the monitoring of drug efficacy. Eventually, such biomarkers may serve as surro- gates for a clinical endpoint. There is intensive research in trying to define potential biomarkers for drug efficacy and safety involving readouts at the level of gene transcription, gene translation or regarding drug-induced metabolic alter- ations [2], all requiring tissue and/or body-fluid sampling. Such markers are, in general, highly sensitive; however, their interpretation is not straightforward requiring sophis- ticated biostatistical tools. In addition, they only provide information at the time point of sampling and are prone to sampling errorsVin particular when analyzing heteroge- neous tissue such as tumors. In comparison, imaging approaches will yield spatially and temporally resolved information in a noninvasive manner. Correspondence to: Markus Rudin, PhD; e-mail: markus-1.rudin@pharma. novartis.com