Drug points: Hepatotoxicity associated with herbal tablets

BMJ 1996; 313 doi: (Published 13 July 1996) Cite this as: BMJ 1996;313:92
  1. P A G M De Smet,
  2. A J M Van Den Eertwegh,
  3. W Lesterhuis,
  4. B H Ch Stricker

    On 29 March 1994 a 69 year old woman who was taking levothyroxine (0.1 mg/day) and ibuprofen (400 mg/day) presented to her general practitioner with nausea, pruritus, dark urine, pale stools, and yellow sclerae after six weeks of daily self treatment with 6-15 herbal tablets. These were labelled Venencapsan and industrially prepared for one pharmacy from horse chestnut leaf, milfoil, celandine, sweet clover, milk thistle, and dandelion root with herb. Liver function was abnormal on 14 April but had improved by 2 May (table 1). She was negative for hepatitis A and B. Ultrasonography and computed tomography showed normal bile ducts without signs of obstruction. Her symptoms disappeared after six weeks, and she restarted the herbal tablets but not ibuprofen (she had stopped both while continuing levothyroxine).

    Table 1

    Results of liver function tests after herbal tablets

    View this table:

    On 27 May she was admitted with jaundice and only levothyroxine was continued. There was no evidence of alcohol abuse, and serum tests for hepatitis A, B, and C; toxoplasmosis; infectious mononucleosis; and cytomegalovirus infection gave negative results. Tests for autoantibodies gave negative results (nuclear antibodies, mitochondrial antibodies) or weakly positive results (smooth muscle antibodies). Endoscopic retrograde cholangiopancreatography showed a normal choledochal and pancreatic duct. Liver biopsy showed mild portal inflammation and focal areas of mild steatosis, ballooning, and Kupffer cell hyperplasia. Recovery was uneventful, and by January 1995 her liver function had returned to normal (table 1).

    The herbal tablets are more suspect in this case than ibuprofen, because ibuprofen had been taken regularly over the past decade and was not restarted after the first episode. Analysis of the tablets did not show pharmaceuticals or phrrolizidine alkaloids, and it remains unclear which constituent(s) might have been responsible. Horse chestnut leaf has been associated once with hepatitis after intramuscular administration.1 Medicinal use of sweet clover can provide up to 30 mg of coumarin per day,2 and this plant lactone has hepatotoxic potential in daily doses of 25 mg or more.3 However, we recovered only 0.1 mg of coumarin per tablet, so our patient was exposed to only 0.6-1.5 mg/day.—P A G M DE SMET, Royal Dutch Association for the Advancement of Pharmacy, the Hague, A J M VAN DEN EERTWEGH, W LESTERHUIS, Merwede Hospital, Dordrecht, B H CH STRICKER, Netherlands Centre for Monitoring Adverse Reactions to Drugs, Rijswijk, Netherlands


    1. 1.
    2. 2.
    3. 3.
    View Abstract