Omeprazole is associated with voriconazole-associated hepatotoxicity: PPI-stratified analysis and a clinically actionable nomogram from an 848-patient cohort.

OBJECTIVE

To evaluate whether individual PPIs (as opposed to PPI use per se) are differentially associated with the risk of VRC-associated hepatotoxicity; to derive omeprazole-specific, clinically actionable VRC trough concentration (Cmin) thresholds; and to develop and internally evaluate a multivariable nomogram that outperforms a Cmin-only model.

METHODS

We conducted a single-centre retrospective study in China (February 2023-February 2025), including 848 non-ICU adults receiving VRC. Multivariable logistic regression assessed the associations between hepatotoxicity and Cmin, PPI type, loading dose, and inflammatory markers, adjusting for potential confounders. Omeprazole-stratified analyses yielded optimal Cmin cut-offs, and nomograms were constructed based on significant predictors.

RESULTS

Overall, 145/848 (17.1%) developed grade ≥2 and 59/848 (6.9%) developed grade ≥3 hepatotoxicity. Among PPI users (61%), omeprazole was the only PPI with a statistically significant adjusted association with hepatotoxicity; associations for other PPIs were not statistically significant. In the omeprazole subgroup, VRC Cmin predicted grade ≥2 and ≥3 events at 4.1 mg/L (AUC 0.688; sensitivity 0.905; specificity 0.492) and 5.5 mg/L (AUC 0.708; sensitivity 0.750; specificity 0.653), respectively. Two nomograms (grade ≥2: Cmin, procalcitonin, loading dose, PPI type; grade ≥3: plus age) achieved AUCs of 0.670 and 0.733, outperforming Cmin-only models (0.602 and 0.608) and showing good calibration and DCA benefit across 0.10-0.60.

CONCLUSION

Omeprazole co-administration was associated with higher VRC-associated hepatotoxicity, consistent with increased VRC exposure (Cmin). Omeprazole-specific Cmin thresholds (4.1 and 5.5 mg/L) support early TDM and PPI optimization when omeprazole is unavoidable.