Incorporating the metabolic ratio (C-_VRC/C-_VNO) into a LASSO-logistic model for predicting voriconazole-induced DILI:...

Area of Science:

• Pharmacology
• Hepatology
• Clinical Prediction Modeling

Background:

• Voriconazole (VRC) is a vital antifungal agent.
• Drug-induced liver injury (DILI) is a significant concern with VRC use.
• Early identification of patients at risk for VRC-DILI is crucial for clinical management.

Purpose of the Study:

• To identify risk factors for VRC-induced DILI.
• To develop and validate a predictive model (nomogram) for VRC-DILI.
• To create a user-friendly web-based tool for clinical application.

Main Methods:

• Retrospective analysis of 250 patients treated with VRC.
• Identification of risk factors using univariate and LASSO-logistic regression.
• Construction and validation of a nomogram prediction model using ROC, calibration curves, and DCA.
• Development of a web-based prediction calculator.

Main Results:

• Five key predictors for VRC-DILI were identified: concomitant bacterial infection, VRC steady-state trough concentration (C-VRC), C-VRC/VRC N-oxide ratio (MR), albumin (ALB), and serum creatinine (Scr).
• The nomogram model demonstrated good predictive performance with an AUC of 0.818.
• The model showed good calibration and clinical applicability, validated by Bootstrap analysis.

Conclusions:

• C-VRC, MR, ALB, and Scr are significant predictors of VRC-DILI.
• The LASSO-logistic regression-based nomogram offers robust predictive capability and clinical utility.
• The web-based calculator enhances clinical convenience for early VRC-DILI risk assessment and management.