Identification of Drugs Associated With Oral Ulcers Based on Pharmacovigilance

Oral ulcers are common, painful lesions that significantly affect quality of life. Certain medications, especially antimetabolites and mTOR inhibitors, have been implicated in their development. However, drug-induced oral ulcers have not been systematically evaluated using large-scale real-world data.

The study employed a dual-validation strategy combining pharmacovigilance signal mining and multivariate logistic regression analysis. By querying specific 'preferred terms' (PTs), adverse event reports in the FAERS database of the Food and Drugs Adminstration of USA, related to oral ulcers were identified. Signal detection was conducted using four disproportionality analysis methods. For the initially screened candidate drugs, a multivariate logistic regression model was constructed, adjusting for potential confounders such as age, gender, weight, reporter type, reporting country and year, to further validate their independent association with oral ulcer reports (P_FDR < .05).

Reports of oral ulcerations have shown a significant and sustained increase since 2010, peaking in 2020. The multi-method validation framework identified 289 statistically significant drug-oral ulceration association signals. Drugs with the highest signal strength included cyclizine, mycophenolate sodium and aredia. Other drugs with significant signals included lactulose, acetaminophen codeine phosphate, leflunomide, omalizumab, everolimus and atezolizumab. Integrating signal detection results, logistic regression analysis further confirmed that multiple drugs were significantly associated with the risk of oral ulcer reports. Drugs with extremely high odds ratios (ORs) included estriol and docetaxel sagent. Strongly associated drugs also included immunosuppressants, antihistamines, combination analgesics, bisphosphonates, sartans and antibiotics.

This study systematically identified approximately 289 drugs significantly associated with oral ulcer risk, including immunosuppressants, cardiovascular drugs and bronchodilators. Based on pharmacovigilance signal detection and regression analysis, patients on long-term treatment with these drugs should receive enhanced oral monitoring. As this is an observational study, further prospective and experimental research is needed to confirm causality.