Variations in parthenolide content and daily dose of feverfew products

Abstract

Variations in the parthenolide content of feverfew products available to consumers were studied.
Feverfew products were analyzed for the content of parthenolide, the purported active component. The actual weight of feverfew was determined only in those products containing dried feverfew leaf. The total daily doses of feverfew leaf and parthenolide were calculated by using the instructions on each product label. Parthenolide content was determined by high-performance liquid chromatography.
The quantity of feverfew leaf in each capsule was similar to that stated on the label and ranged from 25 to 500 mg. Parthenolide content per dosage form varied 150-fold (from 0.02 to 3.0 mg), while percent parthenolide varied 5.3-fold (from 0.14% to 0.74%). If a person consumed the daily dose recommended on the label, intake of dried feverfew leaf would range from 225 to 2246 mg/day, a 10-fold variation, while intake of parthenolide would range from 0.06 to 9.7 mg/day, a 160-fold variation.
Large variations were observed in the parthenolide contents and daily intake as recommended by the labeling in commercial feverfew products.

Full text

REPORTS Parthenolide content and feverfew products
1527
Am J Health-Syst Pharm—Vol 59 Aug 15, 2002
Variations in parthenolide content and daily dose
of feverfew products
MICHAEL H. NELSON, SARAH E. COBB, AND JON SHELTON
MICHAEL H. NELSON, PH.D., is Assistant Professor of Pharmaceutical
Sciences and SARAH E. COBB and JON SHELTON are Pharm.D. degree
candidates, School of Pharmacy, Southwestern Oklahoma State Uni-
versity, Weatherford.
Address correspondence to Dr. Nelson at the School of Pharmacy,
Southwestern Oklahoma State University, 100 Campus Drive,
Weatherford, OK 73096 (nelsonm@swosu.edu).
Copyright © 2002, American Society of Health-System Pharma-
cists, Inc. All rights reserved. 1079-2082/02/0802-1527$06.00.
Variation in the contents and
discrepancies between the la-
bel and actual contents have
been documented for several dietary
supplement products. In 2000, Gurley
et al.1 reported a high variation in the
ephedra alkaloid content of ephedra-
containing dietary supplements.
Other reports dating back as far as
1978 have pointed to considerable
variation in products containing de-
hydroepiandrosterone, kava, and
ginseng.2 These variations present a
threat to the public health and call
into question the classification of po-
tentially harmful substances as di-
etary supplements and the relatively
lax regulation of these substances
under the Dietary Supplement
Health and Education Act (DSHEA)
of 1994.3 The DSHEA does not, for
example, require consistency in the
content of active components and in
the recommended daily doses
among different brands of a dietary
supplement.
For centuries, feverfew (Tanace-
tum parthenium) has been used for
its purported ability to decrease fever
and treat headaches.4,5 Recent clinical
data suggest that feverfew may have a
moderate effect on decreasing the
frequency and severity of migraine
headaches when used prophylactical-
ly, but it does not appear effective for
Abstract: Variations in the parthenolide
content of feverfew products available to
consumers were studied.
Feverfew products were analyzed for
the content of parthenolide, the purport-
ed active component. The actual weight of
feverfew was determined only in those
products containing dried feverfew leaf.
The total daily doses of feverfew leaf and
parthenolide were calculated by using the
instructions on each product label. Par-
thenolide content was determined by
high-performance liquid chromatography.
The quantity of feverfew leaf in each
capsule was similar to that stated on the
label and ranged from 25 to 500 mg. Par-
thenolide content per dosage form varied
150-fold (from 0.02 to 3.0 mg), while per-
cent parthenolide varied 5.3-fold (from
0.14% to 0.74%). If a person consumed the
daily dose recommended on the label, in-
take of dried feverfew leaf would range
from 225 to 2246 mg/day, a 10-fold varia-
tion, while intake of parthenolide would
range from 0.06 to 9.7 mg/day, a 160-fold
variation.
Large variations were observed in the
parthenolide contents and daily intake as
recommended by the labeling in commer-
cial feverfew products.
Index terms: Alternative medicine; Analy-
sis; Chromatography, liquid; Concentration;
Content uniformity; Control, quality; Di-
etary supplements; Dosage; Labeling; Par-
thenolide; Plants; Tanacetum parthenium
Am J Health-Syst Pharm. 2002; 59:1527-
31
the treatment of acute migraines or
for decreasing their duration.5,6 The
beneficial effects of feverfew are be-
lieved to be mostly due to partheno-
lide, a sesquiterpine lactone.4,5 Sever-
al in vitro studies have attempted to
elucidate parthenolide’s mechanism
of action. Capasso7 demonstrated
that aqueous extracts of feverfew
containing parthenolide 50–200 µg/
mL inhibited the metabolism of
arachidonic acid. The antisecretory
properties of feverfew and partheno-
lide have been well documented in
vitro. Groenewegen and colleagues8
identified parthenolide and several
other sesquiterpine lactones as com-
pounds extracted from feverfew that
inhibit the release of serotonin from
platelets. They also found that par-
thenolide and feverfew extract pre-
vented platelet aggregation induced
by several chemicals in vitro.9 Hayes
and Foreman10 reported that fever-
few extract is a novel type of mast-
cell inhibitor, as shown by its antihis-
taminic properties in vitro. Barsby et
al.11 demonstrated the potentially
toxic effect of irreversible inhibition
of vasculature response in rabbit aor-

REPORTS Parthenolide content and feverfew products
1528 Am J Health-Syst Pharm—Vol 59 Aug 15, 2002
liquid chromatographic (HPLC) as-
say.15 Six single-dose samples from
each product were analyzed. Samples
were prepared by extraction in 100
mL of an acetonitrileb:water (9:1)
mixture. Most of the feverfew prod-
ucts were formulated as capsules
containing pure dried feverfew leaf
or dried feverfew leaf and excipients.
Other formulations included tablets,
gelatin capsules, a glycerin extract, and
an alcohol extract (Table 1). Capsules
containing dried feverfew leaf were
opened and the content removed and
weighed before extraction. Gelatin
capsules containing feverfew extract
were assayed by opening the capsule,
removing as much extract as possi-
ble, and adding both the gelatin cap-
sule and the extract to the extraction
solvent; however, we did not attempt
to measure the feverfew mass, if any,
in the gelatin capsules. The tablets
were ground with a mortar and pes-
tle and added to the extraction sol-
vent; liquid formulations were added
directly to the solvent. All samples
were stirred for 30 minutes at room
temperature and then vacuum fil-
tered through a 0.45-µm filter. To ac-
count for solvent evaporation during
ta by feverfew extract and partheno-
lide 50–200 µg/mL. Weber and col-
leagues12 observed that parthenolide
is a selective weak inhibitor of sero-
tonin type 2A receptors but conclud-
ed that the mechanism of action of
parthenolide cannot be entirely ex-
plained by this phenomenon. These
findings are in contrast to the phar-
macologic activity of several pre-
scription antimigraine medications
(e.g., sumatriptan) that are agonists
at serotonin type 1 receptors.
There have been several reports of
feverfew activity and parthenolide
variation in feverfew products. Groe-
newegen and Heptinstall13 reported
that extracts of several commercial
feverfew products inhibited seroto-
nin release from platelets, but the ac-
tivity was less than that claimed on
the labels. Heptinstall et al.14 tested
the parthenolide content and anti-
secretory properties of multiple fe-
verfew products available in Great
Britain. Parthenolide content varied
widely among the products; some
contained no parthenolide at all.
Parthenolide content was found to
be correlated with the degree of
secretory inhibition in platelets,
lending weight to the belief that par-
thenolide is the feverfew component
that contributes most to its effects.
We measured the amount of par-
thenolide in 21 feverfew products
and compared the findings with the
amounts claimed on the products’
labels. In addition, we used the actual
parthenolide content and the dosage
recommendations on the labels to
calculate the quantity of partheno-
lide one would consume daily if fol-
lowing those recommendations.
Methods
Commercially available par-
thenolide (99.5% purity)a was used
as the reference standard. Feverfew
products were purchased at retail
stores in Oklahoma or via pharmacy
and herbal Web sites (Table 1). To
minimize analytical interference by
coeluting compounds, products
containing only feverfew (i.e., single-
ingredient products) were analyzed.
All products were stored at room
temperature (25 °C) and analyzed
before their expiration dates.
The parthenolide content of each
feverfew product was determined by
using a modified high-performance
Table 1.
Feverfew Products Analyzed
Product
No.
Label Feverfew
Content (mg)FormulationLot
Manufacturer
or Distributora
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Sundown Vitamins
Eclectic Institute Inc.
Frontier
Solaray
General Nutrition Corp.
General Nutrition Corp.
Forest Walker Superior Health
Pharmacist’s Ultimate Health
Nature’s Herbs
Nature’s Herbs (Migracin)
Nature’s Herbs
Nature’s Way
Nature’s Way
GAIA Herbs
PharmAssure (Migraplex)
Nature’s Sunshine
Enzymatic Therapy (Migracare)
Julian Whitaker, MD
Vitamin Guy
Source Naturals
Vitamin Power Natural Herbals
769942
4113/4
28316.9320
020810
77767
77767
973640
9071742
710291A
905121
901121A
906356
920995
32506298
C14925
9129848
19174
01426901
88010
NA
000207
Capsule
Capsule
Capsule
Capsule
Gelatin capsule
Capsule
Capsule
Capsule
Gelatin capsule
Capsule
Capsule
Capsule
Glycerin extract
Alcohol extract
Tablet
Capsule
Capsule
Capsule
Capsule
Tablet
Capsule
400 per capsule
125 per capsule
370 per capsule
350 per capsule
143 per capsule
500 per capsule
250 per capsule
380 per capsule
NAb
25 per capsule
384 per capsule
380 per capsule
250 per milliliter
60 per 30 drops
75 per tablet
340 per capsule
100 per capsule
150 per capsule
400 per capsule
200 per tablet
380 per capsule
aProduct trade name included if relevant.
bNA = not available or not stated on the label.

REPORTS Parthenolide content and feverfew products
1529
Am J Health-Syst Pharm—Vol 59 Aug 15, 2002
vacuum filtration, the volume of the
filtered extract was measured and
used in the calculation of partheno-
lide concentration. The residue left
from initial feverfew extraction was
stirred a second time in 100 mL of
extraction solvent and analyzed for
parthenolide content to check for in-
complete extraction (five products
were analyzed in this manner).
The feverfew extracts were ana-
lyzed for parthenolide concentration
by using an HPLC systemc calibrated
with an 8-point parthenolide con-
centration curve prepared on the ba-
sis of synthetic parthenolide (0.2–
100 µg/mL) (recalibrated each day of
analysis). Before injection, a 2-mL
volume of each extract was purified
for HPLC analysis by passage
through a 0.2-µm syringe filter.d The
autoinjector was programmed to de-
liver 20 µL of each purified 2-mL
portion onto a C8 columne equili-
brated at 30 °C. The mobile phase
consisted of water:acetonitrile
(35:65) and had a flow rate of 0.5
mL/min. Under these conditions,
parthenolide eluted at 5.5 minutes.
With the synthetic standard, the
optimal ultraviolet light (UV) ab-
sorption of parthenolide was deter-
mined by diode-array scanning
(190–800 nm) to be 200 nm with a
bandwidth of 10 nm (UV reference
wavelength, 360 nm with a 10-nm
bandwidth). Data collection and
analysis were performed with Chem-
Station software.f To increase the an-
alytical accuracy for extracts with low
concentrations of parthenolide,
weighted linear regression analysis
(1/x2) of the calibration curve was
used.
The parthenolide concentration
of each extract was multiplied by its
volume to determine the quantity of
parthenolide in each extract. Six
samples from each product were an-
alyzed. The percent content of par-
thenolide was calculated by dividing
the actual quantity of parthenolide
in one dosage unit by the weight of
actual dried feverfew leaf. The dai-
ly doses of feverfew and partheno-
lide that would be consumed if one
were to follow the labeling were
also calculated.
Results
The modified HPLC assay was a
rapid and accurate method for ana-
lyzing parthenolide concentration in
the feverfew products.15 Complete
parthenolide extraction was verified
by reextraction of the extract residue
left from the initial extraction of five
products. No residual parthenolide
was detected in the second extrac-
tions (data not shown). Therefore, it
was assumed that essentially all par-
thenolide was extracted from each
product by this method. We assayed
four products at 1, 24, and 48 hours
after performing extractions to ensure
the extraction did not cause substan-
tial degradation of parthenolide. None
of the samples varied in parthenolide
concentration by more than 5%.
All feverfew products were stored
in a dry location at room tempera-
ture and analyzed before their expi-
ration date. It was possible to directly
measure the weight of dried feverfew
leaf only in capsules containing pure
dried feverfew leaf. The quantity of
feverfew leaf in each capsule was sim-
ilar to that stated on the label and
ranged from 25 to 500 mg (Table 2).
Parthenolide content was calcu-
lated in terms of milligrams per dos-
age unit and converted to the per-
centage of the weight of the feverfew
content (Table 3). (For products oth-
er than capsules of dried leaf, this cal-
culation could not be performed be-
cause the feverfew content could not
be directly measured.) The partheno-
lide content per dosage unit ranged
from 0.02 to 3.0 mg, a 150-fold varia-
tion. The percentage of the weight of
the feverfew content represented by
parthenolide ranged from 0.14% to
0.74%, a 5.3-fold variation.
Table 2 shows the amount of fe-
verfew leaf a person would ingest
each day if he or she was to consume
the daily dose recommended on the
label (if the label recommended a
dosage range, the average amount
was used to calculate this dose). In-
take of dried feverfew leaf capsules
would range from 225 mg/day
(product 7) to 2246 mg/day (product
6), a 10-fold variation.
Table 3 shows the variation in dai-
ly parthenolide intake that would oc-
cur if each product was taken accord-
ing to the label instructions. Again, if
the label recommended a dosage
range, the average of that range was
used. (In the case of product 21, this
was difficult to determine, because
the instructions recommended “one
or more capsules per day.”) Exclud-
ing product 13, for which we could
not detect parthenolide, the daily in-
take of parthenolide ranged from
Table 2.
Label and Actual Feverfew Content in Capsule Formulations
Containing Dried Leaf Only
Product
Feverfew Content (mg) Daily Dose
of Feverfew
(mg/Day)b
Actuala
Label
1
2
3
4
6
7
11
12
16
21
400
125
370
350
500
250
384
380
340
380
403 ± 13
129 ± 10
365 ± 31
389 ± 8
499 ± 16
225 ± 1
312 ± 12
387 ± 25
312 ± 13
366 ± 8
1209
258
2190
389
2246
225
1862
1171
1404
366
aMean ± S.D. (n = 6).
bCalculated on the basis of the actual mean content and the recommended daily dose on the label.

REPORTS Parthenolide content and feverfew products
1530 Am J Health-Syst Pharm—Vol 59 Aug 15, 2002
of parthenolide (about 160-fold).
This variation represents an indus-
trywide inconsistency in the content
and preparation of feverfew and in
the daily dose recommended by
manufacturers.
While questions exist as to wheth-
er parthenolide is the sole contribu-
tor to feverfew’s purported efficacy,16
the inconsistencies in parthenolide
content in the products analyzed in
this study are unacceptable. Al-
though not studied here, it would be
reasonable to suspect that other po-
tentially active components of fever-
few are also inconsistently represent-
ed in these products.
Several possibilities may explain
the variation in parthenolide con-
tent. Parthenolide breaks down over
time.14,17 In addition, the source of
the feverfew may have an impact. For
example, feverfew grown in Germa-
ny has a higher parthenolide content
than feverfew grown in Mexico or
Yugoslavia.14 Even though all the
products in this study were pur-
0.06 mg/day (product 9) to 9.7 mg/
day (product 3), a 162-fold variation.
Discussion
In the United States, herbal prepa-
rations are legally considered dietary
supplements but are perceived and
used by most patients as medica-
tions. This use may occur to augment
or even replace prescription drug
therapy. Because of the overwhelm-
ing popularity of dietary supple-
ments, many product brands exists
for any one supplement. In this
study, we analyzed 21 feverfew prod-
ucts for feverfew leaf and partheno-
lide contents and calculated the daily
intake of these two components on
the basis of the label recommenda-
tions. The relatively high variation in
daily parthenolide and feverfew leaf
intake that we identified adds to the
growing body of literature demon-
strating a lack of pharmaceutical
consistency among dietary supple-
ment products. Especially important
is the high variation in the daily dose
chased in the United States and ana-
lyzed before their expiration date (ex-
cept those products whose labeling did
not state an expiration date), the plant
sources of the products studied may
have caused the variations in par-
thenolide content observed.
Feverfew may be moderately ef-
fective in preventing migraine head-
aches, and patients will continue to
use this herb. Feverfew may have a
role in migraine prevention, but
more rigorous scientific evidence of
its efficacy and safety is needed. The
findings of this study suggest that it
would be appropriate to advise pa-
tients who choose to use feverfew to
use one brand exclusively.
The high variation in the daily
doses of parthenolide that consum-
ers receive from feverfew products is
not acceptable. Standardized prepa-
rations of feverfew should be estab-
lished so that patients can be sure
that they are getting the same
amount of active compound each
time they use a feverfew product.
Conclusion
Large variations were observed in
the contents of dried feverfew leaf
and parthenolide, a major active in-
gredient, in commercial feverfew
products.
aSigma-Aldrich Co., St. Louis, MO 63178,
lot 119H3853.
bFisher Scientific International, Inc., Fair
Lawn, NJ 0 7413, lot 962372.
cAgilent 1100 Series system with the follow-
ing components: a vacuum degasser
(G1312A), a quaternary pump (G1311A), an
autoinjector (G1313A), a diode-array detec-
tor (G1315B), and a thermostatted column
compartment (G1316A) (Agilent Technolo-
gies, Palo Alto, CA 94303).
dFisher Scientific International.
eZorbax C8 column, 150 × 3 mm, MAC-MOD
Analytical, Inc., Chadds Ford, PA.
fChemStation software, Hewlett-Packard
Company, Palo Alto, CA.
References
1. Gurley BJ, Gardner SF, Hubbard MA.
Content versus label claims in ephedra-
containing dietary supplements. Am J
Health-Syst Pharm. 2000; 57:963-9.
2. Liberti LE, Der Marderosian A. Evalua-
tion of commercial ginseng products. J
Pharm Sci. 1978; 67:1487-9.
aMean ± S.D. (n = 6).
bPercent parthenolide, by weight, of the actual feverfew content in each dosage form.
cActual daily dose of parthenolide if taken according to the instructions on the product label.
dNA = not available (not possible to measure because the product formulation did not contain pure dried
feverfew leaf).
eBelow the lower limit of detection.
Table 3.
Parthenolide Content in Feverfew Products
Product
Daily Dose
of Parthenolide
(mg/Day)c
Mean ±±
±±
± S.D. %
Parthenolidea,b
Mean ±±
±±
± S.D.
Parthenolide Content
(mg)a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
3.00 ± 0.38
0.49 ± 0.04
1.60 ± 0.12
2.30 ± 0.09
0.45 ± 0.05
0.73 ± 0.05
0.31 ± 0.01
0.56 ± 0.04
0.02 ± 0.01
0.04 ± 0.01
0.75 ± 0.03
1.40 ± 0.16
<0.02e
0.31 ± 0.01
0.16 ± 0.02
0.45 ± 0.03
0.54 ± 0.06
1.30 ± 0.05
0.09 ± 0.02
0.62 ± 0.03
0.77 ± 0.02
0.74 ± 0.08
0.38 ± 0.01
0.44 ± 0.01
0.59 ± 0.02
NAd
0.15 ± 0.01
0.14 ± 0.01
NA
NA
NA
0.24 ± 0.01
0.36 ± 0.03
NA
NA
NA
0.14 ± 0.01
NA
NA
NA
NA
0.21 ± 0.00
9.0
0.5–1.5
9.7
2.3
0.4–0.9
2.2–4.4
0.3
1.1–1.7
0.06
0.8–2.5
4.5
4.2
<0.1
0.3
0.3
1.4–2.7
0.5
1.3
0.2
0.6–1.2
0.8

REPORTS Parthenolide content and feverfew products
1531
Am J Health-Syst Pharm—Vol 59 Aug 15, 2002
3. Dietary Supplement Health and Educa-
tion Act of 1994, Pub. L. No. 104-417, 2,
108 Stat. 4325, 21 U.S.C. §321 (1994).
4. Fetrow CW, Avila J, eds. Professional’s
handbook of complementary and alter-
native medicines. Springhouse, PA:
Springhouse; 1999.
5. Der Marderosian A, ed. Guide to popular
natural products. St. Louis: Facts and
Comparisons; 1999.
6. Murphy JJ, Heptinstall S, Mitchell JR.
Randomised double-blind placebo-
controlled trial of feverfew in migraine
prevention. Lancet. 1988; 2:189-92.
7. Capasso F. The effect of an aqueous ex-
tract of Tanacetum parthenium L. on
arachidonic acid metabolism by rat peri-
toneal leucocytes. J Pharm Pharmacol.
1986; 38:71-2.
8. Groenewegen WA, Knight DW, Heptin-
stall S. Compounds extracted from fever-
few that have anti-secretory activity con-
tain an α-methylene butyrolactone unit. J
Pharm Pharmacol. 1986; 38:709-12.
9. Groenewegen WA, Heptinstall S. A com-
parison of the effects of an extract of fe-
verfew and parthenolide, a component of
feverfew, on human platelet activity in-
vitro. J Pharm Pharmacol. 1990; 42:553-7.
10. Hayes NA, Foreman JC. The activity of
compounds extracted from feverfew on
histamine release from rat mast cells. J
Pharm Pharmacol. 1987; 39:466-70.
11. Barsby RW, Salan U, Knight DW et al.
Feverfew extracts and parthenolide irre-
versibly inhibit vascular responses of the
rabbit aorta. J Pharm Pharmacol. 1991;
44:737-40.
12. Weber JT, O’Connor MF, Haystacks K et
al. Activity of parthenolide at 5HT2A re-
ceptors. J Nat Prod. 1997; 60:651-3.
13. Groenewegen WA, Heptinstall S.
Amounts of feverfew in commercial
preparations of the herb. Lancet. 1986; 1:
44-5.
14. Heptinstall S, Awang DV, Dawson BA et
al. Parthenolide content and bioactivity
of feverfew (Tanacetum parthenium [L.]
Schultz-Bip.). Estimation of commercial
and authenticated feverfew products. J
Pharm Pharmacol. 1992; 44:391-5.
15. Zhou JZ, Kou X, Stevenson D. Rapid ex-
traction and high-performance liquid
chromatographic determination of par-
thenolide in feverfew (Tanacetum par-
thenium). J Agric Food Chem. 1999; 47:
1018-22.
16. Brown AM, Edwards CM, Davey MR et
al. Pharmacological activity of feverfew
(Tanacetum parthenium [L.] Shultz-
Bip.): assessment by inhibition of human
polymorpho-nuclear leukocyte chemilu-
minescence in-vitro. J Pharm Pharmacol.
1997; 49:558-61.
17. Awang DV, Dawson BA, Kindack DG.
Parthenolide content of feverfew (Tan-
acetum parthenium) assessed by HPLC
and 1H-NMR spectroscopy. J Nat Prod.
1991; 54:1516-21.