The trusted source for
healthcare information and
October 2001; Volume 3; 73-76
By Dennis V.C. Awang, PhD, FCIC, Jerry Cott, PhD, and Adriane Fugh-Berman, MD
A recent problematic study claims that pregnant women who ingest herbs may endanger their babies because of colchicine accumulation in placental blood.1 Published in the American Chemical Society’s journal, Chemical Research in Toxicology (CRT), the paper reports that during an experiment to characterize natural anti-inflammatory substances, high concentrations of colchicine (49-763 ng/mL) were identified in the placental blood of women who had used herbal dietary supplements during pregnancy. The investigators then tested ginkgo and echinacea products from the Detroit area and found colchicine (26 ± 3 mcg/tablet) in an unidentified ginkgo product and in smaller quantities in an unidentified echinacea product (2.0 ± 0.5 mcg/tablet). They concluded that women who are pregnant or trying to conceive should avoid herbal products. These results, however, are highly questionable.
A naturally occurring plant alkaloid, colchicine primarily is found in Colchicum species, especially C. autumnale L. (Liliaceae), the autumn crocus or meadow saffron from the Mediterranean (quite different from common culinary saffron, Crocus sativa L., which is from another plant family, Iridaceae). Colchicum species have been used in traditional medicine since Hippocrates’ time, and colchicine currently is used in conventional medicine to treat gout and familial Mediterranean fever, and as an antineoplastic agent.
Colchicine, however, does not occur in either ginkgo (Ginkgo biloba L. [Ginkgoaceae]) or echinacea (Echinacea spp. [Asteraceae]), or in any species related to either. The CRT publication claimed that colchicine was found in "approximately 200 species of plants across 20 genera," but this is not accurate.
The American Botanical Council, an independent non-profit research and education organization, commissioned a search of NAPRALERT, a comprehensive database of the world literature on natural products. The search, conducted by pharmacognocist Norman R. Farnsworth, PhD, revealed that "colchicine has been reported in 91 species of the Lily family (Liliaceae), comprising 28 genera. It also occurs in one species of the Araceae (Arisaema curvatum) and is reported in one species of the Asteraceae (Compositae), i.e., Saussurea sacra. Based on biogenetic considerations, colchicine should never be found outside of the Monocotyledoneae (e.g., Araceae, Liliaceae); thus, the report of its occurrence in Saussurea sacra (Asteraceae) is an anomaly that has not been duplicated by other reports on the chemistry of this species.
The only recognized sources of appreciable quantities of colchicine are Colchicum spp. (chiefly C. autumnale); the glory lily, Gloriosa superba L. (Liliaceae), originating in Africa and Asia (and a popular Thai medicinal plant); and an Indian medicinal plant, Iphigenia indica Kunth & Benth (Liliaceae).
If colchicine has never been reported as a normal constituent of either Ginkgo biloba or any Echinacea species, how could it be identified in these herbal products? There would seem to be no rational incentive for herbal product manufacturers to deliberately adulterate any botanical with either colchicine or colchicine- containing plant material, so contamination would be the most likely explanation.
Improper cleaning of mechanical devices used to powder and package a product containing colchicine could explain the presence of colchicine in the final products. Accidental contamination during plant harvesting is possible but seems unlikely; it would be very strange that plant materials as disparate as ginkgo leaf and echinacea root (presumably) both would be accidentally contaminated with either colchicine or colchicine-containing plant material (usually seeds or corms/rhizomes/tubers).
The possibility of cross-contamination of stored raw material would depend on whether the ginkgo raw material supplier also manufactured colchicine or Colchicum products, the proximity of storage of the materials, handling procedures, etc. There also is the possibility of contamination if the same storage containers were used to store all three botanicals. Identification of the manufacturer(s) of the implicated ginkgo and echinacea products would have been helpful in determining the likelihood of this scenario. Alternatively, the samples themselves could have become contaminated with colchicine in the analytical laboratory where the assays were conducted.
The types or quantities of commercial dietary supplements that were purchased and tested were not described in any way. It is not stated why echinacea and ginkgo were chosen (ginkgo is not commonly used during pregnancy) or how products were chosen. Neither the products tested nor the lot numbers were identified. There is no specification of whether replicate samples were analyzed, and no information on how the sampling of the herbal products was conducted. How can this work be replicated or checked? There is no way for any independent laboratory to verify the findings of this study because experimental details are not given.
Clinical details also are missing. It is stated that five of 24 samples of placental blood contained colchicine (760, 182, 106, 97, and 49 mcg/L); all five came from "women who used herbal supplements." Women who did not use herbal supplements (presumably 19, but the number is not stated) had "little or no" detectable colchicine in placental blood. "Placental blood," by the way, is a meaningless term because the placenta contains separate maternal and fetal circulatory systems. The report states that "fresh blood was collected from the human umbilical cord and placenta." Although fetal blood may be collected easily from the cord, a sample taken from the placenta would be expected to contain a mixture of maternal blood, fetal blood, and placental tissue. Drug levels may be quite different in maternal blood, fetal blood, and placental tissue, and mixing the three together is not useful. Also, what is the source of the "little" colchicine apparently identified in the placental blood of women who said they did not use herbs?
The herbs that were ingested by subjects are not named, and dose and duration of use are not mentioned. Had any of the women with significant colchicine blood levels actually used the ginkgo product analyzed by the authors? Although it is stated that none of the herb-using women were taking pharmaceutical colchicine, apparently they were not asked about Colchicum-containing herbal products. Colchicum is included in several multi-ingredient herbal preparations, such as the French Antigoutteux Rezall (for rheumatism and gout), and the German Colchicum-Strath (for joint disorders).
Questions must be raised about the rigor of the analytical process. The placental blood concentrations of colchicine reported appear to be inconsistent with levels of colchicine identified in products. The ginkgo product was reported to contain 26 mcg of colchicine per tablet. The therapeutic dose of colchicine is around 1 mg/d. Therapeutic serum colchicine levels are 0.3-2.4 ng/mL.2 In another study, chronic dosing with 1 mg/d colchicine resulted in plasma levels of approximately 0.8-7.8 ng/mL.3 This study reported plasma levels of greater than 700 ng/mL! How many 26 mcg ginkgo tablets would one have to consume to reach these plasma levels? A quick calculation, assuming linear kinetics, suggests approximately 10,000 tablets.
More importantly, reported concentrations were so high that, if real, serious adverse effects (including fatalities) would be inevitable.
Colchicine toxicity is dose-dependent. Colchicine poisoning from plants is rare; a recent poisoning case was reported of a 44-year-old male who ingested 40 flowers of Colchicum autumnale L.; he reported nausea, vomiting, and abdominal pain two hours after ingestion, with diarrhea 14 hours after ingestion. Maximal colchicine levels were 4.34 ng/mL; he recovered with supportive care.2
Poisoning from colchicine drugs is more common; severe poisoning may cause multi-organ failure, acute respiratory distress syndrome, cardiogenic shock, and hematological toxicity, with a high mortality rate.4 In a case series of 24 patients admitted to an ICU for acute colchicine poisoning, six of 24 (25%) died. Plasma concentrations on admission did not differ significantly between survivors (4.1 ng/mL [0.7-68.5 ng/mL]) and non-survivors (13.6 ng/mL [3.2-77.0 ng/mL]).
How could any of these mothers or infants in the Petty study have survived so many multiples of a lethal dose? Birth outcomes are not described, but since the investigators did not report any maternal or fetal deaths (or even symptoms related to colchicine toxicity), we can only assume that there were some major errors in the analyses.
Although the CRT article has received media attention as an example of herb-related dangers, this study is not reliable.
Dr. Awang is President, MediPlant Consulting Services, White Rock, British Columbia, and serves on the editorial advisory board of Alternative Therapies in Women’s Health; Dr. Cott is Scientific Director and Chief Science Officer at Scientific Herbal Products, Inc. in College Park, MD.
1. Petty HR, et al. Identification of colchicine in placental blood from patients using herbal medicines. Chem Res Toxicol 2001;14:1254-1258. Available at: http://pubs.acs.org/subscribe/journals/crtoec/browse_asap.html Accessed September 18, 2001.
2. Danel VC, et al. Self-poisoning with Colchicum autumnale flowers. J Toxicol Clin Toxicol 2001;39:409-411.
3. Levy M, et al. The effect of long-term colchicine therapy in patients with recurrent polyserositis on the capacity of blood platelets to synthesize thromboxane A2. Br J Clin Pharmacol 1983;16:191-194.
4. Megarbane B, et al. Prognostic value of plasma colchicine concentrations in acute poisonings [abstract]. J Toxicol Clin Toxicol 2001;39:258. (Abstract #65 from European Association of Poison Centers and Clinical Toxicologists XXI International Conference; Barcelona, May 16-19, 2001).