SMOKE FLAVOURINGS

    EXPLANATION

         Smoke condensates and liquid smoke were considered at the
    nineteenth meeting of the Joint FAO/WHO Committee on Food Additives
    (Annex 1, reference 38). Inadequate information was available at that
    time for an evaluation.

         The present Committee reviewed both the specifications and safety
    data for this group of products. It noted that smoke flavourings were
    complex mixtures of varying composition, primarily prepared by the
    condensation of smoke generated by pyrolysis of certain hardwoods in
    the absence of or in the presence of a limited amount of air. Woods
    commonly used for the preparation of smoke flavourings include oak
    (Quercus spp.), hickory (Carya spp.), beech (Fagus spp.), alder
    (Alnus spp.), and maple (Acer spp.). No toxicity data are
    available on pyroligneous acid preparations which are condensates
    derived from the pyrolysis of wood in the absence of air. The smoke
    flavourings reviewed in this monograph are derived from the smoke
    condensate of wood burned in a limited amount of air.

         The initial smoke condensate separates into an aqueous phase and
    a tarry phase. The smoke condensate may be separated into fractions by
    physical separation techniques or solvent extraction. These fractions
    may be further purified, if necessary, to remove hazardous
    constituents known to be present in smoke. Smoke flavourings include
    smoke condensates, fractions thereof, and mixtures of such fractions.

    SMOKE FLAVOURINGS (AQUEOUS PHASE)

    BIOLOGICAL DATA

    Biochemical aspects

    No information available.

    Toxicological studies

    Special study on carcinogenicity

         An aqueous wood smoke flavouring, which previously had been shown
    to be inactive in a Salmonella typhimurium mutation assay but
    positive in an assay utilizing TK6 human lymphoblasts, was tested in
    the mouse lung adenoma assay. In this study, newborn Swiss Webster
    mice were injected with the smoke flavouring on days 1, 8, and 15, and
    then maintained until 26 weeks of age. The total dose of test
    substance administered ranged from 17.5 µl to 31.5 µl. The test
    substance did not induce lung tumours or rumours at other sites.
    Toxicity was observed at 15 weeks in some of the test animals, which
    included hyperplastic kidneys and abnormalities in the colon and
    rectum (Braun et al, 1986).

    Special studies on mutagenicity

         Smoke flavouring preparations were tested using Salmonella
    typhimurium strains TA98, TA100, TA1535, TA1537, and TA1538 and in
    E. coli WP2 (UVRA), both with and without metabolic activation, at
    doses ranging from 3 to 10,000 µg/plate. Toxicity was observed at the
    highest dose tested without activation. The preparations were not
    mutagenic in these tests (Mortelmans & Eckford, 1980; Rattech, 1981).

         An aqueous wood smoke flavouring which was not mutagenic in the
    Salmonella typhimurium forward mutation assay (using strain TM677),
    at concentrations up to 18 µg/ml (of dissolved solids), induced a
    significant increase in mutation frequency of TK6 human lymphoblasts
    (Braun et al., 1986).

    Acute toxicity

    No information available.

    Short-term studies

    Rats

         Four groups, each containing 10 male and 10 female rats, were
    maintained for 90 days on diets containing 0, 0.3, 2.5, or 20% of a
    liquid smoke preparation (maple or hickory). All diets were adjusted
    to contain 20% added water. Slight growth reduction was observed at

    the high-dose level, but this was associated with decreased feed
    intake. Haematologic and urine analyses at the end of the study showed
    no abnormalities in any of the test groups. At autopsy, organ-weight
    analyses of the principal organs and tissues showed no significant
    changes in absolute organ weights. Gross and microscopic examinations
    of these tissues showed no compound-related effects (WARF, 1961).

         Three groups of 25 male and 25 female rats were maintained for 90
    days on diets containing 0, 0.25, or 2.0% of one liquid smoke
    flavouring or 2.0% of another smoke flavouring preparation. Growth and
    food intake were similar among test and control groups. Haematologic
    values at week 6 and at termination of the study were within normal
    limits in all groups. At autopsy, organ-weight analyses and gross
    pathology and histopathology of the principal organs were carried out.
    No compound-related effects were observed, with the exception of minor
    degenerative changes in the liver and kidneys in one of the 2.0% dose
    groups and slight bone marrow hypoplasia in both 2.0% dose groups
    (WRC, 1963; FCT, 1965).

         Analytical studies on the preparation fed at 0.25 and 2.0% of the
    diet showed the absence of 3,4-benzo(a)pyrene, but the presence of 5
    non-carcinogenic polycyclic aromatic hydrocarbons, namely benzo(a)
    anthracene, carbazole, chrysene, pyrene, and fluouranthene at levels
    of 0.012, 0.20, 0.011, 0.013, and 0.032 ppm of the liquid smoke
    preparation, respectively. The other preparation contained formic,
    propionic, vanillic, and siringic acids, dimethylphenol, methyl
    glyoxal, furfural, acetaldehyde, acetone, ethanol, and benzopyrenes
    (< 1 ppb benzo(a)pyrene, 6 ppb pyrene, 3 ppb flouranthene, and 18 ppb
    phenanthrene (Lijinsky & Shubik, 1965).

         A series of studies was carried out with two liquid smoke
    flavourings. The concentrations of total acids, phenols, and carbonyl
    compounds in these preparations are outlined in Table 1.

    Table 1.  Constituents of liquid smoke 1 (LS 1) and liquid smoke
              2 (LS 2)
                                                                        

    Constituent                       LS 1            LS 2
                                                                        

    Total acids (acetic acid)         1.8 g/l         2.0-2.8 g/l
    Total carboxylic compounds        2.0 g/l         3.0-3.6 g/l
    Phenols                           0.018 g/l       0.9-1.3 g/l
    PAHs (3,4-benzo(a)pyrene)         --              0.2 ppb
                                                                        

         Five groups, each containing 6 male and 6 female Colworth Wistar
    rats, were fed diets containing 1.25, 2.5, 5, 10, or 20% LS 2 or 20%
    LS 1 for 28 days. Another group of 12 male and 12 female rats was fed
    control diets. Body-weight gains of both male and female rats fed 20%
    of either preparation, or 5 or 10% LS 2, were less than those of
    controls. Plasma analysis at termination of the study showed sporadic
    variation in a number of constituents, but these variations were not
    dose-related. However, decreased plasma alkaline phosphatase was
    observed at all dose levels in male rats and female rats in the
    high-dose groups showed elevated thyroxine, (T4) and
    triiodothyronine (T3) levels. Haematologic changes at the high-dose
    level included decreased platelet counts in male rats administered LS
    1 and in haemoglobin and red cell counts in male rats fed LS 2. At
    autopsy, organ-weight analysis showed increased relative liver and
    kidney weights in high-dose male rats fed LS 2 and increased thyroid
    weight in high-dose female rats fed LS 2. Histopathological
    examination of tissues and organs showed no compound-related effects,
    with the exception of severe corticomedullary nephrocalcinosis in
    females fed 20% LS 2 (Parish, 1986a).

         In a paired feeding study, groups of 6 male and 6 female Colworth
    Wistar rats were fed diets containing 2 or 20% LS 1 or LS 2. A control
    group of 12 male and 12 female rats was fed control diet
    ad libitum. Minor differences in weight gain were observed between
    test animals and paired controls. At autopsy, serum chemistry values
    of test animals and paired controls were similar. Organ-weight
    analysis showed significantly, increased liver and kidney weights in
    rats fed liquid smoke when compared to paired controls. Macroscopic
    evaluation of organs and microscopic examination of the livers and
    kidneys showed no compound-related effects (Parish, 1986b).

         Three groups of 20 male and 20 female Carworth Wistar rats were
    maintained on diets containing LS 2 for 13 weeks. The levels of liquid
    smoke in the diet were adjusted biweekly, so that the intake in mg/kg
    b.w. was constant. The levels of liquid smoke in the diet were
    0.5-1.0, 1.5-3.0, and 3.0-6.0%, which were equivalent to daily intakes
    of 0.6, 1.9, and 3.8 g/kg b.w./day for males and 0.7, 2.2, and
    4.4 g/kg b.w./day for females, respectively. Additional groups at the
    high-dose level and the control group were maintained for another 13
    weeks on control diets. No animals died during the course of the
    study. No significant differences in body weights were observed at
    weeks 13 or 26, although there were significant reductions in
    body-weight gain during weeks 4-8 in both male and female rats in the
    high-dose group. At termination of the study, serum analysis showed a
    number of changes in both sexes in the high-dose group. Haematologic
    changes were also observed in this group, which included increased
    haemoglobin and packed RBC volume in male rats and decreased WBC
    counts and lymphocytes in females. Organ-weight analyses showed

    significant increases in relative liver and kidney weights of male and
    female rats in the high-dose group. Relative liver weights of female
    rats were also increased in the 1% group. Histopathology of the
    tissues and organs at week 13 of the study showed no treatment-related
    changes. All the changes observed in rats at 13 weeks were no longer
    apparent after the 13-week recovery period (Parish, 1986c).

         In a study designed to investigate possible effects of smoke
    flavouring on the thyroid, groups of 3 male and 3 female rats received
    diets containing from 0.625 to 20% LS 1 for 28 days. Another group was
    administered 0.025% 2-thiouracil in the diet, and another group of 6
    male and 6 female rats was fed control diet. Body-weight gains of
    female rats fed 20% liquid smoke were significantly lower than those
    of control animals. Food intake of rats in the 2-thiouracil group was
    significantly depressed. At termination, plasma studies, including
    determination of thyroid hormones, were carried out. There were some
    minor changes in plasma constituents in the high-dose group; however,
    thyroid hormones were not affected, except in the 2-thiouracil group.
    No significant histological changes were observed in any of the organs
    of the rats fed LS 1 (Parish, 1986d).

         In a study comparing the toxicity of two types of smoke
    flavourings, 5 groups of 20 male and 20 female rats were maintained
    for 90 days on diets containing 0, 0.3, or 3.0% of a basic aqueous
    condensate or 0.025 or 0.25% of an oil soluble smoke flavouring
    derived from the tar-like precipitate of a freshly prepared
    condensate. Ten animals died during the course of the study, but death
    was not related to the administration of test compound. Weight gain
    was similar in test and control animals. No haematological changes
    were observed at week 5 or at termination of the study. At termination
    of the study, organ-weight analyses and gross pathological and
    histopathological studies of the major organs showed no
    compound-related effects (Eschenberger, 1963).

    Long-term studies

         No information available.

    Observations in man

         No information available.

    LIQUID SMOKE FLAVOURINGS (TARRY EXTRACT)

    BIOLOGICAL DATA

    Biochemical aspects

    No information available.

    Toxicological studies

         Many of the studies summarized here were carried out using a
    liquid smoke flavouring preparation derived from the tarry fraction of
    smoke condensate prepared from alder wood (Alnus incana). Diethyl
    ether was used for the extraction process. Seventy-seven substances
    have been identified in the smoke flavouring. The major components
    were phenols and their alkylated derivatives (alkylated phenols,
    guaicols, and siringols). PAHs were present below 10 ppb, and no
    nitrosamines were detected (limit of detection, 5 ppb) (Miler, 1978).

    Special studies on mutagenicity

         A smoke flavouring derived from hickory was negative in the Ames
    Salmonella microsome plate test, with or without activation, at
    levels of 0.005 µl to 10 µl per plate (Jagannath & Brusick, 1979).

         The smoke flavouring preparation from alder wood was tested using
    Salmonella typhimurium strains TA98, TA100, TA1535, and TA1537, with
    or without metabolic activation, at dose levels ranging from 28.3 to
    2292 µg per plate. No mutagenic effects were observed (Jensen, 1986).

    Acute toxicity1
                                                                        

    Species                 Sex     Route     LD50           Reference
                                              (g/kg b.w.)
                                                                        

    Mouse (Swiss albino)    M       oral      2.8            Fitko, 1979a
                            F                 2.3

    Rat (Wistar)            M                 4.0            Fitko, 1979b
                            F                 3.5

    Pig                     M & F             3.6            Fitko, 1979c
                                                                        

    1    The tested substance was the smoke flavouring preparation from
         alder wood.

    Short-term studies

    Mice

         Swiss albino mice were maintained for 90 days on diets containing
    0, 0.1, 0.5, 1.0, or 2.0% of the smoke flavouring derived from alder
    wood. The number of mice was 35 females and 25 males in each of the
    control and high-dose groups and 12-14 of each sex in each of the
    intermediate groups. Eleven animals died during the course of the
    study. Deaths occurred in all groups and were not compound-related.
    Body-weight gains in all but the lowest-dose animals were increased
    over controls, and were associated with increased food intake.
    Haematologic studies, urinalysis, and clinical chemistry studies at
    weeks 6 and 13 showed no significant differences between the control
    and high-dose animals. At autopsy, organ-weight analysis showed
    significant increases in weights of the kidneys and thyroid.
    Histological examination showed compound-related effects in the liver,
    kidneys, gastrointestinal tract, lungs, and spleen in the test groups,
    with the exception of the 0.1% group. The effects that were noted
    included parenchymatous degeneration of the renal tubules of the
    kidneys, oedema and hyperaemia of the spleen, parenchymatous
    degeneration of the liver, and oedema and hyperaemia in the mucosa of
    the stomach and the submucosa of the small intestine (Fitko, 1979d).

    NOTE:     The pathology terms used in this study and all the other
              studies authored by Fitko that are summarized in this
              monograph are those of the author. No criteria were
              available for the terms used in the pathology reports, hence
              complete evaluations of these studies are difficult because
              of the lack of definitions of the terms used.

    Rats

         Four groups, each containing 15 male and 15 female Wistar rats,
    were fed diets containing 0, 0.1, 0.3, or 1.01 of the tarry fraction
    from a smoke flavouring preparation for 13 weeks. Weight gain was
    significantly decreased in the high-dose group. At autopsy,
    organ-weight analysis showed a significant increase in thyroid weight
    of both male and female rats in the high-dose group. The thyroid
    glands of animals in the high- and medium-dose groups were noticably
    enlarged, compared to the controls. Histological examination indicated
    dose-related changes in the thyroid, characterized by areas of atrophy
    and large acini, a condition similar to colloidal goitre in man
    (Hercules, 1977).

         Five groups of equal numbers of male and female Wistar rats were
    maintained for 90 days on diets containing 0, 0.1, 0.5, 1.0, or 2.0%
    of the smoke flavouring derived from alder wood. The number of rats
    was 35 of each sex in each of the control and high-dose groups and 15
    of each sex in each of the intermediate groups. Haematologic, serum,
    clinical chemistry, and urine analyses carried out at 0, 6, and 13
    weeks showed no significant differences between the control and
    high-dose animals. Body-weight gains were similar in test and control
    animals, except for the high-dose group in which the females showed a
    decrease, and males an increase, in body-weight gain compared to
    controls. Food consumption was higher in the high-dose group than in
    the controls. At autopsy, organ-weight analysis showed significant
    increases in liver, kidney, and heart weights, except in animals in
    the low-dose group. Macroscopic examination of the organs showed the
    following compound-related changes; catharrhal inflamation of the
    mucosa and small intestine and hyperaemia of the mucosa of the stomach
    in dosed groups, with the exception of the low-dose group.
    Histopathological examination revealed, in addition to changes in the
    gastrointestinal tract, changes in the liver (parenchymal degeneration
    and hyperaemia), kidneys (parenchymal degeneration of the renal
    glomerulus), and lungs (hyperaemia). No significant effects were
    observed in the low-dose group (Fitko, 1979e).

    Pigs

         Five groups, each containing 3 male and 3 female "large white"
    pigs, were administered in the diet the smoke flavouring derived from
    alder wood at dose levels equivalent to 0, 200, 600, 1000, or
    1400 mg/kg b.w./day for 90 days. Each dose was given to the animals in
    2 portions. Body-weight gains showed significant decreases with
    increasing dietary concentration of the test compound. However, this
    was associated with decreased food intake. Haematologic, serum,
    clinical chemistry, and urine analyses of the control and high-dose
    animals at weeks 0, 6, and 13 showed no consistent compound-related
    effects, with the exception of a dose-related increase in serum
    bilirubin. A liver function test (BSF retention) of high-dose and
    control animals at week 13 showed no significant differences. At
    autopsy, organ weight analysis showed no compound-related effects
    except for the liver and the uterus. In the case of the liver, there
    was a significant decrease in weight in the lower-dose groups, and an
    increase in weight in the high-dose groups. The uterus showed a
    dose-related increase in weight. Macroscopic examination of the organs
    showed inflammation of the stomach and intestines, as well as changes
    in the kidneys and lungs. Histological examination of the tissues
    showed dose-related changes in the gastrointestinal tract, ranging

    from inflammation to prominent proliferation of lymphoid tissue into
    the submucosa of the stomach and small intestines and oedema of the
    mucosa in these organs. In addition, parenchymal degeneration of the
    liver and renal glomeruli and hyperaemia in the liver and lungs were
    observed. At the lowest dose fed, the only significant effects were
    inflammation of the stomach mucosa and lymphoid cell infiltration in
    the submucosa (Fitko, 1979f).

         In another study, four groups of 3 male and 3 female pigs were
    fed diets containing the smoke flavouring derived from alder wood for
    52 weeks at doses equivalent to 0, 200, 600, or 1000 mg/kg b.w./day.
    The daily dose was given to the test animals in two portions. There
    were no compound-related deaths during the course of the study. There
    was a dose-related decrease in mean body weights, and this was
    associated with decreased feed efficiency. Haematologic studies, serum
    clinical chemistry studies, and urinalyses were performed at 0, 1, 3,
    6, and 12 months on animals in the control and high-dose groups. Serum
    cholesterol and direct bilirubin showed dose-related increases in the
    test animals. There were sporadic variations in haematologic
    parameters and certain serum enzymes (aspartate aminotransferase and
    alkaline phosphatase). Liver function tests (BSP retention) at 0, 6,
    and 12 months were similar at all test periods. At termination of the
    study, organ-weight analysis showed significantly increased relative
    weights of the liver and brain and decreased relative weights of the
    thymus in the high-dose group, and increased relative kidney weights
    in the intermediate group. Macroscopic examination of the organs
    showed inflammation of the gastrointestinal tract of all treated
    animals, and liver necrosis in the high-dose group. Histopathology
    studies showed changes ranging from inflammation of the
    gastrointestinal tract in all animals to liver necrosis in the
    high-dose group. Effects on the kidneys, which included parenchymal
    degeneration, hyperaemia, and necrosis, appeared to be most severe in
    the high-dose group (Fitko, 1979g).

    Long-term studies

    Mice

         Groups of equal numbers of male and female Swiss albino mice were
    maintained for 16 months on diets containing smoke flavouring derived
    from alder wood at dose levels equivalent to 0, 150, 920, or
    1650 mg/kg b.w./day for females, and 0, 150, 760, or 1720 mg/kg b.w./
    day for males. The control group and highest-dose group consisted of
    60 mice of each sex, and the intermediate doses consisted of 25 mice
    of each sex. One hundred thirty-three of the 340 animals died during
    the course of the study. There were no significant differences between
    survival in control and test groups. Gross pathology of 68/133 mice
    autopsied showed inflammation of the gastrointestinal tract in 34

    animals and liver degeneration in 33 animals. These and other
    pathological changes were not dose related. No rumours were found in
    animals dying before sacrifice. Weight gain was slightly reduced in
    the high-dose group compared to the control and appeared to be
    associated with a decrease in feed efficiency. Haematology, serum,
    clinical chemistry studies, and urinalysis were carried out at 1, 3,
    6, 12, and 16 months. There were no significant compound-related
    changes. At termination of the study, organ-weight analysis showed an
    increase in liver weights and a decrease in thyroid weights. Changes
    in weights of other organs appeared to be sporadic and not dose
    related. Histological studies showed a dose-related increase in the
    frequency of fatty degeneration of the liver, with liver necrosis in
    the highest-dose group. Inflammation of the gastrointestinal tract was
    observed in all groups. Other effects observed in the high-dose group
    included lymphocytic infiltration in the adrenals and lungs and
    hypertrophy of mucosa in the uterus. The number of test animals with
    rumours did not differ significantly from the number in the control
    group (the total number of animals examined in each group ranged from
    19 to 24) (Fitko, 1979h).

    Rats

         Groups of rats were maintained for 24 months on test diets
    containing smoke flavouring derived from alder wood at dose levels
    equivalent to 0, 48, 260, or 630 mg/kg b.w./day for males and 59, 360,
    or 850 mg/kg b.w./day for females. The high-dose group consisted of
    118 male and 117 female rats, the intermediate groups of 45 male and
    45 female rats, and the control group of 117 males and 118 females.
    During the study 201 animals died; mortality was equally distributed
    among all groups. Lung disease followed by inflammation of the
    gastrointestinal tract and degeneration of the liver were listed as
    causes of death. No rumours were reported. Weight gain was lower in
    all test groups compared to controls and was associated with decreased
    feed efficiency. Haematology, serum clinical chemistry analyses, and
    urinalyses were carried out at 1, 3, 12, 18, and 24 months. There were
    sporadic changes in various parameters, but these were not dose
    related, and were within normal ranges. At termination of the study
    organ-weight analysis showed weight increases of the uterus, thyroid,
    lungs, and kidneys in females in the high-dose group, and in the
    hypophysis and thymus of males in the high-dose group. Histological
    studies of the tissues showed inflammation of the gastrointestinal
    tract, and necrosis of the liver and kidney in the two high-dose
    groups. There were no significant differences in the frequency
    distribution of any type of observed neoplasms between test and
    control animals. The data were evaluated using the methods described
    by Peto (1974) (Fitko, 1979i).

    Observations in man

         No information available.

    COMMENTS

         Toxicity data are available for liquid smoke flavourings derived
    from the aqueous and tar fractions.

         The fractions were not mutagenic in bacterial systems
    (Salmonella typhimurium and E. coli strains), with or without
    activation. However, one liquid smoke preparation (aqueous fraction)
    which was inactive in the Salmonella test caused a significant
    increase in mutation frequency in the TK6 lymphoblast system. This
    compound was not positive in a lung rumour biossay system utilizing
    newborn mice, although toxic effects were observed.

         Short-term studies were available for several types of liquid
    smoke flavourings (aqueous extract). The major adverse effects noted
    were significant increases in liver and kidney weights of the test
    animals. In one study, these changes were shown to be reversible. Only
    minor histological changes were reported in these animals. Minor
    changes in serum chemistry have also been reported. In these studies
    the no-effect level ranged from 0.25% to 3% of the liquid smoke
    preparation in the diet. It was also noted that in some studies there
    were significant increases in thyroid weights in the high-dose groups.
    However, changes in serum levels of T3 and T4 were reported in
    only one study.

         Almost all the studies with smoke flavourings (tarry extract)
    relate to a single product. This product was tested in a number of
    species (mice, rats, and pigs). The tests include long-term studies in
    mice and rats. The major effects observed in the high-dose groups, in
    all species tested, were decreased body-weight gain and increased
    liver and kidney weights. Histopathological changes were primarily
    related to the gastrointestinal tract, liver and kidneys that included
    inflammation of the gastrointestinal tract and necrosis of the liver
    and kidney. These effects were observed in the groups fed the highest
    levels of the test material, and were minimal or absent in the
    low-dose groups. In lifetime feeding studies in mice and rats, there
    were no compound-related increases in rumour incidence or type in the
    treated animals. However, the lack of any neoplastic foci of the liver
    in rats or mice is unusual in studies of this duration. A complete
    evaluation of these studies is difficult because of the lack of
    definition of the terms used in the pathology report. High levels of
    smoke flavourings (tarry extracts) in the diets of rats also resulted
    in a significant increase in thyroid weights, and in one case
    histological changes were observed.

    EVALUATION

         The Committee viewed the use of smoke flavourings generically,
    keeping in mind that smoke flavourings are a replacement for
    traditional smoking practices, and as such they represent a definite
    improvement in that a large number of potentially toxic compounds are
    eliminated in their production. A similar view has been expressed by
    the Council of Europe (Resolution AP185/2).

         The Committee concluded that so complex a group of products might
    not be amenable to the allocation of an ADI, and that smoke
    flavourings of suitable specifications could be used provisionally to
    flavour foods traditionally treated by smoking. However, as the safety
    data for these products were limited, new or novel uses of smoke
    flavourings should be approached with caution.

         The Committee concluded that detailed information on the
    production and composition of smoke flavourings is required, and that
    it would be desirable to have further safety studies on a well-defined
    spectrum of smoke flavourings.

    REFERENCES

    Braun, A.G., Busby, W.F. Jr., Jackmen, J., Halpin, P.A., &
    Thilly, W.G. (1986). Commercial hickory smoke flavouring is a human
    lymphoblast but does not induce lung adenoma in newborn mice.
    Food Chem. Tox. (in press).

    Eschenberger, A.B. (1963). Biological examination of food additives.
    Unpublished report No. A-675 from Engineering Experiment Station,
    Georgia Institute of Technology, Atlanta, CA, USA. Prepared for Dunn
    Laboratories, Atlanta, GA, USA.

    FCT (1965). Summaries of toxicological data, artificial smoke
    flavourings, II. "Liquid smoke flavour". Safety evaluation by oral
    administration to rats for 90 days. Fd. Cosmet. Tox., 3, 145-147.

    Fitko, R. (1979a). Smoke Extract Flavour. Study of acute oral toxicity
    in the mouse. Unpublished report No. 2 from the Institute of
    Veterinary Sciences, Olsztyn, Poland. Submitted to WHO by Broste
    Industri Copenhagen, Denmark.

    Fitko, R. (1979b). Smoke Extract Flavour. Study of acute oral toxicity
    in the rat. Unpublished report No. 3 from the Institute of Veterinary
    Sciences, Olsztyn, Poland. Submitted to WHO by Broste Industri
    Copenhagen, Denmark.

    Fitko, R. (1979c). Smoke Extract Flavour. Study of acute oral toxicity
    in the pig. Unpublished report No. 4 from the Institute of Veterinary
    Sciences, Olsztyn, Poland. Submitted to WHO by Broste Industri
    Copenhagen, Denmark.

    Fitko, R. (1979d). Smoke Extract Flavour. Ninety-day oral toxicity
    study in the mouse. Unpublished report No. 5 from the Institute of
    Veterinary Sciences, Olsztyn, Poland. Submitted to WHO by Broste
    Industri Copenhagen, Denmark.

    Fitko, R. (1979e). Smoke Extract Flavour. Ninety-day oral toxicity
    study in the rat. Unpublished report No. 6 from the Institute of
    Veterinary Sciences, Olsztyn, Poland. Submitted to WHO by Broste
    Industri Copenhagen, Denmark.

    Fitko, R. (1979f). Smoke Extract Flavour. Ninety-day oral toxicity
    study in the pig. Unpublished report No. 7 from the Institute of
    Veterinary Sciences, Olsztyn, Poland. Submitted to WHO by Broste
    Industri Copenhagen, Denmark.

    Fitko, R. (1979g). Smoke Extract Flavour. Twelve-month oral toxicity
    study in the pig. Unpublished report No. 10 from the Institute of
    Veterinary Sciences, Olsztyn, Poland. Submitted to WHO by Broste
    Industri Copenhagen, Denmark.

    Fitko, R. (1979h). Smoke Extract Flavour. Sixteen-month oral toxicity
    study in the mouse. Unpublished report No. 8 from the Institute of
    Veterinary Sciences, Olsztyn, Poland. Submitted to WHO by Broste
    Industri, Copenhagen, Denmark.

    Fitko, R. (1979i). Smoke Extract Flavour. Twenty-four oral toxicity
    study in the rat. Unpublished report No. 9 from the Institute of
    Veterinary Sciences, Olsztyn, Poland. Submitted to WHO by Broste
    Industri, Copenhagen, Denmark.

    Hercules (1977). Herocsef smoke extract flavour. Three-month oral
    toxicity study in Wistar rats. Unpublished report No. CL76,107,1325,
    from Hercules Powder Co. Ltd., London, England. Submitted to WHO by
    Unilever Research and Engineering, Shambrook, England.

    Jagannath, D.R. & Brusick, D.J. (1979). Mutagenicity evaluation of
    natural hickory smoke flavour - Code 402. Unpublished report No. 20988
    from Litton Bionetics, Inc., Kensington, MD, USA, for Stange Company,
    Chicago, IL, USA.

    Jensen, J.C. (1986), Ames Salmonella/microsome assay with Scansmoke
    concentrate (SEF). Unpublished report from Scantox, Denmark. Submitted
    to WHO by Broste Industri, Copenhagen, Denmark.

    Lijinsky, W. & Shubik, P. (1965). Summaries of toxicological data,
    artificial smoke flavour, the detection and estimation of polycyclic
    hydrocarbons in liquid smoke. Fd. Cosmet. Tox., 3, 145-147.

    Miler, K. (1978). Smoke extract flavours, description of the product
    to be tested in toxicity studies on the mouse, the rat and dog.
    Unpublished report prepared by the Polish Heat and Fat Research
    Institute, Warsaw, Poland. Submitted to WHO by Broste Industri,
    Copenhagen, Denmark.

    Mortelmans, K.E. & Eckford, S. (1980). Microbial mutagenesis testing
    of substances, compound report F76-054, F76-055 and F6-076.
    Unpublished report No. LSV-6909 from S.R.I. International.

    Parish, W.E. (1986a). Ingestion toxicity of liquid smokes, Part 2.
    Four-week rat feeding trial with Oss liquid smoke. Unpublished report
    No. PES 82,1022 from Unilever Research and Engineering. Submitted to
    WHO by Unilever Research and Engineering, Shambrook, England.

    Parish, W.E. (1986b). Ingestion toxicity of liquid smokes, Part 3.
    Four-week rat paired feeding trial with Oss and Zwan liquid smoke.
    Unpublished report No. PES 85,1005 from Unilever Research and
    Engineering. Submitted to WHO by Unilever Research and Engineering,
    Shambrook, England.

    Parish, W.E. (1986c). Ingestion toxicity of liquid smokes.
    Thirteen-week rat feeding trial with Oss Liquid smoke. Unpublished
    report No. D86/057 from Unilever Research and Engineering. Submitted
    to WHO by Unilever Research and Engineering, Shambrook, England.

    Parish, W.E. (19868). Ingestion toxicity of liquid smokes, Part 1.
    Four-week rat feeding trial with Zwan liquid smoke. Unpublished report
    No. PCW 79,1294 from Unilever Research and Engineering. Submitted to
    WHO by Unilever Research and Engineering, Shambrook, England.

    Peto, R. (1974). Guidelines on the analysis of rumour rates and death
    rates in experimental animals (editorial). Brit. J. Cancer,
    29, 101-103.

    Raltech (1981). Ames Salmonella/microsome mutagenicity tests, a smoke
    product. Unpublished report No. 862001 from Raltech Scientific
    Services, Madison, WI, USA for Red Arrow Products Co.,
    Manitowoc, WI, USA.

    WARF (1961). Subacute toxicity study on a liquide smoke flavour.
    Unpublished report from Wisconsin Alumini Research Foundation.
    Submitted to WHO by Red Arrow Products, Co., Manitowoc, WI, USA.

    WRC (1963). Liquid smoke flavour. Safety evaluation by oral
    administration to rats for 90 days. Unpublished report from Woodard
    Research Corporation for the Griffith Lab. Inc.
    




    See Also:
       Toxicological Abbreviations
       SMOKE FLAVOURINGS (JECFA Evaluation)