Skip to main content

Table 7 Potential risks associated with the use of some HDSs used by cancer patients

From: An assessment of the impact of herb-drug combinations used by cancer patients

HDS or combination of concern [no of patients exposed to potential risk]

aComments on previous reports and assessment of risk

Soursop (Graviola) [1]

Herbal products made from the leaves or bark of soursop are sold as alternative treatments for cancer with no clinical evidence but on the basis of experimental studies showing selective cytotoxicity in some cancer cell lines. Their use is associated with atypical Parkinson’s disease, due to the content of acetogenins (especially annonacin-1, a mitochondrial complex I inhibitor) and the neurotoxic alkaloids reticuline and N-methylcoculaurine [4]. These compounds are not present to any great extent in the juice, which the patient had been drinking, so this practice was assessed as not harmful.

Evening primrose oil and Star flower oil in combination [1]

Evening primrose and star flower oils are both sources of gamolenic acid; however, gamolenic acid is a relatively safe substance and the combination is unlikely to be harmful.

Senna [3]

Senna increases intestinal transit and large doses can reduce the absorption of a drug. However, the 3 patients were taking opioids (codeine 2; tramadol 1) which cause constipation, and thus needed senna to counteract their effects.

Fish oil/cod liver oil/EPA/omega-3 fatty acid with garlic in combination [5]

The combination of fish oils (and by inference, unsaturated fatty acids) with garlic has been reported to have increased lipid lowering effects, but these are thought to be generally beneficial.

Apricot kernels [1]

Apricot kernels are promoted as an alternative treatment for cancer with no evidence of efficacy. They contain a glycoside called amygdalin which is toxic in large quantities as it releases cyanide gas, and poisoning has been reported [26]. The patient had not experienced any ADEs and had stopped taking the product, so intervention was not deemed appropriate.

  1. aComments on previous reports are taken from Stockley’s Herbal Medicine Interactions, except where not included, in which case other references are given