Salim Bouchene (1), Lena E. Friberg (1), Diamantis Plachouras (2), Sven Björkman (1), Mats O. Karlsson (1)
(1) Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden (2) 4th Department of Internal Medicine, Medical School, Athens University, Athens, Greece
Objectives: Colistin is used as a salvage therapy for MDR GNB infections and administered as a prodrug, colistimethate sodium (CMS). Characterizing tissue distribution of colistin is a major issue to optimize bacteria kill and avoid toxicity. Whole Body Physiologically Based Pharmacokinetic (WBPBPK) models are increasingly used to predict pharmacokinetic behavior of drugs. The aim of the study is to develop a WPBPK model to describe human PK of CMS and colistin in critically ill patients that could eventually be used to predict bacteria kill in different tissues.
Methods: Thirty-one patients [1,2] with MDR-GNB infections treated with colistin were included in the analysis (10 females; mean age, 62 years; mean creatinine CL, 80ml/min). CMS was intravenously administered at doses of 80, 160, 240 or 480mg every 8 hours through a 15-min infusion. Venous blood was collected after the first and fourth infusions. CMS is cleared renally and via hydrolysis to colistin (assumed to occur in all tissues) whereas colistin is mainly eliminated non-renally [1,2].
A WBPBPK was developed assuming each tissue as a single, perfusion limited and well-stirred compartment. All clearances and Kp values of CMS and colistin were estimated using the prior functionality in NONMEM 7. Prior information was collected from literature[3], in-silico prediction equations [4] and in vitro experiments.
Results: The WBPBPK model described well both CMS and colistin plasma concentrations over time. CMS renal clearance was estimated to be 11.9 L/h whereas non-renal clearance was 1.8 L/h. Colistin non-renal clearance was 2.0 L/h and its renal clearance was 0.13 L/h. Steady-state volumes of distribution of CMS and colistin were 40 L and 30 L, respectively.
Predicted Kp factors were lower for CMS and higher for colistin than the calculated priors. Estimated renal clearance of CMS was higher than the prior literature value [3]. Low tissue distribution of colistin was predicted in the large tissues: adipose, muscle and carcass.
Conclusions: The WBPBPK model described plasma concentration-time profiles for both CMS and colistin. The relatively high Kp values in comparison to those expected for distribution only to interstitial space [5] suggest that colistin enters into cells or binds to cell membranes or interstitial macromolecules.
References:
[1] Plachouras et al., Antimicrob Agents Chemother. 2009
[2] Mohamed et al., Antimicrob Agents Chemother. 2012
[3] Grégoire et al., 21st ECCMID, 2011
[4] Berezhkovskiy, J Pharm Sci. 2004
[5] Björkman et al., J Pharm Sci. 2001
Reference: PAGE 21 (2012) Abstr 2535 [www.page-meeting.org/?abstract=2535]
Poster: Absorption and Physiology-Based PK