Vol. 7, Issue 2, Part G (2025)

Background: In oncology, pharmacovigilance is necessary because of the narrow therapeutic window, complicated regimens and dynamic toxicities of cancer treatments. Conventional methods of pharmacovigilance based on clinical trials and spontaneous reporting fail to capture many of the adverse drug reactions that arise in real-world patient populations. Active pharmacovigilance methods and artificial intelligence present the possibility for overcoming the shortcomings.
Objective: To discuss emerging pharmacovigilance practices in oncology, with a focus on AI-enabled strategies, pharmacist-initiated interventions, and active pharmacovigilance models that supplement clinical trial data.
Methods: A narrative synthesis of the latest literature was performed, with an emphasis on targeted therapies, oral anticancer therapies, and AI-facilitated pharmacovigilance.
Results: Real-world evidence shows unreported ADRs from clinical trials, and under-reporting is still extensive. Pharmacy-led interventions raised reporting of ADRs by >120%, and AP models improved early recognition with no unnecessary discontinuations. Natural language processing (NLP) using AI has promise for the automated identification of ADRs from unstructured electronic health records (EHRs). Hybrid models combining AI, pharmacists, and AP had greater sensitivity and specificity than single-mode approaches.
Conclusion: Oncology pharmacovigilance is headed towards connected, multidisciplinary models that integrate AI, active monitoring, and the active involvement of pharmacists. Such methods can improve ADR detection, maximize treatment safety, and promote equitable cancer care.