Risk Factors for Voriconazole-Associated Hepatotoxicity in Patients with Liver Dysfunction: A Retrospective Nested Case–control Study

Jing Ren¹Xinfeng Cai²Wei Ge³Jinlin Guo⁴Shan Wang⁵Qinhui Wang⁶Linna Liu^1*Le Yang^1*Qi Yang^1*

• ¹Department of Pharmacy, Tangdu Hospital, The Fourth Military Medical University, Xi'an, China
• ²Department of Pharmacy, Shanxi Province Cancer Hospital, Taiyuan, China
• ³Department of Disease Prevention and Control, Tangdu Hospital, The Fourth Military Medical University, Xi'an, China
• ⁴Department of Pharmacy, Shanxi Provincial People’s Hospital, Taiyuan, China
• ⁵Department of Pharmacy, New York University, Langone Hospital–Long Island, Mineola, NY, United States
• ⁶Department of Pharmacy, Tangdu Hospital, The Fourth Military Medical University, xian, China

This is a provisional file, not the final typeset article Introduction: Voriconazole is widely used to prevent and treat invasive aspergillosis. However, its use is restricted by adverse effects, including acute liver injury (ALI). Patients with hepatic insufficiency are often more susceptible to voriconazole-induced liver injury than those with normal hepatic function. The aim of this study was to determine the incidence and risk factors of ALI in patients with mild or moderate liver dysfunction during voriconazole treatment. Methods: This single-center nested case–control study involved adult patients treated with voriconazole for at least 3 days. The Child–Pugh score is now extensively utilized to assess liver damage. The hepatotoxicity of voriconazole was assessed in patients with mild or moderate hepatic insufficiency (Child–Pugh A or B). ALI cases were matched with controls based on age and Child–Pugh score. Basic characteristics were compared between patients who developed ALI and those who did not by performing univariate and multivariate conditional logistic regression analyses. The optimal cutoff condition was determined using a receiver operating characteristic curve. Results: A total of 140 patients (ALI: n = 44; control: n = 96) were enrolled. The incidence of voriconazole-induced ALI in patients with mild or moderate liver dysfunction was 30.6%. The univariate analysis revealed trough voriconazole plasma concentration (VPC), voriconazole treatment duration, activated partial thromboplastin time, and intensive care unit admission as variables for the final analysis. Voriconazole-induced ALI was independently associated with trough VPC (odds ratio [OR]: 1.592, p = 0.013) and voriconazole treatment duration (OR: 1.057, p = 0.005). Notably, the optimal cutoff for treatment duration was 10 days and the recommended trough VPC threshold was 3.81 mg/L. Conclusion: The incidence of voriconazole-induced ALI was higher in patients with mild or moderate liver dysfunction than in the general population. Trough VPC and voriconazole treatment duration are two independent risk factors of ALI. Therefore, voriconazole should be administered with caution to these patients. A lower target trough VPC (<3.81 mg/L) is recommended to minimize the risk of ALI in patients with mild-to-moderate liver dysfunction.