Abnormal Drug Concentrations Due to Use of Herbal Medicines

Several herbal medicines lower the seizure threshold maintained by phenobarbital, offsetting the beneficial anticonvulsant activity. Evening primrose oil is used as a remedy for premenstrual syndrome, diabetic neuropathy, and attention-deficit/hyperactivity disorder. Evening primrose oil contains gamolenic acid that lowers the seizure threshold maintained by several anticonvulsants. Borage oil (starflower) also contains gamolenic acid. Shankhapushpi, an ayurvedic medicine for epilepsy, has adversely affected the effectiveness of phenytoin. Dandekar et al observed 2 patients experience loss of seizure control after self-medication with shankhapushpi. The serum phenytoin concentration dropped from 9.6 µg/mL (38.0 µmol/L) to 5.1 µg/mL (20.2 µmol/L) after ingestion of this herbal product (1 teaspoon 3 times a day).

Warfarin. Warfarin is an anticoagulant with a narrow therapeutic range. The drug has potentially serious consequences if bleeding complications develop or if a subtherapeutic level occurs, thus failing to protect the patient from thromboembolic events. Several herbs interact with warfarin. The herbs that may increase the risk of bleeding (potentiate effects of warfarin) include angelica root, arnica flower, ansine, bogbean, borage seed oil, capsicum, feverfew, garlic, ginger, ginkgo, horse chestnut, licorice root, and willow bark. The herbs with documented interaction with warfarin include Dan Shen, ginseng, Siberian ginseng, Devil's claw, and dong quai, among others.

A 47-year-old man with a mechanical heart valve took warfarin for 5 years and had an average international normalized ratio (INR) of 4. Within 2 weeks of using ginseng, his INR dropped to 1.5, but 2 weeks after discontinuing ginseng use, it returned to 3.3. Fortunately, no adverse effects occurred during the 2 weeks with a subtherapeutic INR. A subtherapeutic INR due to the intake of soy protein in the form of soy milk also has been reported in a 70-year-old man. INR values returned to normal 2 weeks after discontinuation of soy milk. Conversely Dan Shen caused inappropriately increased anticoagulation (INR values ranging from 5.5-8.4) in patients taking warfarin. Apart from inhibition of platelet aggregation, Dan Shen also promotes fibrinolysis due to antithrombin III-like activities. Dan Shen increases the concentration of warfarin owing to a decrease in clearance.

Dong quai is a Chinese medicine used for treatment of menstrual cramps, irregular menses, and menopausal symptoms. A 46-year-old woman with stabilized atrial fibrillation who was taking warfarin experienced a greater than 2-fold rise in prothrombin time (23.5 seconds; baseline, 16.2 seconds) and INR (4.05, baseline value, 1.89) after taking dong quai for 4 weeks. A month later, her INR was 4.9 and the prothrombin time was 27.0 seconds. At that time, the patient admitted taking dong quai. The patient was advised to discontinue dong quai, and her INR was 2.48 and the prothrombin time was 18.5 seconds 4 weeks after withdrawal of the herb. Dong quai contains coumarins, which are natural vitamin K antagonists. Boldo-fenugreek also increases INR and bleeding time in patients taking warfarin.

Licorice. Licorice may offset the ability of spironolactones to reduce blood pressure. Licorice is used as an anti-inflammatory herb and also as a remedy for gastric and peptic ulcers. Carbenoxolone, one of the components of licorice, can elevate blood pressure and cause hypokalemia. However, discontinuation of licorice results in the return of blood pressure to normal.

Significantly Lower Concentrations of Drugs Due to Concurrent Use of St John's Wort

St John's wort is prepared from Hypericum, a perennial aromatic shrub with bright yellow flowers that bloom from June to September. The flowers are believed to be most abundant and brightest around June 24, the day traditionally believed to be the birthday of John the Baptist. Therefore, the name St John's wort became popular for this herbal product. Many chemicals have been isolated from St John's wort, including hypericin, pseudohypericin, quercetin, isoquercitrin, rutin, amentoflavone, hyperforin, other flavonoids, and xanthones. Interestingly melatonin, a human pineal gland hormone, is also found in St John's wort. The mechanism of action of St John's wort is not well established.

Several reports describe unexpected low concentrations of certain therapeutic drugs due to concurrent use of St John's wort. Johne et al reported that 10 days' use of St John's wort resulted in a decrease of trough serum digoxin concentrations by 33% and peak digoxin concentration by 26%. Durr et al confirmed the lower digoxin concentrations in healthy volunteers who concurrently took St John's wort. The authors also demonstrated that St John's wort activates cytochrome P-450 mixed-function oxidase liver enzymes (CYP3A4) responsible for metabolism of digoxin and many other drugs. Barone et al reported 2 cases in which renal transplant recipients started self-medication with St John's wort. Both patients experienced subtherapeutic concentrations of cyclosporine, and in 1 patient, acute graft rejection developed owing to the low cyclosporine concentration. In both patients, termination of the use of St John's Wort returned the cyclosporine concentrations to therapeutic levels. St John's wort also reduced the area under the curve of the HIV-1 protease inhibitor indinavir by a mean of 57% and decreased the extrapolated trough by 81%. A reduction in indinavir concentration of this magnitude could lead to treatment failure. A case report describes an interaction between St John's wort and theophylline. After she began taking St John's wort, a patient who had been taking 300 mg of theophylline twice daily required a dosage increase to 800 mg twice daily to maintain a serum theophylline concentration of 9.2 µg/mL (51 µmol/L). Seven days after discontinuation of St John's wort, her theophylline concentration increased to 19.6 µg/mL (109 µmol/L).

Fugh-Berman and later Fugh-Berman and Ernst have written reviews on interactions between herbs and drugs. The most common interactions between herbs and drugs are summarized in Table 3.(see below)

Unexpected Presence of a Drug in a Patient Who Never Used That Drug: Herbal Medicines Adulterated With Western Medicines

The adulteration of Chinese herbal products with Western drugs is a serious problem. Of 2,069 samples of traditional Chinese medicines obtained from 8 hospitals in Taiwan, 23.7% contained pharmaceuticals, most commonly caffeine, acetaminophen, indomethacin, hydrochlorothiazide, and prednisolone. Nonsteroidal anti-inflammatory drugs and benzodiazepines have been found in many Chinese medicines sold outside Asia. These herbs include Miracle-Herb, tung shueh, and Cuifong Toukuwan. Heavy metal contamination also was found in herbal products. Ko reported that 24 of 254 Asian patented medicines obtained from herbal stores in California contained lead, 36 products contained arsenic, and 35 products contained mercury.

Nelson et al reported a case of aplastic anemia associated with the use of herbal medication in a 12-year-old boy. The authors demonstrated the presence of phenylbutazone in the herbal preparation, but that medication was not listed as an ingredient in the package insert. The boy had a hemoglobin concentration of 8 g/dL (80 g/L), a neutrophil count of 200/mL, and a platelet count of 5,000/mL. These hematologic abnormalities are related to phenylbutazone toxicity. Lau et al reported a case in which a 33-year-old patient had a serum phenytoin concentration of 48.5 µg/mL (192 µmol/L). The patient had a history of a seizure disorder that was managed with 400 mg sodium valproate 3 times a day, 200 mg of carbamazepine twice a day, and 150 mg of phenobarbital every evening. No phenytoin was given. The patient consumed a proprietary Chinese medicine before admission to the hospital. The manufacturer's information leaflets stated that the capsules contained pure Chinese medicines and were effective for controlling epilepsy. Goudie and Kaye reported a case of severe hypoglycemia in a patient (fasting glucose level, 37.8 mg/dL [2.1 mmol/L]) due to use of a Chinese medicine, ZhenQi, for diabetes. Analysis of this herbal medicine showed the presence of glyburide, a sulfonylurea. A sulfonylurea overdose can lead to profound hypoglycemia.