ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, July 1995, p. 1493–1498
Copyright ? 1995, American Society for Microbiology
Vol. 39, No. 7
Pharmacokinetics of Ofloxacin in Serum and Vitreous
Humor of Albino and Pigmented Rabbits
RUTH J. PERKINS,1WEIGUO LIU,1GEORGE DRUSANO,2ASSUMPTA MADU,3
MARTIN MAYERS,3CHRISTIAN MADU,4AND MICHAEL H. MILLER1,5*
Departments of Medicine and Pharmacology, Divisions of Infectious Diseases,1Clinical Pharmacology,2and
Molecular and Cellular Medicine,5Albany Medical College, Albany, New York 12208; the Department
of Ophthalmology, Albert Einstein College of Medicine, Bronx, New York 104613; and Department
of Management Science, Pace University, New York, New York 100384
Received 12 December 1994/Returned for modification 10 February 1995/Accepted 4 May 1995
The purpose of this study was to evaluate the pharmacokinetics of ofloxacin in serum and vitreous humor
samples from albino and pigmented rabbits by using a recently described animal model which permits robust
estimation of parameter values. The drug was administered to rabbits intravenously, multiple vitreous humor
and serum samples were taken from each rabbit, and the vitreous humor and serum samples were assayed by
high-pressure liquid chromatography. The pharmacokinetic parameters were determined with RSTRIP, an
iterative, nonlinear, weighted, least-squares regression program. Eight New Zealand White rabbits and eight
Dutch Belted rabbits (split into single-dose and multiple-dose groups) were investigated in this study. The
value of penetration into the vitreous humor of albino rabbits (n ? 6) was 32.6% ? 2.12%, with terminal-
elimination half-life values of 3.21 and 2.39 h, respectively, for vitreous humor and serum. In pigmented rabbits
after a single dose (n ? 3) and with a steady-state concentration of drug in serum (n ? 4), penetration values
were similar, at 30.4% ? 2.98% and 30.0% ? 4.12%, respectively (P > 0.10). Following a single dose of
ofloxacin, pigmented animals had elimination half-life values from serum and vitreous humor of 2.64 and 4.32
h, respectively. After steady state was achieved, half-life values for serum and vitreous humor were 3.12 and
6.05 h, respectively. By comparing pigmentation, dose mode, and elimination from vitreous humor, the vitreous
humor half-life value for singly dosed albino rabbits (3.21 h) was not more rapid than that for singly dosed
pigmented rabbits (4.32 h [P > 0.10]) but was more rapid than that for multiply dosed pigmented rabbits (6.05
h [P < 0.05]). The excellent penetration of ofloxacin into uninflamed rabbit eyes in conjunction with the higher
levels in blood found in humans and greater activity against coagulase-negative staphylococci suggest that
ofloxacin may be preferable to other quinolones for evaluation in the prophylaxis and treatment of bacterial
We have recently developed and validated an ocular phar-
macokinetic animal model which differs from conventional
methods that obtain a single sample of vitreous humor from
different animals (20, 25, 26, 29). In our model, serial samples
are obtained from the vitreous humor after intravenous drug
administration. This model does not alter the pharmacokinet-
ics of drugs in the eye (24, 25, 29). Serial sampling is prefera-
ble, since it permits more-rigorous pharmacokinetic analysis
and decreases the number of animals which are needed for the
same degree of certainty in the pharmacokinetics parameter
estimates by more than 50-fold compared with the single-sam-
ple approach (20).
Bacterial endophthalmitis is a severe, sight-threatening in-
fection which most commonly occurs after cataract surgery (10,
32). The current therapy of bacterial endophthalmitis includes
the intraocular and systemic administration of antibiotics com-
bined with a pars plana vitrectomy (9, 10, 32, 34, 35). The
National Eye Institute Endophthalmitis Vitrectomy Study
(EVS) is currently evaluating the relative impact of vitrectomy
and systemic antibiotics on the outcome of patients receiving
intraocular antibiotics. However, recent studies with humans
and animals (3, 7, 12, 14, 15, 18, 23, 25, 29, 33, 36) suggest that
quinolone antimicrobial agents penetrate into the vitreous hu-
mor better than the conventional antibiotics that are being
used in the EVS study. Since the EVS will not evaluate the role
of systemic quinolones, pharmacokinetic and outcome studies
with animals may be useful in evaluating the role of systemic
quinolones in the treatment of endophthalmitis.
The purpose of this study was to characterize the penetra-
tion of the quinolone ofloxacin into the vitreous humor of
rabbits. Since differences in the ocular pharmacokinetics of
antibiotics have been described for pigmented versus nonpig-
mented rabbits (8, 28, 30), we compared the pharmacokinetics
of ofloxacin for pigmented versus albino rabbits with previous
studies that used ciprofloxacin. Ofloxacin achieves higher con-
centrations in serum (11) and has been reported to penetrate
into rabbit eyes (18). Relative to ciprofloxacin, ofloxacin is two-
to four-fold more active against coagulase-negative staphylo-
cocci, the most common cause of postoperative bacterial en-
MATERIALS AND METHODS
Animal model. Adult male albino New Zealand White (NZW) and Dutch
Belted (DB) rabbits (Milbrook Farms, Amherst, Mass.) weighing 2 to 3 kg were
used. Prior to experimental use, antibiotic free feed (Hi Fiber ProLab Rabbit
Chow; Lacrose) and water were provided ad libitum.
The animals were sedated with an intramuscular dose of diazepam (2.5 mg)
and were anesthetized with a subcutaneous dose of urethane (1.62 g/kg of body
weight) approximately 45 min prior to antibiotic administration. Anesthesia was
maintained throughout the sampling period, with supplemental intramuscular
ketamine (100 mg/ml) and xylazine (20 mg/ml) (0.1 ml of each given simulta-
neously) as needed. Following anesthesia, a 24-gauge angiocatheter was inserted
* Corresponding author. Mailing address: Department of Medicine,
Albany Medical College, 47 New Scotland Ave., Albany, N.Y. 12208.
Phone: (518) 262-5343. Fax: (518) 262-5727.
into the marginal ear vein to facilitate antibiotic administration, and a second
catheter was inserted into the central artery of the contralateral ear to obtain
Ofloxacin was obtained from the R. W. Johnson Pharmaceutical Research
Institute, Raritan, N.J. A bolus intravenous infusion (40 mg/kg) was administered
through the marginal ear vein, which was followed by a 1-ml flush of 0.9% NaCl.
For animals with steady-state concentrations of drug in serum, four doses (40
mg/kg each) were given every 8 h. Serial samples of blood and vitreous humor
were taken at 0.25, 0.5, 1, 2, 3, 4, 6, and 8 h after drug administration.
For vitreous humor samples, experimental eyes were exposed by using a sterile
lid retractor and were fixated with forceps. A 28-gauge needle affixed to a
low-dose insulin syringe was carefully inserted 4 mm from the limbus by direct
visualization into the center of the vitreous cavity, and 20 ?l was withdrawn. Care
was taken to avoid the lens. Vitreous humor and serum samples were transferred
into a 2-ml microcentrifuge tube and stored frozen at ?20?C prior to the deter-
mination of antibiotic levels. Following the designated sampling period, the
animals were sacrificed with pentobarbital sodium solution (125 mg/kg).
Antibiotic assays. Ofloxacin concentrations were measured by high-pressure
liquid chromatography according to the method described by Borner et al. (6).
The samples were run at 25?C in a C18, 5-?m column (220 by 2.1 mm) packed
with Nucleosil. Samples were prepared by using 20 ?l of serum or 15 ?l of
vitreous humor and by mixing with 130 or 135 ?l of mobile phase, respectively,
to acid precipitate proteins and then filtering. The mobile phase (75% acetoni-
trile, 25% 0.1 M H3PO4[pH 3.82] with concentrated phosphoric acid) was
delivered to the column at a rate of 0.2 ml/min with a Hewlett-Packard series
1050 pump (Wilmington, Del.). Standards were prepared at concentrations of 2,
4, 8, 12, and 24 ?g/ml for serum samples and of 0.1, 0.25, 0.5, 1.0, and 2.0 ?g/ml
for vitreous humor samples. Serum and vitreous humor samples were prepared
with pooled rabbit serum and balanced salt solution, respectively (29). A 100-?l
volume of sample was injected by a Hewlett-Packard series 1050 autosampler
and run serially through a Hewlett-Packard 1040A UV detector (240 to 280 nm)
and a Hewlett-Packard 1046A fluorescence detector (excitation, 280 nm; emis-
sion, 445 nm). Data were collected on a Hewlett-Packard Chemstation (Wilm-
ington, Del.). The antibiotic concentrations were quantitated by using peak
height. The assay sensitivity was 0.01 ?g/ml. The interday coefficient of variation
ranges for serum and vitreous humor were 3.7 to 6.1% and 0.8 to 2.5%, respec-
tively. Linearity over the range examined was excellent, with an r2of ?0.99.
Protein quantitation and characterization. Total protein concentrations in the
vitreous humor samples were determined with the Coomassie protein assay
reagent (Pierce, Rockford, Ill.). Identification and quantitation of proteins in the
vitreous humor were performed by sodium dodecyl sulfate-polyacrylamide gel
electrophoresis (SDS-PAGE) (Mini-protean II cell, model 1000/500 power sup-
ply; Bio-Rad, Hercules, Calif.), and densitometry (model 60S Video Densitom-
eter; BioImage, Ann Arbor, Mich.).
The Coomassie protein assay was performed by placing 1 ?l of sample, 9 ?l of
distilled water, and 240 ?l of Coomassie reagent into each well of a 96-well
microtiter plate. The plate was read on an EL 312e Biokinetics Reader (Bio-Tek
Instruments, Winooski, Vt.) at a filter width of 630 nm. To prevent overloading
of the SDS gels, samples were diluted to a final concentration of ?4 ?g/ml.
Albumin standards (rabbit albumin [Sigma]) were run at concentrations of 0.5, 1,
2, 4, 6, 8, and 10 ?g/ml.
SDS-PAGE. Minigels were run according to the method described by Laemmli
(19). We used a 12% running gel, a 4.5% stacking gel, and a 0.25 M Tris–(1.92
M) glycine–1% SDS buffer. Samples were prepared by using 1 ?l of sample, 4 ?l
of distilled water, and 5 ?l of sample solubilizer. Eight microliters of sample was
loaded onto the gel, which was run at 175 V for 40 to 45 min. The gel was stained
with Coomassie brilliant blue (J. T. Baker, Inc., Danvers, Mass.) for 30 min and
destained with a 5% acetic acid solution. Standards included bovine serum
albumin (0.5, 2, and 4 ?g/ml), rabbit lens protein, and rabbit hemoglobin. Rabbit
lens protein was obtained by homogenizing surgical resected rabbit lenses after
the capsule had been removed. Rabbit hemoglobin was obtained from rabbit
erythrocytes thrice washed in phosphate-buffered saline and lysed in distilled
water; cell fragments were removed by centrifugation at 8,000 ? g (Microfuge
model 5415C; Brinkmann Instruments, Inc., Westbury, N.Y.). A molecular
weight standard (Enprotech, Natick, Mass.) and lens protein (diluted 40?) were
also run with each gel. Concentrations of albumin in vitreous humor and serum
samples were determined by densitometry.
Pharmacokinetic analyses. Pharmacokinetic analyses of the concentration-
time data for plasma and vitreous humor samples following systemic adminis-
tration were performed with RSTRIP (Micromath Scientific Software, Salt Lake
City, Utah), which is an iterative, nonlinear, weighted, least-squares regression
program. The most appropriate pharmacokinetic models were determined by
using the coefficient of determination and the RSTRIP model selection criterion,
which is a modified form of the Akaike (1) information criterion. Noncompart-
mental parameters, especially areas under the concentration-time curve (AUC),
were estimated by using statistical moment theory. Estimations for each expo-
nential coefficient and time constants were computed with the standard devia-
tions of each estimate, along with its 95% confidence range, which was calculated
by using both univariate and support plane approximations for the bounds of the
95% confidence range. On the basis of the coefficient of determination and the
model selection criterion, antibiotic concentration-time data for vitreous humor
and serum samples following intravenous administration were best fit by a two-
compartment model. Other standard pharmacokinetic parameters were deter-
mined by using computer-generated primary coefficients and standard pharma-
cokinetic equations (16).
Protein concentrations and pharmacokinetic parameters were compared by
using a two-sided Student t test.
Pharmacokinetics in albino rabbits following a single dose.
The pharmacokinetics of ofloxacin in eight NZW rabbits was
studied following a single intravenous bolus. To ensure that
multiple paracentesis did not cause a breakdown of the blood-
vitreous humor barrier in any animal, SDS-PAGE and then
densitometric determinations of concentrations of albumin in
the vitreous humor were performed for each ocular sample,
and concentrations of albumin in the vitreous humor were
compared with those in serum (20). Since the concentration of
albumin in the vitreous humor of rabbits and humans is ?1%
of that in serum, a vitreous humor-to-serum value of ?0.01
suggests a breakdown of the blood-eye barrier. Two of eight or
25% of animals had a breakdown of ?1%; data from animals
with a breakdown of ?1% were excluded from analysis. Pre-
vious studies with a larger group of rabbits (n ? 47) showed a
ratio of albumin in vitreous humor to albumin in serum of
?0.01 or no breakdown of the blood-eye barrier in 82% of
Figure 1 shows the pharmacokinetic curves for serum and
vitreous humor of albino animals which received a singe intra-
venous bolus of 40 mg of ofloxacin per kg. Hybrid pharmaco-
kinetic constants (distribution and elimination slopes and in-
tercepts in a two-compartment model) are shown in Table 1.
Concentration-time data are plotted arithmetically to help vi-
sualize the relative penetration of ofloxacin into the vitreous
humor and to compare the pharmacokinetic behavior of this
quinolone with those found in earlier ocular pharmacokinetic
studies with fleroxacin (29) and ciprofloxacin (25). To show the
relationship between slopes (rates) of terminal elimination
from serum and vitreous humor, these data are plotted semi-
logarithmically in the upper panel of Fig. 1. The penetration
ratio calculated from the AUC for vitreous humor (vitreous
humorAUC) over the AUC for serum (serumAUC) was 0.326,
which corresponds to a percent penetration of 32.6% ? 2.12%.
Terminal-elimination rate constants were only slightly greater
for serum (0.29) than for the vitreous humor (0.23), corre-
sponding to elimination rate half-lives (t1/2?s) of 2.39 and 3.21
Pharmacokinetics in pigmented rabbits following a single
dose and at a steady-state concentration of drug in serum.
Figure 2 shows the pharmacokinetic curves for serum and
vitreous humor in pigmented animals which received a single
intravenous bolus of ofloxacin. As with albino animals, one of
four animals studied were excluded from analysis on the basis
of a ratio of albumin in vitreous humor to albumin in serum of
?0.01. Concentration-time data are plotted arithmetically
(lower panel) and semilogarithmically (upper panel). Hybrid
constants are given in Table 1. The penetration value calcu-
lated from the vitreous humorAUC/serumAUCratio was 0.304
(i.e., a percent penetration of 30.4% ? 2.98%). Terminal elim-
ination rate constants for vitreous humor and serum were 0.16
and 0.29, which correspond to t1/2s of 4.32 and 2.64 h, respec-
tively. While elimination rates from the sera of albino and
pigmented animals receiving single doses were virtually iden-
tical, the elimination rate from the vitreous humor following a
single dose was higher for pigmented animals, although not
significantly (4.32 versus 3.21 h [P ? 0.10]).
Figure 3 shows the serum and vitreous pharmacokinetic
curves for pigmented animals with a steady-state concentration
1494PERKINS ET AL.ANTIMICROB. AGENTS CHEMOTHER.
of drug in serum. These animals were given 40 mg/kg every 8
h for four doses. On the basis of a t1/2?from serum of 3.12 h,
?99% of the animals had steady-state concentrations after the
fourth dose. The t1/2?from the vitreous humor was 6.05 h,
which was longer than that for albino rabbits receiving a single
dose (3.21 [P ? 0.05]) but not for pigmented animals receiving
a single dose (4.32 [P ? 0.10]). The percent penetration to the
last time point for pigmented animals with a steady-state con-
centration in serum of 30.0% ? 3.44% did not differ (P ? 0.10)
from that for animals receiving a single dose.
Ocular pharmacokinetic studies of rabbits showed that the
penetration of ofloxacin (31%) into the vitreous humor was
significantly better (P ? 0.05) than that of fleroxacin (28)
(9.6%) or ciprofloxacin (24) (5.45%). The terminal elimination
rate constant for ofloxacin from the vitreous humor (0.23) was
slightly less than that from the serum (0.29); these correspond
to t1/2?s of 3.21 and 2.39 h, respectively. Although in model-
dependent pharmacokinetic analysis the terminal elimination
rates from the serum and the ideal peripheral compartment
are identical (16), actual elimination rates from anatomical
compartments may differ from those from the serum (29).
Consistently, single-dose, ocular pharmacokinetic studies with
albino rabbits demonstrate that the t1/2?s of fleroxacin (29),
ciprofloxacin (25), fluconazole (24), and flucytosine (38) from
the vitreous humor are similar to or just slightly greater than
those from serum.
Several studies with the quinolones pefloxacin, enoxacin,
and ofloxacin have demonstrated that these compounds bind
to pigmented structures in the eye (8, 14, 27). This observation,
combined with the chemical similarity of the quinolones to
chloroquine, a compound with known retinal toxicity, has
prompted considerable interest in the binding of quinolones to
ocular tissue. In unpublished studies, we previously compared
the pharmacokinetics of ciprofloxacin in DB and NZW rabbits
following a single dose of ciprofloxacin (40 mg/kg). The serum
AUCs for DB and NZW rabbits were 11.5 versus 9.3 ?g ? h/ml,
and the percent penetration into the aqueous and vitreous
humors were 21 versus 18% and 5.45% ? 1.73% versus 5.52%
? 0.63%, respectively. Pharmacokinetic studies of ofloxacin in
pigmented rabbits were performed following both a single dose
and at a steady-state concentration of drug in serum. While
animals had steady-state concentrations of drug in serum of
more than 90% after a single dose, multiple-dose studies were
done to ensure saturation of putative binding sites in pig-
mented tissue. The percent penetration values for singly and
multiply dosed pigmented animals did not differ from each
other (30.4% ? 2.98% and 30.0% ? 3.44%, respectively) or
from that for albino rabbits after a single dose (32.6% ? 2.12%
and 30.0% ? 3.44%). Consistently, studies comparing the
pharmacokinetics of ciprofloxacin in pigmented and albino
animals also showed similar values of penetration into the
vitreous humor (unpublished data). t1/2?s of ofloxacin in the
vitreous humor in singly and multiply dosed pigmented animals
were 4.32 and 6.05 h, respectively. The ratios of elimination
rates from the vitreous humor to elimination rates from the
serum were 1.34 and 1.67 for singly dosed albino and singly
dosed pigmented animals, whereas that for multiply dosed
pigmented animals was 2.23 (P ? 0.05). We conclude that
while the percent penetration values are similar, elimination of
ofloxacin is slower for pigmented animals.
There has been considerable interest in the deposition of the
quinolones in the eye because of the chemical similarity of
these drugs to chloroquine. However, there are no data sug-
gesting differences in the penetration of quinolones into the
vitreous humor for pigmented versus nonpigmented rabbits.
FIG. 1. Model-dependent pharmacokinetic curves for albino NZW rabbits
receiving a single dose of ofloxacin. Shown are concentrations in both serum
(squares) and vitreous humor (triangles). The upper and lower panels, respec-
tively, plot data semilogarithmically and arithmetically.
TABLE 1. Hybrid pharmacokinetic constants for serum and
vitreous humor samples from albino and pigmented rabbitsa
27.78 (9.72) 13.69 (2.56)4.45 (2.07)0.35 (0.07)
38.69 (13.78)17.66 (3.71) 3.36 (0.48) 0.28 (0.05)
21.11 (10.51)16.00 (3.27)1.90 (1.26)0.21 (0.05)
?3.82 (0.78)3.85 (0.78) 2.71 (0.80)0.18 (0.04)
?4.55 (0.56)4.71 (0.64) 3.03 (0.96)0.16 (0.04)
?5.22 (1.08)5.24 (1.08) 2.66 (0.56)0.11 (0.04)
aValues are means ? standard deviations. A and B, distribution- and elimi-
nation-phase intercepts, respectively; ? and ?, distribution- and elimination-
phase slopes, respectively.
VOL. 39, 1995 PHARMACOKINETICS OF OFLOXACIN1495
These results are different from those of others following sub-
conjunctive injection that showed that binding to choroid ret-
inas in pigmented rabbits increased ocular concentrations (5,
39). They (5, 39) did not study the effect of pigmentation on
levels of drug in the vitreous humor following systemic admin-
istration. Recent studies by Mounier et al. (31) compared con-
centrations of ofloxacin in the vitreous humor for uninfected
pigmented and nonpigmented rabbits. While concentrations in
the vitreous humor for uninfected animals were somewhat
higher than those for albino rabbits, no data for serum were
provided. As a result, penetration values could not be com-
pared. In vitro studies comparing concentrations of ofloxacin
in the choroid retinas of pigmented and nonpigmented animals
(14) showed higher concentrations for pigmented rabbits.
Studies by Cochereau-Massin and colleagues also showed
higher concentrations for ocular tissues from pigmented versus
nonpigmented animals (8). These investigators compared the
pharmacokinetics of another quinolone, pefloxacin, in pig-
mented versus nonpigmented-rabbit serum, vitreous humor,
aqueous humor, cornea, iris, lens, choroid retina, and sclera
(8). Despite higher concentrations of pefloxacin in ocular tis-
sues from pigmented rabbits, vitreous humorAUC/serumAUC
penetration ratios (0.39 versus 0.45) were similar. While these
studies used a less robust method (single samples from indi-
vidual animals were combined in the so-called naive pooled-
datum method) to measure penetration, these data and studies
with ofloxacin and ciprofloxacin show that values of quinolone
penetration are similar for pigmented and nonpigmented ani-
mals. To compare the effects of pigmentation on ocular elim-
ination of pefloxacin with our data characterizing ofloxacin, we
recalculated the t1/2?s and hybrid constants for pefloxacin using
RSTRIP II. As with ofloxacin, the t1/2?for the vitreous humor
(2.53 h) was slightly longer than that for serum (2.06 h) from
albino animals; the percent penetration values were 32 versus
39% for albino and pigmented animals, respectively (P ?
0.05). The t1/2?for pefloxacin in pigmented animals (4.3 h) was
longer than that in albino animals, with ratios of t1/2in vitreous
humor to t1/2in serum of 1.22 and 1.70 for albino and pig-
mented animals, respectively. Both the percent penetration
values and the t1/2ratios were virtually identical to those for
ofloxacin (31% and 1.34 and 1.67 for singly dosed albino and
pigmented animals, respectively).
In vitro studies characterizing the binding of pharmaceuti-
cals in ocular tissue from pigmented and albino animals by
Menon and colleagues (28) assessed the binding of timolol to
melanin and to melanin-containing tissues obtained from cows,
humans, and rabbits. Binding was directly related and release
was inversely related to the amount of melanin present in
ocular tissue. In vivo and in vitro studies of rabbits by Barza et
FIG. 2. Model-dependent pharmacokinetic curves for pigmented DB rabbits
receiving a single dose of ofloxacin. Shown are concentrations in both serum
(squares) and vitreous humor (triangles). The upper and lower panels, respec-
tively, plot data semilogarithmically and arithmetically.
FIG. 3. Model-dependent pharmacokinetic curves for pigmented DB rabbits
receiving multiple doses of ofloxacin to achieve steady-state concentrations of
drug in serum. Shown are concentrations in both serum (squares) and vitreous
humor (triangles). The upper and lower panels, respectively, plot data semilog-
arithmically and arithmetically.
1496 PERKINS ET AL.ANTIMICROB. AGENTS CHEMOTHER.
al. also showed more binding to tissues from pigmented ani-
mals than to tissues from nonpigmented animals with respect
to clindamycin (5). As with Menon’s studies, differences in
binding appeared to correlate with melanin organelles rather
than melanin itself (28).
In summary, our studies and those of others (8) show that
there are no differences in the penetration values of quino-
lones into the vitreous humor of pigmented and nonpigmented
animals, despite marked binding of quinolones to tissue such
as the iris and choroid retina from pigmented animals. On the
other hand, terminal-elimination rates are lower for pigmented
versus nonpigmented subjects. This observation is consistent
with the studies of Menon et al. showing that drugs which bind
to pigmented retinal tissue are slowly released. We believe that
the differences in elimination rates of ofloxacin and pefloxacin
from the vitreous humor are likely explained by the sink effect
(17). The clinical significance of retinal binding and the asso-
ciated prolongation in elimination from the eye are uncertain.
Preliminary studies suggest that retinal binding of quinolones
is not associated with toxicity.
In previous studies, we showed that quinolones were rapidly
eliminated following direct intravitreal injection. Because of
the relationship between elimination rate constants for serum
and vitreous humor mentioned previously, ocular t1/2?s for
antimicrobial agents including amikacin (26), chloramphenicol
(26), fleroxacin (29), and flucytosine (38) are considerably
longer following systemic administration than those following
direct ocular administration. As a result, we believe systemic
administration is preferable for antimicrobial agents which
penetrate into the vitreous humor.
Finally, it is important to emphasize that our pharmacoki-
netic studies were performed with animals with uninflamed
eyes. Since inflammation increases the penetration of drugs
into the vitreous humor (2, 4, 21, 22), we believe that quino-
lones will show greater penetration in infected animals (studies
in progress). Ofloxacin is more active against coagulase-nega-
tive staphylococci, achieves higher concentrations in serum,
and shows better penetration into the vitreous humor than
ciprofloxacin. We believe that ofloxacin may prove to be more
efficacious than ciprofloxacin in the treatment of acute bacte-
rial endophthalmitis. Moreover, since the quinolones are bac-
tericidal against slowly growing organisms (13) and penetrate
into phagocytes (37), these antimicrobial agents may also be
useful in the therapy of subacute or chronic infections due to
This work was supported in part by the National Eye Institute grant
1. Akaike, H. 1973. A new look at the statistical model identification. IEEE
Trans. Automat. Contr. 1973:716–723.
2. Badenoch, P. R., P. J. McDonald, and D. J. Coster. 1986. Effect of inflam-
mation on antibiotic penetration into the anterior segment of the rat eye.
Invest. Ophthalmol. Visual Sci. 27:958.
3. Barza, M. 1988. Pharmacokinetics and efficacy of the new quinolones
in infections of the eye, ear, nose and throat. Rev. Infect. Dis. 10:S241–
4. Barza, M., A. Kane, and J. Baum. 1983. Comparison of the effects of
continuous and intermittent systemic administration on the penetration of
gentamicin into infected rabbit eyes. J. Infect. Dis. 147:144–148.
5. Barza, M., A. Kane, and J. Baum. 1979. Marked differences between pig-
mented and albino rabbits in the concentration of clindamycin in iris and
choroid-retina. J. Infect. Dis. 139:203–208.
6. Borner, K., E. Borner, and H. Lode. 1992. Determination of sparfloxacin in
serum and urine by high performance liquid chromatography. J. Chro-
7. Bron, A., D. Talon, B. Delbosc, J. M. Estavityer, G. Kaya, and J. Royer. 1987.
La penetration intracamerulaire de l’oflocacine chez l’homme. J. Fr. Oph-
8. Chochereau-Massin, I., J. Bauchet, F. Faurisson, J. M. Vallois, P. Lacombe,
and J. J. Pocidalo. 1991. Ocular kinetics of pefloxacin after intramuscular
administration in albino and pigmented rabbits. Antimicrob. Agents Che-
9. Diamond, J. G. 1981. Intraocular management of endophthalmitis: a sys-
temic approach. Arch. Ophthalmol. 99:96–99.
10. Driebe, W. T., S. Mandelbaum, R. K. Forster, L. K. Schwartz, and W. W.
Culbertson. 1986. Pseudophakic endophthalmitis. Ophthalmology 93:442–
11. Drusano, G. L. 1989. Pharmacokinetics of the quinolone antimicrobial
agents, p. 71–105. In J. S. Wolfson and D. C. Hooper (ed.), Quinolone
antimicrobial agents. American Society for Microbiology, Washington, D.C.
12. El Baba, F. A., M. D. Trousdale, W. J. Gauderman, D. G. Wagner, and
P. E. R. Ligget. 1992. Penetration of oral ciprofloxacin into the aqueous
humor, vitreous and subretinal fluid of humans. Ophthalmology 99:483–
13. Eng, R. H. K., F. T. Padberg, S. M. Smith, E. N. Tan, and C. E. Cherubin.
1991. Bactericidal effects of antibiotics on slowly growing and nongrowing
bacteria. Antimicrob. Agents Chemother. 35:1824–1828.
14. Fukuda, M., and K. Sasaki. 1989. Intraocular dynamic mode differences of
new quinolone antibacterial agents between pigmented and albino rabbit
eyes. Lens Eye Toxic Res. 6:339–351.
15. Fisch, A., C. Lafaix, A. Salvanet, M. Cherifi, and A. Meulemans. 1987.
Ofloxacin in human aqueous humour and lens. J. Antimicrob. Chemother.
16. Gibaldi, M., and D. Perrier. 1982. Pharmacokinetics, 2nd ed., vol 15. New
York: Marcel Dekker.
17. Guy, R. H., and J. Hadgraft. 1981. Theoretical comparisons of release rates
of drugs into sink and nonsink conditions. J. Pharm. Sci. 70:1243–1245.
18. Kanella, D., G. Petrikkos, H. Giamarellou, E. Kavouklis, Z. Chryssoulis, and
L. Abriel. 1990. Comparative humor kinetics of ciprofloxacin, ofloxacin and
pefloxacin into the human eye, p. 475. Abstr. 3rd Int. Symp. New Quino-
lones. Vancouver, Canada, 12 to 14 July, 1990.
19. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of
the head of bacteriophage T4. Nature (London) 227:680–685.
20. Liu, W., G. Drusano, M. Mayers, A. Madu, and M. Miller. 1994. The
determination of robust ocular pharmacokinetic parameters from sparce
data sets using non-linear mixed effects modeling, abstr. 4452, p. 2216. Abstr.
Invest. Ophthalmol. Visual Sci. ARVO, Sarasota, Fla., May 1994.
21. Mandell, B. A., T. A. Meredith, E. Aguilar, A. el-Massry, A. Sawant, and S.
Gardner. 1993. Effects of inflammation and surgery on amikacin levels in the
vitreous cavity. Am. J. Ophthalmol. 115:770–774.
22. Martin, D. F., L. A. Ficker, H. A. Aguilar, S. K. Gardner, L. A. Wilson, and
T. A. Meredith. 1990. Vitreous cefazolin levels after intravenous injection:
effects of inflammation, repeated antibiotic doses and surgery. Arch. Oph-
23. Massin-Cochereau, I., P. Le Hoang, M. Weiser, and F. Rousselie. 1987.
Pefloxacin and severe eye infections. Bull. Soc. Ophtalmol. Fr. 87:735–
24. Mayers, M., Y. Garg, U. Mian, G. Newcomer, W. Liu, G. Drusano, and
M. Miller. 1994. Comparison of fluconazole pharmacokinetics in serum,
aqueous, vitreous and CSF following single dose and serum steady state, p.
2216. Abstr. Invest. Ophthalmol. Visual Sci. ARVO, Sarasota, Fla., May
25. Mayers, M., A. Madu, E. Schwartz, R. Perkins, M. Dezeil, and M. Miller.
1993. Aqueous and vitreous pharmacokinetics of ciprofloxacin and fleroxa-
cin, abstr. no. 2730, p. 1257. Abstr. Invest. Ophthalmol. Visual Sci. ARVO,
Sarasota, Fla., May 1993.
26. Mayers, M., D. Rush, A. Madu, M. Motyl, and M. H. Miller. 1991. Pharma-
cokinetics of amikacin and chloramphenicol in the aqueous humor of rab-
bits. Antimicrob. Agents Chemother. 35:1791–1798.
27. McNally, W., P. DeHart, P. Misiak, R. Toothaker, and T. Chang. 1990.
Quantitative autoradiographic determination of C-enoxacin in eye struc-
ture of pigmented and non-pigmented rats. International Conference and
Workshop on In Vitro and In Vivo Autoradiographic and Correlative
Imaging, 1 to 3 August 1990. University of North Carolina, Chapel Hill,
28. Menon, I. A., G. E. Trope, P. K. Basu, D. C. Wakeham, and S. D. Oersad.
1989. Binding of timolol to iris-ciliary body and melanin: an in vitro model
for assessing the kinetics and efficacy of long-acting antiglaucoma drugs. J.
Ocul. Pharmacol. 5:313–323.
29. Miller, M. H., A. Agocha-Madu, G. Samathanam, D. Ruch, C. Madu, K.
Mathisson, and M. Mayers. 1992. Aqueous and vitreous humor fleroxacin
pharmacokinetics using complete concentration-time data from individual
rabbits. Antimicrob. Agents Chemother. 36:32–38.
30. Mishima, S. 1982. Ocular effects of beta-adrenergic agents. XII Jules Stein
Lecture. Surv. Ophthalmol. 27:187–208.
31. Mounier, M., M. C. Ploy, M. Chauvin, J. P. Adenis, and F. Denis. 1992. Study
of intraocular diffusion of ofloxacin in humans and rabbits. Pathol. Biol.
VOL. 39, 1995PHARMACOKINETICS OF OFLOXACIN1497
32. Olson, J. C., H. W. Flynn, R. K. Forster, and W. W. Culbertson. 1983.
Results in treatment of postoperative endophthalmitis. Ophthalmology 90:
33. Ooishi, M., and M. Miyao. 1990. Ocular pharmacokinetic studies on spar-
floxacin, abstr. no. 42. Abstr. 30th Intersci. Conf. Antimicrob. Agents Che-
mother., Atlanta, Ga., 22 to 24 October 1990. American Society for Micro-
biology, Washington, D.C.
34. Peyman, G. A., and A. Schulman. 1989. Intravitreal drug therapy. Jpn. J.
35. Puliafito, C. A., A. S. Baker, J. Haaf, and C. S. Foster. 1982. Infectious
endophthalmitis. Ophthalmology 89:921–929.
36. Salvanet-Bouccara, A., J. Feys, B. Koffi-Gue, A. Fisch, M. Cherifi, and F.
Forestier. 1989. Bacterial endophthalmitis: treatment with a combination of
fosfomycin and pefloxacin apropos of 9 cases. Bull. Soc. Ophtalmol. Fr.
37. Tulkens, P. M. 1991. Intracellular distribution and activity of antibiotics. Eur.
J. Clin. Microbiol. Infect. Dis. 10:100–106.
38. Walsh, A., D. A. Haft, M. H. Miller, M. R. Loran, and A. H. Friedman. 1978.
Ocular penetration of 5-fluorocytosine. Invest. Ophthalmol. Visual Sci. 17:
39. Young, P., M. Barza, A. Kane, and J. Baum. 1979. Radioactive and bioassay
of intraocular antibiotics: double-assay technique to compare penicillin G,
cefamandole and gentamicin in ocular tissues in vivo. Arch. Ophthalmol.
1498 PERKINS ET AL.ANTIMICROB. AGENTS CHEMOTHER.