Christian Radke (1), Dagmar Horn (2), Christian Lanckohr (3), Michaela Meyer (4), Thomas Eissing (4), Georg Hempel (1)
(1) Department of Pharmaceutical and Medicinal Chemistry, Clinical Pharmacy, Westfälische Wilhelms-Universität Münster, Germany, (2) Department of Pharmacy, University Hospital of Münster, Germany, (3) Department of Anesthesiology, Intensive Care Medicine and Pain Therapy, University Hospital of Münster, Germany, (4) Systems Pharmacology CV, Bayer Technology Services GmbH, Leverkusen, Germany
Objectives: Sepsis is characterized by an excessive release of inflammatory mediators substantially affecting body composition and organ function further augmented by intensive care management (e.g. extensive fluid administration, administration of vasoactive drugs). Thus, drug pharmacokinetics can be altered resulting in uncertainty of pharmacotherapeutic success. Using vancomycin as a renally cleared model drug, a physiologically based pharmacokinetic (PBPK) model for septic patients was developed taking disease related physiological changes into account.
Methods: An extensive literature search was conducted to determine qualitative and quantitative information on physiological alterations described for different septic states and defined according to the official international guideline [1]. Gathered information was consolidated and informed patho-physiological parameter changes were incorporated in the PBPK software PK-Sim® [2], scaling a validated vancomycin model of healthy subjects. Furthermore, an individualized model approach was applied taking readily available patient characteristics into account, such as creatinine clearance (CrCL). Plasma concentrations for model validation, as well as individual patient characteristics were obtained from an ongoing prospective observational study at the University Hospital of Münster.
Results: Literature search yielded more than 100 studies, presenting information on various physiological parameters. However, quantitative data on some parameters are almost entirely lacking (e.g. metabolic enzyme activity, renal blood flow). The literature-based sepsis PBPK model showed a slight improvement when compared to a model considering no physiological alterations, with 59.3% vs 55.6% of the predicted values being within 30% of the observed values. The mean absolute prediction error (MAPE) was 38.0% vs 39.2%, respectively. However, a distinct improvement of the literature model was observed for higher concentrations during the distribution phase. Further individualization of the model significantly improved model performance, with the individualized model showing the best results with 88.9% of all predicted values being within 30% of the observed values and a MAPE of 17.6%.
Conclusions: Available literature information on physiological changes in sepsis improves model prediction regarding vancomycin distribution. For precise predictions in septic patients, individual patient characteristics such as CrCL are required.
References:
[1] Bone, R.C., et al., Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest, 1992. 101(6): p. 1644-55.
[2] Eissing, T., et al., A computational systems biology software platform for multiscale modeling and simulation: integrating whole-body physiology, disease biology, and molecular reaction networks. Front Physiol, 2011. 2: p. 4.
Reference: PAGE 25 (2016) Abstr 5868 [www.page-meeting.org/?abstract=5868]
Poster: Drug/Disease modeling - Absorption & PBPK