Safrole has not previously been evaluated by the Joint FAO/WHO
    Expert Committee on Food Additives.

         Safrole (1,2-methylenedioxy-4-allylbenzene) is the principal
    constituent of oil of sassafras and a minor constituent of many other
    essential oils. The related substance, isosafrole (1,2-methylene-
    dioxy-4-propenylbenzene), also occurs as a minor constituent of many
    essential oils with a distribution similar to that of safrole. Another
    related substance, dihydrosafrole (1,2-methylenedioxybenzene-4-
    propylbenzene), is not known to occur naturally but is formed in the
    production of piperonyl butoxide (International Agency for Research on
    Cancer, 1976).



    Absorption, distribution and excretion

         Safrole was absorbed from the gastrointestinal tract by passive
    diffusion, with the absorption kinetics apparently dependent on its
    lipid solubility as determined in an in situ perfusion method in the
    rat (Fritsch et al., 1975a). In this same procedure, safrole, at a
    level of 2 mg/ml of perfusion medium, reduced the absorption of
    glucose and methionine but not butyric acid (Fritsch et al., 1975b).


         Basic ninhydrin-positive substances were excreted in the urine of
    male rats treated with safrole or isosafrole in doses of 75-300 mg/kg
    i.p. These substances were not seen when dihydrosafrole was
    administered in similar doses. These substances readily decomposed to
    carbonyl-containing compounds. The substances were not identified in
    this study (Oswald et al., 1969), but in a later study (Oswald et al.,
    1971) the safrole metabolites were identified as tertiary amino
    propiophenones, 3-N,N-dimethylamino-1-(3',4'-methylenedioxyphenyl)-1-
    propanone (I), 3-piperidyl-1-(3',4'-methylenedioxyphenyl)-1-propanone
    (II), 3-pyrrolidinyl-1-(3',4'-methylenedioxyphenyl)-1-propanone (III).
    I and III are excreted by the rat and II by both the rat and
    guinea-pig. These safrole metabolites are Mannich bases which are
    believed to be formed by oxidation of the allyl group to yield a vinyl
    ketone which condenses with an available amine (McKinney et al.,
    1972). I, II and III were competitive inhibitors of rat liver
    mitochondrial monoamine oxidase with benzylamine-HCl as a substrate.
    Metabolite III inhibited rat liver, kidney, and brain monoamine
    oxidase with tyramine HCl or serotonin as substrates (Bangdiwala &
    Oswald, 1976).

         Thin-layer chromatography of rat urine and bile following
    intravenous injection of safrole, dihydrosafrole and isosafrole showed
    the metabolites were largely eliminated in the bile after this route
    of administration. Metabolites were not identified (Fishbein et al.,

         C14O2 was excreted by mice and flies treated with safrole or
    dihydrosafrole labelled on the methylene C. Microsome preparations
    from the liver of treated mice or abdomens of treated flies yielded
    formate-C14, indicating a cleavage of the methylenedioxy ring (Casida
    at al., 1966).

         The major urinary metabolites of safrole administration i.p.
    in the rat or guinea-pig were: 1,2-dihydroxyl-4-allylbenzene,
    1'-hydroxysafrole (1,2-methylenedioxy-4-(1-hydroxyallyl)
    benzene) (HOS), 1,2-methylenedioxy-4-(2,3-dihyroxypropyl)
    benzene, 1,2-dihydroxy-4-(2,3-dihydroxypropyl)benzene,
    2-hydroxy-3-(3,4-methylenedioxyphenyl) propanoic acid, and
    3,4-methylenedioxybenzoylglycine. Safrole oxide, administered i.p.,
    produced the same metabolites as safrole, with the exception of the
    first two compounds and the benzoylglycine. This indicates that the
    metabolites containing a propyl side chain are probably formed by
    conversion of the allyl side chain with safrole epoxide as an
    intermediate. Some unchanged safrole epoxide was also found after
    administration of this compound. A small amount of 1,2-methylenedioxy-
    4-(1,2,3-trihydroxypropyl)benzene was found in rats only (Stillwell et
    al., 1974).

         Small doses of [C14]-safrole were absorbed rapidly and excreted
    almost completely via the urine in 24 hours in both man (doses
    0.165 mg or 1.655 mg) and the rat (dose 0.63 mg/kg). A large dose
    (750 mg/kg) in the rat resulted in a decrease in the rate of
    elimination, only 25% being excreted in 24 hours, and plasma and
    tissue levels of safrole and its metabolites were elevated for 48
    hours. 1,2-dihydroxy-4-allylbenzene was the main urinary metabolite in
    both species. The metabolites HOS and 3'-hydroxy-isosafrole were
    detected in the urine of the rat but not in man (Benedetti et al.,

         Rats, mice, hamsters and guinea-pigs excreted a conjugated form
    of HOS in the urine after treatment with safrole i.p. The HOS
    accounted for 30% or more of the safrole dose in male mice and 1-3% of
    the safrole dose in the other species. Pretreatment with phenobarbital
    or 3-methylcholanthrene (3-MC) increased HOS excretion tenfold in
    rats; phenobarbital had no influence on HOS excretion in hamsters or
    guinea-pigs. Bile-duct ligation had little effect on the excretion of
    HOS. The conjugated HOS was cleaved by commercial beta-glucuronidase
    preparations. A small amount of 3'-hydroxy-isosafrole was found in the
    urine; it was believed to arise from rearrangement of HOS. When HOS
    per se was administered to rats, about 40% was excreted as HOS;

    pretreatment with phenobarbital or 3-MC did not markedly alter the
    amount excreted. The percentage of safrole or HOS excreted as HOS
    after p.o. administration was similar to that observed after i.p.
    administration. When safrole was administered in the diet, excretion
    of HOS was 5-10% of the daily dose during the first 18 days and 3-4%
    thereafter. Concurrent administration of 0.1% sodium phenobarbital
    resulted in an average of 7% of the safrole excreted as HOS over an
    11-week period. Forty per cent. of an oral HOS dose was excreted as
    HOS over a seven-month period (Borchert et al., 1973b).

         HOS was not detected in sassafras (Sassafras albidum) oil
    (Sethi et al., 1976).

         Male mice were treated with 185 µmol safrole/100 g bw. Adult mice
    excreted 46% of the dose as HOS, while 21-day-olds excreted only 12%
    as HOS (Drinkwater et al., 1976).

         Administration of tritiated HOS to rats and mice resulted in
    labelled liver macromolecules with specific activity generally in the
    order rRNA=protein, DNA; the nucleosides were not the labelled
    compounds. Hepatic macromolecules isolated from adult female rats had
    55-65% of the specific activities of those obtained from male rats.
    The specific activities of liver macromolecules isolated from adult
    female mice were 10-20 times greater than those from pre-weanling
    (first three weeks of age) male and female mice. HOS was metabolized
    by rat and mouse liver cytosols in a 3'-phospho-adenosine
    5'-phosphosulfate-dependent reaction to an electrophilic reactant
    presumed to be the sulfuric acid ester. HOS was also oxidized by rat
    and mouse liver microsomes to 1'-hydroxysafrole-2',3'-oxide in a
    reduced nicotinamide adenine dinucleotide phosphate-dependent
    reaction. Electrophilic reactivities of various possible safrole
    derivatives in vitro with nucleosides was determined to be in the
    order of 1'-oxosafrole, 1'-acetoxysafrole (ACOS), 1'-acetoxysafrole-
    2',3'-oxide, 1'-hydroxy-2',3'-oxide, safrole-2',3'-oxide,
    1'-oxosafrole-2',3'-oxide. Neither ACOS nor 1'-oxosafrole formation
    were detected in vitro, but ACOS reacted with guanosine-5'-
    monophosphate to yield a reaction product identified as
    0,6-(isosafrol,3'-ul) guanylic acid which in turn yields
    3'-hydroxyisosafrole. 1'-oxosafrole has not been isolated as a
    metabolite, but small amounts of condensation products of the ketone
    with amines appear in the urine of animals treated with safrole or HOS
    (Wislocki et al., 1976).

         HOS and 1'-hydroxyallylbenzene, metabolites of the hepatotoxic
    safrole and the non-hepatotoxic allylbenzene respectively, are
    metabolized in a similar manner. HOS rapidly rearranged under acid
    conditions to 3'-hydroxyisosafrole, whereas the rearrangement of
    1'-hydroxyallylbenzene to cinnamyl alcohol was slow. It was postulated

    that the methylenedioxy group stabilized a carbonium ion by a
    resonance effect and such resonance stability increases the
    electrophilicity of the compound and thus its tissue reactivity
    (Peele & Oswald, 1978).

         By the use of mass spectroscopy, the following metabolites
    of safrole were identified in rat liver epithelial cell cultures:
    eugenal, 2-allylcatechol, 4-(2',3'-epoxypropyl)catechol, 2',3'-
    epoxysafrole, 2',3'-dihydro-2',3'-dihydroxysafrole, HOS, and
    3'-hydroxyisosafrole. HOS, 1',2'-epoxyisosafrole, and 1',2'-dihydro-
    1',2'-dihydroxysafrole were identified as metabolites of isosafrole in
    this system. Cultures of liver cells from female rats produced greater
    quantities of the safrole metabolites than those from male rats
    (Janiaud et al., 1976; Delaforge et al., 1977). The epoxy-diol pathway
    for metabolizing safrole was also found in the adrenal of the rat
    (Doumas & Maume, 1977). With 2',3'-epoxysafrole as the parent compound
    the cell cultures produced 2',3'-epoxy(4-allyl)catechol (Delaforge et
    al., 1977). The following epoxy compounds were identified in the urine
    of rats treated with safrole: 1,2'-epoxyisosafrole, 2',3'-
    epoxyallylcatechol, 1',2'-epoxyallylphenol, 2',3'-epoxyallylphenol,
    and 1'-hydroxy-2',3'-epoxysafrole glucuronide (Delaforge et al., 1977;
    Levi et al., 1977).

    Effects on enzymes and other biochemical parameters

         Safrole, when combined with piperonyl butoxide and other
    methylenedioxyphenol derivatives, inhibited liver aminopyrine
    demethylase and biphenyl-4-hydroxylase activity in microsome
    preparations from mice pretreated with doses of the test compounds
    which were not individually inhibitive (Friedman et al., 1971).

         Safrole and isosafrole, administered to rats, produced liver
    hypertrophy with increases in hepatic biphenyl hydroxylase, nitro
    reductase, glucuronyl-transferase, and cytochrome P-450. Some of the
    actions of safrole were inhibited by actinomycin D and some were not,
    indicating the increased enzyme activity comes both from new enzyme
    protein synthesis and changes in existing enzyme. Addition of safrole
    in vitro to a rat liver microsomal preparation did not increase
    enzyme activities (Parke & Rahman, 1970). Administration of safrole
    and isosafrole also gave rise to a new redox difference absorption
    spectra in liver preparations which was attributed to induction of a
    new hepatic microsomal haemoprotein (Parke & Rahman, 1971). Aryl
    hydrocarbon hydroxylase was induced in liver, lung, kidney, and gut in
    rats fed a diet containing 0.25% isosafrole. There were also increases
    in hepatic biphenyl 2-hydroxylase and 4-hydroxylase activities,
    microsomal protein, liver weight, cytochrome P-450 content, and
    apparent cytochrome b5 content, and a decrease in hepatic aniline
    4'-hydroxylase activity (Lake & Parke, 1972).

         Urinary excretion of 3- and 5-hydroxy derivatives of
    2-acetoamidofluorene (2-AAF) was increased by pretreatment of young
    and adult male rats with three doses of safrole (100 mg/kg per day)
    but not by similar treatment with isosafrole. Safrole also caused an
    increased excretion of the N-hydroxy derivative in adult rats. A
    single dose of 3-MC caused greater increases than safrole. Assay of
    hydroxylated derivatives of 2-AAF formed by liver microsomal
    preparations from rats pretreated as described showed increases in
    amount of ring-hydroxylated derivatives in the order of 3-MC,
    isosafrole, safrole and in the amount of N-hydroxylated derivative in
    the order 3-MC, safrole, isosafrole. Administration of ethionine
    largely inhibited the increased hydroxylation from safrole and
    isosafrole and methionine reversed the inhibition, indicating that
    safrole or isosafrole pretreatment produces new synthesis of liver
    microsomal hydroxylases. Pretreatment of hamsters with safrole had
    little effect on hydroxylation by liver preparations of this species,
    while pretreatment with isosafrole tended to inhibit both N- and some
    ring-hydroxylation (Lotlikar & Wasserman, 1972).

         Safrole, administered at a dietary level of 2% for two weeks to
    rats, induced a P-450 cytochrome, similar to that seen with 3-MC, and
    caused the formation of a stable safrole-metabolite-cytochrome P-450
    complex. The same response was obtained when safrole was incubated
    with control microsomes, NADPH and oxygen. The safrole metabolite
    could be displaced from the complex in vitro by safrole, biphenyl
    and ethylbenzene (Elcombe et al., 1975).

         Safrole, at a level of 0.67 mM in the medium, inhibited protein
    synthesis in isolated adult rat hepatocytes by 75% (Gwynn et al.,
    1979). Safrole did not induce unscheduled DNA synthesis in HeLa cells
    (Martin et al., 1978). Safrole (640 mg/kg i.p.) inhibited mouse
    testicular DNA synthesis, as measured by uptake of tritium-labelled
    thymidine, by 60% (Friedman & Staub, 1976). Adult rat liver epithelial
    cells, maintained in culture for six months, and dosed with safrole
    (0.5 mg in ethanol) four times in eight days, showed high toxicity
    followed by many mitoses. This resulted in a mixed type of cell in the
    culture. Some of the cells maintained their epithelial structure, but
    most appeared transformed (Janiaud & Padieu, 1977).

         The induction of resistant hepatocytes in vivo in rats was
    investigated as a short-term test for carcinogens. Safrole was
    positive in this test but required three doses (150 mg/kg per dose),
    while other carcinogens such as 2-AAF and N-butyl-N-nitrosourea were
    positive with a single dose (Tsuda et al., 1980).

         Sequential biochemical and morphological observations on the
    livers of female rats fed 0.25% safrole in the diet for up to 85 weeks
    showed an early increase in liver weight and activity of drug-
    metabolizing enzymes. With continued administration, the enzyme

    inductive effect disappeared; drug-metabolizing activity was inhibited
    although liver enlargement continued and concentrations of NADPH-
    cytochrome c reductase, cytochrome b5 and microsomal protein were
    elevated. Histochemical and morphological alterations became apparent
    after the loss of the inductive response (Crampton et al., 1977).

         Safrole, in doses of 10 and 295 mg per rat per day,
    dihydrosafrole, in doses of 15 and 360 mg per rat per day, isosafrole,
    in doses of 15 and 375 mg per rat per day, and sassafras oil, in doses
    of 50 and 1130 mg per rat per day, were administered s.c. for the
    first seven post-operative days to partially hepatectomized rats. All
    four substances significantly increased the amount of liver
    regenerated. The same effect was seen when safrole, dihydrosafrole and
    isosafrole, at 0.25% of the diet, were fed for 10 days following
    partial hepatectomy. Sassafras bark tea, at 1.5 or 7% of the diet,
    also increased the amount of liver regenerated (Gershbein, 1977).


    Special studies on carcinogenicity

    Short-term studies


         Topical application of 1.8 µmol of the following substances in
    female mice twice weekly for seven weeks, with subsequent twice weekly
    applications of phorbol-12,13-didecanoate for the remainder of the
    24-week experiment, did not produce a significant increase in the
    incidence of skin papillomas: safrole, HOS, ACOS, 1'-methoxysafrole,
    dihydrosafrole, 1'-hydroxydihydrosafrole, 1'-acetoxydihydrosafrole,
    1'-methoxydihydrosafrole. Groups of 25-30 were used. A positive
    control of 7,12-dimethylbenz(a)anthracene (DMBA), applied once, did
    produce a highly significant increase in skin papilloma incidence
    (Borchert et al., 1973a).

         Groups of 30 female mice were treated topically with the
    following substances in one experiment: 1'-hydroxysafrole-2',3'-oxide,
    HOS, ACOS (1.5) by moles, 1'-oxosafrole, 1'-oxosafrole (1.5) by moles,
    DMBA (0.2) by moles, and in 0.1 ml acetone solvent; and the following
    substances in a second experiment: 1'-hydroxysafrole-2',3'-oxide,
    safrole-2',3'-oxide, 1'-acetoxysafrole-2',3'-oxide, HOS, ACOS,
    1'-oxosafrole, safrole, and solvent. The dose was 30 µmol per
    application except where another dose is indicated in parentheses
    after the compound name. The dose was applied five times a week for
    six weeks and then croton oil (total dose of 0.5 mg in 0.1 ml) acetone
    was applied twice weekly until the experiment was terminated at 36
    weeks. DMBA was applied only once. DMBA, 1'-hydroxysafrole-2',3'-
    oxide, 1'-acetoxysafrole-2',3'-oxide, and safrole-2',3'-oxide were the
    only compounds which produced an increased incidence of papillomas.
    The incidence was 1/60 in the solvent controls, 5/30 with

    1'-acetoxysafrole-2',3'-oxide, 6/30 with safrole-2',3'-oxide, 20/60
    with 1'-hydroxysafrole-2',3'-oxide, and 20/30 with DMBA (Wislocki et
    al., 1977).

         Mice, six to eight weeks of age, were given safrole i.p. three
    times a week for four weeks and killed 24 weeks after the first
    treatment. The doses of safrole were (total) 0.9 and 4.5 g/kg per
    mouse, and groups of 15 mice per sex and dose were used. Safrole did
    not induce pulmonary tumours under the conditions of this study,
    although a number of chemotherapeutic agents including uracil mustard
    were active (Stoner et al., 1973).

    Long-term studies


         Mice were treated with safrole during the lactation period and
    examined for the development of tumours through 49-53 weeks. Safrole
    was administered s.c. in four doses on the 1st, 7th, 14th and 21st day
    of age. Of 12 male mice surviving to weaning, 50% had hepatomas at
    49+ weeks with a total dose of 0.66 mg. With a total dose of 6.6 mg
    safrole and 31 males surviving at weaning, 58% had hepatomas at
    termination. Hepatomas were not observed in safrole-treated females.
    The higher safrole dose also resulted in a slight increase in
    incidence of multiple pulmonary adenomas over that in the controls,
    the occurrence of pulmonary adenocarcinomas, which were not observed
    in the controls, and a slight increase in the incidence of lymphocytic
    and reticular lymphomas. Doses of safrole of 11 and 110 mg on day 1
    were acutely toxic (Epstein et al., 1970).

         Perinatal male mice were given s.c. injections of the following
    substances on days 1, 8, 15 and 22 of age for a total dose of
    3.18 µmol (= 0.52 mg safrole): safrole, HOS, ACOS and 1'-oxosafrole.
    The solvent was trioctanoin and a control group receiving trioctanoin
    was included in the study. The survivors were necropsied at 16 months.
    Group size at necropsy ranged from 31 to 45. Incidence of liver
    carcinomas was: safrole - 51% (14% multiple), HOS - 82%
    (55% multiple), ACOS - 74% (50% multiple), 1'-oxosafrole - 16%
    (0% multiple), control - 13% (5% multiple) (Wislocki et al., 1977).

         Preweanling mice were treated with safrole, HOS, ACOS or solvent
    (trioctanoin) s.c. in a total dose of 9.45 µmol given in four
    injections on days 1, 7, 14 and 21 of age. Groups of 45-50 per sex and
    compound were established at weaning. Surviving males were killed at
    12-14 months and surviving females at 16 months. The incidence of
    liver tumours in the males was 8, 40, 84 and 82% for the controls,
    safrole, HOS and ACOS, respectively. These tumours were multiple in
    more than half of the HOS- and ACOS-treated mice but in only 17% of
    the safrole-treated mice. The incidence of liver tumours in females
    was 0, 0, 16 and 9% for the controls, safrole, HOS and ACOS treated
    mice, respectively (Borchert et al., 1973a).

         In a study similar to the one described above, a total dose of
    4.43 µmol of HOS was administered s.c. to preweanling male mice. The
    dose was divided and administered on days 1, 8, 15 and 22 of age. The
    test groups contained 60 mice and the control group 66. The mice were
    sacrificed at 15 months of age. Six control mice had liver carcinomas;
    none were multiple. Thirty test mice had liver tumours; these tumours
    were multiple in 20 of the mice (Drinkwater et al., 1976).

         Safrole, dihydrosafrole and isosafrole were administered to two
    hybrid strains of mice orally starting at day 7 of age. The test
    substances were administered by stomach tube daily (464, 215 and 464
    mg/kg, respectively) until the mice were weaned at four weeks of age
    and in the diet thereafter (1112, 517 and 1400 ppm (0.1112, 0.0517 and
    0.14%), respectively). Study termination was at approximately 18
    months of age. Eighteen mice of each sex and strain were selected at
    weaning for retention in the study. Hepatomas were observed in 46
    safrole-treated mice, in 19 dihydrosafrole-treated mice and in eight
    isosafrole-treated mice. Pulmonary tumours and lymphomas were also
    observed in the dihydrosafrole- and isosafrole-treated mice. The
    incidences were 18 and three, respectively, with dihydrosafrole and
    four and two, respectively, with isosafrole. The overall incidence of
    the various tumours in control mice was 14/378 with hepatomas, 30/378
    with pulmonary tumours and 18/378 with lymphomas. Hepatoma incidence
    tended to be higher in females than males with safrole but higher in
    males than females with dihydrosafrole and isosafrole (Innes et al.,
    1969). Hyperplasia and carcinomas of the forestomach were increased in
    the females of both strains and in the males of one strain, seven with
    dihydrosafrole. The incidence was 88 and 78% for the females and 41%
    for the males. Maximum incidence in the controls was 28% in the
    females of one strain (Reuber, 1979).

         Safrole (120 µg/g bw per treatment p.o.) was administered to:
    (a) pregnant mice (four times during gestation on days 12, 14, 16 and
    18); (b) lactating mothers (12 times every second day following
    parturition); (c) four-week-old offspring (180 times twice weekly for
    90 weeks); (d) a + b combination treatment; and (e) a + b + c
    combination treatment. All survivors were killed at 94 weeks of age.
    Exposure to safrole in utero produced renal epithelial tumours in 7%
    of the female offspring; none of the other experimental or control
    animals had these tumours. Male offspring (34%) nursed by mothers
    treated with safrole during lactation had hepatocellular tumours but
    not female offspring. However, when safrole was administered post-
    weaning, a significant increase in hepatocellular tumours was observed
    in females (48%) but not in males (8%). Of the tumours observed in the
    females, 86% were hepatocellular carcinomas of which 42% had pulmonary
    metastases (Vesselinovitch et al., 1979).

         Safrole and HOS were fed in the diet to adult mice at final
    levels of 0.50 and 0.55%, respectively, for 12 months. The survivors
    were sacrificed at 17 months. A control group was included and the
    number of mice per sex per group ranged from 45 to 55. It was
    necessary to increase the dietary level of HOS gradually because the
    0.55% level was acutely toxic in uninitiated females. The acute
    toxicity was associated with marked liver toxicity. Both safrole and
    HOS produced depression of weight gain and increased mortality by the
    end of 12 months of treatment. No liver tumours were observed in
    either the male or female controls. Eleven male and 25 female safrole-
    treated mice had liver tumours. No male mice and 30 female HOS-treated
    mice had liver tumours. There were also five s.c. angiosarcomas in the
    female HOS-treated mice (Wislocki et al., 1977).

         The carcinogenicity of dietary safrole was compared to that of
    HOS in one experiment and to that of ACOS in a second experiment in
    male mice. Groups of 25-40 adult mice were used for the test and
    control groups. They were given the test diets for 13 months and then
    control diets for an additional three months. In the first experiment,
    safrole was fed at 0.4 and 0.5% of the diet and HOS at equimolar
    amounts of 0.44 and 0.55%. Mortality between 12 and 16 months was much
    higher in the HOS groups than in the safrole groups. Twenty-seven per
    cent. of the 12-month survivors of safrole feeding had hepatocellular
    carcinomas compared to 14% for HOS and 10% in the controls. In
    addition, 20 mice fed HOS (seven on 0.44%, 13 on 0.55%) had
    interscapular sarcomas. One control and two safrole-fed mice had
    sarcomas in the same area. In the second experiment, safrole was fed
    at 0.4% and ACOS at 0.03% and 0.05 or 0.5% (the text and table differ
    on this level). Administration of the higher dose was terminated at
    seven months because of increased mortality; the 0.03% level also
    caused increased mortality. Three hepatocellular carcinomas were
    observed in the controls, none in the ACOS-treated groups and four in
    the safrole-treated group. No interscapular sarcomas were found in the
    second experiment (Borchert et al., 1973a).


         The carcinogenicity of two possible safrole metabolites,
    1'-oxosafrole and ACOS, was compared by s.c. injection. 18.6 µmol of
    the test compound dissolved in 0.2 ml trioctanoin were injected twice
    weekly for 12 weeks. Groups of 18 rats were used and two experiments
    performed, with one terminated at 19 months and the other at 21
    months. The combined results of the two experiments showed one animal
    with injection-site sarcomas with the vehicle, three with
    1'-oxosafrole and 11 with ACOS (Wislocki et al., 1977).

         The ability of safrole and isosafrole and some derivatives of
    these two substances to induce injection-site sarcomas was compared in
    two experiments. 18.6 µmol of the test compound in 0.2 ml trioctanoin
    was injected s.c. twice weekly for 20 injections and the animals

    killed after 17 or 18 months. The test compounds and number of rats
    (all males) per compound in the pooled experiments were: controls,
    safrole, HOS, ACOS, 3'-hydroxyisosafrole and 3'-acetoxyisosafrole - 36
    rats; isosafrole, 3'-methoxyisosafrole, 2'-bromoisosafrole and
    1'-methoxysafrole - 18 rats. Eleven injection-site sarcomas were
    observed with ACOS, three with HOS and two with 3'-bromoisosafrole.
    None were observed in the controls or in the other test groups
    (Borchert et al., 1973a).

         Groups of 18 male rats were fed, in the diet, 0.22% safrole,
    0.25% 1'-oxosafrole or 0.25% HOS for 17 months, 0.55% HOS for 10
    months, 0.5% safrole for 22 months, or 0.5% safrole plus 1% sodium
    barbital (to stimulate hepatic microsomal enzymes) in the drinking-
    water for 22 months. There were separate control groups for the
    10-17-month and the 22-month feeding periods plus a control group
    which received 1% sodium barbital. All survivors were killed at 22
    months. No hepatic carcinomas or forestomach tumours were observed in
    the control groups, with 0.22% safrole or 0.25% 1'-oxosafrole. One
    hepatic carcinoma was found in the 1% sodium barbital group. In the
    0.25% HOS group there were seven rats with hepatic carcinomas and two
    with papillomas of the forestomach. In the 0.5% safrole group there
    were three hepatic carcinomas; when rats on this safrole level also
    received 1% sodium barbital in the drinking-water there were 12
    hepatic carcinomas. In the 0.55% HOS group there were 16 with hepatic
    carcinomas, two with papillomas and two with carcinomas of the
    forestomach (Wislocki et al., 1977).

         Three rat feeding studies were performed to compare the
    carcinogenicity of safrole, HOS and ACOS administered in the diet for
    8.5-11 months. The rats were killed after 12 or 16 months on
    experiment. The pooled results are as follows (only male rats were
    used): controls, 50 rats, no tumours of liver or forestomach; 0.3%
    safrole, 18 rats, no tumours of liver or forestomach; 0.5% safrole, 50
    rats, one with hepatic adenoma, one with hepatic carcinoma and no
    forestomach tumours; 0.5% HOS, 30 rats, 27 with hepatic carcinomas and
    12 with forestomach papillomas; ACOS, 15 rats with doses from 0.41 to
    0.68%, none with hepatic carcinomas and 10 with forestomach
    papillomas; and 0.41% ACOS, 18 rats, one with hepatic carcinoma, 11
    with forestomach papillomas and one with squamous cell carcinoma of
    the forestomach. The hepatic tumours in the HOS rats were larger and
    more numerous than those in the safrole rats. The livers of the HOS
    rats also showed necrosis, fibrosis and deviations from normal
    architecture, while the livers of the safrole rats showed little
    deviation from normal. The safrole and HOS rats grew more slowly than
    the controls but survival was good. ACOS, at equivalent dose levels,
    caused greater depression of growth and early deaths (Borchert et al.,

    Special studies on mutagenicity

         Safrole was generally inactive in mutagenicity studies in various
    strains of S. typhimurium with or without metabolic activation
    (McCann et al., 1975; Poirier & de Serres, 1979; Cheh et al., 1980;
    Dorange et al., 1977b; Swanson et al., 1979; Wislocki et al., 1977)
    but has occasionally been reported to be weakly positive (Purchase et
    al., 1978; Poirier & de Serres, 1979; Green & Savage, 1978). A special
    activation system was described which yielded positive results with
    safrole in strain TA-1535 (Dorange et al., 1978). Dihydrosafrole and
    isosafrole (Wislocki et al., 1977) and HOS (McCann et al., 1975;
    Poirier & de Serres, 1979; Dorange et al., 1977a; Drinkwater et al.,
    1976; Wislocki et al., 1977) were also negative in this test. ACOS was
    generally positive (McCann et al., 1975; Poirier & de Serres, 1979;
    Drinkwater et al., 1976; Wislocki et al., 1977). 3'-hydroxyisosafrole,
    3'-acetoxyisosafrole and 1'-oxosafrole were not mutagenic for strains
    TA-100 and TA-1535 with or without metabolic activation, but the
    2',3'-oxides of safrole, HOS, ACOS and 1'-oxosafrole were directly
    mutagenic for these strains (Wislocki et al., 1977). The mutagenic
    activity of these 2',3'-oxides (Swanson et al., 1979) and of
    2',3'-epoxysafrole (Dorange et al., 1977a) for TA-1535 was confirmed.
    The safrole ninhydrin-positive metabolite II was positive with
    metabolic activation in TA-1530 and TA-1532 (Green & Savage, 1978).
    Safrole ninhydrin-positive metabolites I and III were inactive with or
    without metabolic activation (Green & Savage, 1978).

         Safrole was positive in the following short-term tests for
    mutagenesis: mammalian cell transformation in culture, Rabin's test-
    degranulation of rough endoplasmic reticulum from rat liver, and
    mouse-sebaceous gland suppression test; and negative in the
    tetrazolium reduction and tissue reaction to s.c. implants in mice
    (Purchase et al., 1978). Safrole was positive in in vitro mutagenic
    assays with E. coli and S. cerevisiae and in the in vivo
    intraperitoneal host-mediated assay with S. typhimurium strain
    TA-1535 or S. cerevisiae (Poirier & de Serres, 1979). Safrole was
    positive in the host-mediated assay with strains TA-1950 and TA-1952
    (Green & Savage, 1978).

         HOS and ACOS were both positive in in vitro tests with
    S. cerevisiae; ACOS was positive with E. coli, but HOS was
    negative (Poirier & de Serres, 1979). The safrole ninhydrin-positive
    metabolite II was positive in the host-mediated assay using
    S. typhimurium strains TA-1950 and TA-1952; the ninhydrin-positive
    metabolites I and III were negative in the host-mediated assay with
    strains TA-1950 and TA-1952 (Green & Savage, 1978).

    Special studies on pharmacology

         Safrole at a dose of 20 mg/kg bw i.p., approximately doubled the
    sleeping time of mice treated with sodium pentobarbital. Isosafrole in
    the same dose was slightly less active. Neither substance had a
    significant effect on ethanol sleeping time (Seto & Keup, 1969).

    Acute toxicity

        Animal         Route   (mg/kg bw)          Reference


       Mouse           Oral       2 350       Jenner et al., 1964

       Rat             Oral       1 950       Jenner et al., 1964


       Mouse           Oral       3 700       Hagan et al., 1965

                       Oral       4 300       Jenner et al., 1964

       Rat             Oral       2 260       Jenner et al., 1964


       Mouse           Oral       2 470       Jenner et al., 1964

       Rat             Oral       1 340       Jenner et al., 1964

    Short-term studies


         The ability of safrole (650 mg/kg), isosafrole (460 mg/kg) and
    dihydrosafrole (770 mg/kg) to induce macroscopic liver damage was
    compared in groups of three male and three female rats with doses of
    approximately one-third the oral LD50 administered daily by stomach
    tube for four days. The changes induced by these three compounds
    ranged from slight discoloration with blunting of lobe edges to marked
    discoloration, fatty appearance and enlargement. The average rating of
    the lesions was the same for the three compounds; there was one death
    in the dihydrosafrole group and two in the isosafrole group (Taylor et
    al., 1964).

         Safrole (250, 500 and 750 mg/kg), dihydrosafrole (250, 500 and
    750 mg/kg) and isosafrole (250 and 500 mg/kg) and corn oil were
    administered daily by stomach tube to groups of 10 rats. Treatment
    time for the test animals ranged from 19 to 46 days, while the
    controls were treated for 106 days. There were no deaths with the 250
    and 500 mg doses of dihydrosafrole and the 250 mg dose of safrole. A
    few deaths, 2/10, 3/10 and 1/10, respectively, resulted from the
    administration of 250 mg isosafrole, 750 mg dihydrosafrole and 500 mg
    safrole. Many deaths, 8/10 and 9/10, respectively, were seen with the
    500 mg dose of isosafrole and the 750 mg dose of safrole.
    Macroscopically, all three compounds induced liver enlargement and
    adrenal enlargement with yellow discoloration, most marked with
    safrole and dihydrosafrole. Microscopically, there was variation in
    hepatic cell size, focal fatty metamorphosis, bile-duct proliferation
    and architectural irregularity. Adrenals from the rats given safrole
    and dihydrosafrole showed an increase in lipid in the cytoplasm of the
    cortex. Safrole was administered to mice at doses of 250 and 500 mg/kg
    daily by stomach tube for 60 days (number of mice not indicated).
    Liver changes similar to those observed in the rat study were observed
    (Hagan et al., 1965).


         Groups of two male and two female dogs received a dose of
    80 mg/kg per day of safrole orally for 26-39 days or 40 mg/kg per day
    for 91-116 days. At 80 mg/kg, one dog died and the others were
    sacrificed moribund. These animals exhibited marked weight loss.
    Skeletal muscle was atrophied in 2/4 dogs. The liver was slightly
    enlarged in 2/4 dogs. Microscopically, the livers showed a heavy
    infiltration of fat, occasional bile-duct proliferation and
    inflammatory-cell infiltration, slight to moderate streaking with
    necrobiosis, and liver cell atrophy. At 40 mg/kg, the dogs exhibited
    emaciation and general weakness; the liver was slightly enlarged in
    3/4 dogs. Microscopically, there was slight fat infiltration, moderate
    cell atrophy and necrobiosis in the liver (Hagan et al., 1967).

    Long-term studies


         Rats (number not reported) were fed safrole or natural oil of
    sassafras at levels of 390 or 1170 ppm (0.039 or 0.117%) in the diet
    for two years. Hepatocarcinomas were found in all the experimental
    groups as well as cellular changes in the kidneys, adrenals, thyroid,
    pituitary, and testes or ovaries at 24 months. Kidneys and livers from
    high level animals sacrificed at 22 months showed evidence of kidney
    congestion but no liver tumours. Blood and urine examinations were
    within normal limits (Abbott et al., 1961).

         Rats were fed riboflavin-deficient diets (four males and 10
    females) or riboflavin-deficient diets plus 1% safrole (five males and
    14 females) for 3.5-13.5 months. Body weight of both sexes fed safrole
    was depressed; liver weight/body weight ratio of these animals was
    elevated. The riboflavin-deficient diet produced fatty degeneration,
    some increase in connective tissue and extensive ceroid deposits in
    the livers of the males. In the females, it produced mild to moderate
    fatty infiltration with no increase in fibrous connective tissues
    until 13.5 months when fibrous connective tissue obscured the normal
    architecture and nodular regeneration appeared. Addition of safrole to
    the diet decreased fatty infiltration in both sexes and produced
    adenomatous changes much earlier in the females (Homburger et al.,

         Groups of 10 male rats were fed control diets containing 5, 10 or
    30% protein and 5% fat or these diets with 0.5% safrole or 0.06%
    butter yellow added. Additional groups received a diet containing 30%
    protein and 15% fat or a riboflavin-deficient diet with 10% protein
    and 5% fat, and these diets with safrole or butter yellow. The rats
    were killed when they appeared moribund (55-634 days). Altering the
    dietary protein level had little effect on the incidence of hepatoma-
    hepatocarcinomas with butter yellow, although the rats on butter
    yellow and the riboflavin-deficient diet showed early mortality.
    Hepatoma-hepatocarcinoma incidence increased with increasing protein
    level in the safrole-treated animals. The safrole rats receiving 5%
    protein showed early mortality. The high fat and riboflavin-deficient
    diets had little effect on safrole toxicity or tumour incidence
    (Homburger et al., 1965).

         Groups of 25 male and 25 female rats were fed 0, 100, 500, 1000
    and 5000 ppm (0, 0.01, 0.05, 0.1 and 0.5%) safrole in the diet for two
    years. Changes in the liver consisting of benign and malignant tumours
    (hepatic cell adenoma, hepatocholangioma, hepatic cell carcinoma and
    hepatocholangiocarcinoma), enlargement of hepatic cells, variation in
    cell size, fatty metamorphosis, architectural irregularity, bile-duct
    proliferation, focal cystic necrosis, focal peripheral margination of
    cytoplasm and minimal coagulation necrosis were observed. The liver
    injury was rated as very slight at 100 ppm (0.01%), slight at 500 ppm
    (0.05%), slight to moderate at 1000 ppm (0.1%), and moderate to severe
    at 5000 ppm (0.5%). Tumour incidence was significantly increased at
    5000 ppm (0.5%), with 14 rats having malignant tumours and five having
    benign tumours versus two and one, respectively, in the controls.
    Tumour incidence in the other groups was: eight benign on 1000 ppm
    (0.1%), two malignant and one benign on 500 ppm (0.05%), and one
    benign on 100 ppm (0.01%). Weight gain and survival were significantly
    decreased on the 5000 ppm (0.5%) level and these rats also showed a
    mild hypochromic anaemia, a slight leucocytosis and a slight
    depression of haematopoiesis in the bone marrow (Long et al., 1963).

         Safrole, dihydrosafrole and isosafrole were fed in the diet at
    levels of 0, 1000, 2500 and 10 000 ppm (0, 0.1, 0.25 and 1%) to groups
    of 10 male and 10 female rats and at a level of 5000 ppm (0.5%) to
    groups of 25 males and 25 females for two years. Growth was depressed
    in both sexes at levels of 2500 ppm (0.25%) and above and in the
    females at 1000 ppm (0.1%) with both safrole and dihydrosafrole. With
    isosafrole, growth was depressed in both sexes at 5000 and 10 000 ppm
    (0.5 and 1%) and in the females at the two lower doses. None of the
    10 000 ppm (1%) safrole test rats survived beyond 62 weeks. These
    animals showed testicular atrophy, changes in the stomach consisting
    of atrophy and atypical regeneration of the mucosa glands with
    associated fibrosis and hyalinization of the surrounding stroma, and
    changes in the liver including tumour formation. The livers were
    enlarged, mottled and irregularly nodular, with single and multiple
    tumour masses. Microscopically, there was hepatic cell enlargement
    which was usually focal and resulted in nodule formation. The nodules
    tended to progress in one of three ways: (1) cystic necrosis, (2)
    cirrhosis, and (3) adenomatoid hyperplasia, leading to formation of
    benign and malignant tumours of the same types as in the study cited
    above. Similar types of liver changes were seen with the lower doses.
    The damage was slight at 1000 ppm (0.1%) and lacked tumours and
    cirrhosis, moderate at 2500 ppm (0.25%) but lacked cirrhosis, and
    severe at 5000 ppm (0.5%) where there were macroscopic cysts and the
    incidence of tumours exceeded the incidence at the higher level. There
    was a statistically significant increase in malignant hepatic primary
    tumours with this level. There was mild hyperplasia of the thyroid at
    5000 ppm (0.5%) and an increase in chronic nephritis with this dose
    and the lower doses. Dose levels of dihydrosafrole of 2500 ppm (0.25%)
    and above induced eosophageal tumours, benign epidermoid papillomas
    and malignant papillary epidermoid carcinomas. No tumours of this type
    were observed in the controls. At 2500 ppm (0.25%), 20% of the rats
    had eosophageal tumours, with 5% of these malignant. At 5000 ppm
    (0.5%), 74% of the rats had such tumours with 32% malignant and, at
    10 000 ppm (1%), 75% of the rats had tumours with 50% malignant.
    Slight liver damage of the same type as seen with safrole was found
    with all test levels of dihydrosafrole; a few tumours were seen, but
    the increase was not statistically significant. Moderate follicular
    atrophy of the spleen was observed at the 5000 and 10 000 ppm (0.5 and
    1%) levels and a moderate increase in chronic nephritis at the two
    lower dose levels. The 10 000 ppm (1%) level of isosafrole increased
    mortality, with none of the rats on this level surviving beyond 11
    weeks of treatment. Slight liver damage of the same type seen with
    safrole was seen at all test levels of isosafrole. There were five
    rats on the 5000 ppm (0.5%) level with primary hepatic tumours, but
    this incidence was not significantly greater than in the controls.
    There was slight thyroid hyperplasia on the 2500 and 5000 ppm (0.25
    and 0.5%) levels and an increased incidence of chronic nephritis on
    the 5000 ppm (0.5%) level (Hagan et al., 1965; Hagan et al., 1967).


         Safrole was administered orally to groups of two male and two
    female dogs at 5 and 20 mg/kg bw for six years. Liver changes, but no
    tumours, were observed with both doses. The changes at the lower dose
    were minimal focal necrosis, bile-duct proliferation, fatty
    metamorphosis, hepatic cell atrophy and leucocytic infiltration. At
    the higher dose, the liver was enlarged with a nodular surface.
    Microscopically, it showed mild post-necrotic cirrhosis characterized
    by focal or generalized nodules, consisting of enlarged hepatic cells
    with slight focal fatty vacuolization, separated by bands of atrophied
    hepatic cells and collagen; slight lymphocytic, Kupffer cell and bile-
    duct proliferation; and minimal focal necrosis. With this dose the
    thyroids were hyperplastic (Hagan et al., 1967).

         Safrole was fed to dogs (number and dose not indicated) for seven
    years. Liver function tests consisting of bromsulfalein excretion
    curve, soluble enzyme content in the central lobe segments, lipid and
    glycogen content of the same segments, and nitroreductase activity
    were performed during the course of the study and at termination. The
    results of these function tests indicated some degree of tissue injury
    during the early part of the study, but a functional adaptation to the
    test substance by the end of the study, although microscopic
    morphological changes were observed at termination (Weinberg &
    Sternberg, 1966).

         The toxicity of safrole was summarized by Opdyke (1974).


         Small doses of safrole are absorbed rapidly and almost completely
    excreted in the urine within 25 hours of exposure in both rat and man.
    1,2-dihydroxy-4-allylbenzene is the main metabolite in both species.

         Safrole and isosafrole administered to rats produces liver
    hypertrophy and induces microsomal enzymes. Safrole is inactive in
    mutagenicity studies with various strains of S. typhimurium with and
    without activation. Safrole was positive in the in vitro mutagenic
    assay with E. coli and S. cerevisiae and in the in vivo
    intraperitoneal host-mediated assay.

         Administration of safrole to mice, either orally or
    subcutaneously, led to a marked increase in the incidence of liver
    tumours. Exposure of mice to safrole in utero produced renal
    epithelial tumours. In the case of rats, chronic administration of
    safrole resulted in progressive dose-dependent liver damage ranging
    from hepatic cell enlargement, nodule formation, cirrhosis adenomatoid
    hyperplasia leading to benign and malignant tumours. No liver tumours
    were reported in dogs fed safrole for six years, but some liver
    changes including bile-duct proliferation were observed.


    No ADI allocated.


    Abbott, D. D. et al. (1961) Chronic oral toxicity of oil of sassafras
         and safrole, Pharmacologist, 3, 62

    Bangdiwala, V. S. & Oswald, E. O. (1976) Metabolic interaction of
         secondary amines and tertiary amino propiophenones with monoamine
         oxidase systems, Chem. Biol. Interactions, 14, 141-148

    Benedetti, M. S., Malnoe, A. & Broillet, A. L. (1977) Absorption,
         metabolism and excretion of safrole in the rat and man,
         Toxicology, 7, 69-83

    Borchert, P. et al. (1973a) 1'-hydroxysafrole, a proximate
         carcinogenic metabolite of safrole in the rat and mouse, Cancer
         Research, 33, 590-600

    Borchert, P. et al. (1973b) The metabolism of the naturally occurring
         hepatocarcinogen safrole to 1'-hydroxysafrole and the
         electrophilic reactivity of 1'-acetoxysafrole, Cancer Research,
         33, 575-589

    Casida, J. E. et al. (1966) Methylene-C14-dioxyphenyl compounds:
         Metabolism in relation to their synergistic action, Science,
         153, 1130-1133

    Cheh, A. M. et al. (1980) A comparison of the ability of frog and rat
         S-9 to activate promutagens in the Ames test, Environmental
         Mutagenesis, 2, 487-508

    Crampton, R. F. et al. (1977) Long-term studies on chemically induced
         liver enlargement in the rat. II. Transient induction of
         microsomal enzymes leading to liver damage and nodular
         hyperplasia produced by safrole and Ponceau MX, Toxicology,
         7, 307-326

    Delaforge, M. et al. (1977) Direct evidence of epoxide pathway for
         natural allylbenzene compounds in adult rat liver cell culture.
         In: A. Frigerio, ed., Recent development in mass spectrometry
         in biochemistry and medicine, New York and London, Plenum
         Press, vol. 1, pp. 521-539

    Dorange, J. L. et al. (1977a) Mutagenicity of metabolic derivatives
         of safrole, Mutation Research, 46 (3), 217

    Dorange, J. L. et al. (1977b) Pouvoir mutagine de métabolites de la
         voie epoxyde-diol du safrol et d'analogues. Etude sur
         Salmonella typhimurium, C.R. Séances Soc. Biol., 171,

    Dorange, J. L. et al. (1978) Comparative survey of microsomal
         activation systems for mutagenic studies of safrole, Mutation
         Research, 53, 179-180

    Doumas, J. & Maume, B. F. (1977) Activation métabolique par le tissu
         surrénalien du rat d'un carcerogine du foie: le safrol, C.R.
         Séances Soc. Biol., 171, 108-114

    Drinkwater, N. R. et al. (1976) Hepatocarcinogenicity of estragole
         (1 allyl-4-methoxybenzene) and 1'-hydroxyestragole in the mouse
         and mutagenicity of 1'-acetoxyestragole in bacteria, J. Natl.
         Cancer Inst., 57, 1323-1331

    Elcombe, C. R. et al. (1975) Studies on the interaction of safrole
         with rat hepatic microsomes, Biochem. Pharmacol., 24,

    Epstein, S. S. et al. (1970) Carcinogenicity testing of selected food
         additives by parenteral administration to infant Swiss mice,
         Toxicol. Appl. Pharmacol., 16, 321-334

    Fishbein, L. et al. (1967) Thin-layer chromatography of rat bile and
         urine following intravenous administration of safrole,
         isosafrole, and dihydrosafrole, J. Chromatog., 29, 267-273

    Friedman, M. A. & Staub, J. (1976) Inhibition of mouse testicular DNA
         synthesis by mutagens and carcinogens as a potential simple
         mammalian assay for mutagenesis, Mutation Research, 37, 67-76

    Friedman, M. A. et al. (1971) Additive and synergistic inhibition of
         mammalian microsomal enzyme functions by piperonyl butoxide,
         safrole and other methylenedioxyphenyl derivatives,
         Experientia, 27, 1052-1054

    Fritsch, P., de Saint Blanquat, G. & Derache, R. (1975a) Absorption
         gastro-intestinale, chez le: Rat, de l'anisle, du transanethole,
         du butylhydroxyanisole et du safrole, Fd Cosmet. Toxicol.,
         13, 359-363

    Fritsch, P., Lamboeuf, Y. & de Saint Blanquat, G. (1975b) Effet de
         l'anisole, de l'anethole, du butylhydroxyanisole et du safrole
         sur l'absorption intestinale chez le rat, Toxicology, 4,

    Gershbein, L. L. (1977) Regeneration of rat liver in the presence of
         essential oils and their components, Fd Cosmet. Toxicol., 15,

    Green, N. R. & Savage, J. R. (1978) Screening of safrole, eugenol,
         their ninhydrin positive metabolites and selected secondary
         amines for potential mutagenicity, Mutation Research, 57,

    Gwynn, J., Fry, J. R. & Bridges, J. W. (1979) The effects of
         paracetamol and other foreign compounds on protein synthesis in
         isolated adult rat hepatocytes, Biochem. Sec. Trans., 2,

    Hagan, E. C. et al. (1965) Toxic properties of compounds related to
         safrole, Toxicol. Appl. Pharmacol., 7, 18-24

    Hagan, E. C. et al. (1967) Food flavourings and compounds of related
         structure. II. Subacute and chronic toxicity, Fd Cosmet.
         Toxicol., 5, 141-157

    Homburger, F., Bogdonoff, P. D. & Kelley, T. F. (1965) Influence of
         diet on chronic oral toxicity of safrole and butter yellow in
         rats, Proc. Soc. Exptl. Biol. Med., 119, 1106-1110

    Homburger, F. et al. (1962) Sex effect on hepatic pathology from
         deficient diet and safrole in rats, Arch. Pathol., 73,

    Innes, J. R. M. et al. (1969) Bioassay of pesticides and industrial
         chemicals for tumorigenicity, J. Natl. Cancer Inst., 42,

    International Agency for Research on Cancer (1976) IARC Monographs on
         the Evaluation of the Carcinogenic Risk of Chemicals to Man. Some
         Naturally Occurring Substances, vol. 10, pp. 231-244

    Janiaud, P. & Padieu, P. (1977) Adult rat liver epithelial cells for
         studies of carcinogenic activity of safrole, Mutation Research,
         46, 221-222

    Janiaud, P. et al. (1976) Etude comparative en culture cellulaire de
         foie de rat du métabolisme de différents analogues et métabolites
         d'un hépatocancerogine naturel: le safrole, C.R. Séances Soc.
         Biol., 170, 1035-1041

    Jenner, P. M. et al. (1964) Food flavourings and compounds of related
         structure. I. Acute oral toxicity, Fd Cosmet. Toxicol., 2,

    Lake, B. G. & Parke, D. V. (1972) Induction of aryl hydrocarbon
         hydroxylase in various tissues of the rat by methylenedioxyphenyl
         compounds, Biochem. J., 130, 86

    Long, E. L. et al. (1963) Liver tumors produced in rats by feeding
         safrole, Arch. Pathol., 75, 595-604

    Lotlikar, P. D. & Wasserman, M. B. (1972) Effects of safrole and
         isosafrole pretreatment on N- and ring-hydroxylation of
         2-acetamidofluorene by the rat and hamster, Biochem. J., 129,

    McCann, J. et al. (1975) Detection of carcinogens as mutagens in the
         Salmonella/microsome test: Assay of 300 chemicals, Proc. Nat.
         Acad. Sci. USA, 72, 5135-5139

    McKinney, J. D. et al. (1972) On the mechanism of formation of Mannich
         bases as safrole metabolites, Bull. Environ. Contamin.
         Toxicol., 7, 305-310

    Martin, C. N., McDermid, A. C. & Garner, R. C. (1978) Testing of known
         carcinogens and noncarcinogens for their ability to induce
         unscheduled DNA synthesis in HeLa cells, Cancer Research, 38,

    Opdyke, D. L. J. (1974) Fragrance raw materials monographs: Safrole,
         Fd Cosmet. Toxicol., 12 (Suppl.), 983-986

    Oswald, E. O., Fishbein, L. & Corbett, B. J. (1969) Metabolism of
         naturally occurring propenylbenzene derivatives. I.
         Chromatographic separation of ninhydrin-positive materials of rat
         urine, J. Chromatog., 45, 437-445

    Oswald, E. O. et al. (1971) Identification of tertiary amino-
         methylenedioxy-propiophenones as urinary metabolites of safrole
         in the rat and guinea pig, Biochem. Biophys. Acta, 230,

    Parke, D. V. & Rahman, H. (1970) The induction of hepatic microsomal
         enzymes by safrole, Biochem. J., 119, 53P

    Parke, D. V. & Rahman, H. (1971) Induction of a new hepatic microsomal
         haemoprotein by safrole and isosafrole, Biochem. J., 123, 9P

    Peele, J. D., jr & Oswald, E. O. (1978) Metabolism of the proximate
         carcinogen 1'-hydroxysafrole and the isomer 3'-hydroxyisosafrole,
         Bull. Environ. Contamin. Toxicol., 19, 396-402

    Poirier, L. A. & de Serres, F. J. (1979) Initial National Cancer
         Institute studies on mutagenesis as a prescreen for chemical
         carcinogens: An appraisal, J. Natl. Cancer Inst., 62, 919-926

    Purchase, I. F. H. et al. (1978) An evaluation of 6 short-term tests
         for detecting organic chemical carcinogens, Br. J. Cancer,
         37, 873-903

    Reuber, M. D. (1979) Neoplasms of the forestomach in mice ingesting
         dihydrosafrole, Digestion, 19, 42-47

    Sethi, M. L. et al. (1976) Identification of volatile constituents of
         Sassafras albidum root oil, Phytochemistry, 15, 1773-1775

    Seto, T. A. & Keup, W. (1969) Effects of alkylmethoxybenzene and
         alkylmethylenedioxybenzene essential oils on pentobarbital and
         ethanol sleeping time, Arch. Int. Pharmacodyn., 180, 232-240

    Stillwell, W. G. et al. (1974) The metabolism of safrole and 2',3'
         epoxysafrole in the rat and guinea pig, Drug Metabolism &
         Deposition, 2, 489-498

    Stoner, G. D. et al. (1973) Test for carcinogenicity of food additives
         and chemotherapeutic agents by the pulmonary tumor response in
         strain A mice, Cancer Research, 33, 3069-3085

    Swanson, A. B. et al. (1979) The mutagenicities of safrole, estragole,
         eugenol, trans-anethole, and some of their known or possible
         metabolites for Salmonella typhimurium mutants, Mutation
         Research, 60, 143-153

    Taylor, J. M., Jenner, P. M. & Jones, W. I. (1964) A comparison of the
         toxicity of some allyl, propenyl, and propyl compounds in the
         rat, Toxicol. Appl. Pharmacol., 6, 378-387

    Tsuda, H., Lee, G. & Farber, E. (1980) Induction of resistant
         hepatocytes as a new principle for a possible short-term
         in vivo test for carcinogens, Cancer Research, 40,

    Vesselinovitch, S. D., Rao, K. V. N. & Mihailovich, N. (1979)
         Transplacental and lactational carcinogenesis by safrole,
         Cancer Research, 39, 4378-4380

    Weinberg, M. S. & Sternberg, S. S. (1966) Effect of chronic safrole
         administration on hepatic enzymes and functional activity in
         dogs, Toxicol. Appl. Pharmacol., 8, 2

    Wislocki, P. G. et al. (1976) The metabolic activation of the
         carcinogen 1'-hydroxysafrole in vivo and in vitro and the
         electrophilic reactivities of possible ultimate carcinogens,
         Cancer Research, 36, 1685-1695

    Wislocki, P. G. et al. (1977) Carcinogenic and mutagenic activities of
         safrole, 1'-hydroxysafrole, and some known or possible
         metabolites, Cancer Research, 37, 1883-1891

    See Also:
       Toxicological Abbreviations