Factors affecting trough concentrations of voriconazole: a dual-center retrospective analysis focusing on loading...

There are significant interindividual variations in voriconazole (VCZ) blood concentrations, which affect both treatment efficacy and safety.

Our goal is to identify the factors influencing VCZ trough concentrations and provide new evidence for individualized dosing.

A total of 281 hospitalized patients receiving VCZ were enrolled. Demographic, liver and kidney function, C-reactive protein (CRP) and other clinical data were collected. Trough concentrations were determined using HPLC. Multivariate linear regression and ordinal logistic regression were employed to identify influencing factors, with predictive performance assessed by ROC curves.

CRP is a significant positive predictor of VCZ trough concentration (B = 0.010, p = 0.006), demonstrating the strongest predictive capability for elevated trough levels (>5 μg/mL) (AUC = 0.8864, p < 0.0001), outperforming conventional liver function indicators such as total bile acids (AUC = 0.6326). Loading dose also showed a significant correlation with increased trough concentration (B = 0.973, p = 0.007) and elevated the risk of supratherapeutic levels (OR = 0.430, p = 0.006). Additionally, weight, albumin, platelet count, and concomitant administration of pantoprazole or dexamethasone were identified as independent influencing factors.

The high variability in VCZ trough concentrations may be partially attributed to factors such as weight, loading dose, liver function, inflammation and concomitant medications. Although loading doses enable rapid efficacy, they increase the risk of supratherapeutic concentrations. Patients in the high-CRP group are more likely to exceed 5.0 μg/mL, demonstrating the best discriminative ability for predicting excessively high concentrations. These findings provide new evidence for VCZ dose optimization.