ABSTRACT Liver cirrhosis is accompanied by pathophysiological changes. Due to multiple absorption, distribution, metabolism and excretion (ADME)‐related pathophysiological alterations, the estimation of the net pharmacokinetics (PK) change in cirrhotic patients is complex. Physiologically based pharmacokinetic (PBPK) modeling is a mechanistic modeling technique that combines knowledge of physiological and drug‐related properties and, thereby, allows the prediction of organism‐specific drug PK. For the integration of pathophysiological changes into a PBPK model, such changes need to be quantified appropriately. To date, published liver cirrhosis pathophysiology repositories contain only average changes for three distinct disease stages limiting clinical applicability. Therefore, the aim of this study was the development of a repository that (1) describes physiological alterations throughout the body during cirrhosis progression, (2) quantifies both mean changes and population variability, and (3) adds parameters of not yet included changes. For this purpose, data was gathered and processed using a Markov‐Chain‐Monte‐Carlo (MCMC)‐based approach that allowed the handling of heterogenous data and information on population variability. The resulting repository, based on 216,609 data points from 68 literature studies and 208,851 patients from the IBM Explorys electronic health records database, encompasses 30 physiological parameters. Integration into a PBPK modeling framework revealed good predictive performance with 96% of all data points for predicted PK parameter ratios lying within a twofold prediction range. In summary, the presented approach provides an advancement in the field of PK modeling in liver cirrhosis patients, possibly facilitating the planning and analysis of clinical studies in these patients and moving towards virtual studies.
Schneider et al. (Fri,) studied this question.