Interactions between herbal medicines and prescribed drugs: a systematic review.

Despite the widespread use of herbal medicines, documented herb-drug interactions are sparse. We have reviewed the literature to determine the possible interactions between the seven top-selling herbal medicines (ginkgo, St John's wort, ginseng, garlic, echinacea, saw palmetto and kava) and prescribed drugs. Literature searches were performed using the following databases: Medline (via Pubmed), Cochrane Library, Embase and phytobase (all from their inception to July 2000). All data relating to herb-drug interactions were included regardless of whether they were based on case reports, case series, clinical trials or other types of investigation in humans. In vitro experiments were excluded. Data were extracted by the first author and validated by the second author. 41 case reports or case series and 17 clinical trials were identified. The results indicate that St John's wort (Hypericum perforatum) lowers blood concentrations of cyclosporin, amitriptyline, digoxin, indinavir, warfarin, phenprocoumon and theophylline; furthermore it causes intermenstrual bleeding, delirium or mild serotonin syndrome, respectively, when used concomitantly with oral contraceptives (ethinylestradiol/desogestrel), loperamide or selective serotonin-reuptake inhibitors (sertaline, paroxetine, nefazodone). Ginkgo (Ginkgo biloba) interactions include bleeding when combined with warfarin, raised blood pressure when combined with a thiazide diuretic and coma when combined with trazodone. Ginseng (Panax ginseng) lowers blood concentrations of alcohol and warfarin, and induces mania if used concomitantly with phenelzine. Garlic (Allium sativum) changes pharmacokinetic variables of paracetamol, decreases blood concentrations of warfarin and produces hypoglycaemia when taken with chlorpropamide. Kava (Piper methysticum) increases 'off' periods in Parkinson patients taking levodopa and can cause a semicomatose state when given concomitantly with alprazolam. No interactions were found for echinacea (Echinacea angustifolia, E. purpurea, E. pallida) and saw palmetto (Serenoa repens). In conclusion, interactions between herbal medicines and synthetic drugs exist and can have serious clinical consequences. Healthcare professionals should ask their patients about the use of herbal products and consider the possibility of herb-drug interactions.