External evaluation of published population pharmacokinetic models of polymyxin B

A systematic evaluation of population pharmacokinetic models for polymyxin B in patients with liver and/or kidney dysfunction

Polymyxin B (PMB) is considered a last-line treatment for multidrug-resistant (MDR) gram-negative bacterial infections. Model-informed precision dosing with population pharmacokinetics (PopPK) models could help to individualize PMB dosing regimens and improve therapy. However, the external prediction ability of the established PopPK models has not been fully elaborated. This study aimed to systemically evaluate eleven PMB PopPK models from ten published literature based on a new independent population, which was divided into four different populations, patients with liver dysfunction, kidney dysfunction, liver and kidney dysfunction, and normal liver and kidney function. The whole data set consisted of 146 patients with 391 PMB concentrations. The prediction- and simulation-based diagnostics and Bayesian forecasting were conducted to evaluate model predictability. In the overall evaluation process, none of the models exhibited satisfactory predictive ability in both prediction- and simulation-based diagnostic simultaneously. However, the evaluation of the models in the subgroup of patients with normal liver and kidney function revealed improved predictive performance compared to those with liver and/or kidney dysfunction. Bayesian forecasting demonstrated enhanced predictability with the incorporation of two to three prior observations. The external evaluation highlighted a lack of consistency between the prediction results of published models and the external validation dataset. Nonetheless, Bayesian forecasting holds promise in improving the predictive performance of the models, and feedback from therapeutic drug monitoring is crucial in optimizing individual dosing regimens.

Pharmacokinetics of polymyxin B in different populations: a systematic review

BACKGROUND AND OBJECTIVES

Despite being clinically utilized for the treatment of infections, the limited therapeutic range of polymyxin B (PMB), along with considerable interpatient variability in its pharmacokinetics and frequent occurrence of acute kidney injury, has significantly hindered its widespread utilization. Recent research on the population pharmacokinetics of PMB has provided valuable insights. This study aims to review relevant literature to establish a theoretical foundation for individualized clinical management.

METHODS

Follow PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, Pop-PK studies of PMB were searched in PubMed and EMBASE database systems from the inception of the database until March 2023.

RESULT

To date, a total of 22 population-based studies have been conducted, encompassing 756 subjects across six different countries. The recruited population in these studies consisted of critically infected individuals with multidrug-resistant bacteria, patients with varying renal functions, those with cystic fibrosis, kidney or lung transplant recipients, patients undergoing extracorporeal membrane oxygenation (ECMO) or continuous renal replacement therapy (CRRT), as well as individuals with obesity or pediatric populations. Among these studies, seven employed a one-compartmental model, with the range of typical clearance (CL) and volume (Vc) being 1.18-2.5L /h and 12.09-47.2 L, respectively. Fifteen studies employed a two-compartmental model, with the ranges of the clearance (CL) and volume of the central compartment (Vc), the volume of the peripheral compartment (Vp), and the intercompartment clearance (Q) were 1.27-8.65 L/h, 5.47-38.6 L, 4.52-174.69 L, and 1.34-24.3 L/h, respectively. Primary covariates identified in these studies included creatinine clearance and body weight, while other covariates considered were CRRT, albumin, age, and SOFA scores. Internal evaluation was conducted in 19 studies, with only one study being externally validated using an independent external dataset.

CONCLUSION

We conclude that small sample sizes, lack of multicentre collaboration, and patient homogeneity are the primary reasons for the discrepancies in the results of the current studies. In addition, most of the studies limited in the internal evaluation, which confined the implementation of model-informed precision dosing strategies.

Effect of continuous renal replacement therapy on the clinical efficacy and pharmacokinetics of polymyxin B in the treatment of severe pulmonary infection

Objective

This study aimed to evaluate the pharmacokinetics of polymyxin B in patients with ventilator-associated pneumonia caused by multi-drug resistant bacteria, and to analyze the effect of continuous renal replacement therapy (CRRT) on pharmacokinetics of polymyxin B.

Methods

Thirty-five patients with ventilator-associated pneumonia caused by multi-drug resistant bacteria admitted to our hospital from June 2021 to January 2022 were selected as the subjects. The patients were divided into the standard group (n = 20) and the non-standard group (n = 15) based on the factors affecting the compliance of polymyxin B plasma concentration. The patients received with polymyxin B and the plasma concentration was monitored. According to the monitoring results, they were divided into the standard group and the non-standard group, to analyze the influencing factors of polymyxin B on the blood concentration. Besides, the patients were then divided into the control group (n = 28) and the observation group (n = 7) according to whether the patients received CRRT treatment. Patients in the control group treated with polymyxin B alone, while patients in the observation group received with polymyxin B and CRRT. The general data of patients in the two groups were compared. The levels of plasma concentration of polymyxin B measured before the next administration (Cmin), peak plasma concentration of polymyxin B measured immediately after end of infusion (Cmax) and intermediate plasma concentration measured 6 h after administration (midpoint of the dosing interval) (C1/2t) were detected and compared between the two groups. Correlation between pharmacokinetics and efficacy was analyzed by Spearman correlation. The incidence of complications and the 28-day mortality rate of the two groups were recorded.

Results

The age, body mass index (BMI) and Acute Physiology and Chronic Health Evaluation II (APACHE II) scores in the non-standard group were higher than these in the standard group (p < 0.05). BMI and APACHE II scores were independent risk factors affecting the pharmacokinetics of polymyxin B in patients with severe pulmonary infection (p < 0.05). There were no significant differences in age, BMI, APACHEII score, alanine aminotransferase level, aspartate aminotransferase level, albumin level, gender and diabetes ratio between the control group and the observation group (p > 0.05). The levels of Cmin, Cmax, and C1/2t in the observation group were lower than these in the control group (p < 0.001). The response rate was 50.00% in the control group and 36.36% in the observation group (p > 0.05). The levels of Cmin, Cmax, and C1/2t in the observation group were no significant correlation with the clinical efficacy (p > 0.05), while these in the control group were positive correlation with the clinical efficacy (r = 0.485, p < 0.05). There was no significant difference in the incidence of skin pigmentation, nephrotoxicity and 28-day mortality between the two groups (p > 0.05).

Conclusion

In patients with ventilator-associated pneumonia not receiving multidrug-resistant bacteria, the rate of achieving blood drug concentration with the usual recommended dose of polymyxin B was satisfactory. However, the proportion of patients with a 6-h plasma concentration exceeding the maximum plasma concentration was high. BMI and APACHE II scores were important factors affecting the pharmacokinetics of polymyxin B. In patients undergoing CRRT, the plasma concentration of polymyxin B was significantly reduced, suggesting that in patients with severe disease, plasma concentration monitoring played an important role in drug efficacy and patient safety. In patients treated with CRRT, the dose of polymyxin B may need to be increased.

Towards precision medicine of long-acting aripiprazole through population pharmacokinetic modelling

Population pharmacokinetic (popPK) models constitute a valuable tool for characterizing the pharmacokinetic properties of once-monthly long-acting injectable aripiprazole (LAI aripiprazole) and quantifying the sources of variability in drug exposure. Our aim is to develop a popPK model of both aripiprazole and its metabolite dehydro-aripiprazole in patients treated with LAI aripiprazole, and to personalize the dosing regimen of aripiprazole across different sub-groups of patients. This is a prospective study investigating the pharmacokinetics of LAI aripiprazole. A total of 93 patients were included, 21 for model development and 71 for external model evaluation. A one-compartment model with linear absorption and elimination adequately described both aripiprazole and dehydro-aripiprazole concentrations. The weight of the patients has been shown to be the factor that most influences the absorption. However, the metabolizing phenotype for CYP2D6 and the concomitant treatment with strong inhibitors of this cytochrome have been shown to be the covariates that most influence total drug exposure. This is the first popPK model developed for LAI aripiprazole that includes aripiprazole and its main active metabolite, dehydroaripiprazole. It provides a personalized dosage recommendation that maximizes the probability of achieving optimal therapeutic concentrations and minimizes the difficulties associated with trial-and-error therapeutic strategies carried out in clinical practice.

External evaluation of the predictive performance of published population pharmacokinetic models of linezolid in adult patients

OBJECTIVES

Several linezolid population pharmacokinetic (popPK) models have been established to facilitate optimal therapy; however, their extrapolated predictive performance to other clinical sites is unknown. This study aimed to externally evaluate the predictive performance of published pharmacokinetic models of linezolid in adult patients.

METHODS

For the evaluation dataset, 150 samples were collected from 70 adult patients (72.9% of which were critically ill) treated with linezolid at our center. Twenty-five published popPK models were identified from PubMed and Embase. Model predictability was evaluated using prediction-based, simulation-based, and Bayesian forecasting-based approaches to assess model predictability.

RESULTS

Prediction-based diagnostics found that the prediction error within ±30% (F30) was less than 40% in all models, indicating unsatisfactory predictability. The simulation-based prediction- and variability-corrected visual predictive check and normalized prediction distribution error test indicated large discrepancies between the observations and simulations in most of the models. Bayesian forecasting with one or two prior observations significantly improved the models' predictive performance.

CONCLUSION

The published linezolid popPK models showed insufficient predictive ability. Therefore, their sole use is not recommended, and incorporating therapeutic drug monitoring of linezolid in clinical applications is necessary.

Pharmacokinetic and pharmacodynamic properties of polymyxin B in Escherichia coli and Klebsiella pneumoniae murine infection models

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 fCmax/MIC index of E. coli fitted best, but was only modestly higher than the R2 of %fT>MIC and fAUC/MIC (R2 0.91-0.93). For K. pneumoniae the fAUC/MIC index had the best fit, which was slightly higher than the R2 of %fT>MIC and fCmax/MIC (R2 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.

Evaluation and Validation of the Limited Sampling Strategy of Polymyxin B in Patients with Multidrug-Resistant Gram-Negative Infection

Polymyxin B (PMB) is the final option for treating multidrug-resistant Gram-negative bacterial infections. The acceptable pharmacokinetic/pharmacodynamic target is an area under the concentration–time curve across 24 h at a steady state (AUC_ss,24h) of 50–100 mg·h/L. The limited sampling strategy (LSS) is useful for predicting AUC values. However, establishing an LSS is a time-consuming process requiring a relatively dense sampling of patients. Further, given the variability among different centers, the predictability of LSSs is frequently questioned when it is extrapolated to other clinical centers. Currently, limited data are available on a reliable PMB LSS for estimating AUC_ss,24h. This study assessed and validated the practicability of LSSs established in the literature based on data from our center to provide reliable and ready-made PMB LSSs for laboratories performing therapeutic drug monitoring (TDM) of PMB. The influence of infusion and sampling time errors on predictability was also explored to obtain the optimal time points for routine PMB TDM. Using multiple regression analysis, PMB LSSs were generated from a model group of 20 patients. A validation group (10 patients) was used to validate the established LSSs. PMB LSSs from two published studies were validated using a dataset of 30 patients from our center. A population pharmacokinetic model was established to simulate the individual plasma concentration profiles for each infusion and sampling time error regimen. Pharmacokinetic data obtained from the 30 patients were fitted to a two-compartment model. Infusion and sampling time errors observed in real-world clinical practice could considerably affect the predictability of PMB LSSs. Moreover, we identified specific LSSs to be superior in predicting PMB AUC_ss,24h based on different infusion times. We also discovered that sampling time error should be controlled within −10 to 15 min to obtain better predictability. The present study provides validated PMB LSSs that can more accurately predict PMB AUC_ss,24h in routine clinical practice, facilitating PMB TDM in other laboratories and pharmacokinetics/pharmacodynamics-based clinical studies in the future.

External evaluation of published population pharmacokinetic models of posaconazole

Population pharmacokinetic (PopPK) models of posaconazole have been established to promote the precision dosing. However, the performance of these models extrapolated to other centers has not been evaluated. This study aimed to conduct an external evaluation of published posaconazole PopPK models to evaluate their predictive performance. Posaconazole PopPK models screened from the PubMed and MEDLINE databases were evaluated using an external dataset of 213 trough concentration samples collected from 97 patients. Their predictive performance was evaluated by prediction-based diagnosis (prediction error), simulation-based diagnosis (visual predictive check), and Bayesian forecasting. In addition, external cohorts with and without proton pump inhibitor were used to evaluate the models respectively. Ten models suitable for the external dataset were finally included into the study. In prediction-based diagnostics, none of the models met pre-determined criteria for predictive indexes. Only M4, M6, and M10 demonstrated favorable simulations in visual predictive check. The prediction performance of M5, M7, M8, and M9 evaluated using the cohort without proton pump inhibitor showed a significant improvement compared to that evaluated using the whole cohort. Consistent with our expectations, Bayesian forecasting significantly improved the predictive per-formance of the models with two or three prior observations. In general, the applicability of these published posaconazole PopPK models extrapolated to our center was unsatisfactory. Prospective studies combined with therapeutic drug monitoring are needed to establish a PopPK model for posaconazole in the Chinese population to promote individualized dosing.

Population pharmacokinetics of intravenous colistin sulfate and dosage optimization in critically ill patients

Aims: To explore the population pharmacokinetics of colistin sulfate and to optimize the dosing strategy for critically ill patients.

Methods: The study enrolled critically ill adult patients who received colistin sulfate intravenously for more than 72 h with at least one measurement of plasma concentration. Colistin concentrations in plasma or urine samples were measured by ultraperformance liquid chromatography tandem mass spectrometry (LC-MS/MS). The population pharmacokinetics (PPK) model for colistin sulfate was developed using the Phoenix NLME program. Monte Carlo simulation was conducted to evaluate the probability of target attainment (PTA) for optimizing dosing regimens.

Results: A total of 98 plasma concentrations from 20 patients were recorded for PPK modeling. The data were adequately described by a two-compartment model with linear elimination. During modeling, creatinine clearance (CrCL) and alanine aminotransferase (ALT) were identified as covariates of the clearance (CL) and volume of peripheral compartment distribution (V2), respectively. In addition, colistin sulfate was predominantly cleared by the nonrenal pathway with a median urinary recovery of 10.05% with large inter-individual variability. Monte Carlo simulations revealed a greater creatinine clearance associated with a higher risk of sub-therapeutic exposure to colistin sulfate. The target PTA (≥90%) of dosage regimens recommended by the label sheet was achievable only in patients infected by pathogens with MIC ≤0.5 mg/L or with renal impairments.

Conclusion: Our study showed that the dose of intravenous colistin sulfate was best adjusted by CrCL and ALT. Importantly, the recommended dosing regimen of 1.0–1.5 million units daily was insufficient for patients with normal renal functions (CrCL ≥80 ml/min) or those infected by pathogens with MIC ≥1.0 mg/L. The dosage of colistin sulfate should be adjusted according to renal function and drug exposure.

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.

Population Pharmacokinetic Evaluation with External Validation of Tacrolimus in Chinese Primary Nephrotic Syndrome Patients

PURPOSE

The generalizability of numerous tacrolimus population pharmacokinetic (popPK) models constructed to promote optimal tacrolimus dosing in patients with primary nephrotic syndrome (PNS) is unclear. This study aimed to evaluate the predictive performance of published tacrolimus popPK models for PNS patients with an external data set.

METHODS

We prospectively collected 223 concentrations from 50 Chinese adult patients with PNS who were undergoing tacrolimus treatment. Data on published tacrolimus popPK models for adults and children with PNS were extracted from the literature. Model predictability was evaluated with prediction-based and simulation-based diagnostics and Bayesian forecasting.

RESULTS

In prediction-based evaluation, none of the 11 identified published popPK models of tacrolimus had met a predefined criteria of a mean prediction error ≤  ± 20%, and the prediction error within ± 30% of the identified models didn't exceed 50%. Simulation-based diagnostics also indicated unsatisfactory predictability. Bayesian forecasting demonstrated amelioration in the model predictability with the inclusion of 2-3 prior observations. Moreover, the predictive performance of nonlinear models was not better than that of one-compartment models.

CONCLUSIONS

The prediction of tacrolimus concentrations for patients with PNS remains challenging; published models are not applicable for extrapolation to other hospitals. Bayesian forecasting significantly improved model predictability and thereby helped to individualize tacrolimus dosing.