Pharmacometrics: modelling and simulation tools to improve decision making in clinical drug development

There is broad recognition within the pharmaceutical industry that the drug development process, especially the clinical part of it, needs considerable improvement to cope with rapid changes in research and health care environments. Modelling and simulation are mathematically founded techniques that have been used extensively and for a long time in other areas than the pharmaceutical industry (e.g. automobile, aerospace) to design and develop products more efficiently. Both modelling and simulation rely on the use of (mathematical and statistical) models which are essentially simplified descriptions of complex systems under investigation. It has been proposed to integrate pharmacokinetic (PK) and pharmacodynamic (PD) principles into drug development to make it more rational and efficient. There is evidence from a survey on 18 development projects that a PK/PD guided approach can contribute to streamline the drug development process. This approach extensively relies on PK/PD models describing the relationships among dose, concentration land more generally exposure), and responses such as surrogate markers, efficacy measures, adverse events. Well documented empirical and physiologically based PK/PD models are becoming available more and more, and there are ongoing efforts to integrate models for disease progression and patient behavior (e.g. compliance) as well. Other types of models which are becoming increasingly important are population PK/PD models which, in addition to the characterization of PK and PD, involve relationships between covariates (i.e. patient characteristics such as age, body weight) and PK/PD parameters. Population models allow to assess and to quantify potential sources of variability in exposure and response in the target population, even under sparse sampling conditions. As will be shown for an anticancer agent, implications of significant covariate effects can be evaluated by computer simulations using the population PK/PD model. Stochastic simulation is widely used as a tool for evaluation of statistical methodology including for example the evaluation of performance of measures for bioequivalence assessment. Recently, it was suggested to expand the use of simulations in support of clinical drug development for predicting outcomes of planned trials. The methodological basis for this approach is provided by (population) PK/PD models together with random sampling techniques. Models for disease progression and behavioral features like compliance, drop-out rates, adverse event dependent dose reductions, etc. have to be added to population PK/PD models in order to mimic the real situation. It will be shown that computer simulation helps to evaluate consequences of design features on safety and efficacy assessment of the drug, enabling identification of statistically valid and practically realisable study designs. For both modelling and simulation a guidance on 'best practices' is currently worked out by a panel of experts comprising representatives from academia, regulatory bodies and industry, thereby providing a necessary condition that model-based analysis and simulation will further contribute to streamlining pharmaceutical drug development processes.