http://www.ama-assn.org/sci-pubs/journals/most/recent/issues/inte/letter_4.htm
Herbal Medicinals: Selected Clinical Considerations,
Focusing on Known or Potential Drug-Herb Interactions
The issue of
herb-drug interactions is an important one for clinicians. However, the recent article by Miller[1] on
this subject contains a number of errors and unsubstantiated statements.
The claim that
Echinacea is hepatotoxic is unreferenced, presumably because there are no
reports of this in the literature.
The unreferenced
statement that ginkgo should be avoided in patients with epilepsy is not
substantiated. The statement that ginkgo toxin is found in both leaf and seed
is true but misleading. Ginkgo seeds, which are eaten on special occasions in
Asia, contain 4’-O-methylpyridoxine, an anti-vitamin B6 neurotoxin. Boiling
inactivates 99% of this toxin. However, during food shortages in China and
Japan, overconsumption of ginkgo seeds resulted in a syndrome called gin-nan
sitotoxism, with a 27% mortality rate.[2]
Ginkgo leaf
extracts, however, are used therapeutically in the West; according to the
author’s own citation, the leaves of ginkgo contain only tiny quantities of
4’-O-methylpyridoxine.[2] Ginkgo leaf extracts have not been shown to have
toxic effects in animals or humans.
Both adverse effects
attributed to Ginkgo biloba were probably drug interactions. The patient who
developed spontaneous hyphema while taking ginkgo was also taking aspirin,[3]
and the patient who developed spontaneous bilateral subdural hematomas was also
taking acetaminophen and ergotamine/caffeine at the time of diagnosis.[4] The concurrent
use of these drugs should have been noted. (Ginkgo alone has been associated
with 1 case of subarachnoid hemorrhage[5] and 1 case of subdural hemorrhage,[6]
but these cases were not referenced in this article.)
There is no reported
association between kava and coma. It is true that the letter cited for this
statement is misleadingly titled “Coma From the Health Food Store: Interaction
Between Kava and Alprazolam,”[7] but the letter reports only lethargy and
disorientation in this patient, not unconsciousness.
Statements in the
story of the androgenized baby are not supported by the references. The
original case report, not cited in the article, described a case of neonatal
androgenization attributed to the maternal use of a product called Siberian
ginseng (Eleutherococcus senticosus),[8] not, as stated in the article by
Miller, “ginseng,” which is the unrelated Panax genus. The letters cited for
the statement that “neonatal androgenization secondary to ginseng has been
debated in the literature in cases in which maternal use of ginseng was
identified as the cause of androgenization of the child”[9,10] actually debated
only whether the product in the preceding case report should have been analyzed
to determine whether it contained Siberian ginseng. In fact, a subsequent
analysis of the preparation (not referenced) revealed that the preparation was
devoid of Siberian ginseng but appeared to contain Periploca sepium, also
called Chinese silk vine.[11] The reference[12] cited for the statement that
“others contend the entity in
question was in fact
a botanically different species, Siberian ginseng” is wrong. The reference
actually only mentions the analysis that unmasked the purported Siberian
ginseng as Periploca as background to an animal experiment that tested the
androgenic qualities of Periploca.
The author conflates
the uses and risks of borage leaves with borage seed oil; these are not
interchangeable entities. It is borage seed oil, not (as stated) borage or
borage oil (both leaf products), that is used as an alternative to evening
primrose oil. It is only borage leaves, however, not borage seed oil, that
contain significant levels of toxic pyrrolizidine alkaloids. (Although small
amounts of pyrrolizidine alkaloids have been isolated from the seeds, borage
seed oil has been found to be devoid of pyrrolizidine alkaloids.[13])
The statement
concerning garlic that “several practitioners have noted elevated international
normalized ratios (INR) and prothrombin times in patients previously stabilized
while taking warfarin,” is not only unreferenced but does not fit with what we
know about the anticoagulant effects of this herb. Garlic prevents platelet
aggregation[14,15] and thus would be expected to alter bleeding times, not
international normalized ratios or prothrombin times.
The caution against
combining phytoestrogens with pharmaceutical estrogens because of possible
estrogen excess is overstated. The strongest phytoestrogen is about 0.1% as
potent as estradiol in binding assays.[16]
Adriane Fugh-Berman, MD
Department of Health
Care Sciences
George Washington
University School of Medicine
2150 Pennsylvania Ave NW
Suite 2B-417
Washington, DC 20037
1. Miller LG. Herbal medicinals: selected clinical
considerations focusing on known or potential drug-herb interactions. Arch
Intern Med. 1998;158:2200-2211.
2. Arenz A, Klein M, Fiehe K, et al. Occurrence
of neurotoxic 4’-O-methylpyridoxine in Ginkgo biloba leaves, Ginkgo medications
and Japanese Ginkgo food. Planta Med. 1996;62:548-551.
3. Rosenblatt M, Mindel J. Spontaneous hyphema
associated with ingestion of Ginkgo biloba extract [letter]. N Engl J Med.
1997;336:1108.
4. Rowin J, Lewis SL. Spontaneous bilateral
subdural hematomas associated with chronic Ginkgo biloba ingestion. Neurology.
1996;46:1775-1776.
5. Vale S. Subarachnoid haemorrhage associated
with Ginkgo biloba [letter].
Lancet.
1998;352:36.
6. Gilbert GJ. Ginkgo biloba [commentary].
Neurology. 1997;48:1137.
7. Almeida JC, Grimsley EW. Coma from the health
food store: interaction between kava and alprazolam [letter]. Ann Intern Med.
1996;12:940-941.
8. Koren G, Randor S, Martin S, Danneman D.
Maternal ginseng use associated with neonatal androgenization [letter]. JAMA.
1990;264:2866.
9. Awang DVC. Maternal use of ginseng and
neonatal androgenization [letter].JAMA. 1991;265:1828.
10. Koren G. Maternal use of ginseng and neonatal
androgenization [reply].JAMA. 1991;265:1828.
11. Awang DVC. Maternal use of ginseng and neonatal
androgenization [letter]. JAMA. 1991;266:363.
12. Waller DP, Martin AM, Farnsworth NR, Awang DVC.
Lack of androgenicity of Siberian ginseng [letter]. JAMA. 1992;267:2329.
13. De Smet PAGM. Borago officinalis. In: De Smet
PAGM, ed. Adverse Effects of Herbal Drugs, Volume II. Berlin, Germany:
Springer-Verlag; 1993.
14. Kiesewetter H, Jung EM, Mrowietz C, et al.
Effect of garlic on platelet aggregation in patients with increased risk of
juvenile ischemic attack. Eur J Clin Pharmacol. 1993;45:333-336.
15. Das I, Khan NS, Sooranna SR. Potent activation
of nitric oxide synthase by garlic: a basis for its therapeutic applications.
Curr Med Res Opin. 1995;13:257-263.
16. Knight DC, Eden JA. A review of the clinical
effect of phytoestrogens.
Obstet Gynecol.
1996;87:897-904.
(Arch Intern
Med. 1999;159:1957-1958)
Rosenblatt and
Mindel[1] reported spontaneous hyphema in a 70-year-old man 1 week after he
began ingesting Ginkgo biloba (40 mg twice daily). The patient had taken
aspirin (325 mg/d) for 3 years without incident; hence, the authors felt that a
temporal relationship strongly implicated ginkgo. I agree. Rowin and Lewis[2]
have also reported spontaneous bilateral subdural hematomas associated with
long-term G biloba use. They acknowledged that their patient had taken
acetaminophen and had a very brief trial of ergotamine/caffeine tablets.
However, given the lack of a temporal relationship and the fact that
acetaminophen, ergotamine, and caffeine have not been shown to be associated
with subdural hematomas, Rowin and Lewis implicated G biloba. Furthermore, the
patient’s bleeding time improved following discontinuation of G biloba. Again,
I agree with their conclusions. Yet another report was recently published
describing a 61-year-old man who developed subarachnoid hemorrhage after
ingesting G biloba (40 mg 3 or 4 times daily).[3] He was not taking any other
medications.
Ginkgo toxin
(4’-O-methylpyridoxine) has been characterized as a neurotoxic anti-vitamin B6
compound that Arenz et al[4] found “present in Ginkgo medications and
...detectable in homeopathic preparations.” Ginkgo toxin is more typically
found in seeds.[5] Arenz et al, however, found it present in both seeds and
leaves, with the highest amount derived from the tree in late July and early
August.[4] In addition, as Fugh-Berman points out, contamination is a problem
with many herbal medicinals. I believe it ill-advised to subject patients to
the effects of a known neurotoxin, especially prone patients (eg, those with epilepsy),
unless we can assure them that their ginkgo product does not contain this
neurotoxin.
Almeida and
Grimsley[6] describe a 54-year-old man who was hospitalized in a lethargic and
disoriented state following ingestion of kava and alprazolam. In the title of their letter, the authors
refer to this state as a “coma”; in the text, they refer to the patient as
being “semicomatose.” I defer to Almeida and Grimsley, who actually saw this
patient, as to what his actual state was. Regardless of the term used, it would
appear that an interaction exists and concomitant use should be avoided.
Ginseng is an herbal
entity that historically has had reports of contamination and considerable
variation in content. In my article, I alerted readers to the controversy and
the contamination issue regarding neonatal androgenization associated with
ginseng. I correctly cited the letters and commentaries in which this debate
was argued, and I also noted that Siberian ginseng did not cause
androgenization. Awang[7] subsequently conducted a mass spectrometric analysis
of the indicted extracts, finding 4-methoxysalicyladehyde, consistent with
Periploca sepium, also referred to as Chinese silk vine. I thank Fugh-Berman
for pointing this out. It not only highlights the problem with products labeled
as containing ginseng but further exemplifies the mislabeling and contamination
issues we are facing with herbal products.
In regard to borage,
I refer Fugh-Berman to reference 115 in my original article.[8] Newall et al
note that evening primrose oil should be used cautiously in patients with
epilepsy and make a similar recommendation for borage oil. I did not refer to
borage leaves, as Fugh-Berman alleges.
Fugh-Berman correctly notes the distinction between borage seed oil and
borage leaves.
I noted garlic’s
effect on platelets in my original article, which of course would affect
bleeding times and not international normalized ratios or prothrombin times.
This provides an example of why we cannot assume we know everything about
herbs. Clinical experience, when we are vigilant to inquire about herb use,
teaches us otherwise. The lesson here is to acknowledge that much more study is
needed.
I do not believe
concomitant use of phytoestrogens with estrogen replacement therapy is well
advised. One must first question why one would want to take these 2 products
together for this indication. Secondly, given that the content of the herbal
product may be in question (eg, mislabeling, contamination), it again does not
seem rational to combine therapies and incur unnecessary risk and expense.
In my article, I
noted that the risk that Echinacea will cause hepatotoxicity is very low
because of medicinal chemistry considerations.
Fugh-Berman elected to ignore this sentence. Traces (0.006%) of the
pyrrolizidine alkaloids (tussilagine and isotussilagine) have been found in
Echinacea angustifolia and Echinacea purpurea, but these lack the
1,2-unsaturated necine ring system associated with hepatotoxicity; hence, this
source characterized Echinacea-induced hepatotoxicity as unlikely.[9] The
absence of this 1,2-unsaturated necine ring system was noted in the original
article.
I commend the
American Medical Association and the ARCHIVES specifically for their
willingness to address alternative medicine and herbal medicinals. They no doubt did so realizing that
reactions would be varied.
Open debate and the
exchange of information and ideas are the hallmarks of scholarly endeavors.
This exchange needs to be conducted in a professional, objective, and
respectful manner so that scholars and clinicians with expertise in herbal
medicinals can bring their knowledge to bear for the benefit of patient care.
It is hoped that with a thoughtful discourse and exchange the defining
boundaries of herb use can be ascertained as they are for other medicinal
interventions. With such a scientific approach, herbal medicinals can assume a
safe and efficacious position among therapies offered to patients.
Lucinda G. Miller,
PharmD, BCPS
Editor, Journal of
Herbal Pharmacotherapy
Amarillo, Tex
Dr Miller currently
serves on an advisory board to Warner-Lambert Co, Morris Plains, NJ.
1. Rosenblatt M, Mindel J. Spontaneous hyphema
associated with ingestion of Ginkgo biloba extract [letter]. N Engl J Med.
1997;336:1108.
2. Rowin J, Lewis SL. Spontaneous bilateral
subdural hematomas associated with chronic Ginkgo biloba ingestion. Neurology.
1996;46:1775-1776.
3. Vale S. Subarachnoid haemorrhage associated
with Ginkgo biloba [letter].
Lancet.
1998;352:36.
4. Arenz A, Klein M, Fiehe K, et al. Occurrence
of neurotoxic 4’-O-methylpridoxine in Ginkgo biloba leaves, Ginkgo medications
and Japanese Ginkgo food. Planta Med. 1996;62:548-551.
5. Newall CA, Anderson LA, Phillipson ID. Ginkgo
monograph. In: Newall CA, Anderson LA, Phillipson JD, eds. Herbal Medicines: A
Guide for Health-Care Professionals. London, England: Pharmaceutical Press;
1996:138-140.
6. Almeida JC, Grimsley EW. Coma from the health
food store: interaction between kava and alprazolam. Ann Intern Med.
1996;125:940-941.
7. Awang DVC. Maternal use of ginseng and
neonatal androgenization [letter].JAMA. 1991;266:363.
8. Newall CA, Anderson LA, Phillipson ID. Borage
monograph. In: Newall CA, Anderson LA, Phillipson JD, eds. Herbal Medicines: A
Guide for Health-Care Professionals. London, England: Pharmaceutical Press;
1996:49.
9. Bisset NG, ed. Herbal Drugs and
Phytopharmaceuticals: A Handbook for Practice on a Scientific Basis. Stuttgart,
Germany: Medpharm Scientific Publishers; 1994:182-183.(Arch Intern Med.
1999;159:1958-1959)