Optimization of Ganciclovir and Valganciclovir Starting Dose in Children by Machine Learning.

BACKGROUND AND OBJECTIVES: Ganciclovir (GCV) and valganciclovir (VGCV) show large interindividual pharmacokinetic variability, particularly in children. The objectives of this study were (1) to develop machine learning (ML) algorithms trained on simulated pharmacokinetics profiles obtained by Monte Carlo simulations to estimate the best ganciclovir or valganciclovir starting dose in children and (2) to compare its performances on real-world profiles to previously published equation derived from literature population pharmacokinetic (POPPK) models achieving about 20% of profiles within the target. MATERIALS AND METHODS: The pharmacokinetic parameters of four literature POPPK models in addition to the World Health Organization (WHO) growth curve for children were used in the mrgsolve R package to simulate 10,800 pharmacokinetic profiles. ML algorithms were developed and benchmarked to predict the probability to reach the steady-state, area-under-the-curve target (AUC0-24 within 40-60 mg × h/L) based on demographic characteristics only. The best ML algorithm was then used to calculate the starting dose maximizing the target attainment. Performances were evaluated for ML and literature formula in a test set and in an external set of 32 and 31 actual patients (GCV and VGCV, respectively). RESULTS: A combination of Xgboost, neural network, and random forest algorithms yielded the best performances and highest target attainment in the test set (36.8% for GCV and 35.3% for the VGCV). In actual patients, the best GCV ML starting dose yielded the highest target attainment rate (25.8%) and performed equally for VGCV with the Franck model formula (35.3% for both). CONCLUSION: The ML algorithms exhibit good performances in comparison with previously validated models and should be evaluated prospectively.

Limited Sampling Strategies to Predict Ganciclovir Exposure after Valganciclovir Administration and to Reduce Monitoring Constraints in Renal Transplant Children.

Valganciclovir, the ganciclovir prodrug, is an antiviral agent used to prevent cytomegalovirus infection in renal transplant children. Therapeutic drug monitoring is still necessary to ensure optimal therapeutic area under the concentration-time curve from 0 to 24 h (AUC0-24) of 40 to 60 µg·h/mL since valganciclovir presents a high pharmacokinetic variability. To calculate ganciclovir AUC0-24 with the trapezoidal method, 7 samples are needed. The objective of this study was to develop and validate a reliable and clinically applicable limited sampling strategy (LSS) for individualizing valganciclovir dose in renal transplant children. Rich pharmacokinetic data from ganciclovir plasmatic dosages measured in renal transplant children who received valganciclovir to prevent cytomegalovirus infection at Robert Debré University Hospital were collected retrospectively. Ganciclovir AUC0-24s were calculated using the trapezoidal method. The LSS was developed using a multilinear regression approach to predict AUC0-24. The patients included were divided into two groups for model development (50 patients) and validation (30 patients). A total of 80 patients were included between February 2005 and November 2018. Multilinear regression models were developed on 50 pharmacokinetic profiles (50 patients) and validated with an independent group of 43 pharmacokinetic profiles (30 patients). Regressions based on samples collected at T1h-T4h-T8h, T2h-T4h-T8h, or T1h-T2h-T8h presented the best AUC0-24 predictive performances with an average difference between reference and predicted AUC0-24 of -0.27, 0.34, and -0.40 µg·h/mL, respectively. In conclusion, valganciclovir dosage adaptation was required in children to achieve the target AUC0-24. Three LSS models using three pharmacokinetic blood samples instead of seven will be useful for individualizing valganciclovir prophylaxis in renal transplant children.

Therapeutic Drug Monitoring of Ganciclovir: Where Are We?

BACKGROUND: Ganciclovir is the mainstay of therapy for the prophylaxis and treatment of Cytomegalovirus. However, therapy with this antiviral agent is hindered by side effects such as myelosuppression, which often leads to therapy cessation. Underdosing, as an attempt to prevent side effects, can lead to drug resistance and therapy failure. Therapeutic drug monitoring (TDM) has been used to overcome these problems. The purpose of this narrative review was to give an overview of ganciclovir TDM, available assays, population pharmacokinetic models, and discuss the current knowledge gaps. METHODS: For this narrative review, a nonsystematic literature search was performed on the PubMed database in April 2021. The following search terms were used: ganciclovir, valganciclovir, pharmacokinetics, pharmacodynamics, population pharmacokinetics, therapeutic drug monitoring, bioassay, liquid chromatography coupled with tandem mass spectrometry, liquid chromatography, chromatography, spectrophotometry, and toxicity. In addition, the reference lists of the included articles were screened. RESULTS: The most common bioanalysis method identified was liquid chromatography coupled with tandem mass spectrometry. There are different models presenting ganciclovir IC50; however, establishing a pharmacokinetic/pharmacodynamic target for ganciclovir based on preclinical data is difficult because there are no studies combining dynamic drug exposure in relation to inhibition of viral replication. The data on ganciclovir TDM show large interindividual variability, indicating that TDM may play a role in modifying the dose to reduce toxicity and prevent treatment failure related to low concentrations. The main hurdle for implementing TDM is the lack of robust data to define a therapeutic window. CONCLUSIONS: Although the pharmacokinetics (PK) involved is relatively well-described, both the pharmacodynamics (PD) and pharmacokinetic/pharmacodynamic relationship are not. This is because the studies conducted to date have mainly focused on estimating ganciclovir exposure, and owing to the limited therapeutic options for CMV infections, future studies on ganciclovir are warranted.

Population Pharmacokinetics and Bayesian Estimation of the Area Under the Concentration-Time Curve for Ganciclovir in Adult Chinese Renal Allograft Recipients After Valganciclovir Administration.

Valganciclovir (VGCV) is the prodrug of ganciclovir (GCV). The objective of this study was to establish a population pharmacokinetic model (PPK) of GCV to investigate the PK characteristics of GCV after administration of VGCV in adult Chinese renal allograft recipients. Seventy Chinese renal allograft recipients were given 450 mg (n = 41) or 900 mg (n = 29) VGCV daily. Blood samples were drawn 0-24 hours after 5 days' therapy, and GCV plasma levels were determined. The PPK model was constructed using nonlinear mixed-effects modeling, and the Bayesian estimation of AUC0-24h was constructed for an individual patient based on limited plasma samples. The PK of GCV was best described by a 2-compartment model with a first-order absorption process. The CL/F, V2 /F, Q/F, V3 /F, Ka , and lag time of GCV were 15.8 ± 0.71 L/h, 10.9 ± 2.38 L, 3.98 ± 0.40 L/h, 167 ± 44.0 L, 0.23 ± 0.0078 1/h, and 0.93 ± 0.017 hours, respectively. Clearance of creatinine was found to have a significant impact on the CL/F of GCV (P < .01). Sampling strategies consisted of plasma concentrations 0 and 2 and 0, 2, and 4 hours after VGCV administration were shown to be suitable for the estimation of the GCV AUC0-24h . The PPK model was acceptable and can describe the PK of GCV in Chinese renal transplant patients administered VGCV. The AUC0-24h of GCV in Chinese renal transplant patients can be calculated by a limited sampling strategy method.

Optimization of Ganciclovir use in allogeneic hematopoietic cell transplant recipients - the role of therapeutic drug monitoring.

Introduction: Cytomegalovirus (CMV) is an opportunistic infectious complication that can occur after allogeneic hematopoietic cell transplantation (HCT). The mainstay of treatment and prevention of this infection is ganciclovir and its ester prodrug valganciclovir. There is conflicting evidence on the clinical utility of routine ganciclovir therapeutic drug monitoring (TDM) as a means to optimize treatment.Areas covered: This review aims to describe the current knowledge of the pharmacokinetic and pharmacodynamic characteristics of ganciclovir and valganciclovir, and to explore the evidence and challenges surrounding ganciclovir TDM within the allogeneic HCT cohort.Expert opinion: Ganciclovir TDM is important to optimize efficacy in selected patient groups where there are variable pharmacokinetic factors or inadequate response to treatment. However, defined pharmacokinetic exposures which correlate with treatment efficacy and toxicity remain elusive. Prospective clinical studies in specific patient groups are required to clarify this issue. Alternative TDM targets such as the intracellular ganciclovir triphosphate should be explored as they may prove to have better correlation with clinical outcomes and adverse effects. With recent advances in CMV immune monitoring, novel approaches integrating TDM with specific CMV immune phenotyping in a predictive model will be advantageous in optimizing ganciclovir dosing by combining TDM with a risk stratification approach.

Limited sampling strategy for the estimation of the area under the concentration-time curve for ganciclovir in Chinese adult renal allograft recipients.

OBJECTIVES: Valganciclovir (VGCV) treatment is recommended for the prevention of cytomegalovirus (CMV) infection in renal allograft recipients. The aim of the present study is to investigate the pharmacokinetic characteristics of ganciclovir (GCV) after administration of VGCV in Chinese adult renal allograft recipients and estimate the exposure to GCV using limited sampling strategy (LSS). METHODS: Forty Chinese renal allograft recipients were given 450 mg or 900 mg VGCV daily. Blood samples were drawn before treatment and 0.5, 1, 1.5, 2, 3, 4, 6, 8, 12, and 24 h after 5 days of VGCV therapy, and the plasma concentrations of VGCV and GCV were determined using a liquid chromatography-mass spectrometry assay. The major pharmacokinetic parameters for GCV and VGCV were determined using a noncompartmental assay. Multiple stepwise linear regression analysis was conducted to establish a model equation for the estimation of the GCV AUC0-24 h in Chinese patients using LSS. RESULTS: In the 450 and 900 mg groups, the Cmax for VGCV was 0.2 ± 0.10 and 0.4 ± 0.16 mg/L, respectively; the Cmax for GCV was 4.2 ± 1.1 and 8.6 ± 1.6 mg/L, respectively; and the AUC0-24 h for GCV was 28.4 ± 8.4 and 60.7 ± 17.5 mg·h/L, respectively. For the establishment of LSS models, 40 patients were divided into the training group (n = 24) and validation group (n = 16). The model equations used for the calculation of AUC0-24 h for GCV were established in the training group by using multiple linear regression assay. Equations including AUC = 8.1 + 29.7 × C0 + 5.7 × C4 (r2 = 0.91) and AUC = - 0.4 + 11.0 × C0 + 2.1 × C2 + 13.7 × C8 (r2 = 0.98) were acceptable. The %MPE and %MAPE values obtained from the validation group for the two model equations were 5.89 ± 14.5% and 12.1 ± 9.53%, and - 1.30 ± 4.40% and 3.28 ± 3.11%, respectively. CONCLUSIONS: The LSS models that included C0 and C4 or C0, C2, and C8 in the estimation of AUC0-24 h for GCV had favorable performance and can be used for therapeutic drug monitoring in the prevention of CMV infection using VGCV in Chinese renal allograft recipients.

Pediatric Dosing of Ganciclovir and Valganciclovir: How Model-Based Simulations Can Prevent Underexposure and Potential Treatment Failure.

Intravenous ganciclovir and oral valganciclovir are effective in the prevention and treatment of pediatric cytomegalovirus (CMV) infection but various dosing regimens are used in medical practice. Population pharmacokinetic (PopPK) model-based simulations were used to propose a new ganciclovir pediatric dosing algorithm for regulatory review and to evaluate the approved valganciclovir pediatric dosing algorithm against published dosing recommendations derived from quantitative approaches. Oral valganciclovir (mg = 7 × body surface area (BSA) × creatinine clearance according to the Schwarz formula (CrCLS) daily) and i.v. ganciclovir (mg = 3 × BSA × CrCLS daily) are effective in reaching ganciclovir target exposure for the prevention of CMV (area under the concentration-time curve (AUC)0-24 40-60 µg âˆ™ hour/mL) in most pediatric patients across the full pediatric age range. In contrast, ganciclovir and valganciclovir dosing based on body weight, as commonly used in medical practice, leads to underexposure, particularly in younger pediatric patients. This example shows that model-based dosing algorithms built on clinical pharmacology and implemented using good modeling practice can prevent underexposure and reduce the risk of treatment failure in pediatric patients.

Valganciclovir Pharmacokinetics in Patients Receiving Oral Prophylaxis Following Kidney Transplantation and Model-Based Predictions of Optimal Dosing Regimens.

BACKGROUND AND OBJECTIVES: Valganciclovir is used as oral prophylaxis for cytomegalovirus (CMV) infection in kidney transplant recipients. However, limited pharmacokinetic data exist to guide dosing in this patient group. This study aimed to describe the population pharmacokinetics of valganciclovir in a large sample of kidney transplant recipients and predict optimal dosing based on Monte Carlo simulations. METHODS: Therapeutic drug monitoring (TDM) data from adult kidney transplant recipients who received valganciclovir prophylaxis during a 10-year study period were collected retrospectively. A non-parametric pharmacokinetic analysis and Monte Carlo simulations to determine the probabilities of reaching an area under the drug concentration-time curve (AUC) target of 40-50 mg·h/L with various dosing regimens at different levels of renal function were conducted using the Pmetrics™ package for R. RESULTS: This study included 792 ganciclovir concentration measurements derived from 97 patients. A one-compartment oral absorption model best described the data. The final covariate model was as follows: CL(ganciclovir) = TVCL × (CLCR/51)0.75, where CL is the clearance, TVCL is the typical value of ganciclovir clearance, creatinine clearance (CLCR) according to the Cockcroft-Gaultt equation and 51 is the mean CLCR determined in the study. In the simulations, the probability of reaching the targeted AUC was insufficient when using the recommended dosing regimens for prophylaxis, especially in patients with impaired renal function at CLCR < 50 mL/min. CONCLUSIONS: Higher doses of valganciclovir corrected to renal function are suggested for use as oral prophylaxis for CMV infection in kidney transplant recipients. Further study is required to establish TDM targets to ensure adequate drug concentrations while avoiding potentially toxic drug exposures.