The Committee was aware that canthaxanthin has been used as a
    direct food additive, as a feed additive, and as an
    orally-administered pigmenting agent for human skin in both
    pharmaceutical and cosmetic applications. It was evaluated for
    acceptable daily intake at the tenth and eighteenth meetings of the
    Committee (Annex 1, references 13 and 35).

         The present Committee was asked to review the safety of
    canthaxanthin as a food additive because of reports of crystalline
    deposition in the retina during its use as an orally-administered skin
    pigmenting agent. The dose that resulted in this deposition was within
    the ADI established by the Committee at its eighteenth meeting
    (Annex 1, reference 35).

         Canthaxanthin is a diketo carotenoid pigment with an orange-red
    colour. It occurs in the edible mushroom, chanterelle (Cantharellus
    cinnabarinus), in the plumage and organs of flamingoes, the scarlet
    ibis (Guara rubra), and the roseate spoonbill (Ajaja ajaja), and
    in various crustacea and fish (trout, salmon) (Haxo, 1950; Fox,
    1962a, b; Thommen & Wackernagel, 1963).


    Biochemical aspects

    Absorption, distribution, and excretion


         Adult rats were fed a range of oral doses of canthaxanthin (not
    specified) for 13 and 20 weeks, respectively. Canthaxanthin
    accumulated in fat and some organs, notably the liver and spleen. Only
    slight depletion of canthaxanthin from fat occurred over a period of 1
    month, indicating very slow elimination from this tissue
    (Hoffmann-La Roche, 1986).

         Groups of rats were given daily doses of 0.6, 6, or 60 mg
    canthaxanthin/kg b.w. daily for five weeks. Highest organ
    concentrations were found in the liver and spleen, the tissue levels
    corresponding to the three dietary dose levels being 0.9, 12, and
    125 µg/g liver, and 2.6, 50, and 67 µg/g spleen, respectively. Levels
    in other organs were much lower (0.2-1.5 µg/g at the highest dose
    level). After discontinuing administration of canthaxanthin, tissue
    levels in the adrenals and small intestine fell to one-quarter to
    one-third of their original levels over a period of 2 weeks (adrenals)
    or 1 month (intestine) (Hoffmann-La Roche, 1986).

         When rats were fed daily doses of 50-60 mg canthaxanthin/kg b.w.
    for 9 weeks, the concentration in the eye remained at approximately
    0.1 µg/g with no further accumulation. After administration of doses
    of 1.2, 2.0, 3.4, 5.6, 9.8, 16.7, or 28.4 mg canthaxanthin/kg feed
    (equivalent to 1.4 mg/kg b.w./day at the top-dose level) for 20 weeks,
    maximum concentrations in the eye were found to be about 0.01 µg/g;
    the residual levels in the eyes fell to 0.002 µg/g over a 4-week
    period after removal of canthaxanthin from the diet
    (Hoffmann-La Roche, 1986).


         Canthaxanthin earlier was reported not to exhibit provitamin A
    activity (Hoffmann-La Roche, 1966). However, in recent studies in
    chickens, canthaxanthin was shown to be converted to vitamin A in
    small amounts. Groups of 15 broiler chickens, 37 days old, were given
    diets containing 8.9, 18, or 35 mg canthaxanthin/kg feed along with 0,
    300, or 600 i.u. vitamin A/kg feed. At each dietary level of vitamin
    A, canthaxanthin caused a dose-related increase in plasma and liver
    concentrations of both vitamin A and canthaxanthin. The lowest level
    of canthaxanthin (8.9 mg/kg feed), in the absence of dietary vitamin
    A, yielded higher plasma and liver vitamin A levels than did a diet
    containing 600 i.u. vitamin A/kg in the absence of canthaxanthin
    (Hoffmann-La Roche, 1986).

         Distribution studies with radiolabelled canthaxanthin (dose not
    specified) fed to laying hens resulted in deposition of 30-40% of the
    dose in the egg yolk and 7% in body tissues; 42-54% was excreted.
    During a 10-day period of depletion after withdrawal of canthaxanthin,
    the body stores were mobilized and excreted via eggs and excrement.
    The canthaxanthin concentration in the fat was reported to be low and
    to remain constant (Hoffmann-La Roche, 1986).


         The distribution of canthaxanthin in the tissues was investigated
    in dogs which had received 50, 100, or 250 mg/kg b.w./day for 52
    weeks, corresponding to total doses of 200, 400, or 1,100 g
    respectively. The highest mean tissue concentration was seen in
    adipose tissue (24 µg/g in the top-dose group). Relatively high
    concentrations were also seen in adrenals (15.1 µg/g), skin
    (9.62 µg/g) and liver (8.1 µg/g) in the low-dose group. The total
    amount of canthaxanthin extracted from 8 eyes of treated dogs was
    0.1-0.4 pg, but ophthalmological examinations were not performed, so
    it was not possible to determine whether crystalline deposits had
    formed (Hoffmann-La Roche, 1986).


         A pharmacokinetic study was performed in which plasma levels of
    canthaxanthin were measured at intervals after multiple dosing of
    human volunteers. Ten subjects were each given 1 mg canthaxanthin 6
    times a day for 5 days, corresponding to a total dose of 30 mg. A
    further ten subjects received 8 mg canthaxanthin 6 times a day for 2
    days, total dose 96 mg. Blood samples were taken at the start and at
    12-hour intervals for 8 days, and canthaxanthin concentrations were
    determined by HPLC. The elimination half-life was calculated as 4.5
    days in each group and the proportion of the dose absorbed was
    estimated to be 12 and 9%, respectively. The calculated steady state
    plasma concentrations of canthaxanthin after daily ingestion of 6 mg
    (6 times 1 mg) or 48 mg (6 times 8 mg) was calculated as 1843 or
    10,346 µg/l, respectively (Kubler, 1986).

    Toxicological studies

    Special studies on carcinogenicity


         When given orally, canthaxanthin exhibited no promotional
    activity in mice treated dermally with dimethylbenz(a)anthracene or
    benzo(a)pyrene (Mathews-Roth, 1982, Santamaria et al., 1982).

         Oral doses of 6680 mg/kg b.w./day gave some protection against
    the skin carcinogenicity of regular exposure to UV radiation
    (Mathews-Roth, 1982).


         Groups of rats were treated with canthaxanthin at doses of
    0, 250, 500, or 1000 mg/kg/day for 104 weeks. A summary report stated
    that some changes in blood chemistry parameters were observed, mainly
    at the intermediate and high-dose levels, relating to a (non-
    specified) liver effect. No changes in tumour incidence were seen
    at up to 78 weeks of treatment, but after 104 weeks there was a
    slight, but non-dose related, increase in benign liver tumours in
    treated females. Full details of this study were not available; the
    report stated that the significance and nature of these findings will
    be evaluated in an additional study aimed at determining the no-effect
    level (Hoffmann-La Roche, 1986).

    Special studies on ocular toxicity

         Preliminary results of studies in rabbits on the ocular effects
    of canthaxanthin (dose not specified) did not reveal any deposits in
    the retina, but small alterations (prolongation) in dark adaptation
    were observed. The significance of these results is controversial
    (Hoffmann-La Roche, 1986).

    Special studies on reproduction

         A three-generation study using 0 or 1000 ppm canthaxanthin in the
    diet revealed no adverse effects in any generation (Hoffmann-La Roche,

         No adverse effects on reproductive function were observed when
    doses of 0, 250, 500, or 1000 mg canthaxanthin/kg b.w./day were given
    throughout 3 generations (Hoffmann-La Roche, 1986).

    Acute toxicity

    Species      Route     LD50               Reference
                           (mg/kg b.w.)

    Mouse        oral      10,000             Hoffmann-La Roche, 1966

    Short-term studies


         Groups of three male and three female dogs received 0, 100, or
    400 mg/kg b.w. canthaxanthin daily for 15 weeks. No significant
    effects were noted on body weight of control or test groups or on
    their general health (Hoffmann-La Roche, 1966).

    Long-term studies

         See also "Special studies on carcinogenicity"


         In a preliminary outline report of an 80-week study in mice in
    which the animals received 0, 250, 500, or 1000 mg canthaxanthin/kg
    b.w./day, it was stated that no signs of systemic toxicity and no
    changes of any tumour incidence were seen that could be related to
    treatment (Hoffmann-La Roche, 1986).


         Groups of 25-30 male and female rats received 0, 0.5, 2, or 5%
    canthaxanthin in their diets for 93-98 weeks. No adverse effects were
    noted on food consumption or weight gain. Mortality and tumour
    incidence were not increased (Hoffmann-La Roche, 1966).


         Oral doses of 0, 50, 100, or 250 mg canthaxanthin/kg b.w./day for
    52 weeks were well tolerated and no adverse effects attributable to
    treatment were observed (Hoffmann-La Roche, 1986).

    Observations in man

         Six out of a group of 42 patients with a history of urticaria
    suffered a recurrence of their symptoms within 23 hours after an oral
    challenge with 410 mg canthaxanthin taken as three divided doses over
    3 hours (Juhlin, 1981).

         The ingestion of doses of about 30-120 mg canthaxanthin daily
    (approximately 0.4-1.7 mg/kg b.w./day) for 3 months to several years
    in medicinal or oral sun-tanning preparations has been associated with
    a retinopathy in some individuals characterised by glistening, golden
    crystals in the inner layers of the retina, up to 10-14 µm in size
    (Boudreauit et al., 1983; Cottin et al., 1984; Ros et al.,
    1985). The crystalline deposits occur mainly in a ring between 5° and
    10° around the fovea, less numerous in the fovea and rarely in the
    foveola (Cortin et al., 1982). Occasionally, deposits have been
    reported nasally of the disc (Metge et al., 1984) or scattered in
    the posterior fundus (cited in Daicker et al., 1987) and in one case
    only in the periphery of the fundus in the inner layer of a
    retinoschisis (Cortin et al., 1982). A total dose of 75-178 g
    canthaxanthin has been found to be effective in 50% of subjects and
    numerous cases have now been described (Franco et al., 1985;
    Hennekes et al., 1985; McGuiness & Beaumont, 1985; Meyer et al.,
    1985; Philipp, 1985; Saraux & Laroche, 1983; Weber et al., 1985a,b;
    Weber & Goerz, 1985).

         In most cases, pigment deposition is not associated with any
    detectable functional changes, but occasionally there have been
    complaints of dazzle or blurred vision (Cortin et al., 1984;
    Hennekes et al., 1985; Philipp, 1985); visual field defects have
    been described in only one report (Ros et al., 1985). The EOG is
    normal or subnormal and dark adaptation may be delayed; scotopic
    vision after exposure to glare is reduced while the ERG is normal or
    with b-wave changes (BoudreauIt et al., 1983; Merge et al., 1984;
    McGuiness & Beaumont, 1985; Weber et al., 1985b; Hennekes et al.,
    1985; Philipp, 1985).

         Twenty-five patients were re-examined 2-10 months after therapy
    with canthaxanthin and ß-carotene was discontinued. Dark adaptation
    and ERG had normalized, but the crystalline retinopathy and pigment
    epithelial defects showed no signs of reversibility (Weber &
    Goerz, 1986).

         Increased susceptibility to retinal deposits has been associated
    with age and a number of clinical factors, including focal
    epitheliopathy, ocular hypertension, and possibly co-administration of
    5-carotene (Cortin et al., 1984).

         In a biostatistical evaluation of 253 cases having received
    treatment with canthaxanthin, of whom 33 (15%) had retinal deposits,
    the median yearly dose in subjects free from visible retinal deposits
    was 5.3 g, whereas the corresponding figure for the group with pigment
    deposits was 14.4 g. The lowest dose at which deposits were recorded
    was 7 g canthaxanthin/year and no retinal deposits were found in
    patients receiving less than 30 mg/day (Hoffmann-La Roche, 1986).

         The eyes of a female patient, aged 72 years, who had retinal
    deposits and who had died under anaesthesia, were examined by light
    and electron microscopy, and the extracted pigment was examined by
    mass and proton magnetic resonance spectroscopy. There were red,
    birefringent crystals in the inner layers of the entire retina,
    particularly large and numerous perifoveally where they were
    clinically visible. The crystals were located in the inner neuropil
    where an atrophy of the inner parts of the Muller cells was observed.
    The compound was identical to canthaxanthin and the retina contained
    up to 42 µg/g tissue besides a minor amount of other carotenoids. Of
    the other ocular tissue, only the ciliary body contained measureable
    amounts of canthaxanthin (Daicker et al., 1987).

         Canthaxanthin was measured at autopsy in the tissues of 38
    people, aged 22 to 96 years, none of whom were known to have received
    canthaxanthin therapeutically or in sun-tanning preparations. The
    tissues examined were mesenteric and sub-cutaneous fat, skin, liver,

    spleen, and blood serum. The highest concentrations were found in
    omentum and sub-cutaneous fat (mean concentrations, 0.2 and 0.3 µg/g,
    respectively). The mean concentrations in other tissues were:
    liver, 0.08 µg/g; skin and spleen, 0.04 µg/g; and serum, 0.024 µg/ml
    (Hoffmann-La Roche, 1986).

         Fat samples from mesenterium and omentum and a liver sample were
    taken at autopsy from a 71-year-old woman who had died of bronchial
    carcinoma. The patient had previously ingested approximately 45 mg
    canthaxanthin/day for four years (total dose approximately 65 g) in a
    sun-tanning preparation. The concentrations of canthaxanthin in
    omentum and mesenteric fat were 270 µg/g and 158 µg respectively;
    lower levels of 5 µg/g were found in the liver (Hoffmann-La Roche,

         Biopsy samples of orange-coloured fat (omentum) were obtained
    from two patients undergoing surgery. In one case the woman had taken
    a total dose of about 6 g canthaxanthin in a tanning preparation
    during 1983/84 and had stopped this intake 1-1.5 years before the
    biopsy; fat and serum canthaxanthin levels were 49 µg/g and 69 µg/l,
    respectively. In the second case the patient had taken a total dose of
    approximately 16 g over 2.5-3 years and the concentration in omentum
    was 34 µg/g (Hoffmann-La Roche, 1986).

         The Committee viewed the observation of crystal deposition in the
    retina as new data that warranted a complete review of this compound.
    Most of the new data from animal studies were available only as
    summaries, and therefore could not be used as a basis for evaluation.
    The Committee noted that, although pigment accumulation in the eye had
    been demonstrated analytically following oral administration of
    canthaxanthin to rats and dogs, no ophthalmoscopy had been carried out
    and no animal model for the human condition had been developed. In
    man, however, an estimate was made of the minimal level of exposure
    resulting in fundal pigment deposition in the retina.

         In the light of all these considerations, the previous ADI was
    made temporary and reduced.

         When setting an ADI, the Committee does not consider therapeutic
    use, which is a matter for clinical judgement. The cosmetic use of
    canthaxanthin as an orally-administered skin pigmenting agent was not
    anticipated when the ADI was established at the eighteenth meeting.
    Therefore, it is not included in the temporary ADI established at the
    present meeting; the temporary ADI applies only to the food and feed
    additive uses of canthaxanthin.


    Estimate of temporary acceptable daily intake for man

         0-0.05 mg/kg b.w. (based on the minimal effect level for pigment
    deposition in the retina of human subjects, to which a 10-fold safety
    factor was applied).

    Further work or information


         1. Details of the long-term studies in mice and rats for which
    summary data were submitted, including ophthalmological data where

         2. Clarification of the factors that influence deposition in the
    eye, including the establishment of the threshold dose, the influence
    of dose and duration of exposure, the reversibility of pigment
    accumulation, and the investigation of potential animal models.

         3. Clarification of whether pigment deposition is causally
    related to impaired ocular function.


    Boudreauit, G., Cortin, P., Corriveau, L.A., Rousseau, A.P.,
    Tardif, Y., & Malenfant, M. (1983). La rétinopathie ŕ la
    canthaxanthine: 1. Etude clinique de 51 consommateurs.
    Can. J. Ophtalmol., 18, 325-328.

    Cortin, P., Corriveau, L.A., Rousseau, A.P., Tardif, Y., Malenfant,
    M., & Boudreauit, G. (1982). Maculopathie en paillettes d'or.
    Can. J. Ophtalmol., 17, 103-106.

    Cortin, P., Boudreauit, G., Rousseau, A.P., Tardif, Y., &
    Malenfant, M. (1984). La rétinopathie ŕ la canthaxanthine;
    2. Facteurs prétdisposants. Can. J. Ophtalmol., 19, 215-219.

    Daicker, B., Schiedt, K., Adnet, J.J., & Bermond, P. (1987).
    Canthaxanthin retinopathy; Light and electron microscopic
    investigation and physicochemical analysis. Graefe's Arch. Clin. Exp.
    Ophthalmol. (in press).

    Fox, D.L. (1982a). Comp. Biochem. Physiol., 6, 1.

    Fox, D.L. (1982b). Comp. Biochem. Physiol., 6, 305.

    Franco, J.L., Adenis, J.P., Mathon, C., & Lebraud, P. (1985).
    Un nouveau cas de maculopathie en paillettes d'or. Bull. Soc.
    Ophtal. Fr., 85, 1035-1037.

    Haxo, F. (1950). Botan. Gaz., 122, 228.

    Hennekes, R., Weber, U., & Kuchle, H.J. (1985). Ueber
    canthaxanthin-schaden der netzhaut. Z. prakt. Augenheilk., 6, 7-9.

    Hoffmann-La Roche (1966). Canthaxanthin. Unpublished report submitted
    to WHO by F. Hoffman-La Roche & Co., Basle, Switzerland.

    Hoffmann-La Roche (1986). Canthaxanthin. Unpublished report submitted
    to WHO by F. Hoffman-La Roche & Co., Basle, Switzerland.

    Juhlin, L. (1981). Recurrent urticaria; clinical investigation of 330
    patients. Br. J. Dermatol., 104, 369-381.

    Kubler, W. (1986). Biokinetic evaluation of canthaxanthin plasma
    levels after multiple doses of 1 mg and 8 mg canthaxanthin.
    Unpublished report submitted to WHO by F. Hoffman-La Roche & Co.,
    Basle, Switzerland.

    Mathews-Roth, M.M. (1982). Antitumour activity of beta-carotene,
    canthaxanthin and phytoene. Oncology, 39, 33-37.

    McGuiness R. & Beaumont, P. (1985). Gold dust retinopathy after the
    ingestion of canthaxanthin to produce skin-bronzing. Med. J. Aust.,
    143, 622-623.

    Metge, P., Maudirac-Bonnefoy, C., & Bellaube, P. (1984). Théraurismose
    rétinienne ŕ la canthaxanthine. Bull. Mem. Soc. Fr. Ophtalmol.,
    95, 547-549.

    Meyer, J.J., Bermond, P., Pournaras, C., & Zoganas, L. (1985).
    Canthaxanthin. Langzeiteinnahme und sehfunktion beim menschen.
    Dtsch. Apoth. Zeitung, 125, 1053-1057.

    Philipp, W. (1985). Carotinoid-einlagerungen in der netzhaut.
    Klin. Mbl. Augenheilk., 187, 439-440.

    Ros, A.M., Leyon, H., & Wennersten, G. (1985). Crystalline retinopathy
    in patients taking an oral drug containing canthaxanthin.
    Photodermatol., 2, 183-185.

    Santamaria, L., et al. (1982). Médecine, biologie et environnement,
    10, 359.

    Saraux, H. & Laroche, L. (1983). Maculopathie ŕ papillottes d'or aprčs
    absorption de canthaxanthine. Bull. Soc. Opht. Ft., 83, 1273-1275.

    Thommen, H. & Wackernagel, H. (1963). Isolation and identification of
    canthaxanthin in the lesser flamingo (Phoenicolnaias minor).
    Biochim. Biophys. Acta., 69, 387-396.

    Weber, U. & Goerz, G. (1985). Augenscheden durch carotinoid-einnahme.
    Dtsch Aerzteblatt., 82, 181-182.

    Weber, U., Goerz, G., & Hennekes, R. (1985a). Carotinoid-retinopathie:
    I. Morphologische und funktionelle befunde. Klin. Mbl. Auggenheilk.,
    186, 507-511.

    Weber, U., Hennekes, R., & Goerz, G. (1985b). Carotinoid-retinopathie:
    II. Elektrophysiologische befunde bei 23 carotinoid-behandelten
    patlenten. Klin. Mbl. Auggenheilk., 186, 507-511.

    Weber, U. & Goerz, G.  (1986). Carotinoid-retinopathie. III.
    Reversibilitat. Klin. Mbl. Auggenheilk, 188, 20-22.

    See Also:
       Toxicological Abbreviations
       Canthaxanthin (WHO Food Additives Series 26)
       Canthaxanthin (WHO Food Additives Series 35)
       Canthaxanthin (WHO Food Additives Series 44)
       CANTHAXANTHIN (JECFA Evaluation)