Artificial intelligence-driven pharmacovigilance analysis of polypharmacy and adverse events in breast cancer survivors treated with antidepressants.

[BACKGROUND] Breast cancer survivors often require concurrent treatment with antidepressants and anticancer drugs, raising concerns about potential drug-drug interactions and complex safety issues.

[METHODS] This retrospective pharmacovigilance study analyzed adverse event (AE) reports from the FDA Adverse Event Reporting System database spanning from 2014 to 2024. The analysis focused on breast cancer survivors who were receiving five or more concomitant medications and had reported two or more AEs. An artificial intelligence-driven clustering methodology was employed to identify distinct patterns of AE co-reporting.

[RESULT] Among 12 076 eligible patients, representing 255 020 medication records and 89 369 AE records, three distinct patient subgroups (2.1% of the total) with unique AE co-reporting profiles were identified. The patients in Cluster 1, predominantly middle-aged males, mainly treated with mirtazapine, antipsychotics, chemotherapy, and anti-HER2 monoclonal antibodies, showed high co-reporting rates for gastrointestinal symptoms, extrapyramidal disorders, peripheral neuropathy, and seizures. The patients in Cluster 2, comprising young females, mainly used citalopram combined with chemotherapy and anti-HER2 drugs, co-reported severe gastrointestinal AEs. The patients in Cluster 3, consisting of middle-aged to elderly females, mainly received citalopram, chemotherapy, anti-HER2 monoclonal antibodies, and steroid hormone drugs exhibited extremely high co-reporting rates of respiratory AEs.

[CONCLUSION] Certain polypharmacy regimens with antidepressants in specific breast cancer survivor subgroups may lead to complex safety signals. Personalized risk evaluation and vigilant medication safety strategies are crucial for these distinct, non-representative patients.