Purpose: To develop a system to estimate the risk of herb-drug interactions that includes the available evidence from clinical and laboratory studies, transparently delineates the algorithm for the risk estimation, could be used in practice settings and allows for adaptation and update.
Methods: We systematically searched Drugbank, Transformer, Drug Information Handbook, European and German Pharmacopoeia and MEDLINE for studies on herb-drug interactions of five common medicinal plants (coneflower, ginseng, milk thistle, mistletoe and St. John's wort). A diverse set of data were independently extracted by two researchers and subsequently analysed by a newly developed algorithm. Results are displayed in the form of interaction risk categories. The development of the algorithm was guided by an expert panel consensus process.
Results: From 882 publications retrieved by the search, 154 studies were eligible and provided 529 data sets on herbal interactions. The developed algorithm prioritises results from clinical trials over case reports over in vitro investigations and considers type of study, consistency of study results and study outcome for clinical trials as well as identification, permeability, bioavailability, and interaction potency of an identified herbal perpetrator for in vitro investigations. Risk categories were assigned to and dynamically visualised in a colour-coded matrix format.
Conclusions: The novel algorithm allows to transparently generate and dynamically display herb-drug interaction risks based on the available evidence from clinical and laboratory pharmacologic studies. It provides health professionals with readily available and easy updatable information about the risk of pharmacokinetic interactions between herbs and oncologic drugs.
Keywords: Anticancer drugs; Assessment algorithm; Complementary medicine; Herb-drug interactions; Herbal supplements.