Limitations of Antibiotic MIC-Based PK-PD Metrics: Looking Back to Move Forward

Multiomics informed mathematical model for meropenem and tobramycin against hypermutable Pseudomonas aeruginosa

BACKGROUND

Hypermutable P. aeruginosa isolates frequently display resistance emergence during treatment. Mechanisms of such resistance emergence have not been explored using dynamic hollow-fiber studies and multiomics informed mathematical modeling.

METHODS

Two hypermutable and heteroresistant P. aeruginosa isolates, CW8 (MICmeropenem=8 mg/L, MICtobramycin=8 mg/L) and CW44 (MICmeropenem=4 mg/L, MICtobramycin=2 mg/L), were studied. Both isolates had genotypes resembling those of carbapenem- and aminoglycoside-resistant strains. Achievable lung fluid concentration-time profiles following meropenem at 1 or 2 g every 8 h (3-h infusion) and tobramycin at 5 or 10 mg/kg body weight every 24 h (0.5-h infusion), in monotherapy and combinations, were simulated over 8 days. Total and resistant bacterial counts were determined. Resistant colonies and whole population samples at 191 h were whole-genome sequenced, and population transcriptomics performed at 1 and 191 h. The multiomics analyses informed mechanism-based modeling of total and resistant populations.

RESULTS

While both isolates eventually displayed resistance emergence against all regimens, the high-dose combination synergistically suppressed resistant regrowth of only CW8 up to ∼96 h. Mutations that emerged during treatment were in pmrB, ampR, and multiple efflux pump regulators for CW8, and in pmrB and PBP2 for CW44. At 1 h, mexB, oprM and ftsZ were differentially downregulated in CW8 by the combination. These transcriptomics results informed inclusion of mechanistic synergy in the mechanism-based model for only CW8. At 191 h, norspermidine genes were upregulated (without a pmrB mutation) in CW8 by the combination, and informed the adaptive loss of synergy in the model.

CONCLUSION

Multiomics information enabled mechanism-based modeling to describe the bacterial response of both isolates simultaneously.

IMPORTANCE

Pseudomonas aeruginosa causes serious bacterial infections in people with cystic fibrosis (pwCF), and has numerous resistance mechanisms. Current empirical approaches to informing antibiotic regimen selection have important limitations. This study exposed two P. aeruginosa clinical isolates to concentration-time profiles of meropenem and tobramycin as would be observed in lung fluid of pwCF. The combination elicited different bacterial count profiles between the isolates, despite similar bacterial baseline characteristics. We found differences between the isolates in the expression of a key resistance mechanism against meropenem at 1 h, and expression that implied a loss of cell membrane permeability for tobramycin without the expected DNA mutation. This information enabled mathematical modeling to accurately describe all bacterial profiles over time. For the first time, this multiomics informed modeling approach using DNA and RNA data was applied to a hollow-fiber infection study. Using bacterial molecular insights with mechanism-based mathematical modeling has high potential for ultimately informing personalised antibiotic therapy.

Machine Learning Prediction and Validation of Plasma Concentration-Time Profiles

Recent research has increasingly focused on using machine learning for covariate selection in population pharmacokinetics (PPK) analysis. However, few studies have explored the prediction of plasma concentration profiles of drugs using nonlinear mixed-effect models combined with machine learning. This gap includes limited validation of prediction accuracy and applicability to diverse patient populations and dosing conditions. This study addresses these gaps by using remifentanil as a model drug and applying machine learning models to predict plasma concentration profiles based on virtual and real-world data. We created various training data sets for the virtual data by clustering based on the size and diversity of the test data set. Our results demonstrated high prediction accuracy for virtual and real-world data sets using Random Forest models. These results suggest that machine learning models are effective for large-scale data sets and real-world data with variable dosing times and amounts per patient. Considering the efficiency of machine learning, it offers a fit-for-purpose approach alongside traditional PPK methods, potentially enhancing future pharmacokinetic and pharmacodynamic studies.

In vitro pharmacokinetic/pharmacodynamic modeling of the effect of mucin on polymyxin B activity against Acinetobacter baumannii

The antibacterial efficacy of polymyxins in lungs may be impacted by mucin. The aim of this study was to characterize in vitro the effect of mucin on polymyxin B (PMB) activity against two multidrug-resistant Acinetobacter baumannii strains isolated from a patient before (AB121-D0) and after colistin treatment (AB122-D12), using a pharmacokinetic/pharmacodynamic (PK/PD) modeling approach. PMB binding to mucin was characterized by ultracentrifugation in cation-adjusted Mueller-Hinton broth (CAMHB) supplemented with 1% mucin. Time-kill (TK) experiments were performed in CAMHB, with 1% mucin or without as control, and with PMB total concentrations ranging from 0.25 to 512 mg/L based on the strain's minimum inhibitory concentration (MIC). For each strain, TK data were modeled based on unbound PMB concentrations. Bacterial resistance to PMB was investigated via MIC and whole genome sequencing from bacteria that regrew in the presence of antibiotics at the end of the TK experiments. PMB unbound fraction increased nonlinearly from 6% to 60% when total concentration increased from 0.5 to 512 mg/L. In addition to binding to PMB, mucin had an impact on PMB activity, which differed between the two strains. For AB121-D0, PMB activity increased in the presence of mucin resulting in a reduction of the bacterial regrowth, whereas for AB122-D12, a decrease in PMB activity was observed. Mutations in genes involved in PMB resistance appeared randomly and explained only partially the bacterial regrowth observed in TK with antibiotics. This study showed that PMB binding to mucin had a real and important impact but was not the only factor explaining the impaired PMB efficacy in the presence of mucin.

The tradeoffs between persistence and mutation rates at sub-inhibitory antibiotic concentrations in Staphylococcus aureus

The rational design of the antibiotic treatment of bacterial infections employs these drugs to reach concentrations that exceed the minimum needed to prevent the replication of the target bacteria. However, within a treated patient, spatial and physiological heterogeneity promotes antibiotic gradients such that the concentration of antibiotics at specific sites is below the minimum needed to inhibit bacterial growth. Here, we investigate the effects of sub-inhibitory antibiotic concentrations on three parameters central to bacterial infection and the success of antibiotic treatment, using in vitro experiments with Staphylococcus aureus and mathematical and computer-simulation models. Our results, using drugs of six different classes, demonstrate that exposure to sub-inhibitory antibiotic concentrations alters bacterial growth dynamics, increases the mutation rate to antibiotic resistance, and decreases the production of persister cells thereby reducing persistence levels. Understanding this trade-off between mutation rates and persistence levels resulting from sub-inhibitory antibiotic exposure is crucial for optimizing, and mitigating the failure of, antibiotic therapy.

IMPORTANCE

Much of the research on antibiotics and antibiotic treatment has focused on drug concentrations sufficient to prevent the growth of bacteria. These concentrations, however, are not always reached everywhere in the body. Here, we look at the effects of exposure to these low concentrations of antibiotics on the common clinically important pathogen Staphylococcus aureus. We confirm a previous finding that sub-inhibitory antibiotic exposure decreases the total growth and the growth rate of the bacteria. Moreover, we demonstrate that the level of persistence, an important mechanism for bacteria to survive antibiotics, is decreased due to sub-inhibitory exposure. However, we find that the rate of generation of resistant mutants is substantially increased. Taken together, these results reveal an important trade-off that emerges as a consequence of bacteria being exposed to sub-inhibitory concentrations of antibiotics.

Simulation-based evaluation of the impact of dose fractionation study design on antibiotic PKPD analyses

Objectives

To evaluate the impact of antibiotic dose fractionation study design on pharmacokinetic/pharmacodynamic (PK/PD) indices and PKPD model estimation.

Methods

PKPD models for meropenem and polymyxin B (PMB) were applied to (i) simulate various dose fractionation studies in mice to derive PK/PD indices and efficacy targets and (ii) perform stochastic simulations and estimations evaluating which efficacy assessment times, in addition to 24 h, would improve the estimation of drug effect parameters.

Results

The R 2 values of PK/PD indices were primarily influenced by reductions of the dosing intervals for meropenem and by decreases of the lowest total daily dose for PMB. For certain study designs (e.g. frequent administration of higher meropenem doses), R 2 values for fT > MIC and fAUC/MIC were similar. Efficacy target magnitudes were also sensitive to the selected doses. Additional efficacy assessment times improved parameter accuracy (e.g. 40% reduction in relative root mean squared error of PMB effect slope). The model parameter accuracy was more affected by the selection of time points for meropenem, which included resistance, than for PMB. Efficacy measurements in the first hours after treatment start (e.g. 2 and 6 h), in addition to 24 h, were essential for resistance characterization.

Conclusions

The choice of doses and fractionations impacted PK/PD index selection and efficacy target magnitude. Depending on the antibiotic, the dose or fractionation selection appeared to be the most critical. Early treatment efficacy measurements were beneficial to PKPD model-based analyses, particularly to describe resistance processes.

Enhancing antibiotic therapy through comprehensive pharmacokinetic/pharmacodynamic principles

Antibiotic therapy relies on understanding both pharmacokinetics (PK) and pharmacodynamics (PD), which respectively address drug absorption, distribution, and elimination, and the relationship between drug concentration and antimicrobial efficacy. This review synthesizes decades of research, drawing from in-vitro studies, in-vivo models, and clinical observations, to elucidate the temporal dynamics of antibiotic activity. We explore how these dynamics, including concentration-effect relationships and post antibiotic effects, inform the classification of antibiotics based on their PD profiles. Additionally, we discuss the pivotal role of PK/PD principles in determining optimal dosage regimens. By providing a comprehensive overview of PK/PD principles in antibiotic therapy, this review aims to enhance understanding and improve treatment outcomes in clinical practice.

Translating PK-PD principles into improved methodology for clinical trials which compare intermittent with prolonged infusion of beta-lactam antibiotics

Based on the fact that beta-lactam antibiotics demonstrate time-dependent killing, different dosing strategies have been implemented to increase the time that free (f) (unbound) antibiotic concentrations remain above the Minimal Inhibitory Concentration (MIC), including prolonged and continuous infusion. Multiple studies have been performed that compared continuous with traditional intermittent infusion to improve outcomes in patients with severe sepsis and/or septic shock. These studies have yielded inconsistent results for patients as measured by clinical response to treatment and mortality due to heterogeneity of included patients, pathogens, dosing strategies and the absence of Therapeutic Drug Monitoring (TDM). The MERCY and BLING III studies failed to show a difference in mortality between patients randomized to receive continuous and intermittent infusion of beta-lactam antibiotics.
 A deeper understanding of pharmacokinetic (PK) and pharmacodynamic (PD) mechanisms that occur in critically ill patients should guide us in dose optimization and improvement in methodology for future clinical trials.

Bicarbonate Within: A Hidden Modulator of Antibiotic Susceptibility

Since its standardization, clinical antimicrobial susceptibility testing (AST) has relied upon a standard medium, Mueller-Hinton Broth/Agar (MHB/A), to determine antibiotic resistance. However, this microbiologic medium bears little resemblance to the host milieu, calling into question the physiological relevance of resistance phenotypes it reveals. Recent studies investigating antimicrobial susceptibility in mammalian cell culture media, a more host-mimicking environment, demonstrate that exposure to host factors significantly alters susceptibility profiles. One such factor is bicarbonate, an abundant ion in the mammalian bloodstream/tissues. Importantly, bicarbonate sensitizes methicillin-resistant Staphylococcus aureus (MRSA) to early-generation β-lactams used for the treatment of methicillin-susceptible S. aureus (MSSA). This "NaHCO3-responsive" phenotype is widespread among US MRSA USA300/CC8 bloodstream and skin and soft tissue infection isolates. Translationally, β-lactam therapy has proven effective against NaHCO3-responsive MRSA in both ex vivo simulated endocarditis vegetation (SEV) and in vivo rabbit infective endocarditis (IE) models. Mechanistically, bicarbonate appears to influence mecA expression and PBP2a production/localization, as well as key elements for PBP2a functionality, including the PBP2a chaperone PrsA, components of functional membrane microdomains (FMMs), and wall teichoic acid (WTA) synthesis. The NaHCO3-responsive phenotype highlights the critical role of host factors in shaping antibiotic susceptibility, emphasizing the need to incorporate more physiological conditions into AST protocols.

Model-based translation of results from in vitro to in vivo experiments for afabicin activity against Staphylococcus aureus

BACKGROUND

Translation of experimental data on antibiotic activity typically relies on pharmacokinetic/pharmacodynamic (PK/PD) indices. Model-based approaches, considering the full antibiotic killing time course, could be an alternative.

OBJECTIVES

To develop a mechanism-based modelling framework to assess the in vitro and in vivo activity of the FabI inhibitor antibiotic afabicin, and explore the ability of a model built on in vitro data to predict in vivo outcome.

METHODS

A PK/PD model was built to describe bacterial counts from 162 static in vitro time-kill curves evaluating the effect of afabicin desphosphono, the active moiety of the prodrug afabicin, against 21 Staphylococcus aureus strains. Combined with a mouse PK model, outcomes of afabicin doses of 0.011-190 mg/kg q6h against nine S. aureus strains in a murine thigh infection model were predicted, and thereafter refined by estimating PD parameters.

RESULTS

A sigmoid Emax model, with EC50 scaled by the MIC described the afabicin desphosphono killing in vitro. This model predicted, without parameter re-estimation, the in vivo bacterial counts at 24 h within a ±1 log margin for most dosing groups. When parameters were allowed to be estimated, EC50 was 38%-45% lower in vivo, compared with in vitro, within the studied MIC range.

CONCLUSIONS

The developed PK/PD model described the time course of afabicin activity across experimental conditions and bacterial strains. This model showed translational capacity as parameters estimated on in vitro time-kill data could well predict the in vivo outcome for a wide variety of doses in a mouse thigh infection model.

PK/PD modelling and simulation of longitudinal meropenem in vivo effects against Escherichia coli and Klebsiella pneumoniae strains with high MICs

BACKGROUND

Carbapenem-resistant bacteria pose a threat to public health. Characterising the pharmacokinetics-pharmacodynamics (PKPD) of meropenem longitudinally in vivo against resistant bacteria could provide valuable information for development and translation of carbapenem-based therapies.

OBJECTIVES

To assess the time course of meropenem effects in vivo against strains with high MIC to predict PK/PD indices and expected efficacy in patients using a modelling approach.

METHODS

A PKPD model was built on longitudinal bacterial count data to describe meropenem effects against six Escherichia coli and Klebsiella pneumoniae strains (MIC values 32-128 mg/L) in a 24 h mouse thigh infection model. The model was used to derive PK/PD indices from simulated studies in mice and to predict the efficacy of different infusion durations with high-dose meropenem (2 g q8 h/q12 h for normal/reduced kidney function) in patients.

RESULTS

Data from 592 mice were available for model development. The estimated meropenem concentration-dependent killing rate was not associated with differences in MIC. The fraction of time that unbound concentrations exceeded EC50 (fT>EC50, EC50 = 1.01 mg/L) showed higher correlations than fT>MIC. For all investigated strains, bacteriostasis at 24 h was predicted for prolonged infusions of high-dose meropenem monotherapy in >90% of patients.

CONCLUSIONS

The developed PKPD model successfully described bacterial growth and meropenem killing over time in the thigh infection model. For the investigated strains, the MIC, determined in vitro, or MIC-based PK/PD indices, did not predict in vivo response. Simulations suggested prolonged infusions of high-dose meropenem to be efficacious in patients infected by the studied strains.

Model-informed precision dosing: State of the art and future perspectives

Model-informed precision dosing (MIPD) stands as a significant development in personalized medicine to tailor drug dosing to individual patient characteristics. MIPD moves beyond traditional therapeutic drug monitoring (TDM) by integrating mathematical predictions of dosing and considering patient-specific factors (patient characteristics, drug measurements) as well as different sources of variability. For this purpose, rigorous model qualification is required for the application of MIPD in patients. This review delves into new methods in model selection and validation, also highlighting the role of machine learning in improving MIPD, the utilization of biosensors for real-time monitoring, as well as the potential of models integrating biomarkers for efficacy or toxicity for precision dosing. The clinical evidence of TDM and MIPD is discussed for various medical fields including infection medicine, oncology, transplant medicine, and inflammatory bowel diseases, thereby underscoring the role of pharmacokinetics/pharmacodynamics and specific biomarkers. Further research, particularly randomized clinical trials, is warranted to corroborate the value of MIPD in enhancing patient outcomes and advancing personalized medicine.

A retrospective cohort analysis of factors influencing continuous antibiotic therapy with ampicillin

AIMS

There are limited data on ampicillin/sulbactam, both for continuous infusion and for use in critically ill patients. We aimed to identify factors that help predict ampicillin plasma levels during continuous antibiotic therapy in intensive care patients.

MAIN METHODS

We retrospectively reviewed and retrieved a large dataset of patients who received continuous ampicillin infusion with therapeutic drug monitoring between 2015 and 2022. Patients initially received standard dosing (single shot of 2/1 g followed by continuous infusion of 6/3 g ampicillin/sulbactam per day), which was then adjusted based on the results of regular therapeutic drug monitoring and according to a target range of 30-60 mg/l (equivalent to four to eight times the minimum inhibitory concentration of ampicillin for Enterobacterales).

MAIN RESULTS

466 measurements from 225 patients (152 male, mean age 61 years) were analyzed. Initial measurements of ampicillin plasma levels were below the predefined optimal therapeutic range in 50 %, within the range in 30 % and above the range in 20 %. Target attainment increased to 70 % by the 4th measurement. There was a significant negative correlation between ampicillin plasma levels and estimated glomerular filtration rate (eGFR) (r = -0.74; p < 0.001) and, to a lesser extent, with height (r = -0.31; p < 0.001). Based on multiple linear regression, eGFR and body weight or height were the factors accounting for 63 % of the variability in the data.

SIGNIFICANCE

To optimise target achievement, ampicillin dosing in critically ill patients requires a personalized approach based on renal function. Importantly, patients with normal or augmented eGFR require higher standard doses of ampicillin/sulbactam.

Application of MIDD to accelerate the development of anti-infectives: Current status and future perspectives

This review examines the role of model-informed drug development (MIDD) in advancing antibacterial and antiviral drug development, with an emphasis on the inclusion of host system dynamics into modeling efforts. Amidst the growing challenges of multidrug resistance and diminishing market returns, innovative methodologies are crucial for continuous drug discovery and development. The MIDD approach, with its robust capacity to integrate diverse data types, offers a promising solution. In particular, the utilization of appropriate modeling and simulation techniques for better characterization and early assessment of drug resistance are discussed. The evolution of MIDD practices across different infectious disease fields is also summarized, and compared to advancements achieved in oncology. Moving forward, the application of MIDD should expand into host system dynamics as these considerations are critical for the development of "live drugs" (e.g. chimeric antigen receptor T cells or bacteriophages) to address issues like antibiotic resistance or latent viral infections.

Pharmacokinetics and pharmacodynamics of two in-feed chlortetracycline regimens provided to beef cattle

Plasma chlortetracycline (CTC) concentration data were subjected to Monte Carlo simulation of area under the concentration curve (AUC) values related to bovine respiratory disease pathogen MIC distributions to evaluate target attainment rates. Crossbred Hereford heifers were randomly assigned into two treatment groups. Treatment group (A) received chlortetracycline (CTC) at a target dose of 22 mg/kg of bodyweight daily for 5 consecutive days (n = 8) and group (B) received CTC at 350 mg/head per day (1.5 ± 0.2 mg/kg based on actual bodyweights) for seven consecutive days (n = 8). Non-compartmental analysis was used to calculate plasma-free drug CTC area under the concentration curves. The mean observed (±SD) free drug AUC values were 4.18 (±1.72) μg × h/mL and 0.30 (±0.06) μg × h/mL for treatment groups A and B, respectively. The probability of target attainment for AUC24/MIC values of 25 and 12.5 was modeled using Monte Carlo simulations. Treatment group A achieved >90% target attainment (AUC24/MIC of 25) at an MIC of 0.06 μg/mL, whereas treatment group B displayed only 12.6% target attainment (AUC24/MIC of 12.5) at the lowest MIC evaluated (0.015 μg/mL). Both in-feed CTC regimens failed to obtain a reasonable target attainment rate in light of expected MIC distributions of potential pathogens.

A Practical In Silico Method for Predicting Compound Brain Concentration-Time Profiles: Combination of PK Modeling and Machine Learning

Given the aging populations in advanced countries globally, many pharmaceutical companies have focused on developing central nervous system (CNS) drugs. However, due to the blood-brain barrier, drugs do not easily reach the target area in the brain. Although conventional screening methods for drug discovery involve the measurement of (unbound fraction of drug) brain-to-plasma partition coefficients, it is difficult to consider nonequilibrium between plasma and brain compound concentration-time profiles. To truly understand the pharmacokinetics/pharmacodynamics of CNS drugs, compound concentration-time profiles in the brain are necessary; however, such analyses are costly and time-consuming and require a significant number of animals. Therefore, in this study, we attempted to develop an in silico prediction method that does not require a large amount of experimental data by combining modeling and simulation (M&S) with machine learning (ML). First, we constructed a hybrid model linking plasma concentration-time profile to the brain compartment that takes into account the transit time and brain distribution of each compound. Using mouse plasma and brain time experimental values for 103 compounds, we determined the brain kinetic parameters of the hybrid model for each compound; this case was defined as scenario I (a positive control experiment) and included the full brain concentration-time profile data. Next, we built an ML model using chemical structure descriptors as explanatory variables and rate parameters as the target variable, and we then input the predicted values from 5-fold cross-validation (CV) into the hybrid model; this case was defined as scenario II, in which no brain compound concentration-time profile data exist. Finally, for scenario III, assuming that the brain concentration is obtained at only one time point, we used the brain kinetic parameters from the result of the 5-fold CV in scenario II as the initial values for the hybrid model and performed parameter refitting against the observed brain concentration at that time point. As a result, the RMSE/R2-values of the brain compound concentration-time profiles over time were 0.445/0.517 in scenario II and 0.246/0.805 in scenario III, indicating the method provides high accuracy and suggesting that it is a practical method for predicting brain compound concentration-time profiles.

Antibiotic dose optimisation in the critically ill: targets, evidence and future strategies

PURPOSE OF REVIEW

To highlight the recent evidence for antibiotic pharmacokinetics and pharmacodynamics (PK/PD) in enhancing patient outcomes in sepsis and septic shock. We also summarise the limitations of available data and describe future directions for research to support translation of antibiotic dose optimisation to the clinical setting.

RECENT FINDINGS

Sepsis and septic shock are associated with poor outcomes and require antibiotic dose optimisation, mostly due to significantly altered pharmacokinetics. Many studies, including some randomised controlled trials have been conducted to measure the clinical outcome effects of antibiotic dose optimisation interventions including use of therapeutic drug monitoring. Current data support antibiotic dose optimisation for the critically ill. Further investigation is required to evolve more timely and robust precision antibiotic dose optimisation approaches, and to clearly quantify whether any clinical and health-economic benefits support expanded use of this treatment intervention.

SUMMARY

Antibiotic dose optimisation appears to improve outcomes in critically ill patients with sepsis and septic shock, however further research is required to quantify the level of benefit and develop a stronger knowledge of the role of new technologies to facilitate optimised dosing.

The effects of single and multiple resistance mechanisms on bacterial response to meropenem

OBJECTIVES

Meropenem is commonly used against Pseudomonas aeruginosa. Traditionally, the time unbound antibiotic concentration exceeds the MIC (fT>MIC) is used to select carbapenem regimens. We aimed to characterize the effects of different baseline resistance mechanisms on bacterial killing and resistance emergence; evaluate whether fT>MIC can predict these effects; and, develop a novel Quantitative and Systems Pharmacology (QSP) model to describe the effects of baseline resistance mechanisms on the time-course of bacterial response.

METHODS

Seven isogenic P. aeruginosa strains with a range of resistance mechanisms and MICs were used in 10-day hollow-fiber infection model studies. Meropenem pharmacokinetic profiles were simulated for various regimens (t1/2,meropenem = 1.5 h). All viable counts on drug-free, 3 × MIC, and 5 × MIC meropenem-containing agar across all strains, five regimens, and control (n = 90 profiles) were simultaneously subjected to QSP modeling. Whole genome sequencing was completed for total population samples and emergent resistant colonies at 239 h.

RESULTS

Regimens achieving ≥98%fT>1×MIC suppressed resistance emergence of the mexR knockout strain. Even 100%fT>5 × MIC failed to achieve this against the strain with OprD loss and the ampD and mexR double-knockout strain. Baseline resistance mechanisms affected bacterial outcomes, even for strains with the same MIC. Genomic analysis revealed that pre-existing resistant subpopulations drove resistance emergence. During meropenem exposure, mutations in mexR were selected in strains with baseline oprD mutations, and vice versa, confirming these as major mechanisms of resistance emergence. Secondary mutations occurred in lysS or argS, coding for lysyl and arginyl tRNA synthetases, respectively.

DISCUSSION

The QSP model well-characterized all bacterial outcomes of the seven strains simultaneously, which fT>MIC could not.

Comparison of vancomycin assays in patients undergoing hemodialysis

Vancomycin is a glycopeptide antibiotic mainly excreted by glomerular filtration. Therefore, patients undergoing hemodialysis tend to accumulate its crystalline degradation product, which has been associated with cross-reaction in commercial immunoassays. The aim of this study was to assess the performance of two commercial immunoassays for measuring vancomycin levels in patients undergoing hemodialysis. This method-comparison study enrolled patients undergoing hemodialysis at two hospitals in Porto Alegre, Brazil. Vancomycin serum concentrations measured by Chemiluminescent Microparticle Assay (CMIA) and measured by Kinetic Interaction of Microparticles in Solution (KIMS) were compared with Liquid Chromatography coupled with Tandem Mass Spectrometry (LC-MS/MS). A total of 64 samples from 42 patients and 54 samples from 23 patients were included in CMIA and KIMS groups. Both measurements were highly correlated with LC-MS/MS, with Spearman rank correlation coefficient r = 0.840 (p < 0.001) and r = 0.926 (p < 0.001), respectively. No deviation of linearity was observed (p = 0.81 and p = 0.49, respectively). The mean difference between CMIA and LC-MS/MS was -1.19 μg/mL and between KIMS and LC-MS/MS was -2.28 μg/mL. LC-MS/MS measured levels were, on average, 2.64 % higher than CMIA and 8.81 % higher than KIMS. CMIA and KIMS revealed accurate commercial methods to measure vancomycin serum concentrations in patients undergoing hemodialysis.

Forecasting antimicrobial resistance evolution

Antimicrobial resistance (AMR) is a major global health issue. Current measures for tackling it comprise mainly the prudent use of drugs, the development of new drugs, and rapid diagnostics. Relatively little attention has been given to forecasting the evolution of resistance. Here, we argue that forecasting has the potential to be a great asset in our arsenal of measures to tackle AMR. We argue that, if successfully implemented, forecasting resistance will help to resolve the antibiotic crisis in three ways: it will (i) guide a more sustainable use (and therefore lifespan) of antibiotics and incentivize investment in drug development, (ii) reduce the spread of AMR genes and pathogenic microbes in the environment and between patients, and (iii) allow more efficient treatment of persistent infections, reducing the continued evolution of resistance. We identify two important challenges that need to be addressed for the successful establishment of forecasting: (i) the development of bespoke technology that allows stakeholders to empirically assess the risks of resistance evolving during the process of drug development and therapeutic/preventive use, and (ii) the transformative shift in mindset from the current praxis of mostly addressing the problem of antibiotic resistance a posteriori to a concept of a priori estimating, and acting on, the risks of resistance.

Dose optimization of β-lactam antibiotics in children: from population pharmacokinetics to individualized therapy

INTRODUCTION

β-Lactams are the most widely used antibiotics in children. Their optimal dosing is essential to maximize their efficacy, while minimizing the risk for toxicity and the further emergence of antimicrobial resistance. However, most β-lactams were developed and licensed long before regulatory changes mandated pharmacokinetic studies in children. As a result, pediatric dosing practices are poorly harmonized and off-label use remains common today.

AREAS COVERED

β-Lactam pharmacokinetics and dose optimization strategies in pediatrics, including fixed dose regimens, therapeutic drug monitoring, and model-informed precision dosing are reviewed.

EXPERT OPINION/COMMENTARY

Standard pediatric doses can result in subtherapeutic exposure and non-target attainment for specific patient subpopulations (neonates, critically ill children, e.g.). Such patients could benefit greatly from more individualized approaches to dose optimization, beyond a relatively simple dose adaptation based on weight, age, or renal function. In this context, Therapeutic Drug Monitoring (TDM) and Model-Informed Precision Dosing (MIPD) emerge as particularly promising avenues. Obstacles to their implementation include the lack of strong evidence of clinical benefit due to the paucity of randomized clinical trials, of standardized assays for monitoring concentrations, or of adequate markers for renal function. The development of precision medicine tools is urgently needed to individualize therapy in vulnerable pediatric subpopulations.

Impact of nutritional factors on in vitro PK/PD modelling of polymyxin B against various strains of Acinetobacter baumannii

The main objective of this study was to assess the effect of rich artificial cation-adjusted Mueller-Hinton broth (CAMHB) on the growth of three strains of Acinetobacter baumannii (ATCC 19606 and two clinical strains), either susceptible or resistant to polymyxin B (PMB), and on PMB bactericidal activity. A pharmacokinetic (PK)/pharmacodynamic (PD) modelling approach was used to characterize the effect of PMB in various conditions. Time-kill experiments were performed using undiluted CAMHB or CAMHB diluted to 50%, 25% and 10%, with or without Ca2+ and Mg2+ compensation (known to affect PMB activity), and with PMB concentrations ranging from 0.25 to 256 mg/L based on the strain's MIC. For each strain, time-kill replicates were modelled using NONMEM. Unexpectedly, dilution of CAMHB by up to 10-fold did not affect the growth rate of any of the three strains in the absence of PMB. However, the bactericidal activity of PMB increased with medium dilution, resulting in a reduction in the apparent bacterial regrowth of the various strains observed after a few hours. Data for each strain were well characterized by a PK/PD model, with two bacterial subpopulations with different susceptibility to PMB (more susceptible and less susceptible). The impact of medium dilution and cation compensation showed relatively high, unexplained between-strain variability. Further studies are needed to characterize the mechanism underlying the medium dilution effect.

Preserving the efficacy of antibiotics to tackle antibiotic resistance

Different international agencies recognize that antibiotic resistance is one of the most severe human health problems that humankind is facing. Traditionally, the introduction of new antibiotics solved this problem but various scientific and economic reasons have led to a shortage of novel antibiotics at the pipeline. This situation makes mandatory the implementation of approaches to preserve the efficacy of current antibiotics. The concept is not novel, but the only action taken for such preservation had been the 'prudent' use of antibiotics, trying to reduce the selection pressure by reducing the amount of antibiotics. However, even if antibiotics are used only when needed, this will be insufficient because resistance is the inescapable outcome of antibiotics' use. A deeper understanding of the alterations in the bacterial physiology upon acquisition of resistance and during infection will help to design improved strategies to treat bacterial infections. In this article, we discuss the interconnection between antibiotic resistance (and antibiotic activity) and bacterial metabolism, particularly in vivo, when bacteria are causing infection. We discuss as well how understanding evolutionary trade-offs, as collateral sensitivity, associated with the acquisition of resistance may help to define evolution-based therapeutic strategies to fight antibiotic resistance and to preserve currently used antibiotics.

Antimicrobial susceptibility of Mycobacterium abscessus and treatment of pulmonary and extra-pulmonary infections

BACKGROUND

Mycobacterium abscessus (MAB) is the mycobacterial species least susceptible to antimicrobials. Infections are difficult to treat, and cure rates are below 50% even after a combination of 4-5 drugs for many months.

OBJECTIVES

To examine antimicrobial susceptibilities and treatment recommendations in light of what is known about mechanisms of resistance and pharmacodynamics/pharmacokinetics (PK/PD) interactions.

SOURCES

Original papers on the topics of 'antimicrobials', 'susceptibility', 'treatment', and 'outcome' from 2019 onwards, in the context of the evidence brought by the guidelines published in 2020 for pulmonary infections.

CONTENT

MAB is susceptible in vitro to only a few antimicrobials. Breakpoints were set by the Clinical and Laboratory Standards Institute and are revised by the European Committee on Antimicrobial Susceptibility Testing for epidemiological cut-off values. Innate resistance is due to multiple resistance mechanisms involving efflux pumps, inactivating enzymes, and low drug-target affinity. In addition, MAB may display acquired resistance to macrolides and amikacin through mutations in drug binding sites. Treatment outcomes are better for macrolide-based combinations and MAB subspecies massiliense. New compounds in the family of cyclines, oxazolidinones, and penem-β-lactamase inhibitor combinations (described in another paper), as well as bedaquiline, a new antituberculous agent, are promising, but their efficacy remains to be proven. PK/PD studies, which are critical for establishing optimal dosing regimens, were mainly done for monotherapy and healthy individuals.

IMPLICATIONS

Medical evidence is poor, and randomized clinical trials or standardized cohorts are needed to compare outcomes of patients with similar underlying disease, clinical characteristics, and identified MAB subspecies/sequevar. Microbiological diagnosis and susceptibility testing need to be harmonized to enable the comparison of agents and the testing of new compounds. Testing antimicrobial combinations requires new methods, especially for PK/PD parameters. Molecular testing may help in assessing MAB resistance prior to treatment. New antimicrobials need to be systematically tested against MAB to find an effective antimicrobial regimen.

Modeling of Effectiveness of N3-Substituted Amidrazone Derivatives as Potential Agents against Gram-Positive Bacteria

Prediction of the antibacterial activity of new chemical compounds is an important task, due to the growing problem of bacterial drug resistance. Generalized linear models (GLMs) were created using 85 amidrazone derivatives based on the results of antimicrobial activity tests, determined as the minimum inhibitory concentration (MIC) against Gram-positive bacteria: Staphylococcus aureus, Enterococcus faecalis, Micrococcus luteus, Nocardia corallina, and Mycobacterium smegmatis. For the analysis of compounds characterized by experimentally measured MIC values, we included physicochemical properties (e.g., molecular weight, number of hydrogen donors and acceptors, topological polar surface area, compound percentages of carbon, nitrogen, and oxygen, melting points, and lipophilicity) as potential predictors. The presence of R1 and R2 substituents, as well as interactions between melting temperature and R1 or R2 substituents, were also considered. The set of potential predictors also included possible biological effects (e.g., antibacterial, antituberculotic) of tested compounds calculated with the PASS (Prediction of Activity Spectra for Substances) program. Using GLMs with least absolute shrinkage and selection (LASSO), least-angle regression, and stepwise selection, statistically significant models with the optimal value of the adjusted determination coefficient and of seven fit criteria were chosen, e.g., Akaike's information criterion. The most often selected variables were as follows: molecular weight, PASS_antieczematic, PASS_anti-inflam, squared melting temperature, PASS_antitumor, and experimental lipophilicity. Additionally, relevant to the bacterial strain, the interactions between melting temperature and R1 or R2 substituents were selected, indicating that the relationship between MIC and melting temperature depends on the type of R1 or R2 substituent.

Revisiting the Checkerboard to Inform Development of β-Lactam/β-Lactamase Inhibitor Combinations

A two-dimensional "checkerboard" array employing systematic titration (e.g., serial two-fold dilutions) is a well-established in vitro method for exploring the antibacterial effects of novel drug combinations. Minimum inhibitory concentrations (MICs) on the checkerboard are isoeffective points at which the antibiotic potency is the same. Representations of checkerboard MIC curves for a β-lactam and β-lactamase inhibitor combination are used in hypothetical "thought experiments" and reveal the ways in which current practices can be improved. Because different types of response (i.e., independence vs. additivity vs. one effective agent; interaction vs. noninteraction) produce different MIC curves, data from different strains/isolates should not be pooled indiscriminately, as the composition of a pooled dataset will influence any derived pharmacokinetic/pharmacodynamic (PK/PD) index. Because the β-lactamase inhibitor threshold concentration (CT) parameter is a function of the β-lactam partner dosing regimen, it is not possible to derive a universal PK/PD index target based on CT. Alternative susceptibility testing methods represent different planes through the checkerboard; a fixed ratio method is less prone to bias for all β-lactam and β-lactamase inhibitor combinations. Susceptibility test MICs will often not reflect the sensitivity of the strain/isolate to the β-lactamase inhibitor, so the use of these MICs to normalize PK/PD indices is inappropriate.

Identifying and mathematically modeling the time-course of extracellular metabolic markers associated with resistance to ceftolozane/tazobactam in Pseudomonas aeruginosa

Extracellular bacterial metabolites have potential as markers of bacterial growth and resistance emergence but have not been evaluated in dynamic in vitro studies. We investigated the dynamic metabolomic footprint of a multidrug-resistant hypermutable Pseudomonas aeruginosa isolate exposed to ceftolozane/tazobactam as continuous infusion (4.5 g/day, 9 g/day) in a hollow-fiber infection model over 7-9 days in biological replicates (n = 5). Bacterial samples were collected at 0, 7, 23, 47, 71, 95, 143, 167, 191, and 215 h, the supernatant quenched, and extracellular metabolites extracted. Metabolites were analyzed via untargeted metabolomics, including hierarchical clustering and correlation with quantified total and resistant bacterial populations. The time-courses of five (of 1,921 detected) metabolites from enriched pathways were mathematically modeled. Absorbed L-arginine and secreted L-ornithine were highly correlated with the total bacterial population (r -0.79 and 0.82, respectively, P<0.0001). Ribose-5-phosphate, sedoheptulose-7-phosphate, and trehalose-6-phosphate correlated with the resistant subpopulation (0.64, 0.64, and 0.67, respectively, P<0.0001) and were likely secreted due to resistant growth overcoming oxidative and osmotic stress induced by ceftolozane/tazobactam. Using pharmacokinetic/pharmacodynamic-based transduction models, these metabolites were successfully modeled based on the total or resistant bacterial populations. The models well described the abundance of each metabolite across the differing time-course profiles of biological replicates, based on bacterial killing and, importantly, resistant regrowth. These proof-of-concept studies suggest that further exploration is warranted to determine the generalizability of these findings. The metabolites modeled here are not exclusive to bacteria. Future studies may use this approach to identify bacteria-specific metabolites correlating with resistance, which would ultimately be extremely useful for clinical translation.

The Tradeoffs Between Persistence and Mutation Rates at Sub-Inhibitory Antibiotic Concentrations in Staphylococcus aureus

The rational design of the antibiotic treatment of bacterial infections employs these drugs to reach concentrations that exceed the minimum needed to prevent the replication of the target bacteria. However, within a treated patient, spatial and physiological heterogeneity promotes antibiotic gradients such that the concentration of antibiotics at specific sites is below the minimum needed to inhibit bacterial growth. Here, we investigate the effects of sub-inhibitory antibiotic concentrations on three parameters central to bacterial infection and the success of antibiotic treatment, using in vitro experiments with Staphylococcus aureus and mathematical-computer simulation models. Our results, using drugs of six different classes, demonstrate that exposure to sub-inhibitory antibiotic concentrations not only alters the dynamics of bacterial growth but also increases the mutation rate to antibiotic resistance and decreases the rate of production of persister cells thereby reducing the persistence level. Understanding this trade-off between mutation rates and persistence levels resulting from sub-inhibitory antibiotic exposure is crucial for optimizing, and mitigating the failure of, antibiotic therapy.

A low-footprint, fluorescence-based bacterial time-kill assay for estimating dose-dependent cell death dynamics

Dose-response curves that describe the relationship between antibiotic dose and growth rate in bacteria are commonly measured with optical density (OD) based assays. While being simple and high-throughput, any dose-dependent cell death dynamics are obscured, as OD assays in batch culture can only quantify a positive net change in cells. Time-kill experiments can be used to quantify cell death rates, but current techniques are extremely resource-intensive and may be biased by residual drug carried over into the quantification assay. Here, we report a novel, fluorescence-based time-kill assay leveraging resazurin as a viable cell count indicator. Our method improves upon previous techniques by greatly reducing the material cost and being robust to residual drug carry-over. We demonstrate our technique by quantifying a dose-response curve in Escherichia coli subject to cefotaxime, revealing dose-dependent death rates. We also show that our method is robust to extracellular debris and cell aggregation. Dose-response curves quantified with our method may provide a more accurate description of pathogen response to therapy, paving the way for more accurate integrated pharmacodynamic-pharmacokinetic studies.

Translational PK/PD for the Development of Novel Antibiotics-A Drug Developer's Perspective

Antibiotic development traditionally involved large Phase 3 programs, preceded by Phase 2 studies. Recognizing the high unmet medical need for new antibiotics and, in some cases, challenges to conducting large clinical trials, regulators created a streamlined clinical development pathway in which a lean clinical efficacy dataset is complemented by nonclinical data as supportive evidence of efficacy. In this context, translational Pharmacokinetic/Pharmacodynamic (PK/PD) plays a key role and is a major contributor to a "robust" nonclinical package. The classical PK/PD index approach, proven successful for established classes of antibiotics, is at the core of recent antibiotic approvals and the current antibacterial PK/PD guidelines by regulators. Nevertheless, in the case of novel antibiotics with a novel Mechanism of Action (MoA), there is no prior experience with the PK/PD index approach as the basis for translating nonclinical efficacy to clinical outcome, and additional nonclinical studies and PK/PD analyses might be considered to increase confidence. In this review, we discuss the value and limitations of the classical PK/PD approach and present potential risk mitigation activities, including the introduction of a semi-mechanism-based PK/PD modeling approach. We propose a general nonclinical PK/PD package from which drug developers might choose the studies most relevant for each individual candidate in order to build up a "robust" nonclinical PK/PD understanding.

A minimal PBPK model to accelerate preclinical development of drugs against tuberculosis

Introduction: Understanding drug exposure at disease target sites is pivotal to profiling new drug candidates in terms of tolerability and efficacy. Such quantification is particularly tedious for anti-tuberculosis (TB) compounds as the heterogeneous pulmonary microenvironment due to the infection may alter lung permeability and affect drug disposition. Murine models have been a longstanding support in TB research so far and are here used as human surrogates to unveil the distribution of several anti-TB compounds at the site-of-action via a novel and centralized PBPK design framework. Methods: As an intermediate approach between data-driven pharmacokinetic (PK) models and whole-body physiologically based (PB) PK models, we propose a parsimonious framework for PK investigation (minimal PBPK approach) that retains key physiological processes involved in TB disease, while reducing computational costs and prior knowledge requirements. By lumping together pulmonary TB-unessential organs, our minimal PBPK model counts 9 equations compared to the 36 of published full models, accelerating the simulation more than 3-folds in Matlab 2022b. Results: The model has been successfully tested and validated against 11 anti-TB compounds-rifampicin, rifapentine, pyrazinamide, ethambutol, isoniazid, moxifloxacin, delamanid, pretomanid, bedaquiline, OPC-167832, GSK2556286 - showing robust predictability power in recapitulating PK dynamics in mice. Structural inspections on the proposed design have ensured global identifiability and listed free fraction in plasma and blood-to-plasma ratio as top sensitive parameters for PK metrics. The platform-oriented implementation allows fast comparison of the compounds in terms of exposure and target attainment. Discrepancies in plasma and lung levels for the latest BPaMZ and HPMZ regimens have been analyzed in terms of their impact on preclinical experiment design and on PK/PD indices. Conclusion: The framework we developed requires limited drug- and species-specific information to reconstruct accurate PK dynamics, delivering a unified viewpoint on anti-TB drug distribution at the site-of-action and a flexible fit-for-purpose tool to accelerate model-informed drug design pipelines and facilitate translation into the clinic.

Enhancing the antimicrobial activity of silver nanoparticles against ESKAPE bacteria and emerging fungal pathogens by using tea extracts

The sale of antibiotics and antifungals has skyrocketed since 2020. The increasing threat of pathogens like ESKAPE bacteria (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter spp.), which are effective in evading existing antibiotics, and yeasts like Candida auris or Cryptococcus neoformans is pressing to develop efficient antimicrobial alternatives. Nanoparticles, especially silver nanoparticles (AgNPs), are believed to be promising candidates to supplement or even replace antibiotics in some applications. Here, we propose a way to increase the antimicrobial efficiency of silver nanoparticles by using tea extracts (black, green, or red) for their synthesis. This allows for using lower concentrations of nanoparticles and obtaining the antimicrobial effect in a short time. We found that AgNPs synthesized using green tea extract (G-TeaNPs) are the most effective, causing approximately 80% bacterial cell death in Gram-negative bacteria within only 3 hours at a concentration of 0.1 mg mL-1, which is better than antibiotics. Ampicillin at the same concentration (0.1 mg mL-1) and within the same duration (3 h) causes only up to 40% decrease in the number of S. aureus and E. cloacae cells (non-resistant strains). The tested silver nanoparticles also have antifungal properties and are effective against C. auris and C. neoformans, which are difficult to eradicate using other means. We established that silver nanoparticles synthesized with tea extracts have higher antibacterial properties than silver nanoparticles alone. Such formulations using inexpensive tea extracts and lower concentrations of silver nanoparticles show a promising solution to fight various pathogens.

Characterization of Pseudomonas aeruginosa resistance to ceftolozane-tazobactam due to ampC and/or ampD mutations observed during treatment using semi-mechanistic PKPD modeling

A double ampC (AmpCG183D) and ampD (AmpDH157Y) genes mutations have been identified by whole genome sequencing in a Pseudomonas aeruginosa (PaS) that became resistant (PaR) in a patient treated by ceftolozane/tazobactam (C/T). To precisely characterize the respective contributions of these mutations on the decreased susceptibility to C/T and on the parallel increased susceptibility to imipenem (IMI), mutants were generated by homologous recombination in PAO1 reference strain (PAO1- AmpCG183D, PAO1-AmpDH157Y, PAO1-AmpCG183D/AmpDH157Y) and in PaR (PaR-AmpCPaS/AmpDPaS). Sequential time-kill curve experiments were conducted on all strains and analyzed by semi-mechanistic PKPD modeling. A PKPD model with adaptation successfully described the data, allowing discrimination between initial and time-related (adaptive resistance) effects of mutations. With PAO1 and mutant-derived strains, initial EC50 values increased by 1.4, 4.1, and 29-fold after AmpCG183D , AmpDH157Y and AmpCG183D/AmpDH157Y mutations, respectively. EC50 values were increased by 320, 12.4, and 55-fold at the end of the 2 nd experiment. EC50 of PAO1-AmpCG183D/AmpDH157Y was higher than that of single mutants at any time of the experiments. Within the PaR clinical background, reversal of AmpCG183D, and AmpDH157Y mutations led to an important decrease of EC50 value, from 80.5 mg/L to 6.77 mg/L for PaR and PaR-AmpCPaS/AmpDPaS, respectively. The effect of mutations on IMI susceptibility mainly showed that the AmpCG183D mutation prevented the emergence of adaptive resistance. The model successfully described the separate and combined effect of AmpCG183D and AmpDH157Y mutations against C/T and IMI, allowing discrimination and quantification of the initial and time-related effects of mutations. This method could be reproduced in clinical strains to decipher complex resistance mechanisms.

An improved PKPD modeling approach to characterize the pharmacodynamic interaction over time between ceftazidime/avibactam and colistin from in vitro time-kill experiments against multidrug-resistant Klebsiella pneumoniae isolates

In contrast to the checkerboard method, bactericidal experiments [time-kill curves (TKCs)] allow an assessment of pharmacodynamic (PD) interactions over time. However, TKCs in combination pose interpretation problems. The objective of this study was to characterize the PD interaction over time between ceftazidime/avibactam (CZA) and colistin (CST) using TKC against four multidrug-resistant Klebsiella pneumoniae susceptible to both antibiotics and expressing a widespread carbapenemase determinant KPC-3. In vitro TKCs were performed and analyzed using pharmacokinetic/pharmacodynamic (PKPD) modeling. The general pharmacodynamic interaction model was used to characterize PD interactions between drugs. The 95% confidence intervals (95%CIs) of the expected additivity and of the observed interaction were built using parametric bootstraps and compared to evaluate the in vitro PD interaction over time. Further simulations were conducted to investigate the effect of the combination at varying concentrations typically observed in patients. Regrowth was observed in TKCs at high concentrations of drugs alone [from 4 to 32× minimum inhibitory concentrations (MIC)], while the combination systematically prevented the regrowth at concentrations close to the MIC. Significant synergy or antagonism were observed under specific conditions but overall 95%CIs overlapped widely over time indicating an additive interaction between antibiotics. Moreover, simulations of typical PK profile at standard dosages indicated that the interaction should be additive in clinical conditions. The nature of the PD interaction varied with time and concentration in TKC. Against the four K. pneumoniae isolates, the bactericidal effect of CZA + CST combination was predicted to be additive and to prevent the emergence of resistance at clinical concentrations.

The clinical application of beta-lactam antibiotic therapeutic drug monitoring in the critical care setting

Critically ill patients have increased variability in beta-lactam antibiotic (beta-lactam) exposure due to alterations in their volume of distribution and elimination. Therapeutic drug monitoring (TDM) of beta-lactams, as a dose optimization and individualization tool, has been recommended to overcome this variability in exposure. Despite its potential benefit, only a few centres worldwide perform beta-lactam TDM. An important reason for the low uptake is that the evidence for clinical benefits of beta-lactam TDM is not well established. TDM also requires the availability of specific infrastructure, knowledge and expertise. Observational studies and systematic reviews have demonstrated that TDM leads to an improvement in achieving target concentrations, a reduction in potentially toxic concentrations and improvement of clinical and microbiological outcomes. However, a small number of randomized controlled trials have not shown a mortality benefit. Opportunities for improved study design are apparent, as existing studies are limited by their inclusion of heterogeneous patient populations, including patients that may not even have infection, small sample size, variability in the types of beta-lactams included, infections caused by highly susceptible bacteria, and varied sampling, analytical and dosing algorithm methods. Here we review the fundamentals of beta-lactam TDM in critically ill patients, the existing clinical evidence and the practical aspects involved in beta-lactam TDM implementation.

Ceftolozane-Tazobactam against Pseudomonas aeruginosa Cystic Fibrosis Clinical Isolates in the Hollow-Fiber Infection Model: Challenges Imposed by Hypermutability and Heteroresistance

Pseudomonas aeruginosa remains a challenge in chronic respiratory infections in cystic fibrosis (CF). Ceftolozane-tazobactam has not yet been evaluated against multidrug-resistant hypermutable P. aeruginosa isolates in the hollow-fiber infection model (HFIM). Isolates CW41, CW35, and CW44 (ceftolozane-tazobactam MICs of 4, 4, and 2 mg/L, respectively) from adults with CF were exposed to simulated representative epithelial lining fluid pharmacokinetics of ceftolozane-tazobactam in the HFIM. Regimens were continuous infusion (CI; 4.5 g/day to 9 g/day, all isolates) and 1-h infusions (1.5 g every 8 hours and 3 g every 8 hours, CW41). Whole-genome sequencing and mechanism-based modeling were performed for CW41. CW41 (in four of five biological replicates) and CW44 harbored preexisting resistant subpopulations; CW35 did not. For replicates 1 to 4 of CW41 and CW44, 9 g/day CI decreased bacterial counts to <3 log10 CFU/mL for 24 to 48 h, followed by regrowth and resistance amplification. Replicate 5 of CW41 had no preexisting subpopulations and was suppressed below ~3 log10 CFU/mL for 120 h by 9 g/day CI, followed by resistant regrowth. Both CI regimens reduced CW35 bacterial counts to <1 log10 CFU/mL by 120 h without regrowth. These results corresponded with the presence or absence of preexisting resistant subpopulations and resistance-associated mutations at baseline. Mutations in ampC, algO, and mexY were identified following CW41 exposure to ceftolozane-tazobactam at 167 to 215 h. Mechanism-based modeling well described total and resistant bacterial counts. The findings highlight the impact of heteroresistance and baseline mutations on the effect of ceftolozane-tazobactam and limitations of MIC to predict bacterial outcomes. The resistance amplification in two of three isolates supports current guidelines that ceftolozane-tazobactam should be utilized together with another antibiotic against P. aeruginosa in CF.

Optimizing the Use of Beta-Lactam Antibiotics in Clinical Practice: A Test of Time

Despite their limitations, the pharmacokinetics (PK) and pharmacodynamics (PD) indices form the basis for our current understanding regarding antibiotic development, selection, and dose optimization. Application of PK-PD in medicine has been associated with better clinical outcome, suppression of resistance, and optimization of antibiotic consumption. Beta-lactam antibiotics remain the cornerstone for empirical and directed therapy in many patients. The percentage of time of the dosing interval that the free (unbound) drug concentration remains above the minimal inhibitory concentration (MIC) (%fT > MIC) has been considered the PK-PD index that best predicts the relationship between antibiotic exposure and killing for the beta-lactam antibiotics. Time dependence of beta-lactam antibiotics has its origin in the acylation process of the serine active site of penicillin-binding proteins, which subsequently results in bacteriostatic and bactericidal effects during the dosing interval. To enhance the likelihood of target attainment, higher doses, and prolonged infusion strategies, with/or without loading doses, have been applied to compensate for subtherapeutic levels of antibiotics related to PK-PD changes, especially in the early phase of severe sepsis. To minimize resistance and maximize clinical outcome, empirical therapy with a meropenem loading dose followed by high-dose-prolonged infusion should be considered in patients with high inoculum infections presenting as severe (Gram negative) sepsis. Subsequent de-escalation and dosing of beta-lactam antibiotics should be considered as an individualized dynamic process that requires dose adjustments throughout the time course of the disease process mediated by clinical parameters that indirectly assess PK-PD alterations.

Therapeutic drug monitoring of carbapenem antibiotics in critically ill patients: an overview of principles, recommended dosing regimens, and clinical outcomes

INTRODUCTION

The importance of antibiotic treatment for sepsis in critically ill septic patients is well established. Consistently achieving the dose of antibiotics required to optimally kill bacteria, minimize the development of resistance, and avoid toxicity is challenging. The increasing understanding of the pharmacokinetic and pharmacodynamic (PK/PD) characteristics of antibiotics, and the effects of critical illness on key PK/PD parameters, is gradually re-shaping how antibiotics are dosed in critically ill patients.

AREAS COVERED

The PK/PD characteristics of commonly used carbapenem antibiotics, the principles of the application of therapeutic drug monitoring (TDM), and current as well as future methods of utilizing TDM to optimally devise dosing regimens will be reviewed. The limitations and evidence-base supporting the use of carbapenem TDM to improve outcomes in critically ill patients will be examined.

EXPERT OPINION

It is important to understand the principles of TDM in order to correctly inform dosing regimens. Although the concept of TDM is attractive, and the ability to utilize PK software to optimize dosing in the near future is expected to rapidly increase clinicians' ability to meet pre-defined PK/PD targets more accurately, current evidence provides only limited support for the use of TDM to guide carbapenem dosing in critically ill patients.

What's the Matter with MICs: Bacterial Nutrition, Limiting Resources, and Antibiotic Pharmacodynamics

The MIC of an antibiotic required to prevent replication is used both as a measure of the susceptibility/resistance of bacteria to that drug and as the single pharmacodynamic parameter for the rational design of antibiotic treatment regimes. MICs are experimentally estimated in vitro under conditions optimal for the action of the antibiotic. However, bacteria rarely grow in these optimal conditions. Using a mathematical model of the pharmacodynamics of antibiotics, we make predictions about the nutrient dependency of bacterial growth in the presence of antibiotics. We test these predictions with experiments in broth and a glucose-limited minimal media with Escherichia coli and eight different antibiotics. Our experiments question the sufficiency of using MICs and simple pharmacodynamic functions as measures of the pharmacodynamics of antibiotics under the nutritional conditions of infected tissues. To an extent that varies among drugs: (i) the estimated MICs obtained in rich media are greater than those estimated in minimal media; (ii) exposure to these drugs increases the time before logarithmic growth starts, their lag; and (iii) the stationary-phase density of E. coli populations declines with greater sub-MIC antibiotic concentrations. We postulate a mechanism to account for the relationship between sub-MICs of antibiotics and these growth parameters. This study is limited to a single bacterial strain and two types of culture media with different nutritive content. These limitations aside, the results of our study clearly question the use of MIC as the unique pharmacodynamic parameter to develop therapeutically oriented protocols. IMPORTANCE For studies of antibiotics and how they work, the most-often used measurement of drug efficacy is the MIC. The MIC is the concentration of an antibiotic needed to inhibit bacterial growth. This parameter is critical to the design and implementation of antibiotic therapy. We provide evidence that the use of MIC as the sole measurement for antibiotic efficacy ignores important aspects of bacterial growth dynamics. Before now, there has not been a nexus between bacteria, the conditions in which they grow, and the MIC. Most importantly, few studies have considered sub-MICs of antibiotics, despite their clinical importance. Here, we explore these concentrations in-depth, and we demonstrate MIC to be an incomplete measure of how an infection will interact with a specific antibiotic. Understanding the critiques of MIC is the first of many steps needed to improve infectious disease treatment.

Prediction of Compound Plasma Concentration-Time Profiles in Mice Using Random Forest

Pharmacokinetic (PK) parameters such as clearance (CL) and volume of distribution (Vd) have been the subject of previous in silico predictive models. However, having information of the concentration over time profile explicitly can provide additional value like time above MIC or AUC, etc., to understand both the efficacy and safety-related aspects of a compound. In this work, we developed machine learning models for plasma concentration-time profiles after both i.v. and p.o. dosing for a series of 17 in-house projects. For explanatory variables, MACCS Keys chemical descriptors as well as in silico and experimental in vitro PK parameters were used. The predictive accuracy of random forest (RF), message passing neural network, 2-compartment models using estimated CL and Vdss, and an average model (as a control experiment) was investigated using 5-fold cross-validation (5-fold CV) and leave-one-project-out validation (LOPO-V). The predictive accuracy of RF in 5-fold CV for i.v. and p.o. plasma concentration-time profiles was the best among the models studied, with an RMSE for i.v. dosing at 0.08, 1, and 8 h of 0.245, 0.474, and 0.462, respectively, and an RMSE for p.o. dosing at 0.25, 1, and 8 h of 0.500, 0.612, and 0.509, respectively. Furthermore, by investigating the importance of the in vitro PK parameters using the Gini index, we observed that the general prior knowledge in ADME research was reflected well in the respective feature importance of in vitro parameters such as predicted human Vd (hVd) for the initial distribution, mouse intrinsic CL and unbound fraction of mouse plasma for the elimination process, and Caco2 permeability for the absorption process. Also, this model is the first model that can predict twin peaks in the concentration-time profile much better than a baseline compartment model. Because of its combination of sufficient accuracy and speed of prediction, we found the model to be fit-for-purpose for practical lead optimization.

Should Airway Interstitial Fluid Be Used to Evaluate the Pharmacokinetics of Macrolide Antibiotics for Dose Regimen Determination in Respiratory Infection?

Macrolide antibiotics are important drugs to combat infections. The pharmacokinetics (PK) of these drugs are essential for the determination of their optimal dose regimens, which affect antimicrobial pharmacodynamics and treatment success. For most drugs, the measurement of their concentrations in plasma/serum is the surrogate for drug concentrations in target tissues for therapy. However, for macrolides, simple reliance on total or free drug concentrations in serum/plasma might be misleading. The macrolide antibiotic concentrations of serum/plasma, interstitial fluid (ISF), and target tissue itself usually yield very different PK results. In fact, the PK of a macrolide antibiotic based on serum/plasma concentrations alone is not an ideal predictor for the in vivo efficacy against respiratory pathogens. Instead, the PK based on drug concentrations at the site of infection or ISF provide much more clinically relevant information than serum/plasma concentrations. This review aims to summarize and compare/discuss the use of drug concentrations of serum/plasma, airway ISF, and tissues for computing the PK of macrolides. A better understanding of the PK of macrolide antibiotics based on airway ISF concentrations will help optimize the antibacterial dose regimen as well as minimizing toxicity and the emergence of drug resistance in clinical practice.

All Roads Lead to Rome: Enhancing the Probability of Target Attainment with Different Pharmacokinetic/Pharmacodynamic Modelling Approaches

In light of rising antimicrobial resistance and a decreasing number of antibiotics with novel modes of action, it is of utmost importance to accelerate development of novel treatment options. One aspect of acceleration is to understand pharmacokinetics (PK) and pharmacodynamics (PD) of drugs and to assess the probability of target attainment (PTA). Several in vitro and in vivo methods are deployed to determine these parameters, such as time-kill-curves, hollow-fiber infection models or animal models. However, to date the use of in silico methods to predict PK/PD and PTA is increasing. Since there is not just one way to perform the in silico analysis, we embarked on reviewing for which indications and how PK and PK/PD models as well as PTA analysis has been used to contribute to the understanding of the PK and PD of a drug. Therefore, we examined four recent examples in more detail, namely ceftazidime-avibactam, omadacycline, gepotidacin and zoliflodacin as well as cefiderocol. Whereas the first two compound classes mainly relied on the ‘classical’ development path and PK/PD was only deployed after approval, cefiderocol highly profited from in silico techniques that led to its approval. Finally, this review shall highlight current developments and possibilities to accelerate drug development, especially for anti-infectives.

Comparison of the relative efficacy of β-lactam/β-lactamase inhibitors and carbapenems in the treatment of complicated urinary tract infections caused by ceftriaxone-non-susceptible Enterobacterales: a multicentre retrospective observational cohort study

BACKGROUND

Treating complicated urinary tract infections (cUTIs) caused by ESBL-producing Enterobacterales represents a significant clinical challenge. The present study was thus developed to explore the relative efficacy of β-lactam/β-lactamase inhibitors (BLBLIs) and carbapenems for the treatment of hospitalized patients suffering from cUTIs caused by BLBLI-susceptible ceftriaxone-non-susceptible Enterobacterales.

METHODS

Data from 557 patients from four Chinese teaching hospitals diagnosed with cUTIs caused by ceftriaxone-non-susceptible Enterobacterales from January 2017 to May 2022 were retrospectively assessed.

RESULT

The 30 day rate of treatment failure, defined by unresolved symptoms or mortality, was 10.4% (58/557). Independent predictors of 30 day treatment failure included immunocompromised status, bacteraemia, septic shock, lack of infection source control and appropriate empirical treatment. When data were controlled for potential confounding variables, BLBLI treatment exhibited a comparable risk of 14 day (OR 1.61, 95% CI 0.86-3.00, P = 0.133) and 30 day treatment failure (OR 1.45, 95% CI 0.66-3.15, P = 0.354) relative to carbapenem treatment for the overall cohort of patients. In contrast, BLBLI treatment in immunocompromised patients was associated with an elevated risk of both 14 day (OR 3.18, 95% CI 1.43-7.10, P = 0.005) and 30 day treatment failure (OR 3.06, 95% CI 1.07-8.80, P = 0.038) relative to carbapenem treatment.

CONCLUSIONS

These results suggested that carbapenem treatment may be superior to BLBLI treatment for immunocompromised patients suffering from cUTIs caused by ceftriaxone-non-susceptible Enterobacterales species. However, these results will need to be validated in appropriately constructed randomized controlled trials to ensure appropriate patient treatment.

Optimization of Therapy and the Risk of Probiotic Use during Antibiotherapy in Septic Critically Ill Patients: A Narrative Review

Optimizing the entire therapeutic regimen in septic critically ill patients should be based not only on improving antibiotic use but also on optimizing the entire therapeutic regimen by considering possible drug-drug or drug-nutrient interactions. The aim of this narrative review is to provide a comprehensive overview on recent advances to optimize the therapeutic regimen in septic critically ill patients based on a pharmacokinetics and pharmacodynamic approach. Studies on recent advances on TDM-guided drug therapy optimization based on PK and/or PD results were included. Studies on patients <18 years old or with classical TDM-guided therapy were excluded. New approaches in TDM-guided therapy in septic critically ill patients based on PK and/or PD parameters are presented for cefiderocol, carbapenems, combinations beta-lactams/beta-lactamase inhibitors (piperacillin/tazobactam, ceftolozane/tazobactam, ceftazidime/avibactam), plazomicin, oxazolidinones and polymyxins. Increased midazolam toxicity in combination with fluconazole, nephrotoxic synergism between furosemide and aminoglycosides, life-threatening hypoglycemia after fluoroquinolone and insulin, prolonged muscle weakness and/or paralysis after neuromuscular blocking agents and high-dose corticosteroids combinations are of interest in critically ill patients. In the real-world practice, the use of probiotics with antibiotics is common; even data about the risk and benefits of probiotics are currently spares and inconclusive. According to current legislation, probiotic use does not require safety monitoring, but there are reports of endocarditis, meningitis, peritonitis, or pneumonia associated with probiotics in critically ill patients. In addition, probiotics are associated with risk of the spread of antimicrobial resistance. The TDM-guided method ensures a true optimization of antibiotic therapy, and particular efforts should be applied globally. In addition, multidrug and drug-nutrient interactions in critically ill patients may increase the likelihood of adverse events and risk of death; therefore, the PK and PD particularities of the critically ill patient require a multidisciplinary approach in which knowledge of clinical pharmacology is essential.

What Is the Best Vancomycin Therapeutic Drug Monitoring Parameter to Assess Efficacy? A Critical Review of Experimental Data and Assessment of the Need for Individual Patient Minimum Inhibitory Concentration Value

Therapeutic drug monitoring is recommended for the use of vancomycin, but a recent widely publicized US medical society consensus statement has changed the suggested optimal method(s) of dose adjustment. Specifically, 24 h area under the curve (AUC₂₄)-based monitoring is has been recommended for vancomycin in preference to monitoring of trough concentrations. One reason cited for this change is the claim that AUC₂₄ is a superior correlate to efficacy than trough (Cmin). Evidence from a number of retrospective analyses have been critically reviewed and determined to have weaknesses. This narrative review focuses on the experimental studies performed in vivo in animal models of infection and in vitro to determine the extent to which these data may provide a compelling distinction between pharmacokinetic/pharmacodynamics (PKPD) parameters that may translate to clinical use in therapeutic drug monitoring. Animal in vivo studies have been presented at conferences, but no original peer reviewed studies could be found that compare various PKPD parameters. These conference proceeding findings were supportive but unconvincing, even though they were favorably presented subsequently in review articles and clinical practice guidelines. In vitro data are somewhat conflicting, but the range of concentrations may play a role in the discrepancies found. It has been suggested that MIC may be assumed to have a value of 1 mg/L; however, it can be demonstrated that this assumption may lead to considerable discrepancy from results with an actual MIC value. The AUC₂₄ parameter has been weighed against the percentage of time above the MIC (%T > MIC) as a comparative PKPD parameter, yet this may be an inappropriate comparison for vancomycin since all clinically useful dosing provides 100% T > MIC. Regardless, there is a distinction between clinical TDM parameters and PKPD parameters, so, in practice, the change to AUC₂₄:MIC based on animal experiments and in vitro evidence for vancomycin may be premature.

Pharmacokinetics of Antimicrobials in Children with Emphasis on Challenges Faced by Low and Middle Income Countries, a Clinical Review

Effective antimicrobial exposure is essential to treat infections and prevent antimicrobial resistance, both being major public health problems in low and middle income countries (LMIC). Delivery of drug concentrations to the target site is governed by dose and pharmacokinetic processes (absorption, distribution, metabolism and excretion). However, specific data on the pharmacokinetics of antimicrobials in children living in LMIC settings are scarce. Additionally, there are significant logistical constraints to therapeutic drug monitoring that further emphasize the importance of understanding pharmacokinetics and dosing in LMIC. Both malnutrition and diarrheal disease reduce the extent of enteral absorption. Multiple antiretrovirals and antimycobacterial agents, commonly used by children in low resource settings, have potential interactions with other antimicrobials. Hypoalbuminemia, which may be the result of malnutrition, nephrotic syndrome or liver failure, increases the unbound concentrations of protein bound drugs that may therefore be eliminated faster. Kidney function develops rapidly during the first years of life and different inflammatory processes commonly augment renal clearance in febrile children, potentially resulting in subtherapeutic drug concentrations if doses are not adapted. Using a narrative review approach, we outline the effects of growth, maturation and comorbidities on maturational and disease specific effects on pharmacokinetics in children in LMIC.

Optimizing Clinical Outcomes Through Rational Dosing Strategies: Roles of Pharmacokinetic/Pharmacodynamic Modeling Tools

Significant progress in previous decades has led to several methodologies developed to facilitate the design of optimal antimicrobial dosing. In this review, we highlight common pharmacokinetic/pharmacodynamic (PKPD) modeling techniques and their roles in guiding rational dosing regimen design. In the early drug development phases, dose fractionation studies identify the PKPD index most closely associated with bacterial killing. Once discerned, this index is linked to clinical efficacy end points, and classification and regression tree analysis can be used to define the PKPD target goal. Monte Carlo simulations integrate PKPD and microbiological data to identify dosing strategies with a high probability of achieving the established PKPD target. Results then determine dosing regimens to investigate and/or validate the findings of randomized controlled trials. Further improvements in PKPD modeling could lead to an era of precision dosing and personalized therapeutics.

Overcoming Antibiotic Resistance with Novel Paradigms of Antibiotic Selection

Conventional antimicrobial susceptibility tests, including phenotypic and genotypic methods, are insufficiently accurate and frequently fail to identify effective antibiotics. These methods predominantly select therapies based on the antibiotic response of only the lead bacterial pathogen within pure bacterial culture. However, this neglects the fact that, in the majority of human infections, the lead bacterial pathogens are present as a part of multispecies communities that modulate the response of these lead pathogens to antibiotics and that multiple pathogens can contribute to the infection simultaneously. This discrepancy is a major cause of the failure of antimicrobial susceptibility tests to detect antibiotics that are effective in vivo. This review article provides a comprehensive overview of the factors that are missed by conventional antimicrobial susceptibility tests and it explains how accounting for these methods can aid the development of novel diagnostic approaches.