833. Butylated hydroxytoluene (WHO Food Additives Series 35)

BUTYLATED HYDROXYTOLUENE (BHT)

First draft prepared by
Ms Elizabeth Vavasour
Toxicological Evaluation Division, Bureau of Chemical Safety
Food Directorate, Health Protection Branch, Health Canada
Ottawa, Ontario, Canada

Explanation

Biological data
Biochemical aspects
Absorption, distribution, and excretion
Biotransformation
Effects on enzymes and other biochemical parameters

Toxicological studies
Acute toxicity studies
Short-term toxicity studies
Long-term toxicity/carcinogenicity studies
Reproductive toxicity studies
Special studies on teratogenicity
Special studies on genotoxicity
Special studies on hepatotoxicity
Special studies on nephrotoxicity
Special studies oil pulmonary toxicity
Special studies on haemorrhagic effects
Special studies on effects on the thyroid
Special studies on effects on the immmune system
Special studies on potentiation or inhibition of cancer
Special studies on other effect
Observations in human
Comments
Evaluation
References

1. EXPLANATION

Butylated hydroxytoluene (BHT) was previously evaluated by the
Committee at its sixth, eighth, ninth, seventeenth, twentieth,
twenty-fourth, twenty-seventh, thirtieth, and thirty-seventh meetings
(Annex 1, references 6, 8, 11, 32, 41, 53, 62, 73, and 94). At the
thirty-seventh meeting, the temporary ADI of 0 - 0.125 mg/kg bw,
established at the thirtieth meeting, was extended pending the results
of a study designed to elucidate the role of hepatic changes in the
development of hepatic carcinomas observed in Wistar rats following
in utero and lifetime exposure to BHT.

The results of the above study were reviewed at the present
meeting. In addition, new data relating to the previously-noted
effects of BHT on the lung, liver, kidney, clotting mechanisms and
promotion/inhibition of carcinogenesis, new long-term and reproductive
toxicity studies, genotoxicity assays and human observations were
reviewed.

The following consolidated monograph is a compilation of studies
from the previous monographs and monograph addenda and those reviewed
for the first time at the present meeting.

2. BIOLOGICAL DATA

2.1 Biochemical aspects

2.1.1 Absorption, distribution, and excretion

2.1.1.1 Mice

A single oral dose of ¹⁴C-BHT given to male and female mice
resulted in rapid absorption and distribution of ¹⁴C to the tissues.
Excretion of ¹⁴C was mainly in the faeces (41-65%) and urine
(26-50%), with lesser amounts in expired air (6-9%). The half-life for
a single dose in the major tissues studied (stomach, intestines,
liver, and kidney) was 9-11 h. When daily doses were given for 10
days, the half-life for ¹⁴C in the tissues examined was 5-15 days.
Metabolism was characterized by oxidation at one or both of the
tert-butyl groups, followed by formation of the glucuronide
conjugate, and excretion in the urine, or by excretion of the free
acid in the faeces. More than 43 metabolites were present in the urine
and faeces of mice (Matsuo et al., 1984).

2.1.1.2 Rats

When ¹⁴C-BHT was administered to rats, 80-90% of the ¹⁴C was
excreted in the urine and faeces within 96 h, and less than 0.3% of
the ¹⁴C was in expired air. Most of the ¹⁴C-BHT was excreted as
the free acid in faeces with lesser amounts in the urine. More than 43
metabolites were present in the urine and faeces of rats
(Matsuo et al., 1984).

Male F344 rats were fed BHA/BHT mixtures at levels of 0/0,
0.5/0.05, 1.0/0.1, or 2.0/0.2% in the diet and the levels of the
compounds were determined in adipose tissue after 1, 2, or 4 months.
The BHT levels found in adipose tissue were 1.4, 2.9, or 7.8 mg/kg,
respectively, in the dosed animals. On an equivalent dose basis, BHT
accumulated to ten times the level of BHA. However, neither showed any
progressive accumulation with time. Considering the rat adipose tissue
data, the mean intake of BHT by humans and the corresponding level of
BHT found in adipose tissue from 6 humans (0.12 mg/kg), previous
observations that accumulation of BHT in the adipose tissue on a
dose/body weight basis is greater in humans than in rats were
confirmed (Conacher et al., 1986).

Groups of male and female rats were maintained on diets
containing 0 or 0.5% BHT for a period of 35 days, and then for a
period on diets free of BHT. During the period on the test diets,
groups of rats were killed at 5-day intervals and fat and liver
removed for BHT analysis. During the period on BHT-free diets, rats
were killed at 2-day intervals to measure loss of BHT from the fat and
liver. There was no clear evidence of progressive accumulation of BHT

in fat during the period of administration of the test compound. BHT
levels in the fat reached a maximum level (55 mg/kg in males, 65 mg/kg
in females) within 10 days of exposure to BHT. Thereafter there was
considerable fluctuation in the observed levels. The levels of BHT in
liver were very low, the maximum BHT levels being about 5 mg/kg in
males, and 1.5 mg/kg in females. The biological half-life of BHT in
fat and liver was estimated to be 7 to 10 days (Gage, 1964).

Groups of two rats (one male, one female) were given 1 to 5 oral
doses of 44 mg/kg bw BHT on alternate days and rats in each group were
killed 24 h after the final dose. The range of the total dose
accounted for was 92 to 104% in males and 93 to 99% in females. There
was an indication of sex difference in the route of excretion, females
excreting 19-43% of the radioactivity in urine and males only 3-15%.
Eight days after administration of 5 doses, 92% of the radioactivity
had been excreted by males and 97% by females. Subcutaneous
administration of graded doses of BHT to female rats revealed
substantial faecal excretion but the rate of excretion decreased with
increasing dose. There was no evidence of accumulation of ¹⁴C-BHT in
the body under the conditions of repeated oral dosage (Tye et al.,
1965).

Rats were given single oral doses (1-100 mg/rat) of ¹⁴C-BHT and
approximately 80 to 90% of the dose was recovered within four days in
the urine and faeces. Of the total radioactivity, 40% and 25% appeared
in the urine of females and males, respectively. After 4 days,
approximately 3.8% of the dose was retained, mainly in the alimentary
tract. A substantial portion of the radioactivity was found in the
bile collected from two rats (one male, one female) over a period of
40 h. The relatively slow excretion of BHT is probably attributable to
enterohepatic circulation rather than to tissue retention (Daniel &
Gage, 1965).

The liver and body fat of rats fed a diet containing 0.5% BHT for
35 days were analyzed. The concentration of BHT in the liver never
rose above 5 mg/kg in males and 1.5 mg/kg in females. In the body fat,
the level fluctuated around 30 mg/kg in males and 45 mg/kg in females.
Fat from rats returned to BHT-free diet showed a progressive fall in
the concentration of BHT, the half-life being about 7 to 10 days.

The daily excretion of radioactivity in urine and faeces was
studied in rats given an oral dose of ¹⁴C-labelled BHT (12 mg/kg
bw). Excretion became negligible by the sixth day after administration
when about 70% of the injected dose had been recovered. Less than 1%
was excreted as carbon dioxide in the expired air. About 50% of the
radioactivity was excreted in the bile during the 24-hour period
following the oral dose (Daniel & Gage, 1965).

The BHT content of fat and liver of rats given diets containing
0.5 or 1% BHT for periods up to 35 and 50 days, respectively, were
analyzed. With 0.5% BHT in the diet, a level of approximately 30 mg/kg
in the fat was reached in males and 45 mg/kg in females, with
approximately 1-3 mg/kg in the liver. With 1% BHT, the level in the
fat was 50 mg/kg in males and 30 mg/kg in females. On cessation of
treatment, the level of BHT in fat fell with a half-life of 7 to 10
days (Daniel & Gage, 1965).

The level of BHT in the fat reached a plateau at approximately
100 mg/kg after 3 to 4 days when daily doses of 500 mg/kg bw were
given by intubation. A daily dose of 200 mg/kg bw for one week
produced a level of about 50 mg/kg (Gilbert & Golberg, 1965).

Rats were given doses of 100 µg of ³H-labelled BHT
intraperitoneally and the urinary output of radioactivity was measured
for 4 consecutive days. After 4 days of dosing, 34.5% of the injected
radioactivity was recovered in urine (Ladomery et al., 1963). The
same dose of BHT (100/µg) labelled with ¹⁴C was given to rats, and
34% of the radioactivity was excreted in the urine in the first four
days, in close agreement with the previous result using tritiated BHT
(Ladomery et al., 1967a).

After a single parenteral dose (100 µg) of ¹⁴C-BHT, rats
excreted 3235% of the radioactivity in the urine, and 35-37% in the
faeces, in a 4-day period. The intestinal contents together with the
gut wall contained most of the remaining activity. Biliary excretion
was rapid, and the radioactive material in bile was readily absorbed
from the gut, suggesting a rapid enterohepatic circulation
(Ladomery et al., 1967b).

White male Wistar rats (290-350 g) were administered
[¹⁴C]-labelled BHT, or its alcohol [BHT-CH₂OH] or its aldehyde
[BHT-CHO] or acid [BHT-COOH] derivative by i.v. or i.p. injection. The
overall excretion of BHT and its related compounds excreted in urine
and faeces was studied for a 5-day period, and biliary excretion
monitored for 120426 h after i.p. injection. For the low doses of the
compounds tested (100 µg), there were no significant differences in
the total recovery of ¹⁴C during the 5 days urinary and faecal
excretion and 120-126 h biliary excretion. However, there were
differences in ratio of urinary to faecal excretion of ¹⁴C. The
major metabolite present in early bile after i.p. injection of the
labelled compounds was BHT-COOH or its ester glucuronide. Late bile
after acid hydrolysis showed BHT-COOH to be the major ¹⁴C component
(Holder et al., 1970a).

Temporal concentration changes in BHT and BHT-quinone methide
were studied following a single oral dose of 800 mg BHT/kg bw in rats,
which were monitored in plasma, the GI tract, and some adipose tissues
over a 48-hour period using GLC. Groups of 4 or 5 male Sprague-Dawley
rats were sacrificed at 0.5, 1, 3, 6, 12, 18, 24, 30, 36 or 48 h after
administration of BHT. The amount of BHT in the gastric contents
stayed constant for 12 h, then declined rapidly to 24 h, followed by a
gradual decline to 48 h. BHT was not detected in the plasma at any of
the time points, while BHT-quinone methide showed a rapid rise between
12 and 18 h after administration, followed by a gradual decline to
48 h. BHT was detected in epididymal, subcutaneous, perirenal and
brown dorsal adipose tissue soon after administration. A peak
concentration was noted in epididymal and subcutaneous fat at 18 h,
but did not show this trend in perirenal or brown dorsal fat.

In a second experiment, in which groups of 5 or 6 rats were
sacrificed 4-7 or 24-27 h after administration of BHT, a single oral
dose of 800 mg/kg bw resulted in a significant increase in the weight
of stomach contents in both fasted and non-fasted rats at 4-7 h. By
24-27 h after administration, there was no difference in the weight of
stomach contents between non-fasted treated animals and controls. This
study suggested that high single doses of BHT caused a delay in
gastric emptying which was reflected in the delay in plasma levels of
metabolite.

In another experiment, 800 mg/kg bw BHT was instilled
intraduodenally to 5 anaesthetized rats and 30 minutes later the
concentrations of BHT and its metabolites were measured in blood from
the portal vein and descending aorta, and in liver and epididymal
adipose tissue. BHT and/or BHT radical was detected in all but the
descending aorta, while BHT-quinone methide was not present in any of
the samples collected at 30 minutes after administration (Takahashi,
1990).

2.1.1.3 Chickens

One-day old chicks were given ¹⁴C-BHT at a level of 200 mg/kg
in the feed for 10 weeks. At broiler age, edible portions had residues
amounting to 1-3 mg/kg of BHT and metabolites. Similar diets given to
laying hens produced residues in eggs of 2 mg/kg after 7 days, the
level thereafter remaining constant (Frawley et al., 1965a).

When feed containing 500 or 100 mg/kg BHT was given to laying
hens, residues of 20 and <5 mg/kg were found in the fat fraction of
eggs, respectively. In the broiler chicken, over a period of 21 weeks,
the residues in body fat were 55 mg/kg on the 500 mg/kg diet and
< 5 mg/kg on the 100 mg/kg diet (Van Stratum & Vos, 1965).

2.1.1.4 Humans

Four human male subjects were administered a single dose of
approximately 40 mg ¹⁴C-labelled BHT. About 75% of the administered
radioactivity was excreted in the urine. About 50% of the dose
appeared in the urine in the first 24 h, followed by a slower phase
which probably represented the release of the compounds or their
metabolites stored in tissues. In humans, the bulk of the
radioactivity was excreted as the ether insoluble glucuronide of a
metabolite in which the ring methyl group and one tert-butyl methyl
group were oxidized to carboxyl groups, and a methyl group on the
other tert-butyl group was also oxidized, probably to an aldehyde
group. BHT-acid, free and conjugated, was a minor component of the
urine and the mercapturic acid was virtually absent. The rapidity of
the first phase of the urinary excretion in humans suggested that
there was no considerable enterohepatic circulation as had been
observed in the rat (Daniel et al., 1967).

Reported values of BHT in body fat were 0.23 ± 0.15 mg/kg
(11 individuals, residents of the United Kingdom) and 1.30 ±
0.82 mg/kg (12 individuals, residents of the United States of America)
(Collings & Sharratt, 1970).

Based on reported BHT levels in human fat in Japan, United
Kingdom and USA and the calculated dietary intakes of BHT, a
bioconcentration factor in humans (BCF, wet weight basis) of 0.36 was
calculated for BHT. This BCF was 45 times higher than that calculated
for the rat. In comparison, the BCF for total DDT was calculated at
1279 (Geyer et al., 1986).

The disposition of single oral doses of BHT was compared in
humans and rats. A single oral dose of 0.5 mg/kg bw of BHT was
ingested by 7 healthy male volunteers after fasting overnight. Blood
samples were taken after 0, 15, 30, 45, 60, 75, 90, 120, 150, 180 and
240 minutes. Total urine and faeces were collected for 2 days. In
another experiment, 5 healthy female volunteers ingested 0.25 mg/kg bw
of BHA and one week later 0.25 mg/kg bw of BHT. After another week,
0.25 mg/kg bw of BHT plus 0.25 mg/kg bw of BHA were given
simultaneously. After each dosing, blood samples were taken as
described above. Similar experiments were conducted in male Wistar
rats, except that the doses used were 20, 63, or 200 mg/kg bw of BHT.
In rats, peak plasma concentrations of BHT (0.2, 0.3, and 2.3 µg/ml)
were seen after 2.6 h. Simultaneous administration of BHA produced
significantly lower plasma concentrations between 0.5 and 3 h. Large
variations were seen in humans in plasma levels of BHT. The mean peak
plasma level was 0.09/µg/ml, reached after 1.5 h. The plasma
concentrations were not influenced by simultaneous administration of
BHA. In the rat, approximately 2% of the dose was excreted as BHT-COOH
in the urine (equal amounts of conjugated and unconjugated compound)

and 10% as BHT in the faeces in 4 days. In humans, 2.8% of the dose
was found in the urine as BHT-COOH (mainly conjugated) and no BHT
could be detected in the faeces. On a comparative dose basis, it seems
that BHT in plasma reaches a higher level in humans than in rats
(Verhagen et al., 1989).

2.1.2 Biotransformation

2.1.2.1 Mice

The oxidative metabolism of BHT by liver microsomes from 3 inbred
mouse strains, NGP/N, A/J and MA/MyJ was compared. The strain order
shown is the order of increasing susceptibility of these mice to BHT
lung tumour promotion which correlates with their increasing ability
to produce BHT-BuOH, by hydroxylation of BHT at one of the
tert-butyl groups. Four weekly i.p. injections of BHT selectively
induced the BHT oxidation pathway leading to formation of BHT-BuOH
(Thompson et al., 1989).

The metabolism of BHTOOH was examined to assess the role of
reactive intermediates in mediating tumour promotion in mouse skin.
Incubation of BHTOOH with either isolated neonatal mouse keratinocytes
or a cell-free haematin system resulted in the generation of
the BHT-phenoxyl radical. Only one non-radical metabolite of
BHTOOH-BHT-quinol was detected in keratinocytes, while incubation of
BHTOOH with haematin produced several metabolites: oxacyclopentenone,
BHT-quinone, BHT, BHT-stilbene quinone, and BHT-quinone methide. In
contrast to the action of BHTOOH, topical application of epidermal
doses of BHT-quinol, BHT-quinone, BHT-stilbene quinone, as well as BHT
itself to mouse skin, did not induce epidermal ornithine decarboxylase
activity (Taffe et al., 1989).

Co-administration of BHA (200 mg/kg bw) with a subtoxic dose of
BHT (200 mg/kg bw) enhanced the lung toxicity of BHT in male mice. BHA
co-administration significantly increased the radioactivity covalently
bound to lung macromolecules at 4-8 h after [¹⁴C]-BHT. The
pretreatment also reduced the rate of in vitro metabolism of BHT in
mouse liver supernatant. The co-administration of BHA and BHT caused a
decrease in metabolism of BHT in the liver with the result that the
lung was exposed to a larger amount of BHT (Yamamoto et al., 1988).

2.1.2.2 Rats

Examination of the biliary metabolites from i.v. and i.p. doses
of small amounts of ¹⁴C-BHT, showed the presence of four principal
metabolites. Some 34 to 53% of the ¹⁴C-label in the bile was
identified as 3-5-di- t-butyl-4-hydroxy-benzoic acid, which was
probably present as the ester glucuronide. The other metabolites
present were 3,5-di- t-butyl-4-hydroxybenzaldehyde, 3-5-di- t-butyl-
4-hydroxybenzyl alcohol, and 1,2-bis (3,5-di- t-butyl-4-hydroxyphenyl)
ethane (Ladomery et al., 1967b).

Rats were given a single dose of ¹⁴C-BHT and urine and
bile collected for periods ranging from 48 to 96 h. Faeces were
also collected during this same period. About 19% to 59% of the
radioactivity appeared in the urine during this period, and 26% to 36%
in the bile. The major metabolites in the urine were 3,5-di- tert-
butyl-4-hydroxybenzoic acid, both free (9% of the dose), as well as
glucuronide (15%) and S-(3,5-di- tert-butyl-4-hydroxybenzyl)-
N-acetylcysteine. The ester glucuronide and mercaptic acid were also
the major metabolites in rat bile. Free 3,5-di- tert-butyl-4-
hydroxybenzoic acid was the major metabolite in faeces (Daniel
et al., 1968).

Male Wistar rats were dosed intraperitoneally with 200 mg/kg bw
BHT-acid or 2,6-di- tert-butyl phenol (DBP). Faeces and urine were
collected for 5 days after administration of BHT-acid and for 3 days
after administration of DBP. Bile was collected from rats treated with
BHT-acid for 24 h after administration. Following administration of
BHT-acid, the metabolites DBP, 2,6-di- tert-butyl- p-benzoquinone
(BBQ), 2,6-di- tert-butylhydroquinone (BHQ) glucuronide and BHT-acid
glucuronide were identified in the faeces and bile. This suggested
that BHT-acid, considered as a main metabolic end-product of BHT,
was metabolized to the quinone and hydroquinone following its
decarboxylation to form the di-substituted phenol, DBP. An alternative
route for formation of the quinone, BBQ, was a homolysis reaction of
the O-O bond of the hydroperoxide which is catalyzed by rat liver
cytochrome P-450 (Yamamoto et al., 1991).

2.1.2.3 Rabbits

Rabbits were given single or repeated doses of BHT in the range
of 400-800 mg/kg bw. About 16% of the dose was excreted as ester
glucuronide and 19% as ether glucuronide. Unconjugated phenol (8%),
ethereal sulfate (8%) and a glycine conjugate (2%) were also excreted.
Excretion of all detectable metabolites was essentially complete 3 to
4 days after administration of the compound and about 54% of the dose
was accounted for as identified metabolites (Dacre, 1961).

The metabolism of butylated hydroxytoluene (BHT) orally
administered to rabbits in single doses of 500 mg/kg bw was studied.
The metabolites 2,6-di- tert-butyl-4-hydroxymethylphenol
(BHT-alcohol), 3,5-di- tert-butyl-4-hydroxy-benzoic acid (BHT-acid)
and 4,4'-ethylene- bis-(2,6-di-tert-butylphenol) were identified. The
urinary metabolites of BHT comprised 37.5% as glucuronides, 16.7%
as ethereal sulfates and 6.8% as free phenols; unchanged BHT was
present only in the faeces (Akagi & Aoki, 1962a); 3,5-di- tert-butyl-
4-hydroxybenzaldehyde (BHT-aldehyde) was also isolated from rabbit
urine (Aoki, 1962). The main metabolic pathway was confirmed by
administering BHT-alcohol to rabbits and isolating BHT-aldehyde,
BHT-acid, the ethylene- bis derivative and unchanged BHT-alcohol in
the urine (Akagi & Aoki, 1962b).

2.1.2.4 Ruminants

When [¹⁴C]BHT was activated in vitro by the prostaglandin H
synthase system in microsomes from ram seminal vesicles or by
horseradish peroxidase, significant covalent binding to protein could
be detected. BHT-quinone methide was detected at only minor
concentrations, therefore an intermediate free radical was suggested
as an active metabolite. Addition of BHA to the medium greatly
increased the formation of BHT-quinone methide and covalent binding to
proteins (Thompson et al., 1986).

2.1.2.5 Humans

A group of 8 men each received 100 mg of BHT on two occasions
separated by a 4-day interval. Urine was collected for 24 h after BHT
administration. The metabolites were identified as BHT-COOH and
benzoyl-glycine. In another study in which two adults were given 1.0 g
of BHT, BHT-COOH and its ester glucuronide were the only major
metabolites identified in urine (Holder et al., 1970b).

2.1.2.6 Combined species

The metabolism of BHT was studied with liver and lung microsomes
from rats and mice. Two main metabolic processes occurred,
hydroxylation of alkyl substituents and oxidation of the aromatic ¼
electron system. The former led to the 4-hydroxymethyl product
(BHT-CH₂OH) and a primary alcohol resulting from hydroxylation of a
t-butyl group (BHT- tBuOH). Additional metabolites were produced by
oxidation of BHT-CH₂OH to the corresponding benzaldehyde and benzoic
acid derivatives. Hydroxylation of BHT- tBuOH occurred at the
benzylic methyl position, and the resulting diol was oxidized further
to the hydroxybenzaldehyde derivative. Oxidation of the ¼ system led
to BHT-quinol (2,6-di- t-butyl-4-hydroxy-4-methyl-2,5-cyclo-
hexadienone), BHT-quinone (2,6-di- t-butyl-4-benzoquinone), and
BHT-quinone methide (2,6-di- t-butyl-4-methylene-2,5-cyclohexa-
dienone) probably via the hydroperoxide (BHTOOH). Derivatives of the
quinol and quinone with a hydroxylated t-butyl group were also
formed. Quantitative data demonstrated that BHT-CH₂OH was the
principal metabolite in rat liver and lung microsomes. The mouse
produced large amounts of both BHT-CH₂OH and BHT- tBuOH in these
tissues. The metabolite profile was similar in rat liver and lung.
Mouse lung, however, produced more quinone relative to other
metabolites than mouse liver (Thompson et al., 1987).

The in vitro peroxidase-catalyzed covalent binding of BHT to
microsomal protein and the formation of BHT-quinone methide was
enhanced by addition of BHA. Several other phenolic compounds commonly
used in food also enhanced the metabolic activation of BHT. Microsomes

from lung, bladder, kidney medulla and small intestine of various
animal species, including humans, were also able to support this
interaction of BHA and BHT using either hydrogen peroxide or
arachidonic acid as the substrate (Annex 1, reference 95).

Phenobarbital (PB) pretreatment of Sprague-Dawley rats and A/J
mice had little or no effect on respective quinone methide formation
from BHT or BHT-BuOH in pulmonary microsome preparations, but resulted
in a 6-37 fold induction of this activity in hepatic microsomes from
these species. PB administration had little effect on the two-step
oxidation of BHT to QM-OH in pulmonary microsomes of the mouse, while
in hepatic microsomes, PB pre-treatment resulted in a greater than
100-fold increase in this activity. This was found to be mainly due to
a greater than 100-fold increase in the initial tert-butyl
hydroxylation step in mouse liver microsomes. The enhancement was
somewhat higher than for 7-pentoxyresorufin O-dealkylase activity,
demonstrating that tert-butyl hydroxylation could serve as a
specific marker for the enzyme. The results also showed that pulmonary
microsomes from mice, but not rats, had a relatively high constitutive
P-450 activity for tert-butyl hydroxylation of BHT, supporting
the proposal that this metabolite was involved in BHT-induced
pneumotoxicity. Two cytochrome P-450 inhibitors, SKF 525-A and
metapyrone, inhibited the conversions of BHT to QM and QM-OH to a
similar extent in PB-treated mouse liver microsomes; the terpenoid
alcohol cedrol was found to selectively inhibit BHT conversion to
QM-OH. This compound has been found previously to inhibit
pneumotoxicity when administered to mice prior to treatment with BHT.
In untreated mouse liver microsomes, BHT hydroxylation to BHT-MeOH
showed the greatest activity, with oxidation to QM second and
hydroxylation to BHT-BuOH last (Bolton & Thompson, 1991).

2.1.3 Effects on enzymes and other biochemical parameters

2.1.3.1 Mice

Mice (BALB/c strain) were maintained on a diet containing 7.5 g
BHT/kg of feed. After 3 weeks on the test diet, there was an enhanced
activity in plasma esterases which persisted throughout the
experimental period of 20 weeks. Following electrophoretic separation
of the esterases, the increased enzyme activity was shown to be
located in two specific bands (Tyndall et al., 1975).

Dietary administration of BHT to male Swiss Webster mice resulted
in a marked increase in hepatic microsomal epoxide hydrolase and
glutathione-S-transferase (Hammock & Ota, 1983).

2.1.3.2 Rats

Feeding experiments were carried out on 45 pairs of weanling male
rats for 5 to 8 weeks with diets containing 0, 10 or 20% lard
supplements to which 0.01, 1 or 5 g BHT/kg had been added. At
0.01 g/kg, no changes were observed in any of the serum constituents
studied, while at 5 g/kg an increase in the serum cholesterol level
was seen within 5 weeks. Female rats fed for 8 months a diet
containing a 10% lard supplement with 1 g BHT/kg showed increased
serum cholesterol levels, but no other significant changes. Diets
containing 5 g BHT/kg in 10% and 20% lard supplements fed to female
rats for the same period increased serum cholesterol, phospholipid and
mucoprotein levels (Day et al., 1959).

In further work with rats, it was found that increased output of
urinary ascorbic acid accompanied liver enlargement induced by BHA or
BHT in onset, degree and duration, being rapid but transient with BHA,
and slower in onset but more prolonged with BHT (Gaunt et al.,
1965a). The simultaneous stimulation of processing enzyme activity,
increase in urinary ascorbic acid output, and increase in relative
liver weight brought about by BHT was unaffected by 14 days of dietary
restriction, and all these changes except liver weight were reversible
during 14 days' recovery on normal diet (Gaunt et al., 1965b).

Rats given BHT by daily intubation showed increased activity of
some liver microsomal enzymes. Stimulation of enzyme activity
correlated with an increase in relative liver weight. The threshold
dose for these changes in enzyme activity in female rats was below
25-75 mg BHT/kg bw/day. The storage of BHT in fat appeared to be
influenced by the activity of the processing enzymes. In rats given
500 mg/kg bw/day, the level of BHT in fat attained values of 230 mg/kg
in females and 162 mg/kg in males by the second day, by which time the
relative liver weight and processing enzyme activities had become
elevated. Thereafter, liver weight and enzyme activity continued to
rise but the BHT content of fat fell to a plateau of about 100 mg/kg
in both sexes (Gilbert & Golberg, 1965).

Groups of 12 SPF Carworth rats equally divided by sex were
administered BHT dissolved in arachis oil daily for one week, at dose
levels of 50, 100, 200 or 500 mg/kg bw. A group of 8 rats served as
control. The animals were killed 24 h after the final dose, and
histological and biochemical studies (glucose-6-phosphatase: and
glucose-6-phosphate dehydrogenase) made on the livers of all animals.
A histochemical assessment of the livers of test animals was also
carried out. BHT caused an increase in liver weight in males at dose
levels of 100 mg/kg bw and greater, and in females at 200 mg/kg bw and
greater. BHT caused a decrease in glucose-6-phosphatase activity in
females at dose levels greater than 100 mg/kg bw, and an increase in
glucose-6-phosphate dehydrogenase in both males and females at the
highest dose tested. In another study, rats were dosed according to

the above schedule and then maintained for 14 to 28 days following the
final dosing. By day 28, no biochemical changes were observed, and
relative liver weights returned to normal by day 14 (Feuer et al.,
1965).

Rats fed diets containing BHT at levels of 100 to 5000 mg/kg for
12 days showed liver enlargement, as well as increased activity of
liver microsomal biphenyl-4-hydroxylase, at all levels except the
lowest level of 100 mg/kg. Enzyme activity was not significantly
altered at 5000 mg/kg BHT fed for one day (Creaven et al., 1966).

Rats (male and female Carworth Farm SPF) were given an oral dose
of BHT equivalent to 500 mg/kg bw. Dosing was from 1 to 5 days, and
rats varying in size from 100-400 g body weight were used. Microsomal
preparations from the livers of treated rats were assayed for BHT
oxidase, an enzyme that metabolizes BHT to the BHT alcohol
(2,6-di- tert-butyl-4-methylphenol to 2,6-di- tert-butyl-4-hydroxy
methylphenol). Treatment of female rats with BHT (500 mg/kg bw/day for
5 days) caused a six fold increase in the activity of the enzyme/gram
of liver and a 35% increase in relative liver weight, both being
prevented by actinomycin D. The induction was more pronounced in males
than in females, and the induction of the enzyme, low in rats in the
100 g body-weight range, reached a maximum in rats in the 200 g
body-weight range, and fell in larger animals (300-400 g range)
(Gilbert & Golberg, 1967).

Groups of female Alderly Park SPF rats were maintained on diets
containing 0%, 0.01%, 0.1%, 1% or 5% BHT for periods up to 28 days,
and then on diets free of BHT for 56 days. Animals were killed in
groups of four, 2 being used for enzyme assay (aminopyrene
demethylase) and 2 for electron microscopy. The increase in enzyme
activity was directly related to the dietary level of BHT. No
detectable increase was observed at the lowest level over the 28-day
feeding period. Following withdrawal of BHT from the diet, the enzyme
level returned to normal in all test animals. The degree of
endoplasmic reticulum proliferation was proportional to the amount of
BHT in the diet and the duration of feeding at the 5% and 1% level. At
the 0.1% level, there was a transient rise in smooth endoplasmic
reticulum. No proliferation was observed at the 0.01% level. Following
removal of BHT from the diet, there was a rapid disappearance of the
proliferated smooth endoplasmic reticulum.

In a second study, groups of rats were fed diets containing 1%
BHT for 10 days, and then for a second period of 10 days after an
interval of 20 days on a normal diet. The animals were killed in
groups of five at 10, 30, 40, 42 or 47 days. Livers were removed for
aminopyrene demethylase assay and electron microscopy. Enzyme activity
did not differ significantly following both 10-day periods of
administration of BHT. Electron microscopy showed similar smooth
endoplasmic reticulum response during both these periods (Botham
et al., 1970).

Microsomal preparations from livers of rats, dosed daily with
450 mg/kg bw BHT for up to 7 days, showed an increased capacity to
incorporate labelled amino acids, when compared to preparations from
controls. BHT also stimulated the in vivo incorporation of amino
acids, mainly into the proteins of the endoplasmic reticulum
(Nievel, 1969).

A group of 23 female SPF rats (Wistar strain) was administered
500 mg/kg BHT dissolved in rape-seed oil, for 11 days. The control
group was administered rape-seed oil alone. Groups of 7 rats were
killed following administration of the final dosing. The remaining
rats were maintained without further exposure to BHT, and killed on
days 28 or 63 of the study. Livers of the rats were examined for
weight, DNA content and number of cell nuclei. Treatment with BHT
resulted in enlargement of the liver, with a concomitant increase in
its DNA content, and in the number and ploidy of its nuclei. The liver
mass returned to normal within two weeks. However, the DNA content of
the liver of BHT-treated animals remained elevated up to the time of
termination of this study, and there was no reduction in the total
number of nuclei or the degree of ploidy (Hermann et al., 1971).

Two groups of 10 male and 10 female rats (Alderly Park, SPF
Wistar strain) were dosed by stomach tube with 200 mg/kg bw/day of BHT
dissolved in maize oil for 7 days. Four rats/sex dosed with an
equivalent amount of maize oil served as controls. Urinary ascorbic
acid excretion was measured in urine samples collected following 5
days on the test compound. The animals were killed 24 h after the
final dose and the livers removed for biochemical assays (aminopyrine
demethylase-AMPM, hexobarbitone oxidase-HO, cytochrome P-450, and
glucose-6-phosphatase), and electron microscopy. Another group of
treated rats was maintained for a 7-day recovery period, and a similar
battery of liver studies was carried out. Administration of BHT
resulted in an increase of urinary excretion of ascorbic acid which
remained constant throughout the treatment period. Following cessation
of BHT treatment there was a gradual return towards control values.
There were significant sex differences in some of the biochemical
responses to BHT, with the exception of the glucose-6-phosphatase
activity. Female rats showed a marked increase in APDM and HO
activity, which was not observed in male rats. Cytochrome P-450 levels
were increased in both males and females. The biochemical parameters,
with the exception of APDM activity in female rats, returned to normal
following the 7-day recovery period. Electron microscopy showed
significant proliferation of the smooth endoplasmic reticulum of the
hepatic cells. No other morphological changes were detected
(Burrows et al., 1972).

BHT (500 mg/kg bw/day) was administered by gavage to groups of
young Wistar male and female rats for 7 days and the animals were
housed in metabolism cages. Control animals received corn oil vehicle
only. They were then sacrificed and liver enzymes (aniline-

4-hydroxylase, biphenyl-4-hydroxylase, ethyl morphine N-demethylase,
and 4-methyl umbelliferone glucuronyl transferase) were assayed and the
cytochrome P-450-CO interaction spectrum evaluated. Urinalysis using
GC was conducted to assay for D-glucaric acid, D-glucuronic acid,
1-gulonic acid, xylitol and L-ascorbic acid. Administration of BHT
enhanced all the parameters measured with the exception of the hepatic
microsomal protein content. BHT was a more potent inducer of
xenobiotic metabolism in female rats (Lake et al., 1976).

BHT in the diet of Sprague-Dawley rats resulted in a marked
decrease in the NADPH-cytochrome P450 reductase activity of isolated
liver microsomal preparations. This effect was not observed when BHT
was added in vitro to liver microsomes (Rikans et al., 1981).

Dietary BHT was also shown to affect the carboxylation process in
the conversion of rat liver microsomal protein to prothrombin
(Takahashi & Hiraga, 1981a).

Rats fed a diet containing 0.4% BHT showed an increase in GSH-S
transferase activity in the liver, but not in lungs and kidneys.
GSH-reductase levels were increased in liver and lungs
(Partridge et al., 1982).

BHT at 300-6000 mg/kg in the diet caused a dose-dependent
increase in gamma-glutamyl transpeptidase in normal F344 male rats.
However, cytosolic glutathione S-transferase was only enhanced at
dietary concentrations of 3000 or 6000 mg/kg (Furukawa et al.,
1984).

Groups of 4 male F344 rats were pretreated with buthionine
sulfoximine, a glutathione-depleting agent (900 mg/kg bw), and after
1 h given intraperitoneal injections of BHT (100, 250, 400, or
500 mg/kg bw). A dose-related elevation of serum GOT and GPT
activities was observed. BHT or buthionine sulfoximine alone had no
effect. In contrast, pretreatment with cysteine (100-200 mg/kg bw)
inhibited the elevation of serum enzyme activities at a toxic dose of
BHT (1000 mg/kg bw) (Nakagawa, 1987).

Supplementation of AAF-containing diets with 0.3% BHT, which
affords protection against AAF hepatocarcinogenesis in high-fat fed
Sprague-Dawley rats, protected and/or induced total hepatic nuclear
envelope cytochrome P-450 content. Short-term feeding with AAF without
BHT resulted in a marked loss of total hepatic nuclear envelope P-450
(Carubelli & McCay, 1987). Immunological studies showed that BHT
enhanced the AAF-dependent induction of P-450c, but not P-450d. BHT by
itself had no effect on these nuclear envelope enzymes
(Friedman et al., 1989).

Administration of 0.5% BHT in the diet of male Wistar rats for 2
weeks increased UDP-glucuronosyl transferase activity in liver
microsomes for several substrates to 236-269% of controls. The amounts
of UDP-glucuronosyl transferase protein and associated mRNA in liver
microsomes were also increased, paralleling the increases in enzyme
activity. In addition to induction of hepatic activity, BHT treatment
resulted in increased activity in microsomes from the kidney and small
intestine (Kashfi et al., 1994).

2.1.3.3 Rabbits

Acute effects on electrolyte excretion, similar to those
described for large doses of BHA were also obtained following
administration of BHT at doses of 500-700 mg/kg bw. No such effects
were observed at lower dosage levels (Denz & Llaurado, 1957).

2.1.3.4 Monkeys

Groups of 2-4 juvenile rhesus monkeys (Macaca mulatta) were fed
BHT dissolved in corn oil at dose levels of 0, 50, or 500 mg/kg bw for
4 weeks. Blood samples were taken prior to treatment and then at
weekly intervals from the control and test animals in the high-dose
group, and from test animals in the low-dose group at the end of the
4-week period, for determination of total plasma cholesterol, lipid
phosphorus and triglyceride. Liver biopsies were taken from the test
animals in the high group at two weeks. At the end of the test period
all animals were fasted for 24 h and sacrificed, and liver and blood
samples obtained. Liver samples were analyzed for succinic
dehydrogenase and susceptibility to peroxidation. Extracted liver
lipids were analyzed for total cholesterol, lipid phosphorus and
triglycerides. Total cholesterol levels in plasma and liver were
significantly lowered. Lipid phosphorus levels in the plasma were
increased at the high-dose level, as were cholesterol:lipid phosphorus
ratios in the plasma and liver. The susceptibility of liver lipids to
oxidation was reduced in the high-dose group (Branen et al., 1973).

2.2 Toxicological studies

2.2.1 Acute toxicity studies

The results of acute toxicity studies with BHT are summarized in
Table 1.

Acute oral, intraperitoneal (mice) and eye irritation (rabbits)
and skin irritation (rats) were measured for 7 breakdown products of
BHT. All compounds tested were less toxic than the parent compound
(Conning et al., 1969).

Table 1. Acute toxicity studies with BHT

Animal Route LD₅₀ Approximate Reference
(mg/kg bw) lethal dose
(mg/kg bw)

Rat oral > 1700-1970 - Deichmann et al., 1955

Cat oral - 940-2100 Deichmann et al., 1955

Rabbit oral - 2100-3200 Deichmann et al., 1955

Guinea-pig oral - 10 700 Deichmann et al., 1955

Rat oral 2450 Karplyuk, 1959

Mouse oral 2000 Karplyuk, 1959

As shown in Table 2, the LD₅₀ (i.p.) for BHT showed
considerable differences for strains of inbred and non-inbred male
mice. In all cases death occurred 4 to 6 days after administration of
BHT, and was accompanied by massive edema and haemorrhage in the lung
(Kawano et al., 1981).

Table 2. Variation in LD₅₀ with strains of mice (Kawano et al., 1981)

Strain LD₅₀ (mg/kg bw)

DBA/2N (inbred) 138
BALB/cNnN (inbred) 1739
C57BL/6N (inbred) 917
ICR-JCL (non-inbred) 1243

2.2.2 Short-term toxicity studies

2.2.2.1 Mice

In order to estimate the MTD for BHT in mice, groups of 5/sex
B6C3F₁ mice, 6-week old, were fed diets containing 0, 3100, 6200,
12 500, 25 000 or 50 000 mg BHT/kg of feed for 7 weeks. Each animal was
weighed twice weekly. Gross necropsy was performed on all animals. One
female mouse in the 25 000 mg/kg group, and 1 male and 4 female mice
in the 50 000 mg/kg group died before the end of the study.

Body-weight decrements, mostly dose-related, were noted in all
treatment groups compared with controls. Histopathologic examination
of male mice receiving 25 000 mg/kg revealed a very small amount of
centrilobular cytoplasmic vacuolization of hepatocytes which was not
observed in females receiving 12 500 mg/kg (high-dose females were not
examined) (NCI, 1979).

A similar study was conducted to establish the dose levels for a
subsequent carcinogenicity study. Groups of B6C3F₁ mice (10/sex)
received BHT in the diet at concentrations of 0.25%, 0.5%, 1%, 2%, or
4% for 10 weeks. Twenty mice/sex were used for the control group. Both
male and female mice receiving the highest dose of BHT experienced a
retardation of body-weight gain which exceeded 10% of control values.
In addition, histopathological examination of mice in the 4% group,
revealed marked starvation atrophy of the spleen, heart and kidneys.
None of these changes were noted in mice at the next lower dose level
(2% diet) or any of the other groups. The MTD was considered to be 2%
diet (Inai et al., 1988).

Groups of male and female C₃H mice (17-39 mice/group), 6-10
weeks old, were maintained for 10 months on a semi-synthetic diet
containing 0.05 or 0.5% BHT. Control groups were maintained on
BHT-free semi-synthetic diet or commercial lab chow. At the end of the
test period, the liver and lungs were excised and inspected grossly
for proliferative lesions. Of the proliferative lesions considered to
be clearly identifiable as tumours, approximately 50% were examined
microscopically. Mice maintained on diets containing BHT had lower
body weights than controls. Male mice fed BHT showed an increase in
liver tumours, compared to controls. Histologically, the tumours were
identified as hepatocellular adenomas. No increase was observed in
female mice. The reported incidence of liver tumours in male C₃H
mice was 38% (10/26) in the 0.5% BHT group, 58% (15/26) in the 0.05%
BHT group, 5% (2/37) in the control semi-synthetic diet group, and 18%
(7/38) in the control lab chow group.

In a study in which C₃H mice were maintained on diets
containing 0.5% BHT for one month followed by lab chow for 10 months,
or control diet (BHT-free) for one month followed by lab chow for 10
months, the incidence of liver tumours in the two groups of male mice
were 9% (3/35) and 17% (5/29), respectively. Dietary BHT did not
result in an increased incidence of lung tumours in either male or
female mice.

In another study in which male BALB/c mice were maintained for
one year on a BHT-free diet, or diets containing 0.05% BHT or 0.5%
BHT, the incidence of liver tumours were 13% (4/30), 14% (6/43), and
7% (2/28) for the respective groups (Lindenschmidt et al., 1986).

It has previously been demonstrated that the incidence of
spontaneously-occurring hepatic tumours in C₃H mice is modified by
sex, population density, level of dietary protein, and caloric intake.
Historical dam were not available for a 10-month study. The incidence
of hepatic tumours in a 12-month study in this strain ranged from
6-13% for females, and 41-68% for males (Peraino et al., 1973).
Thus, the reported incidence of hepatocellular tumours was not
significantly different from other controls of similar age in studies
with the same inbred strain.

2.2.2.2 Rats

Feeding experiments conducted for 20 or 90 days indicated that
rats did not find food containing respectively 0.5 or 1% BHT
palatable. However, the animals ingested food so treated more readily
if these concentrations were attained gradually. Paired feeding
experiments with groups of 5 or 10 rats for 25 days demonstrated that
diets containing 0.8 and 1% BHT reduced the daily intake of food below
control values. A level of 1% in the diet retarded weight gain
(Deichmann, 1955).

BHT (0.3%) in the diet of pregnant rats that had been kept for 5
weeks on a diet deficient in vitamin E produced no toxic symptoms,
while 1.6% caused drastic loss of weight and fetal death (Ames
et al., 1956).

BHT was fed to rats (12/group) for 7 weeks at a level of 0.1% BHT
in diets containing 20% or 10% lard supplement. With the 20% lard
supplement diet, significant reduction of the initial growth rate and
mature weight of male rats was observed. No effect was noted in female
or male rats receiving the 10% lard supplement diet. A paired feeding
experiment showed that this inhibition of growth was a direct toxic
effect of BHT and could not be explained by a reduction in the
palatability of the diet. At this level BHT produced a significant
increase in the weight of the liver, both absolute and relative to
body weight. Rats under increased stress showed significant loss of
hair from the top of the head. The toxic effect of BHT was greater if
the fat load in the diet was increased. Anophthalmia occurred in 10%
of the litters (Brown et al., 1959).

BHT administered to rats at 250 mg/kg bw/day for 68 to 82 days
caused fatty infiltration of the liver and reduction in body-weight
gain (Karplyuk, 1959).

Groups of 6 weanling rats (3/sex) were fed diets containing a 20%
lard supplement and BHT at levels of 0, 0.1, 0.2, 0.3, 0.4 or 0.5% for
6 weeks. BHT reduced the growth rate, especially in males, the effect
becoming significant at 0.3% BHT. It also increased the absolute liver
weight and the ratio of liver weight to body weight in both sexes, the
latter effect becoming significant at 0.2% BHT. BHT increased the

ratio of left adrenal weight to body weight in male rats but had no
consistent effect in females. There were no histological changes in
the adrenal attributable to treatment. All dietary levels of BHT
increased the serum cholesterol and the concentration of cholesterol
was directly proportional to the BHT level. There was also a
significant increase in the concentration of adrenal cholesterol. BHT
produced no significant changes in the concentration of total or
percentage esterified liver cholesterol, total liver lipid or
concentration of total polyunsaturated fatty acids in the liver
(Johnson & Hewgill, 1961).

Rats fed diets supplemented with 20% lard, and containing 0, 0.2,
0.3, 0.4 or 0.5% (dry weight) BHT for 6 weeks, showed an increase in
serum cholesterol that was directly related to the level of dietary
BHT. BHT increased the relative weight of the male adrenal and also
caused a significantly greater decrease in growth rate of male as
compared to the female. Increased liver weight in test animals was
paralleled by increased absolute lipid content of the liver (Johnson &
Hewgill, 1961).

In another study, rats were maintained on diets containing 0.5%
dietary BHT in the presence or absence of a 20% lard supplement. BHT
increased the basic metabolic rate, the concentration of body
cholesterol and the rate of synthesis of body and liver cholesterol,
and reduced the total fatty acid content of the body, irrespective of
the presence or absence of dietary lard. In the animals fed BHT
without lard, BHT increased the rate of synthesis and turnover of body
and liver fatty acids and reduced the growth rate. These effect
occurred to a greater extent in animals fed BHT with lard (Johnson &
Holdsworth, 1968).

Groups of 8 young rats were fed diets containing 19.9% casein and
0, 0.02 or 0.2% BHT for 8 weeks. The experiment was repeated with
16.6% casein in the diet of further groups for 4 weeks and again with
9.6% casein (and no added choline) for 7 weeks. In all three instances
BHT caused stimulation of growth and improved protein efficiency. The
nitrogen content of the liver was, however, greatly reduced in
BHT-treated animals, except when the level of BHT was reduced to
0.02%. Recovery of hepatic protein after fasting (details not given)
was also impaired in rats on 0.2% BHT. Liver lipid content was
increased at 0.2%, but not at 0.02% BHT. A dietary level of 0.2% BHT
also increased the adrenal weight and ascorbic acid content, although
if recalculated on the basis of weight of gland, there was no
significant difference. The increase in adrenal ascorbic acid was
interpreted as indicating a stress imposed on the organism by BHT
(Sporn & Schöbesch, 1961).

Groups of 20 male and 20 female rats fed 1% BHT in the diet for
10 weeks showed recovery both in liver to body weight ratios and in
morphological appearance of the liver cells within a few weeks after
restoring the animals to a normal diet (Goater et al., 1964).

Groups of 48 weanling rats (24/sex) were given diets containing 0
or 0.1% BHT for periods of up to 16 weeks. Measurements of growth
rate, food consumption, weight and micropathological examination of
organs at autopsy revealed no difference between treated and control
groups. However, increase in relative liver weight and in the weight
of the adrenals was produced without histopathological evidence of
damage. Biochemical measurements and histochemical assessments of
liver glucose phosphatase and glucose 6-phosphate dehydrogenase
activities revealed no difference from the control group
(Gaunt et al., 1965a).

Groups of rats (16/sex) were fed diets containing 20% fat and BHT
levels of 0, 0, 0.03, 0.1 or 0.3% BHT for 10 weeks. No definite effect
on body weight was observed at any level in females and there was only
a slight depression in males at the 0.3% level. There was no
significant effect on blood cholesterol level in either sex after
feeding BHT at any of the levels. Four males at the 0.3% and two at
the 0.1% level died during the experiment. Two deaths occurred among
females at 0.3%. Only one male rat died in both control groups
(Frawley et al., 1965b).

In order to determine the MTD for BHT in rats, groups of F344
rats (5/sex) were given diets containing 0, 6200, 12 500, 25 000 or
50 000 mg BHT/kg diet for 7 weeks. Body weights were determined twice
weekly. Gross necropsy was performed on all animals in the study. All
of the male and female rats in the 50 000 mg/kg group died before the
end of the study. With the exception of one male in the 12 500 mg/kg
group, all of the animals in the other treatment groups and control
groups survived to the end of the study. Body weights at week 7 of the
study showed a dose-related decrement, with animals in the 25000 mg/kg
group weighing only 38% to 44% of control values. At dietary levels of
12 500 mg/kg, there was a slight increase in haematopoiesis in both
sexes (NCI, 1979).

2.2.2.3 Dogs

Mild to moderately severe diarrhoea was induced in a group of
4 dogs fed BHT at doses of 1.4-4.7 g/kg bw every 2 to 4 days for
4 weeks. Postmortem examination revealed no significant gross
pathological changes. No signs of intoxication and no gross or
histopathological changes were observed in dogs fed doses of
0.17-0.94 g/kg bw, 5 days/week for a 12-month period (Deichmann
et al., 1955).

2.2.2.4 Monkeys

Groups of 3 infant or juvenile monkeys (Macaca mulatta)
received BHT at doses of 0, 50 or 500 mg/kg bw/day for 4 weeks. Blood
analysis (complete cell count, serum sodium and potassium, bilirubin,
cholesterol and GOT) was carried out weekly, as was a complete

urinalysis. Liver biopsies were taken from the juvenile monkey at 2
weeks, following a 24-h fast. At the end of the test period, all
animals were fasted 24 h and sacrificed. Tissues from all major organs
were prepared for light and electron microscopy. Liver tissue was
also analyzed for protein, RNA and cytochrome P-450. Microsomal
preparations prepared from the livers were used to measure
nitroanisole demethylase and glucose-6-phosphatase activity. Urine and
blood values of test and control animals were similar. Histological
evaluation of all organs other than the liver from either infant or
juvenile monkeys did not indicate any compound-related changes. Test
animals receiving BHT showed hepatocytomegaly and enlargement of
hepatic cell nuclei. The hepatocytes of treated animals showed
moderate proliferation of the endoplasmic reticulum. Lipid droplets
were also prominent in cytoplasm of these hepatic cells. There was
fragmentation of the nucleolus in 15% of the hepatic cells in the test
animals in the high-dose group. DNA, RNA and cytochrome P-450 levels
in the liver of test and control animals were similar. BHT-treated
juveniles showed an increase in nitroanisole demethylase activity
which increased with time. The enzyme activity was unaffected in
infant monkeys. Glucose-6-phosphatase activity declined in juvenile
monkeys but was unchanged in infant monkeys (Allen & Engblom, 1972).

2.2.3 Long-term toxicity/carcinogenicity studies

2.2.3.1 Mice

Groups of 60 FAF male mice were maintained on semi-synthetic
diets containing 0, 0.25 or 0.5% BHT. The mean life-span of the test
animals was significantly greater than controls, being 17.0 ± 5.0 and
20.9 ± 4.7 months respectively for the 0.25% and 0.5% BHT, as compared
to 14.5 ± 4.6 months for controls (Harman, 1968).

A group of 18, 8-week old male BALB/C mice fed BHT at a level of
0.75% for a period of 12 months, developed marked hyperplasia of the
hepatic bile ducts with an associated subacute cholangitis (Clapp
et al., 1973).

Eleven mice (BALB/C strain) were maintained on a diet containing
0.75% BHT for a period of 16 months. The incidence of lung tumours in
the test group was 63.6%, compared with 24% in controls (Clapp
et al., 1974). However, a repeat of this study using a larger group
of test animals, showed that BHT had no effect on the incidence of
lung tumours in either sex (Clapp et al., 1975).

Groups of 48 mice (CFI strain) equally divided by sex were
maintained on diets containing 1000 mg/kg BHT. At week 4, one group
was then fed a diet containing 2500 mg/kg BHT, and at week 8, another
group was fed a diet containing 5000 mg/kg BHT. The animals were

maintained on these diets until 100 weeks of age. There was no
statistically significant reduction in survival of animals on the BHT
diet, although survival was poorer in males at the high-dose level
during the last quarter of the study. Animals dying or sacrificed
during the course of the study showed greater centrilobular cytomegaly
and karyomegaly than controls. Bile duct hyperplasia was only observed
in 3/141 test animals. There was no significant difference in the
incidence of malignant tumours in the high-dose group and control.
However, there was an increased incidence of lung neoplasia in treated
mice (75%, 74%, 53% and 47% in the 5000, 2500, 1000 mg/kg and control
groups, respectively). There were no morphological features to
distinguish the lung tumours in treated mice from those in controls.
There was also an apparent increase in benign ovarian tumours in
BHT-treated female mice, since none were observed in control animals
(Brooks et al., 1976).

BHT was administered in the diet at levels of 0, 3000, or
6000 mg/kg to groups of 20 (control) or 50 (treated) male and female
B6C3F₁ mice for 107108 weeks. The mice were observed twice daily for
signs of ill-health. Physical examinations were performed each month
and body weights were recorded at least once a month. Gross and
microscopic examination of 28 major organs, tissues and gross lesions
including the liver, thyroid and forestomach, was performed on all
animals at the end of the study and all animals dying on test where
possible. Peripheral blood smears were made for all animals where
possible. The body weights of treated male and female mice were lower
than the control mice throughout the study [numerical data not
provided]. The magnitude of the body-weight decrements was
dose-related. Administration of BHT in the diet resulted in similar or
improved survival in the treated groups compared with controls. At
termination, survival in male mice was 60%, 86% and 92% and in females
85%, 82% and 90%, for the control, low-and high-dose groups,
respectively. There was a marked dose-related increase in the
incidence of hepatocytomegaly and non-neoplastic lesions of the liver
(peliosis, hepatocellular degeneration/necrosis and cytoplasmic
vacuolation) in males but not in females. The incidence of
hepatocellular adenoma or carcinoma was not significantly increased in
either treated male or female mice, although there was a small
increase in the incidence of combined adenomas and carcinomas in the
treated females (1/20, 4/46 and 5/49 for the control, low- and
high-dose females, respectively) which was not statistically
significant. The historical control incidence for hepatocellular
neoplasms was not provided. The incidence of alveolar/bronchiolar
carcinomas or adenomas in female mice was significantly higher than
controls (5%) in the low dose (35%), but not the high dose (14%). The
historical control incidence was 4.7% for alveolar/bronchiolar
adenomas or carcinomas. Chronic ingestion of BHT in the diet was

related to a significant reduction in the incidence of sarcomas of
multiple organs in female mice. Four adenomas of the eye/lacrimal
gland were observed in high-dose males (8%) and in 2 low-dose females
but not in corresponding controls. The historical incidence of this
tumour in male mice was 1.2%. Since the lacrimal gland was only
examined microscopically in animals with grossly apparent lesions, the
report states that the lacrimal gland tumours could not be clearly
related to BHT administration (NCI, 1979).

Groups of B6C3F₁ mice (100/sex), were fed diets containing 0,
200, 1000 or 5000 mg/kg BHT for 96 weeks, followed by a basal diet for
8 weeks. At the end of the test period the surviving animals were
killed. A complete autopsy was carried out, and the principal organs
and tissues were examined microscopically. Mice that died during the
course of the study were also autopsied. In addition, terminal blood
samples were collected for haematological examination and serum
clinical biochemistry. Urine samples were also examined. During the
course of the study, food consumption was similar for test and control
groups. Body weights of females in the 1000 and 5000 mg/kg groups were
lower than controls, as was the body weight of males in the 5000 mg/kg
group. There were minor changes in the absolute weight of some organs
in the high-dose groups (salivary glands, heart and kidney). In males,
the serum GOT and GPT levels in the 5000 mg/kg group were higher than
controls. No other compound-related effects were observed in the
haematological, serum and urine analysis. Neoplastic lesions were
reported in both test and control animals. The tumours that occurred
with greatest frequency were adenomas of the lungs, hyperplastic
nodules and hepatocellular carcinomas of the liver and malignant
lymphomas. However, there was no statistically significant difference
between the BHT-treated and control groups for the incidence of any
type of tumour (Shirai et al., 1982).

BHT was administered in the diet at concentrations of 0, 1% or 2%
to groups of B6C3F₁ mice (50/sex) for 104 weeks. After the treatment
period, all the surviving mice were given basal diet for an additional
16 weeks. Treated animals underwent a 16-week recovery period prior to
pathological examination. Mean body weights of the treated male and
female mice were lower than those of controls. The body-weight
decrements were dose-related in both sexes and were more marked in
female mice. Treatment with BHT was found to improve survival in a
dose-related manner in both males and females. At the end of treatment
(104 weeks), the percent survival for male/female mice was 40%/58%
(control), 64%/81% (1% diet) and 74%/89% (2% diet). In male mice
administered BHT, there was a statistically significant increase in
the incidence of either a hepatocellular adenoma (19%, 38% and 53% in
control, low- and high-dose groups, respectively) or a focus of
hepatocellular alteration, showing a clear dose-response relationship.
No increases in the number of female mice with hepatocellular adenomas
or foci of altered hepatocytes were noted. The incidence of male mice

with other tumours and the incidence of female mice with any
tumour were not significantly increased as a consequence of BHT
administration. There was a dose-related tendency to lower incidence
of lymphoma and leukemia in both males and females (Inai et al.,
1988).

2.2.3.2 Rats

Groups of rats (15/sex) given diets containing 1% lard and
0.2, 0.5 or 0.8% BHT for 24 months showed no specific signs of
intoxication, and micropathological studies were negative. For the
group given a diet containing 0.5% BHT, the BHT was dissolved in lard
and then heated for 30 minutes at 150°C before incorporation in the
diet. There were no effects on weight gain or blood constituents and
micropathological studies of the main organs were negative. The
feeding of 0.8% BHT was followed in both male and female rats by a
subnormal weight gain and by an increase in the weight of the brain
and liver and some other organs in relation to body weight.
Micropathological studies were negative in this group also. BHT had no
specific effect on the number of erythrocytes and leucocytes, or on
the concentration of haemoglobin in the peripheral blood. A number of
rats of both sexes died during this experiment, but the fatalities
were not treatment-related. Micropathological studies supported this
observation. At 0.5%, BHT had no effect on the rat reproductive cycle,
the histology of the spleen, kidney, liver and skin, or on the weight
of the heart, spleen or kidney. There was no significant increase in
mortality of rats fed a diet containing 0.1% BHT and 10% hydrogenated
coconut oil for a period of two years (Deichmann et al., 1955).

Groups of JCL strain rats (20/sex/group), reared under a barrier
system and 4 weeks of age at the start of the study, received BHT at
0, 0.005, 0.062 or 0.32% in the diet. Of each group of 40, 15 received
compound for a "lifetime", 10 for 24 months, and 5 each for 3, 6 or 12
months. At the interim and final kills, liver, kidney, heart, spleen,
thyroid and caecum weight were determined as were haematology, serum
biochemistry, urinalysis, and histological investigation of the
tissues. At 24 months, heart, liver, kidney, spleen, pituitary,
thyroid, adrenal, testes, prostate and brain were weighed,
haematological and biochemical measurements conducted, and
histopathology done. There was an increase in liver weight, serum
cholesterol, serum K⁺ and histological changes in liver and kidney
at the 0.32% dietary level. There was no change in quantity of food
intake, body-weight gain, mortality during feeding or mean life span
and no finding suspicious of tumour induction. There was no indication
of a dose-related trend in tumour prevalence in either 24-month or
"lifetime" groups. The tumours found were said to be typical of those
described in aged rats. It is to be noted that the number of surviving

rats was small and that tumour data included both lifetime groups and
animals dead or sacrificed moribund during the 6, 12 and 24-month
feeding. The available data do not list the number of each of the
individual tissues examined, although the number of rats is listed.
The data show a tendency for a decreased number of tumours per rat at
higher BHT levels (Hiraga, 1978).

Groups of Fischer 344 rats (50/sex) received 3000 mg/kg or
6000 mg/ kg BHT in the diet for 105 weeks. The compound was mixed with
autoclaved lab meal containing 4% fat. A control group of 20
animals/sex received lab meal only. The animals were observed twice
daily for signs of ill-health. Physical examinations were performed
each month and body weights were recorded at least once a month. Gross
and microscopic examination of 28 major organs, tissues and gross
lesions including the liver, kidney, thyroid and forestomach, were
performed on all animals at the end of the study and all animals dying
on test where possible. Peripheral blood smears were made for all
animals where possible. There was a dose-related decrease in body
weights of treated male and female rats throughout the study. There
was no significant effect of BHT on mortality and no difference in the
incidence of various neoplasms between treated and control groups. The
incidence of adenomas of the pituitary was significantly reduced in
female rats with BHT administration. The incidence of focal alveolar
histiocytosis was elevated in treated male and female rats in a
dose-related manner compared with controls. The effect was more
pronounced in females than in males (NCI, 1979).

Groups of 7-week old Wistar rats (57/sex/group) were maintained
on diets containing 0.25 or 1% of BHT for 104 weeks. Control groups
consisted of 36 rats/sex. At the end of the test period, the surviving
animals were killed and a complete autopsy was carried out and the
weights of liver, spleen and kidneys were taken. The principal organs
and tissues were examined microscopically. Terminal blood samples were
collected for haematological examination and serum clinical
biochemistry. Survival in test groups was between 40% and 68%. A
significant increase in the mortality of the high-dose males was noted
after week 96 of the study. Food intake was similar for test and
control animals, but body-weight gain was significantly reduced in
high-dose males up to week 60 and in high-dose females for most of the
study. Increases in the mean absolute and relative liver weights were
observed in all treated animals, and decreases in the absolute and
relative spleen weights were observed in the treated females.
Dose-related changes in serum triglyceride (reduction) and GGT
(increase) in treated males and in total blood cholesterol (increase)
in treated females were noted. No significant morphological changes
were observed in the liver which were attributable to BHT treatment. A
variety of tumours were noted on histopathological examination at the
end of the study, with no dose-related response in either type of
tumour or total number as compared to controls. The incidence of
hyperplastic nodules in the liver and of pancreatic carcinomas in

female rats and of pituitary adenomas and adeno-carcinomas in both
sexes of test animals was higher than in controls. However, with the
exception of the incidence of pituitary adenomas in the low-dose
females, these differences were not significantly different from
controls. Since this effect was not dose-related, it was concluded
that BHT, under the conditions of this test. was not carcinogenic
(Shibata et al., 1979; Hirose et al., 1981).

Groups of 60, 40, 40, or 60 Wistar rats of each sex (F₀
generation) were fed BHT in the diet at doses of 0, 25, 100, or
500 mg/kg bw/day, respectively. The F₀ rats were mated after 13
weeks of dosing. The F₁ groups consisted of 100, 80, 80, and 100
F₁ rats, respectively, of each sex from the offspring from each
group. Because of an adverse effect on the kidney in the parents, the
concentration of BHT in the highest dose group was lowered to
250 mg/kg bw/day in the F₁ generation. The study was terminated when
rats in the F₁ generation were 144 weeks of age. Parameters studied
were food consumption (weekly), body weight, appearance, and
mortality. Autopsy and complete histopathological examinations were
performed on all animals dying during the study, or sacrificed
in extremis or at termination.

All animals consuming BHT experienced a dose-related increase in
survival. In both sexes differences (p <0.001) in longevity were
seen. The average body weights of the F₁ pups at birth in the mid-
and high-dose groups were slightly lower than those in the control
group. Body weights of all dosed animals from weaning through the
entire experiment were lower than those of control animals. In the
low-, mid-, and high-dose groups, the reductions in body weight were
for males/females 7%/5%, 11%/10%, and 21%/16%. Food intake was
comparable for all groups. Clinical appearance and behaviour were
reported to be normal for all animals. The high-dose males voided a
slightly reddish urine. Haematological parameters were reported to be
unchanged by BHT treatment, but no data were given. Serum triglycerides
were reduced in both sexes and cholesterol was somewhat elevated in
females only.

Histological studies indicated an increase in hepatocellular
carcinomas in male rats and an increase in hepatocellular adenomas in
both male and female rats. Most liver tumours were found during
terminal sacrifice at 141-144 weeks. One hepatocellular carcinoma was
found in a control male at 117 weeks and one in a high-dose male at
132 weeks. The remainder of the carcinomas occurred at terminal
sacrifice. The first adenoma was noted in a high-dose male at 115
weeks. Tables 3 and 4 summarize the data on mortality and the
appearance of adenomas and carcinomas of the liver. Data on mortality
and tumour incidence in different groups were analyzed using the
procedure of Peto et al. (1980). The dose-related increases in the
number of hepatocellular adenomas were statistically significant

(at p < 0.05) in male F₁ rats, when all groups were tested for
heterogenicity or analysis of trend. The increase in hepatocellular
adenomas and carcinomas in treated female F₁ rats was statistically
significant only for adenomas (at p <0.05) in the analysis for trend.
Reports on the spontaneous incidence of hepatocellular neoplasms in
Wistar rats from the laboratory performing this study, as well as
other European laboratories, indicated that it was usually less than
3% (Solleveld et al., 1984; Deerberg et al., 1980; Olsen et al.,
1984). The median life-span for animals in these studies ranged from
28-36 months for males and 28-33 months for females. Other sites
reported to have a slight but not statistically significant increase
in neoplastic lesions were as follows: thyroid, pancreas, ovary,
uterus, thymus, reticuloendothelium system, and mammary gland.

Non-neoplastic lesions occurred incidentally and showed no
relationship to BHT treatment, with the exception of lesions of the
liver, which showed a dose-related increased incidence of bile duct
proliferation and cysts in males, and focal cellular enlargement in
females. At the highest dose (250 mg/kg bw/day), there was no adverse
effect on the kidney (Olsen et al., 1986).

A long-term study was initiated in order to investigate the role
of hepatic changes in the development of hepatocellular carcinomas in
rats following in utero/lifetime exposure to BHT. The dosing regimen
of the study and strain of rat used were similar to those in the
two-generation study conducted by Olsen et al. (1986). Groups of
6 male and 48 female Wistar rats, aged 13 weeks and 9 weeks,
respectively, were fed BHT in the diet at doses of 0, 25, 100 or
500 mg/kg bw/day for 3 weeks prior to mating (13 weeks in Olsen
study). The rats were then mated on a 1:8 ratio for up to 21 days.
When pregnancy had been established by abdominal palpation, dams were
removed to individual cages. On day 20 of gestation, 5 pregnant rats
were sacrificed for assessment of body and liver weights and liver
histopathology. The pups were delivered by Caesarian section and
retained for assessment of a number of parameters. The remaining
females (20 dams in the control and high-dose groups; 24 dams in the
low- and mid-dose groups) were allowed to deliver normally. On day 6
post-partum, litters were either culled or augmented to comprise 8
pups, at the same time maximizing the number of robust males in each
litter. At weaning (3 weeks), 4 pups from each of 5 litters per group
were selected randomly, maximizing the numbers of males, for
assessment of a number of parameters. The male pups from the remaining
litters (approximately 60/group) were selected to continue in the
study and were placed in one of 4 groups corresponding to the diets
fed to their parents, with the exception that the high dose was
reduced to 250 mg/kg bw/day as in the Olsen study. Interim kills were
conducted at 1, 6, 11, or 16 months. The study was terminated 22
months after the F₁ male rats were placed on test diets. All animals
were observed daily throughout the study for clinical or behavioural
signs of toxicity. F₀ females were palpated from pregnancy to
weaning and F₁ males were palpated from 15 months to termination.

Body weights were recorded approximately every 2 weeks and food
consumption monitored every few days. Animals dying during the study
or at scheduled sacrifices were subjected to gross necropsy. At each
of the scheduled sacrifices the following parameters were measured in
all groups of F₁ fetuses and weanlings and F₁ male rats: body and
liver weights; hepatic enzyme activities for glucose-6-phosphatase,
epoxide hydrolase, glutathione-S-transferase, ethoxyresorufin
O-deethylase and pentoxyresorufin O-depentylase; hepatic content of
total cytochrome P-450, total glutathione, total microsomal and
cytosolic protein; and histopathological changes in the liver. In
addition, immunochemical staining of liver sections for cytochrome
P-450 1A and 2B and epoxide hydrolase was performed in the control and
high-dose groups at all scheduled sacrifices. Cellular proliferation
in the liver was measured in 5 animals per dose group at 4 weeks and
all subsequent sacrifices using pulse labelling techniques with
bromodeoxyuridine (only control and high-dose results were reported).
At the 11, 16 and 22-month sacrifices, a number of parameters in
addition to those listed above were monitored: histopathological
changes in the adrenal, kidney and thyroid; assessment of distribution
of hepatic glucose-6-phosphatase and gamma-glutamyl transferase by
histochemical staining; and serum thyroxine concentration (presented
for 16 and 22 months only).

In the first 5 weeks of BHT administration, a reduction in
body-weight gain was noted in the high-dose males. This trend started
prior to BHT administration and continued after BHT administration was
initiated. Body-weight gain in all other treatment groups was similar
to that in controls. No treatment-related effects on the health of the
animals were noted. During pregnancy and lactation, there was no
difference in food consumption between treated and control female
rats, and body weights of the dams were similar at weaning. At the
sacrifice on day 20 of gestation, both absolute and relative liver
weights of the dams were increased in a dose-related manner,
statistically significant at the high dose. The body weights of the
females, both including and excluding their litters, were similar in
all groups. Histopathological examination of the liver revealed mild
enlargement of centrilobular hepatocytes and eosinophilia in 4/5
high-dose (500 mg/kg bw/day) animals, and 1/5 low-dose (25 mg/kg
bw/day) animals, consistent with induction of mixed function oxidase
activity. A decrease in the mitotic index of hepatocytes from dams
receiving 100 and 500 mg/kg bw/day was noted; the significance of this
result was unclear.

Table 3. Mortality (and combined adenomas and carcinomas of the liver) in F₁ rats (Olsen et al., 1986)

BHT Effective
(mg/kg number Number of deaths during weeks^a
bw/day) of rats

0-90 91-104 105-113 114-118 119-126 127-132 133-140 141-144 Total

MALES
0 100 20 10 13 (1)/8 11 10 (1)/12 16 (2)/100
25 80 8 11 6 3 (1)/13 11 8 20 (1)/80
100 90 8 12 3 2 10 7 (2)/11 (4)/27 (6)/80
250 99 7 7 6 (1)/4 (2)/8 (2)/13 (1)/10 (20)/44 (26)/99

FEMALES
0 100 16 15 18 (2)/8 11 7 8 17 (2)/100
25 79 10 9 4 6 13 (1)/10 (1)/8 (1)/19 (3)/79
100 80 5 17 5 5 (1)/7 (1)/9 (1)/11 (3)/21 (6)/80
250 99 9 5 11 12 8 5 (2)/10 (12)/39 (14)/99

^a Figures in parenthesis are combined adenomas and carcinomas occurring during that period.
Table 4. Incidence of hepatocellular nodular hyperplasia, adenomas,
and carcinomas (Olsen et al., 1986)

BHT Effective No. Nodular
(mg/kg bw/day) of rats hyperplasia Adenomas Carcinomas

MALES
0 100 2 1 1
25 80 0 1 0
100 90 2 5 1
250 99 2 18^a 8^b

FEMALES
0 100 2 2 0
25 79 0 3 0
100 80 4 6 0
250 99 5 12^c 2^d

^a Over-all test for heterogeneity, p< 0.001, chi-square
18.17, 3 df.
Test lot trend, p< 0.001, chi-square = 17.97, 1 df.
^b Over-all test for heterogeneity, p < 0,05. chi-square
11.12, 3 df.
Test for trend, p< 0.01, chi-square = 9.40. 1 df.
^c Over-all test for heterogeneity, not significant, chi-square =
5.20, 3 df.
Test for trend, p< 0.05, chi-square = 4.99, 1 df
^d Over-all test for heterogeneity, not significant, chi-square =
2.87, 3 df.
Test for trend, not significant, chi-square = 2.59, 1 df.

Reproductive parameters were as follows: mating index was in the
range 54-65%; gestation index, 93-96%; viability index (birth to
day 6), 93-96%. No effect of treatment was evident in these
parameters. The absolute numbers of resorptions/dam were also similar
in treated and control groups. There was a slight decrease in the
numbers of pups/litter in the low- and high-dose groups, but a
dose-related trend was not evident. Body weights of the pups from the
high-dose group were significantly lower than controls at birth (10%),
and at days 6 (12%) and 21 (21%) of lactation. Mortality of the pups
remaining after culling, to day 21 of lactation was 2%, 8%, 12% and
15%, in order of ascending dose. Since the culling of pups at day 6
was non-random, these data could not be used for an unbiased
evaluation of reproductive function in BHT-treated animals.

Body weights of the F₁ males which continued in the study were
lower in the high-dose group, compared with controls, throughout the
22-month treatment period. During the first year of the study, the
difference was in the range of 10-20%. Lower body weights which
differed from controls by about 5% were also noted in the mid-dose
animals during the first half of the treatment period. No adverse
reactions to treatment or effects on food consumption were noted.
Subjective assessment of the coats of rats indicated that there was
less age-related deterioration in the 100 and 250 mg/kg bw/day groups
than in controls. At the scheduled sacrifices, consistent dose-related
increases in relative, but not absolute liver weights were observed,
which were statistically significant at the high dose.

A dose-related incidence of enlargement and eosinophilia of the
centrilobular hepatocytes was observed consistently at the scheduled
sacrifices, starting at 6 months. This was indicative of proliferation
of the SER, consistent with an induction of mixed function oxidases.
Immunohistochemical staining of liver sections from control and
high-dose rats revealed a marked increase in hepatocellular content
and distribution of cytochrome P-450 2B with BHT treatment which
persisted throughout the study. No such alteration was seen in the
staining patterns for cytochrome P-450 1A or epoxide hydrolase.
Histochemical staining revealed a marked induction of gamma-glutamyl
trans-peptidase activity in the periportal hepatocytes of nearly all
of the high-dose rats, starting at 11 months of treatment. This effect
was noted to a lesser extent in the mid-dose group. The distribution
of glucose-6-phosphatase activity in the treated rats was normal.
There was no evidence of treatment-related bile duct hyperplasia or
inflammatory cell infiltrate in the portal tract. Some rats with
altered hepatocellular foci (AHF) were noted in treated groups at 11
months and in all groups at 16 and 22 months. There was no treatment
relationship in the incidence of specific types of foci. At 22 months,
there was a higher incidence of eosinophilic and basophilic foci in
the high-dose group. Histochemical staining of liver sections revealed
a small number of high-dose animals with glucose-6-phosphatase-
deficient AHF which was statistically significant. No treatment-
related increase was noted in the incidence of AHF staining positive
for GGT At 22 months, there was also a significant increase in the
number of rats with hepatic nodules in the high-dose group (6/19
animals compared with none in the other groups).

No increases in the rate of hepatocellular proliferation were
detected as a result of BHT administration at any point in the study
commencing from 4 weeks post-weaning. It is of interest to note that
Clayson et al. (1993) observed an increase in hepatocellular
proliferation between 2 and 4 days after initiation of treatment of
male Wistar rats with 0.5% dietary BHT. Such a transient increase
would be difficult to detect with the widely spaced sampling times
used in this study.

A number of hepatic enzyme activities and other parameters
related to xenobiotic metabolism were altered as a result of BHT
treatment. Total cytochrome P-450 content was increased by 30-60% in
the high-dose animals starting at 21 days of age. Dose-related
increases were noted in epoxide hydrolase, glutathione-S-transferase
and pentoxyresorufin O-depentylase (PROD) activities, starting at 21
days of age, which were statistically significant in the mid- and
high-dose groups. The increases in PROD activity were very large,
10-25 fold in the mid-dose, and 20-80 fold in the high-dose groups.
Modest dose-related increases in ethoxyresomfin O-deethylase
activities were noted which were not statistically different from
controls. No effects on hepatic glutathione levels or glucose
6-phosphatase activity were noted throughout the study.

Histopathological examination of the kidneys revealed a reduction
in severity of chronic progressive nephropathy which affected the rats
in all groups from 11 months. No effects on the adrenal were noted,
although in a nearly identical, but invalidated study conducted in the
same laboratory (Robens Institute, 1989) cytomegaly of cells of the
zona fasciculata was observed in the mid- and high-dose groups at
weaning and at 4 weeks post-weaning, but not at subsequent time
points. In the present study, histopathology of the adrenal was
conducted starting at 11 months post-weaning. Evidence of thyroid
hyper-activity, characterized by reduction of follicular size, absence
or reduction of colloid, irregularities in the follicular outline,
hyperaemia and increase in the number of follicular cells was noted
starting at 11 months in both the mid-dose group (mild changes
affecting 75-82% of the rats) and the high-dose group (marked changes
affecting 100% of the rats). This effect was probably secondary to the
effects of BHT on the liver. Serum thyroxine levels in treated rats
did not differ from controls.

The demonstrated effects on hepatic enzyme induction and
consequent thyroid hyperactivity in the mid- and high-dose groups
together with the tumour data from the Olsen study suggested a NOEL of
25 mg/kg bw/day (Price, 1994).

BHT was added to the diet of male F344 rats for periods of up to
110 weeks. In the first experiment, a group of 36 male F344 rats
received basal NIH-07 diet only, while groups of 21 rats received
diets containing 300, 1000, 3000 or 6000 mg/kg BHT. Four animals were
randomly selected from each group for measurement of altered
hepatocellular foci at 12, 36, 48 or 76 weeks.

The remaining animals were sacrificed at 76 weeks. In a second
experiment, groups of 27 male F344 rats were fed basal diet, or diets
containing 12 000 mg/kg BHT or 12 000 mg/kg BHA for 110 weeks, at
which time all surviving animals were sacrificed. In both experiments,
body weights were recorded every four weeks. Gross necropsy was
performed on all animals. The livers were weighed and submitted for
histopathology along with tumours and gross lesions from other organs.

All the rats in experiment 1 survived the entire treatment period
(76 weeks). In experiment 2, the survival of rats in control and
treated groups started to decline after 84 weeks, but no
treatment-related trend was evident. At the end of the treatment
periods, rats fed 3000, 6000 or 12000 mg/kg BHT had significantly
lower body weights compared with respective controls. The body-weight
decrement for these three groups was approximately 10%. Rats fed
6000 mg/kg BHT for 76 weeks had significantly increased absolute and
relative liver weights compared with controls. In rats fed 12000 mg/kg
BHT for 110 weeks, the absolute liver weights were significantly
decreased and relative liver weights were comparable to controls. The
density of iron storage deficient hepatocellular foci in the one
affected rat per group was slightly but not significantly increased in
BHT-treated animals at 48 and 76 weeks; the incidence and size of foci
were slightly decreased at 110 weeks in rats receiving 12000 mg/kg
BHT. No hepatocellular carcinomas were detected in any of the groups.
Hepatocellular adenomas were detected in all groups, including
controls, with no treatment-related trend in incidence. There was also
no treatment-related effect of BHT on the incidence of grossly
observable tumours in specific organs (Williams et al., 1990a).

2.2.4 Reproductive toxicity studies

2.2.4.1 Mice

Diets containing 0.1 or 0.5% BHT together with two dietary levels
of lard (10 or 20%) were given to mice. The 05% level of BHT produced
slight but significant reduction in mean pup weight and total litter
weight at 12 days of age. The 0.1% level of BHT had no such effect.
Out of 7754 mice born throughout the reproductive life span of the
mothers, none showed anophthalmia, although 12 out of the 144 mothers
were selected from an established anophthalmic strain (Johnson, 1965).

The chronic ingestion of 05% BHT by pregnant mice and their
offspring resulted in a variety of behavioural changes. Compared to
controls, BHA-treated pregnant mice showed increased exploration,
decreased sleeping, decreased self-grooming, slower learning, and a
decreased orientation reflex. BHT-treated offspring showed decreased
sleeping, increased social and isolation-induced aggression, and a
severe deficit in learning (Stokes & Scudder, 1974).

A 3-generation study was performed in mice for the evaluation of
reproductive, developmental, and behavioural effects. Groups of 10
Crj:CD-1 mice/sex received BHT in the diet at concentrations of 0,
0.015%, 0.045%, 0.135%, or 0.405% (equivalent to 0, 20, 70, 200 or
610 mg/kg bw/day) starting at 5 weeks of age. At 9 weeks of age, the
mice were mated on a 1:1 basis for a period of 5 days. The pups
resulting from these matings were weaned at 4 weeks of age and 10
mice/sex/group were randomly selected to continue in the study. Males
and females were mated at 9 weeks as in the previous generation.

Reproductive parameters measured for each of the F₁ and F₂ groups
were: number of litters and pups, litter size and weight, sex ratio,
pup weights on lactation days 0, 4, 7, 14, and 21, and survival to day
21. Neurobehavioural parameters measured at various times during
the lactation period were: surface righting, negative geotaxis,
cliff avoidance, swimming behaviour and olfactory orientation.
Administration of BHT in the diet to mice adversely affected only
one of the measured reproductive parameters. Body-weight gain was
consistently reduced from day 7 to day 21 of lactation in the F₁
high-dose pups, but no body weight differences were noted in the F₂
pups compared with controls. In the F₂ males, lower scores were
assigned to the treated groups for the 180° turn in the open field
trial but this effect was not apparent in the F₂ female pups.
Consequently, no toxicologically significant effects of treatment were
apparent with BHT at the dietary concentrations tested in this study
(Tanaka et al., 1993).

2.2.4.2 Rats

Weanling rats (16/sex) were fed a diet containing 20% lard and 0,
300, 1000 or 3000 mg/kg BHT and mated at 100 days of age (79 days on
test). Ten days after weaning of the first litter, the animals were
again mated to produce a second litter. The offspring (16 females and
8 males) were mated at 100 days of age. Numerous function and clinical
tests including serum cholesterols and lipids were performed on the
parents and the first filial generation up to 28 weeks and gross and
microscopical examination at 42 weeks. At the 3000 mg/kg dietary
level, 10-20% reduction in growth rate of parents and offspring was
observed. A 20% elevation of serum cholesterol levels was observed
after 28 weeks, but no cholesterol elevation after 10 weeks. A 10-20%
increase in relative liver weight was also observed upon killing after
42 weeks on diet. All other observations at 3000 mg/kg and all
observations at 1000 mg/kg and 300 mg/kg were comparable with control.
All criteria of reproduction were normal. No teratogenic effects were
detected (Frawley et al., 1965b).

Similar results to those obtained with the parental and first
filial generations were also obtained with the second filial
generation. Examination of two litters obtained from the latter at 100
days of age revealed no effects except a reduction of mean body weight
at the 3000 mg/kg level. The offspring were examined for litter size,
mean body weight, occurrence of stillbirth, survival rate and gross
and microscopic pathology (Kennedy et al., 1966).

Breeding pairs of Sprague-Dawley rats (200-220 g) received Purina
chow supplemented with 0.125%, 0.25% or 0.5% BHT ad libitum
beginning from the week before mating and continuing in females
through lactation and weaning of the pups. Growth rates and survival
were adversely affected. Pre-weaning pups born of mothers at the

highest dose level weighed significantly less than controls at ages 7,
14 and 21 days. The total number of pups dying on study, born of dams
receiving 0.25 or 0.5% BHT, was significantly higher than controls.
Behavioural tests were conducted, consisting of righting reflex,
pivoting, cliff avoidance, startle response, swimming, open field,
running, wheel activity, roto-rod, active avoidance, position
discrimination and passive avoidance. For pre-weaning testing, no
differences were noted at the low- or mid-dose groups. At the
high-dose level, there was significant increase in surface righting
time, delayed forelimb swimming development and a trend to less
activity in open field tests. In post-weaning tests, males in the
0.25% BHT group showed an effect on passive avoidance, with more
partial re-entries into compartments where shocked. For all other
tests, there were no statistical differences, suggesting that BHT had
no effect on basic motor coordination, active avoidance acquisition,
or extinction performance (Brunner et al., 1978).

Groups of 46 rats, 6-week old (Wistar outbred, SPF) were fed
diets containing 0, or 0.5 to 0.9% BHT so that the dietary intake of
BHT was equivalent to 500 mg/kg bw/day during the course of the study.
At week 19, the F₀ generation was mated. Twenty-four hours after
birth of F₁ rats, the size of the litters was reduced to 8, and
half of the litters were cross-fostered. The body weight of parents
and offspring and the developmental events of offspring were
monitored during the course of the study, as well as the reproductive
performance of the F₀ rats. Auditory and visual function and
locomotive coordination tests were carried out on tile F₁
generation. The F₁ animals were autopsied at day 25 of age, and a
histological examination made of the brains. Body weights and weight
gain of test animals were reduced when compared to controls, and this
persisted during gestation. The duration of pregnancy, average body
weight, and litter size were similar for test and control animals. The
average body weight and weight gain of the F₁ offspring were
significantly reduced in pups nursed by dosed mothers. Pups exposed
in utero to BHT also showed a relatively slower development than
controls when fostered with non-dosed mothers. Pups exposed to BHT
in utero and/or mothers milk showed alterations in the behavioural
patterns examined as well as higher incidence in average number of
dead cells in the brain (Meyer & Hansen, 1980).

Detailed comments were submitted by the Chemical Manufacturers
Association (CMA) (1983) on studies of the effect of BHT on
reproduction and teratogenicity. The major comments were concerned
with the studies of Brunner et al. (1978) and Vorhees et al.
(1981) previously reviewed by JECFA in 1980 (Annex 1, reference 54) as
well as the study by Meyer & Hansen (1980).

In the case of the Brunner et al. and Vorhees et al. studies,
it was concluded that the study showed normal pup survival and
development in pups raised by dams on diets containing 0.125% BHT.
Normal post-weaning development was observed in pups raised by dams
on diets containing 0.25% BHT, although increased post-weaning
mortality occurred in pups raised by dams on the 0.25% and 0.5% diet;
developmental delays occurred in pups in the 0.5% group. In the case
of the Meyer and Hansen (1980) study, developmental delays were seen
in rats raised by dams on diets containing 0.5% BHT. At the 0.25% and
0.5% level, the effect may be due either to toxic effects of BHT on
the dam, or direct toxicity during lactation. A number of questions
were also raised about the design of the Brunner and Vorhees study.
These are: (i) the pup selection, in which all litters of fewer than 8
live pups were discarded; and (ii) the excess mortality was reported
in terms of pup count rather than affected litters. The data from this
study have been audited by the U.S. FDA (1983). It was concluded that
the raw data support the authors' observations of increased mortality
in the mid-dose and high-dose BHT offspring. However, excess mortality
occurred in a limited number of litters. For example, in the 0.5%
group, of the 60 deaths reported in 19 litters, 49 of the deaths
occurred in 5 litters; in the 0.25% group, of the 42 deaths, 21
occurred in 2 litters, and at the 0.125% level, of the 12 deaths, 11
occurred in 1 litter. It was also noted that in the high dose-group,
there was an increased number of litters with 8 pups or less, and no
litters larger than 12 pups. In the other dose groups, the litter size
was comparable to controls.

In the case of the Meyer & Hansen (1980) study, the CMA comments
noted that the level of BHT used in the study caused toxicity in the
dams, which appeared to affect the pups directly or indirectly.
Reports of teratogenicity studies and/or 1-generation reproductive
toxicity studies in several strains of mice and rats as well as a
3-generation reproductive toxicity study in rats were also submitted
in the comments to support a "no effect" level of 0.1% BHT in the
diet.

Groups of 60, 40, 40, or 60 Wistar rats of each sex, 7 weeks of
age, were fed a semi-synthetic diet containing BHT so that the dietary
intake was equivalent to 0, 25, 100, or 500 mg/kg bw/day, respectively.
After 13 weeks on the test diet the rats were mated.

Food consumption was similar in all groups. Male and female rats
in the high-dose group showed a significant decrease in body weight
which persisted throughout the study. Gestation rate was similar for
all test groups. The litter size and number of males per litter were
significantly lower in the 500 mg/kg bw/day group than in the
controls. Viability was similar in test groups and in the control

group during the lactation period. The average birth weight of the
pups in the 500 and 100 mg/kg bw/day groups was slightly lower than in
the controls. During the lactation period, BHT caused a significantly
lower dose-related body-weight gain (5%, 7%, and 41% lower body weight
for the 25, 100, and 500 mg/kg bw/day groups, respectively, as
compared to controls) (Olsen et al., 1986).

A study was initiated to determine the maximum dietary dose of
BHT tolerated by female rats exposed prior to and through pregnancy,
and by pups similarly exposed in utero and until weaning. Groups of
3 male and 16 female Wistar rats were administered BHT in the diet
corresponding to 0, 500, 750, or 1000 mg/kg bw/day for 3 weeks before
mating. At least 8 females per group were dosed during the pregnancy,
and until weaning (21 days after the delivery). After mating, the
males and the remaining females were autopsied. No effect of treatment
was seen on blood-clotting times in these animals. Food consumption of
treated females was considerably higher than controls from the fourth
week of the study onwards. No significant effect was seen on body
weight although a dose-related trend to reduction was apparent. No
effects were seen on general health except for fur discoloration in
treated animals.

Successful mating occurred less frequently in rats pretreated
with 1000 mg/kg bw/day of BHT than in the other groups. No major
differences were observed between the groups of pregnant females. The
weight gain in rats treated with the two highest doses appeared to be
inhibited in the last week of the pregnancy. There was no significant
difference between litter number or litter weight between pups born of
control or treated animals, although a dose-related trend towards
reduction in litter size was seen. No evidence of teratogenic effects
of BHT was provided.

Litter sizes were standardized to 8 pups if possible. At weaning,
the dams treated with 1000 mg/kg bw/day of BHT had lower body weights
and very little body fat was observed at autopsy. Pups from the dams
treated with the lowest BHT dose were markedly stunted in their
growth, but appeared healthy. Pups from dams treated with the two
highest doses were severely stunted, showed poor fur condition, and
were less active. It was noted that in BHT-treated animals, where the
litter size was less than 8, the average pup weight was generally
considerably greater. This implies that the reduced weight gains in
litters of normal size were associated with poor milk production
rather than BHT toxicity. Pups from two litters from each dose group
were maintained on control diet for 4 weeks after weaning. Pups born
to dams receiving BHT-containing diets remained of lower body weight
than control pups. Pups from the two highest dose groups continued to
show poor condition. Treatment with BHT caused a marked increase in
liver weight in all dams. The liver weights were almost 10% of the
body weights, the maximum degree of enlargement possible in rats. The
relative liver weights of pups from BHT treated dams were not
different from controls (Robens Institute, 1989).

2.2.4.3 Chickens

When BHT was fed at a level of 0.125% for 34 weeks to a group of
10 pullets, no differences in fertility, hatchability of eggs or
health of chicks in comparison with a similar control group were
found. The eggs of the antioxidant-treated birds contained more
carotenoids and vitamin A than those of the controls (Shellenberger
et al., 1957).

2.2.4.4 Monkeys

A group of 6 adult female rhesus monkeys were maintained on a
test diet containing a mixture of BHT and BHA that provided an intake
equivalent to 50 mg BHT/kg bw/day and 50 mg BHA/kg bw/day. Another
group of 6 adult female rhesus monkeys were used as controls. The
monkeys were fed the diet for one year prior to breeding and then for
an additional year, including a 165-day gestation period. Haematologic
studies including haemoglobin, haematocrit, total and differential
WBC, cholesterol, Na⁺, K⁺, total protein, serum GPT, and GOT, were
carried out at monthly intervals. Body weights were taken at monthly
intervals. Records of menstrual cycles were maintained through the
test period.

After one year the females were bred to rhesus males not
receiving test diets. During pregnancy complete blood counts were done
on days 40, 80, 120 and 160 of gestation and on days 30 and 60
post-partum. A total of 5 infants were born to the experiment monkeys
and 6 to the control monkeys. Haematological evaluations were made on
infants of the test and control monkeys at days 1, 5, 15, 30 and 60,
and observations of the infants were continued through two years of
age. Two experimental and 2 control infants, 3 months of age, were
removed from their mothers for 1 month of psychological home cage
observations.

No clinical abnormalities were observed in parent or offspring
during the period of study. The gestation of test animals was free of
complications and normal infants were delivered. Adult females
continued to have normal infants. Infants born during the exposure
period remained healthy, with the exception of one infant that died
from unrelated causes. Home cage observations at the third month of
life did not reveal any behavioural abnormalities (Allen, 1976).

2.2.5 Special studies on teratogenicity

In a study on the embryotoxicity of BHT, 3 dosing schedules were
employed: single doses (1000 mg/kg bw) on a specific day of gestation,
repeated daily doses (750 mg/kg bw) from the time of mating throughout
pregnancy, and daily doses (250-500 mg/kg bw for mice and 500 and
700 mg/kg bw for rats) during a 7- to 10-week period before mating,

continuing through mating and gestation up to the time the animals
were killed. No significant embryotoxic effects were observed on
examination of the skeletal and soft tissues of the fully developed
fetuses as well as by other criteria. Reproduction and postnatal
development were also unaffected (Clegg, 1965).

2.2.6 Special studies on genotoxicity

The results of genotoxicity studies with BHT are summarized in
Table 5.

At a concentration as low as 10 µg/ml (optimal 50-100/µg/ml), BHT
exerted a strong inhibitory effect on cell-to-cell dye transfer
(lucifer yellow transfer) in cultures of SV-40-transformed Djungarian
hamster fibroblasts. The effect was reversible. BHT shared this effect
with a series of well known tumour promoters (Budunova et al.,
1989).

An extensive database of genotoxicity studies for BHT, including
those documented above, were reviewed in a paper by Bombard et al.
(1992). The authors concluded that the majority of evidence indicated
a lack of potential for BHT to induce point mutations, chromosomal
aberrations, or to interact with or damage DNA, and that BHT does not
represent a genotoxic risk to humans.

The ability of BHT and its metabolites to induce cleavage of
supercoiled plasmid DNA was studied in pUC18 by agarose gel
electrophoresis. Dose-related cleavage of DNA was noted with
BHT-quinone in the range 10⁶-10^-3M, and to a lesser extent using
higher doses of BHT-aldehyde and BHT-peroxyquinone. BHT, BHT- quinone
methide and other BHT metabolites had no effect on the plasmid DNA.
Free radical scavengers were effective against the effects of
BHT-quinone and -peroxyquinone, but not against BHT-aldehyde.
Formation of superoxide radical as measured by a reduction in
cytochrome c, was noted only with BHT-quinone (Nagai et al., 1993).

The addition of BHT (5 to 20 µg/plate) caused a two-fold increase
in the mutagenic potency of aflatoxin B₁ using Salmonella
typhimurium strains TA98 and TA100, with and without activation
(Shelef & Chin, 1980).

The addition of 50-250 µg of BHT/plate inhibited 3,2'-dimethyl-
4-aminobiphenyl-induced mutagenicity in Salmonella typhimurium
strains TA98 and TA100 in the presence of rat liver S-9 fraction
(Reddy et al., 1983a).

Table 5. Results of genotoxicity assays with BHT

Test System Test Object Concentration of Results Reference
BHT

Point Mutation

Ames test¹ S. typhimurium 0.015-0.6% Negative Brusick 1975
TA1535, TA1537
TA1538

Ames test¹ S. typhimurium Negative Hageman et al.
TA97, TA102 1988
TA104, TA100

Ames test¹ S. typhimurium 100-10 000 Negative Williams et al.
TA98, TA100 µg/plate 1990b
TA1535, TA1537
TA1538

Ames test¹ S. typhimurium 100-1000 µg/plate Negative Yoshida 1990
TA98, TA100

Ames test¹ S. typhimurium 10 µg/plate Negative Detringer et al.
TA 98 1993

Host-mediated assay ICR Swiss mouse/ 30-1400 mg/kg - Negative SRI 1972
S. typhimurium acute
G46, TA1530 30-500 mg/kg -
subacute

Mammalian cell rat liver epithelial 60-90 µg/ml Negative Williams et al.
gene mutation cell (line 18), 1990b
HGPRT locus

Table 5 (cont'd)

Test System Test Object Concentration of Results Reference
BHT

Sex-linked recessive Drosophila 2.0 × 10^-6 µg Negative² Prasad &
lethal melanogaster Kamra 1974

Sex-linked Drosophila 5% diet Negative Brusick 1975
recessive lethal melanogaster

Clastogenic effects and chromosomal aberrations

Chromosomal human WI-38 2.5-250 µg/ml Positive² SRI 1972
aberration assay (embryonic lung
cells)

Sex chromosome loss Drosophila 2.0 × 10^-6 Negative² Prasad &
melanogaster Kamra 1974

Micronucleus assay rat bone marrow 30, 90, 1400 Negative SRI 1972
mg/kg (acute and
subacute)

Dominant lethal assay Sprague-Dawley rat 30, 900, 1400 Negative Brusick 1975
mg/kg bw (acute)
30, 250, 500 Positive
mg/kg bw/day
(subacute)

Dominant lethal assay male Sprague-Dawley 50, 150, 500 Positive² Sheu et al.
rats mg/kg bw/day 1986
1% diet Negative
male mice

Table 5 (cont'd)

Test System Test Object Concentration of Results Reference
BHT

Heritable male mice 1% diet Negative Sheu et al.
translocation assay 1986

DNA interactions

Mitotic recombination Saccharomyces 0.6-2.4% Negative Brusick 1975
cerevisiae D4

Mitotic recombination Saccharomyces 30, 900, 1400 Negative SRI 1972
- Host-mediated cerevisiae D3/ICR mg/kg bw (acute
Swiss mouse 30, 250, 500 Negative
mg/kg bw/day
(subacute)

Sister chromatid CHO cells 1 - 1000 µg/ml Negative Williams et al.
exchange¹ 1984

DNA excision repair UV-irradiated human ? Positive Daugherty
synthesis lymphocytes 1978

DNA repair test hepatocyte primary 0.01 - 10 µg/ml Negative Williams et al.
culture 1990b

¹ Both with and without rat liver S9 metabolic activation.
² A discussion of this result is contained in Bombard et al. (1992).

The addition of 100-250/µg of BHT/plate was shown to inhibit
3,2'-dimethyl-4-aminobiphenyl-induced mutagenicity in Salmonella
typhimurium strains TA98 and TA100 Mutagenicity was further
inhibited by use of S-9 preparations from rats fed dietary BHT (0.6%)
as compared to S-9 preparation from rats fed BHT-free diets
(Reddy et al., 1983b).

BHT was a moderately effective inhibitor of benzidine-induced
mutagenicity in Salmonella typhimurium strain TA98, activated with a
hamster liver S-9 fraction (Josephy et al., 1985).

BHT (0.11-11 /µM) protected against DNA damage induced in rat
hepatocytes by AAF or N-hydroxy AAF as shown by a marked reduction of
unscheduled DNA synthesis. BHT also inhibited AAF-induced DNA damage
in human hepatocytes. In addition, rats pre-treated with 0.5% BHT in
the diet for 10 days provided hepatocytes which exhibited less
unscheduled DNA synthesis than did hepatocytes from control rats when
these cells were exposed to either AAF or N-hydroxy AAF (Chipman &
Davies, 1988).

2.2.7 Special studies on hepatotoxicity

2.2.7.1 Mice

Groups of male ddY mice treated perorally with BHT (200-800 mg/kg
bw) in combination with an inhibitor of GSH synthesis, buthionine
sulfoximine (BSO, 1 h before and 2 h after BHT, 4 mmol/kg bw/dose,
i.p.) developed hepatotoxicity characterized by an increase in SGPT
activity and centrilobular necrosis of hepatocytes. The hepatotoxic
response was both time- and dose-dependent. BHT (up to 800 mg/kg bw)
alone produced no evidence of liver injury. Drug metabolism inhibitors
such as SKF-525A, piperonyl butoxide, and carbon disulfide prevented
the hepatotoxic effect of BHT given in combination with BSO while
inducers of drug metabolism such as phenobarbital tended to increase
hepatic injury. The results suggested that BHT was activated by a
cytochrome-P-450-dependent metabolic reaction and that the hepatotoxic
effect was caused by inadequate rates of detoxification of the
reactive metabolite in mice depleted of hepatic GSH by BSO
administration. Based on studies with structural BHT analogues, the
authors suggested that a BHT-quinone methide may play a role in the
hepatotoxicity in mice (Mizutani et al., 1987).

2.2.7.2 Rats

Female (albino Wistar) rats, initial body weight 120-130 g, were
maintained on a diet containing 0 or 0.4% BHT, for 80 weeks. After one
week on the test diet, significant increases were observed in liver
weight, microsomal protein, cytochrome P-450, cytochrome b5,
NADPH-cytochrome c reductase, biphenyl-4-hydroxylase and ethyl
morphine N-demethylase but not aniline-4-hydroxylase. Total liver

protein, succinic dehydrogenase and glucose-6-phosphatase were
slightly decreased. There was little change in this pattern during the
period of the study. Rats removed from the BHT test diet at the end of
the test period and maintained on BHT free diet for 18 days, showed a
return to normal for many liver parameters except for cytochrome b5,
cytochrome c reductase and ethylmorphine N-demethylase. Histological
changes at the end of 80 weeks feeding consisted of centrilobular cell
enlargement, which was reversible following 18 days on a BHT-free
diet. The only ultrastructural change was a prolileration of smooth
endoplasmic reticulum (Gray et al., 1972).

Groups of female rats (80-100 g) received 0.4% BHT with corn oil
mixed in ground lab chow and were sacrificed at intervals of 1, 8, 16,
32, or 80 weeks, and compared with controls. Samples of liver were
taken for biochemical, histochemical, and morphological studies. To
examine for reversibility of hepatic changes, control diet was
administered for 18 days to a group of 4 rats following 80 weeks of
BHT administration. After one week on BHT, there was a marked liver
enlargement with relative liver weight increased up to 35% and with an
increase in drug metabolizing activities and NADP cytochrome c
reductase activity. After 18 days of removal from the 80-week
treatment, there was only a slight increase in liver weight. The
effect was therefore reversible. Histologically, after BHT treatment
the liver was characterized by enlarged centrilobular hepatocytes,
with a heterogenous appearance of this zone. Ultrastructurally, there
was a proliferation of smooth endoplasmic reticulum. The authors noted
that although the evidence of liver injury was equivocal, there were
two features that were also seen with many hepatotoxins and
hepatocarcinogens: depression of glucose-6-phosphatase activity and
cell enlargement. However, there were no lysosomal changes
characteristic of cytologic injury, and effects were reversible
(Crampton et al., 1977).

Groups of 8 male Wistar rats were given diets containing 0, 0.1,
0.25, 0.5, or 0.75% BHT for 30 days. BHT did not induce cellular
proliferation in the liver, urinary bladder or thyroid after 30 days
as measured by the [³H]thymidine labelling index or mitotic index.
In a second experiment, groups of 8 rats were treated with 0.5%
dietary BHT for 2, 4, 8, 10, or 14 days. This treatment led to a
time-limited increase in liver cell [³H]thymidine-labelling index
that subsided to control values within 8 days. This increase in
[³H]thymidine labelling in the liver was accompanied by an
une