117 HIV Protease: Can Better Inhibitors Be Found? Dale J Kempf Global Pharma Res and Devt, Abbott Labs, Abbott Park, IL, USA

The introduction of HIV protease inhibitors (PIs) in the mid 1990s revolutionized the treatment of HIV infection.  Eight PIs are currently available, and more are in development, particularly for use in individuals infected with drug-resistant virus.  Nonetheless, the usefulness of the PI class is still limited by GI side effects, relatively high pill burden, drug-drug interactions (affecting both PI and concomitant medication pharmacokinetics), metabolic complications (e.g., hyperlipidemia, insulin resistance, hyperbilirubinemia), and concern over long-term manifestations (lipodystrophy and cardiovascular disease).  Although major advances have been realized in optimizing PI efficacy and identifying mechanisms of resistance development, a detailed understanding of many of the PI side effects has proven challenging.  The identification of new PIs without the above limitations will require the establishment of mechanistically relevant preclinical assays.  This lecture will focus on efforts from our laboratories directed toward the development and validation of novel models of PI side effects.  A microarray-based hyperlipidemia model that correlates with cholesterol elevations, and an in vivo hyperbilirubinemia model that recapitulates bilirubin elevations produced by PIs clinically, have been developed and used to discriminate potential PI candidate compounds.  Studies in human hepatocytes and co-transfected human PXR/PXRE reporter systems have provided unique insights into the origins of drug-drug interactions.  The challenges of establishing relationships between chemical structure and biological property, and combining multiple favorable properties into novel PIs will be discussed.