c mice

Toxicology Letters, 54 (1990) 299-308

299

Elsevier TOXLET

02471

Phototoxic potential of quinolone antibacterial agents in Balb/c mice

Nobuhiko Wagai, Fumie Yamaguchi, Masayasu Sekiguchi and Katsuhiko Tawara Research Institute. Daiichi Pharmaceutical Co., Ltd., Tokyo (Japan) (Received

22 May 1990)

(Accepted

27 July 1990)

Key words: Phototoxicity;

Quinolone

derivatives;

Mouse; Ultraviolet-A

SUMMARY The phototoxic

potentials

mice were orally (CPFX),

of quinolone

antibacterial

nalidixic

acid (NA),

administered

lomefloxacin

(LMFX)

and DR-3355

let-A (UVA) for 4 h (21.6 joules/cmr). parameter, marked

cutaneous

phototoxic tration

phototoxic

however,

changes

reactions

CPFX,

OFLX

were characterized

of inflammatory

of OFLX),

on the ears, whereas grossly

doses of LMFX,

at 19, 102, 143, 553, 619 and 741 mg/kg,

respectively. DR-3355,

OFLX,

for overt damage, CPFX,

LMFX,

OFLX

Thus, the phototoxic

to ultravio-

as a major

phototoxic

NA and ENX caused

and DR-3355 reactions

none.

by edema and infil-

and CPFX potencies

caused

on the ears. These

tissue surrounding

DR-3355

ciprofloxacin

exposed

and histopathologically

into the connective

ENX, NA, OFLX,

tested were: LMFX > ENX, NA>

(OFLX),

and immediately

also caused phototoxic

by erythema,

neutrophils,

in Balb/c strain mice. The

ofloxacin

ended. At doses of 200 mg/kg,

and DR-3355

cells, especially

The 50% erythema-inducing

(s-isomer

(ENX),

The ears were examined

0, 24 and 48 h after irradiation

At 800 mg/kg,

agents were investigated enoxacin

the cartilage. were calculated

of the quinolones

CPFX.

INTRODUCTION

Quinolone antibacterial agents are widely used as oral preparations. Quinolones have been under development worldwide since the discovery of nalidixic acid in the 1960s and there has been great progress in broadening their antibacterial spectrum against both gram-negative and gram-positive organisms, and also in improving their pharmacokinetic profiles (absorption, distribution, metabolism and excretion). However, they occasionally show side-effects in a very low percentage of patients, e.g. gasAddress for correspondence: 13, Kitakasai

0378-4274/90/S

I-Chome,

Nobuhiko

Edogawa-ku,

Wagai, Tokyo

Research

Institute,

Daiichi

Pharmaceutical

134, Japan.

3.50 @ 1990 Elsevier Science Publishers

B.V. (Biomedical

Division)

Co., Ltd.,

16-

300

trointestinal such as nausea/vomiting, central nervous symptoms, such as insomnia and dizziness, or side-effects on the skin, such as rash and photosensitivity [ 11. Nalidixic acid (NA) is a well-known photosensitizer. In the early 1970s a considerable number of cases of photosensitization due to NA were reported in clinical practice [2-61, and there have recently been some case-reports of photosensitivity due to other newer quinolones such as enoxacin (ENX) or ciprofloxacin (CPFX) [7-91. In a previous paper [lo], we reported precise experimental conditions for phototoxicity studies and demonstrated a simple method for detecting the phototoxic potentials of some photosensitizers such as NA and chlorpromazine by using Balb/c strain mice. The purpose of this study was to compare the phototoxic potentials of several newer quinolones used clinically or under clinical testing in the same mouse model. MATERIALS AND METHODS

Chemicals

Nalidixic acid (NA), enoxacin (ENX), ciprofloxacin (CPFX), lomefloxacin (LMFX), ofloxacin (OFLX) and DR-3355 (structural formulae shown in Fig. 1) were used in this study. These compounds were synthesized or extracted from commercial tablets or powder in our laboratory. DR-3355 was synthesized through optical resolution of the racemic intermediate of ofloxacin [ 111. E.xperimental

animals

Female Balb/c strain mice, 67 weeks old and weighing 17.1-18.8 g (Charles River Japan Inc., Atsugi, Japan) were used in this study. These mice were housed 46 per cage in plastic cages and maintained in an air-conditioned room (temperature, 2125°C; relative humidity, 40-70s) with free access to commercial laboratory chow (MB- 1, Funabashi Farm Ltd, Funabashi, Japan) and tap water. 0

COOH

0

0

F

C-N

CH. AH.CHI

Nalidixic

Acid(NA)

CH a0N3

LomefioxacinlLMFX)

Ofioxacin(OFLX)

0 COOH

0

*HCI*HIO

Enoxacin(ENX)

Ciprofloxacin(CPFX) Fig. 1.Chemical structure of quinolone antibacterial agents.

DR-3355

301

Light source and irradiation Details of the light source and the method of irradiation have been described previously [lo]. Briefly, the light source of ultraviolet-A (UVA) consisted of a bank of 10 Toshiba FL20SBLB ‘Black Light’ tubes; 3-mm-thick pane of glass (Floatglass, Asahi Glass, Tokyo, Japan) was used as a filter to eliminate wavelengths below 320 nm. Each mouse was placed alone in a partitioned chamber (4 x 8 x 4 cm) covered by a window pane under the light source and with an electric fan at a distance from the light source for ventilation. Up to 12 mice could be irradiated at once, and they stayed prone for 4 h. The distance from the light source to the backs of the mice was 15 cm. The dose of radiation was 21.6 joule/cm*, as measured with a model of a UVX Digital Radiometer fitted with a UVX-36 sensor responding at 365 nm (UVP Inc., San Gabriel, CA, U.S.A.). Phototoxic effects of quinolone derivatives Each mouse, fasted overnight, was orally administered a quinolone antibacterial agent (10 mg/kg body wt.), and was immediately exposed to UVA for 4 h in a box placed on the central irradiation area. Each chemical except for NA was dissolved or suspended in distilled water. NA was suspended in 0.5% sodium carboxymethylcellulose (CMC). Two groups of mice were used as controls, one receiving the test compound without UVA irradiation, and the other the UVA irradiation with the vehicle alone. Overt damage to the ears as a major phototoxic parameter was examined 0, 24 and 48 h after the completion of irradiation. The phototoxic reaction was judged to be positive when the ears exhibited significant erythema or obvious reddening according to the method of Kuzuna et al. [ 121. Histopathology Immediately after the last observation of the ears (48 h), the mice were sacrificed; their ears were excised, fixed in 10% buffered formalin, and then calcified and embedded in paraffin blocks. Hematoxylin-and-eosin preparations of processed sections were prepared for microscopic examination. RESULTS

Phototoxic eflects of quinolone derivatives As shown in Table I, at doses of 200 mg/kg, LMFX, ENX and NA caused positive reactions. The mice treated with LMFX and ENX showed marked erythema on the ears immediately after the completion of the irradiation (0 h); these reactions lasted 48 h after the completion of irradiation. All mice treated with NA also showed significant erythema on the ears at 0 h, and the number of mice with positive reactions decreased over the course of time. On the other hand, mice treated with a dose of 200 mg/kg of CPFX, OFLX or DR-3355 showed no phototoxic reaction, but elevation of the dose of CPFX, OFLX and DR-3355 to 800 mg/kg caused marked erythe-

302

ma on the ears. These cutaneous phototoxic reactions induced by a 800 mg/kg dose of OFLX or DR-3355 also lasted for 48 h after the completion of irradiation. The peak of the reaction induced by CPFX was observed at 48 h. Neither the non-irradiated nor the iradiated control mice showed any erythema on the ears. Dose-dependency of phototoxic reaction Female Balb/c mice, 67 weeks old, were administered quinolones at the oral doses indicated in Figure 2, 10 mice being given each dose except for LMFX (n = 4), after overnight fasting and were immediately exposed to UVA for 4 h. The incidence of erythema (s) on their ears at 24 h was plotted as the ordinate against the log dose. TABLE

I

PHOTOTOXIC LONE

REACTIONS

DERIVATIVES

INDUCED

BY ORAL

ADMINISTRATION

OF SEVERAL

QUINO-

IN MICE

Treatment

No. of

UVA

No. of mice with erythema

mice used 0 LMFX

ENX

NA

OFLX

+

(200 mg/kg)

(200 mg/kg)

(200 mg/kg)

(200 mg/kg) (800 mg/kg)

DR-3355

(20 mg/kg)

CPFX

(800 mg/kg)

* Erythema

6

6

6

0

0

0

8

8

7

7

0

0

0

+ _

4

4

4

2

4

0

0

0

+ _

4

0

0

0

4

0

0

0

+ -

4

4

4

4

4

0

0

0

+ _

4

0

0

0

4

0

0

0

+ +

Distilled water CMC

6 4

4

+ + -

(200 mg/kg)

48 h*

+ _

+ -

(800 mg/kg)

24

on the ears was observed

6

6

5

5

4

0

0

0

4

0

0

0

4 4

0 2

0 3

0 4

4

0

0

0

4

0

0

0

4

0

0

0

0, 24 and 48 h after the completion

of the irradiation

(see Materials

and Methods). Mice were orally administered dissolved

or suspended

quinolones

and immediately

exposed

to UVA for 4 h. Each quinolone

in distilled water except for NA. NA was suspended

in CMC.

was

.. / 8/’ P

303

A

*

0

0

0

//

A

0 LMFX A ENX q NA

.

OFLX

.

DR.3355

.

CPFX

Fig. 2. Dose-dependency of phototoxic reactions caused by several quinolones. Mice were orally administered quinolones and immediately exposed to UVA for 4 h. Ten mice were used for each dose except for LMFX (n = 4 mice). The incidence (k) of erythema at 24 h after the completion of the irradiation was plotted (as the ordinate) against the log dose.

The dose-response curves of these quinolones tested in this study were proved by the Litchfield and Wilcoxon method to be linear, and the 50% erythema-inducing dose (EIDss, mg/kg) was thus calculated for each. As shown in Table II, the EIDso values of LMFX, ENX, NA, OFLX, DR-3355 and CPFX were 19, 102, 143, 553, 619 and 741 mg/kg, respectively. These results enabled the photototoxic activities of these quinolones to be divided roughly into 3 groups: stronger, medium and weaker. The stronger group contained only LMFX, the medium group consisted of ENX and NA, and the weaker group of OFLX, DR-3355 and CPFX. TABLE II THE 50% ERYTHEMA-INDUCING

DOSE (EID,,) OF QUINOLONE DERIVATIVES IN MICE

Quinolone

EIDso (mg/kg)

95% Confidence limit (mg/kg)

LMFX ENX NA OFLX DR-3355 CPFX

19 102 143 553 619 741

lo-37 89-l 17 107-190 483634 43&890 522-1052

EIDSo was calculated by the Litchtield and Wilcoxon method.

304

Histopathoiogy

Figure 3B shows the histopathological findings of cutaneous phototoxic reactions observed on the ears of mice treated orally with ENX (200 mg/kg) plus UVA irradiation. Merked edematous changes, vascular ectasia and considerable infiltration of inflammatory cells, especially neutrophils, were observed in the connective tissue surrounding the cartilage. Almost the same findings were observed in mice given LMFX (200 mg/kg), NA f200 mg/kg), CPFX (800 mg/kg), OFLX (800 mg/kg) or DR-3355 (800 mg/kg) together with UVA irradiation (data not shown), No histopathological changes were observed in non-irradiated mice receiving quinolones or irradiated mice receiving the vehicle alone (Fig. 3A). DISCUSSION

Phototoxicity tests using animal model systems are potentially more predictive than those using simpler organisms or photosensitized oxidation [13,14]. There are several test systems employing guinea-pigs and hairless mice [15,16]. Although Epstein and Wintroub [17] mentioned the need to take into account several disadvantages of in vivo testing, improvements could be made methodologically in factors such as whether enough ultraviolet light is reaching the appropriate target tissue, the appropriateness of the wavelengths, and the timing of irradiation. Our previous report [lo] showed the precise experimental conditions for phototoxicity studies and demonstrated the simple method for detecting the phototoxic potentials of NA and chlorpromazine which are known clinically as ‘photosensitizers’ [Z6,181. In this study, this method was used to compare the phototoxic potentials of several quinolone antibacterial agents, The present study demonstrated that the administration of quinolone derivatives followed by irradiation with UVA induced marked cutaneous phototoxic reactions on the ears of Balb/c strain mice. These phototoxic reactions were characterized grossly by erythema and histopathologically by edema and infiltration of inflammatory cells, especially neutrophils, in the connective tissue surrounding the cartilage. In the first experiment, the phototoxic potential of 6 quinolones was investigated by oral administration of these compounds (200 and/or 800 mg/kg) plus UVA irradiation for 4 h. At doses of 200 mg/kg, LMFX, NA and ENX caused marked erythema on the ears, but CPFX, OFLX and DR-3355 showed no erythema at all. However, elevation of the doses of CPFX, OFLX or DR-3355 to 800 mg/kg caused marked erythema on the ears (Table I). In the second experiment, we compared the phototoxic potential of these 6 quinolones more precisely by calculating the EIDs0 values of these compounds. The incidence of erythema at 24 h (most of the mice treated with each quinolone showed positive reactions in the first experiment) was plotted on the ordinate against the log dose on the abscissa. The dose-response curves were obtained and proved to be almost linear by the Lichfield and Wilcoxon method (Fig. 2). The EIDSo of LMFX, ENX, NA, OFLX, DR-3355 and CPFX were calculated

Fig. 3. Histopathological (B) ENX (200 mg/kg

sections of ears of (A) UVA-treated p.o.) plus UVA-irradiated

were stained with hematoxylin

mice receiving

mice. Specimens

vehicle (distilled water) alone,

were taken

and eosin ( x 240). Other details are provided

48 h after irradiation

in the text.

and

306

as 19, 102, 143, 553, 619 and 741 mg/kg, respectively (Table II). These results indicated that the phototoxic potency of quinolone antibacterial agents tested in this study decreased in the order: LMFX > ENX, NA > OFLX, DR-3355, CPFX. Ferguson et al. [19] demonstrated that the phototoxic potential of CPFX was greater than that for OFLX. The discrepancy between our results and those of Ferguson et al. is probably due to the difference in the route of quinolone administration. They injected the quinolones intraperitoneally (150 mg/kg) and subsequently irradiated with UVA for 4 h, and repeated this treatment for 5 days. The pharmacokinetic profile, especially absorption, of CPFX is probably inferior to that of OFLX. To clarify this point, we performed a complementary experiment. Mice were treated intravenously with 100 mg/kg doses of CPFX, OFLX and DR-3355 and were immediately exposed to UVA for 4 h. The mice treated with OFLX and DR-3355 showed no erythema on the ears for 48 h after the completion of irradiation, but 5 of 6 mice treated with CPFX had erythema at 24 h (Table III). These findings indicate that the absorption of CPFX is not as good as that of OFLX or DR-3355, and that the phototoxic potential of CPFX is substantially greater than that of OFLX or DR-3355. Furthermore, Goto et al. [20] compared the serum level of CPFX with those of OFLX and other new quinolones and showed that the maximum level of CPFX was about half that of OFLX, and that the maximum serum levels of CPFX and OFLX were reached at 15 and 5 min, respectively, when these compounds were administered orally at a dose of 1 or 2 mg/animal in ICR strain mice. Generally, phototoxic reactions are initiated by absorption of photoenergy and

TABLE

III

PHOTOTOXIC REACTIONS AND CPFX IN MICE

INDUCED

Treatment

UVA

BY INTRAVENOUS

No. of

INJECTIONS

OF OFLX,

DR-3355

No. of mice with erythema

mice used

(100 mg/kg)

OFLX

DR-3355(100

mg/kg)

(100 mg/kg)

CPFX

Vehicle * Erythema Female

24

48 h*

+

6

0

0

0

_

6

0

0

0

+

6

0

0

0

_

6

0

0

0

+

6

I

5

1

_

6

0

0

0

+

6

0

0

0

on the ears was observed

0, 24 and 48 h after the completion

Balb/c mice, 6 weeks old and weighing

(10 ml/kg) and immediately

0

17.2-18.0

of irradiation.

g, were injected

exposed to UVA for 4 h. Each quinolone

intravenously

was dissolved

with quinolones

in 0.1 N NaOH-saline.

307

occur via photochemical reactions. The photoproducts including free radicals induce toxic reactions dose-dependently, Kornhauser et al. [21] reviewed the mechanism of phototoxic reactions. The electronically excited molecule, located within or near a cell, may elicit a phototoxic response through several mechanisms. Femandez et al. [22] showed by a photohemolytic experiment that the phototoxicity induced by NA was the result of oxygen-dependent reactions, and that both singlet oxygen and hydroxyl radicals were required for erythrocyte hemolysis but not for photoproducts. The cause of the differences in phototoxic reactivity observed in the 6 quinolones tested in this study is not yet clear. Further studies will be necessary to clarify the mechanism and cause of the different degrees of phototoxic reactivity induced by combined treatment with quinolones and UVA irradiation. In summary, the administration of quinolone plus UVA irradiation caused cutaneous phototoxic reactions on the ears of Balb/c mice. This phototoxic reaction was almost dose-dependent. The phototoxic potency of quinolones tested in this study decreased in the order: LMFX > ENX, NA > CPFX, OFLX, DR-3355. ACKNOWLEDGEMENT

This work was supported by Daiichi Pharmaceutical

Co., Ltd., Tokyo.

REFERENCES 1 Halkin,

H. (1988) Adverse

2 Ramsay,

effects of the fluoroquinolones.

C.A. and Obreshfova,

5233528. 3 Baes, H. (1968) Photosensitivity 4 Birkett,

D.A., Garretts,

acid. Br. J. Dermatol.

Rev. Infect. Dis. 10, Suppl. 1, S2588S261.

B. (1974) Photosensitivity caused by nalidixic

M. and Stevenson,

from nalidixic

acid. Dermatologica

C.J. (1969) Phototoxic

acid. Br. J. Dermatol.

91,

136,6164.

bullous

eruptions

due to nalidixic

81, 342-344.

5 Burry, J.N. (1974) Persistent

phototoxicity

due to nalidixic

acid (Letters

to Editor).

Arch.

Dermatol.

109,263. 6 Brauner, 7 Matsui,

G.J. (1975) Bullous photoreaction C., Takahashi,

vity. Rinsho-hifu 8 Kawabe,

S. and Morohashi,

to nalidixic

acid. Am. J. Med. 58,576580.

M. (1987) Three cases of enoxacin-induced

photosensiti-

(Jpn) 41, 579-582.

Y., Sakakibara,

S. and Mizuno,

N. (1987) Photoallergic

reaction

due to enoxacin.

Hifu-rin-

sho (Jpn) 29,51&53. 9 Smith, G.M., loxacin,

Chemother. 10 Wagai,

Leyland,

M.J., Farrell,

a new 4-quinolone

in the treatment

A.M. (1986) Preliminary of febrile neutropenic

evaluation

patients.

of ciprof-

J. Antimicrob.

18, Suppl. D, 1655174.

N., Yamaguchi,

for detecting 11 Hayakawa,

I.D. and Geddes,

antibiotic, F., Tawara,

phototoxic

potentials

H., Atarashi,

thesis and antibacterial

K. and Onodera,

S., Yokohama, activities

T. (1989) Studies on experimental

of drugs in Balb/c mice, J. Toxicol. S., Imamura,

of optically

M., Sakano,

active ofloxacin.

conditions

Sci. 14, 1977204.

K. and Furukawa,

Antimicrob.

Agents

M. (1986) SynChemother.

29,

163-164. 12 Kuzuna,

S., Tamura,

by ear skin erythema 13 Daniels,

S. and Kawai,

K. (1983) Phototoxicity

F. (1965) A simple microbiological

vest. Dermatol.

of antiinflammatory

agents,

evaluated

in mice. Ensho (Jpn) 3,515-516.

44, 259-263.

method

for demonstrating

phototoxic

compounds.

J. In-

308

14 Johnson,

B.E., Walker,

E.M. and Hetherington,

toxicity.

Skin Models,

Methods

15 Forbes,

P.D., Davies,

R.E. and Urbach,

tive effects of anthracene

A.M. (1986) In Vitro Models

to Study. Springer-Verlag.

F. (1976) Phototoxicity

and 8-methoxypsoralen

for Cutaneous

Photo-

Inc., New York, pp. 264281. and photocarcinogenesis:

in the skin of mice. Food Cosmet.

compara-

Toxicol.

14, 303-

306. 16 Harber,

L.C. (1969) Use of guinea-pigs

Effects of Ultraviolet

Radiation.

17 Epstein,

J.H. and Wintroub,

18 Epstein,

S. (1968) Chlorpromazine

Dermatol.

J., Walker,

J. Dermatol. terial therapy

In: F. Urbach

(Ed.), The Biologic

B.U. (1985) Photosensitivity photosensitivity:

due to drugs. phototoxic

Drug 30,42257.

and photoallergic

reactions.

Arch.

E.M. and Johnson,

B.E. (1989) Quinolone

antibiotic-induced

phototoxicity.

Br.

120, 291.

S., Ogawa, activity

M., Miyazaki,

S., Kaneko,

Y. and Kuwabara,

of BAY o 9867 (ciprofloxacin),

(Tokyo)

21 Kornhauser,

a new pyridone

S. (1985) In vitro and in vivo antibaccarboxylic

acid derivative.

Chemo-

33, S-7, 18-30.

A., Warner,

cellular and molecular ington,

studies.

New York, pp. 291-295.

98, 354363.

19 Ferguson, 20 Goto,

in photobiologic

Pergamon,

W. and Giles, Jr., A. (1987) Light-induced

level. In: Dermatotoxicity,

3rd ed. Hemisphere

dermal

toxicity:

Publishing

effects on the

Corporation,

Wash-

D.C., 377412.

22 Fernindez,

E.B., Cardenas,

gen dependent

A.M.G.

photohemolysis.

and Martinez,

Farmaco

G.S. (1987) Phototoxicity

Ed. Sci. 42,68ll690.

from nalidixic

acid: oxy-