Model-informed precision dosing (MIPD) is indicated for compounds with narrow therapeutic range and complex pharmacokinetics (PK) such as voriconazole. |
To guide evidence-based application of MIPD for voriconazole, a workflow for external evaluation and application of MIPD was implemented and it was shown that the aggregate model of voriconazole had a better predictive performance than the ‘standard’ nonlinear mixed-effects models but no evaluated published models were able to adequately describe the highly variable PK (especially after oral administration), indicating the need for more (semi-)mechanistic models combining in vitro and in vivo data to characterize the complex PK behavior for successful implementation of MIPD. |
The clinical need for MIPD was demonstrated by high intra-and interindividual variability and low PK target attainment after standard dosing, and simulations for a representative patient illustrated how a MIPD workflow could be designed to improve target attainment. Predictive performance substantially improved by including the most recent voriconazole observations in MIPD. |
1 Introduction
2 Materials and Methods
2.1 Clinical Database for Voriconazole
Study 1 | Study 2 | Study 3 | Study 4 | Study 5 | Study 6 | Study 7 | Study 8 | Study 9 | |
---|---|---|---|---|---|---|---|---|---|
Study design | Prospective | Prospective | Prospective | Prospective | Prospective | Prospective | Prospective | Prospective | Prospective |
References | [37] | [67] | [66] | [94] | [43] | [95] | [96] | [45] | [97] |
Study population | Healthy volunteers | Healthy volunteers | Healthy volunteers | Healthy volunteers | Hospitalized patients | Healthy volunteers | Healthy volunteers | Healthy volunteers | Healthy volunteers |
Number of individuals | 10 | 12 | 15 | 20 | 31 | 20 | 17 | 8 | 8 |
Voriconazole dosing | Single/multiple | Single | Single | Single | Single/multiple | Single | Single | Single/multiple | Single |
Voriconazole dosing regimens | 2 × 6 mg/kg IV q12h, 2 × 4 mg/kg IV q12h, 3 × 200 mg PO q12h | 100 mg IV, 400 mg IV | 50 mg IV/PO, 400 mg IV/PO | 400 mg IV/PO | 2 × 400 mg PO q12h, 200 mg PO q12h | 400 mg PO | 400 mg PO | 2 × 400 mg PO q12h, 15 × 200 mg PO q12h | 400 mg IV/PO |
Study visits | 7 | 4 | 4 | 2 | 1 or 2 | 2 | 3 | 6 | 1 |
Sampling schedulea,b | Dense and sparse | Dense | Dense | Dense | Dense | Dense | Dense | Dense and sparse | Dense |
Analytical method | HPLC–UV/VIS | HPLC–MS/MS | HPLC–MS/MS | HPLC–MS/MS | HPLC–MS/MS | HPLC–MS/MS | HPLC–MS/MS | HPLC–MS/MS | HPLC–MS/MS |
Lower limit of quantification [mg/L] | 0.15 | 0.0006 | 0.001 | 0.05 | 0.02 | 0.05 | 0.05 | 0.03 | 0.003 |
Continuous covariates, [unit] median (range) | |||||||||
Age [years] | 28.0 (21.0–46.0) | 30.0 (23.0–52.0) | 28.0 (22.0–52.0) | 25.0 (20.0–38.0) | 57.0 (20.0–72.0) | 25.0 (19.0–37.0) | 27.0 (22.0–35.0) | 29.5 (22.0–36.0) | 31.0 (24.0-46.0) |
Total body weight [kg] | 75.2 (65.1–83.1) | 82.4 (64.0–96.0) | 70.3 (55.1–96.3) | 70.6 (58.0–103) | 75.0 (50.0–124) | 74.0 (47.0–85.0) | 80.0 (68.0–93.0) | 82.5 (59.3–88.0) | 78.0 (55.3–89.5) |
Categorical covariates, [%] | |||||||||
Male | 100 | 91.7 | 73.3 | 60.0 | 54.8 | 65.0 | 100 | 75.0 | 50.0 |
gCYP2C19 phenotype frequencyc (gPM/gIM/gNM/gRM/gUM/NA) | 10.1/30.0/10.0/10.0/–/40.0 | –/16.7/41.7/25.0/16.7 | –/26.7/20.0/40.0/13.3 | 20.0/40.0/10.0/30.0/– | –/22.6/48.4/29.0/–/– | 15.0/45.0/25.0/15.0/–/– | 11.8/35.3/23.5/29.4/–/47.1 | –/–/–/–/–/100.0 | –/–/12.5/–/12.5/75.0 |
2.2 Identification and Implementation of Nonlinear Mixed-Effects Pharmacokinetic Models of Voriconazole
Aggregate model | Model 1 (M1) | Model 2 (M2) | |
---|---|---|---|
Study design | Meta-analysis of six published population PK models | Prospective | Prospective |
Number of study participants | 800 (pooled) | 55 | 151 |
Study population | Healthy adults (n = 119), patients with invasive fungal infections (n = 426), patients with lung/liver transplant or malignancy (n = 35), pediatrics (n = 194), adolescents (n = 26) | Patients with invasive fungal infections | Patients with invasive fungal infections |
Voriconazole dosing | Approved dosing regimen | Approved dosing regimen | Approved dosing regimen |
Study visits, median (range) | NA | 3 (1–9) | NR |
Samples | > 10,742 (pooled) | 505 | 406 |
Samples per patient, median (range) | NR | 8 (1–47) | NR |
Sampling schedulea,b (number of patients) | Dense (NR), sparse (NR) | Dense (35), sparse (20) | Dense (7), sparse (144) |
Continuous covariates, [unit] median (range) | |||
Age [years] | NR (2–99) | 58 (23–78) | 61 (18–99) |
Total body weight [kg] | NR (10–125) | 68 (42–125) | 59.1 ± 7.8 (35.0–80.0)c |
Alkaline phosphatase [U/L] | NR | NR | 104 (2.0–693.0) |
Categorical covariates, [%] | |||
Male | NR | 71.0 | 68.9 |
gCYP2C19 phenotype frequencyd (gPM/gIM/gNM/gUM) | NR | NR | 12.6/43.0/42.4/1.99 |
Omeprazole coadministration | NR | 29.1 | 58.9 |
Dexamethasone coadministration | NR | NR | 59.6 |
Rifampicin coadministration | NR | NR | NR |
Grade 3 cholestasis | NR | 12.7 | NR |
Model | NLME PK (aggregate model) | NLME PK | NLME PK |
Structural model | 2-compartment model with parallel linear and nonlinear elimination | 1-compartment model with linear elimination | 1-compartment model with linear elimination |
Identified covariates | Total body weight (allometric scaling), lean body weight, CYP2C19 gPM, CYP enzyme inhibitor/inducer | Rifampicin coadministration, severe hepatic cholestasis | Age, alkaline phosphatase, CYP2C19 gPM |
2.3 External Model Evaluation Framework
2.3.1 Deterministic Simulations
2.3.2 Model Performance Evaluation
2.3.3 Bayesian Forecasting
2.4 Clinical Impact of Bayesian Forecasting
2.5 Software
3 Results
3.1 External Evaluation Datasets
3.2 External Model Evaluation
Characteristic | Aggregate model | Model 1 (M1) | Model 2 (M2) | |||
---|---|---|---|---|---|---|
(nsamples) | MPE [%] (95% CI) | RMSE [%] | MPE [%] (95% CI) | RMSE [%] | MPE [%] (95% CI) | RMSE [%] |
Overall (n = 5334) | 36.7 (24.9, 48.4) | 438 | 151 (137, 164) | 524 | 92.5 (81.3, 104) | 429 |
Route of administration | ||||||
IV (n = 2604) | − 1.73 (− 5.01, 1.56) | 85.4 | 124 (112, 136) | 335 | 40.2 (32.9, 47.4) | 192 |
PO (n = 2730) | 73.3 (50.7, 95.9) | 607 | 176 (153, 200) | 655 | 142 (122, 163) | 569 |
Population | ||||||
Healthy volunteers (n = 5005) | 30.3 (18.0, 42.5) | 443 | 153 (139, 167) | 530 | 93.0 (81.2, 105) | 434 |
Patients (n = 329) | 134 (98.2, 169) | 354 | 111 (66.5, 156) | 427 | 85.5 (50.4, 121) | 334 |