919. Sarafloxacin (WHO Food Additives Series 41)

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

SARAFLOXACIN

First draft prepared by
Dr P.L. Chamberlain
Center for Veterinary Medicine
Food and Drug Administration, Rockville, Maryland, USA

1. Explanation
2. Biological data
2.1 Biochemical aspects
2.1.1 Absorption, distribution, and excretion
2.1.2 Biotransformation
2.2 Toxicological studies
2.2.1 Acute toxicity
2.2.2 Short-term toxicity
2.2.3 Long-term toxicity and carcinogenicity
2.2.4 Genotoxicity
2.2.5 Reproductive toxicity
2.2.6 Special studies on microbiological effects
2.2.7 Special studies on ecotoxicity
2.3 Observations in humans
3. Comments
4. Evaluation
5. Acknowledgements
6. References

1. EXPLANATION

Sarafloxacin is a fluoroquinolone antibacterial agent that acts
by inhibiting the activity of DNA gyrase. It is used for treatment and
control of bacterial infections in poultry caused by Escherichia
coli and Salmonella spp. Sarafloxacin has not previously been
reviewed by the Committee.

The structure of sarafloxacin is shown in Figure 1.

2. BIOLOGICAL DATA

2.1 Biochemical aspects

2.1.1 Absorption, distribution, and excretion

Mice

Three groups of 12 female mice were treated with
¹⁴C-sarafloxacin base, as follows: Animals in the first two groups
were given a single dose of 10 mg/kg bw, one group intravenously and
the other by gavage; animals in the third group were given a dose of
100 mg/kg bw by gavage. Urine and faeces were collected from the mice
daily for three days. Compliance with the principles of GLP was not
required for this study. The quality and design of the study was
consistent with current scientific standards.

Absorption of the parent drug was estimated from the content in
0-24-h urine samples: 48% (range, 27-73%) of the parent drug was
absorbed after the 10 mg/kg bw dose and 34% (range, 29-38%) after the
100 mg/kg dose. Within three days of administration of the single
intravenous dose, 49% was excreted in the urine and about 44% was
eliminated in the faeces. After oral administration of the same dose,
urinary excretion accounted for about 25% and faecal excretion for
80%. Mice given 100 mg/kg bw orally eliminated 18% in the urine and
74% in the feces. Almost all of the radiolabel was excreted during the
first 24 h after oral or intravenous administration (Merits & Bopp,
1985a).

Rats

Five groups of 18 Sprague-Dawley rats of each sex were treated
with sarafloxacin as follows: One group received a single intravenous
dose of 20 mg/kg bw; three groups received a single oral dose of 20,
75, or 275 mg/kg bw; and animals in the fifth group received an oral
dose of 1000 mg/kg bw daily for 14 consecutive days. Blood samples
were collected from four rats in each group just before treatment and
0.5, 1, 2, 4, 6, 8, 12, and 24 h after treatment on day 1 for the
groups receiving the single dose and on days 1 and 14 for the 14-day
treatment group. Plasma and urine samples were assayed for
sarafloxacin base by high-performance liquid chromatography. The
pharmacokinetic parameters determined from this study are presented in
Table 1. A comparison of the 0 to infinity area under the
concentrationœtime curve (AUC) after a single intravenous or oral dose
of 20 mg/kg bw sarafloxacin indicated that its bioavailability was
about 12%. A plot of the AUC against dose was linear up to 275 mg/kg
bw but deviated from linearity at 1000 mg/kg bw. Compliance with the
principles of GLP was not required for this study. The quality and
design of this study were consistent with current scientific standards
(Patterson, 1985).

Table 1. Pharmacokinetic parameters of sarafloxacin in rats

Dose Route V_d Elimination T_max C_max K_a K_e Apparent
(mg/kg (L/kg) t_1/2 (h) (h) (µg/ml) (h^-1) (h^-1) body clearance
bw) (ml/min per kg)

20 i.v. 5.3 2.0 - - - 0.3 30
20 Oral 60 3.0 1.0 0.3 3.0 0.3 270
75 Oral 70 2.0 2.0 0.6 1.0 0.4 470
275 Oral 250 7.0 2.0 0.9 2.0 0.1 420
1000 Oral 400 6.0 1.0 2.0 2.0 0.1 820
1000 Oral once/ 110 6.0 2.0 8.0 1.0 0.1 200
d for 14 d

i.v., intravenous

In a summary report of another study, Sprague-Dawley rats (number
and sex not stated) were given 10 mg/kg bw ¹⁴C-sarafloxacin orally.
Within three days, about 37% of the radiolabelled dose had been
excreted in the urine, while about 52% was recovered in faeces (Bopp,
1985a)

Rabbits

The absorption, metabolism, and excretion of ¹⁴C-labelled
sarafloxacin was studied in three-month-old female New Zealand white
rabbits. Two groups of three animals per group were treated orally by
gavage with 10 mg/kg bw of ¹⁴C-sarafloxacin base. A third group of
three animals received the same dose by intravenous administration.
Blood samples were collected 1, 3, 6, 12, and 24 h after oral
administration from animals in one of the groups, and urine and faeces
were collected daily for five days from animals in the other groups.
Compliance with the principles of GLP was not required. The quality
and design were consistent with current scientific standards.

Within five days of oral administration, about 11% of the dose
was eliminated in the urine and about 79% in the faeces. Urinary
excretion after intravenous administration indicated that about 16% of
the oral dose had been systemically absorbed (Merits & Bopp, 1985b).

Dogs

Three groups of 14 dogs (strain, age, and sex not stated) were
given daily oral doses of 5, 25, or 125 mg/kg bw sarafloxacin base by
capsule. After one month, six dogs from each group were killed, and
plasma and cerebrospinal fluid were collected for high-performance
liquid chromatography. The remaining dogs continued to be treated
daily for a total of 90 days. The pharmacokinetic parameters
determined from this study are presented in Table 2. Compliance with
the principles of GLP was not required. The quality and design were
consistent with current scientific standards. It was also shown that
the dispositional kinetics of sarafloxacin in the dog are independent
of dose and treatment duration, while absorption of sarafloxacin
becomes less efficient with increasing dose (Granneman, 1985a) .

The tissue distribution of ¹⁴C-sarafloxacin base after a single
oral dose of 10 mg/kg bw was studied in four adult male beagle dogs.
The concentrations of radiolabel in tissues 2 and 6 h after treatment
are shown in Table 3. Compliance with the principles of GLP was not
required for this study. The quality and design were consistent with
current scientific standards (Bopp, 1985b).

The bioavailability of an oral dose of 200 mg sarafloxacin base,
equal to 20 mg/kg bw, was studied in six adult female dogs. The
compound was administered as a suspension, a solution, or a capsule.
Compliance with the principles of GLP was not required for this study.
The quality and design were consistent with current scientific
standards.

Table 2. Pharmacokinetics of sarafloxacin base after oral administration to dogs

Dose Mean half-life (h)^a AUC (µg h/ml) Mean ratio
(mg/kg bw) cerebrospinal
2 doses 24 doses 79 doses 2 doses 24 doses 79 doses fluid:plasma^b

5 5 6 6 9 9 10 0.2

25 5 5 6 30 31 30 0.2

125 5 6 6 104 108 106 0.2

^a Samples taken 1, 3, 6, and 24 h after treatment
^b Samples taken about 24 h after treatment

The bioavailability of the suspension and capsule were similar.
The values for the 0-32-h mean AUC for these formulations were 27 and
23 µg h/ml, respectively. The mean AUC for the solution was 52 µg
h/ml. The author cited the results of other studies which showed that
the bioavailability of an oral 10 mg/kg bw dose of the solution in
comparison with an equal intravenous dose was 58-70%. The relationship
between dose and bioavailability for the capsule formulation appears
to be non-linear or log-linear. The suspension resulted in AUC values
that were about one-half those obtained with the solution; however, in
a study with a dose of 10 mg/kg bw, the formulations were equivalent.
The basis for these differences was not readily apparent (Granneman,
1985b).

In another study, reported as a summary, dogs (breed, sex, and
number not stated) were given an oral or intravenous dose of 10 mg/kg
bw dose of ¹⁴C-sarafloxacin base. The extent of absorption was
estimated to be 73% on the basis of the AUCs and 89% on the basis of a
comparison of volumes of distribution. About 54% of the radiolabelled
dose was recovered from urine and about 27% from faeces. About 30% of
the intravenous dose was also eliminated in the faeces, suggesting
that biliary secretion is a factor in the disposition of the compound
(Bopp, 1985c).

Table 3. Concentrations (µg equivalents/g or ml) of
radiolabel in tissues of male dogs after oral
administration of ¹⁴C-sarafloxacin at a dose of
10 mg/kg bw

Tissue 2h 6h 24h

Liver 14 12 2
Kidney 16 14 1
Lung 6 5 1
Brain 0.4 0.7 0.3
Fat^a 0.6 0.5 0.6
Muscle^a 5 6 1
Bone^b 3 3 2
Retina/uvea 15 43 45
Blood 3 3 0.4
Plasma 3 3 0.4
Bile 154 454 420
Urine 89 412 188

^a Percent dose in muscle and fat calculated by
assuming that those tissues represent 46% and
10% of the body weight, respectively
^b Rib including marrow

Humans

Grannemann (1985b) cited data from a clinical study in which
capsules of the same lot given to the dogs were administered to
humans. The recoveries in urine ranged from 24% at 1.3 mg/kg bw to 10%
at 10 mg/kg bw. As the recoveries of sarafloxacin in human urine are
considered to provide an approximate estimate of absorption, since the
urine is the predominant route of elimination in humans, the results
indicate that the absorption rates in humans are considerably lower
than those in dogs.

A single oral dose of 100, 200, 400, or 800 mg sarafloxacin was
administered to 22 healthy male volunteers ranging in age from 20 to
39 years. Blood samples were taken before treatment and 0.25, 0.5, 1,
1.5, 2, 3, 4, 6, 8, 10, 12, 16, 24, 28, 32, and 48 h after treatment.
Urine was collected at 0-4, 4-8, 8-12, 12-16, 16-24, 24-32, and 32-48
h. Compliance with the principles of GLP was not required for this
study. The quality and design were consistent with current scientific
standards.

The plasma concentrations peaked 1.5-4 h after treatment and
declined biphasically, the terminal phase becoming dominant about 12 h
after treatment. The means of the individual peak concentrations after
the 100, 200, 400, and 800 mg doses were 140, 180, 240, and 350 ng/ml,
respectively. The corresponding dose-normalized peak concentrations
were 106, 62, 44, and 34 ng/ml per mg/kg. The average dose-normalized
AUC values for the 100, 200, 400, and 800 mg doses were 860, 570, 410,
and 350 ng h/ml per mg/kg, respectively. The decreases in the values
of the dose-normalized AUC and peak concentrations as a function of
dose provide evidence that the efficiency of absorption deceased by a
factor of about 3 as the dose was increased. The average terminal
phase half-lives were 9, 9, 10, and 11 h at the 100, 200, 400, and 800
mg doses, respectively

Elimination occurred mainly by renal excretion of unchanged drug.
The average renal clearances of the 100, 200, 400, and 800 mg doses
were 280, 290, 290, and 260 ml/min, respectively. Little difference
was seen between the groups. The average urinary recovery of unchanged
drug was 19, 14, 10, and 7% of the administered 100, 200, 400, and 800
mg doses, respectively. The extent of absorption of sarafloxacin
decreased from about 27-34% for the 100-mg dose to 11-13% for the
800-mg dose (Granneman, 1985c).

2.1.2 Biotransformation

Mice

The biotransformation of sarafloxacin after oral and intravenous
administration to mice in the study described above is shown in Table
4 (Merits & Bopp, 1985a).

Table 4. Metabolites of sarafloxacin identified in 24-h excreta of mice

Identity Mean percent of ¹⁴C dose Mean percent of ¹⁴C dose
of 10 mg/kg bw of 100 mg/kg bw
(oral)
Oral Intravenous

Urine Faeces Urine Faeces Urine Faeces

Unknown 0.3 0.2 1 ND 0.4 1
Unknown 0.1 0.1 0.3 0.2 ND ND
Sarafloxacin glucoronide 6 ND 9 ND 5 0.5
Sarafloxacin-N-acetyl 0.2 0.1 1 < 0.1 0.1 ND
Sarafloxacin 15 79 32 43 11 71

Sum 21.6 79.4 43.3 43.2 16.5 72.5
Total 101 86.5 89

ND, not detected

Rabbits

The biotransformation of sarafloxacin after oral and intravenous
administration to rabbits in the study described above is shown in
Table 5 (Merits & Bopp, 1985b).

Table 5. Metabolites of sarafloxacin identified in excreta of rabbits

Identity Mean percent of ¹⁴C dose

Oral Intravenous

Urine Faeces Urine Faeces

Sarafloxacin glucoronide 0.8 ND 3 ND
Unknown 0.3 ND 2 ND
Unknown < 0.1 ND 0.2 ND
Sarafloxacin-3'-oxo 0.2 < 0.1 2 0.2
Sarafloxacin-N-acetyl 0.3 ND 3 ND
Sarafloxacin 9 76 61 24

ND, not detected

Dogs

In the summary report cited in section 2.1.1, about 79% of the 10
mg/kg bw dose of ¹⁴C-sarafloxacin base was excreted as unmetabolized
parent drug in urine and faeces. In bile, the unchanged parent drug
and its glucuronide were found in about equal proportions (Bopp,
1985c).

Humans

The pharmacokinetics and metabolism of sarafloxacin were studied
in two groups of six volunteers given a single oral dose of 100 or 200
mg sarafloxacin and two groups of five volunteers given a single oral
dose of 400 or 800 mg. Compliance with the principles of GLP was not
required for this study. The quality and design were consistent with
current scientific standards.

The metabolism of sarafloxacin appears to involve mainly
oxidative degradation of the piperazinyl substituent, first producing
3'-oxo-sarafloxacin. Subsequent oxidation produces an ethylene
diamine-substituted quinolone, which in turn is oxidized to an
aminoquinolone. The plasma concentrations of the ethylene
diamine-substituted quinolone parallel those of the parent drug, but
the average AUC for the quinolone was consistently only about 6% that
of sarafloxacin. The concentration of the aminoquinolone in plasma and

urine was considerably lower than that of the ethylene
diamine-substituted quinolone. Owing to its weak fluorescence,
3'-oxo-sarafloxacin was not detected in plasma. In urine, the major
drug-related peak was sarafloxacin, accounting for 75-80% of all
urinary metabolites. After sarafloxacin, the predominant metabolite in
urine was tentatively identified as 3'-oxo-sarafloxacin, which
occurred at concentrations that were typically one-third to one-fourth
those of sarafloxacin. The total urinary recovery of parent drug plus
metabolites was low and dose-dependent, decreasing from 24 to 10% as
the dose increased from 100 to 800 mg. The extent of the decrease was
similar to that in the dose-normalized AUC. Collectively, the
aminoquinolone, the ethylene diamine-substituted quinolone, and their
conjugates accounted for < 7% of the urinary excretion (Granneman,
1985c).

2.2 Toxicological studies

2.2.1 Acute toxicity

The results of studies of the acute toxicity of sarafloxacin are
presented in Table 6. Compliance with the principles of GLP was not
required for these studies. The quality and design were consistent
with current scientific standards.

Table 6. Acute toxicity of orally administered
sarafloxacin in male rodents

Species Formulation LD₅₀ Reference
(mg/kg bw)

Mouse Suspension 18 000 Hahn (1991)
Suspension > 8000 Hahn (1991b)
Capsule > 8000 Hahn (1991c)
Suspension > 8000 Majors (1985)
Suspension > 5000 Hahn (1991)
Suspension > 5000 Hahn (1991e)

Rat Suspension > 8000 Hahn (1991a)

2.2.2 Short-term toxicity

Mice

In a study of dietary palatability, sarafloxacin was administered
to four groups of five CD-1 mice of each sex, four to five weeks old,
as a dietary admixture for 15 consecutive days. The diets contained
about 5000, 10 000, 25 000, or 50 000 mg/kg feed of sarafloxacin,
providing doses equivalent to 1250, 2500, 6250, or 12 500 mg/kg bw per
day, respectively, of the base. An untreated control group of five

males and five females was fed a basal diet. Compliance with the
principles of GLP was not required for this study. The quality and
design were consistent with current scientific standards.

Survival, general condition, clinical signs, body weight, food
consumption, and body-weight gain were evaluated. No overt signs of
toxicity or mortality were reported in animals consuming diets
containing up to 10 000 mg/kg feed. Decreased feed consumption and
body-weight gain were the only treatment-related effects observed in
rats consuming the diets containing 25 000 and 50 000 mg/kg feed
(Weltman, 1989).

Rats

In a study of dietary palatability, sarafloxacin was administered
to four groups of five CD rats of each sex, four to five weeks old, as
a dietary admixture for two weeks. The diets contained about 1000,
5000, 10 000, and 50 000 mg sarafloxacin base per kg feed, providing
doses equal to 15, 480, 850, or 3000 mg/kg bw per day, respectively.
An untreated control group of five males and five females was fed a
basal diet. Compliance with the principles of GLP was not required for
this study. The quality and design were consistent with current
scientific standards.

Clinical signs, body weight, food consumption, and body-weight
gain were evaluated. No overt signs of toxicity or mortality were
reported in animals consuming diets containing up to 10 000 mg/kg
feed. Alopecia, emaciation, dehydration, decreased feed consumption,
and body-weight gain were treatment-related effects observed in rats
consuming 50 000 mg/kg feed (Weltman, 1988).

In a range-finding study, sarafloxacin was administered to six
groups of four CD rats of each sex, six weeks old, by gavage at doses
of 20, 50, 125, 275, 650, or 1500 mg/kg bw per day for 14 or 15 days.
The untreated controls received 10-ml doses of the vehicle, 0.2%
hydroxypropylmethylcellulose. Compliance with the principles of GLP
was not required for this study. The quality and design were
consistent with current scientific standards. Clinical signs,
ophthalmological parameters, body weight, food consumption, and
clinical and anatomical (gross and microscopic) pathological
appearance were evaluated. No overt signs of toxicity or deaths were
observed in rats treated at up to 650 mg/kg bw. At the highest dose,
the only abnormality noted was light-coloured faeces (Fort & Buratto,
1984).

A 90-day study with a one-month interim kill was conducted in
four groups of 25 Sprague-Dawley rats of each sex, which were given
sarafloxacin at doses of 20, 75, 280, or 1000 mg/kg bw per day by
gavage. Animals in an untreated control group received 10-ml doses of
the vehicle, 0.2% hydroxypropylmethylcellulose. After one month of
treatment, 10 animals of each sex per group were randomly selected for
the interim kill. The remaining animals continued receiving the drug
daily for a total of 90 days, when 10 rats of each sex per group were

killed and necropsied. The remaining animals were killed and
necropsied after a 30-day recovery period. This study was conducted in
accordance with the principles of GLP.

Clinical observations, body weight, food consumption, and
ophthal-moscopic changes were evaluated, and urinalysis, haematology,
clinical chemistry, organ weights, and anatomical (gross and
microscopic) examinations were carried out. The only treatment-related
effect observed in animals treated for one month was grossly enlarged
caeca in those at the intermediate and high doses, but no microscopic
alterations were detected in these enlarged caeca. The
treatment-related effects in animals treated for 90 days included
grossly enlarged caeca in males at 75 mg/kg bw per day and higher and
in females at 280 mg/kg bw per day and higher. No microscopic
alterations were detected in these enlarged caeca. Caecal enlargement
was not present in the rats that were necropsied at the end of the
one-month recovery period. At gross necropsy, swollen ears were
reported in rats treated for 90 days, in two rats at the low dose, one
at 75 mg/kg bw per day, one at 280 mg/kg bw per day, and three at the
high dose. Auricular chondritis was seen histologically in the three
females with swollen ears at the high dose. Microscopically, the
auricular chondritis was characterized by nodular cartilagenous
proliferation and by a cellular infiltrate of predominantly
mononuclear cells. Swollen ears were not reported in treated or
control animals at the one-month interim kill, but they were observed
during clinical examinations of control male and female animals
throughout the treatment period. As data were not available on
individual animals, it was not possible to determine how many animals
per group had swollen ears on clinical examination. At necropsy, the
incidences were 0, 2, 1, 1, and 3 animals in the control group and at
20, 75, 280, and 1000 mg/kg bw per day, respectively. In the absence
of a clear doseœresponse relationship for the incidence at necropsy,
this finding could not be attributed to treatment. Auricular
chondritis was observed histologically in three females at the high
dose. Three males at the high dose died during the study, and one of
these deaths may have been related to treatment; however, the cause of
death could not be determined owing to the presence of moderate
autolysis in several tissues from this animal. The other two deaths
were attributed to treatment accidents. The NOEL was 280 mg/kg bw per
day on the basis of auricular chondritis at 1000 mg/kg bw per day
(Creighton & Pratt, 1985a,b).

Dogs

In a range-finding study, six groups of two young adult (age not
stated) beagle dogs of each sex were given sarafloxacin in gelatin
capsules at doses of 8, 20, 50, 125, 300, or 800 mg/kg bw per day for
two weeks. Two positive control groups were included: in one, dogs
received nalidixic acid in a gelatin capsule at a dose of 50 mg/kg bw
per day, and in the other dogs received nalidixic acid at a dose of
125 mg/kg bw per day for two weeks. The negative control group
received empty gelatin capsules. Compliance with the principles of GLP

was not required for this study. The quality and design were
consistent with current scientific standards.

Clinical signs, body weight, food consumption, clinical
pathology, and anatomical (gross and microscopic) pathology were
evaluated. The treatment-related effects in the sarafloxacin-treated
groups included lachrymation and emesis (at 8-125 mg/kg bw per day);
emesis, salivation, lachrymation, reduced activity, dehydration,
increased serum activity of alanine and aspartate aminotransferases
and alkaline phosphatase, a biliary sediment characteristic of
glucuronated drug in the gall-bladder, and centrilobular necrosis of
the liver (at 300 mg/kg bw per day). Evidence of hepatotoxicity was
also found microscopically in the livers of dogs at 125 mg/kg bw per
day. One female at the highest dose died, and the liver of this animal
showed moderate vacuolar degeneration. Flattening of the angle of the
radial-carpal joint when the limb is in a weight-bearing position was
observed from day 7-8 until the end of the study in both front legs of
both dogs at 800 mg/kg bw per day. No microscopic articular lesions
were observed. Decreased food consumption and body-weight gain,
increased serum alanine aminotrasferase activity and biliary sediment
characteristic of glucuronated drug in the gall-bladder were also
observed in this group.

The animals treated with 50 mg/kg bw per day nalidixic acid had
emesis, loose stools, bilirubinuria, increased serum activity of
alanine and aspartate aminotransferases and alkaline phosphatase,
lachrymation, dehydration, and centrilobular necrosis of the liver.
Emesis, salivation, lachrymation, soft stools, dehydration,
bilirubinuria, increased serum alanine aminotransferase activity,
weight loss, centrilobular necrosis of the liver, and decreased
activity were observed in both dogs at 125 mg/kg bw per day nalidixic
acid. In addition, the male experienced tremors, seizures, dyspnoea,
and flattening of the angle of the radialœcarpal joint in both front
feet from day 7 until the end of the study. No microscopic articular
lesions were observed (Kimura & Pratt, 1984).

In another range-finding study, five groups of one
three-month-old beagle dog of each sex were given sarafloxacin in
gelatin capsules at doses of 2, 20, 50, 125, or 300 mg/kg bw per day
for two weeks. Two positive control groups were included, receiving
nalidixic acid at a dose of 50 or 125 mg/kg bw per day. The negative
control group received empty gelatin capsules. Compliance with the
principles of GLP was not required for this study. The quality and
design were consistent with current scientific standards.

Clinical signs, food consumption, body weight, and the results of
ophthalmological examinations and clinical and anatomical (gross and
micro-scopic) pathology were evaluated. In the sarafloxacin-treated
groups, one male at 300 mg/kg bw per day was killed on day 8 in a
moribund condition, and decreased food consumption and body-weight
gain were observed in animals at this dose. Emesis and flattening of
the angle of both radial-carpal joints were observed at 125 and 300
mg/kg bw per day. Moderate to severe vesicular arthropathic changes of

the articular cartilage were observed microscopically in animals at
300 mg/kg bw per day. These lesions were present on the proximal and
distal femoral extremities and the proximal humeral and tibial
surfaces. The female at the high dose experienced a convulsion-like
state on day 8. In the nalidixic acid-treated groups, one male at 125
mg/kg bw per day was killed on day 12 in a moribund condition. The
treatment-related effects in both dose groups included hepatotoxicity
(increased activity of serum liver enzymes and mild-to-severe liver
necrosis and degeneration), flattening of the radial-carpal joints,
and vesicular arthropathic changes of the articular cartilage, which
were of equal severity and in the same locations as in the
sarafloxacin-treated dogs (Dudley & Buratto, 1984).

A 90-day study of toxicity with a one-month interim kill was
conducted in groups of seven 9-14-month-old beagle dogs of each sex
which received sarafloxacin base in gelatin capsules at doses of 5,
25, or 125 mg/kg bw per day. A control group received empty gelatin
capsules. After 30 days of treatment, three dogs of each sex per group
were killed and necropsied; the remaining animals continued to receive
treatment for 90 days before they were killed and necropsied. The
study was conducted in accordance with the principles of GLP. Clinical
signs, body weight, and food consumption were evaluated, and
electroretinography, ophthalmoscopy, electrocardiography, clinical
pathology, organ weighing, and anatomical (gross and microscopic)
pathology were carried out.

Males treated for one month had a dose-related decrease in
body-weight gain. Food consumption was difficult to assess owing to
numerous instances of spilled food, so it was not known whether the
decrease in body-weight gain was due to decreased food intake or to
treatment. The mean serum globulin concentration was decreased from
control values in all treated males and females, and the differences
were statistically significant for males at all doses; however, the
decreases did not show a doseœresponse trend in animals of either sex,
nor were the lower concentrations outside the normal range for this
species. Decreased globulin concentrations have been reported
elsewhere in animals treated with antimicrobial drugs and have been
attributed to reduced stimulation of the immune system after reduction
of the resident microbial population. The finding is thus considered
to be a treatment-related effect.

Numerous instances of food spilling by both treated and control
animals were reported during treatment for 90 days, and animals at the
intermediate and high doses were involved in a greater number of
instances; however, no significant difference in body-weight gain was
seen between days 0 and 91. The mean serum globulin concentrations
were comparable to those of controls for males at the low and
intermediate doses but were slightly decreased for males at the high
dose. The mean serum globulin concentrations of females at the
intermediate and high doses were statistically significantly lower
than the control values, whereas the value for those at the low dose
was comparable to that of controls. The NOEL was 5 mg/kg bw per day on
the basis of decreased serum globulin concentrations in females at the

intermediate and high doses after 90 days of treatment (Kimura &
Tekeli, 1985a,b).

A 90-day study was conducted in groups of six four-month-old
beagle dogs which received gelatin capsules containing sarafloxacin
base at 0 or 200 mg/kg bw per day. An additional two groups of four
animals of each sex were given gelatin capsules containing 10 or 50
mg/kg bw per day. The study was conducted in accordance with the
principles of GLP. During the first two weeks of the study,
sarafloxacin, as the hydrochloride salt, was administered as the
actual weight, without regard to the concentration of free base. This
resulted in actual doses of the free base that were about 80% of those
intended, i.e. 8, 40, and 160 mg/kg bw per day. For the remainder of
the study, the doses were adjusted to the intended original doses of
10, 50, and 200 mg/kg bw per day (US Food and Drug Administration,
1995). The Committee considered that the lower values corresponded to
actual intake during the study. Clinical signs, body weight, and food
consumption were evaluated, and ophthalmos-copy, electrocardiography,
electroretinography, urinalysis, haematology, clinical chemistry,
organ weighing, and anatomical (gross and microscopic) pathology were
carried out.

The treatment-related effects included erythema of the ear flaps
and muzzle in animals at the intermediate and high doses. During weeks
9-14, 50% of animals at the high dose were affected, and generalized
erythema was observed in one male at this dose. Swelling around the
eyes, eyelids, and earflaps was seen in two animals of each sex at the
high dose during the first three weeks of treatment. The swelling
occurred 2-6 h after treatment but was not evident the next day
(before treatment). An increased incidence of ocular discharge was
observed in females at the high dose. No other treatment-related
effects were found. A slight decrease in mean serum globulin
concentrations was observed in males at the high dose; in females, a
statistically significant decrease in mean serum globulin
concentration was observed at the intermediate and high doses; the
mean serum globulin concentration of females at the low dose was
comparable to that of controls. The investigator proposed that the
decreases in serum globulin were due an effect on the gastrointestinal
flora. Thus, a treatment-related reduction in the flora may have
caused a secondary reduction in immunoglobulin and acute-phase protein
synthesis in response to diminished antigenic stimulation. The
Committee considered the effect to be related to treatment. The NOEL
was 8 mg/kg bw per day of sarafloxacin base on the basis of the facial
swelling and erythema observed in males and females at the
intermediate and high doses and the decreased serum globulin
concentrations in females at these doses (Kiorpes, 1991).

2.2.3 Long-term toxicity and carcinogenicity

Mice

Sarafloxacin was administered to groups of 60 CD-1 mice of each
sex as a dietary admixture at 1000, 5000, or 20 000 mg/kg feed
(equivalent to 150, 750, and 3000 mg/kg bw per day). An additional 10
animals of each sex were included in each group for haematological
evaluations and sacrifice at 52 weeks. The carcinogenicity phase was
terminated at 78 weeks because of high mortality. The study was
conducted in accordance with the principles of GLP. Mortality,
clinical signs, body weights, food consumption, haematology and
anatomical (gross and microscopic) pathology were evaluated.

Mortality was increased in mice of each sex at the intermediate
and high doses, the survival at 78 weeks being 23% for males at the
high dose, 28% for females at the high dose, 45% for males at the
intermediate dose, and 40% for females at the intermediate dose; the
survival of mice at the low dose was comparable to that of controls.
Abdominal distension and increased faecal volume and moisture were
noted consistently in animals that died prematurely. Nephrotoxic
effects (epithelial vacuolation with tubular dilatation and chronic
interstitial nephritis) were observed in females at the intermediate
and high doses. Gall-bladder calculi and urolithiasis were found in
males at the high dose. Caecal dilatation was observed in males and
females at all doses, and caecal torsion was also observed in males
and females at the intermediate and high doses. The effect was
attributed to a local effect of large doses of the compound on caecal
microflora. No treatment-related pathological effects were observed in
animals at the low dose. There was no evidence of carcinogenicity
(Procter et al., 1991).

Rats

Sarafloxacin was incorporated into the feed of Sprague-Dawley
rats at concentrations of 1000, 10 000, or 25 000 mg/kg. Twenty rats
of each sex were used to test toxicity (52 weeks) and 65 of each sex
to test carcinogenicity (104 weeks). Two control groups of the same
numbers of rats of each sex were included in both phases of the study.
Five satellite groups consisting of 10 rats of each sex (two control
groups and three treated groups) were included for analysis of the
plasma concentrations of the drug after about 52 and 103 weeks and for
laboratory investigations after about 13 and 38 weeks of treatment.
The study was conducted in accordance with the principles of GLP.

Clinical examinations, mortality, food consumption, and
body-weight gain were evaluated in the toxicity phase, with
ophthalmoscopic examinations, urinalyses, clinical chemistry,
haematology, organ weighing, and anatomical (gross and microscopic)
pathology. Drug intake over 52 weeks was equal to 61, 670, and 1700
mg/kg bw per day. A treatment-related decrease in mean body-weight
gain was observed in animals at the high dose, despite the fact that
overall (weeks 1-52) food intake was greater than in controls by 5% in

males and 11% in females. Statistically significant increases in blood
urea nitrogen and creatinine concentrations were observed in females
at the high dose at week 52 and in males at the high dose at week 51.
The total protein values were statistically significantly decreased in
comparison with controls for males at all doses at each sampling
period (weeks 25/26 and weeks 51/52). The decreased values were
characterized by a statistically significant decrease in globulin and
a relatively unchanged albumin concentration. Total protein and
globulin concentrations were statistically significantly decreased in
females at the intermediate and high doses at week 51/52 only. The
absolute and relative kidney weights of females at the high dose were
statistically significantly increased in comparison with controls, and
a statistically significant increase in relative pituitary weights was
observed in these animals, with gross enlargement of the pituitary in
6/20 rats. A slight increase in relative pituitary weight was observed
in females at the intermediate dose. Dilatation of the caecum was
observed in nearly all rats treated with 10 000 or 25 000 ppm
sarafloxacin. The treatment-related histopathological findings
included tubular nephropathy in 10/20 males and females at the high
dose and in 1/20 females at the intermediate dose. Tubular nephropathy
was characterized by the presence of filamentous crystalline material
in the collecting ducts with associated basophilia and dilitation of
both cortical and medullary tubules. The severity varied, the most
severe cases being associated with interstitial inflammatory
infiltrates, fibrosis, rare focal tubular necrosis, and occasional
apoptosis or mitosis in the tubules. In addition, minimal glomerular
sclerosis and focal granulomatous inflammation were observed. The
tubular nephropathy was different from that of chronic progressive
nephropathy, which was characterized by dilated renal tubules with
proteinaceous casts, mononuclear infiltration, and focal fibrosis with
atrophic tubules. Other changes seen in chronic progressive
nephropathy were mineralization of the pelvis, occasional cysts, and
sclerotic glomeruli. No histopathological changes were observed in
caecae that were grossly dilated. Therefore, as in the 90-day study,
this effect was attributed to a local effect of large doses of the
compound on caecal microflora. The NOEL was 61 mg/kg bw per day on the
basis of the nephrotoxicity observed in males and females at the
intermediate and high doses and the decreased body-weight gain of
males and females at the high dose (Smith, 1990).

The parameters evaluated in the carcinogenicity phase were
clinical signs, body-weight gain, food consumption, ophthalmological
and haematological parameters, organ weights, and anatomical (gross
and microscopic) pathological changes. The mean intake of sarafloxacin
in the carcinogenicity phase was 54, 580, and 1500 mg/kg bw per day.
Reduced body-weight gain and tubular nephropathy were seen in females
at the high dose, which were considered to be related to treatment.
Tubular nephropathy similar to that seen in the 52-week study was also
observed in males at the intermediate and high doses. Increased
relative and absolute kidney weights were observed in males and
females at the high dose, and a treatment-related increase in relative
kidney weight was observed in females at the intermediate dose.
Treatment-related dilatation of the caecum was observed in males and

females at all doses. No histopathological changes were seen in caecae
that were grossly dilated, but the Committee considered this to be a
treatment-related effect. There was no evidence of carcinogenicity
(Smith et al., 1991).

2.2.4 Genotoxicity

The results of assays for the genotoxicity of sarafloxacin are
summarized in Table 7. All of the studies were conducted in accordance
with the principles of GLP.

2.2.5 Reproductive toxicity

(i) Multigeneration reproductive toxicity

Rats

A three-generation study of reproductive toxicity was conducted
in Sprague-Dawley rats, each generation consisting of 30 males and 30
females per group. The animals were treated orally by gavage with
sarafloxacin base at 75, 275, or 1000 mg/kg bw per day, beginning a
minimum of 70 days before breeding. A control group consisting of 30
males and 30 females received daily 10-ml doses of the vehicle, 0.2%
hydroxypropylmethylcellulose. Each generation was permitted to produce
up to two litters. The study was conducted in accordance with the
principles of GLP. Adult animals were observed for clinical signs,
mortality, reproductive performance, body weights (weekly during
gestation and lactation), food consumption (weekly during gestation
and lactation), and organ weights; anatomical (gross and microscopic)
examinations were made post mortem. The litter parameters evaluated
were live birth and viability indices, sex ratios, general physical
condition, deaths, live litter size, and body weights; weaned pups
that were not selected for breeding in the subsequent generation were
necropsied.

Soft stools and/or whitish faeces resulting from faecal excretion
of sarafloxacin were observed in the parental animals at the
intermediate and high doses in all three generations. Gross necropsy
of the F₀ animals that died during the study or were killed as
scheduled revealed red contents in the gastrointestinal tract and/or
red foci in the stomach; however, most of these observations were made
in animals that had died prematurely. In animals killed at scheduled
sacrifice, the occurrence of these lesions did not show a clear
doseœresponse relationship and was sporadic in all groups, including
the controls (one animal). These lesions were not considered to be
related to treatment. Microscopic examinations were not performed on
the grossly affected intestinal portions of the animals that died or
on the affected stomachs. The parental animals of the second
generation at the high dose that died also had red contents in the
intestine, but histopathological examination was not performed. Some
animals at the low and intermediate doses had gastric and intestinal
lesions at scheduled sacrifice, but these also were not examined
microscopically. In female parental animals of the first generation at

Table 7. Results of genotoxicity testing of sarafloxacin

Test system Test object Concentration Results Reference

In vitro
Forward mutation^a Chinese hamster 100-1000 Positive (+S9) Young (1985)
ovary cells µg/ml Negative (-S9)
(hprt locus)

Unscheduled DNA Rat primary 1-500 µg/ml Positive Cifone (1985)
synthesis hepatocytes

Chromosomal Chinese hamster 120-2000 Positive (+S9) Hemalatha
aberrations ovary cells µg/ml (1988)
50-800 µg/ml Negative (-S9)

In vivo/in vitro
Unscheduled DNA Rat primary 250-2500 Negative^c Cifone (1988)
synthesis hepatocytes mg/kg orally^b

Micronucleus Mouse bone 2000, 4000, Negative Diehl (1994)
formation^d marrow 8000 mg/kg
bw orally^e

S9, 9000 × g fraction from rat liver
^a With and without S9
^b Three rats per concentration
^c Justification for use of the solvent/vehicle was not provided, nor were the criteria
by which the doses were selected (e.g. preliminary range finding study).
^d Five males and five females at the low and intermediate doses, and bone-marrow cells
harvested after 24 h; 15 males and 15 females at the high dose, and cells harvested
at 24, 48, and 72 h after treatment
^e The high dose represented the maximum applicable dose. Neither general nor bone-marrow
toxicity was observed.

the intermediate and high doses, the absolute and relative liver
weights were significantly decreased in comparison with controls. The
relative liver weights were also significantly decreased in male and
female parental animals of the second generation and in males of the
third generation at the intermediate and high doses. Parental females
of the third generation at the high dose had decreased relative liver
weights. Because this effect was seen in all three generations and was
dose-related, it is considered to be related to treatment with 275
mg/kg bw per day and higher. The NOEL for maternal toxicity was 75
mg/kg bw per day on the basis of decreased liver weights in rats at
the intermediate and high doses. No treatment-related effects were
observed on reproductive parameters, litter parameters, or fetal
morphology at doses up to 1000 mg/kg bw per day (Lehrer, 1991).

(ii) Developmental toxicity

Rats

The developmental toxicity of sarafloxacin was evaluated by daily
oral administration of the compound to pregnant rats during days 6œ15
of gestation. Four groups of 20 pregnant Sprague-Dawley rats were
treated orally by gavage with 20, 75, 280, or 1000 mg/kg bw per day. A
control group of 20 pregnant females received a daily 10-ml dose of
the vehicle, 0.2% hydroxymethyl-cellulose, on the same schedule as the
treated animals. On day 20 of gestation, the rats were killed and the
fetuses removed. The study was conducted in accordance with the
principles of GLP. Dams were evaluated for clinical signs, body weight
(on gestation days 6, 9, 12, 15, and 20), food consumption (on
gestation days 6-9, 9-12, and 12-15), numbers of implants, and percent
of nonviable implants. The parameters evaluated in the litters were
sex ratio, group mean weights, and external, visceral, and skeletal
morphology.

No maternal toxicological effects were seen, and there was no
evidence of teratogenicity. The investigator noted that the results of
a study of the bioavailability of sarafloxacin in rats (see above)
indicated that at 20 mg/kg bw per day about 12% of a single oral dose
of sarafloxacin was absorbed. Comparable figures for absorption of the
doses of 75, 275, and 1000 mg/kg bw per day would be about 6.5, 7, and
4%, respectively. Therefore, the intended 50-fold range of test doses
in this study more closely approximated 16-fold (2.3-37 mg/kg bw per
day). Although no maternal toxic effects were seen, the highest dose
administered was considered to be sufficiently high to conclude that
sarafloxacin is not teratogenic in rats. The NOEL for maternal
toxicity and teratogenicity was 1000 mg/kg bw per day (Lehrer, 1985;
Patterson, 1985).

Rabbits

A study of developmental toxicity was conducted in three groups
of 18 artifically inseminated New Zealand white rabbits given
sarafloxacin by gavage once daily on gestation days 6-18 at doses of
15, 35, or 75 mg/kg bw per day. A concurrent control group of 18

females received a daily 2-ml dose of 0.5% aqueous methylcellulose
vehicle. The dams were observed for clinical signs and survival. The
ovaries were examined for the number of corpora lutea, and the uteri
were examined for the location of viable and nonviable fetuses, early
and late resorptions, and the total number and distribution of
implantation sites. The litter parameters evaluated were group mean
weight, sex ratio, and external, visceral, and skeletal morphology.

Fourteen females aborted between gestation days 21 and 29; three
of the females were receiving the low dose, four the intermediate
dose, and seven the high dose. These abortions were considered to be
related to treatment as a secondary effect of the maternal toxicity.
On gestation day 6, all females at the high dose and most of those at
the low and intermediate doses showed treatment-related decreases in
defaecation and urination. Soft stools were observed in two animals at
the low dose, four at the intermediate dose, and 10 at the high dose.
Diarrhoea was observed in single animals at the low and intermediate
doses and in four animals at the high dose. Body-weight gain was
slightly decreased in animals at the low dose in comparison with
control values during the last six days of treatment and during the
first six days after cessation of treatment. Animals at the
intermediate dose lost weight throughout the treatment period and
during the initial six days after cessation of treatment. Loss of body
weight throughout the gestation period was severe in animals at the
high dose.

External examination showed that six fetuses from one litter at
the high dose had malformations, reported as carpal and/or tarsal
flexure. Visceral examination revealed that five fetuses from one
litter at the high dose had malformations, reported as hydrocephaly.
Six fetuses from one litter at the high dose had skeletal
malformations, reported as cartilagenous skeletal anomalies. Three
malformations (two external and one skeletal) were observed in two
litters at the intermediate dose. The numbers of malformations and the
numbers of litters affected at the low dose were comparable to those
in controls. The only parameters not affected by treatment were the
mean numbers of corpora lutea, implantation sites, viable fetuses per
litter, and mean postimplantation loss at scheduled removal of
fetuses. A dose-related decrease in mean fetal weight occurred at
doses of 35 and 75 mg/kg bw per day. The NOEL for teratogenicity and
fetotoxicity was 15 mg/kg bw per day. The teratogenic effects were
considered to be secondary to maternal toxicity and not directly
attributable to treatment. No NOEL for maternal toxicity was
identified (Lehrer & Tekeli, 1986).

2.2.6 Special studies on microbiological effects

In vitro

In a study of several microbiological end-points, the minimum
inhibitory concentrations (MICs) of sarafloxacin were determined
against human clinical bacterial isolates (Table 8); the effect of
inoculum size on the potency of sarafloxacin against human clinical

isolates was investigated in vitro (Table 9); and the effect of pH
on the potency of sarafloxacin against human clinical isolates of
aerobic (Table 10) and anaerobic bacteria (Table 11) was investigated
in vitro. Compliance with the principles of GLP was not required for
this study. The quality and design were consistent with current
scientific standards (Prabhavathi, 1984).

The microbiological activity of four metabolites of sarafloxacin,
N-acetyl sarafloxacin, N-formylsarafloxacin, 3'-oxosarafloxacin,
and a sulfamic acid conjugate of sarafloxacin, was determined in MIC
assays with Staphylococcus spp., Streptococcus spp.,
Micrococcus spp., E. coli, Klebsiella pneumoniae, Providencia
stuartii, Pseudomonas spp., Acinetobacter calcoaceticus, and
Lactobacillus cosei. The MIC₅₀ values for the metabolites varied
with the organism tested but were in general significantly higher than
those seen with sarafloxacin against E. coli. Therefore, the
metabolites were less microbiologically active than the parent
compound (Dr Stephan Schutte, Global Project and Registration Manager,
Fort Dodge Animal Health Holland, personal communication, 1998).

The frequency of the development of resistance to sarafloxacin by
human clinical isolates of E. coli Juhl, S. aureus 730a, and
P. aeruginosa 5007 was studied in two ways. In one method, overnight
broth cultures were grown from single colonies. A 0.1-ml sample of the
undiluted culture and 0.1 ml of 10-fold dilutions of the cultures were
plated on agar plates containing four or eight times the MIC of
sarafloxacin. Viable cells were counted on drug-free plates. The
plates were incubated for 72 h at 35°C, and the colonies on plates
containing the drug were counted. Resistant colonies were picked and
streaked on medium containing four or eight times the MIC of
sarafloxacin in order to confirm resistance, and the MIC of
sarafloxacin against resistant mutants was then determined. Finally,
the resistant mutants were subcultured for 10 consecutive transfers on
drug-free medium to determine the stability of the resistance.

The second method involved transfer of the test organisms to
broth containing increasing concentrations of sarafloxacin. The
inoculum used for each step of the serial transfer was a 10^-3
dilution of the broth from the well containing the highest
concentration of antibiotic that allowed growth about equal to that in
the control well, containing no antibiotic. This procedure was
repeated for 10 transfers, and each organism was also streaked for
isolation on agar plates to verify its purity and presumptive
identity. After the final transfer, the susceptibility of the organism
to sarafloxacin was determined by the agar dilution method. Resistant
mutants obtained by this procedure were subcultured to drug-free broth
medium for 10 consecutive transfers in order to determine the
stability of the resistance, and the MICs were determined by the agar
dilution method. The results of these studies are shown in Tables 12
and 13. Compliance with the principles of GLP was not required for
this study. The quality and design were consistent with current
scientific standards (Prabhavathi, 1984).

Table 8. Minimum inhibitory concentrations (MICs) for sarafloxacin
against human clinical isolates (10⁴ bacteria/inoculation)

Organism No. MIC₅₀ MIC₉₀
(µg/ml) (µg/ml)

Staphylococcus aureus 70 0.25 0.25
Staphylococcus epidermidis 43 0.25 0.5
Staphylococcus spp. 12 0.25 0.5
Streptococcus pyogenes (group A) 13 0.5 2
Streptococcus agalactiae (group B) 13 1 2
Streptococcus (group C) 3 0.5 1
Streptococcus (group D; enterococcus) 58 2 4
Streptococcus (group G) 5 0.5 1
Streptococcus pneumoniae 5 1 2
Corynebacterium spp. 10 2 16
Pseudomonas aeruginosa 53 0.25 1
Pseudomonas spp. 3 0.125 0.5
Pseudomonas cepacia 1 2 2
Pseudomonas maltophilia 7 1 4
Acinetobacter anitratus 8 0.5 0.5
Achromobacter xylosoxidans 1 16 16
Acinetobacter spp. 5 0.125 0.125
Aeromonas hydrophila 2 <0.031 <0.031
Cedecea davisae 1 0.062 0.062
Chromobacterium spp. 1 0.25 0.25
Miscellaneous gram-negative bacteria 1 0.062 0.062
Escherichia coli^a 140 <0.031 0.062
Enterobacter aerogenes^a 10 0.062 0.125
Enterobacter cloacae^a 18 <0.031 0.5
Enterobacter agglomerans 1 0.062 0.062
Klebsiella penumoniae 34 0.062 0.125
Klebsiella oxytoca 4 <0.031 4
Klebsiella rhinoscleromatis 2 <0.031 <0.031
Citrobacter freundii 6 <0.031 0.5
Citrobacter diversus 2 <0.031 <0.031
Citrobacter spp. 7 0.062 4
Proteus mirabilis 43 0.25 0.5
Morganella morganii 12 0.062 0.25
Providencia rettgeri 7 0.125 0.25
Providencia stuartii 13 0.062 0.25
Provindencia spp. 5 0.25 0.5
Serratia marcescens 1 0.25 0.5
Serratia liquefaciens 3 0.25 0.25
Shigella flexneri 4 <0.031 0.062
Shigella dysenteriae 1 <0.031 <0.031
Shigella boydii 1 0.5 0.5
Shigella sonnei 3 <0.031 <0.031

Table 8 (continued)

Organism No. MIC₅₀ MIC₉₀
(µg/ml) (µg/ml)

Salmonella typhimurium 5 <0.031 <0.031
Salmonella choleraesuis 1 0.62 0.062
Salmonella arizonae 2 <0.031 <0.031
Salmonella spp. 10 <0.031 0.062
Yersinia enterocolitica 2 <0.031 <0.031
Hafnia alvei 4 <0.031 0.125
Haemophilus influenzae 21 0.125 0.125
Neisseria gonorrhoeae 8 <0.015 0.5
Campylobacter fetus 4 2 4
Legionella spp. 3 0.25 0.25
Bacteroides fragilis^a 17 2 8
Bacteroides disiens^a 1 4 4
Bacteroides melaninogenicus^a 1 2 2
Bacteroides thetaiotaomicron^a 4 4 8
Bacteroides vulgatus^a 1 8 8
Bacteroides spp.^a 2 4 16
Clostridium difficile 1 8 8
Clostridium perfringens^a 8 0.5 1
Fusobacterium spp.^a 1 2 2
Peptostreptococcus spp.^a 3 0.125 1
Peptococcus spp.^a 2 0.5 1
Propionibacterium spp.^a 1 2 2
Veillonella spp. 1 4 4

^a Possible constituent of human intestinal microflora

Table 9. Effect of inoculum size on potency of sarafloxacin in vitro

Organism No. of Geometric mean MIC (µg/ml)
strains

10⁵ CFUs/ml 10⁷ CFUs/ml

E. coli^a 7 0.018 0.05
K. pneumoniae 3 0.04 0.19
S. marcescens 3 0.12 0.25
C. freundii 2 0.015 0.02
E. cloacae 3 0.03 0.19
P. mirabilis 2 0.25 0.25
P. vulgaris 1 0.06 0.12
M. morganii 2 0.08 0.25
P. stuartii 2 0.17 0.17
P. rettegeri 1 0.12 0.25
Acinetobacter spp. 3 0.04 0.05
P. aeruginosa 8 0.16 0.35
S. aureus 5 0.12 0.21
S. epidermidis 3 0.12 0.25
S. faecalis 5 1.3 7

CFU, colony-forming units
^a Possible constituent of human intestinal microflora

Table 10. Effect of pH on the potency of sarafloxacin against
aerobic bacteria in vitro

Organism No. of Geometric mean MIC (µg/ml)
strains
pH 6.5 pH 7.2 pH 8.0

E. coli^a 7 0.07 0.03 0.03
K. pneumoniae 3 0.06 <0.03 <0.03
S. marcescens 3 0.63 0.32 0.15
C. freundii 2 0.09 0.04 0.03
E. cloacae 3 0.08 0.04 0.04
P. mirabilis 2 0.5 0.25 0.25
P. vulgaris 1 0.5 0.12 0.12
M. morganii 2 0.25 0.09 0.12
P. stuartii 2 0.98 0.35 0.17
P. rettegeri 1 0.5 0.12 0.06
Acinetobacter spp. 3 0.25 0.12 0.20
P. aeruginosa 8 1 0.55 0.65
S. aureus 5 0.16 0.16 0.16
S. epidermidis 3 0.12 0.12 0.15
S. faecalis 5 1.51 1.15 1.15

^a Possible constituent of human intestinal microflora

Table 11. Effect of pH on the potency of sarafloxacin against
anaerobic bacteria in vitro

Organism No. of Geometric mean MIC (µg/ml)
strains
pH 6.6 pH 7.3 pH 8.1

B. fragilis^a 6 5.6 1.6 1.4
Bacteroides spp.^a 7 7.1 3.8 3.9
Fusobacterium spp.^a 2 2.2 3.1 1.6
Clostridium^a 3 0.4 0.4 0.4
C. Difficile^a 1 6.2 6.2 6.2
Peptococcus and 5 0.4 0.6 0.7
Peptostreptococcus
spp.^a

^a Possible constituents of human intestinal microflora

Table 12. Frequency of spontaneous resistance of human clinical
isolates to sarafloxacin

Organism (strain) Selected at Selected at
4 × MIC 8 × MIC

E. coli Juhl 1.0 × 10^-8 <2.0 × 10^-9
S. aureus CMX730a 1.2 × 10^-7 1.3 × 10^-8
P. aeruginosa 5007 3.2 × 10^-6 3.9 × 10^-7

The effects of sarafloxacin on five E. coli strains of human
origin were assessed in a gastrointestinal model system in vitro.
The study was conducted in accordance with the principles of GLP. The
strains of E. coli were obtained from the National Collection of
Type Cultures, London, United Kingdom, and Abbott Laboratories, North
Chicago, USA. The strains were E. coli NCTC 8603, 8761, 8783, 9437,
and ATCC 25922. The model was designed to simulate possible
inactivation of sarafloxacin by degradation and protein binding after
residues of food are ingested. Thus, 1 ml of an aqueous solution of
sarafloxacin was added to 9 ml of cooked meat medium, and the mixture
was incubated for 0.5 h at 37°C to simulate binding of residues to
tissues. Gastric fluid (sodium chloride, pepsin, and hydrochloric
acid) was then added in order to decrease the pH of the mixture to
about 3, and the mixture was incubated for 1 h at 37°C to simulate
conditions in the stomach. Intestinal fluid (monobasic potassium
phosphate, sodium hydroxide, pancreatin, and sodium hydroxide) and
bile salts (sodium cholate and sodium deoxycholate) were then added to
the mixture, which raised the pH to about 7. This mixture was allowed
to equilibrate under anaerobic conditions for 4-6 h at 37°C before
being inoculated with about 10⁶ bacteria from a fresh overnight
culture of E. coli. This mixture was incubated at 37°C for 18 h
under strictly anaerobic conditions.

Plate counts were performed on aliquots of the organism
suspension that was used as inoculum, the model contents immediately
after inoculation, and the model contents 18 h after inoculation. For
each strain, 10 individual colonies were picked off and subcultured.
The MIC of sarafloxacin was determined for each isolate by inoculation
of about 10⁵ bacteria into 1 ml of broth. The medium contained
sarafloxacin at 8, 4, 2, 1, 0.5, 0.25, 0.125, 0.625, 0.0313, 0.0156,
or 0.0078 µg/ml. The inoculated tubes were incubated aerobically
(unless otherwise stated) at 37°C overnight and assessed visually for
growth. The MIC was defined as the lowest concentration of
antimicrobial agent that resulted in complete inhibition of visible
growth. The MIC₅₀ was defined as the concentration of the
antimicrobial agent at which at least 50% of the tested strains were
inhibited. On the basis of the results of a range-finding study, the
doses selected for the definitive study were 0, 0.025, 0.07, 0.1, and
0.4 µg/ml.

The model gastrointestinal tracts were supplemented with
sarafloxacin over the range of 0.025-0.4 µg/ml. A later
high-performance liquid chromatography analysis of the model contents
revealed that the actual sarafloxacin concentrations were 10-fold
higher than the target doses (i.e. 0.25, 0.7, 1, and 4 µg/ml). No
strains survived at the highest concentration after 18 h of
incubation. Strains NCTC 8761, 8783, and 9434 survived at 1 µg/ml,
strain NCTC 8603 survived at 0.7 µg/ml, and strain ATCC 25922 survived
at 0.25 µg/ml. The MICs of test isolates from each strain selected
from models containing the highest concentration of sarafloxacin that
permitted survival were determined. The survival of bacteria in the
model system and changes in the sensitivity of the isolates to
sarafloxacin are shown in Table 14 (McConville, 1992a,b).

The effects of sarafloxacin on five strains of Bacteroides
fragilis and five strains of Bifidobacterium spp. of human origin
were assessed in the same gastrointestinal tract model system
described above. The study was conducted in accordance with the
principles of GLP. Table 15 shows the survival of B. fragilis
strains in model gut supplemented with sarafloxacin over the range of
2-16 µg/ml. Strains NCTC 9343, NCTC 9344, and NCTC 11625 survived at
concentrations < 16 µg/ml. Strains NCTC 8560 and NCTC 10581
survived at 8 µg/ml. All of the strains grew well in the absence of
sarafloxacin. The MICs of isolates of each strain obtained from the
model gastrointestinal tracts containing the highest concentration of
sarafloxacin that permitted survival were then determined.

For Bifidobacterium spp., the doses were chosen such that a
toxic effect of sarafloxacin could be observed and selective pressure
applied to surviving bacteria. B. adolenscentis, B. infantis,
B. breve, and B. longum all survived at concentrations < 16
µg/ml; B. angulatum survived at concentrations < 8 µg/ml,
although survival was about 0.01% of the inoculum. The MICs of these
isolates and the MIC₅₀ values are shown in Table 16. The
investigators proposed that the similar sensitivity of the isolates is
due to saturation of the model at concencentrations of sarafloxacin
> 8 µg/ml.

The results show that sarafloxacin is more active against
facultative anaerobe strains of E. coli than against the obligate
anaerobes B. fragilis and Bifidobacterium spp. The finding that
strains grew in the presence of higher concentrations of sarafloxacin
in the model than in broth culture suggests that sarafloxacin was
partially unavailable in the model. The 'unavailability' factor ranged
from 3 to 12 for E. coli and from 2 to 4 for B. fragilis and was
essentially 1 for Bifidobacterium spp. (McConville, 1992b).

Table 13. Susceptibility of resistant mutants obtained from experiments designed to determine the frequency of spontaneous resistance to
sarafloxacin

Organism Susceptibility of resistant Susceptibility of resistant Susceptibility of resistant Susceptibility of resistant
mutants from studies of mutants after 10 subcultures mutants obtained by mutants from serial
frequency of spontaneous in quinolone-free medium culturing in increasing subculture in quinolone-
resistance concentrations of containing medium after
quinolones 10 subcultures in quinolone-
free medium

No. of MIC No. of MIC No. of MIC No. of MIC
mutants (µg/ml) mutants (µg/ml) mutants (µg/ml) mutants (µg/ml)

E. coli Juhl 1 0.5^a 1 0.5 NR 0.25 NR 0.12
S. aureus CMX 730a 5 2^b 4 2-4 NR 2 NR 2
P. aeruginosa 5007 6 8^c 5 8 NR 4 NR 2

NR, not reported
^a MIC before induction of resistance was < 0.03 µg/ml
^b MIC before induction of resistance was 0.25 µg/ml
^c MIC before induction of resistance was 0.5 µg/ml

Table 14. MIC (in µg/ml) determination on E. coli isolates from model gastrointestinal tracts
containing sarafloxacin

E. coli strain MIC of parent strain MIC₅₀ (µg/ml) of 10 isolates Concentration of
(µg/ml) after incubation in model sarafloxacin (µg/ml)
in model from which
Aerobic Anaerobic Without With isolates were
sarafloxacin^a sarafloxacin obtained

NCTC 8603 0.0625 0.0625 0.0625 0.0313 0.7
NCTC 8761 0.5 0.25 1 0.5 1
NCTC 8783 0.125 0.125 0.0625 0.125 1
NCTC 9434 0.125 0.125 0.0625 0.0625 1
ATCC 25922 0.0625 0.0625 0.0625 0.125 0.25

^a Only one isolate tested

Table 15. MIC (in µg/ml) determination on B. fragilis isolates from model gastrointestinal tracts
containing sarafloxacin

B. fragilis MIC of parent strain MIC₅₀ (µg/ml) of 10 isolates Concentration of
strain (µg/ml) after incubation in model sarafloxacin (µg/ml)
in model from which
Without With isolates were
sarafloxacin^a sarafloxacin obtained

NCTC 8560 4 4 4 8
NCTC 9343 8 2 4 16
NCTC 9344 4 4 4 16
NCTC 10581 4 2 4 8
ATCC 11625 4 4 2 16

^a Only one isolate tested

Table 16. MIC (in µg/ml) determination on Bifidobacteria isolates from model gastrointestinal tracts
containing sarafloxacin

Bifidobacteria MIC of parent strain MIC₅₀ (µg/ml) of 10 isolates Concentration of
strain (µg/ml) after incubation in model sarafloxacin (µg/ml)
in model from which
Without With isolates were
sarafloxacin^a sarafloxacin obtained

B. adolescentis 8 8 8 16
B. infantis 8 8 8 16
B. angulatum 8 8 8 8 and 4^b
B. breve 16 16 16 16
B. longum > 16 > 16 > 16 16

^a Only one isolate tested
^b Five isolates each were obtained from model gastrointestinal tracts containing 4 and 8 µg/ml
sarafloxacin.

The activities of five quinolones - ciprofloxacin, lomefloxacin,
oxofloxacin, sparfloxacin, and DU-6859 - against 320 anerobic
bacterial strains isolated from human patients with infections were
determined by an agar dilution method in vitro. The tested organisms
were 50 strains of Peptostrepto-coccus, 30 strains of Clostridium
perfringens, 50 strains of Clostridium difficile, 100 strains of
Bacteroides fragilis, 50 strains of Prevotella and
Porphyromonas, and 40 strains of Fusobacterium. The most sensitive
strains to all of the quinolones tested were Clostridium
perfringens (MIC₅₀, 0.008-1 µg/ml) and Peptostreptococcus
(MIC₅₀, 0.008-4 µg/ml) (Nord, 1996).

2.2.7 Special studies on ecotoxicity

Because use of sarafloxacin as a therapeutic agent in fish
farming has been considered, various aspects of its ecotoxicity have
been evaluated. Adsorption and desorption and the effect on bacteria
present in sediments are components of such an assessment that may be
relevant to the microbiological assessment of foodborne residues. The
results of those studies are summarized in an environmental report
submitted to the Committee as part of the dossier. The study of
sorption and desorption was conducted in accordance with the
principles of GLP, but the study on the effect of sarafloxacin on
bacteria in sediment was not.

A study of adsorption and desorption was conducted in silty clay
loam (pH 5.4), sandy clay loam (pH 6.0), and loam (pH 8.3). The
results indicate that sarafloxacin hydrochloride is readily sorbed and
would be considered immobile on the three soil types tested. The
adsorption coefficients (K_d) for sarafloxacin were 8400 in silty
clay loam, 7400 in sandy clay loam, and 143 000 in loam. The
desorption capacity of the compound in all three soils was negligible,
with average values of 0.03, 0.02, and 0.63%, respectively.

The effect of sarafloxacin on bacteria present in sediment was
tested in ocean sediment from a source where there had been no
previous sarafloxacin use. The numbers of bacteria from samples
containing 0, 3, 30, or 300 µg/ml of sarafloxacin hydrochloride
equalled 10⁶ to 10⁷ colony forming units, indicating no inhibitory
effect. The observed lack of inhibition could be due to the strong
sorption of sarafloxacin hydrochloride and its consequent
unavailability to bacteria present in the sediment. The unavailability
of sarafloxacin hydrochloride in sediment to bacteria could also help
explain the apparent resistance of sarafloxacin to biodegradation
(Duke, 1990).

2.3 Observations in humans

The safety of single oral doses of sarafloxacin was studied in
groups of healthy male volunteers. Six subjects received 100 mg
sarafloxacin, six received 200 mg, five received 400 mg, and five
received 800 mg. The adverse events reported most frequently were
dizziness and asthenia, although the increase in incidence was not

dose-related. Emotional lability, somnolence, and hiccoughs were the
only adverse events reported by those receiving the lowest dose.
Compliance with the principles of GLP was not required for this study.
The quality and design were consistent with current scientific
standards (Tolman, 1986).

The safety of oral administration of sarafloxacin for seven
consecutive days was studied in groups of six healthy male volunteers
who received doses of 100 or 200 mg twice daily, as a slurry to
maximize exposure of the surface of the stomach. Compliance with the
principles of GLP was not required for this study. The quality and
design were consistent with current scientific standards. Asthenia,
vasodilatation, anxiety, dizziness, and nervousness were reported by
the treated subjects but not those receiving a placebo. Somnolence was
the most frequently reported adverse event in both the treated and
placebo groups. There were no clinically significant changes in
haematological, clinical chemical, coagulation, or urinary parameters,
nor were clinically significant changes seen in physical,
ophthalmological, or neurological examinations or on
electrocardiograms or electroencephalograms (Tolman, 1986).

The safety of oral administration of sarafloxacin for seven
consecutive days was also studied in groups of six healthy male
volunteers who received 100 mg every 12 h, 200 mg every 12 h, or
100 mg every 6 h. The most frequently reported adverse events were
asthenia (eight reports, 20%) and dizziness (six reports, 15%). The
most frequently reported adverse events in the group receiving placebo
were asthenia (six reports, 17%) and somnolence (four reports, 11%).
Compliance with the principles of GLP was not required for this study.
The quality and design were consistent with current scientific
standards (Tolman, 1988).

3. COMMENTS

The Committee considered data from studies of pharmacodynamics,
pharmacokinetics, metabolism, acute and short-term toxicity,
carcinogenicity, genotoxicity, reproductive toxicity, and
developmental toxicity, special studies on microbiological effects and
ecotoxicity, and observations in humans. All of the studies were
conducted according to appropriate standards for study protocol and
conduct.

The absorption, metabolism, and excretion of sarafloxacin were
studied in mice, rats, rabbits, dogs, and humans. After oral
administration, the percent absorption ranged from a low of 10% in
humans given a single 800-mg dose to a high of 70% in dogs given a
single 10-mg/kg bw dose. In mice, rats, and rabbits, the primary route
of excretion was the faeces. The parent drug accounted for 80-90% of a
radiolabelled dose in urine and faeces of mice, rats, rabbits, and
dogs given an oral dose of 10 mg/kg bw, indicating that sarafloxacin
undergoes little metabolism in these species. In humans given single
oral doses of 100-800 mg, sarafloxacin accounted for 75-80% of total
urinary recovery. A metabolite, 3'-oxo-sarafloxacin, comprised about

15% of the recovery. In all species studied, a decreased fraction of
the dose was absorbed at high doses.

Orally administered sarafloxacin was slightly hazardous in
studies of acute toxicity in mice and rats, with LD₅₀ values on the
order of > 5000 to > 8000 mg/kg bw.

In a 90-day study of toxicity with a one-month interim kill, rats
were treated with sarafloxacin at 0, 20, 75, 280, or 1000 mg/kg bw per
day by gavage. The only treatment-related effect observed in animals
treated for one month was grossly enlarged caeca in animals at the
intermediate and high doses. In animals treated for 90 days, the
treatment-related effects included grossly enlarged caeca in males at
doses > 75 mg/kg bw per day. No microscopic pathological changes
were detected in these enlarged caeca. At necropsy, swollen ears were
reported in rats treated for 90 days, with none in controls, two
animals at the low dose, one each at the two intermediate doses, and
three animals at the high dose. The Committee concluded that this
finding was of no toxicological significance. Auricular chondritis was
observed histologically in three females at the high dose. Three
deaths occurred in this study, all among rats at the high dose; one
may have been related to treatment, but the presence of autolysis in
several tissues from this animal made it impossible to determine the
cause of death. The NOEL was 20 mg/kg bw per day on the basis of
grossly enlarged caeca at doses > 75 mg/kg bw per day.

The potential for sarafloxacin to induce arthropathy in dogs was
evaluated in two two-week pilot studies. Arthropathy, characterized by
flattening of the angle of the radial-carpal joint, with no
microscopic evidence of articular lesions, was observed in young adult
(age not stated) dogs given sarafloxacin at 800 mg/kg bw per day in
gelatine capsules. Similar arthropathy was seen in three-month-old
dogs given 125 or 300 mg/kg bw per day in gelatine capsules for two
weeks. Moderate to severe vesicular arthropathic changes of the
articular cartilage were observed microscopically in dogs receiving
300 mg/kg bw per day. The NOEL was 50 mg/kg bw per day on the basis of
the arthropathic effect of sarafloxacin in young dogs.

A 90-day study with a one-month interim kill was conducted in
9œ14-month-old dogs. Three groups of seven dogs of each sex received
sarafloxacin at 0, 5, 25, or 125 mg/kg bw per day in gelatine
capsules. Decreases in mean serum globulin concentration were observed
in males and females at all doses after 28 days of treatment. After 90
days of treatment, statistically significant decreases were seen in
females at the intermediate and high doses and in males at the high
dose. A dose-related decrease in body-weight gain was observed in
males. The NOEL was 5 mg/kg bw per day on the basis of decreased mean
serum globulin concentrations at higher doses.

A 90-day study was conducted in groups of four four-month-old
dogs of each sex, which received 10 or 50 mg/kg bw per day in gelatine
capsules; a third group of six dogs of each sex received 200 mg/kg bw
per day. During the first two weeks of the study, sarafloxacin (as the

hydrochloride salt) was administered as the actual weight, without
regard to the concentration of free base. This resulted in actual
doses that were about 80% of the intended free base doses (i.e. 8, 40,
and 160 mg/kg bw per day). For the remainder of the study, the doses
were adjusted to the target doses. After 90 days of treatment, a
significant decrease in mean serum globulin was observed in females at
the intermediate and high doses. The mean serum globulin concentration
of females at the low dose was comparable to the control value.
Treatment-related toxicity included erythema of the earflaps and
muzzle in males and females at the intermediate and high doses.
Generalized erythema was observed in one male at the high dose.
Swelling around the eyes, eyelids, and earflaps was also seen in dogs
at the high dose. The NOEL was 8 mg/kg bw per day on the basis of
decreased serum globulin, facial swelling, and erythema at higher
doses.

A study of carcinogenicity was conducted in groups of 60 mice of
each sex which received sarafloxacin in the diet at 0, 1000, 5000, or
20 000 mg/ kg feed. The study was terminated after 78 weeks owing to
high mortality in the animals at the intermediate and high doses. The
treatment-related toxicity included nephrotoxicity in females at the
intermediate and high doses and gall-bladder calculi and urolithiasis
in males at the high dose. Caecal dilatation was observed in all
treated males and females, and caecal torsion was observed in males
and females at the intermediate and high doses. No treatment-related
toxicity was observed in those given the low dose. There was no
evidence of carcinogenicity.

A combined long-term study of toxicity and carcinogenicity was
conducted in rats. Sarafloxacin was incorporated into the feed at 0,
1000, 10 000, or 25 000 mg/kg of feed. The toxicity phase (52 weeks)
included 20 animals of each sex per group, and the carcinogenicity
phase (104 weeks) included 65 animals of each sex per group. In the
toxicity phase, the drug intake was equal to 61, 670, or 1700 mg/kg bw
per day. A treatment-related decrease in mean body-weight gain was
observed in males and females at the high dose. Increased blood urea
nitrogen and creatinine concentrations were observed in females and
males at the high dose at weeks 51 and 52, respectively. Total protein
values were decreased in comparison with controls for males at all
doses at each sampling period. These decreased values were
characterized by a significant decrease in globulin concentration with
a relatively unchanged albumin concentration. Total protein and
globulin concentrations were statistically significantly decreased in
females at the intermediate and high doses at week 51/52 only. The
absolute and relative kidney weights were significantly increased in
females at the high dose. Tubular nephropathy was seen in 10/20 males
and females at the high dose and in 1/20 females at the intermediate
dose. Dilatation of the caecum was observed in most rats treated with
sarafloxacin at 10 000 or 25 000 mg/kg of feed. No histopathological
changes were observed in caeca that were grossly dilated. The NOEL was
61 mg/kg bw per day. In the carcinogenicity phase, drug intake was
equal to 54, 580, or 1500 mg/kg bw per day. The signs of toxicity were

similar to those found in the toxicity phase of the study. There was
no evidence of a carcinogenic effect.

Sarafloxacin induced mutation in Chinese hamster ovary cells
in vitro, unscheduled DNA synthesis in rat primary hepatocytes, and
chromosomal aberrations in Chinese hamster ovary cells. It did not
induce unscheduled DNA synthesis in rat primary hepatocytes in
vitro/in vivo or micronuclei in mouse bone marrow in vivo. The
Committee concluded that sarafloxacin is genotoxic in vitro but not
in vivo.

The reproductive toxicity of sarafloxacin was studied in a
three-generation study in rats treated by gavage with 0, 75, 275, or
1000 mg/kg bw per day beginning a minimum of 70 days before breeding.
Each generation was comprised of 30 males and 30 females per group.
The relative liver weights were significantly decreased in males and
females of the second parental generation at the intermediate and high
doses, in males of the second parental generation at the intermediate
and high doses, and in females of the second generation at the high
dose. No treatment-related effects were observed on reproductive or
litter parameters or fetal morphology at doses up to 1000 mg/kg bw per
day, the highest dose tested. The NOEL for parental toxicity was 75
mg/kg bw per day on the basis of decreased liver weights in males and
females at higher doses.

The developmental toxicity of sarafloxacin was evaluated by daily
oral administration of the compound to pregnant rats during days 6œ15
of gestation. Four groups of 20 pregnant rats were treated by gavage
with sarafloxacin base at doses of 0, 20, 75, 280, or 1000 mg/kg bw
per day. There was no evidence of maternal toxicity or teratogenicity
at any dose.

In a study of developmental toxicity, rabbits were given doses of
0, 15, 35, or 75 mg/kg bw per day by gavage during days 6-18 of
gestation, but the highest dose was considered to be inappropriate for
evaluating teratogenicity owing to excessive maternal toxicity,
manifested by decreased body weight, abortions, and decreased
defaecation and urination. Maternal toxicity was also observed in
animals at the low and intermediate doses. External, visceral, and
skeletal malformations and fetotoxicity were observed in fetuses at
the intermediate and high doses. The NOEL for teratogenicity and
fetotoxicity was 15 mg/kg bw per day. No NOEL for maternal toxicity
was identified. The teratogenic effects in this study were considered
to be secondary to maternal toxicity and not directly attributable to
treatment with sarafloxacin.

Data on humans reviewed by the Committee consisted of reports of
side-effects in healthy male volunteers enrolled in clinical trials of
the safety of oral doses of sarafloxacin in comparison with placebo.
The doses ranged from 100 to 800 mg/person per day for one to seven
consecutive days. Effects on gastrointestinal microflora were not
evaluated in these studies. The reported side-effects included

asthenia, vasodilation, anxiety, dizziness, and nervousness or
somnolence, which were observed sporadically at all doses.

Sarafloxacin belongs to a group of antimicrobial fluoroquinolones
that are primarily active against aerobic gram-negative bacteria. In
humans, this characteristic is used therapeutically for selective
elimination of potential aerobic and facultative anaerobic pathogens
from the gastrointestinal tract while preserving the predominant
anaerobic bacterial intestinal flora that protect the gastrointestinal
tract from invasion or overgrowth by potentially pathogenic bacteria.

Several studies on the microbiological activity of sarafloxacin
in vitro were evaluated by the Committee. In one study, MIC₅₀ and
MIC₉₀ values were determined for 735 human clinical isolates of 65
genera at an inoculation density of 10⁴ bacteria. The effect of
sarafloxacin against 210 bacterial strains was assessed, 14 of which
were identified as possible constituents of the human intestinal
microflora. The most sensitive were Escherichia coli, and
Enterobacter cloacae, each with MIC₅₀ values < 0.031 µg/ml. The
most sensitive relevant organism was Peptostreptococcus spp., with
an MIC₅₀ value of 0.125 µg/ml. The least sensitive relevant organism
was Bacteroides vulgatus, with an MIC₅₀ value of 8 µg/ml. In a
study designed to evaluate the effect of inoculum size on the potency
of sarafloxacin in vitro, the geometric mean MIC value was
calculated for sarafloxacin against 50 clinical isolates of human
origin. At inoculation densities of 10⁵ and 10⁷ colony-forming
units per ml, geometric mean MIC values of 0.018 and 0.05 µg/ml,
respectively, were reported for E. coli, the most sensitive
organism.

The microbiological activities of four potential metabolites of
sarafloxacin were determined. The MIC₅₀ values varied with the
species tested, but were generally significantly higher than those for
sarafloxacin against E. coli. Therefore, the Committee concluded
that the microbiological activity of the metabolites against relevant
strains of bacteria found in the human gastrointestinal tract would be
significantly lower than that of sarafloxacin.

The frequency of spontaneous resistance of human clinical
isolates to sarafloxacin was studied in E. coli, S. aureus, and
P. aeruginosa cultured on agar plates containing sarafloxacin at
four and eight times the MIC and by repeatedly transferring organisms
into broth containing increasing concentrations of sarafloxacin.
Stable, resistant mutants arose.

The effects of sarafloxacin against five strains each of
E. coli, B. fragilis, and Bifidobacterium spp. of human origin
were tested in an in-vitro gastrointestinal model system designed to
simulate possible inactivation of sarafloxacin by degradation and
binding in food. E. coli strains grew in the presence of higher
concentrations of sarafloxacin in the model than in broth culture. The
'unavailability' factor ranged from 3 to 12, indicating that
sarafloxacin binds to organic matter in the gastrointestinal model and

decreases the sensitivity of the E. coli strains to the compound.
The factor ranged from 2 to 4 for B. fragilis and was essentially 1
for Bifidobacterium spp. The results of this study suggest that
sarafloxacin is less available in the gastrointestinal model than in
broth culture.

The effect of pH on the potency of sarafloxacin in vitro was
studied with aerobic and anaerobic bacterial isolates of human origin.
In general, the geometric mean MIC value of those organisms considered
to be potential constituents of the human gastrointestinal tract
increased as the pH decreased.

The Committee noted that the organisms most sensitive to the
antimicrobial effects of sarafloxacin, namely E. coli and
Enterobacter cloacae, comprise approximately 1% of the total
gastrointestinal bacterial population and are considered to make a
minimal contribution to colonization resistance in the
gastrointestinal tract. The available microbiological data did not
permit a full evaluation of the relevant bacteria of the human
gastrointestinal tract, as very few such strains were tested. On the
basis of a published report on the activity of related
fluoroquinolones against relevant human intestinal anerobic bacteria,
however, Clostridium perfringens and Peptostreptococcus spp. were
shown to be the most sensitive strains tested. In relevant bacteria,
the lowest MIC₅₀ for sarafloxacin was seen in Peptostreptococcus
spp., with a value of 0.125 µg/ml. The Committee considered these data
to be limited because only three strains rather than the preferred 10
were tested, but they were sufficient to support an ADI.

4. EVALUATION

The upper limit of the ADI based on the antimicrobial activity of
sarafloxacin was calculated on from the formula described on p. 28 as
follows:

Upper limit 0.125 µg/g^a × 220 g
=
of ADI 0.70^b × 2^c × 60 kg

= 0.33 µg/kg bw

^a For the purpose of this evaluation, this is the MIC₅₀ for three
strains of human isolates of Peptostreptococcus spp., which was
the most sensitive strain of the relevant bacteria of human
gastrointestinal microflora tested with sarafloxacin.

^b The fraction of the dose available to act upon microorganisms in
the colon was based on studies in humans in which approximately
70% of a 100-mg oral dose of sarafloxacin was not absorbed.

^c A safety factor of 2 was used because of the limited MIC data
available on the sensitive, relevant bacteria of the human
gastrointestinal tract.

The Committee established an ADI of 0-0.3 µg/kg bw on the basis
of the microbiological activity of sarafloxacin against the most
sensitive, relevant constituent of the human gastrointestinal flora
for which limited adequate microbiological data were available, i.e.
Peptostreptococcus spp. This ADI provides a margin of safety of
17 000 when compared with the lowest toxicological NOEL of 5 mg/kg bw
per day in the 90-day study in dogs.

5. ACKNOWLEDGMENTS

The following individuals at the US Food and Drug Administration
are acknowledged for their assistance with the preparation of the
first draft of this monograph: Dr Carl Cerniglia, Microbiologist,
National Center for Toxicological Research; Dr Robert Condon,
Biostatistician, Center for Veterinary Medicine; Dr Anna Fernandez,
Toxicologist, Center for Veterinary Medicine; Dr Louis T. Mulligan,
Toxicologist, Center for Veterinary Medicine; Dr Terry Peters,
Pathologist, Center for Drug Evaluation and Research; and Dr Leonard
Schechtman, Genetic Toxicologist, Center for Veterinary Medicine.

6. REFERENCES

Bopp, B.A. (1985a). Abbott-56620 drug metabolism report No. 7. Summary
of Abbott-56620 drug metabolism reports for Abbott-56620. Drug
metabolism report No. 4. The absorption, distribution, metabolism and
excretion of Abbott-56620-¹⁴C base in rats. Unpublished report No.
PPRd/85/059. Submitted to WHO by Fort Dodge Animal Health Holland,
Weesp, Netherlands.

Bopp, B.A. (1985b) The distribution of radioactivity in the tissues of
dogs after oral administration of Abbott-55620-¹⁴C base. Unpublished
report No. PPRd/85/153. Submitted to WHO by Fort Dodge Animal Health
Holland, Weesp, Netherlands.

Bopp, B.A. (1985c) Abbott-56620 drug metabolism report No. 7. Summary
of Abbott-56620 drug metabolism reports for Abbott-56620. Drug
metabolism report No. 1. The metabolism and pharmacokinetics of
Abbott-56620 ¹⁴C-base in dogs. Unpublished report No. PPRd/85/059.
Submitted to WHO by Fort Dodge Animal Health Holland, Weesp,
Netherlands.

Brumfitt, W., Franklin, I., Grady, D., Hamilton-Miller, J.M.T. &
Iliffe, A. (1984) Changes in the pharmacokinetics of ciprofloxacin and
fecal flora during administration of a 7-day course to human
volunteers. Antimicrob. Agents Chemother., 26, 757œ761.

Cifone, M.A. (1985) Evaluation of A-56620, lot 66-298-AL in the rat
primary hepatocyte unscheduled DNA synthesis assay. Unpublished report
No. LBI-20991. Submitted to WHO by Fort Dodge Animal Health Holland,
Weesp, Netherlands.

Cifone, M.A. (1988) In vivo/in vitro rat primary hepatocyte
unscheduled DNA synthesis assay of Abbott-56620. Unpublished report
No. PPRd/88/238. Submitted to WHO by Fort Dodge Animal Health Holland,
Weesp, Netherlands.

Creighton, J.M. & Pratt, M.C. (1985a) Three-month toxicity (with
one-month interim kill) study of Abbott-56620 administered orally to
rats. One-month interim kill report. Unpublished report No.
PPRd/85/043. Submitted to WHO by Fort Dodge Animal Health Holland,
Weesp, Netherlands.

Creighton, J.M. & Pratt, M.C. (1985b) Three-month toxicity (with
one-month interim kill) study of Abbott-56620 administered orally to
rats. Final report. Unpublished report No. PPRd/85/085. Submitted to
WHO by Fort Dodge Animal Health Holland, Weesp, Netherlands.

Diehl, M.S. (1994) Mouse micronucleus assay of sarafloxacin
hydrochloride. Unpublished report No. R&D/93/884. Submitted to WHO by
Fort Dodge Animal Health Holland, Weesp, Netherlands.

Dudley, R.E. & Buratto, B. (1984) Abbott 56620 dosage range finding
toxicity study in immature dogs. Unpublished study No. TB84-192.
Submitted to WHO by Fort Dodge Animal Health Holland, Weesp,
Netherlands.

Duke, K.M. (1990) Use of sarafloxacin hydrochloride in fish farming as
a therapeutic agent: An environmental expert report. Unpublished
report No. N0972-3805. Submitted to WHO by Fort Dodge Animal Health
Holland, Weesp, Netherlands.

Fort, F.L. & Buratto, B. (1984) Abbott-56620 dosage range finding
study in rats. Unpublished study No. TA84-188. Submitted to WHO by
Fort Dodge Animal Health Holland, Weesp, Netherlands.

Granneman, G.R. (1985a) The pharmacokinetics of Abbott-56620 base
after oral administration of 5, 25 and 125 mg/kg/day doses in dog.
Unpublished report No. PPRd/85/141. Submitted to WHO by Fort Dodge
Animal Health Holland, Weesp, Netherlands.

Granneman, G.R. (1985b) Plasma Abbott-56620 levels in dog after oral
200 mg (base) doses of solution, suspension and capsule formulations.
Unpublished report No. PPRd/85/258. Submitted to WHO by Fort Dodge
Animal Health Holland, Weesp, Netherlands.

Granneman, G.R. (1985c) The pharmacokinetics and metabolism of
Abbott-56620 base in man after single oral 100, 200, 400 and 800 mg
doses. Unpublished report No. PPRd/85/213. Submitted to WHO by Fort
Dodge Animal Health Holland, Weesp, Netherlands.

Hahn, K.R. (1991a) Acute oral toxicity evaluation of Abbott-56620 in
male mice and male rats. Unpublished report No. R&D/91/623. Submitted
to WHO by Fort Dodge Animal Health Holland, Weesp, Netherlands.

Hahn, K.R. (1991b) Acute oral toxicity evaluation of Abbott-56620 in
male mice. Unpublished report No. R&D/91/624. Submitted to WHO by Fort
Dodge Animal Health Holland, Weesp, Netherlands.

Hahn, K.R. (1991c) Acute oral toxicity evaluation of Abbott-56620
capsule formulation in male mice. Unpublished report No. R&D/91/625.
Submitted to WHO by Fort Dodge Animal Health Holland, Weesp,
Netherlands.

Hahn, K.R. (1991d) Acute oral toxicity evaluation of Abbott-56620 in
male mice. Unpublished report No. R&D/91/626. Submitted to WHO by Fort
Dodge Animal Health Holland, Weesp, Netherlands.

Hahn, K.R. (1991e) Acute oral toxicity evaluation of Abbott-56620 in
male mice. Unpublished report No. R&D/91/627. Submitted to WHO by Fort
Dodge Animal Health Holland, Weesp, Netherlands.

Hemalatha, M. (1988) Mutagenicity test on Abbott-56620 in an in
vitro cytogenetic assay measuring chromosomal aberration frequencies
in Chinese hamster ovary (CHO) cells. Unpublished report No.
PPRd/88/309. Submitted to WHO by Fort Dodge Animal Health Holland,
Weesp, Netherlands.

Kimura, E.T. & Pratt, M.C. (1984) Abbott-56620 dosage range finding
oral toxicity study in young adult dogs. Unpublished study No.
TB84-189. Submitted to WHO by Fort Dodge Animal Health Holland, Weesp,
Netherlands.

Kimura, E.T. & Tekeli, S. (1985a) Three month toxicity study of
Abbott-56620 administered orally to young adult dogs (with one-month
interim kill). One-month interim kill report. Unpublished report No.
PPRd/85/003. Submitted to WHO by Fort Dodge Animal Health Holland,
Weesp, Netherlands.

Kimura, E.T. & Tekeli, S. (1985b) Three month toxicity study of
Abbott-56620 administered orally to young adult dogs (with one-month
interim kill). Final report. Unpublished report No. PPRd/85/086.
Submitted to WHO by Fort Dodge Animal Health Holland, Weesp,
Netherlands.

Kiorpes, A.L. (1991) Thirteen-week capsule toxicity study with
Abbott-56620 in dogs. Unpublished report No. R&D/91/128. Submitted to
WHO by Fort Dodge Animal Health Holland, Weesp, Netherlands.

Lehrer, S.B. (1985) Evaluation of the effects of orally administered
Abbott-56620 on the embryonic and fetal development of the rat
(segment II, TFR). Unpublished report No. PPRd/85/076. Submitted to
WHO by Fort Dodge Animal Health Holland, Weesp, Netherlands.

Lehrer, S.B. (1991) Three-generation reproduction study of
sarafloxacin (Abbott-56620) administered orally in rats. Unpublished
report No. R&D/91/166. Submitted to WHO by Fort Dodge Animal Health
Holland, Weesp, Netherlands.

Lehrer, S.B. & Tekeli, S (1986) A teratology study in rabbits with
A-56620. Final report. Unpublished report No. PPRd/86/061. Submitted
to WHO by Fort Dodge Animal Health Holland, Weesp, Netherlands.

Majors, K.R. (1985) Exploratory acute oral toxicity study of
Abbott-56620 (DNA gyrase inhibitor) in male mice. Unpublished report
No. PPRd/85/098. Submitted to WHO by Fort Dodge Animal Health Holland,
Weesp, Netherlands.

McConville, M. (1992a) Effects of sarafloxacin on relevant human
enteric bacteria under gut-like conditions. Unpublished interim report
No. IRI 8739 (dated 24/04/92). Submitted to WHO by Fort Dodge Animal
Health Holland, Weesp, Netherlands.

McConville, M. (1992b) Effects of sarafloxacin in relevant human
enteric bacteria under gut-like conditions. Unpublished final report
No. IRI 8739 (dated 10/11/92). Submitted to WHO by Fort Dodge Animal
Health Holland, Weesp, Netherlands.

Merits, I. & Bopp, B.A. (1985a) The absorption, metabolism and
excretion of Abbott-56620-14C base in mice. Unpublished report No.
PPRd/85/188. Submitted to WHO by Fort Dodge Animal Health Holland,
Weesp, Netherlands.

Merits, I. & Bopp, B.A. (1985b) The absorption, metabolism and
excretion of Abbott-56620-14C base in rabbits. Unpublished report No.
PPRd/85/070. Submitted to WHO by Fort Dodge Animal Health Holland,
Weesp, Netherlands.

Patterson, S.E. (1985) Pharmacokinetics and bioavailability of
Abbott-56620 base in rats following single or multiple dose
administration of Abbott-56620 or Abbott-56620 base-lactobionate
complex. Unpublished report No. PPRd/85/107. Submitted to WHO by Fort
Dodge Animal Health Holland, Weesp, Netherlands.

Prabhavathi, F. (1984) In vitro antibacterial potency of the
aryl-fluoroquinolone, Abbott 56620. Unpublished report No.
PPNC/84/278. Submitted to WHO by Fort Dodge Animal Health Holland,
Weesp, Netherlands.

Procter, B.G., Salame, R. & Noveroske, J.W. (1991) A dietary
carcinogenicity study of A-56620 in the albino mouse. Unpublished
report No. BRL-84079. Submitted to WHO by Fort Dodge Animal Health
Holland, Weesp, Netherlands.

Smith, S.Y. (1990) A dietary chronic toxicity and carcinogenicity
study of Abbott-56620 in the albino rat: Chronic toxicity phase.
Unpublished report No. R&D/90/217. Submitted to WHO by Fort Dodge
Animal Health Holland, Weesp, Netherlands.

Smith, S.Y., Reid, S. & Noveroske, J.W. (1991) A dietary chronic
toxicity and carcinogenicity study of Abbott-56620 in the albino rat.
Carcinogenicity phase. Unpublished report No. R&D/91/369. Submitted to
WHO by Fort Dodge Animal Health Holland, Weesp, Netherlands.

Tolman, K.G. (1986) Safety and pharmacokinetics of rising single oral
doses of Abbott-56620. Unpublished report on protocol No. M84-075.
Submitted to WHO by Fort Dodge Animal Health Holland, Weesp,
Netherlands.

Tolman, K.G. (1988) Safety and pharmacokinetics of multiple oral doses
of Abbott-56620. Unpublished report No. PPRD/87/238. Submitted to WHO
by Fort Dodge Animal Health Holland, Weesp, Netherlands.

US Food and Drug Administration (1995) Freedom of information summary.
Sarafloxacin water soluble powder (sarafloxacin hydrochloride). For
the control of mortality associated with E.coli in growing broiler
chickens and turkeys. Washington DC.

Weltman, R.H. (1988) Two-week dietary palatability study with
Abbott-56620 in rats. Unpublished report No. PPRd/88/278. Submitted to
WHO by Fort Dodge Animal Health Holland, Weesp, Netherlands.

Weltman, R.H. (1989) Dietary palatability study with Abbott-56620 in
mice. Unpublished report No. PPRd/89/162. Submitted to WHO by Fort
Dodge Animal Health Holland, Weesp, Netherlands.

Young, R.R. (1985) Evaluation of A-56620 Lot-66-298-AL in the
CHO/HGPRT forward mutation assay. Unpublished report No. LBI-22207.
Submitted to WHO by Fort Dodge Animal Health Holland, Weesp,
Netherlands.

    See Also:
       Toxicological Abbreviations
       SARAFLOXACIN (JECFA Evaluation)