Beyond class assumptions: a clinical decision framework for navigating drug-supplement interactions in oncology

Background Dietary supplement use is common among patients with cancer, with nearly half reporting concurrent use alongside conventional therapy. Despite widespread use, drug–supplement interactions in oncology are often difficult to interpret due to heterogeneous evidence, outdated case reports, reliance on in vitro and animal data, and a traditional focus on drug class effects rather than drug-specific vulnerability. Translation of preclinical findings is further complicated by host factors such as inflammation, pharmacogenomic variability, and the gut microbiome. The purpose of this review is to provide a mechanistic overview of drug–supplement interactions in oncology and characterize the existing literature to support clinical decision-making. Methods This narrative review synthesizes current literature on pharmacokinetic and pharmacodynamic interactions between dietary supplements and anticancer therapies, focusing on key mechanistic domains, including drug metabolism (CYP450 and UDP-glucuronosyltransferase enzymes), transporter systems, and immune modulation, and clinically relevant uncertainty. Interaction risk was evaluated across major anticancer drug classes, including tyrosine kinase inhibitors, hormonal therapies, oral chemotherapies, and immune checkpoint inhibitors. Results Interaction vulnerability varies substantially by drug class but is not uniform across individual agents. Tyrosine kinase inhibitors have substantial interaction potential due to CYP3A4 dependence, exposure-response relationships, and, for some agents, narrow therapeutic margins, with vulnerability varying by agent. Hormonal therapies present mechanism-specific risks, particularly CYP2D6-mediated activation of tamoxifen. Oral chemotherapies demonstrate variable interaction potential, with capecitabine associated with increased bleeding risk and cyclophosphamide dependent on CYP-mediated activation. Immune checkpoint inhibitors are less susceptible to metabolic interactions but present interpretation challenges related to immune modulation and emerging microbiome-related evidence, where theoretical risks often exceed clinical evidence. Across drug classes, many reported interactions remain theoretical, are derived from non-clinical data, or lack standardized product verification, limiting clinical interpretation. Conclusion Drug–supplement interaction risk in oncology is best understood through a mechanism-based, drug-specific framework rather than isolated supplement-based or class-level approaches. A structured assessment incorporating drug metabolism, supplement activity, host factors, and clinical context may improve risk stratification and patient counseling, although management remains challenging in the absence of robust clinical data. Routine medication reconciliation and increased patient disclosure remain essential to minimizing avoidable harm and supporting safe, individualized cancer care.