Download Pharmacokinetic interactions related to the chemical structures of

Journal of Antimicrobial Chemotherapy (1996) 37, Suppl. A 41-55
Pharmacokinetic interactions related to the chemical structures of
fluoroquinolones
Yasuyuki Mizuki, Iwao Fujiwara and Toshikazo Yamaguchi
Department of Pharmacokinetics, Developmental Research Laboratories, Dainippon
Pharmaceutical Co., Ltd, Enoki-cho 33-94, Suita, Osaka 564, Japan
Fluoroquinolone derivatives interact with methylxanthines (theophylline, caffeine)
and metallic ion-containing drugs to different degrees. The rat appears to be a
suitable model for predicting such interactions in man. It has been possible to
determine the relationship between the chemical structure of thefluoroquinoloneand
the magnitude of the interaction. Fluoroquinolones with a bulky substituent at the
position 8, such as sparfloxacin, lomefloxacin and fleroxacin, are less prone to
interact with theophylline than those without an 8-substituent, such as enoxacin. This
substituent determines the planarity of the wholefluoroquinolonemolecule and the
interaction tends to be more significant for planar fluoroquinolones. Furthermore,
a 4'-nitrogen atom in the 7-piperazinyl group is essential for the interaction to occur.
The nitrogen atom is possibly the site that binds cytochrome P-450, which catalyses
theophylline metabolism. The reduction in bioavailability of fluoroquinolones by
concurrent administration of aluminium hydroxide is more striking for derivatives
with fewer substituents on the essential structure and on the piperazinyl group, such
as norfloxacin, ciprofloxacin and enoxacin. Substitution at the 5-position diminishes
the interaction, which suggests that the 5-substituent may affect the formation and/or
stability of unabsorbable chelate complex which is the probable cause of the
interaction. These findings are potentially useful in designing fluoroquinolones less
prone to drug interactions.
Introduction
The fluoroquinolone antibacterial agents are characterised by a broad antibacterial
spectrum, good oral absorption and wide tissue distribution. They have become widely
used in the treatment of a number of infectious diseases. However, as their clinical
use has increased, so have the number of reports of drug interactions, some of which
have serious clinical consequences. These include pharmacokinetic interactions, such
as the inhibition of quinolone absorption by metallic cation-containing compounds
(e.g. antacid and vitamin preparations), and the inhibition of the metabolism of
methylxanthines (theophylline and caffeine); pharmacodynamic interactions, such as
induction of convulsions by the non-steroidal anti-inflammatory drug fenbufen; and
physicochemical interactions due to incompatibilities between quinolones, such as
pefloxacin and ciprofloxacin, and some penicillin antibiotics (Davey, 1988; Wijnands &
Vree, 1988; Janknegt, 1990).
41
O3O5-7453/96/37AO41 + 15 $12.00/0
© 1996 The British Society for Antimicrobial Chemotherapy
42
Y. Mizukl et al.
In contrast, there have been few reports of drug interactions with sparfloxacin, since
its availability in Japan. Compared with other fluoroquinolones currently available,
sparfloxacin is more active against Gram-positive bacteria, chlamydiae and
mycobacteria (Nakamura et al., 1989). It also has a favourable pharmacokinetic profile
including good tissue distribution and a long half-life (Shimada, Nogita & Ishibashi,
1993). These properties can be attributed to particular aspects of its structural formula
(Andriole, 1993). Similarly, the structure activity properties of a quinolone influences
its propensity to interact with other drugs. This review focuses on the pharmacokinetic
interactions of fluoroquinolones, including sparfloxacin, with methylxanthines
(theophylline and caffeine) and aluminium hydroxide.
Methylxanthines
Theophylline
Studies in humans. Concomitant administration of enoxacin and theophylline was
reported to cause headache and abdominal discomfort (Maesen et al., 1984; Wijnands,
Van Herwaarden & Vree, 1984). Subsequently, enoxacin was found to decrease the
theophylline clearance, leading to increased plasma theophylline concentrations
(Wijnands, Vree & Van Herwaarden, 1985). Thereafter, ciprofloxacin, pefloxacin
(Wijnands, Vree & Van Herwaarden, 1986) and tosufloxacin (Niki et al., 1988a)
were also reported to increase plasma or serum theophylline concentrations, although
no marked clinical symptoms were noted. A weak interaction with theophylline,
without accompanying clinical effects, was also found for both ofloxacin and
norfloxacin (Niki et al., 1987), yet no significant interaction with theophylline were
detected for lomefloxacin (Niki et al., 1988*), fleroxacin (Niki et al., 1990) and
sparfloxacin (Okimoto et al., 1991). The structures of these fluoroquinolones are shown
in Figure 1.
Theophylline is primarily eliminated by hepatic metabolism to 3-methylxanthine
(3-MX), 1-methyluric acid (1-MU) and 1,3-dimethyluric acid (1,3-DMU), which are
mainly excreted in the urine (Grygiel & Birkett, 1980). Rogge et al. (1988) reported that
concurrent administration of enoxacin with theophylline resulted in a significant
decrease in the urinary excretion of these three metabolites. However, co-administration
of fleroxacin or sparfloxacin, which had no influence on plasma concentrations of
theophylline as described above, resulted in no significant change in the urinary
concentrations of theophylline metabolites (Niki et al., 1990; Yamaki et al., 1991).
Using a NADPH-generating system, Sarkar et al. (1990) examined the effect of
several fluoroquinolones on the in-vitro metabolism of theophylline by the human liver
microsome fraction. The yield of each theophylline metabolite was lowest with enoxacin
(the extent of the inhibition was 3-MX > 1-methylxanthine (an intermediate to
1-MU > 1,3-DMU), intermediate with ciprofloxacin and highest with norfloxacin.
These results concur with the degree of interaction reported in vivo with these agents.
These findings indicate that the interaction between fluoroquinolones and theophylline
is the result of inhibition of cytochrome P-450.
Studies in rats. We have examined the mechanism of the theophylline-quinolone
interaction using a rat model (Mizuki et al., 1989*, 1991). When the extent of
the fluoroquinolone-theophylline interaction was assessed by their effects on the
43
Phannacoklnetic interactions of fluoroquinolones
0
•C00H
COOH
Lomefloxacin
CiprofToxacin
COOH
NH 2 0
COOH
F
r ji
F
F
CHJCHJF
Tbsufloxacin
Fleroxacdn
Sparfloxacin
Figure 1. Chemical structures of selected fluoroquinolones.
elimination half-life (t{) of theophylline, the quinolones could be ranked as follows:
enoxacin > ciprofloxacin > norfloxacin > ofloxacin > sparfloxacin (Table I), which
essentially reflects the order of magnitude of the interaction in humans. Therefore, the
rat appears to be a suitable model for predicting this interaction in man.
Table I. Plasma elimination half-lives (T\) of theophylline after single oral administration
(15 mg/kg) in rats treated with various fluoroquinolones 2 h before, and 2 h and 6 h after
the theophylline dose, and comparison of increases in calculated peak plasma concentration
(C™,) at steady state in rats with observed (CM, of theophylline in humans (Mizuki et al.,
19896)
Theophylline*
Treatment
Control
Enoxacin
Ciprofloxacin
Norfloxacin
Ofloxacin
Sparfloxacin
Oxo-enoxacin
Dose*
(mg/kg/dose)
—
300, po
50, ip
50, ip
50, po
100, po
50, ip
(h)
3.8
7.4
6.8
5.3
4.2
4.0
4.7
±0.1
± 0.8
± 1.1
±0.5
± 0.2
± 0.2
± 0.4
Increase in theophylline
relative to control
Rat
Human r
Human'
1
1.6
1.4
1.2
1.1
1.0
1.1
1
1.6
1.5
1.1
1.1
1.0
—
1
2.1
1.2
—
0.9
—
—
•The doses of the fluoroquinolones were adjusted so that their peak plasma concxntrations ranged
from 5 to 10 mg/L.
*Each tj value is the mean ± S.E.M. for five animals.
The C», increases in humans were reported by 'Niki el al. (1987) and AVijnands, Vree & Van
Herwaarden (1986).
44
Y. Miznki et al.
Unchanged theophylline, 1-MU and 1,3-DMU are excreted in the urine of rats given
theophylline. Co-administration of enoxacin with theophylline caused a significant
increase in the urinary excretion of unchanged theophylline and a significant decrease
in the two metabolites (Mizuki et al., 1989a). However, no changes were observed when
sparfloxacin was co-administered with theophylline in this model (Mizuki et al., 1991).
Inhibition of theophylline metabolism by enoxacin has also been demonstrated in vitro
using the isolated rat liver cell system; the inhibition was greater than that seen with
ofloxacin (Mulder et al., 1988).
Relationship between interaction and chemical structure. Using the rat model, an
attempt has been made to relate the chemical structure of fluoroquinolones with their
propensity to interact with theophylline (Table II). Quinolones with a bulky substituent
at position 8, e.g. fluoro or methoxy group (derivatives nos 1-11), were less prone to
interact with theophylline than those with a hydrogen at position 8 (derivatives nos
12-15) which in turn were less likely to inhibit the metabolism of theophylline than
quinolones with no substituent at position 8 (derivatives no. 16 and no. 17). In contrast,
the substituent at either position 1 or 5 and the position and number of methyl groups
Table n . Plasma elimination half-lives (tj) of theophylline after a single oral administration
(15 mg/kg) in rats treated with various fluoroquinolone derivatives 2 h before, and 2 h and 6 h
after the theophylline dose
Re
F
COOH
Derivatives 18-23
Derivative
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
R,
R3
A
A
A
A
A
A
A
A
NH 2
NH 2
C2H3
(CH2)2F
A
A
C2H5
C2H3
A
Substituent
R.
Rr
R4
Rs
Theophylline
Dose*
7V
(mg/kg/dose) (control = 1)
NH 2
F
CH,
F
H
H
H
H
F
F
F
F
F
OCH,
F
F
F
F
F
CH,
CH 3
H
H
CH,
CH,
CH 3
CH,
CH,
H
H
H
H
CH,
H
H
H
H
H
H
CH,
H
CH,
H
H
H
CH,
CH,
H
H
H
H
H
100,
100,
100,
50,
50,
50,
50,
50,
50,
50,
300,
po
po
po
ip
po
po
po
po
po
po
po
1.05
1.18
1.18
1.11
1.03
0.97
1.16
0.95
1.13
1.18
1.08
H
H
H
NH 2
H
H
H
H
H
H
H
CH,
H
H
CH,
H
H
H
H
CH,
50,
50,
50,
75,
ip
ip
po
ip
1.79
1.39
1.87
1.61
NHj
Pharmacoldnetic interactions of
16
17
C2H5
C2H5
H
H
—'
18
A
NH 2
F
19
A
F
F
20
A
NH 2
F
A
H
22
A
H
f
23
A
H
C
21
e
H
H
H
CH3
NH 2
CH,
NH 2
CH3
(trans)
NH 2
CH3
CH3
flnoroqidnolones
H
H
45
300, po
100, po
1.95
2.37
4
300, po
1.00
4
50, po
1.32
d
300, po
1.29
(cis)
*"
(cis)
CH3
i
300, po
3.39
NH 2
CH3
(trans)
NH 2
CH3
CH3
i
300, po
4.21
d
100, po
1.68
NH 2
(.ds)
(cis)
•The doses of the derivatives were adjusted so that their maximum plasma concentrations ranged from 5
to lOmg/L.
'Mean tj during coadministration/mean control tj (n = 4 or 5).
'Naphthyridine:
COOH
^'-Aminopyrrondine.
in the piperazine ring at position 7 were found to scarcely affect the degree of
interaction.
The results of in-vitro metabolic studies are presented in Table III. An enoxacin
derivative which lacks the carboxyl group at position 3 (no. 24) had equivalent
inhibitory activity to enoxacin (no. 16) whereas a derivative which lacks the terminal
nitrogen atom in the 7-piperazinyl group of enoxacin (no. 25) demonstrated no
inhibitory activity (Mizuki et al., 1989a). Similarly, derivatives which have a carbonyl
group at a neighbouring position to the terminal nitrogen atom (no. 26 & 27) showed
minimal inhibitory activity (unpublished data).
Our previous studies of the inhibitory effect of different quinolone substrates on rat
liver microsomal enzymes (Mizuki et al., 1989a) suggested that the nitrogen atom is the
portion that binds to the trivalent iron ion (Fe3 + ) (Poulos et al., 1985) in the catalytic
centre of cytochrome P-450. Being electrophilic, Fe3+ is capable of binding basic
functional groups (Dawson, Andersson & Sono, 1982), such as a nitrogen atom, at the
sixth ligand. This binding could hinder the approach of theophylline to the active centre
of cytochrome P-450, resulting in reduced theophylline metabolism. When the carbonyl
group was introduced adjacent to the nitrogen, the basicity of the nitrogen disappeared
due to the electron-attracting effect of the carbonyl group and hence the affinity between
the nitrogen and Fe3+ possibly decreased. This hypothesis was strongly supported by
the electrostatic potential distribution calculated by the molecular orbital method.
Y. Miruki et al.
46
Table HI. Inhibitory effects of enoxacin derivatives on in-vitro theophylline
metabolism to 1,3-DMU by liver microsomes of rats pretreated with
3-methylcholanthrepane. The initial concentrations of theophylline and the
derivatives were 0.25 and 0.50 mM, respectively. Each value is the mean of
duplicate experiments
Derivative
no.
Structure
% Inhibition (%)
47
Enoxacin
16
COOH
HN.
24
40
C2H6
25
0
No inhibition
COOH
26
COOH
27
COOH
CH3CO-N
Figure 2 illustrates the isopotential contours of enoxacin (no. 16) and its 3'-oxo
derivative (no. 26), which has a carbonyl group adjacent to the nitrogen. The
isopotential shape of derivative no. 26, which exhibited less of an inhibitory effect
Phannacokinetic interactions of fluoroquinolones
47
3'-oxo-enraacin
26
Figure 2. Electrostatic potentials of 3'-oxo-enoxacin (no. 26) and enoxacin (no. 16) based on the AMI
molecular orbital calculation (Mizuki et al., 1996). Lines in shadowed and open areas represent an isopotential
contour at — I and + 1 kcal/mole, respectively.
in vitro compared with enoxacin, obviously differs from enoxacin around the nitrogen,
i.e. the potential Fe3* binding site (Mizuki et al., 1996).
In addition, the substituent at position 8 was found to affect the degree of interaction
between theophylline and fluoroquinolone. Derivative no. 11, which has a fluorine atom
at position 8, showed only a weak interaction with theophylline (Table II) despite
having similar physicochemical properties (e.g water solubility, log P and pKa),
pharmacokinetic properties in rats, and electrostatic potentials to those of enoxacin (no.
16) (data not shown). When the stereochemistry of these two compounds was studied
by X-ray crystallography and by the molecular orbital method, enoxacin was found
to have a highly planar conformation, since its essential skeleton naphthyridine
ring was present on the same plane as the piperazine at its 7 position (Figure 3).
48
Y. Mlzuki et al.
Derivative no. 11, on the other hand, exhibited steric hindrance between its essential
skeleton quinolone ring and its piperazine substituent, which resulted in a lack of planar
conformation (Figure 3) (Mizuki et al., 1996). Miyamoto et al. (1990) reported X-ray
crystallographic data on sparfloxacin (no. 1), which possesses afluorineatom at position
8 and exhibited no interaction with theophylline (Table II), showing that the compound
had a bulky conformation toward the quinoline plane similar to derivative no. 11. The
steric conformation of pefloxacin (no. 14), based on X-ray crystallography (Baenziger,
Fox & Modak, 1986), and norfloxacin (no. 13) calculated from pefloxacin data (data
not shown), estimated their bulkiness as intermediate between enoxacin and
sparfloxacin. This correlates with the degree to which these agents interact with
theophylline (Table II). These results suggest that it is necessary for the quinolone to
have a planar conformation in order to approach cytochrome P-450, bind to Fe3 + and
inhibit theophylline metabolism. The conformation of theophylline has also been
reported to be highly planar (Sutor, 1958).
The above discussion has focused on fluoroquinolones with a 7-piperazine group
but according to our study results, it also holds true for fluoroquinolones which have
a 7-pyrrolidine substituted with an amino group (unpublished data, see Table II).
Caffeine
Although adverse symptoms are clinically less obvious, the plasma or serum t| of
caffeine, like that of theophylline, is prolonged by concurrent administration of
certain fluoroquinolones. The likelihood of an interaction with caffeine has been
8-fluorinated derivative
Enoxacin
11
16
Figure 3. Molecular structures of 8-fluorinated derivative (no. 11) optimised by the AM 1 molecular orbital
calculation and enoxacin (no. 16) based on X-ray crystallography (Mizuki et al., 1996).
Pharmacoklnetic interactions of
fluoroqulnolones
49
ranked as follows: enoxacin > ciprofloxacin = norfloxacin > ofloxacin = lomefloxacin
(Fuhr et al., 1992a), which is similar to the situation for theophylline. By conducting
in-vitro studies on caffeine metabolism in human liver microsome preparation to yield
paraxanthine (3-N-demethylation) and 1,3-dimethylxanthine (7-N-demethylation),
Fuhr et al. (1990) have shown that the cause of the interaction is competitive inhibition
of caffeine metabolism by fluoroquinolones. Therefore, the relationship between
chemical structure and the degree of interaction is thought to be the same as for
theophylline. The enzyme CYP1A2 has been shown to be involved in the metabolism
of caffeine (Butler et al., 1989; Berthou et al., 1991; Fuhr et al., 1992A).
Fluoroquinolones have only been reported to have a marked metabolic interaction with
methylxanthine drugs, which suggests that fluoroquinolones specifically inhibit
CYP1A2.
Metallic cation-containing compounds
Aluminium hydroxide
Studies in humans. When fluoroquinolones are administered orally concurrently
with aluminium hydroxide, the plasma or serum concentration and urinary
recovery of fluoroquinolone are decreased as a result of decreased bioavailability.
Shimada et al. (1991) compared the peak serum concentrations, AUCs and
urinary recovery of fluoroquinolones administered to healthy volunteers at a dose
of 200 mg both with and without aluminium hydroxide 1 g. The magnitude of
interaction was least for sparfloxacin and fleroxacin, intermediate for ofloxacin
and lomefloxacin, and greatest for norfloxacin and enoxacin. In an earlier study,
the AUC of ciprofloxacin 750 mg was 84.5% lower after co-administration with
aluminium hydroxide 2.4 g (Frost et al., 1989). Thus, it appears that, similar
to the situation with theophylline, the degree of interaction between aluminium
hydroxide and the quinolone is dependent upon the chemical structure of the
quinolone.
Chelate formation involving metallic ions and 3-carboxyl and 4-carbonyl groups
on the quinolone, leading to inhibition of quinolone absorption, has been proposed
as the cause of the interaction (Polk, 1989). This concept was seemingly based on
the reports that nalidixic acid, which possesses these functional groups, interacted
with various metallic ions and this resulted in changes in its physicochemical properties
(UV spectrum, partition ratio to carbon tetrachloride and permeability to artificial
membrane) (Nakano, Yamamoto & Arita, 1973; Coif, Goodfield & Williams, 1984).
Through the use of 13C-NMR, Shimada et al. (1992) revealed that the 3-carboxyl
and 4-carbonyl groups of lomefloxacin played a role in its interaction with Al3 + and
Mg^+ ions.
Studies in rats. Compared with rats given enoxacin, ofloxacin or norfloxacin
20 mg/kg alone, the AUC and urinary recovery of each fluoroquinolone were
significantly lower in rats concomitantly given aluminium hydroxide 50 mg/kg
(Okazaki, Kurata & Tachizawa, 1988). The degree of decrease was in the order of
norfloxacin > enoxacin > ofloxacin. The amount of ofloxacin remaining in the digestive
tract was greater when the quinolone was co-administered with aluminium hydroxide
and thus the fraction absorbed was lower. Ofloxacin and norfloxacin were found to
form a stable chelate complex with Al3 + .
Y. Mizuki et at.
50
In a similar study, rats were given 20 mg/kg of various fluoroquinolones with or
without aluminium hydroxide 100 mg/kg and the fluoroquinolone AUC calculated
(unpublished data). The decrease in AUC with co-administration of aluminium
hydroxide ranged from 9 to 69% and the ranking of the magnitude of the interaction
was
norfloxacin = ciprofloxacin > enoxacin > ofloxacin > lomefloxacin >
sparfloxacin (Table IV), which corresponded to the degree of interaction observed in
healthy human volunteers. Therefore, the rat appears to be a useful model for studying
this interaction.
Relationship between interaction and chemical structure. To elucidate the relationships
between chemical structure and interaction with aluminium hydroxide, various
fluoroquinolone derivatives were administered alone and with aluminium hydroxide to
rats and the fluoroquinolone AUCs calculated (Table V). The major findings were:
(i) the interaction is less significant for derivatives with a naphthyridine ring rather than
Table IV. Peak plasma or scrum concentrations (C^,) and areas under plasma or serum
concentration-lime curve (AUC) of fluoroquinolones after single oral administration in rats
(20 mg/kg) and humans (200 mg) treated concurrently with oral aluminium hydroxide
(100 mg/kg and 1000 mg, respectively)
Rat (plasma)"
c_
Human (serum)*
(mg/L)
AUC
(mg • h/L)
(mg/L)
AUC
(mg • h/L)
0.68 ±0.18
2.00 ± 0.34
1.45
6.7
0.14 ±0.02
(W
0.62 ± 0.07
(69)
<0.1
(100)
0.2
(97)
1.26 ±0.25
2.49 ±0.19
0.32 ± 0.09
(75)
0.79 ± 0.09
(68)
—
—
1.62 ±0.34
6.34 ± 0.78
2.26
11.4
0.59 ±0.10
(64)
2.67 ± 0.44
(58)
0.46
(80)
1.8
(84)
Ofloxacin
Ofloxacin plus
aluminium hydroxide
5.57 ± 0.42
13.39 ± 1.51
2.05
10.4
1.70 ±0.22
(69)
7.90 ± 0.51
(45)
0.60
(61)
3.4
(53)
Lomefloxacin
Lomefloxacin plus
aluminium hydroxide
4.37 ±0.58
13.46 ± 0.98
2.03
9.1
1.54 ±0.12
(65)
7.90 ±0.51
(41)
0.60
(70)
3.4
(63)
Sparfloxacin
Sparfloxacin plus
aluminium hydroxide
1.73 ±0.23
5.91 ±0.83
0.87
21.1
1.21 ±0.29
(30)
4.15 ±0.60
(30)
0.68
(22)
13.7
(35)
Treatment
Norfloxacin
Norfloxacin plus
aluminium hydroxide
Ciprofloxacin
Ciprofloxacin plus
aluminium hydroxide
Enoxacin
Enoxacin plus
aluminium hydroxide
•Mean ± S.E.M. (n => 5 to 7).
'Data reported by Shimada et al. (1991).
'Percent of decrease (%).
51
Pharmacokinetic interactions of ftuoroquinolones
Table V. E>ecreases in AUC of fluoroquinolone derivatives after single oral administration
(20mg/kg) in rats treated concurrently with oral aluminium hydroxide (lOOmg/kg)
Re
O
F.
Derivative
no.
13
12
17
24'
28'
29
14
3C
9
11
8
1
31
32
7
3
A
A
COOH
R,
R5
R.
Ry
R.
R5
AUC decreases*
(control = 1)
QH,
H
H
H
H
H
H
H
H
H
H
H
NH 2
F
F
F
NH 2
H
H
H
H
H
H
—
CH 3
H
H
CH 3
H
CH 3
CH 3
H
H
CH 3
H
H
H
H
H
—
H
CH 3
CH 3
H
H
H
H
H
CH 3
H
CH 3
H
H
H
H
—
H
H
H
H
H
H
CH 3
H
H
H
H
0.69
0.68
0.58
0.02
0.65
0.49
0.45
0.45
0.41
0.32
0.31
0.30
0.25
0.11
0.10
0.09
Substituent
A
C 2 H,
C 2 H,
C2HS
A
C 2 H,
C2H5
A
A
A
A
A
A
A
4
b
b
H
H
F
F
F
F
F
F
F
F
'1 — (mean AUC dunng co-administration/mean control AUC).
'Naphthyridine.
COOH
52
Y. Mlzukl et al.
a quinoline ring; (ii) a cyclopropyl group at position 1 rather than an ethyl group
decreases the likelihood of interaction; (iii) a 3-carboxyl group is essential for the
interaction; (iv) the presence of a substituent at position 5 lessens the interaction;
(v) a methyl group in the 7-piperazine ring weakens the interaction; (vi) the interaction
occurs even in the absence of a 7-substituent; and (vii) an 8-fluoro group decreases the
extent of the interaction. The important conclusion is that the more substituents there
are in the essential skeleton and the piperazine group, the less significant the interaction.
In addition, point (iii) supports the view that the interaction is caused by chelation
involving substituents at positions 3 and 4.
A significant positive correlation was observed between the stability constants,
measured by potentiometric titration of the chelate of fluoroquinolones with Al3 + , and
decreases in AUC (Table V & Figure 4). This indicates that the greater the stability of
the chelate, the more significant the interaction and suggests that the chelate is
minimally absorbable. Two derivatives with a 5-amino group, including sparfloxacin,
were off the correlation line negative to the y-axis (Figure 4), indicating that the
interaction with these derivatives was less significant despite the stability of their
chelates. This may suggest that the chelate of the 5-amino derivatives has essentially
different properties to the chelates with other derivatives.
Iron and calcium-containing drugs
Metallic cation-containing drugs other than aluminium hydroxide, including ferrous
sulphate, ferrous-glycine-sulphate and calcium carbonate have been reported to
interact with fluoroquinolones. These metals also form a chelate complex with
fluoroquinolones.
When co-administered with fluoroquinolones in humans, Fe2 + -containing drugs
decreased the AUC of ciprofloxacin, ofloxacin (Lode et al., 1989), norfloxacin and
sparfloxacin (Kanemitsu et al., 1994) by 48, 36, 51 and 28%, respectively, compared
Stability constant with Al
Figure 4. Correlation between stability constant with Al'* (logK.K:) and decrease in AUC of
fluoroquinolones ( # ) and 5-amino-fluoroquinolones (O) (data unpublished).
Pharmacokinetic interactions of
fluoroquinolones
S3
with the control. Although these studies were not performed under the same conditions
and therefore direct comparisons are difficult, it appears that the degree of interaction
is high for ciprofloxacin and norfloxacin and low for the other fluoroquinolones. The
extent of the interaction with Fe2 + -containing drugs seems to be similar to or slightly
lower than that with Al3 + -containing drugs.
The interaction between fluoroquinolones and Ca2 + -containing drugs appears to be
less significant than that with other metallic cations; calcium carbonate administered
5 min before ingestion of ciprofloxacin (Frost et al., 1992) or 2 h before norfloxacin
(Flor et al., 1990) lowered the AUC of the quinolone by only 42 and 4%, respectively,
compared with the control.
In rats, the sparfloxacin AUC was found to decrease by 15% and 0% when
co-administered with ferrous sulphate (100 mg/kg) or calcium gluconate (200 mg/kg),
respectively, whereas the enoxacin AUC was 51% and 7% lower, respectively. The
highest stability constants of the chelate between the metallic ion and the two
quinolones were observed with Al 3+ followed by Fe 2 * and then Ca2 + , which
corresponds well with the degree of interaction observed in vivo. As above, the
formation of the chelate complex conceivably plays an important role in the interaction
between the quinolones and metallic cation-containing drugs. However, it has not yet
been possible to fully elucidate the relationship between the chelate and the number of
substituents on the essential structure of the fluoroquinolone and on the piperazinyl
moiety. For an overview of absorption interactions with fluoroquinolones, see
Lomaestro & Bailie (1995).
Conclusions
Fluoroquinolone derivatives interact with methylxanthines (theophylline, caffeine) and
metallic ion-containing drugs to different degrees. The rat appears to be a suitable model
for predicting the magnitude of these interactions in man. The stereochemistry and the
electron density distribution of fluoroquinolones are important determinants of the
interaction with theophylline. Sparfloxacin, which has a bulky substituent at position
8 and therefore lacks a planar conformation, does not appear to inhibit the metabolism
of theophylline significantly and thus no specific precautions are required for patients
who receive both drugs concomitantly. This is of importance for a quinolone used in
the treatment of respiratory tract infections.
As regards metallic cation-containing drugs, the stability of the chelate with the
fluoroquinolone is an important determinant of the degree of interaction. The decrease
in fluoroquinolone absorption appears to be greatest with Al3 + -containing preparations
followed by Fe2 + - and then Ca2 + -containing compounds. These findings are potentially
useful when designing fluoroquinolones less prone to drug interactions.
References
Andriole, V. T. (1993). The future of the quinolones. Drugs 45, Suppl. 3, 1-7.
Baenziger, N. C , Fox, C. L. & Modak, S. L. (1986). The structure of silver pefloxacin, an
antibiotic related to nalidixic acid. Ada Crystallography C42, 1505-9.
Berthou, F.F Flinois, J., Ratanasavanh, D., Beaune, P., Riche, C. & Guillouzo, A. (1991).
Evidence for the involvement of several cytochromes P-450 in the first steps of caffeine
metabolism by human liver microsomes. Drug Metabolism and Disposition 19, 561-7.
54
Y. Mizuki et al.
Butler, M. A., Iwasaki, M., Guengerich, F. P. & Kadlubar, F. F. (1989). Human cytochrome
P-450PA (P-450IA2), the phenacetin O-deethylase, is primarily responsible for the hepatic
3-demethylation of caffeine and N-oxidation of carcinogenic arylamines. Proceedings of the
National Academy of Sciences of the United States of America 86, 7696-700.
Cole, A., Goodfield, J., Williams, D. R. & Midgley, J. M. (1984). The complexation of transition
series metal ions by nalidixic acid. Inorganica Chimica Ada 92, 91-7.
Davey, P. G. (1988). Overview of drug interactions with the quinolones. Journal of Antimicrobial
Chemotherapy 22, Suppl. C, 97-107.
Dawson, J. H., Andersson, L. A. & Sono, M. (1982). Spectroscopic investigations of ferric
cytochrome P-450-CAM ligand complexes. Journal of Biological Chemistry 257, 3606-17.
Flor, S., Guay, D. R. P., Opsahl, J. A., Tack, K. & Matzke, G. R. (1990). Effects of
magnesium-aluminum hydroxide and calcium carbonate antacid on bioavailability of
ofloxacin. Antimicrobial Agents and Chemotherapy 34, 2436-8.
Frost, R. W., Lasseter, K. C , Noe, A. J., Shamblen, E. C. & Lettieri, J. T. (1992). Effects of
aluminum hydroxide and calcium carbonate antacids on the bioavailability of ciprofloxacin.
Antimicrobial Agents and Chemotherapy 36, 830-2.
Frost, R. W., Lettieri, J. T., Noe, A., Shamblen, E. C. & Lasseter, K. (1989). Effect of aluminum
hydroxide and calcium carbonate antacids on ciprofloxacin bioavailability. Clinical
Pharmacology and Therapeutics 45, 165.
Fuhr, U., Anders, E. M., Mahr, G., Sorgel, F. & Staib, A. H. (1992a). Inhibitory potency of
quinolone antibacterial agents against cytochrome P450IA2 activity in vivo and in vitro.
Antimicrobial Agents and Chemotherapy 36, 942-48.
Fuhr, U., Doehmer, J., Battula, N , Wolfel, C , Kudla, C , Keita, Y. et al. (19926).
Biotransformation of caffeine and theophylline in mammalian cell lines genetically
engineered for expression of single cytochrome P450 isoforms. Biochemical Pharmacology 43,
225-35.
Fuhr, U., Wolff, T., Harder, S., Schymanski, P. & Staib, A. H. (1990). Quinolone inhibition
of cytochrome P-450-dependent caffeine metabolism in human liver microsomes. Drug
Metabolism and Disposition 18, 1005-10.
Grygiel, J. J. & Birkett, D. J. (1980). Effect of age on pattern of theophylline metabolism. Clinical
Pharmacology and Therapeutics 28, 456-62.
Janknegt, R. (1990). Drug interactions with quinolones. Journal of Antimicrobial Chemotherapy
26, Suppl. D, 7-29.
Kanemitsu, K., Hori, S., Yanagawa, A. & Shimada, J. (1994). Effect of ferrous sulfate on the
pharmacokinetics of sparfloxacin. Chemotherapy (Tokyo) 42, 6-13.
Lode, H., Stuhlert, P., Depperman, K., Mainz, S. & Borner, K. (1989). Pharmacokinetics
interactions between oral ciprofloxacin, ofloxacin and ferro-salts. In Program and Abstracts
of the Twenty-Ninth Jnterscience Conference on Antimicrobial Agents and Chemotherapy,
Houston, TX, 1989. Abstract 213. American Society for Microbiology, Washington, DC.
Lomaestro, B. M. & Bailie, G. R. (1995). Absorption interactions with fluoroquinolones: 1995
update. Drug Safety 12, 314-33.
Maesen, F. P., Teengs, J. P., Baur, C. & Davies, B. I. (1984). Quinolones and raised plasma
concentrations of theophylline. Lancet ii, 530.
Miyamoto, T., Matsumoto, J., Chiba, K.., Egawa, H., Shibamori, K., Minamida, A. et al.
(1990). Synthesis and structure-activity relationships of 5-substituted 6,8-difluoroquinolones,
including sparfloxacin, a new quinolone antibacterial agent with improved potency. Journal
of Medicinal Chemistry 33, 1646-56.
Mizuki, Y., Fujiwara, I., Yamaguchi, T. & Sekine, Y. (1996). Structure-related inhibitory effect
of antimicrobial enoxacin and derivatives on theophylline metabolism by rat liver
microsomes. Antimicrobial Agents and Chemotherapy, in press.
Mizuki, Y., Kamaura, M., Yamaguchi, T., Sekine, Y. & Hashimoto, M. (1989a). Interaction of
enoxacin with theophylline in rats. Arzneimittel-Forschung/Drug Research 39, 593-7.
Mizuki, Y., Yamaguchi, T., Sekine, Y. & Hashimoto, M. (1989A). Interaction of enoxacin with
theophylline in rats. Chemotherapy (Tokyo) 37, 1071-9.
Mizuki, Y., Yamaguchi, T., Sekine, Y. & Hashimoto, M. (1991). Effect of sparfloxacin on
pharmacokinetics of theophylline in rats. Xenobiotic Metabolism and Disposition (Tokyo) 6,
67-72.
PharmacoVinetic interactions of
fluoroqulnolones
55
Mulder, G. J., Nagelkerke, J. F., Tijdens, R. B., Wijnands, W. J. A. & Van der Mark, E. J.
(1988). Inhibition of the oxidative metabolism of theophylline in isolated rat hepatocytes
by the quinolone antibiotic enoxacin and its metabolite oxoenoxacin, but not by
ofloxacin. Biochemical Pharmacology 37, 2565-8.
Nakamura, S., Minami, A., Nakata, K., Kurobe, N., Kouno, K., Sakaguchi, Y. et al. (1989).
In vitro and in vivo antibacterial activities of AT-4140, a new broad-spectrum quinolone.
Antimicrobial Agents and Chemotherapy 33, 1167-73.
Nakano, M., Yamamoto, M. & Arita, T. (1973). Interactions of aluminum, magnesium, and
calcium ions with nalidixic acid. Chemical and Pharmaceutical Bulletin 26, 1505-10.
Niki, Y., Soejima, R., Kawane, H., Sumi, M. & Umeki, S. (1987). New synthetic quinolone
antibacterial agents and serum concentration of theophylline. Chest 92, 663-9.
Niki, Y., Sumi, M., Tsukiyama, K., Moriya, O., Umeki, S. & Soejima, R. (1988a). Effect of
T-3262 on serum concentration of theophylline. Chemotherapy (Tokyo) 36, Suppl. 9, 201-7.
Niki, Y., Tasaka, Y., Kishimoto, T , Nakajima, M., Tsukiyama, K., Nakagawa, Y. et al.
(1990). Effect of fleroxacin on serum concentration of theophylline. Chemotherapy (Tokyo)
38, Suppl. 2, 364-71.
Niki, Y., Tasaka, Y., Tsukiyama, K., Umeki, S., Watanabe, M. & Soejima, R. (19886). Effect
of NY-198 on serum concentration of theophylline. Chemotherapy (Tokvo) 36, Suppl. 2,
251-5.
Okazaki, O., Kurata, T. & Tachizawa, H. (1988). Studies on the mechanism of pharmacokinetic
interaction of aluminum hydroxide, an antacid, with new quinolones in rats. Xenobiotic
Metabolism and Disposition (Tokyo) 3, 1219-24.
Okimoto, N., Niki, Y., Sumi, M., Nakagawa, Y. & Soejima, R. (1991). Effect of sparfloxacin
on serum concentration of theophylline. Chemotherapy (Tokyo) 39, Suppl. 4, 158-60.
Polk, R. (1989). Drug-drug interactions with ciprofloxacin and other fluoroquinolones. American
Journal of Medicine 87, Suppl. 5A, 76-81.
Poulos, T. L., Finzel, B. C , Gunsalus, I. C , Wagner, G. C. & Kraut, J. (1985). The 2.6-A
crystal structure of Pseudomonas putida cytochrome P-450. Journal of Biological Chemistry
260, 16122-30.
Rogge, M. C , Solomon, W. R., Sedman, A. J., Welling, P. G., Toothaker, R. D. & Wagner,
J. G. (1988). The theophylline-enoxacin interaction: I. Effect of enoxacin dose size on
theophylline disposition. Clinical Pharmacology and Therapeutics 44, 579-87.
Sarkar, M., Polk, R. E., Guzelian, P. S., Hunt, C. & Karnes, H. T. (1990). In vitro effect of
fluoroquinolones on theophylline metabolism in human liver microsomes. Antimicrobial
Agents and Chemotherapy 34, 594—9.
Shimada, J., Nogita, T. & Ishibashi, Y. (1993). Clinical pharmacokinetics of sparfloxacin. Clinical
Pharmacokinetics 25, 358-69.
Shimada, J., Saito, A., Shiba, K.., Hojo, T., Kaji, M., Hori, S. et al. (1991). Pharmacokinetics
and clinical studies on sparfloxacin. Chemotherapy (Tokyo) 39, Suppl. 4, 234—44.
Shimada, J., Shiba, K., Oguma, T., Miwa, H., Yoshimura, Y., Nishikawa, T. et al. (1992). Effect
of antacid on absorption of the quinolone lomefloxacin. Antimicrobial Agents and
Chemotherapy 36, 1219-24.
Sutor, D. J. (1958). The structures of the pyrimidines and purines. VI. The crystal structure of
theophylline. Ada Crystallography 11, 83-7.
Wijnands, W. J. A., Van Herwaarden, C. L. A. & Vree, T. B. (1984). Enoxacin raises plasma
theophylline concentrations. Lancet ii, 108-9.
Wijnands, W. J. A. & Vree, T. B. (1988). Interaction between the fluoroquinolones and the
bronchodilator theophylline. Journal of Antimicrobial Chemotherapy 22, Suppl. C, 109-14.
Wijnands, W. J. A., Vree, T. B. & Van Herwaarden, C. L. A. (1985). Enoxacin decreases the
clearance of theophylline in man. British Journal of Clinical Pharmacology 20, 583-8.
Wijnands, W. J. A., Vree, T. B. & van Herwaarden, C. L. A. (1986). The influence of quinolone
derivatives on theophylline clearance. British Journal of Clinical Pharmacology 22, 677-83.
Yamaki, K., Miyatake, H., Taki, F., Suzuki, R., Takagi, K. & Satake, T. (1991). Studies on
sparfloxacin (SPFX) against respiratory tract infections and its effect on theophylline
pharmacokinetics. Chemotherapy (Tokyo) 39, Suppl. 4, 280-5.