INTERNATIONAL PROGRAMME ON CHEMICAL SAFETY
WORLD HEALTH ORGANIZATION
TOXICOLOGICAL EVALUATION OF CERTAIN
VETERINARY DRUG RESIDUES IN FOOD
WHO FOOD ADDITIVES SERIES: 43
Prepared by the Fifty-second meeting of the Joint FAO/WHO
Expert Committee on Food Additives (JECFA)
World Health Organization, Geneva, 2000
IPCS - International Programme on Chemical Safety
ANTIMICROBIAL AGENT
The Committee examined the results of studies of antimicrobial
activity in vitro in which relevant gut microflora were used to
calculate ADIs for the antimicrobial agent that was reviewed. The
equation used for deriving the ADIs is based on that described in the
report of the forty-seventh meeting of the Committee (Annex 1,
reference 125).
The upper limit of the ADI was derived as follows:
Upper limit MIC50 (µg/g × MCC (g)
of ADI = ----------------------
(µg/kg bw) FA × SF × BW (kg)
where:
MIC50 = The minimum concentration of an antimicrobial drug that
completely inhibits the growth of 50% of cultures of a
given microorganism, as judged by the naked eye, after
a given period of incubation. For the purposes of the
evaluation, the MIC50 value(s) for the strain(s) of
the relevant species tested. Alternatively, the lowest
MIC50 value for the most sensitive species can be
used.
MCC = Mass of colonic content; a value of 220 g is used.
FA = Fraction of an oral dose available to act upon
microorganisms in the colon
SF = The safety factor used to account for uncertainty about
the amount and relevance of data for MIC values
available for review; may range from 1 to 10, a value
of 1 being used when extensive, relevant
microbiological data are available.
BW = Body weight; a value of 60 kg is used.
THIAMPHENICOL (addendum)
First draft prepared by F.R. Ungemach
Institute of Pharmacology, Pharmacy and Toxicology
Veterinary Faculty, University of Leipzig, Leipzig, Germany
Explanation
Biological data
Short-term study of toxicity
Study of carcinogenicity
Study of teratogenicity in rabbits
Special studies on microbiological effects
Comments
Evaluation
References
1. EXPLANATION
Thiamphenicol was evaluated by the Committee at its forty-seventh
meeting (Annex 1, reference 125), at which time a temporary ADI of 0-6
µg/kg bw was established on the basis of a NOEL of 1.25 mg/kg bw per
day for maternal toxicity in a study of teratogenicity in rabbits and
by applying a safety factor of 200. The ADI was designated 'temporary'
since only a summary report of the study on carcinogenicity in rats
was available. In order to evaluate the carcinogenic potential of
thiamphenicol further, the Committee requested detailed reports giving
the individual data from the carcinogenicity study and from the
range-finding study from which the doses were derived.
Since the previous evaluation, the results of a 13-week study of
toxicity and of a subsequent two-year study of the carcinogenicity of
thiamphenicol in rats have been published, and the data for individual
animals in both studies have been submitted. The additional
information is summarized and discussed in this addendum. Furthermore,
the study of developmental toxicity in rabbits and the studies on the
microbiological effects of thiamphenicol which were used by the
Committee at its forty-seventh meeting to establish the toxicological
and microbiological ADIs were re-evaluated to determine the most
appropriate toxicological and microbiological end-points.
2. BIOLOGICAL DATA
2.1 Short-term study of toxicity
In a study of toxicity with repeated doses, which was designed to
find the range of doses to be used in a subsequent study of
carcinogenicity, Fischer 344 rats were treated orally with
thiamphenicol for 13 weeks. The appropriate doses for this study were
determined in two pilot studies of single and repeated oral doses of
three to four weeks' duration. Data on individual animals were not
submitted. After a single oral dose of 4, 6, 8, or 10 g/kg bw of
thiamphenicol given by gavage to groups of five rats of each sex, no
relevant clinical signs, no suppression of body-weight gain, and no
pathological findings were observed during the 14-day observation
period, indicating that this substance has low acute toxicity. When
groups of two to five male and female rats were given 0, 500, 1000,
2000, or 2500 mg/L of thiamphenicol in drinking-water (equal to 0, 67,
130, 270, and 330 mg/kg bw per day) for three to four weeks,
body-weight gain was severely suppressed at doses of 1000 mg/L and
higher. At the lowest dose, there was slight suppression of
body-weight gain and slight pathological alterations, such as thymic
atrophy, dilatation of the caecum, suppression of haematopoiesis, and
testicular degeneration. A dose of 500 mg of thiamphenicol per litre
of drinking-water was selected as the highest level for the 13-week
study (Ando et al., 1997).
The 13-week study was not performed in accordance with good
laboratory practice, but the data for individual animals were provided
for evaluation, allowing an independent assessment. In this study,
groups of 12 Fischer 344 rats of each sex received drinking-water ad
libitum containing thiamphenicol at a dose of 0, 125, 250, or 500 mg/L
for 13 weeks, equal to 9, 17, and 36 mg/kg bw per day for males and
12, 21, and 39 mg/kg bw per day for females. The animals were observed
daily for clinical signs of adverse reactions. Body weight and food
and water consumption were recorded weekly. At the end of treatment,
all survivors were checked for haematological and serum parameters,
organ weights and gross and histopathological appearance.
No deaths were recorded in any of the groups. A dose-related
reduction in body-weight gain was observed in all treated groups,
reaching significance at the intermediate and high doses. Animals at
the high dose also showed increased relative liver and kidney weights,
decreased relative thymus weight, slight suppression of bone-marrow
erythropoiesis and testicular lesions with vacuolation of germinal
epithelium, spermatogranulomas, and reduced sperm count in the
epididymides; similar effects were observed occasionally at the
intermediate dose. Slight changes in haematological parameters
(increased mean corpuscular volume, decreased erythrocyte and platelet
counts) and serum biochemical parameters (decreased total protein,
calcium, and cholesterol in females, increased albumin:globulin ratio,
urea nitrogen, and alkaline phosphatase in females) were seen in a
dose-dependent manner at the two higher doses. Thiamphenicol at all
three doses caused enlargement of the caecum, which is a common
side-effect of long-term oral administration of antibacterial agents
in rabbits. The lowest dose caused no other relevant substance-related
adverse in organs or tissues, and the minor haematological and serum
bioc, were almost within the range of normal variations and were
inconsistent with regard to dose and sex.
A dose of 250 mg/L of thiamphenicol in drinking-water (equal to
17 mg/kg bw per day) was slightly toxic, whereas 125 mg/L (equal to
9 mg/kg bw per day) caused no histopathological lesions in the target
organs of thiamphenicol, such as bone marrow and testicular tissue.
The minor changes in haematological and serum biochemical parameters
at the lowest dose were biologically insignificant. The NOEL was 125
mg/L, 9 mg/kg bw per day (Ando et al., 1997; Maekawa, 1998a).
2.2 Study of carcinogenicity
The doses used in a two-year carcinogenicity study in Fischer 344
rats were based on the findings of the dose range-finding study
described above. The previous Committee had access only to a summary
report of the results of this study, but detailed individual data have
since been submitted which allowed independent assessment of the
results and of the conclusions summarized in a recent publication. The
design of the study did not conform to good laboratory practice, but
the study was of adequate quality and duration. Randomly allocated
groups of 50 rats of each sex received thiamphenicol in their
drinking-water at a concentration of 0, 150, or 250 mg/L for 104
weeks; the drinking-water was supplied ad libitum. On the basis of
water consumption, the doses of thiamphenicol were calculated to be
equal to 5 and 11 mg/kg bw per day for males and 7 and 14 mg/kg bw per
day for females. The complete dissolution of thiamphenicol in the
water at all doses and the stability of the solutions at room
temperature were checked. The plasma concentrations of thiamphenicol
were not determined. The animals were observed daily for clinical
signs and deaths, and food and water consumption and body weights were
checked once a week for the first 13 weeks and then every four weeks.
Four weeks after cessation of the treatment, all survivors were killed
and autopsied. The organs were weighed and gross and microscopic
anatomical pathological examinations were performed.
Reduced body-weight gain was observed at the high dose, which
resulted in siginificantly lower body weights of females at the end of
the experiment. Food consumption and water intake were not
significantly different in the treated and control groups and remained
constant throughout the experiment. The mean length of survival of
thiamphenicol-treated animals was not different from that of controls;
in fact, by the end of the experiment, slightly more treated animals
(72-78%) than controls (68% of males and 72% of females) were still
alive. The only difference in organ weights was increased pituitary
weights in females at the high dose. The incidence and severity of
non-neoplastic lesions such as nephropathy, bile-duct proliferation,
microgranuloma in liver, bone marrow, and lymph nodes, extramedullary
haematopoiesis in spleen, and atrophy of accessory genital glands were
similar in all groups, including the controls. These pathological
findings were considered to be spontaneous and typical for aged
Fischer 344 rats. The histopathological survey showed no evidence of
effects on target organs such as bone marrow and testicular tissue.
Furthermore, no changes were reported that reflected the slight,
biologically insignificant alterations in some haematological and
biochemical parameters observed at the lowest dose in the 13-week
study in rats.
The total tumour incidence was 100% in males in all groups,
whereas in females the percentage of tumour-bearing animals showed a
nonsignificant dose-related increase, from 84% in controls to 86% at
the low dose and 90% at the high dose. Tumours were found in many
organs and tissues in all groups, and the organ distribution and
histological characteristics of the tumours were similar to those
reported to occur spontaneously in rats of this strain, with the
exception of a higher incidence of histiocytic sarcomas of the bone
marrow in some control and treated animals. The reason for the more
frequent occurrence of this rare tumour remains unclear, but it does
not appear to be related to treatment in view of the higher incidence
in controls and the lack of dependence on dose. The incidence of
pituitary adenomas was increased in females, from 12/50 in controls to
16/49 at the low dose and, significantly, to 19/50 in high-dosed
animals. The more frequent occurrence of this tumour was, however, not
associated with any increase of the incidence of preneoplastic lesions
or differences in induction time as compared with controls. In males,
the incidence of pituitary adenomas was decreased in a dose-dependent
manner. At the high dose, a slight statistically nonsignificant
increase in the incidence of thyroid follicular adenomas and subcutis
fibromas was observed in males and increases in the incidences of
thyroid C-cell adenomas, clitoral gland adenomas and uterine
endometrial stomal polyps in females; however, the incidences of these
tumours were within the range of the spontaneous incidence in Fischer
344 rats.
The Committee concluded that thiampenicol is not carcinogenic in
Fischer 344 rats when given continuously in the drinking-water at
slightly toxic doses up to 14 mg/kg bw per day for two years. The NOEL
was 5 mg/kg bw per day (Kitamura et al., 1997; Maekawa, 1998b).
2.3 Study of developmental toxicity in rabbits
The study of developmental toxicity in rabbits treated by daily
oral administration of thiamphenicol at a dose of 0, 1.25, 2.5, or 5
mg/kg bw on days 6-18 of gestation, which was evaluated by the
Committee at its forty-seventh meeting (Annex 1, reference 125), was
re-evaluated. In this study, thiamphenicol caused mild toxicity in the
dams, as indicated by a dose-dependent suppression bw throughout the
administration period. This effect reached a maximum within the first
week and was partly reversed after cessation of treatment. This slight
maternal toxicity was not associated with overt toxic effects in
embryos or fetuses. Decreased mean fetal weight was seen at the
highest dose as compared with controls, but this effect was presumed
to be the consequence of increased litter weight and a larger number
of fetuses (Sisti et al., 1994).
At its forty-seventh meeting, the Committee used the NOEL for
maternal toxicity of 1.25 mg/kg bw per day to establish a temporary
ADI, although the body-weight gain of the pregnant animals was
significantly decreased at this dose at days 7-9 of treatment and did
not reach the values of the control group until the end of the
experiment. Rabbits are, however, known to be highly sensitive to
orally administered antimicrobial agents that cause gastroenteritis
and poor digestion, with negative consequences on body-weight gain.
2.4 Special studies on microbiological effects
The Committee at its forty-seventh meeting evaluated the
antibacterial activity of thiamphenicol on the basis of published
literature (Annex 1, reference 125). No new data have been provided.
In the available studies, the minimum inhibitory concentrations (MICs)
of a wide range of pathogens and of bacterial species representative
of human gut flora indicate that anaerobic bacterial strains are
especially sensitive to thiamphenicol. Actinomyces, Fusobacterium and
Propionibacterium were found to be the most sensitive species, with
MIC50 values of 0.5-1.0 µg/ml. The Committee at its forty-seventh
meeting used the mean MIC50 of 1.68 µg/ml for all 261 strains of 16
bacterial species of human origin included in the study as the
microbiological end-point to calculate the ADI. According to the most
recent approach of the Committee for the assessment of microbiological
risks of residues of antibacterial substances (see section 2.3 of the
report of the present meeting: Annex 1, reference 140), the
microbiological ADI should be established on the basis of the MIC50
of the most sensitive, relevant bacterial species of the human gut
flora. Fusobacterium, with a MIC50 of 0.5 µg/ml, was considered to be
one of the most relevant microorganisms of the human gut flora, of
which 20 strains had been tested.
3. COMMENTS
In a pilot study to determine the range of doses to be used in a
carcinogenicity study, thiamphenicol was administered in the
drinking-water to Fischer 344 rats of each sex at a dose of 0, 125,
250, 0r 500 mg/L (equal to 9, 17, and 36 mg/kg bw per day in males and
to 12, 21, and 39 mg/kg bw per day in females), for 13 weeks.
Thiamphenicol caused no deaths. At 125 and 250 mg/L in drinking-water,
numerous adverse effects were reported, including dose-dependent
suppression of body-weight gain, a slight reduction in erythropoiesis
in bone marrow, reduced erythrocyte and platelet counts, decreased
thymus weight and increased liver and kidney weights relative to body
weight, testicular lesions with degeneration of the seminiferous
tubules, reduced sperm count, and spermatogranulomas in the
epididymides. In addition, minor changes were observed in serum
biochemical parameters, with reduced total serum protein, an increased
albumin:globulin ratio, and decreased cholesterol concentration. In
all treated animals, thiamphenicol caused dose-related enlargement of
the caecum, which is a common side-effect of long-term oral treatment
of rodents with antibiotics. At the lowest dose, no other relevant
treatment-related adverse effect was seen on tissues or organs or on
haematological or serum biochemical parameters. The results indicated
that 250 mg/L in drinking-water (equal to 17 mg/kg bw per day) was
slightly toxic, whereas 125 mg/L (equal to 9 mg/kg bw per day) was the
NOEL.
On the basis of the findings in the pilot study, doses of
thiamphenicol of 0, 125, and 250 mg/L in drinking-water (equal to 5
and 11 mg/kg bw per day in males and 7 and 14 mg/kg bw per day in
females) were selected for the two-year study of carcinogenicity in
Fischer 344 rats. The highest dose caused a reduction in body-weight
gain, whereas food and water consumption remained unchanged. The mean
survival time of thiamphenicol-treated animals was not different from
that of controls. The incidence and severity of non-neoplastic lesions
were similar in all groups, and the lesions were considered to be
spontaneous and typical of aged Fischer 344 rats. The weight of the
pituitary gland was increased in females. The incidence of malignant
tumours was not significantly increased in treated animals over that
in controls, and the distribution of tumours by organ and the
histological characteristics and frequency of the tumours observed in
all groups were similar to those reported to occur spontaneously in
this strain of rats. The incidences of tumours at some sites were
slightly increased in animals at the highest dose but did not exceed
the spontaneous incidence reported in historical controls. The
increased incidence of pituitary adenomas reached significance only in
females and was not associated with any increase in the incidence of
preneoplastic lesions or differences in induction time.
On the basis of these findings, the Committee concluded that
there was no evidence for the carcinogenicity of thiamphenicol in the
two-year study in rats. This conclusion was further supported by the
absence of genotoxic effects of thiamphenicol in five tests for
genotoxicity in vitro and a test for micronucleus formation in mice
treated in vivo that had been evaluated by the Committee at its
forty-seventh meeting. At its present meeting, the Committee noted
that a histopathological survey had shown no evidence that
thiamphenicol was toxic in target tissues such as bone marrow and
testis. Furthermore, no changes were reported that reflected the
slight, biologically insignificant alterations in some of the
haematological and biochemical parameters observed at the lowest dose
in the 13-week study of toxicity in rats. The NOEL in the two-year
carcinogenicity study was therefore 5 mg/kg bw per day.
The Committee reconsidered the study of developmental toxicity in
rabbits that was used at the forty-seventh meeting to establish the
ADI, on the basis of the NOEL of 1.25 mg/kg bw per day for maternal
toxicity. The Committee considered that this effect was not an
appropriate toxicological end-point because of the known sensitivity
of rabbits to orally administered antimicrobial agents. The Committee
considered that the NOEL of 5 mg/kg bw per day in the two-year study
of carcinogenicity in rats was the most relevant toxicological
end-point.
The Committee further reconsidered the microbiological data on
thiamphenicol that were reviewed at its forty-seventh meeting. At that
time, the mean MIC50 of all 261 strains of 16 bacterial species of
human origin that had been studied was used to calculate the ADI. The
most recent approach of the Committee for assessing the
microbiological risks of residues of antimicrobial substances is
described in section 2.3 of the report of the present meeting (Annex
1, reference 140). At its present meeting, the Committee calculated
the upper limit of the ADI on the basis of antimicrobial activity from
the formula given on p. using the MIC50 of the most sensitive,
relevant bacterial species of the human gut flora, Fusobacterium,
which was 0.5 µg/ml:
0.5 µg/ga × 220 gb
Upper limit of ADI = ------------------------
0.4c × 1d × 60 kge
= 4.6 µg/kg bw
where a is the MIC50 value for Fusobacterium spp. as the most
sensitive, relevant bacterial species, b is the mass of the colonic
content, c is the bioavailable fraction of thamphenicol in the
gastrointestinal tract, d is the safety factor, and e is the human
body weight.
4. EVALUATION
The Committee established an ADI of 0-5 µg/kg bw on the basis of
the microbiological end-point. This ADI is one-tenth the toxicological
ADI of 0-50 µg/kg bw that would be derived on the basis of the NOEL of
5 mg/kg bw per day in the two-year study of carcinogenicity in rats
and a safety factor of 100.
5. REFERENCES
Ando, J., Ishihara, R., Imai, S., Takano, S., Kitamura, T., Takahashi,
M., Yoshida, M. & Maekawa, A, (1997) Thirteen-week subchronic toxicity
study of thiamphenicol in F344 rats. Toxicol. Lett., 91, 137-146.
Kitamura, T., Ando, J., Ishihara, R., Iijima, T., Nishimura, S.,
Yoshida, M., Takahashi, M. & Maekawa, A. (1997) Lack of
carcinogenicity of thiamphenicol in F344 rats. Food Chem. Toxicol.,
37, 1075-1080.
Maekawa, A. (1998a) Thirteen-week toxicity study of thiamphenicol in
rats. Individual data. Unpublished report. Submitted to WHO by Sasaki
Institute, Tokyo, Japan.
Maekawa, A. (1998b) Carcinogenicity study of thiamphenicol in rats.
Individual data. Unpublished report. Submitted to WHO by Sasaki
Institute, Tokyo, Japan.