OBJECTIVES:Polymyxin B has been used for many years to treat Acinetobacter baumannii, but little is known about its pharmacodynamics (PD). We aimed to describe the PD of polymyxin B in treating A. baumannii infections. METHODS:Using the murine neutropenic thigh and lung infection models we determined the magnitude of the pharmacokinetic (PK)/PD index correlating with efficacy for eight A. baumannii strains. PD was analysed using the Emax model to determine PD targets. Using published human PK data the PTAs were calculated. RESULTS:In the thigh infection model, stasis, 1-log10 and 2-log10 kill were reached for all strains. Median (range) fAUC/MIC (area under the unbound concentration-time curve divided by the MIC) targets for stasis, 1-log10 kill and 2-log10 kill were 2.1 (1.0-11), 2.9 (1.0-15) and 4.0 (1.1-20), respectively. In contrast, in the lung model, 2-log10 kill was reached in 2/8 strains only, and there was insufficient killing at tolerated exposures to determine PD fAUC/MIC targets. For the standard human dosing regimen, PTAs derived from the thigh model were inadequate at the USCAST (United States Committee on Antimicrobial Susceptibility Testing) clinical breakpoint and a protein binding of 90%. CONCLUSIONS:Whereas polymyxin B treatment had good efficacy in the murine thigh infection model, in the lung infection model its effect was limited. PD targets, with MICs of circulating A. baumannii isolates of ≤2 mg/L, are not reached with a standard human dosing regimen and will probably exceed threshold values for toxicity. These data suggest that the efficacy of polymyxin B as monotherapy for A. baumannii infections is questionable.

BACKGROUND:Taniborbactam (VNRX-5133) is a novel boronate-based β-lactamase inhibitor that directly inhibits all four classes of β-lactamases. We studied the pharmacodynamics of taniborbactam in combination with cefepime against β-lactamase-producing Enterobacterales and Pseudomonas aeruginosa. METHODS:In vitro, cefepime/taniborbactam combination was assessed with a checkerboard broth microdilution method against two ESBL-producing Enterobacterales and one AmpC- and one VIM-producing P. aeruginosa isolates (cefepime MIC 16-256 mg/L). In vivo, neutropenic infected mice were treated with cefepime, every 2 h for 24 h, alone or in combination with taniborbactam at q2h, q4h and q8h dosing intervals. Single dose escalation and dose-fractionation experiments were conducted in order to describe plasma pharmacokinetics and pharmacodynamics of taniborbactam, respectively. RESULTS:In vitro, reversal of phenotypic resistance to cefepime was found at taniborbactam ≤0.03 and 0.25 mg/L for ESBL-producing E. coli and K. pneumoniae, and at 0.125 and 2 mg/L for VIM- and AmpC constitutively-producing P. aeruginosa, respectively. In vivo, cefepime alone marginally reached stasis against Enterobacterales and AmpC-producing P. aeruginosa. Taniborbactam restored cefepime's static effect against all isolates and its 1 log10 kill effect against all strains except the K. pneumoniae isolate. The percentage of time above free concentration threshold (%fT > Ct) best described taniborbactam efficacy (R2 = 0.50-0.80). At high cefepime exposures, area under the free concentration-time curve (fAUC) performed equally well (R2 0.49-0.70). A 40%-50% and 60%-100% fT > Ct of taniborbactam was associated with stasis and 1 log10 kill, respectively, at taniborbactam concentrations where reversal of cefepime resistance was found in vitro. CONCLUSIONS:Taniborbactam restored cefepime's activity against resistant Gram-negative bacteria in a time- and concentration-dependent manner at low and higher cefepime exposures, respectively.

BACKGROUND:NOSO-5O2 is the first clinical candidate of a new antimicrobial class, the odilorhabdins. The pharmacodynamics of NOSO-502 was studied to establish the magnitude of the pharmacodynamic index (PDI) and make human dose predictions. METHODS:In vitro experiments using different types of media were performed in time-kill curves and a pharmacokinetic model. In vivo experiments were conducted in the neutropenic murine thigh infection model. Six E. coli (MIC 1-8 mg/L) and two K. pneumoniae (MIC 1-2 mg/L) strains were used. 24 h bacteriostatic and 1- and 2-log10 kill effects were related to fAUC0-24/MIC and fAUC0-24/MIC per length of dosing interval (fAUC0-24/MIC·1/tau). Human pharmacokinetic parameters were predicted using interspecies allometric scaling and used to simulate the dose needed to reach the bacteriostatic PDI target for E. coli. RESULTS:The in vitro activity of NOSO-502 was dependent on the media and the strength of Mueller-Hinton Broth II (MHBII) used such that fAUC0-24/MIC ratios were higher when measured in 100% MHBII than 50% MHBII. In vivo for E. coli, the fAUC0-24/MIC for bacteriostatic effect and 1-log10 reduction in bacterial count were 10.7 ± 10.9 and 18.2 ± 16.5, respectively. The final human predicted parameters of the model had CV values of <20%. The human dose required to achieve the bacteriostatic fAUC0-24/MIC for each E. coli strain varied from 149 to 1717 mg/day. CONCLUSIONS:A combination of the use of PDI targets and prediction of human pharmacokinetics allowed effective doses of NOSO-502 in man to be estimated.

OBJECTIVES:To expand a translational pharmacokinetic-pharmacodynamic (PK/PD) modelling approach for assessing the combined effect of polymyxin B and minocycline against Klebsiella pneumoniae. METHODS:A PK/PD model developed based on in vitro static time-kill experiments of one strain (ARU613) was first translated to characterize that of a more susceptible strain (ARU705), and thereafter to dynamic time-kill experiments (both strains) and to a murine thigh infection model (ARU705 only). The PK/PD model was updated stepwise using accumulated data. Predictions of bacterial killing in humans were performed. RESULTS:The same model structure could be used in each translational step, with parameters being re-estimated. Dynamic data were well predicted by static-data-based models. The in vitro/in vivo differences were primarily quantified as a change in polymyxin B effect: a lower killing rate constant in vivo compared with in vitro (concentration of 3 mg/L corresponds to 0.05/h and 57/h, respectively), and a slower adaptive resistance rate (the constant in vivo was 2.5% of that in vitro). There was no significant difference in polymyxin B-minocycline interaction functions. Predictions based on both in vitro and in vivo parameters indicated that the combination has a greater-than-monotherapy antibacterial effect in humans, forecasting a reduction of approximately 5 and 2 log10 colony-forming units/mL at 24 h, respectively, under combined therapy, while the maximum bacterial load was reached in monotherapy. CONCLUSIONS:This study demonstrated the utility of the PK/PD modelling approach to understand translation of antibiotic effects across experimental systems, and showed a promising antibacterial effect of polymyxin B and minocycline in combination against K. pneumoniae.

BACKGROUND:Antibiotic combination therapy is increasingly used to treat MDR pathogens. In vitro studies suggest that the polymyxin B/rifampicin combination might be synergistic. Therefore, the pharmacodynamics of rifampicin as monotherapy and combined with polymyxin B were studied in Escherichia coli- and Klebsiella pneumoniae-infected mice. METHODS:The rifampicin pharmacokinetics (oral doses 0.5-64 mg/kg) in murine plasma were studied to estimate the exposures to rifampicin. These exposures were subsequently correlated with the antibacterial effect in a sigmoid maximum-effect model. The minimum exposures needed for a static, 1 log10 and 2 log10 kill effect in two E. coli and two K. pneumoniae strains were determined for monotherapy and the combination. The pharmacodynamic interactions between polymyxin B and rifampicin were assessed using Loewe additivity and Bliss independence in both an E. coli and a K. pneumoniae strain. RESULTS:Rifampicin monotherapy resulted in a static effect in E. coli but not against K. pneumoniae. When combined with polymyxin B, rifampicin fAUC/MIC needed for stasis, 1 log10 and 2 log10 kill effect decreased with increasing polymyxin B exposures for all strains. Synergy was confirmed in Loewe additivity (interaction indices 0.11-0.51 for E. coli and 0.04-0.19 for K. pneumoniae) and Bliss independence (267% and 863%). Maximal killing (>2 log10 kill) in combination therapy was found at rifampicin/polymyxin B fAUC/MIC of 0.68/32.56 for E. coli and 0.169/16.28 for K. pneumoniae. CONCLUSIONS:These in vivo studies confirmed that there is a clear synergistic effect between polymyxin B and rifampicin, which was stronger for the K. pneumoniae strain than for the E. coli strain.

Background: NOSO-5O2 is the first clinical candidate of a new antimicrobial class-the odilorhabdins. The pharmacodynamics of NOSO-502 were studied in vitro and in vivo to establish the pharmacodynamic index (PDI) driver. Methods: A dilutional pharmacokinetic system was used for in vitro experiments. In vivo experiments were conducted in the neutropenic murine thigh infection model. Three Escherichia coli and two Klebsiella pneumoniae strains were used. NOSO-502 pharmacokinetics (PK) in mice were described in a population PK model. Dose fractionation and escalation exposures were performed and bacteriostatic and 1 log(10) and 2 log(10) kill effects were determined using an E-max model. The coefficient of determination (R2) was used to estimate the variance that might be due to regression with various PDIs (fAUC/MIC, fC(max)/MIC, %fT(>MIC) and fAUC/MIC per length of dosing interval fAUC/MIC1/tau). Results: Exposure fractionation in both in vitro and in vivo studies indicated that 6-hourly dosing resulted in a greater reduction in K. pneumoniae bacterial burden compared with either 12- or 24-hourly dosing. Similar observations were made with three E. coli strains in vitro and a single strain of E. coli in vivo. For K. pneumoniae, R-2 for fAUC/MIC1/tau was 0.8376 in vitro and 0.6001 in vivo, which was greater than R-2 with other PDIs. Analysis of pooled in vitro data from three strains of E. coli produced an R-2 of 0.7073 for fAUC/MIC and 0.6100 for fAUC/MIC1/tau. Conclusions: NOSO-502 exhibits both dose- and time-dependent in vitro and in vivo activity against both E. coli and K. pneumoniae.

AIMS:The beta-lactam antibiotic temocillin is increasingly used to treat extended-spectrum beta-lactamase (ESBL-producing) strains; however, its protein binding is complex. This study aims to predict unbound temocillin concentrations in various participant groups to determine its impact on the probability of target attainment (PTA) and to improve dosing recommendations. METHODS:The plasma pharmacokinetics were analysed using non-linear mixed-effects modelling. Data from individuals in four groups: healthy volunteers (HV), urinary tract infection patients (UTI), ventriculitis patients and sepsis-ICU patients were included. Simulations were performed to compare the PTA for different dosing regimens and participant-groups. RESULTS:A two-compartment protein-binding model best fitted the 1085 concentrations (543 unbound, 542 total). Temocillin clearance was influenced by creatinine clearance, serum albumin (ALB) and C-reactive protein (CRP). For 2 g q8h intermittent infusion, the PTAs at an MIC of 16 mg/L were 2.3%, 39.5%, 10.0% and 72.5%, for HV, UTI, ventriculitis and sepsis-ICU patients, respectively. The effects of the covariates on the PTA were simulated for two example patients with intermittent infusion: the PTAs at an MIC of 8 mg/L for a sepsis-ICU patient (CRP 300 mg/L, albumin 15 g/L) and a mild-UTI patient (CRP 30 mg/L, albumin 35 g/L) were 94.3% and 62.4%, respectively. Continuous infusion consistently outperformed intermittent infusion in achieving the desired pharmacodynamic target (time above MIC). CONCLUSIONS:Our study underscores the significant variation in temocillin clearance and unbound fractions among different participant groups, challenging the efficacy of traditional 2 g q12h dosing. For patients with enhanced renal function and lower inflammation, continuous infusion emerges as a more effective strategy to achieve optimal target attainment.

Successful antimicrobial therapy depends on achieving optimal drug concentrations within individual patients. Inter-patient variability in pharmacokinetics (PK) and differences in pathogen susceptibility (reflected in the minimum inhibitory concentration, [MIC]) necessitate personalised approaches. Dose individualisation strategies aim to address this challenge, improving treatment outcomes and minimising the risk of toxicity and antimicrobial resistance. Therapeutic drug monitoring (TDM), with the application of population pharmacokinetic (popPK) models, enables model-informed precision dosing (MIPD). PopPK models mathematically describe drug behaviour across populations and can be combined with patient-specific TDM data to optimise dosing regimens. The integration of machine learning (ML) techniques promises to further enhance dose individualisation by identifying complex patterns within extensive datasets. Implementing these approaches involves challenges, including rigorous model selection and validation to ensure suitability for target populations. Understanding the relationship between drug exposure and clinical outcomes is crucial, as is striking a balance between model complexity and clinical usability. Additionally, regulatory compliance, outcome measurement, and practical considerations for software implementation will be addressed. Emerging technologies, such as real-time biosensors, hold the potential for revolutionising TDM by enabling continuous monitoring, immediate and frequent dose adjustments, and near patient testing. The ongoing integration of TDM, advanced modelling techniques, and ML within the evolving digital health care landscape offers a potential for enhancing antimicrobial therapy. Careful attention to model development, validation, and ethical considerations of the applied techniques is paramount for successfully optimising antimicrobial treatment for the individual patient.

Data published on the oral clavulanic acid pharmacokinetics in the pediatric population is lacking. This research aimed to describe clavulanic acid disposition following oral and intravenous administration and to provide insights into clavulanic acid exposure based on threshold concentrations for (pre-)term neonates and infants. This pooled population pharmacokinetic study combined four datasets for analysis in NONMEM v7.4.3. Clavulanic acid exposure was simulated using the percentage of time above the threshold concentrations (%fT > CT). Multiple dosage regimens and amoxicillin/clavulanic acid dosage ratios were evaluated. The cohort consisted of 89 (42 oral, 47 intravenous) subjects (403 samples) with a median (range) postnatal age 54.5 days (0-365), gestational age 37.4 weeks (23.0-41.7), and current bodyweight 3.9 kg (0.6-9.0). A one-compartment model with first-order absorption best described clavulanic acid pharmacokinetics with postnatal age as a covariate on the inter-individual variability of clearance. Oral bioavailability was 24.4% in neonates up to 10 days of age. An oral dosing regimen 90 mg/kg/day amoxicillin/clavulanic acid (4:1 ratio) resulted in 40.2% of simulated patients achieving 100% fT > CT,2mg/L. An amoxicillin/clavulanic acid ratio of 4:1 is preferred for neonatal oral regimens due to the higher exposure along the entire %fT > CT range (0-100%) as ratios higher than 4:1 might result in inadequate exposure. Our results highlight substantial exposure differences (%fT > CT) when using threshold concentrations of 1 mg/L vs. 2 mg/L. This first population pharmacokinetic model for clavulanic acid in neonates may serve as a foundational step for future research, once more precise clavulanic acid targets become available.

Objectives WCK 4282 is a novel combination of cefepime 2 g and tazobactam 2 g being developed for the treatment of infections caused by piperacillin/tazobactam-resistant ESBL infections. The dosing regimen for cefepime/tazobactam needs to be optimized to generate adequate exposures to treat infections caused by ESBL-producing pathogens resistant to both cefepime and piperacillin/tazobactam.Methods We developed pharmacokinetic population models of cefepime and tazobactam to evaluate the optimal dose adjustments in patients, including those with augmented renal clearance as well as various degrees of renal impairment, and also for those on intermittent haemodialysis. Optimal doses for various degrees of renal function were identified by determining the PTA for a range of MICs. To cover ESBL-producing pathogens with an cefepime/tazobactam MIC of 16 mg/L, a dosing regimen of 2 g q8h infused over 1.5 h resulted in a combined PTA of 99% for the mean murine 1 log10-kill target for the cefepime/tazobactam combination.Results We found that to adjust for renal function, doses need to be reduced to 1 g q8h, 500 mg q8h and 500 mg q12h for patients with CLCR of 30-59, 15-29 and 8-14 mL/min (as well as patients with intermittent haemodialysis), respectively. In patients with high to augmented CLR (estimated CLCR 120-180 mL/min), a prolonged 4 h infusion of standard dose is required.Conclusions The suggested dosing regimens will result in exposures of cefepime and tazobactam that would be adequate for infections caused by ESBL-producing pathogens with a cefepime/tazobactam MICs up to 16 mg/L.

Drain-associated intracerebral infections are life-threatening emergencies. Their treatment is challenging due to the limited penetration of antibiotics to the site of infection, resulting in potentially inadequate exposure. The emergence of multidrug-resistant pathogens might force the use of off-label intrathecal (IT) doses of antibiotics. We reviewed the literature on general aspects determining intrathecal dosing regimen, using pharmacometric knowledge. We summarised clinical experience with IT doses of antibiotics that are usually not used intrathecally, as well as the outcome of the cases and concentrations reached in the cerebrospinal fluid (CSF). Factors determining the IT regimen are the size of the ventricle system and the CSF drainage volume. With regard to pharmacometrics, pharmacokinetic/pharmacodynamic indices are likely similar to those in non-cerebral infections. The following number (N) of cases were described: benzylpenicillin (>50), ampicillin (1), ceftazidime (2), cephaloridine (56), ceftriaxone (1), cefotiam (1), meropenem (57), linezolid (1), tigecycline (15), rifampicin (3), levofloxacin (2), chloramphenicol (3) and daptomycin (8). Many side effects were reported for benzylpenicillin in the 1940–50s, but for the other antibiotics, when administered correctly, all side effects were minor and reversible. These data might help when choosing an IT dosing regimen in case there is no alternative option due to antimicrobial resistance.

The number of pharmacokinetic (PK) models of meropenem is increasing. However, the daily role of these PK models in the clinic remains unclear, especially for critically ill patients. Therefore, we evaluated the published meropenem models on real-world ICU data to assess their suitability for use in clinical practice. All models were built in NONMEM and evaluated using prediction and simulation-based diagnostics for the ability to predict the subsequent meropenem concentrations without plasma concentrations (a priori), and with plasma concentrations (a posteriori), for use in therapeutic drug monitoring (TDM). Eighteen PopPK models were included for evaluation. The a priori fit of the models, without the use of plasma concentrations, was poor, with a prediction error (PE)% of the interquartile range (IQR) exceeding the ±30% threshold. The fit improved when one to three concentrations were used to improve model predictions for TDM purposes. Two models were in the acceptable range with an IQR PE% within ±30%, when two or three concentrations were used. The role of PK models to determine the starting dose of meropenem in this population seems limited. However, certain models might be suitable for TDM-based dose adjustment using two to three plasma concentrations.

The in vitro/in vivo correlation of antifungal combination testing is necessary in order to assess the efficacy of combination regimens. We, therefore, attempted to correlate in vitro chequerboard testing of posaconazole (POS) and amphotericin B (AMB) with the in vivo outcome of combination therapy against experimental candidiasis in a neutropenic murine model. The AMB + POS combination was tested against a Candida albicans isolate. In vitro, a broth microdilution 8 × 12 chequerboard method with serial two-fold drug dilutions was used. In vivo, CD1 female neutropenic mice with experimental disseminated candidiasis were treated with i.p. AMB and p.o. POS alone and in combination at three effective doses (ED20, ED50 and ED80 corresponding to 20%, 50% and 80% of maximal effect, respectively). CFU/kidneys after 2 days were determined. The pharmacodynamic interactions were assessed based on Bliss independence interaction analysis. In vitro, a Bliss antagonism of −23% (−23% to −22%) was observed at 0.03–0.125 mg/L of AMB with 0.004–0.015 mg/L of POS, while a Bliss synergy of 27% (14%–58%) was observed at 0.008–0.03 mg/L of AMB with 0.000015–0.001 mg/L of POS. In vivo, Bliss synergy (13 ± 4%) was found when an AMB ED20 of 1 mg/kg was combined with all POS ED 0.2–0.9 mg/kg, while Bliss antagonism (35–83%) was found for the combinations of AMB ED50 2 mg/kg and ED80 3.2 mg/kg with POS ED80 of 0.9 mg/kg. Free drug serum levels of POS and AMB in in vivo synergistic and antagonistic combinations were correlated with the in vitro synergistic and antagonistic concentrations, respectively. Both synergistic and antagonistic interactions were found for the AMB + POS combination. POS compromised the efficacy of high effective AMB doses and enhanced low ineffective AMB doses. In vitro concentration-dependent interactions were correlated with in vivo dose-dependent interactions of the AMB + POS combination. In vivo interactions occurred at free drug serum levels close to in vitro interacting concentrations.

Background Although polymyxin B has been in use since the late 1950s, there have been limited studies done to unravel its pharmacokinetics (PK) and pharmacodynamics (PD) index. Methods We determined, in neutropenic infected mice, the PK, plasma protein binding and PK/PD index best correlating with efficacy for Escherichia coli and Klebsiella pneumoniae strains. Results The pharmacokinetic profile showed non-linear PK; dose was significantly correlated with absorption rate and clearance. The inhibitory sigmoid dose-effect model for the fC(max)/MIC index of E. coli fitted best, but was only modestly higher than the R-2 of %fT(>)(MIC) and fAUC/MIC (R-2 0.91-0.93). For K. pneumoniae the fAUC/MIC index had the best fit, which was slightly higher than the R-2 of %fT(>)(MIC) and fC(max)/MIC (R-2 0.85-0.91). Static targets of polymyxin B fAUC/MIC were 27.5-102.6 (median 63.5) and 5.9-60.5 (median 11.6) in E. coli and in K. pneumoniae isolates, respectively. A 1 log kill effect was only reached in two E. coli isolates and one K. pneumoniae. The PTA with the standard dosing was low for isolates with MIC >0.25 mg/L. Conclusions This study confirms that fAUC/MIC can describe the exposure-response relationship for polymyxin B. The 1 log kill effect was achieved in the minority of the isolates whereas polymyxin B PK/PD targets cannot be attained for the majority of clinical isolates with the standard dosing regimen, indicating that polymyxin B may be not effective against serious infections as monotherapy.

AbstractBackgroundStandard antibiotic dosing is not suitable for critically ill patients, due to altered pharmacokinetics (PK) in these patients. Knowledge of protein binding is important for optimizing antibiotic exposure because only the unbound fraction is pharmacologically active. If unbound fractions can be predicted, minimal sampling techniques and less costly methods can be routinely used.MethodsData from the DOLPHIN trial, a prospective randomized clinical trial that included critically ill patients, were used. Total and unbound ceftriaxone concentrations were determined using a validated UPLC-MS/MS method. A non-linear saturable binding model was made using 75% of the trough concentrations and validated on the remaining data. Our model and previously published models were tested for their performance for subtherapeutic (<1 mg/L) and high (>10 mg/L) unbound concentrations.ResultsIn total, 113 patients were sampled [Acute Physiology And Chronic Health Evaluation version 4 (APACHE IV) score 71 (IQR 55–87), albumin 28 g/L (IQR 24–32)]. This resulted in 439 samples (trough = 224, peak = 215). Unbound fractions were significantly different between samples taken at trough and peak times [10.9% (IQR 7.9–16.4) versus 19.7% (IQR 12.9–26.6), P < 0.0001], which was not explained by concentration differences. Our model and most literature models showed good sensitivity and low specificity to determine high and subtherapeutic ceftriaxone trough concentrations using only the total ceftriaxone and albumin concentrations.ConclusionsCeftriaxone protein binding is not concentration related in critically ill patients. Existing models show good ability to predict high concentrations, but low specificity in predicting subtherapeutic concentrations.

PurposeA population pharmacokinetic model of fosfomycin was developed in healthy volunteers after intravenous administration, and different dosing regimens were evaluated in terms of the probability of target attainment for Escherichia coli using both plasma and urinary pharmacokinetic/pharmacodynamic targets.MethodsEight healthy men received fosfomycin as both intermittent 8 g q8h and continuous infusion 1 g/h with a loading dose of 8 g in a crossover study design. Dense sampling was conducted during both regimens. Population pharmacokinetic modelling was performed using NONMEM. Monte Carlo simulations were conducted to evaluate the Probability of Target Attainment (PTA) of different dosing regimens using bactericidal (AUC(24h)/MIC of 83 and 75%T->MIC) and bacteriostatic (AUC(24h)/MIC of 25) plasma targets and bacteriostatic (AUC(24h)/MIC of 3994) urine target.ResultsA total of 176 plasma and 86 urine samples were available for PK analysis. A two-compartment model with a urine compartment best described the data. Glomerular filtration rate (GFR) showed a significant correlation with renal clearance and was implemented in the final model. Simulation results show that the dose of 4 g q8h reached 100% of PTA using bactericidal and bacteriostatic targets for MIC up to 16 mg/L.ConclusionFor the clinical breakpoint of 32 mg/L, the standard dosing regimen (4 g q8h) might not be sufficient to reach the bactericidal target. Higher dosing of 8 g q8h as an intermittent infusion or 0.75 g/h as a continuous infusion might be required. Continuous infusion resulted in better attainment of the %T->MIC target than intermittent infusion.

OBJECTIVES:Antibiotic dosing is not optimal in the ICU. Our recent trial investigated the effect of model-informed precision dosing (MIPD) of beta-lactam antibiotics and ciprofloxacin and showed no significant differences in clinical outcomes in all patients. This study aimed to identify subgroups of patients in which the MIPD of these antibiotics could be beneficial for clinical outcomes. METHODS:We analysed data from the DOLPHIN randomized controlled trial, which compared MIPD to standard dosing of beta-lactam antibiotics and ciprofloxacin in 388 ICU patients. We divided patients into subgroups based on baseline characteristics and assessed the effect of MIPD on 28-day mortality, 6-month mortality, change in sequential organ failure assessment (delta-SOFA), and ICU length of stay (LOS). RESULTS:We found a lower 28-day mortality in patients with a SOFA below 8 randomized to MIPD (OR 0.40; 95% CI 0.17-0.88). However, patients with a higher SOFA show an increased 28-day mortality (OR 1.94; 95% CI 1.07-3.59) in the MIPD group. ICU LOS was increased in patients receiving MIPD with a SOFA below 8 (IRR 1.36; 95% CI 1.01-1.83) and those receiving MIPD for ceftriaxone (IRR 1.76; 95% CI 1.24-2.51). Patients receiving a dose recommendation within 24 hours show a trend towards decreased ICU LOS (IRR 0.77; 95% CI 0.52-1.16) and higher delta-SOFA (estimate -1.19; 95% CI -2.98-0.60). CONCLUSIONS:ICU patients with a SOFA below 8 using MIPD had an increased ICU LOS but a lower 28-day mortality. Fast dose recommendations using MIPD of beta-lactam antibiotics and ciprofloxacin needs to be investigated in ICU patients.

AimsTo describe the pharmacokinetics (PK) of cefotaxime as pre‐emptive treatment in critically ill adult patients, including covariates and to determine the probability of target attainment (PTA) of different dosage regimens for Enterobacterales and Staphylococcus aureus.MethodsFive samples were drawn during 1 dosage interval in critically ill patients treated with cefotaxime 1 g q6h or q4h. PK parameters were estimated using NONMEM (v7.4.2). The percentage of patients reaching 100% fT>MICECOFF was used to compare different dosage regimens for Enterobacterales and S. aureus.ResultsThis study included 92 patients (437 samples). The best structural model was a 2‐compartment model with a combined error, interindividual variability on clearance, central volume and intercompartmental clearance. Correlations between interindividual variability were included. Clearance increased with higher estimated glomerular filtration rate (eGFR; creatinine clearance) and albumin concentration. For Enterobacterales, 1 g q8h reached 95% PTA and continuous infusion (CI) of 4 g 24 h−1 100% PTA at the highest eGFR and albumin concentration. For S. aureus the predefined target of 95% PTA was not reached with higher eGFR and/or albumin concentrations. CI of 6 g 24 h−1 for S. aureus resulted in a minimum of 99% PTA.ConclusionCefotaxime PK in critically ill patients was best described by a 2‐compartment model with eGFR and albumin concentration as covariates influencing clearance. For Enterobacterales 1 g q8h or CI of 4 g 24 h−1 was adequate for all combinations of eGFR and albumin concentration. For S. aureus CI of 6 g 24 h−1 would be preferred if eGFR and albumin concentration exceed 80 mL min−1 and 40 g L−1 respectively.