Model-based comparison of antibiotic activity at target-site concentrations: a case study with ceftaroline and lefamulin for soft-tissue infections
Wisse van Os (1), Anh Duc Pham (2), Sabine Eberl (1), Iris Minichmayr (1), J. G. Coen van Hasselt (2), Markus Zeitlinger (1)
(1) Department of Clinical Pharmacology, Medical University of Vienna, Austria, (2) Division of Systems Pharmacology & Pharmacy, Leiden Academic Centre for Drug Research, Leiden University, Netherlands
Objectives: Comparing antibiotic activity of different drugs using traditional pharmacokinetic-pharmacodynamic (PK-PD) targets has limitations: the targets may have been derived using different pathogens, experimental systems and PD endpoints, and information on the time-course of antibiotic concentrations and effects is lost. Moreover, both the targets and the probability of target attainment analyses which they underlie are often based on plasma PK, whereas infection-site PK drives the effect. We thus propose an approach to compare the activity of antibiotics based on target-site PK-PD, integrating target-site PK models, PD models based on in vitro time-kill data, and antimicrobial susceptibility distributions. We exemplify the approach by comparing the activity of lefamulin (LEF) and ceftaroline (CPT) against methicillin-resistant Staphylococcus aureus (MRSA) at soft-tissue concentrations. LEF is a novel antibiotic with potential utility for treatment of soft-tissue infections [1]. CPT is approved for soft-tissue infections and its MRSA coverage based on plasma PK is superior to alternatives [2,3].
Methods:
Population PK: A population PK model for LEF in plasma, subcutaneous adipose tissue and skeletal muscle tissue was developed using data from a microdialysis study in 12 healthy volunteers [4]. A saturable protein binding model was incorporated to convert total to unbound plasma concentrations [5]. Microdialysis data were modelled using an integrated dialysate-based approach [6], and various tissue distribution models were evaluated. For CPT, a published model based on data from a similar soft-tissue PK study was used [7].
In vitro PD: Six MRSA strains were exposed to LEF and CPT at concentrations 0.125-8 times the minimum inhibitory concentration (MIC), in twofold increments. Samples were taken during 24 h. A joint PD model for all isolates was developed. The bacterial population was assumed to consist of growing, drug-susceptible and resting, insusceptible subpopulations. Linear, power and Emax models were evaluated to describe concentration-effect relationships. The MIC was tested as a covariate on exposure-response parameters using linear and power functions. Estimates of bacteria-related parameters (initial and maximum population size, growth rate) were shared for the two drugs. The PK and PD models were linked and stochastic simulations were performed.
Results:
PK model: LEF plasma PK was best described by a three-compartment model, in line with literature [5,8]. To describe soft-tissue PK, a model with two additional compartments per tissue was superior to a single tissue compartment or scaling interstitial space fluid concentrations to plasma model compartments. The need for peripheral tissue compartments may reflect intracellular LEF accumulation [9].
PD model: Sigmoid Emax models best described the in vitro effect of both antibiotics. Regrowth was observed for CPT and was best described by a time- and concentration-dependent reduction of the maximum effect. A power relationship between the potency parameter (EC50) and the MIC allowed the description of the data from all strains simultaneously.
Simulations: In simulations with common MIC values for the respective antibiotics, standard doses of ceftaroline fosamil (600 mg q12h) resulted in lower bacterial counts at 24 h than standard doses of LEF (150 mg q12h), regardless of whether simulations were based on unbound plasma PK or tissue PK. However, higher LEF activity was observed at soft-tissue PK compared to plasma PK, resulting in >1 log10 differences in median bacterial count for some MIC values. In contrast, CPT activity was higher at plasma PK than soft-tissue PK. Consequently, LEF doses required to match the effect of standard doses of CPT were lower when considering soft-tissue PK (275 mg q12h) compared to plasma PK (500 mg q12h) in simulations with the MRSA MIC90 value for each antibiotic [10].
Conclusions: CPT appeared to have superior activity to LEF against MRSA at soft-tissue concentrations. Substantial dose increases of LEF, not supported by safety data, would be required to match CPT activity. The proposed approach compares antibiotic activity given target-site PK and simultaneously considers the full time course of antibiotic PK and PD. The framework might be applied to explore the potential of antibiotics for specific indications or in special populations with unique target-site PK, such as obese patients.
References:
[1] Prince WT et al. Antimicrob Agents Chemother 57(5) (2013), 2087-94.
[2] Cristinacce et al. Diagn Microbiol Infect Dis 99(4) (2021), 115292.
[3] Housman et al. Int J Antimicrob Agents 44(3) (2014), 235-241.
[4] Zeitlinger M et al. J Antimicrob Chemother 71(4) (2016), 1022-1026.
[5] Rubino CM et al. Antimicrob Agents Chemother 59(1) (2015), 282-8.
[6] Tunblad K et al. Pharm Res 21(9) (2004), 1698-707.
[7] Helfer VE et al. Antimicrob Agents Chemother 66(9) (2022), e00741-22.
[8] Zhang L et al. J Antimicrob Chemother 74(Suppl 3) (2019), iii27-iii34.
[9] Wicha WW et al. J Antimicrob Chemother 74(Suppl 3) (2019), iii11-iii18.
[10] European Committee on Antimicrobial Susceptibility Testing. Data from the EUCAST MIC distribution website, last accessed 15 March 2023. http://www.eucast.org.