IPCS INCHEM Home


        INTERNATIONAL PROGRAMME ON CHEMICAL SAFETY

        WORLD HEALTH ORGANIZATION



        TOXICOLOGICAL EVALUATION OF CERTAIN
        VETERINARY DRUG RESIDUES IN FOOD



        WHO FOOD ADDITIVES SERIES 41





        Prepared by:
          The 50th meeting of the Joint FAO/WHO Expert
          Committee on Food Additives (JECFA)



        World Health Organization, Geneva 1998




    RECOMBINANT BOVINE SOMATOTROPINS (addendum)

    First draft prepared by
    Professor F.R. Ungemach
    Institute of Pharmacology, Pharmacy and Toxicology
    Veterinary Faculty, University of Leipzig, Leipzig, Germany
    and
    Dr N.E. Weber
    Food and Drug Administration, Center of Veterinary Medicine
    Rockville, MD, USA

    1.   Explanation
    2.   Biological data
         2.1  Use of antibiotics
         2.2  Concentrations of bovine somatotropin and insulin-like
              growth factor in tissues and milk
              2.2.1  Tissues
              2.2.2  Residues of insulin-like growth factor in milk
              2.2.3  Assays
              2.2.4  Bioavailability and bioactivity of insulin-like
                     growth factor  residues in milk
         2.3  Expression of lentiviruses and prion proteins
              2.3.1  Somatotropins and the immune system
              2.3.2  Effect of bovine somatotropin on expression of
                     retroviruses
              2.3.3  Effect of recombinant bovine somatotropin on prion
                     proteins
         2.4  Cow's milk and insulin-dependent type I diabetes mellitus in
              childhood
    3.   Comments
    4.   Evaluation
    5.   References

    1.  EXPLANATION

         The four analogues of bovine somatotropins, somagrebove,
    sometribove, somavubove, and somidobove, that are produced by
    recombinant DNA techniques were evaluated by the Committee at its
    fortieth meeting (Annex 1, reference 104). At that time, the Committee
    allocated an ADI 'not specified' for recombinant bovine somatotropins
    (rbSTs) because of the lack of activity of orally administered rbSTs
    and insulin-like growth factor I (IGF-I) and the low concentrations
    and non-toxic nature of the residues of these compounds, even after
    very high doses, resulting in an extremely large margin of safety for
    humans consuming dairy products from rbST-treated cows. The Committee
    concluded that use of these drugs according to good practice in
    veterinary medicine did not represent a dietary hazard to human health
    and that there was no need to specify a numerical ADI. Accordingly,
    MRLs 'not specified' were recommended for milk and edible tissues of
    cattle. The Codex Alimentarius Commission, when considering adoption
    of these recommended MRLs at its twenty-second session in 1997,
    postponed a decision pending re-evaluation of rbSTs by the Expert

    Committee to consider scientific information that had become available
    since its previous evaluation.

         Information was submitted by organizations and individuals
    relating to the following concerns:

    *    increased use of antibiotics, with a higher rate of violative
         drug residues in milk due to a possible increase in the incidence
         of mastitis in rbST-treated cows,

    *    the possibility that increased concentrations of IGF-I in the
         milk of rbST-treated cows increase cell division and the growth
         of tumours in humans,

    *    the potential effect of rbST on the expression of certain viruses
         in cattle, particularly the retroviruses,

    *    the possibility that the incubation period of bovine spongiform
         encephalopathy (BSE) is shortened by an IGF-I-induced increase in
         the production of pathogenic prion proteins, and

    *    the possibility that early exposure of human newborns to milk
         from rbST-treated cows increases the risk for developing
         insulin-dependent diabetes mellitus.

    2.  BIOLOGICAL DATA

    2.1  Use of antibiotics

         The induction by rbST of an increased incidence of mastitis and
    somatic-cell count in the milk of treated cows was not reviewed by the
    Committee at its fortieth meeting, as these effects on animal health
    were considered outside its terms of reference. At the present
    meeting, the Committe considered results reported in the literature
    and the results of a programme to monitor the effects of sometribove
    (Posilac(R)) on mastitis and animal health in the United States. The
    Committee confirmed that these effects of rbST on animal health and
    the resulting treatment per animal with any medication are issues
    outside the terms of reference of the Committee. The results of the
    monitoring programme with regard to the percentage of milk discarded
    because of drug residues above the approved limit, as a consequence of
    antibiotic use, were, however, considered by the Committee. The
    programme was initiated by the US Food and Drug Administration at the
    time sometribove was approved, in November 1993 and began to be
    commercialized, in February 1994. The objectives of this programme
    were to determine whether the incidence of masitits and antibiotic use
    were manageable under actual conditions of use and whether the
    labelling of rbSTs was adequate (US Food and Drug Administration,
    1996). The programme was designed to address the following areas
    (Collier, 1996):

    *    the incidence of mastitis and responses related to herd health
         (not within the terms of reference of the Committee),

    *    treatment of 28 herds with rbST-treated cows with any medication
         (not within the terms of reference of the Committee), and

    *    the percent of milk discarded because of drug residues above the
         allowed limit in key dairy states representing at least 50 % of
         US milk production.

    The programme was closely monitored by the US Food and Drug
    Administration and performed according to the sponsor's standard
    operating procedures for quality assurance. The US Food and Drug
    Administration (1996) confirmed that the integrity of the findings was
    acceptable and that the records and analyses showed excellent
    fidelity.

         The programme for tracking milk residues in key dairy states
    before and after the approval of sometribove was designed to reveal
    whether any increase in drug residues above the allowed limit in milk
    was associated with an increased frequency of use of antibiotics for
    mastitis and other health problems in rbST-treated herds of dairy cows
    (Veenhuizen  et al., 1996). The results of the milk monitoring
    programme for the two years before commercial use of sometribove
    (1992-93) were compared with those for the two years after its launch
    on the market (1994-95). Residues in milk were traced according to the
    National Drug Residue Milk Monitoring Program in which the contents of
    all milk tanker trucks are sampled routinely. The data set represented
    more than 50% of the total US milk supply. The data were analysed
    quarterly by comparing the milk discarded before and after the launch
    of sometribove.

         As seen in Figure 1, no change was observed in 1994 after
    sometribove was approved. The average percent of milk discarded was
    0.06% in 1992 and 1993 and 0.07% in 1994 after the launch; in 1995,
    the number of violations increased slightly but significantly to
    0.09%. This increase coincided, however, with a change in the
    screening procedures in most states to include more sensitive tests.
    Veenhuizen  et al. (1996) submitted data to the US Food and Drug
    Administration on 17 May 1996 showing that in New York State there had
    been no significant change in the rate of discard of milk in the two
    years after approval of Posilac in comparison with the two years
    before approval, but that in that State the same testing protocol had
    been used throughout the four-year period. The values were 0.062%
    before approval and 0.064% afterwards. The company that launched
    sometribove reported that rbST was purchased by nearly 37% of the
    farms in the State, and these farms represented about 50% of the
    State's cows. These reports indicate that:

    *    no product-related increase in residues above the approved limit
         had occurred after commercialization of sometribove;

    *    the rate of positive results was even slightly lower than those
         for antibiotics in grade A milk in the United States; and

    *    use of sometribove will have no effect on the safety of milk and
         dairy products due to violative drug residues resulting from a
         slightly higher rate of medication of rbST-treated animals, as
         measured in the milk monitoring programme.

    FIGURE 1

    It was concluded that use of rbST would not result in a higher risk to
    human health due to the use of antibotics to treat mastitis and that
    the increased potential for drug residues in milk could be managed by
    practices currently in use by the dairy industry and by following the
    directions for use (US Food & Drug Administration, 1996).

    2.2  Concentrations of bovine somatotropin and insulin-like growth 
    factor in tissues and milk

    2.2.1  Tissues

         The concentrations of bST and IGF-I were measured in tissues of
    cattle that had been treated with a 14-day sustained-release product
    containing the natural variant of rbST, somavubove (Choi  et al., 
    1997), in two experiments. In the first experiment, three groups of 12
    beef cattle (except at the low dose, to which only six animals were
    exposed) with an average weight of 450 kg were treated by subcutaneous
    injection for 20 weeks. The controls received no treatment or vehicle,
    one group received 250 mg rbST each week, and another received 500 mg

    rbST at two-week intervals. The controls and animals at the high dose
    were further divided into groups receiving low- and high-energy feed.
    The stated total doses were 5 and 10 g rbST, respectively; however,
    the dose for both groups calculated from the stated regimens would be
    5 g. Two weeks after the final treatment, the animals were slaughtered
    and muscle samples were obtained and stored at -20°C for analysis. In
    the second experiment, four groups of beef cattle were used: a control
    group which received no drug or vehicle, a group receiving a
    sustained-release low dose (0.42 mg/kg bw; 0.03 mg/kg bw per day), a
    group receiving an intermediate dose (0.84 mg/kg bw; 0.06 kg bw per
    day), and one receiving a high dose (1.26 mg/kg bw; 0.09 mg/kg bw per
    day). The treated groups were given the drug by subcutaneous injection
    every two weeks for 24 weeks, to give total doses of 2.3, 4.5, and 6.8
    g, respectively. Two weeks after the final treatment, the animals were
    slaughtered, and samples of muscle, kidney, liver, and fat were taken
    and stored at -20°C.

         The frozen samples were assayed for bST and IGF-I residues by
    radioimmunoassay (RIA) in which 5 g of tissue were extracted in
    acidœethanol for muscle and acetic acid for kidney, liver, and fat.
    The RIA procedures involved standard double-antibody techniques and
    iodinated tracers. The detection limit for the assays (the amount that
    could be distinguished from zero concentration with 95% confidence)
    was 0.17 ng/g bST and 0.61 ng/g IGF-I. The coefficients of variation
    for the two assays were 6% or less, and the recoveries from liver,
    kidney, and fat were 64% for bST and 84% for IGF-I; similar recoveries
    were obtained in muscle samples. A summary of the results is given in
    Table 1. The authors concluded that two weeks after administration of
    two doses of rbST for extended times, the tissue concentrations of bST
    and IGF-I were not significantly different from those in untreated
    animals.

    2.2.2  Residues of insulin-like growth factor in milk

         Extensive information on the residues of bST and IGF-I in the
    milk of rbST-treated cows was evaluated by the Committee at its
    fortieth meeting. IGF-I is a normal but highly variable constituent of
    bovine milk, the concentration depending on the state of lactation,
    nutritional status, and age. Over an entire lactation, the IGF-I
    concentrations range from 1 to 30 ng/ml, with the highest
    concentrations in colostrum and a constant decline thereafter.
    Multiparous animals have higher concentrations of IGF-I in milk than
    primoparous cows (Burton  et al., 1994). Bulk milk from cows not
    given rbST had IGF-I concentrations of 1œ9 ng/ml (Juskevich & Guyer,
    1990), and the fortieth Meeting cited an average value of 3.7 ng/ml
    for untreated cows. In milk from rbST-treated cows, the concentrations
    of IGF-I ranged from 1 to 13 ng/ml in most studies and were about
    25-70% greater than those in untreated animals (Burton  et al., 
    1994). The fortieth meeting reported an average IGF-I concentration of
    5.9 ng/ml, and the increase was significant, even though most of the
    concentrations were < 10 ng/ml.


        Table 1.  Concentrations of bovine somatotropin (bST) and insulin-like growth factor (IGF-I) in tissues after treatment with 
    recombinant bovine somatotropin

                                                                                                              

    First experiment

    Tissue     No. of    Concentration (ng/g; mean ± SD)
               samples                                                                                        
                         Controls                       Low dose                    High dose
                         bST             IGF-I          bST           IGF-I         bST            IGF-I
                                                                                                                
    Muscle     12        1.9 ± 1.8       88 ± 21        1.5 ± 1.6     130 ± 25      3.3 ± 2.2      110 ± 32
                                                                                                              

                                                                                                                                             

    Second experiment

    Tissue     No. of    Concentration (ng/g; mean ± SD)
               samples                                                                                                                       
                         Controls                       Low dose                    Intermediate dose            High dose
                                                                                                                                                
                         bST             IGF-I          bST           IGF-I         bST            IGF-I         bST            IGF-I
                                                                                                                                             

    Muscle     5         3.4 ± 1.5       45 ± 6.5       4.9 ± 1.5     35 ± 15       3.8 ± 2.0      40 ± 5.1      1.5 ± 0.86     55 ± 19
    Fat        4         5.1 ± 1.7       210 ± 85       9.3 ± 5.2     200 ± 65      4.8 ± 1.9      200 ± 53      11 ± 12        340 ± 230
    Liver      4         5.2 ± 0.59      350 ± 23       3.6 ± 1.7     390 ± 130     5.4 ± 1.2      380 ± 170     4.6 ± 2.0      290 ± 88
    Kidney     4         3.6 ± 1.1       910 ± 130      4.5 ± 1.6     1000 ± 140    4.5 ± 1.8      820 ± 120     3.9 ± 0.94     980 ± 220
                                                                                                                                             
    

         Since the original work was reviewed, few additional data have
    appeared in the literature or in reports made available by the
    sponsors. The manufacturer of Posilac(R) - previously identified as
    sometribove, which is a form of rbST approved in a number of countries
    - submitted additional information. In a study designed to determine
    the concentrations of IGF-I in retail milk samples, the concentrations
    in milk specifically labelled as coming from cows that had not been
    treated with bST were compared with those in unlabelled milk. While
    the sponsor assumed that all of the unlabelled milk was from cows that
    had been treated with bST, the extent to which this was true was not
    ascertained. The study was conducted under US Food and Drug
    Administration GLP regulations (21CFR 58). The labelled and unlabelled
    status was determined of 127 of 129 retail milk samples collected as
    2% fat cow's milk in 51 retail outlets in 34 cities in Wisconsin,
    Minnesota, and Iowa, USA, from four goats and 125 cows, and the
    samples were analysed by RIA for IGF-I. For 78 samples, the farmer had
    certified that rbST had not been used. As rbST is not approved for use
    in goats in the United States, it is assumed that the goat-milk
    samples were from untreated animals. The results are shown in Table 2
    (Eppard  et al., 1994). The values for IGF-I are not unlike those
    reported in the monograph prepared at the fortieth meeting of the
    Committee. Although the contribution of milk from rbST-treated cows to
    the unlabelled milk is unknown, the results indicate that the IGF-I
    concentrations in retail milk did not increase in the first year after
    the launch of rbST.


    Table 2.  Concentrations of insulin-like growth factor in milk
    certified by farmers as coming from cows not treated with
    recombinant bovine somatotropin (labelled) and from unlabelled milk

                                                                    

                     Insulin-like growth factor (ng/ml)
                                                                    
                     Labelled           Unlabelled 
                     (n = 78)           (n = 45)
                                                                    

    Raw mean         4.3 ± 0.09         4.5 ± 0.12
    Logea            1.47 ± 0.044       1.55 ± 0.031     p = 0.01769
    Antilog (95%     4.4 (4.0-4.7)      4.7 (4.4-5.0)
    confidence 
    interval)
                                                                    

    a Least square means adjusted for state where purchased and dairy


    2.2.3  Assays

         Because of variations in the IGF-I values reported in different
    studies, questions have been raised in submissions to the Committee
    about their accuracy. Lower values were found in some studies because
    the assay used involved acid-ethanol extraction; an assay involving
    acidic gel filtration has been reported to allow better recovery. In a
    comparison of the two assays for determining IGF-I in pre- and
    post-partum mammary secretions, the acid-ethanol assay was found to
    result in a 24 ± 6.6% underestimate in comparison with acid gel
    filtration (Vega  et al., 1991). Use of the acidœethanol assay to
    recover 125I-IGF-I in bovine serum and colostrum, which also contain
    binding proteins, gave values of 86 ± 6% and 88 ± 7% recovery by the
    antibody, respectively (Hadsell, 1991). The difference is due to the
    fact that the binding protein with affinity for IGF-I competes with
    the antibody used in the RIA procedure; however, gel filtration
    removes the hormone that is extensively dissociated by the acid common
    to both assays. The studies suggest that the acidœethanol assay
    probably results in 15-25% underestimates of the concentrations in
    milk and plasma. The relative difference in milk will probably not
    affect a decision, as the concentrations are low, although higher
    control values may affect the inferences. Although incomplete removal
    of IGF-binding proteins or differences in the source of standards
    might affect the reported results, the Committee considered these
    factors immaterial.

    2.2.4  Bioavailability and bioactivity of insulin-like growth factor
    residues in milk

         At its fortieth meeting, the Committee concluded that many of the
    physiological effects of rbSTs are mediated by bovine IGF-I, which is
    structurally identical to human IGF-I. It noted that there is
    substantial endogenous synthesis of IGF-I, mainly in the liver but
    also in human milk, saliva, and pancreatic secretions. It further
    concluded that IGF-I has no bioactivity when administered orally to
    normal and hypophysectomized rats at doses up to 2 mg/kg bw per day,
    as dietary IGF-I is degraded by digestive enzymes and is not active in
    the upper gastrointestinal tract.

         Concern has been expressed that widespread use of rbSTs in dairy
    production would lead to a sustained increase in the concentrations of
    IGF-I in bulk cow's milk and that, if IGF-I survives digestion, the
    greater exposure of consumers would cause adverse health effects
    (Hansen  et al., 1997). For a quantitative assessment of risk, the
    slight increases in the concentration of IGF-I in milk from
    rbST-treated cows should be compared with the physiological variations
    in the concentrations of this growth factor during lactation (see
    section 2.2.2) and with the concentrations in human breast milk, in
    the secretions of the gastrointestinal tract, and in serum.

         The concentrations of IGF-I are 8-28 ng/ml in colostrum and
    5-10 ng/ml thereafter (Zumkeller, 1992; Burton  et al., 1994),
    indicating that breast-fed newborns are usually exposed to IGF-I at
    concentrations equal to or higher than those in milk from rbST-treated
    cows. Assuming a daily intake of 1.5 L of milk from rbST-treated cows
    with an average IGF-I concentration of 6 ng/ml, the amount of IGF-I
    ingested would be 9000 ng/day. The additional daily ingestion of IGF-I
    over that of people drinking milk from untreated animals, with an
    average IGF-I concentration of 4 ng/ml or 6000 ng/1.5 L, would be 3000
    ng. The slightly increased IGF-I concentrations would contribute to
    the endogenous concentrations of IGF-I in the gastrointestinal tract
    of consumers; however, the main site of IGF-I production in animals
    and humans is the liver. It is also produced in the human
    gastrointestinal mucosa and is found in saliva, bile, and pancreatic
    juice (Olanrewaju  et al., 1992). The average concentrations of IGF-I
    in the five human gastrointestinal secretions were 0.9 nmol/L in
    saliva, 3.5 nmol/L in gastric juice, 24.6 nmol/L in jejunal chyme, 3.6
    nmol/L in pancreatic juice, and 0.9 nmol/L in bile (Chaurasia  et 
     al., 1994). On the basis of a molecular mass of 7.5 kDa for IGF-I
    (Zumkeller, 1992) and the volume of each of the fluids produced
    (Vander et al., 1990), the total volume of IGF-I emptying into the
    gastrointestinal tract is 383 000 ng/day. Table 3 shows that the
    amount of endogenous IGF-I emptied into the gastrointestinal tract
    daily is more than (383 000/9000) 42 times greater than the amount
    present in 1.5 L of milk from rbST-treated cows. The 9000 ng value is
    2.3% of the estimated daily gastrointestinal secretion of IGF-I in
    adults. The additional daily ingestion of 3000 ng IGF-I over that in
    milk from untreated animals thus represents 0.78% of the
    gastrointestinal secretion.


    Table 3. Concentration of insulin-like growth factor (IGF-I) in 
    gastrointestinal tract secretions

                                                                     

    Secretion            Volume      Concentration      Total IGF-I
                         (ml/day)a   (average; ng/ml)   secreted (ng)
                                                                     

    Jejunal chyme         1500           184.5            276 750
    Pancreatic juice      1500            27.0             40 500
    Gastric juice         2000            26.2             52 400
    Bile                   500             6.8              3 400
    Saliva                1500             6.8             10 200

                                                                     

    Adapted from Baumann (1995)
    a From Vander et al. (1990)

         In contrast to the conclusion of the Committee at its fortieth
    meeting, that IGF-I is completely and rapidly degraded in the
    gastrointestinal tract, milk-borne IGF-I may escape digestion by
    proteases and therefore be bioactive in the intestine (Hansen
     et al., 1997) or even be absorbed as intact peptide into the
    systemic circulation (Epstein, 1996). In a study designed to
    investigate the potential therapeutic use of IGF-I and to determine
    whether oral formulations were feasible, the degradation of
    125I-IGF-I was determined in various segments of the
    gastrointestinal tract of rats  in vivo and  in vitro (Xian
     et al., 1995). The extent of degradation was measured by one of
    three methods: receptor binding, immunoprecipitation, and
    trichloracetic acid precipitation. Two segments of the duodenum,
    ileum, or whole stomach and part of the colon were ligated in an
    anaesthetized male Sprague-Dawley rat that had been fasted for 24 h. A
    bolus of labelled IGF-I (8.6 ng/ml in 0.2% bovine serum albumin in
    saline) was injected into each segment and incubated for various times
    up to 1 h. The reactions were stopped, and the flushed luminal
    contents were examined for intact IGF-I. In a parallel set of
    experiments, the flushed luminal contents from each of the four gut
    segments were used as a source of degredation enzymes  in vitro. The
    results are shown in Table 4. The most rapid degradation occurred in
    the duodenum and ileum and in their contents  in vitro, followed by
    the stomach and then by the colon. In all cases the values seen
     in vitro were equal to or greater than those seen  in vivo. 

         The authors also examined the effectiveness of slowing the
    degradation rate of IGF-I in the gut by protecting the molecule in
    several ways. Casein at a concentration of 10 mg/ml conferred > 90%
    protection on stomach contents in both the trichloracetic acid and
    receptor assays; in duodenal flushings, casein conferred 80%
    protection against IGF-I degradation in the trichloracetic acid assay
    but only 36% protection in the receptor assay at a maximal casein
    concentration of 40 mg/ml. The half-life of IGF-I in the upper
    gastrointestinal tract in the receptor assay increased from 2-3 min in
    the absence of casein to 35 min in its presence. While these results
    appear to demonstrate significant protection by casein at a
    concentration similar to those in milk, the authors acknowledge that
    the observed effects could be explained by simple competition by the
    additional proteins for degradation by proteases in the segments. The
    results demonstrate that biological receptor-stimulating  activity,
    which is the best indicator of biological activity, is dramatically
    reduced even with good protection from protease activity. The authors
    noted with regard to milk residues of IGF-I that 'the protective
    effect of casein makes irrelevant the argument that human saliva
    contains IGF-I at concentrations greater than the quantities that
    would be consumed in milk. As the IGF-I produced by salivary glands is
    free IGF-I, without protective effect of casein, it is unlikely to
    survive digestion.' (Hansen  et al., 1997). That argumentation
    neglects the following facts:


        Table 4.  Half-lives of intact 125I-insulin-like growth factor (IGF-I) in ligated Sprague-Dawley rat gut 
    segments in vivo and in the flushed contents in vitro

                                                                                                              

    Test for            Half-life (min)
    intact IGF-I                                                                                              
                        Duodenum/ileum                Stomach                       Colon
                                                                                                              
                        In vivo        In vitroa      In vivo        In vitro       In vivo        In vitro
                                                                                                              

    Trichloracetic      2              2              8              50             38             > 60

    Antibody binding    2                             5                             33

    Receptor binding    2              2              2.5            3              16             ND
                                                                                                              

    ND, not determined
    a Antibody and membrane receptor values reported as receptor binding
    

    *    Saliva is not the only source of IGF-I in the gastrointestinal
         tract: most is secreted into the gut, and the high concentration
         in intestinal chyme indicates that  IGF-I is secreted in
         substantial amounts by the mucosa throughout the gastrointestinal
         tract (Olanrewaju  et al., 1992; Chaurasia  et al., 1994).

    *    Casein has a flexible structure and is readily degraded in the
         stomach and small bowel (Xian  et al., 1995). Thus, the
         protective effect is present only in the upper gastrointestinal
         tract.

    *    The half-life of IGF-I in the intestine in the presence of casein
         is only 35 min (Xian  et al., 1995). Therefore, less than 5% of
         an initial dose of IGF-I will survive for more than 2 h during
         passage through the upper gastrointestinal tract.

    *    In the presence of casein ingested with milk, endogenous IGF-I in
         the gastrointestinal tract will also be protected.

    Therefore, even if casein has a limited protective effect, the amount
    of bioactive IGF-I ingested with milk from rbST-treated cows would
    still be negligible.

         Because of the protective effect of casein, some IGF-I might
    escape digestive degradation and be absorbed intact. The absorption of
    large (1 mg/kg) oral doses of 125I-labelled recombinant human IGF-I
    was studied in fasted adult rats, with trichloracetic acid
    precipitation of plasma proteins. The baseline bioavailability of the
    administered IGF-I was 9.3% of the dose, but this was was increased by
    co-administration of 4 mg/kg aprotinin (47%) and 10 mg/kg casein
    (67%). RIA of the plasma confirmed the bioavailability of IGF-I in
    this model, and the administered radiolabel was found in the form of
    high-molecular-mass complexes (Kimura  et al., 1997). It should be
    noted, however, that the receptor assay, which is the most accurate,
    was not used.

         The relatively good bioavailability of intact IGF-I in this adult
    rat model is in contrast to the lack of bioactivity of orally
    administered IGF-I in adult animals (Annex 1, reference 104) and to
    the results of studies with neonatal animals which have an incomplete
    mucosal barrier and reduced intestinal proteolytic activity (Burrin,
    1997). Studies in neonatal rats and piglets indicated that although
    30% of an orally administered dose of 125I-IGF-I can be recovered in
    the intestinal mucosa there is limited absorption into the peripheral
    circulation (Phillips  et al., 1995; Donovan  et al., 1997). When
    suckling transgenic rats ingested 1000-fold higher concentrations of
    des(1,3) human IGF-I, no des(1,3)-IGF-I was detected in the plasma of
    their pups (Burrin, 1997). Furthermore, in newborn calves and piglets
    given large doses of IGF-I in milk replacers, no substantial increase
    in the plasma concentration of this growth factor was found (Donovan
     et al., 1997; Hammon & Blum, 1997; Houle  et al., 1997). In one
    study with newborn calves fed milk replacer, a small amount of orally

    administered 125I-IGF-I was detected in plasma (Baumrucker  et 
     al., 1992); however, the increase was observed only three days after
    administration and in only three of six animals. Even in newborns,
    therefore, IGF-I is absorbed to only a small extent, and absorption is
    unlikely in adults.

         Furthermore, the amount absorbed should be compared with the
    normal concentrations of IGF-I in human serum, which show considerable
    variation with age. The values are lowest in infants under two years
    of age, then increase steadily to reach a maximum in late puberty, and
    afterwards decrease to the adult values (Table 5). Assuming a blood
    volume that is 5% of the body weight (Ganong, 1971), the serum load of
    IGF-I can be calculated to be 50 000 ng in a 15-kg child, 714 000 ng
    in a 60-kg adult, and 1 220 000 ng in a 50-kg teenager. The total
    IGF-I production in adults has been estimated at 107 ng per day
    (Guler  et al., 1989). These amounts should be compared with the 9000
    ng IGF-I in 1.5 L of milk, which constitutes only 0.09% of the daily
    IGF-I production. Since only one-third of the concentrations in milk
    can be attributed to IGF-I due to rbST treatment and only a small
    amount if any will be absorbed, the milk-borne IGF-I that reaches the
    systemic circulation is negligible and this small amount is
    immediately sequestered by unsaturated binding proteins.


    Table 5. Concentrations (ng/ml) of insulin-like growth factor (IGF-I) 
    in human plasma

                                                                        

    Age                           Males                Females
                                                                      
                                  Mean     Range       Mean     Range
                                                                        

    0-2 years                      42      14-98        56     14-238
    3-5 years                      56     59-210        84     21-322
    6-10 years                     98     28-308       182     56-364
    Before puberty > 10 years     126     84-182       182     70-280
    Early puberty                 210    140-240       224     84-392
    Late puberty                  364    224-462       434    224-686
    Adult > 23 yeras              112     42-266       140     56-308
                                                                        

    From Schaff-Blass et al. (1984)

         Concern has been expressed about the possible adverse effects on
    the health of consumers exposed to increased concentrations of IGF-I
    in milk from rbST-treated cows (Epstein, 1996; Hansen  et al., 1997).
    The most important potential adverse effects of IGF-I arise from the
    fact that it is a mitogen for a number of cell types and has been
    associated with the growth of tumours including those of the colon,
    breast, and lung and osteosarcoma (Pines  et al., 1985; Macaulay,
    1992; National Institute of Health, 1995). The mitogenic effect could

    also result in proliferative reactions locally in the gut. Thus,
    orally administered IGF-I increased the cellularity of the intestinal
    mucosa of rats  in vivo (Olanrewaju  et al., 1992) and increased the
    rate of proliferation in cultures of human duodenal epithelial crypt
    cells (Challacombe & Wheeler, 1994). Since IGF-I receptors can be
    detected throughout the epithelium of the intestine, with a high
    density in the colon (Laburthe  et al., 1988), and the incidence of
    colorectal cancer is increased in acromegalic patients who have
    excessively high concentrations of free IGF-I in their plasma (Ezzat &
    Melmed, 1991), concern has been expressed that increased
    concentrations of milk-borne IGF-I may increase the risk for colon
    cancer. Although the normal biological effects of IGF-I mean that it
    could promote the growth of tumours, this hazard would become a risk
    only if there were adequate exposure of consumers to IGF-I. Since
    exposure to IGF-I in milk from rbST-treated cows is negligible when
    compared with endogenous IGF-I production, it is extremely unlikely
    that IGF-I residues cause any systemic or local mitogenic reaction.

    2.3  Expression of lentiviruses and prion proteins

    2.3.1  Somatotropins and the immune system

         Somatotropin enhances the immune system in many species,
    including cattle (Comens-Keller  et al., 1995). The primary effect
    appears to be altered responsiveness of the immune system, although
    substantive evidence of this effect is lacking (Burton  et al., 
    1994), and information on changes in cytokine concentrations or
    secretion and on their binding sites are needed in order to define the
    nature of the immune-enhancing effects of somatotropins. The studies
    reported in the literature differ with regard to the source of
    somatotropin, the treatment schedule, and the age of the animals. The
    differences between findings  in vitro and  in vivo may be due to
    the release  in vivo of mediators such as IGFs and cytokines, which
    are not present  in vitro (Kelley, 1989). A better understanding of
    somatotropin-mediated immune enhancement as a homeorhetic regulator of
    the overall health and disease resistance of animals is needed.
    Lymphocytes from rbST-treated cows have a greater average maximum
    lymphoblastogenic response to rbST than to other mitogens around the
    time of parturition (Comens-Keller  et al., 1995), and this effect
    might prevent mastitis and the other infectious diseases that occur
    during this period of immunosuppression.

    2.3.2  Effect of bovine somatotropin on the expression of retroviruses

         Concern has been expressed that the immunomodulatory effect of
    bST might affect retroviral expression in treated animals and thus
    cause resurgence of latent retroviral and lentiviral infections in the
    ruminant population and the presence of these viruses in somatic cells
    in milk. The concern is based largely on a review by Lerondelle et al.
    (1994), who discussed the evidence for induction of these viruses in
    small ruminants by steroid hormones and the evidence for induction of
    lentiviruses in other species by hormones including growth hormone and

    IGF-I, and an unpublished study by Lerondelle  et al. (1996), who
    investigated the effects of rbST on the expression of caprine
    arthritis encephalitis virus (CAEV) in goats. This virus belongs to
    the group of lentiviruses which, like maediœvisna, can infect small
    ruminants.

         Ruminant lentiviruses are of interest for at least three reasons.
    First, they may cause disease in persons who drink milk. Second, if
    use of rbST increases the prevalence of ruminant viruses, by the
    presence of rbST itself or by the action of IGF-I, the small
    additional amounts of growth hormone in the milk of treated cows could
    also affect the retroviruses that attack the human immune system,
    HIV-1 and HIV-2. Finally, the severity or kinetics of expression of
    the disease in ruminants might be increased.

         The study of Lerondelle and coworkers (1996) attempted to address
    the question of whether rbST increases the expression of CAEV. Viral
    expression was measured by assaying reverse transcriptase activity in
    milk cells, clinical examination of the udders and joints of the
    animals at the beginning and end of the study, and evidence of
    infection in an immunodiffusion assay. Twelve pregnant Saanen goats
    that were seronegative for CAEV were given an intramammary injection
    of monocytes infected  in vitro with the Cork strain of CAEV at the
    time of drying off. Seven weeks after giving birth, four goats were
    treated daily with 5 mg/goat rbST (sometribove), a second group was
    treated daily with 10 mg/goat thyroxine, and the control group was
    untreated. The compounds were administered in suspension in sterile
    water for 30 days, followed by a 45-day observation period. Milk
    samples were taken to measure reverse transcriptase activity on days
    7, 14, 21, and 28 of treatment and three times during the observation
    period. Examination of the udders and joints and immunodiffusion tests
    were carried out at the beginning and end of the study. Milk
    production and milk-cell counts were evaluated every two days. The
    occurrence of CAEV and the onset of effects are shown in Table 6. More
    positive cultured cells were seen in the controls than after treatment
    with either hormone. Perhaps the most striking effect is the lack of
    increase in the rate of infectivity and even a suggestion of a
    decrease after treatment with rbST, particularly after the first milk
    sampling.

         When the results were expressed as a ratio of transcriptase
    activity to the number of cells considered to contain virus (Table 7),
    there was no correlation between the effect of rbST or thyroxine and
    the activity of reverse transcriptase in the milk samples. In fact,
    there was no evidence of increased transcriptase activity in any
    group. In particular, even though the group receiving rbST had a lower
    initial rate of infection than the other two groups, including the
    controls, the rate of infection was not increased, as measured by the
    number of positive cultures or increased transcriptase activity. The
    authors interpreted their results as showing a tendency to increased
    viral expression with increased milk production. The results for rbST
    were, however, biased by the fact that only two animals were included
    in the final evaluation. The authors concluded that, owing to the


        Table 6.  Onset of appearance of cytopathic effect (in days) for each of 10 milk samplings from groups of four goats given no treatment or
    thyroxine or recombinant bovine somatotropin (rbST), as a function of the period of hormonal treatment

                                                                                                                                         

    Treatment        Milk sampling
                                                                                                                                         
                     Before treatment                                                 During treatment                  After treatment
                                                                                                                                         
                     1                  2              3               4              5         6              7        8        9
                                                                                                                                         

    Control          4                  8              6               10             6         6              6        8        10
                     4                  6              6               6              8         6              6        6        -
                     4                  6              4               6              4         4              6        4        4
                     6                  8              10              6              10        6              -        -        -
    Mean ± SD (n)    6 ± 1.91 (12)                     6.4 ± 1.71 (15)                          7 ± 2.39 (8)

    Thyroxine
                     8                  6              10              4              6         6              -        -        6
                     6                  4              4               4              4         4              4        4        ND
                     4                  4              4               4              4         4              6        4        ND
                     4                  8              6               4              4         4              6        4        ND
    Mean ± SD (n)    5.67 ± 2.06 (12)   4.53 ± 0.92 (15)               4.25 ± 0.71 (8)

    rbST             8                  -              -               -              -         -              -        -        -
                     4                  8              8               -              8         -              -        -        4
                     ND                 -              -               -              -         -              -        -        -
                     4                  4              4               4              4         4              4        4        4
    Mean ± SD (n)    5.71 ± 2.41 (7)                   4.8 ± 1.79 (5)                           4 (4)
                                                                                                                                         

    ND, not determined
    

    Table 7. Number of samples found to contain virus by the reverse 
    transcriptase-positive culture ratio in groups of four goats given 
    no treatment, thyroxine, or recombinant bovine somatotropin (rbST)

                                                                       

    Treatment      Before         During        After          Total
                   treatment      treatment     treatment
                                                                       

    Control        3/6            4/7           3/4            10/17

                   3/6            3/5           1/3            7/14
                   6/6            4/5           5/6            15/17
                   4/4            2/8           3/6            9/18
    Total          16/22          13/25         12/19          41/66

    Thyroxine      0/3            2/4           0/3            2/10
                   4/4            5/7           6/6            15/17
                   6/6            8/8           6/6            20/20
                   6/6            8/8           4/6            18/20
    Total          16/19          23/27         16/21          55/67

    rbST           2/4            0/6           0/1            2/11
                   6/6            7/8           2/5            15/19
                   0/2            0/4           0/3            0/8
                   4/4            4/5           5/6            12/14
    Total          12/16          11/23         7/15           29/52
                                                                       


    heterogeneity of the effects on milk production and the small number
    of animals tested, the evidence for a promoting effect of rbST on
    viral expression was insufficient.

         These results provide no evidence that treatment of cows infected
    with lentiviruses with rbST will cause resurgence of viral infections
    or pose any risk to human health. Lentiviruses are retroviruses that
    replicate only in activated cells of the immune system. They may stay
    dormant for months or years before they gradually wear down the immune
    system to the point of collapse. The phylogenetic tree of lentiviruses
    includes the subfamily of bovine immunodeficiency virus (also called
    bovine leukaemia virus, BLV) and the subfamily of HIV-1 and HIV-2
    which are the causative agents of AIDS. BLV and HIV are widely
    separated phylogenetically (Robertson, 1997). BLV is not known to
    cause disease in humans although it can infect human cells  in 
     vitro, where host defence mechanisms are not present (Georgiades
     et al., 1978; van der Maaten & Miller, 1990); infection of human
    cells  in vivo could not be demonstrated, and all attempts to obtain
    direct evidence of infection in exposed human populations have yielded
    negative results (Straub, 1981; van der Maaten & Miller, 1990).
    Failure to infect humans has also been shown for other ruminant
    lentiviruses, such as CAEV and maediœvisna (Straub, 1981). Excretion

    of virus with milk somatic cells can infect the offspring of infected
    cows. Such transmission can be effectively blocked by procedures
    similar to pasteurization, which destroy the virus at 60°C within 30 s
    (Abramova  et al., 1974; Baumgartner  et al., 1976; van der Maaten &
    Miller, 1990). BLV therefore cannot induce disease in humans and is
    completely inactivated by routine pasteurization. Furthermore, the
    company that sells sometribove has reported that there is no
    indication that the incidence of BLV infection has increased in cattle
    after eight years of continuous use of rbST in Mexico and Brazil and
    four years of use in the United States (Collier & Kowalczyk, 1998),
    although this statement is not further qualified.

         An increase in the expression of HIV in humans who ingest milk
    from rbST-treated cows is extermely unlikely because of the negligible
    residues of rbST and IGF-I. Treatment of AIDS patients for six weeks
    with recombinant human growth hormone and IGF-I had no effect on the
    HIV titres in peripheral blood mononuclear cells, on the CD3, CD4, or
    CD8 counts in peripheral blood, or on serum HIV p24 antigen
    concentrations (Waters  et al., 1996).

    2.3.3  Effect of recombinant bovine somatotropin on prion proteins

         Concern has been expressed that rbST treatment could increase the
    risk for BSE in dairy cows (Hansen  et al., 1997). Little evidence is
    available to support this concern, and that which has been provided is
    indirect. It is currently thought that the infectious agent of BSE is
    a prion protein (Prusiner, 1982). Prion proteins are found normally in
    all animals and are encoded by a prion-protein gene. BSE is associated
    with a post-translationally modified protease-resistant prion protein,
    which differs in its three-dimensional structure from the normal
    protease-sensitive prion protein. Normal prion proteins are bound to
    membranes on the surface of all nerve cells, some lymphocytes, and
    other tissues (Prusiner, 1991). To date, no function has been ascribed
    to normal prion protein. The most widely accepted theory of BSE is the
    conversion of normal prion protein to the abnormal protease-resistant
    form, which in turn causes more normal prion protein to convert to
    protease resistance. The mechanisms of the conversion to the
    disease-causing protease-resistant prion protein are not clearly
    understood. In contrast to the normal form, the protease-resistant
    form cannot be turned over and builds up in cells, forming large
    oligomers which are observed as plaques (amyloids) in the brains of
    affected individuals (Gajdusek, 1993).

         IGF-I increased the production of prion protein mRNA  in vitro 
    in a rat phaeochromocytoma cell line (PC12 cells), with a rather flat
    dose-response curve showing a 40% increase at 10 ng/ml and a doubling
    at 100 ng/ml (Lasmézas  et al., 1993). In transgenic mice that
    harbour multiple copies of the prion protein gene, the progession of
    scrapie was accelerated (Prusiner, 1991). It has been speculated that
    the increased IGF-I concentrations in rbST-treated cows could increase
    prion protein production and possibly speed up the progression of BSE;
    however, no studies are available that directly address the question
    of whether rbST or IGF-I increases the formation of normal prion

    protein or its pathogenic protease-resistant mutant in the brains of
    the cattle, and the possibility of a link between rbST treatment and
    BSE is highly speculative.

    2.4  Cow's milk and insulin-dependent type I diabetes mellitus in 
    childhood

         Epidemiological studies have shown that environmental factors
    such as chemicals, viral infections, a short duration of
    breastfeeding, and early dietary exposure of newborns to cow's milk
    increase the risk for insulin-dependent type I diabetes mellitus
    (IDDM) by about 1.5 times (Scott, 1990; Dahlquist  et al., 1991;
    Jorgensen  et al., 1991; Virtanen  et al., 1993; Gerstein, 1994;
    Verge  et al., 1994; Virtanen  et al., 1994). IDDM develops as a
    consequence of autoimmune destruction of the insulin-producing b cells
    of the pancreatic islets. The precise trigger of the autoimmune
    reaction is unknown, but it is assumed to be a genetically acquired
    immune defect in susceptible individuals (Gerstein, 1994). The
    epidemiological evidence indicates that IDDM is geographically and
    temporally related to neonatal feeding with cow's milk and that
    avoidance of cow's milk during the first few month of life can protect
    genetically predisposed individuals (Gerstein, 1994; Verge  et al., 
    1994).

         Serological evidence supports the theory that the immune defect
    is triggered by exposure to proteins in cow's milk (Gerstein, 1994).
    It has been postulated that, in newborns, milk proteins cross the
    immature gut wall, initiating an immune response that cross-reacts
    with a beta-cell surface antigen (Verge  et al., 1994). Children aged
    five to nine, who have an intact intestinal barrier, are not at risk
    of acquiring IDDM by drinking cow's milk (Dahlquist  et al., 1991).
    The possible triggering factors in cow's milk have not been
    identified. Casein is unlikely to be involved, since replacement of
    milk proteins by casein in the diets of rats susceptible to diabetes
    completely prevented the disease (Jorgensen  et al., 1991). Increased
    concentrations of immunoglobulin A antibodies to cow's milk and
    b-lactoglobulin appear to be associated with an increased risk for
    IDDM (Dahlquist  et al., 1991; Virtanen  et al., 1994).

         It is unlikely that exposure of human newborns to milk from
    rbST-treated cows would increase the risk for IDDM for the following
    reasons:

    *    The composition of milk from rbST-treated cows is well within the
         normal variations observed during lactation,

    *    The slightly increased IGF-I concentrations can be excluded as a
         triggering factor because bovine and human IGF-I are identical
         and because the concentrations of IGF-I in breast milk are equal
         to or initially higher than those found in cow's milk.

    3.  COMMENTS

     Use of antibiotics 

         After reviewing the available information, the Committee
    considered the risk for mastitis associated with use of rbST as an
    issue of animal health that is not within its terms of reference;
    however, the possibly increased use of antibiotics was considered. A
    post-approval monitoring programme was established in the United
    States to address the following issues:

    *    the incidence of mastitis and responses related to herd health
         (not within the terms of reference of the Committee),

    *    treatment with any medications of 28 herds of rbST-treated cows
         (not within the terms of reference of the Committee), and

    *    the percent of milk discarded because of violative drug residues
         in key dairy states representing at least 50% of United States
         milk production.

    In New York State, the percentage of milk discarded because of the
    presence of antibiotic residues was not significantly changed after
    the introduction of rbST. In other states, however, a small but
    statistically significant, increase was observed in 1995, which
    coincided with a change to a more sensitive testing method in those
    states. The Committee concluded that the use of rbST would not result
    in a higher risk to human health due to the use of antibiotics to
    treat mastitis and that the increased potential for drug residues in
    milk could be managed by practices currently in use within the dairy
    industry and by following the directions for use.

     Concentrations of insulin-like growth factor in milk and tissues

         Insulin-like growth factor (IGF-I) is a normal component of milk
    and is found in abundance in a variety of body fluids (Table 8). The
    presence and concentrations of IGF-I were the source of much of the
    scientific discussion during the original review of bST undertaken at
    the fortieth meeting of the Committee and in submissions to the
    present one. The information that was reviewed is summarized in FAO
    Food and Nutrition Paper No. 41/5 (Annex 1, reference 106). The
    concentrations of IGF-I in milk are variable and have been shown to
    depend on the state of lactation, nutrition, and age.

         Methods for assaying IGF-I were considered by the Committee.
    Although incomplete removal of IGF-binding proteins or differences in
    the source of standards and extraction methods might affect the
    reported values, these factors were considered not to alter any
    conclusions materially. The relatively high values previously reported
    in milk were considered to reflect inadequate extraction.

    Table 8. Insulin-like growth factor in milk and body fluids

                                                             

    Medium                                  Concentration 
                                            (ng/ml)
                                                             
    Milk
      Human                                 5-10
        Colostrum                           8-28
      Bovine (bulk milk)
        Untreated                           1-9
        rbST-treated                        1-13
      Plasma
        Child                               17-250
        Adolescent                          180-780
        Adult                               120-460
      Gastrointestinal secretions (human)
        Saliva                              6.8
        Gastric juice                       26
        Pancreatic juice                    27
        Bile                                6.8
        Jejunal chyme                       180

      Daily production of adult humans      107 ng/d
                                                             


         Since the previous evaluation, few additional data on residues
    have appeared in the literature or in reports provided by interested
    parties; however, the manufacturer of sometribove submitted additional
    information on the concentrations of IGF-I in retail milk after the
    approval of rbST in the United States. The results showed no
    difference in the IGF-I concentrations of milk certified as derived
    from cows not treated with rbST and of unlabelled milk; however, the
    percentage of the unlabelled milk that was derived from cows receiving
    rbST was not specified.

         Concern has been expressed that any rbST-induced increase in the
    concentration of IGF-I in milk would contribute to the endogenous
    levels of IGF-I in the gastrointestinal tract and serum if it were not
    biodegraded and were absorbed. In rats, IGF-I is rapidly degraded in
    the gastrointestinal tract; however, a protective effect of casein on
    IGF-I was demonstrated in these studies. It was postulated that the
    retarded degradation leads to increased serum levels of IGF-I (as
    shown in one study in rats) and to prolonged exposure of the gut. The
    Committee noted that seven days' oral administration of high doses of
    IGF-I in milk replacer did not increase the circulating concentrations
    of IGF-I in newborn calves and piglets, indicating that significant
    absorption of IGF-I is unlikely to occur under physiological
    conditions. In view of the decreased rate of degradation in the small
    intestine of rats in the presence of casein, the concentrations of the
    growth factor would probably decrease to less than 5% of their initial

    values within 2 h, so that milk-borne IGF-I would not be expected to
    contribute to the concentrations of IGF-I in the large intestine.

         If 1.5 L of milk are ingested per day, the average intake of
    IGF-I will be 6000 ng in milk from untreated cows containing an
    assumed IGF-I concentration of 4 ng/ml and 9000 ng in milk from rbST-
    treated cows with an approximate average concentration of 6 ng/ml. It
    has been calculated that IGF-I in gastrointestinal secretions amounts
    to about 380 000 ng/day. Therefore, the additional amount of IGF-I in
    1.5 L of milk from rbST-treated cows would represent only about 0.8%
    of the IGF-I secreted in the gastrointestinal tract. The total amount
    of IGF-I in serum has been calculated to range from 50 000 to 1 220
    000 ng, depending on age. The total daily IGF-I production in adult
    humans has been estimated to be 107 ng. Therefore, the daily amount
    of IGF-I ingested with milk from rbST-treated cows would represent
    less than 0.09% of the daily production of adults. Even if all of the
    milk-borne IGF-I were absorbed, the additional amount would be
    negligible.

         When sustained-release rbST was administered to cattle once every
    two weeks for a total of 20 weeks, the tissue concentrations of rbST
    and IGF-I two weeks after the final dose were not significantly
    increased.

         The Committee concluded that any increase in the concentration of
    IGF-I in milk from rbST-treated cows is orders of magnitude lower than
    the physiological amounts produced in the gastrointestinal tract and
    in other parts of the body. Thus, the concentration of IGF-I would not
    increase either locally in the gut or systemically, and the potential
    for IGF-I to promote tumour growth would not increase when milk from
    rbST-treated cows was consumed; there is thus no appreciable risk for
    consumers.

     Expression of retroviruses

         Concern that treatment of cattle with rbST would increase the
    expression of retroviruses, including BLV, were addressed in
    experiments with caprine arthritis encephalitis virus in goats.
    Infectivity was not increased when measured as numbers of infected
    cells, and there was no evidence of increased reverse transcriptase
    activity. These studies provide no evidence that rbST affects the
    expression of BLV, a lentivirus in cattle. Furthermore it has been
    shown that BLV is destroyed in simulated pasteurization conditions by
    heating milk to 60 .C for  30 s. In addition, there is no evidence
    of human susceptibility or response to ruminant retroviruses.

     Expression of prion proteins

         Concern has been expressed that treatment with rbST could shorten
    the incubation period for BSE. This hypothesis is based on results
    obtained with a neuronal cell line  in vitro indicating increased
    formation of prion protein mRNA in response to IGF-I. Furthermore,
    increased formation of prion protein shortened the incubation period

    for scrapie in transgenic mice harbouring multiple copies of the gene
    that codes for prion proteins. No data were available, however, to
    address directly the question of whether rbST or IGF-I increases the
    formation of normal prion protein or its pathogenic protease-resistant
    mutant in the brains of cattle. The Committee considered that the
    possibility of a link between rbST treatment and BSE was highly
    speculative.

     Risk for insulin-dependent diabetes mellitus

         Exposure of human newborns to cow's milk increases the risk for
    insulin-dependent diabetes mellitus by about 1.5-fold. The Committee
    considered whether exposure of newborns to milk from rbST-treated cows
    further increases this risk. It concluded that, because of its
    unchanged composition, the milk of rbST-treated cows would not pose an
    additional risk to the development of insulin-dependent diabetes
    mellitus.

    4.  EVALUATION

         On the basis of the following:

    *    insignificant changes in the quantities of milk discarded after
         testing for antibiotic residues following the introduction of
         rbST into commercial use;

    *    low concentrations of rbST and IGF-I residues in milk;

    *    degradation of IGF-I in the gut and its abundance in gut
         secretions;

    *    the extremely low concentrations of ingested IGF-I in comparison
         with endogenous production;

    *    lack of evidence that rbST stimulates the expression of
         retroviruses;

    *    lack of a direct link between rbST treatment and BSE; and

    *    the absence of significant changes in the composition of milk
         from rbST-treated cows that could contribute to an additional
         risk for the development of insulin-dependent diabetes mellitus,

    the Committee concluded that rbST can be used without any appreciable
    risk to the health of consumers. The Committee reaffirmed its previous
    ADIs1 and MRLs2 'not specified' for somagrebove, sometribove,
    somavubove, and somidobove.

    5.  REFERENCES

    Abramova, E.M., Kondratev, V.S. & Sytinskii, I.A. (1974) The
    biochemistry of leucosis in cattle.  Vet. Bull., 44, 689-711.

    Baumann, D.M. (1995)  IGF-I Fact Sheet, Department of Animal Science,
    Cornell University, Ithaca, NY, USA.

    Baumgartner, L.E., Olson, C. & Onuma, M. (1976) Effect of
    pasteurization and heat treatment on bovine leukemia virus.  J. Am. 
     Vet. Med. Assoc., 169, 1189-1191.

    Baumrucker, R.R., Hadsell, D.J., Skarr, T.C., Campbell, P.G. & Blum,
    J.W. (1992) Insulin-like growth factors (IGFs) and IGF binding
    proteins in mammary secretions: Origins and implications in neonatal
    physiology. In: Picciano, M.F. & Lonnerdal, B., eds,  Mechanisms 
     Regulating Lactation and Infant Nutrient Utilization, New York,
    Wiley-Liss, pp. 285-308.

    Burrin, D.G. (1997). Is milk-borne insulin-like growth factor-l
    essential for neonatal development?  J. Nutr., 127, 975S-979S.

    Burton, J.L., McBride, B.W., Block, E., Glimm, D.R. & Kenelly, J.J.
    (1994) A review of bovine growth hormone.  Can. J. Anim. Sci., 74,
    167-201.

    Challacombe, D.N. & Wheeler, E.E. (1994) Safety of milk from cows
    treated with bovine somatotropin.  Lancet, 344, 815-816.

    Chaurasia, O.P., Marcuard, S.P. & Seidel, E.R. (1994) Insulin-like
    growth factor l in human gastrointestinal exocrine excretion.
     Regul. Pept., 50, 113-119.

    Choi, J., Choi, M.J., Kim, C., Ha, J., Hong, A., Ji, Y. & Chang, B.
    (1997). The effect of recombinant bovine somatotropin (rbST)
    administration on residual BST and insulin-like growth factor I levels
    in various tissues of cattle.  J. Food Hyg. Soc. Jpn, 38, 225-232.

    Codex Alimentarius Commission (1997)  Report of the Twenty-second 
     Session of the Codex Alimentarius Commission, Geneva, 23-28 June 
     1997 (FAO Document ALINORM 97/37), Rome, Food and Agriculture
    Organization of the United Nations.

    Collier, R.J. (1996) Post-approval evaluation of Posilac bovine
    somatotropin in commercial dairy herds (100-USA-COW-RJC-93-051).
    Unpublished report from Protiva, Monsanto Co. Submitted to WHO by
    Protiva, Monsanto Co., St Louis, Missouri, USA.

    Collier, R.J. & Kowalczyk, D.F. (1998) Human health risk of
    retroviruses in cattle and bovine somatotropine. Unpublished report
    from Protiva, Monsanto Co. Submitted to WHO by Protiva, Monsanto Co.,
    St Louis, Missouri, USA.

    Comens-Keller, P.G., Eppard, P.J. & Collier, R.L. (1995). Evaluation
    of somatotropin as a homeorhetic regulator of immunity. In: Ivan, M.,
    ed.,  Animal Science Research and Development: Moving Toward a New 
     Century. Proceedings of the 75th Anniversary Meeting of the 
     Canadian Society of Animal Science, Ottawa, pp. 79-94.

    Dahlquist, G., Blom, L. & Lönnberg, G. (1991) The Swedish childhood
    diabetes study: A multivariate analysis of risk determinants for
    diabetes in different age groups.  Diabetologica, 34, 757-762.

    Donovan, S.M., Chao, J.C., Zijlstra, R.T. & Odle, J. (1997) Oral
    administered iodinated recombinant human insulin-like growth factor-I
    (125I-rhIGF-I) is poorly absorbed by the neonatal piglet.  J. 
     Pedriat. Gastroenterol. Nutr., 24, 174-182.

    Eppard, P.J., Collier, R.J., Hintz, R.L., Veenhuizen, J.J. & Baile,
    C.A. (1994) Survey of milk insulin-like growth factor in retail milk
    samples. Unpublished report No. 100-USA-COW-RJC-94-074 from Protiva,
    Monsanto Co. Submitted to WHO by Protiva, Monsanto Co., St Louis,
    Missouri, USA.

    Epstein, S.S. (1996) Unlabeled milk from cows treated with
    biosynthetic growth hormones: A case of regulatory abdication.  Int. 
     J. Health Serv., 26, 173-185.

    Ezzat, H. & Melmed, S. (1991) Clinical review 18: Are patients with
    acromegaly at increased risk for neoplasia?  J. Clin. Endocrinol. 
     Metab., 72, 245-249.

    Gajdusek, D.C. (1993) Genetic control of nucleation and polymerization
    of host precursors to infectious amyloids in the transmissible
    amyliodoses of brain.  Br. Med. Bull., 49, 913-931.

    Ganong, W.F., ed. (1971)  Review of Medical Physiology, Los Altos,
    Lange Medical Publications, p. 5.

    Georgiades, J.A., Billiau, A. & Vanderschueren, B. (1978) Infection of
    human cell cultures with bovine visna virus.  J. Gen. Virol., 38,
    375-381.

    Gerstein, H.C. (1994) Cow's milk exposure and type I diabetes
    mellitus.  Diabetes Care, 17, 13-19.

    Guler, H.P., Zapf, J., Schmid, C. &, Froesch, E.R. (1989) Insulin-like
    growth factors I and II in healthy man. Estimations of half-lives and
    production rates.  Arch. Endocrinol., 121, 753-758.

    Hadsell, D.L. (1991)  Insulin-like Growth Factor (IGF) Receptors in 
     the Bovine Mammary Gland: Regulation, Mitogenic Action, and Role in 
     Mammary Secretion of IGF-I, PhD Thesis, Pennsylvania State
    University, Philadelphia, Pennsylvania, pp. 108-114,

    Hammon, H. & Blum, J.W. (1997) The somatotropic axis in neonatal
    calves can be modulated by nutrition, growth hormone and
    Long-R3-IGF-I.  Am. J. Physiol., 273, E130-E138.

    Hansen, M., Halloran, J.M., Groth, E., III & Lefferts, L.Y. (1997)
    Potential public health impacts of the use of recombinant bovine
    somatotropin in dairy production. Unpublished report. Submitted to WHO
    by Consumer Policy Institute, Yonkers, New York, USA.

    Houle, V.M., Sschroeder, E.A., Odle, J. & Donovan, S.M. (1997) Small
    intestinal disaccharidase activity and ileal villus height are
    increased in piglets consuming recombinant human insulin-like growth
    factor-I.  Pediatr. Res., 42, 78-86.

    Jorgensen, K.D., Joner, G. & Hanssen, K.F. (1991) Relationship between
    cows' milk consumption and incidence of childhood IDDM.  Diabetes 
     Care, 14, 1081-1083.

    Juskevich, J.C. & Guyer, C.G. (1990) Bovine growth hormone: Human food
    safety evaluation.  Science, 249, 875-884.

    Kelley, K.W. (1989) Growth hormone, lymphocytes and macrophages.
     Biochem. Pharmacol., 38, 705-713.

    Kimura, T., Murakawa, Y., Ohno, M., Ohtani, S. & Higaki, K. (1997)
    Gastrointestinal absorption of recombinant human insulin-like growth
    factor-I in rats.  J. Pharmacol. Exp. Ther., 283, 611-618.

    Laburthe, M., Rouyer-Fessard, C. & Gammeltoft, S. (1988) Receptors for
    insulin-like growth factors I and II in the rat gastrointestinasl
    epithelium.  Am. J. Physiol., 254, 457-462.

    Lasmézas, C., Deslys, J.P. & Dormont, D. (1993) Recombinant human
    growth hormone and insulin-like growth factor I induce PrP gene
    expression in PC12 cells.  Biochem. Biophys. Res. Commun., 196,
    1163-1169.

    Lerondelle, C., Lena, P. & Mornex, J.F. (1994) [Hormonal control of
    retroviral infections.]  Renc. Rech. Ruminants, 1, 45œ50 (in French).

    Lerondelle, C., Guiguen, F., Fornazero, C., Gounel, F., Leroux, C.,
    Chastang, J., Pardo, E., Bolland, P., Bruyas, S., Cordier, G. &
    Mornex, J.F. (1996) Effects of recombinant bovine somatotropin and of
    thyroxine on the expression of lentiviruses in ruminants. Unpublished
    report from INRA Veterinary School of Lyon and INSERM Hôpital Louis
    Pradel, Lyon, France. Submitted to WHO by Protiva, Monsanto Co., St
    Louis, Missouri, USA.

    van der Maaten, M.J. & Miller, J.M. (1990) Bovine leucosis. In:
     Virus Infections of Ruminants, Amsterdam,  Elsevier Science
    Publishers, pp. 419-429.

    Macaulay, V.M. (1992). Insulin-like growth factors and cancer.  Br. 
     J. Cancer, 65, 311-320.

    National Institute of Health (1995) NIH conference: Insulin-like
    growth factors and cancer.  Ann. Intern. Med., 122, 54-59.

    Olanrewaju, H., Patel, L. & Seidel, E.R. (1992) Trophic action of
    intraileal infusion of insulin-like growth factor I: Polyamine
    dependence.  Am. J. Physiol., 263, E282-E286.

    Phillips, A.F., Rao, R., Anderson, G.G., McCracken, G.M., Lake, M. &
    Koldovsky, O. (1995) Fate of insulin-like growth factors I and II
    administered orogastrically to suckling rats.  Pediatr. Res., 37,
    586-592.

    Pines, A., Rozen, M.B. & Gilat, T. (1985) Gastrointestinal tumors in
    acromegalic patients.  Am. J. Gastroenterol., 80, 286-289.

    Prusiner, S.B. (1982) Molecular biology of prion diseases.  Science, 
    216, 136.

    Prusiner, S.B. (1991) Molecular biology of prion diseases.  Science, 
    252, 1515-1522.

    Robertson, D.L. (1997) The evolution of AIDS viruses. From the
    Internet

    Schaff-Blass, E., Burstein, S. & Rosenfield, R.L. (1984) Advances in
    diagnosis and treatment of short stature, with special reference to
    the role of growth hormone.  J. Pediatr., 104, 801.

    Scott, F.W. (1990) Cow milk and insulin-dependent diabetes mellitus:
    Is there a relationship?  Am. J. Clin. Nutr., 51, 489-491.

    Straub, O.C. (1981) Enzootic bovie leucosis. In: Gibbs, E.P.J., ed.,
     Virus Diseases of Food Animals, New York, Academic Press, pp.
    683-718.

    US Food and Drug Administration (1996) CVM Update: VMAC endorses
    post-approval program for Posilac(R), 18 December 1996.

    Vander A.J., Sherman, J.H. & Luciana, D.S., eds (1990)  Human 
     Physiology, The Mechanisms of Body Function, 5th Ed., New York,
    McGraw-Hill Publishing Co., p. 15.

    Veenhuizen, J.J., Hintz, R.L., Hemenover, M.L., Kowalczyk, D.F. &
    Collier, R.J. (1996) Post-approval monitoring program for Posilac
    bovine somatotropin: Tracking program of milk discarded due to drug
    residue violations. Unpublished report No. 100-USA-Cow-RLH-93-053 from
    Protiva, Monsanto Co. Submitted to WHO by Protiva, Monsanto Co., St
    Louis, Missouri, USA.

    Vega, J.R., Gibson, C.A., Skaar, T.C., Hadsell, D.L. & Baumrucker,
    C.R. (1991) Insulin-like growth factor (IGF)-I and II and IGF binding
    proteins in serum and mammary secretions during the prepartum period
    and early lactation in dairy cows.  J. Anim. Sci., 69, 2538-2547.

    Verge, C.F., Howard, N.J., Irwig, L., Simpson, J.M., Mackerras, D. &
    Silink, M. (1994) Environmental factors in childhood IDDM.  Diabetes 
     Care, 17, 1391-1389.

    Virtanen, S.M., Räsänen, L., Ylönen, K., Aro, A., Clayton, D.,
    Langholz, B., Pitkäniemi, J., Savilahti, E., Lounamaa, R., Tuomolehto,
    J. & Akerblom, H.K. (1993) Early introduction of dairy products
    associated with increased risk of IDDM in Finnish children.
     Diabetes, 42, 1786-1790.

    Virtanen, S.M., Saukkonen, T., Savilahti, E., Ylönen, K., Räsänen, L.,
    Aro, A., Knip, M., Tuomolehto, J. & Akerblom, H.K. (1994) Diet, cow's
    milk protein antibodies and the risk of IDDM in Finnish children.
     Diabetologica, 37, 381-387.

    Waters, D., Danska, J., Hardy, K., Koster, F., Qualls, C., Nickell,
    D., Nightingale, S., Gesundheit, N., Watson, D. & Schade, C. (1996)
    Recombinant human growth hormone, insulin-like growth factor 1, and
    combination therapy in AIDS-associated wasting.  Ann. Intern. Med., 
    125, 865-872.

    Xian, C.L., Shoubridge, C.A. & Read, L.C. (1995) Degradation of IGF-I
    in the adult rat gastrointestinal tract is limited by a specific
    antiserum or the dietary protein casein.  J. Endocrinol., 146,
    215-225.

    Zumkeller W. (1992) Relationship between insulin-like growth factor-I
    and II and IGF-binding proteins in milk and the gastrointestinal
    tract: Growth and development of the gut.  J. Pediatr. 
     Gastroenterol., 15, 357-369.
    


    See Also:
       Toxicological Abbreviations