BioCentury
ARTICLE | Translation in Brief

Residential predictions

PK/PD model incorporating time-dependent kinetics can predict compound efficacy

April 30, 2015 7:00 AM UTC

Researchers at AstraZeneca plc and Stony Brook University have developed a PK/PD model that incorporates how long a drug sits on its target - its residence time - to predict the efficacy and safety of candidate compounds. The team believes the model's greatest value will be for compounds that have slow binding kinetics, such as antibacterials that require high and sustained target engagement to produce a therapeutic effect.

Peter Tonge and Stewart Fisher thought a model that accounts for the on- and off-rates of a drug-target interaction could be a more accurate predictor of a compound's activity in vivo than traditional models, which rely on IC50, Ki or Kd values. Tonge is professor of chemistry at Stony Brook. Fisher is director of microbial therapeutics at the Broad Institute of MIT and Harvard, and was executive director of biology in the infectious diseases unit at AZ when the study was performed...