Pharmacokinetic/Pharmacodynamic Modelling and Monte Carlo Simulations to Predict Cytomegalovirus Viral Load in Pediatric Transplant Recipients Treated with (val)Ganciclovir.

Abstract

Background and objectives: Cytomegalovirus (CMV) infection poses significant challenges in pediatric transplant recipients. Ganciclovir and its prodrug valganciclovir are primary treatments because of their potent antiviral effects. Balancing efficacy and toxicity is particularly critical in children. This study aimed to develop a pharmacokinetic/pharmacodynamic (PK/PD) model for (val)ganciclovir and assess the relationship between area under the concentration-time curve (AUC) and CMV viral loads via Monte Carlo simulations.

Methods: We conducted a retrospective analysis including 184 viral load samples from 36 transplanted children treated with ganciclovir/valganciclovir. We developed a population pharmacodynamic model using Monolix and performed Monte Carlo simulations to assess viral load decline with varying AUCs. Internal validation was performed using goodness-of-fit plots and bootstraps.

Results: We used a viral turnover model with stimulated degradation to model the pharmacodynamic data. Model validation showed no bias or misspecification. Simulations indicated that maintaining an AUC_0-24 ≥ 40 mg·h/L achieved an 85.4% probability of undetectable viral load after 28 days of therapy. An AUC_0-24 > 30 mg·h/L provided 80.9% probability of reducing viral loads by - 1 log after 2 weeks. AUC_0-24 values > 60 mg·h/L offered minimal incremental benefits.

Conclusion: The pharmacodynamic model accurately predicted observed data. Simulations indicated that maintaining a ganciclovir plasma AUC_0-24 around 40-60 mg·h/L maximized antiviral efficacy. An AUC_0-24 > 60 mg·h/L might increase the risk of adverse events without providing additional efficacy.