Vincent Aranzana-Climent (1,2), Julien M. Buyck (1,2), Lena E. Friberg (3), Younes Smani (4,5), Jerónimo Pachón-Diaz (4), Emma Marquizeau (1,2), William Couet (1,2,6), Nicolas Grégoire (1,2)
(1) Université de Poitiers, Pharmacologie des anti-infectieux, Poitiers, France, (2) INSERM U1070 - Pharmacologie des anti-infectieux, Poitiers, France, (3) Uppsala University, Department of Pharmaceutical Biosciences, Uppsala, Sweden, (4) Institut of Biomedicine of Seville (IBiS), Seville, Spain, (5) University Hospital Virgen del Rocio/CSIC/University of Seville, Seville, Spain, (6) CHU de Poitiers, Service de Pharmacologie-Toxicologie, Poitiers, France
Objectives:
Acinetobacter baumannii is one of the most difficult to treat multi-drug resistant (MDR) pathogens responsible for opportunistic nosocomial infections all over the world [1]. It can cause a broad range of infections, the deadliest being ventilator-associated pneumonia and bloodstream infections [2], and has the ability to become resistant to a wide variety of drugs [3]. In face of these resistances, neglected and disused antibiotics like polymyxins (colistin and polymyxin B) may be used, especially in combination with other antibiotics, as the last line of defence against MDR A. baumannii [4]. In a preliminary checkerboard screening study (data not shown), polymyxin B (PMB) and minocycline (MIN) combination was shown to be synergistic on polymyxin-resistant A. baumannii clinical isolates.
To further investigate this synergistic combination, a polymyxin-resistant clinical isolate (CR17) was selected to be thoroughly studied through development of a semi-mechanistic pharmacokinetic-pharmacodynamic (PK/PD) model.
The main objective of this study was to investigate the determinants of the PMB + MIN synergy against CR17 observed in checkerboard experiments
Methods:
A polymyxin-resistant A. baumannii clinical isolate CR17 (MIC: PMB 8 mg/L; MIN 4 mg/L) was studied [5].
Heteroresistance to PMB and MIN was evaluated by plating a high inoculum (~109 CFU/mL) on plates containing 8 x MIC of drug (resistant subpopulation) and on drug free plates (total population) and counting after 24h at 37°C.
Fitness cost was evaluated by inoculating a 96 well plate with ~106CFU/mL of bacteria that grew on drug-free and drug-containing plates and reading OD at 600nm over 24h.This enabled calculation of a growth rate constant as described earlier [6].
Single drug and combination time-kill experiments (TKE) with determination of total bacterial count at 0, 3, 8, 24 and 30h at concentrations ranging from 1/16 to 4 x MIC for MIN and from 1/128 to 1 x MIC for PMB. Simultaneously, population analysis profiles (PAPs) were performed by culture and bacterial count on PMB-containing (8 x MIC) plates.
A mechanistic bacterial life cycle model based on results of heteroresistance and fitness cost experiments was built. For single drug effect model multiple PK/PD models taken from literature [7] were tested and the interaction between the two drugs was explored using the “Global PharmacoDynamic Interaction model (GPDI)” [8].
Analysis was performed using NONMEM v7.4.3 with Laplacian algorithm [9], PsN [10] and R [11].
Results:
CR17 did not exhibit heteroresistance to MIN but to PMB (mean frequency: 5.07 *10-6, n=6).
No fitness cost was found, mean growth rate for colonies that grew on drug-free plates was 1.105 h-1 (n=11) and 1.092 h-1 (n=11) for colonies that grew on plates containing 8 x MIC of PMB.
A total of 253 TKE were performed. In single drug TKE, no effect was observed at concentrations <1 x MIC MIN while at concentrations >= 1 x MIC a concentration-independent effect was observed, with complete bacterial killing at 30h. PMB alone exhibited a fast concentration-dependent effect followed by regrowth at all tested concentrations. When combining MIN and PMB, total bacterial killing at 30h was observed for concentrations as low as 1/4 x MIC MIN + 1/16 x MIC PMB.
Data were adequately described by a model including a phenotypic switch to resting [12] form at high bacterial concentrations, two subpopulations (PMB-susceptible and PMB-resistant), a sigmoidal Emax effect of MIN, a slope power effect model of PMB and adaptive resistance to PMB of both subpopulations. Synergy consisted of PMB increasing MIN killing of both subpopulations, and MIN increasing PMB killing of the susceptible subpopulation. Simulations showed that while significantly improving the fit to data, the potentiation effect of MIN on PMB could be accounted for a marginal part of the total effect, while the potentiation effect of PMB on MIN was essential to the observed effect.
Conclusions:
Combining MIN and PMB in vitro proved to be efficient against a polymyxin-resistant A. baumannii clinical isolate. The developed PK-PD model enabled us to quantify the synergistic effect between MIN and PMB.
This study could serve as a proof of concept that using targeted experiments to inform model building and then use advanced PK/PD models built on combination TKE+PAPs data is a good methodology to gain insights on the in vitro PK/PD of antibiotic combinations.
References:
[1] Antunes LCS, Visca P, Towner KJ. Acinetobacter baumannii: evolution of a global pathogen. Pathog Dis 2014;71:292–301. doi:10.1111/2049-632X.12125.
[2] Dijkshoorn L, Nemec A, Seifert H. An increasing threat in hospitals: multidrug-resistant Acinetobacter baumannii. Nature Reviews Microbiology 2007;5:939–51. doi:10.1038/nrmicro1789.
[3]. Fournier P-E, Vallenet D, Barbe V, Audic S, Ogata H, Poirel L, et al. Comparative Genomics of Multidrug Resistance in Acinetobacter baumannii. PLOS Genetics 2006;2:e7. doi:10.1371/journal.pgen.0020007.
[4] Nation RL, Li J, Cars O, Couet W, Dudley MN, Kaye KS, et al. Framework for optimisation of the clinical use of colistin and polymyxin B: the Prato polymyxin consensus. The Lancet Infectious Diseases 2015;15:225–34. doi:10.1016/S1473-3099(14)70850-3.
[5] López-Rojas R, Jiménez-Mejías ME, Lepe JA, Pachón J. Acinetobacter baumannii Resistant to Colistin Alters Its Antibiotic Resistance Profile: A Case Report From Spain. J Infect Dis 2011;204:1147–8. doi:10.1093/infdis/jir476
[6] Bleibtreu A, Gros P-A, Laouénan C, Clermont O, Le Nagard H, Picard B, et al. Fitness, Stress Resistance, and Extraintestinal Virulence in Escherichia coli. Infect Immun 2013;81:2733–42. doi:10.1128/IAI.01329-12.
[7] Nielsen EI, Friberg LE. Pharmacokinetic-Pharmacodynamic Modeling of Antibacterial Drugs. Pharmacol Rev 2013;65:1053–90. doi:10.1124/pr.111.005769.
[8] Wicha SG, Chen C, Clewe O, Simonsson USH. A general pharmacodynamic interaction model identifies perpetrators and victims in drug interactions. Nature Communications 2017;8:2129. doi:10.1038/s41467-017-01929-y.
[9]. Beal SL et al. 1989-2019. NONMEM Users Guides. Icon Development Solutions, Ellicott City, Maryland, USA.
[10] Lindbom L, Pihlgren P, Jonsson EN. PsN-Toolkit–a collection of computer intensive statistical methods for non-linear mixed effect modeling using NONMEM. Comput Methods Programs Biomed. 2005 Sep;79(3):241-57
[11] R Core Team (2018). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL: https://www.R-project.org/.
[12] Rittershaus ESC, Baek S-H, Sassetti CM. The Normalcy of Dormancy: Common Themes in Microbial Quiescence. Cell Host & Microbe 2013;13:643
Reference: PAGE 28 (2019) Abstr 9164 [www.page-meeting.org/?abstract=9164]
Poster: Drug/Disease Modelling - Infection