External evaluation of population pharmacokinetic models for continuous administration of meropenem in critically ill adult patients

External validation of population pharmacokinetic models of vancomycin in postoperative neurosurgical patients

OBJECTIVE

Vancomycin is commonly used in the prevention and treatment of intracranial infections in postoperative neurosurgical patients with narrow therapeutic window and large pharmacokinetic variations. Several population pharmacokinetic (PPK) models of vancomycin have been established for neurosurgical patients. But comprehensive external evaluation has not been performed for almost all models. The objective of this study was to evaluate the predictive ability of published vancomycin PPK models in adult postoperative neurosurgical patients using an independent dataset.

METHOD

PubMed, Embase and China National Knowledge Internet databases were searched to identify published vancomycin PPK models in adult postoperative neurosurgical patients. Prediction-based and simulation-based diagnostics were used to evaluate model predictability. Bayesian forecasting was used to assess the influence of prior concentration on model prediction performance.

RESULT

A total of 763 vancomycin plasma concentrations from 493 postoperative neurosurgical patients were included in the external dataset. Eight population pharmacokinetic models of vancomycin in postoperative neurosurgical patients were included and evaluated. The model by Zhang et al. exhibited the best predictive performance in prediction-based diagnostics and prediction-corrected visual predictive checks, followed by the model by Shen et al. The predictive performance of other models was not satisfactory. The normalized predictive distribution error test shows that none of the models is suitable to describe our data. The predictive performance of vancomycin models was obviously improved by maximum a posteriori Bayesian forecasting.

CONCLUSION

The published PPK models for adult postoperative neurosurgical patients show extensive variation in predictive performance in our patients. Although it is challenging to recommend initial doses of vancomycin from these predictive models, the combination of model-based prediction and therapeutic drug monitoring can be used for dose optimization.

Systematic Evaluation of Pharmacokinetic Models for Model-Informed Precision Dosing of Meropenem in Critically Ill Patients Undergoing Continuous Renal Replacement Therapy

The altered pharmacokinetics of renally cleared drugs such as meropenem in critically ill patients receiving continuous renal replacement therapy (CRRT) might impact target attainment. Model-informed precision dosing (MIPD) is applied to individualize meropenem dosing. However, most population pharmacokinetic (PopPK) models developed to date have not yet been evaluated for MIPD. Eight PopPK models based on adult CRRT patients were identified in a systematic literature research and encoded in NONMEM 7.4. A data set of 73 CRRT patients from two different study centers was used to evaluate the predictive performance of the models using simulation and prediction-based diagnostics for i) a priori dosing based on patient characteristics only and ii) Bayesian dosing by including the first measured trough concentration. Median prediction error (MPE) for accuracy within |20%| (95% confidence intervals including zero) and median absolute prediction error (MAPE) for precision ≤ 30% were considered clinically acceptable. For a priori dosing, most models (n = 5) showed accuracy and precision MPE within |20%| and MAPE <35%. The integration of the first measured meropenem concentration improved the predictive performance of all models (median MAPE decreased from 35.4 to 25.0%; median MPE decreased from 21.8 to 4.6%). The best predictive performance for intermittent infusion was observed for the O'Jeanson model, including residual diuresis as covariate (a priori and Bayesian dosing MPE within |2%|, MAPE <30%). Our study revealed the O'Jeanson model as the best-predicting model for intermittent infusion. However, most of the selected PopPK models are suitable for MIPD in CRRT patients when one therapeutic drug monitoring sample is available.

External Evaluation of Population Pharmacokinetic Models for Precision Dosing: Current State and Knowledge Gaps

Predicting drug exposures using population pharmacokinetic models through Bayesian forecasting software can improve individual pharmacokinetic/pharmacodynamic target attainment. However, selecting the most adapted model to be used is challenging due to the lack of guidance on how to design and interpret external evaluation studies. The confusion around the choice of statistical metrics and acceptability criteria emphasises the need for further research to fill this methodological gap as there is an urgent need for the development of standards and guidelines for external evaluation studies. Herein we discuss the scientific challenges faced by pharmacometric researchers and opportunities for future research with a focus on antibiotics.

The Current Status and Future Perspectives of Beta-Lactam Therapeutic Drug Monitoring in Critically Ill Patients

Beta-lactams (BL) are the first line agents for the antibiotic management of critically ill patients with sepsis or septic shock. BL are hydrophilic antibiotics particularly subject to unpredictable concentrations in the context of critical illness because of pharmacokinetic (PK) and pharmacodynamics (PD) alterations. Thus, during the last decade, the literature focusing on the interest of BL therapeutic drug monitoring (TDM) in the intensive care unit (ICU) setting has been exponential. Moreover, recent guidelines strongly encourage to optimize BL therapy using a PK/PD approach with TDM. Unfortunately, several barriers exist regarding TDM access and interpretation. Consequently, adherence to routine TDM in ICU remains quite low. Lastly, recent clinical studies failed to demonstrate any improvement in mortality with the use of TDM in ICU patients. This review will first aim at explaining the value and complexity of the TDM process when translating it to critically ill patient bedside management, interpretating the results of clinical studies and discussion of the points which need to be addressed before conducting further TDM studies on clinical outcomes. In a second time, this review will focus on the future aspects of TDM integrating toxicodynamics, model informed precision dosing (MIPD) and “at risk” ICU populations that deserve further investigations to demonstrate positive clinical outcomes.

Individualised antimicrobial dose optimisation: a systematic review and meta-analysis of randomised controlled trials

BACKGROUND

Therapeutic drug management (TDM) and model-informed precision dosing (MIPD) allow dose individualisation to increase drug effectivity and reduce toxicity.

OBJECTIVES

Evaluate the available evidence on the clinical efficacy of individualised antimicrobial dosing optimisation.

METHODS

Data sources: Pubmed, Embase, Web of Science, and Cochrane Library databases from database inception to the 11th of November 2022.

STUDY ELIGIBILITY CRITERIA

Published peer-reviewed Randomised Controlled Trials (RCTs).

PARTICIPANTS

Human subjects aged ≥18 years receiving an antibiotic or antifungal drug.

INTERVENTIONS

Patients receiving individualised antimicrobial dose adjustment. Assessment of risk of bias: Cochrane risk-of-bias tool for randomised trials (RoB2). Methods of data synthesis: Primary outcome was the risk of mortality. Secondary outcomes included target attainment, treatment failure, clinical and microbiological cure, length of stay, treatment duration and adverse events. Effect sizes were pooled using a random-effects model. Statistical heterogeneity was assessed by inconsistency testing (I²).

RESULTS

Ten RCTs were included in the meta-analysis (1,241 participants; n= 624 in the TDM group, n = 617 in the control group). Individualised antimicrobial dose optimisation was associated with a numerical decrease in mortality (RR = 0.86; 95% CI 0.71-1.05), without reaching statistical significance. Moreover, it was associated with significantly higher target attainment rates (RR = 1.41; 95% CI, 1.13-1.76) and a significant decrease in treatment failure (RR = 0.70; 95% CI, 0.54-0.92). Individualised antimicrobial dose optimisation was also associated with improvement, but not significant in clinical cure (RR = 1.33; 95% CI, 0.94-1.33) and microbiological outcome (RR = 1.25; CI, 1.00-1.57), as well as with a significant decrease in the risk of nephrotoxicity (RR = 0.55; 95% CI, 0.31-0.97).

CONCLUSIONS

This meta-analysis demonstrates that target attainment, treatment failure, and nephrotoxicity were significantly improved in patients who underwent individualised antimicrobial dose optimisation. However, it did not show a significant decrease in mortality, clinical cure or microbiological outcome.

Meropenem Model-Informed Precision Dosing in the Treatment of Critically Ill Patients: Can We Use It?

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.

Systematic Evaluation of Voriconazole Pharmacokinetic Models without Pharmacogenetic Information for Bayesian Forecasting in Critically Ill Patients

Voriconazole (VRC) is used as first line antifungal agent against invasive aspergillosis. Model-based approaches might optimize VRC therapy. This study aimed to investigate the predictive performance of pharmacokinetic models of VRC without pharmacogenetic information for their suitability for model-informed precision dosing. Seven PopPK models were selected from a systematic literature review. A total of 66 measured VRC plasma concentrations from 33 critically ill patients was employed for analysis. The second measurement per patient was used to calculate relative Bias (rBias), mean error (ME), relative root mean squared error (rRMSE) and mean absolute error (MAE) (i) only based on patient characteristics and dosing history (a priori) and (ii) integrating the first measured concentration to predict the second concentration (Bayesian forecasting). The a priori rBias/ME and rRMSE/MAE varied substantially between the models, ranging from −15.4 to 124.6%/−0.70 to 8.01 mg/L and from 89.3 to 139.1%/1.45 to 8.11 mg/L, respectively. The integration of the first TDM sample improved the predictive performance of all models, with the model by Chen (85.0%) showing the best predictive performance (rRMSE: 85.0%; rBias: 4.0%). Our study revealed a certain degree of imprecision for all investigated models, so their sole use is not recommendable. Models with a higher performance would be necessary for clinical use.

External Evaluation of Population Pharmacokinetic Models to Inform Precision Dosing of Meropenem in Critically Ill Patients

Routine clinical meropenem therapeutic drug monitoring data can be applied to model-informed precision dosing. The current study aimed to evaluate the adequacy and predictive capabilities of the published models with routine meropenem data and identify the dosing adaptations using a priori and Bayesian estimation. For this, 14 meropenem models for the external evaluation carried out on an independent cohort of 134 patients with 205 meropenem concentrations were encoded in NONMEM 7.3. The performance was determined using: 1) prediction-based and simulation-based diagnostics; and 2) predicted meropenem concentrations by a priori prediction using patient covariates only; and Bayesian forecasting using previous observations. The clinical implications were assessed according to the required dose adaptations using the meropenem concentrations. All assessments were stratified based on the patients with or without continuous renal replacement therapy. Although none of the models passed all tests, the model by Muro et al. showed the least bias. Bayesian forecasting could improve the predictability over an a priori approach, with a relative bias of −11.63–68.89% and −302.96%–130.37%, and a relative root mean squared error of 34.99–110.11% and 14.78–241.81%, respectively. A dosing change was required in 40.00–68.97% of the meropenem observation results after Bayesian forecasting. In summary, the published models couldn’t adequately describe the meropenem pharmacokinetics of our center. Although the selection of an initial meropenem dose with a priori prediction is challenging, the further model-based analysis combining therapeutic drug monitoring could be utilized in the clinical practice of meropenem therapy.

Using a Validated Population Pharmacokinetic Model for Dosing Recommendations of Continuous Infusion Piperacillin for Critically Ill Adult Patients

BACKGROUND AND OBJECTIVE: Piperacillin is a broad-spectrum β-lactam antibiotic commonly prescribed in intensive care units. Many piperacillin population pharmacokinetic models have been published, but few underwent an external evaluation. External evaluation is an important process to determine a model's capability of being generalized to other hospitals. We aimed to assess the predictive performance of these models with an external validation dataset.

METHODS

Six models were evaluated with a dataset consisting of 30 critically ill patients (35 samples) receiving piperacillin by continuous infusion. Models were subject to prediction-based (bias and imprecision) and simulation-based evaluations. When a model had an acceptable evaluation, it was used for dosing simulations to evaluate the probability of target attainment.

RESULTS

Bias and imprecision ranged from - 35.7 to 295% and from 22.7 to 295%, respectively. The models of Klastrup et al. and of Udy et al. were acceptable according to our criteria and were used for dosing simulations. Simulations showed that a loading dose of 4 g followed by a maintenance dose of 16 g/24 h of piperacillin infused continuously was necessary to remain above a pharmacokinetic-pharmacodynamic target set as a minimal inhibitory concentration of 16 mg/L in 90% of patients, for a median patient with a creatinine clearance of 76 mL/min.

CONCLUSIONS

Despite the considerable variation in the predictive performance of the models with the external validation dataset, this study was able to validate two of these models and led to the elaboration of a dosing nomogram for piperacillin by continuous infusion that can be used by clinicians in intensive care units.

Dexmedetomidine in Children on Extracorporeal Membrane Oxygenation: Pharmacokinetic Data Exploration Using Previously Published Models

BACKGROUND

Dexmedetomidine is a sedative and analgesic increasingly used in children supported with extracorporeal membrane oxygenation (ECMO). No data is available to describe the pharmacokinetics (PK) of dexmedetomidine in this population.

METHODS

We performed a single-center prospective PK study. Children <18 years old, supported with ECMO, and on a dexmedetomidine infusion as part of their management were prospectively included. PK samples were collected. Dexmedetomidine dosing remained at the discretion of the clinical team. Six population PK models built in pediatrics were selected. Observed concentrations were compared with population predicted concentrations using the PK models.

RESULTS

Eight children contributed 30 PK samples. None of the PK models evaluated predicted the concentrations with acceptable precision and bias. Four of the six evaluated models overpredicted the concentrations. The addition of a correction factor on clearance improved models' fit. Two of the evaluated models were not applicable to our whole population age range because of their structure.

CONCLUSION

Most of the evaluated PK models overpredicted the concentrations, potentially indicating increased clearance on ECMO. Population PK models applicable to a broad spectrum of ages and pathologies are more practical in pediatric critical care settings but challenging to develop.

Aminoglycosides in the Intensive Care Unit: What Is New in Population PK Modeling?

Background: Although aminoglycosides are often used as treatment for Gram-negative infections, optimal dosing regimens remain unclear, especially in ICU patients. This is due to a large between- and within-subject variability in the aminoglycoside pharmacokinetics in this population. Objective: This review provides comprehensive data on the pharmacokinetics of aminoglycosides in patients hospitalized in the ICU by summarizing all published PopPK models in ICU patients for amikacin, gentamicin, and tobramycin. The objective was to determine the presence of a consensus on the structural model used, significant covariates included, and therapeutic targets considered during dosing regimen simulations. Method: A literature search was conducted in the Medline/PubMed database, using the terms: ‘amikacin’, ‘gentamicin’, ‘tobramycin’, ‘pharmacokinetic(s)’, ‘nonlinear mixed effect’, ‘population’, ‘intensive care’, and ‘critically ill’. Results: Nineteen articles were retained where amikacin, gentamicin, and tobramycin pharmacokinetics were described in six, 11, and five models, respectively. A two-compartment model was used to describe amikacin and tobramycin pharmacokinetics, whereas a one-compartment model majorly described gentamicin pharmacokinetics. The most recurrent significant covariates were renal clearance and bodyweight. Across all aminoglycosides, mean interindividual variability in clearance and volume of distribution were 41.6% and 22.0%, respectively. A common consensus for an optimal dosing regimen for each aminoglycoside was not reached. Conclusions: This review showed models developed for amikacin, from 2015 until now, and for gentamicin and tobramycin from the past decades. Despite the growing challenges of external evaluation, the latter should be more considered during model development. Further research including new covariates, additional simulated dosing regimens, and external validation should be considered to better understand aminoglycoside pharmacokinetics in ICU patients.

Model-Informed Therapeutic Drug Monitoring of Meropenem in Critically Ill Patients: Improvement of the Predictive Ability of Literature Models with the PRIOR Approach

BACKGROUND AND OBJECTIVE

To improve the predictive ability of literature models for model-informed therapeutic drug monitoring (TDM) of meropenem in intensive care units, we propose to tweak the literature models with the "prior approach" using a subset of the data. This study compares the predictive ability of both literature and tweaked models on TDM concentrations of meropenem in critically ill patients.

METHODS

Blood samples were collected from patients of an intensive care unit treated with intravenous meropenem. Data were split six times into an "estimation" and a "prediction" datasets. Population pharmacokinetic (popPK) models of meropenem were selected from literature. These models were run on the "estimation" dataset with the $PRIOR subroutine in NONMEM to obtain tweaked models. The literature and tweaked models were used a priori (with covariate only) and with Bayesian fitting to predict each individual concentration from the previous concentration(s). Their respective predictive abilities were compared using median relative prediction error (MDPE%) and median absolute relative prediction error (MDAPE%).

RESULTS

The total dataset was composed of 115 concentrations from 58 patients. For each of the six splits, the "estimation" and the "prediction" datasets were respectively composed of 44 and 14 patients or 45 and 13 patients. Six popPK models were selected in the literature. MDPE% and MDAPE% were globally lower for the tweaked than for the literature models, especially for a priori predictions.

CONCLUSION

The "prior approach" could be a valuable tool to improve the predictive ability of literature models, especially for a priori predictions, which are important to optimize dosing in emergency situations.