A Regression Approach to Visual Predictive Checks for Population Pharmacometric Models

Estimating optimal therapeutic drug levels of anti-tuberculosis medications based on treatment safety and effectiveness

Background

Therapeutic drug ranges (TDR) for standard anti-tuberculosis (TB) treatment have been determined based on expected drug levels at least 2 hours after taking the dose. In this study we constructed TDR for TB drug levels based on minimizing drug toxicity and maximizing treatment effectiveness.

Methods

Participants were followed prospectively in the Regional Prospective Observational Research in Tuberculosis (RePORT)-Brazil observational cohort study. We focused on participants with culture-confirmed drug-susceptible pulmonary TB who underwent standard TB therapy. TDR were estimated for each TB drug separately: isoniazid (INH), rifampin (RIF), ethambutol (EMB), and pyrazinamide (PZA). TDR were defined as drug concentrations that were both safe and effective: safety was defined as the probability of having an ADR of at most 5%, while effectiveness was defined as a probability of at least 95% of not having either TB treatment failure or recurrence.

Results

There were 765 plasma samples from 448 patients; 110 (24.6%) were people with HIV, 9 (2.0%) had a grade 3 or higher ADR, and 15 (3.3%) had treatment failure/recurrence. Higher drug concentrations of INH, RIF and EMB were associated with increased odds of having ADR. High concentrations of INH suggested protection against treatment failure/recurrence. Estimated therapeutic drug range for INH (2.3-8.2 µg/ml) and for RIF (0.5-7.5 µg/ml) differed from the currently recommended drug ranges (3-5 µg/ml and 8-24 µg/ml, respectively). Estimates for PZA and EMB were similar to the currently recommended values.

Conclusions

Our estimated upper end TDR were higher for INH and lower for RIF compared to currently recommended ranges.

RPEM: Randomized Monte Carlo parametric expectation maximization algorithm

Inspired from quantum Monte Carlo, by sampling discrete and continuous variables at the same time using the Metropolis-Hastings algorithm, we present a novel, fast, and accurate high performance Monte Carlo Parametric Expectation Maximization (MCPEM) algorithm. We named it Randomized Parametric Expectation Maximization (RPEM). We compared RPEM with NONMEM's Importance Sampling Method (IMP), Monolix's Stochastic Approximation Expectation Maximization (SAEM), and Certara's Quasi-Random Parametric Expectation Maximization (QRPEM) for a realistic two-compartment voriconazole model with ordinary differential equations using simulated data. We show that RPEM is as fast and as accurate as the algorithms IMP, QRPEM, and SAEM for the voriconazole model in reconstructing the population parameters, for the normal and log-normal cases.

Population pharmacokinetics of Amisulpride in Chinese patients with schizophrenia with external validation: the impact of renal function

Introduction: Amisulpride is primarily eliminated via the kidneys. Given the clear influence of renal clearance on plasma concentration, we aimed to explicitly examine the impact of renal function on amisulpride pharmacokinetics (PK) via population PK modelling and Monte Carlo simulations. Method: Plasma concentrations from 921 patients (776 in development and 145 in validation) were utilized. Results: Amisulpride PK could be described by a one-compartment model with linear elimination where estimated glomerular filtration rate, eGFR, had a significant influence on clearance. All PK parameters (estimate, RSE%) were precisely estimated: apparent volume of distribution (645 L, 18%), apparent clearance (60.5 L/h, 2%), absorption rate constant (0.106 h-1, 12%) and coefficient of renal function on clearance (0.817, 10%). No other significant covariate was found. The predictive performance of the model was externally validated. Covariate analysis showed an inverse relationship between eGFR and exposure, where subjects with eGFR= 30 mL/min/1.73 m2 had more than 2-fold increase in AUC, trough and peak concentration. Simulation results further illustrated that, given a dose of 800 mg, plasma concentrations of all patients with renal impairment would exceed 640 ng/mL. Discussion: Our work demonstrated the importance of renal function in amisulpride dose adjustment and provided a quantitative framework to guide individualized dosing for Chinese patients with schizophrenia.

Population Pharmacokinetics of Piperacillin/Tazobactam Across the Adult Lifespan

BACKGROUND AND OBJECTIVE

Piperacillin/tazobactam is one of the most frequently used antimicrobials in older adults. Using an opportunistic study design, we evaluated the pharmacokinetics of piperacillin/tazobactam as a probe drug to evaluate changes in antibacterial drug exposure and dosing requirements, including in older adults.

METHODS

A total of 121 adult patients were included. The population pharmacokinetic models that best characterized the observed plasma concentrations of piperacillin and tazobactam were one-compartment structural models with zero-order input and linear elimination.

RESULTS

Among all potential covariates, estimated creatinine clearance had the most substantial impact on the elimination clearance for both piperacillin and tazobactam. After accounting for renal function and body size, there was no remaining impact of frailty on the pharmacokinetics of piperacillin and tazobactam. Monte Carlo simulations indicated that renal function had a greater impact on the therapeutic target attainment than age, although these covariates were highly correlated. Frailty, using the Canadian Study of Health and Aging Clinical Frailty Scale, was assessed in 60 patients who were ≥ 65 years of age.

CONCLUSIONS

The simulations suggested that adults ≤ 50 years of age infected with organisms with higher minimum inhibitory concentrations may benefit from continuous piperacillin/tazobactam infusions (12 g/day of piperacillin component) or extended infusions of 4 g every 8 hours. However, for a target of 50% fT + minimum inhibitory concentration, dosing based on renal function is generally preferable to dosing by age, and simulations suggested that patients with creatinine clearance ≥ 120 mL/min may benefit from infusions of 4 g every 8 hours for organisms with higher minimum inhibitory concentrations.

Population Pharmacokinetics of Moxifloxacin in Children

BACKGROUND/OBJECTIVE

Moxifloxacin is a fluoroquinolone that is commonly used in adults, but not children. Certain clinical situations compel pediatric clinicians to use moxifloxacin, despite its potential for toxicity and limited pharmacokinetics (PK) data. Our objective was to further characterize the pharmacokinetics of moxifloxacin in children.

METHODS

We performed an opportunistic, open-label population PK study of moxifloxacin in children < 18 years of age who received moxifloxacin as part of standard care. A set of structural PK models and residual error models were explored using nonlinear mixed-effects modeling. Covariates with known biological relationships were investigated for their influence on PK parameters.

RESULTS

We obtained 43 moxifloxacin concentrations from 14 participants who received moxifloxacin intravenously (n = 8) or orally (n = 6). The dose of moxifloxacin was 10 mg/kg daily in participants ≤ 40 kg and 400 mg daily in participants > 40 kg. The population mean clearance and mean volume of distribution were 18.2 L/h and 167 L, respectively. The oral absorption was described by a first-order process. The estimated extent of oral bioavailability was highly variable (range 20-91%). Total body weight was identified as a covariate on clearance and volume of distribution, and substantially reduced the random unexplained inter-individual variability for both parameters. No participants experienced suspected serious adverse reactions related to moxifloxacin.

CONCLUSION

These data add to the existing literature to support use of moxifloxacin in children in certain situations; however, further prospective studies on the safety and efficacy of moxifloxacin are needed.

External Evaluation of a Bayesian Warfarin Dose Optimization Based on a Kinetic-Pharmacodynamic Model

Warfarin is a representative anticoagulant with large interindividual variability. The published kinetic-pharmacodynamic (K-PD) model allows the prediction of warfarin dose requirement in Swedish patients; however, its applicability in Japanese patients is not known. We evaluated the model's predictive performance in Japanese patients with various backgrounds and relationships using Bayesian parameter estimation and sampling times. A single-center retrospective observational study was conducted at Tokyo Women's Medical University, Medical Center East. The study population consisted of adult patients aged >20 years who commenced warfarin with a prothrombin time-international normalized ratio (PT-INR) from June 2015 to June 2019. The published K-PD model modified by Wright and Duffull was assessed using prediction-corrected visual predictive checks, focusing on clinical characteristics, including age, renal function, and individual prediction error. The external dataset included 232 patients who received an initial warfarin daily dose of 3.2 ± 1.28 mg with 2278 PT-INR points (median [range] follow-up period of 23 d [7-28]). Prediction-corrected visual predictive checks carried a propensity for underprediction. Additionally, age >60 years, body mass index ≤25 kg/m², and estimated glomerular filtration rate ≤60 mL/min/1.73 m² had a pronounced tendency to underpredict PT-INR. However, Bayesian prediction using four prior observations reduced underprediction. To improve the prediction performance of these special populations, further studies are required to construct a model to predict warfarin dose requirements in Japanese patients.

Pharmacometric modeling to explore 4F-PCC dosing strategies for VKA reversal in patients with INR below 2

The indicated dose of 4-factor prothrombin complex concentrate (4F-PCC) for urgent vitamin K antagonist (VKA) reversal in patients with an international normalized ratio (INR) of 2 to 4 is 25 IU/kg, but there is no indicated dose for INR <2. We explored 4F-PCC dosing strategies for baseline INR <2. Clinical trial data were used to develop pharmacometric models for Factor X (FX) and FII, accounting for covariates including baseline INR. FX and FII levels over time were simulated for mean baseline INR levels of the clinical trial participants plus baseline INRs 3.1, 1.9, and 1.6. For each INR, 200 virtual male patients were simulated to evaluate 4F-PCC doses of 35, 25, 20, 15, 12.5, and 10 IU/kg. Given an elevated bleeding risk with VKA therapy in Japanese vs Western populations, results were stratified by Japanese and non-Japanese patients. Target levels of FX and FII were ≥50% activity at 30 minutes after dosing in ≥80% of patients. FX- and FII-time models were developed with 1088 FX observations from 193 patients and 1074 FII observations from 192 patients. Model-based simulations indicated that at baseline INR 3.1, ≥80% of patients achieved ≥50% FX and FII activity with 25 IU/kg and 20 IU/kg 4F-PCC, respectively; at baseline INR 1.9, corresponding doses were 20 IU/kg and 15 IU/kg 4F-PCC, and at baseline INR 1.6, corresponding doses were 15 IU/kg, and 10 IU/kg 4F-PCC. Trends in Japanese and non-Japanese patients were similar. In conclusion, low 4F-PCC doses (15-20 IU/kg) may be sufficient to achieve hemostatic levels of FX and FII in Japanese and non-Japanese patients with baseline INR <2.

Use of Real-World Data and Pharmacometric Modeling in Support of Lacosamide Dosing in Pediatric Patients Under 4 Years of Age

The antiepileptic drug lacosamide (LCM) is approved in the United States and the European Union as monotherapy as well as adjunctive therapy for the treatment of focal seizures in children ≥4 years of age and adults. Using real-world therapeutic drug monitoring data, we performed a pharmacometric analysis for 315 pediatric patients (>1 month to <18 years of age) who received lacosamide as both monotherapy and adjunctive therapy. Population pharmacokinetic modeling was performed using nonlinear mixed-effects modeling with a 1-compartment structural model with linear elimination, where clearance and volume of distribution were allometrically scaled for body weight, with no further need for age-associated maturation functions. A covariate analysis for age, sex, race, and coadministration of other antiepileptic drugs identified phenobarbital and felbamate to significantly increase lacosamide clearance (1.71- and 1.46-fold, respectively). Based on the developed population pharmacokinetic model, simulations were performed in virtual pediatric patients to explore age-associated dose requirements to match lacosamide exposure in patient groups of different age with the exposure achieved in children ≥4 year of age with the weight-based dosing recommendations provided by the US Food and Drug Administration. Based on this approach, our analysis suggested that children ≥3 years of age needed the same dose as recommended by the US Food and Drug Administration for children ≥4 years of age (12 mg/kg/d), while children 1 to 3 years of age may need 13 to 14 mg/kg/d and infants between 1 month and 1 year of age may need 15 to 18 mg/kg/d (based on their actual age) to match the exposure seen in children ≥4 years of age.

Pan-cancer population pharmacokinetics and exposure-safety and -efficacy analyses of atezolizumab in patients with high tumor mutational burden

We retrospectively investigated the pharmacokinetics and exposure-efficacy/safety relationships of single-agent atezolizumab based on tissue tumor mutational burden (tTMB) status (high vs low [≥16 vs <16 mutations/megabase]) in a pan-tumor population from seven clinical trials. Data sources included the OAK, POPLAR, BIRCH, FIR, IMvigor210, IMvigor211, and PCD4989g studies; 986 of 2894 treated patients (34%) had TMB data. Exposure metrics were obtained using a prior two-compartment intravenous-infusion population-pharmacokinetics model, merged with prognostic, biomarker, efficacy, and safety variables. Baseline demographic/clinical characteristics and prognostic factors were well balanced between patients with high (n = 175) and low (n = 811) tTMB. Exposure was similar in the high- and low-tTMB subgroups, with no difference seen in the evaluable vs total treated populations. The objective response rate (ORR) was 29.7% vs 13.4%, complete response rate was 6.9% vs 3.2%, and median duration of response (95% CI) was 29.0 (18.6-NE) months vs 15.9 (12.5-20.5) months for patients with high-tTMB vs low-tTMB tumors, respectively. A flat exposure-efficacy relationship was seen for ORR in patients with high-tTMB based on the cycle 1 minimum atezolizumab concentration and area under the serum concentration time curve (AUC). A nonsignificant exposure-safety profile was seen for grade 3/4 adverse events and adverse events of special interest based on the AUC of atezolizumab in the high-tTMB population. tTMB is an additional predictive biological factor affecting response to atezolizumab, and quantitative investigations of atezolizumab exposure and relationships of exposure with safety and efficacy support the use of a 1200-mg, every 3-week regimen in a tumor-agnostic high-tTMB population.

Population pharmacokinetics, exposure-safety, and immunogenicity of atezolizumab in pediatric and young adult patients with cancer

BACKGROUND

The iMATRIX-atezolizumab study was a phase I/II, multicenter, open-label study designed to assess the safety and pharmacokinetics of atezolizumab in pediatric and young adult patients. We describe the pharmacokinetics (PK), exposure-safety, and immunogenicity of atezolizumab in pediatric and young adults with metastatic solid tumors or hematologic malignancies enrolled in this study.

METHODS

Patients aged < 18 years (n = 69) received a weight-adjusted dose of atezolizumab (15 mg/kg every 3 weeks [q3w]; maximum 1200 mg); those aged ≥ 18 years (n = 18) received a flat dose (1200 mg q3w). A prior two-compartment intravenous infusion input adult population-PK (popPK) model of atezolizumab was used as a basis to model pediatric data.

RESULTS

A total of 431 atezolizumab serum concentrations from 87 relapse-refractory pediatric and young adult patients enrolled in the iMATRIX-atezolizumab study were used for the popPK analysis. The dataset comprised predominantly patients aged < 18 years, including two infants aged < 2 years, with a wide body weight and age range. The clearance and volume of distribution estimates of atezolizumab were 0.217 L/day and 3.01 L, respectively. Atezolizumab geometric mean trough exposures were ~ 20% lower in pediatric patients versus young adults; this was not clinically meaningful as both groups achieved the target concentration (6 μg/mL). Safety was similar between pediatric and young adult patients with no exposure-safety relationship observed. Limited responses (4/87) precluded an exposure-response assessment on outcomes. A comparable rate (13% vs 11%) of atezolizumab anti-drug antibodies was seen in pediatric and young adult patients.

CONCLUSIONS

These findings demonstrate a similar exposure-safety profile of atezolizumab in pediatric and young adult patients, supportive of weight-based dosing in pediatric patients.

TRIAL REGISTRATION

NCT02541604.

Development of visual predictive checks accounting for multimodal parameter distributions in mixture models

The assumption of interindividual variability being unimodally distributed in nonlinear mixed effects models does not hold when the population under study displays multimodal parameter distributions. Mixture models allow the identification of parameters characteristic to a subpopulation by describing these multimodalities. Visual predictive check (VPC) is a standard simulation based diagnostic tool, but not yet adapted to account for multimodal parameter distributions. Mixture model analysis provides the probability for an individual to belong to a subpopulation (IPmix) and the most likely subpopulation for an individual to belong to (MIXEST). Using simulated data examples, two implementation strategies were followed to split the data into subpopulations for the development of mixture model specific VPCs. The first strategy splits the observed and simulated data according to the MIXEST assignment. A shortcoming of the MIXEST-based allocation strategy was a biased allocation towards the dominating subpopulation. This shortcoming was avoided by splitting observed and simulated data according to the IPmix assignment. For illustration purpose, the approaches were also applied to an irinotecan mixture model demonstrating 36% lower clearance of irinotecan metabolite (SN-38) in individuals with UGT1A1 homo/heterozygote versus wild-type genotype. VPCs with segregated subpopulations were helpful in identifying model misspecifications which were not evident with standard VPCs. The new tool provides an enhanced power of evaluation of mixture models.