J Infect Chemother xxx (xxxx) xxx

Contents lists available at ScienceDirect

Journal of Infection and Chemotherapy

journal homepage: http://www.elsevier.com/locate/jic

Original Article

Clinical pharmacokinetics of flomoxef in prostate tissue and dosing

considerations for prostatitis based on site-specific pharmacodynamic
target attainment
Kogenta Nakamura a, Kazuro Ikawa b, *, Genya Nishikawa a, Ikuo Kobayashi a,
Motoi Tobiume c, Miho Sugie a, Hiroyuki Muramatsu a, Shingo Morinaga a,
Keishi Kajikawa a, Masahito Watanabe a, Kent Kanao a, Tetsushu Onita b,
Norifumi Morikawa b
a
Department of Urology, Aichi Medical University School of Medicine, Nagakute, Aichi, 480-1195, Japan
b
Department of Clinical Pharmacotherapy, Hiroshima University, 1-2-3 Kasumi, Minami-ku, Hiroshima, 734-8551, Japan
c
Department of Urology, Asahi Rousai Hospital, Nagoya, Aichi, 488-8585, Japan

a r t i c l e i n f o a b s t r a c t

Article history: Flomoxef is used to treat bacterial prostatitis; however, its prostatic pharmacokinetics have not been fully
Received 19 July 2019 clarified. Flomoxef (500 or 1000 mg) was administered to patients with benign prostatic hypertrophy
Received in revised form (n ¼ 54). After a 0.5-h infusion, venous blood samples were drawn at time points of 0.5e5 h, and prostate
22 August 2019
tissue samples were collected at time points of 0.5e1.5 h during transurethral resection of the prostate.
Accepted 29 August 2019
The drug concentrations in plasma and prostate tissue were analyzed pharmacokinetically and used for a
Available online xxx
stochastic simulation to predict the probability of attaining pharmacodynamic target in prostate tissue.
Showing dose linearity in the prostatic pharmacokinetics, flomoxef rapidly penetrated into prostate
Keywords:
Flomoxef
tissue, with a prostate/plasma ratio of 0.48e0.50 (maximum drug concentration) and 0.42e0.55 (area
Pharmacokinetics under the drug concentration-time curve). Against the tested populations of Escherichia coli, Klebsiella
Pharmacodynamics and Proteus species isolates, 0.5-h infusion of 1000 mg three times daily achieved a 90% expected
Prostatitis probability of attaining the bactericidal target (70% of the time above the minimum inhibitory concen-
tration [MIC]) in prostate tissue. The site-specific pharmacodynamic-based breakpoint (the highest MIC
at which the target-attainment probability in prostate tissue was >90%) values were 0.25 mg/L (MIC for
90th percentile of E. coli and Klebsiella species) for 500 mg four times daily and 0.5 mg/L (MIC90 of Proteus
species) for 1000 mg four times daily. These results help to fully characterize the prostatic pharmaco-
kinetics of flomoxef, while also helping to rationalize and optimize the dosing regimens for prostatitis
based on site-specific pharmacodynamic target attainment.
© 2019 Japanese Society of Chemotherapy and The Japanese Association for Infectious Diseases.
Published by Elsevier Ltd. All rights reserved.

1. Introduction strains producing b-lactamase enzymes especially extended-

Flomoxef is a b-lactam antibiotic, also called oxacephem, which
has antibacterial activities against a variety of Gram-positive and
Gram-negative aerobic and anaerobic bacteria. It is approved in
Japan, China, Korea and Taiwan [1]. However, flomoxef has gained
attention from regions other than the approved countries [2,3]
because it is stable and effective against worldwide problematic
* Corresponding author.
E-mail address: ikawak@hiroshima-u.ac.jp (K. Ikawa).
spectrum b-lactamases (ESBL) [4,5]. Flomoxef is used to treat
various infections including bacterial prostatitis [1,6]. It is also used
for antibacterial prophylaxis in prostatic surgery or biopsy [7,8],
particularly when the patient is intolerant or refractory to quino-
lone antibiotics or the causative bacterial strains are ESBL-
producing.
Antibiotics act at the site of infection. Therefore, it is important
to understand the pharmacokinetics of flomoxef in the prostate in
order to improve its efficacy in treating prostatitis. There have been
reports measuring flomoxef concentrations in human prostate [9]
as well as in other body sites including lung [10] and sputum

https://doi.org/10.1016/j.jiac.2019.08.019
1341-321X/© 2019 Japanese Society of Chemotherapy and The Japanese Association for Infectious Diseases. Published by Elsevier Ltd. All rights reserved.

Please cite this article as: Nakamura K et al., Clinical pharmacokinetics of flomoxef in prostate tissue and dosing considerations for prostatitis
based on site-specific pharmacodynamic target attainment, J Infect Chemother, https://doi.org/10.1016/j.jiac.2019.08.019
2 K. Nakamura et al. / J Infect Chemother xxx (xxxx) xxx
[11]. However, no attempt has been made to describe and evaluate
the site-specific pharmacokinetics and pharmacodynamics with
mathematical modeling and stochastic simulation.
The antibacterial and therapeutic effects of flomoxef correlate
with the exposure time for which the drug concentrations remain
above the minimum inhibitory concentration for the bacterium
(T > MIC) [1,12,13], particularly at the site of action. Therefore, in
order to define effective doses and optimize the treatment regi-
mens for prostatitis, it is important to estimate the pharmacody-
namic exposure in the prostate by predicting the probability that
the T > MIC target is attained.
The present study thus aimed to characterize the prostatic
pharmacokinetics of flomoxef and to estimate its site-specific
pharmacodynamic target attainment.
2. Materials and methods
2.1. Study design and subject population
This study was a prospective and open-label study of flomoxef
conducted at Aichi Medical University Hospital and its related fa-
cilities between February 2014 and February 2015. The study pro-
tocol and informed consent form were in compliance with the
Declaration of Helsinki (World Medical Association) and were
reviewed and approved by the ethics committee (Aichi Medical
University School of Medicine).
Study subjects were male patients with benign prostatic hy-
pertrophy who underwent transurethral resection of the prostate.
The patient inclusion criteria were as follows: older than 20 years
old, amenable to antibacterial prophylaxis for postoperative in-
fections, and willing and able to provide their written informed
consent. Patients were excluded who were hypersensitive to b-
lactams and who had preoperative pyuria or bacteriuria.
2.2. Drug administration and sample collection
The dose of flomoxef was assigned to 500 mg in the former
period (February to June, 2014) and 1000 mg in the latter period
(June 2014 to February 2015). After preoperative drug administra-
tion by a 0.5-h intravenous infusion, venous blood samples were
scheduled to be drawn 0.5, 1, 1.5, 3 and 5 h. Plasma was immediately
separated from the blood by centrifugation. Prostate tissue samples
were scheduled to be collected during transurethral resection of
the prostate 0.5, 1 and 1.5 h after the start of the infusion. The
resected prostate was immediately washed with physiological sa-
line. All of the plasma and prostate samples were stored at 40  C
until concentration measurement.
2.3. Measurement of flomoxef concentrations in plasma and
prostate tissue
The total concentrations of flomoxef in plasma and prostate
tissue were measured by high-performance liquid chromatography
(HPLC) with minor modifications of the methods of Konaka et al.
[14] Prostate tissue samples (0.5 g) were homogenized using an
overhead mixer with 2 vol (1 mL [w/v]) of double-distilled water.
The prostate tissue homogenate was centrifuged, and the super-
natant was collected. The tissue supernatants or plasma samples
(200 mL each) were then added to 400 mL of methanol, and the
mixture was vortexed and centrifuged. Next, the supernatants
(20 mL) were injected into an HPLC system. The HPLC employed a
C18 column at a temperature of 40  C and detected flomoxef at a
wavelength of 254 nm ultraviolet absorbance. The mobile phase
consisted of a mixture of 60 mmol/L sodium phosphate buffer (pH
6.0) containing 5 mmol/L tetrabutylammonium hydroxide and
Please cite this article as: Nakamura K et al., Clinical pharmacokinetics of flomoxef in prostate tissue and dosing considerations for prostatitis
based on site-specific pharmacodynamic target attainment, J Infect Chemother, https://doi.org/10.1016/j.jiac.2019.08.019
acetonitrile (85:15 [v/v]) with a flow rate of 1 mL/min. The quan-
tification limits for flomoxef were 0.25 mg/L and 1.5 mg/kg in
plasma and prostate tissue, respectively, and both calibration
curves were linear up to 250 mg/L and 150 mg/kg, respectively. The
interday and intraday accuracy (as absolute values of relative errors
of the means) and precision (as coefficients of variations) were
within 10%.
2.4. Noncompartmental pharmacokinetic analysis
The Cmax was defined as the maximum observed concentration
of flomoxef. Based on the trapezoidal rule using the MULTI program
[15], the area under the drug concentration-time curve from 0 to
infinity (AUC0e∞) was estimated as the actual area from 0 to 1.5 h
(AUC0e1.5) plus the extrapolated area (AUC1.5e∞ ¼ C1.5/lZ, where C1.5
is the drug concentration at 1.5 h and lZ is the terminal slope on a
loge scale). The drug penetration ratio into prostate tissue was
defined as the Cmax or AUC ratio, which are often used as indices of
drug distribution into tissues. The specific gravity of prostate tissue
was defined as 1 (kg ¼ L).
2.5. Simultaneous pharmacokinetic modeling
All of the flomoxef concentration data for plasma and prostate
tissue were simultaneously fitted to a three-compartment phar-
macokinetic model (1, central; 2, peripheral; 3, prostate), as this
general model has been used for prostatic pharmacokinetic ana-
lyses [16e19]. In this population pharmacokinetic modeling using
the NONMEM program (version 7.4; ICON Public Limited Company,
Dublin, Ireland), the fixed-effects parameters were as follows: V1,
V2 and V3 are the volumes of distribution of the central, peripheral
and prostate compartments (L), CL is the clearance (L/h). K12, K21,
K13 and K31 are the first-order transfer rate constants connecting
the compartments (1/h). The interindividual variability was
modeled with an exponential error model: qi ¼ q*exp(hi), where qi
is the fixed-effects parameter for the i-th subject, q is the mean
value of the fixed-effects parameter in the population, and h is a
random interindividual variable, which is normally distributed
with a mean zero and variance u2. The residual variability was
modeled with a proportional error model: Cobs, ij ¼ Cpred, ij*(1 þ εij),
where Cobs, ij and Cpred, ij denote the j-th observed and predicted
concentrations for the i-th subject, and ε is a random intra-
individual error, which is normally distributed with a mean zero
and variance s2.
The adequacy of the developed population model was assessed
by goodness-of-fit plots. To assess the reliability and stability of the
estimated parameters, a nonparametric bootstrap method was
performed using the Perl-speaks-NONMEM program [20]. The 95%
confidence intervals of the parameters from 1000 bootstrap repli-
cates were compared with the estimates of the developed popu-
lation model.
2.6. Site-specific pharmacodynamic simulation
A set of fixed-effects parameters qi (V1, V2, V3, CL, K12, K21, K13
and K31) for flomoxef was randomly generated 5000 times by the
$SIMULATION command in NONMEM, according to each mean
estimate and interindividual variance of the developed model. The
set of eight qi values gave the three-compartment model equations
and simulated the flomoxef concentration in prostate tissue. The
time point at which the drug concentration coincided with a MIC
value (0.125e64 mg/L) was determined, and the duration of drug
exposure above the MIC for bacteria (T > MIC) was calculated as the
cumulative percentage of 24 h. Using the same method as for the
earlier simulation [21,22], the total concentration in prostate tissue
 K. Nakamura et al. / J Infect Chemother xxx (xxxx) xxx
was not adjusted for the free fraction because the protein binding of
flomoxef in the prostate is currently unknown. As the protein
binding of flomoxef in human serum has been reported to be
relatively low (approximately 30%) [23], even a moderately higher
or lower protein binding in the prostate compared to serum was
not expected to significantly affect this simulation.
Based on the simulation, the probability of attaining the phar-
macodynamic target in prostate tissue was defined as the propor-
tion achieving 40% T > MIC (bacteriostatic target) or 70% T > MIC
(bactericidal target) [4,12,13] among 5000 estimates. The proba-
bility at a specific MIC was then multiplied by the fraction of the
clinical isolate population at each MIC category, and the sum of the
individual products was determined as the expected population
probability of attaining the pharmacodynamic target in prostate
tissue (%). MIC distributions for flomoxef against common bacteria
causing prostatitis were derived from nationwide susceptibility
surveillance data in Japan (Fig. 1) [24,25]. The MICs for the 50th
percentile (MIC50) and 90th percentile (MIC90) of the clinical iso-
lates were 0.125 and 0.25 mg/L for both Escherichia coli and Kleb-
siella species, and 0.25 and 0.5 mg/L for Proteus species,
respectively.
3. Results
The demographic values (mean ± standard deviation) for the
study subjects in the flomoxef 500 mg (n ¼ 25) and 1000 mg
(n ¼ 29) groups were as follows: age, 71.2 ± 6.4 and 73.0 ± 5.5 years
old; body weight, 62.2 ± 8.7 and 62.3 ± 8.4 kg; height, 165.4 ± 7.5
and 162.2 ± 6.0 cm; serum creatinine, 0.90 ± 0.22 and
0.95 ± 0.28 mg/dL; blood urea nitrogen, 16.3 ± 4.8 and
17.2 ± 3.6 mg/dL. No significant differences between the two groups
were noted in any demographic values, and all of the subjects had a
normal renal and hepatic function.
The noncompartmental pharmacokinetic parameters are sum-
marized in Table 1. For flomoxef 500 mg, the mean Cmax values were
38.9 mg/L in plasma and 19.0 mg/kg in prostate tissue. The mean
AUC0e1.5 and AUC0e∞ values were 31.1 and 53.1 mgh/L in plasma,
and 16.8 and 23.3 mgh/kg in prostate tissue, respectively. For
flomoxef 1000 mg, the corresponding values increased 1.80e2.11
times (nearly 1000 mg/500 mg ratio of 2), showing dose linearity in
the prostatic pharmacokinetics. For flomoxef 500 and 1000 mg, the
mean prostate/plasma ratios were 0.48e0.50 in Cmax, 0.54e0.58 in
AUC0e1.5, and 0.42e0.48 in AUC0e∞.
The pharmacokinetic model parameters of flomoxef in the
three-compartment model are listed in Table 2. The interindividual
variability (h) for V2, K12 and K13 was finally fixed as 0 in the
Fig. 1. Minimum inhibitory concentration (MIC) distributions for flomoxef against
clinical isolates of Escherichia coli (n ¼ 230; MIC50 ¼ 0.125 mg/L, MIC90 ¼ 0.25 mg/L),
Klebsiella species (n ¼ 183; MIC50 ¼ 0.125 mg/L, MIC90 ¼ 0.25 mg/L) and Proteus
species (n ¼ 99; MIC50 ¼ 0.25 mg/L, MIC90 ¼ 0.5 mg/L).
Please cite this article as: Nakamura K et al., Clinical pharmacokinetics of flomoxef in prostate tissue and dosing considerations for prostatitis
based on site-specific pharmacodynamic target attainment, J Infect Chemother, https://doi.org/10.1016/j.jiac.2019.08.019
3
NONMEM modeling because it was very small (<0.0001). For V1, V3,
CL, K21 and K31, there was no significant covariate that correlated
well with the individual pharmacokinetic parameters and
explained their interindividual variability. The parameter estimates
were all in the range of the 95% confidence intervals obtained using
the bootstrap method.
The observed flomoxef concentrations in plasma and prostate
tissue were in good correspondence with the individual predicted
concentrations (Fig. 2). Using the mean fixed-effects parameters
(Table 2), the simulation curves fitted well with all of the mean
measurements for plasma and prostate tissue in the flomoxef 500
and 1000 mg groups (Fig. 3). Based on the simulation curves, the
mean AUC0e∞ values for 500 mg were estimated to be 52.1 mgh/L
in plasma and 24.6 mgh/kg in prostate tissue, and the prostate/
plasma ratio was 0.47.
The probabilities of attaining bacteriostatic and bactericidal
targets in prostate tissue, at specific MICs, are given in Fig. 4.
Regarding the bacteriostatic target of 40% T > MIC (Fig. 4a), the site-
specific pharmacodynamic-based breakpoint MIC (the highest MIC
at which the target-attainment probability in prostate tissue was
90%) values were as follows: 0.25 mg/L for 500 mg twice daily,
0.5 mg/L for 500 mg three times daily and 1000 mg twice daily,
1 mg/L for 500 mg four times daily and 1000 mg three times daily,
and 2 mg/L for 1000 mg four times daily. Regarding the bactericidal
target of 70% T > MIC (Fig. 4b), the breakpoint MIC values were as
follows: 0.125 mg/L for 500 mg three times daily, 0.25 mg/L for both
500 mg four times daily and 1000 mg three times daily, and 0.5 mg/
L for 1000 mg four times daily. Neither regimens of 500 mg twice
daily nor 1000 mg twice daily achieved a 90% probability.
The expected population probabilities of attaining the phar-
macodynamic targets in prostate tissue against bacterial pop-
ulations of E. coli, Klebsiella and Proteus species isolates are shown
in Table 3. For the bacteriostatic target, except for the flomoxef
regimen of 500 mg twice daily, all of the tested regimens achieved
an expected probability of 90% in prostate tissue against the three
bacterial populations. However, for the bactericidal target, only the
regimens of 1000 mg three times daily and 1000 mg four times
daily achieved an expected probability of 90% against the three
bacterial populations.
4. Discussion
The present study investigated the prostatic pharmacokinetics
of flomoxef in patients with benign prostatic hypertrophy and
defined the drug penetration from the systemic circulation into the
tissue (a prostate/plasma ratio of around 0.5). The study also con-
ducted simultaneous population pharmacokinetic modeling of
flomoxef concentrations in plasma and prostate tissue and per-
formed stochastic simulation to estimate the probabilities of
attaining the pharmacodynamic targets in prostate tissue. Aiming
for the bactericidal target of 70% T > MIC, 0.5-h infusion of 1000 mg
three times daily (3 g/day) achieved a 90% expected probability
against the tested populations of bacteria that cause prostatitis. The
bactericidal regimens required for a prostatic pharmacodynamic-
based breakpoint of 0.25 mg/L (MIC90 of the E. coli and Klebsiella
species isolates) and 0.5 mg/L (MIC90 of the Proteus species isolates)
were 500 mg four times daily (2 g/day) and 1000 mg four times
daily (4 g/day), respectively.
In the noncompartmental pharmacokinetic analysis, the mean
prostate/plasma concentration ratios were 0.524 (¼ 19.8 [mg/kg]/
37.8 [mg/L]) at 0.5 h and 0.480 (¼11.9/24.8) at 1 h after the end of
0.5-h infusion of 1000 mg (Table 1 and Fig. 3). Hoshi et al. reported
the mean prostate/plasma concentration ratios in 27 patients with
benign prostatic hypertrophy to be 0.423 (¼ 18.6/44.0) at 0.5 h and
0.456 (¼ 13.4/29.4) at 1 h after the end of one-shot injection of
4 K. Nakamura et al. / J Infect Chemother xxx (xxxx) xxx

Table 1
Noncompartmental pharmacokinetic parameters of flomoxef after 0.5-h infusions.

Sample type and parameter Mean ± standard deviation

Flomoxef 500 mg (n ¼ 25) Flomoxef 1000 mg (n ¼ 29) Ratio of mean values (Flomoxef 1000 mg/500 mg)

Plasma
Cmax (mg/L) 38.9 ± 8.8 76.6 ± 13.8 1.97
AUC0e1.5 (mgh/L) 31.1 ± 6.9 62.6 ± 12.0 2.01
AUC0e∞ (mgh/L) 53.1 ± 15.8 111.9 ± 30.1 2.11
Prostate
Cmax (mg/kg) 19.0 ± 4.7 35.3 ± 9.3 1.86
AUC0e1.5 (mgh/kg) 16.8 ± 2.3 31.3 ± 4.9 1.86
AUC0e∞ (mgh/kg) 23.3 ± 5.4 41.9 ± 6.5 1.80
Prostate/plasma ratio
Cmax ratio 0.50 ± 0.19 0.48 ± 0.15
AUC0e1.5 ratio 0.58 ± 0.11 0.54 ± 0.12
AUC0e∞ ratio 0.48 ± 0.12 0.42 ± 0.12

Cmax, maximum observed concentration; AUC0e1.5, area under the concentration-time curve from 0 to 1.5 h; AUC0e∞, AUC from 0 to infinity.

Table 2 1000 mg [9]. The mean prostate/plasma ratios were about 5e20%
Population pharmacokinetic parameters of flomoxef in three-compartment model. higher in the present study than in the study of Hoshi et al., as the
Parameter Estimate 95% confidence interval (bootstrap method) administration methods (infusion or one-shot injection) may have
Fixed-effects parameter
varied the drug distribution. Compared with our previous results
qV1 (L) 6.60 4.77e9.35 concerning other b-lactam antibiotics, the mean prostate/plasma
qV2 (L) 4.88 3.16e5.26 AUC ratios were higher for flomoxef than for doripenem (0.187)
qV3 (kg) 2.32 1.16e2.96 [16], meropenem (0.177) [17] and piperacillin (0.36) [18] but lower
qCL (L/h) 9.64 6.56e12.9
than for pazufloxacin (0.98) [19].
qK12 (1/h) 1.86 0.930e2.33
qK21 (1/h) 2.19 1.10e2.97 In the compartmental pharmacokinetic modeling, the individ-
qK13 (1/h) 2.34 1.13e3.89 ual predicted concentrations in both plasma and prostate tissue
qK31 (1/h) 13.3 9.21e19.9 were in good agreement with the observed drug concentrations
Interindividual variability (exponential error model) (Fig. 2). The simulation curves obtained using the mean fixed-
hV1 0.00512 0.00132e0.00957
hV2 0 None
effects parameters had a good fit with all of the mean measure-
hV3 0.0415 0.0223e0.0809 ments (Fig. 3). These results indicate the validity and reliability of
hCL 0.0521 0.0318e0.1129 our population pharmacokinetic model for flomoxef. To date, the
hK12 0 None information on pharmacokinetic modeling of flomoxef has been
hK21 0.0630 0.0152e0.1082
limited. Flomoxef in plasma was shown to follow a two-
hK13 0 None
hK31 0.0145 0.00246e0.0917 compartment model, and the mean model parameters were V1
Residual variability (proportional error model) (L) ¼ 7.24e9.30 L, K12 (1/h) ¼ 1.18e1.83, K21 (1/h) ¼ 1.48e2.23 and
ε 0.0264 0.0041e0.0391 K10 (1/h) ¼ 1.85e2.01 after intravenous administration (bolus in-
V1, V2 and V3, volumes of distribution of the central, peripheral and prostate com- jection, or 1-h or 2-h infusion) of 1000 mg [1]. Although an exact
partments; CL, clearance from the central compartment; K12, K21, K13 and K31, comparison between these earlier results and the current results
transfer rate constants connecting compartments. was difficult, the corresponding pharmacokinetic parameter values
(Table 2) were V1 (L) ¼ 6.60 (mean) and 4.77e9.35 (95% confidence
interval), K12 (1/h) ¼ 1.86 and 0.930e2.33, K21 (1/h) ¼ 2.19 and

Fig. 2. Scatterplots of the observed concentrations of flomoxef in plasma (269 sam- Fig. 3. Observed concentrations (mean ± standard deviation, n ¼ 54) and simulation
ples) and prostate tissue (152 samples) versus the individual predicted concentrations curves for plasma and prostate tissue after 0.5-h infusions of flomoxef (500 and
based on the pharmacokinetic model parameters (Table 2). The line represents unity 1000 mg). The simulation curves were generated using the mean fixed-effects pa-
(Y ¼ X). rameters (Table 2).

Please cite this article as: Nakamura K et al., Clinical pharmacokinetics of flomoxef in prostate tissue and dosing considerations for prostatitis
based on site-specific pharmacodynamic target attainment, J Infect Chemother, https://doi.org/10.1016/j.jiac.2019.08.019
 K. Nakamura et al. / J Infect Chemother xxx (xxxx) xxx 5

Fig. 4. Probabilities of attaining the bacteriostatic target (40% T > MIC, a) and bactericidal target (70% T > MIC, b) for flomoxef in prostate tissue at specific MICs using twice-daily
(b.i.d.), three-times-daily (t.i.d.) and four-times-daily (q.i.d.) regimens. The dotted lines represent 90% probability.

Table 3
Expected probabilities of target attainment in prostate tissue against bacterial populations (Fig. 1) using different flomoxef regimens.

Flomoxef regimen Expected probability of target attainment (%) in prostate tissue

(0.5-h infusion) Escherichia coli Klebsiella species Proteus species

Bacteriostatic target of 40% T > MIC

500 mg twice daily (1 g/day) 91.9 93.0 83.9
500 mg three times daily (1.5 g/day) 95.2 96.1 91.0
500 mg four times daily (2 g/day) 96.4 97.3 93.5
1000 mg twice daily (2 g/day) 97.1 97.7 95.6
1000 mg three times daily (3 g/day) 97.0 97.7 95.0
1000 mg four times daily (4 g/day) 97.5 97.9 96.7
Bactericidal target of 70% T > MIC
500 mg twice daily (1 g/day) 42.1 42.9 28.9
500 mg three times daily (1.5 g/day) 70.4 71.5 50.1
500 mg four times daily (2 g/day) 85.6 86.7 73.8
1000 mg twice daily (2 g/day) 91.8 92.8 85.2
1000 mg three times daily (3 g/day) 95.4 96.4 90.6
1000 mg four times daily (4 g/day) 96.6 97.4 94.0

1.10e2.97, and K10 (1/h) ¼ CL/V1 ¼ 1.46 and 0.702e2.70. The mean
values of the earlier results were a little different from those of the
present study. However, the earlier values were in the range of the
95% confidence interval for our parameter estimates. Therefore, the
model and its parameter estimates were considered to be adequate
and have a good predictive performance for pharmacodynamic
evaluation use.
Using the developed model parameters, this study then pre-
dicted the flomoxef concentrations in prostate tissue in order to
estimate their site-specific pharmacodynamic target attainment for
different dosing regimens. The results of the stochastic simulation
based on the prostate concentration (Table 3) indicated that 0.5-h
infusion of 1000 mg three times daily (3 g/day) was capable of
bactericidal activity in the prostate and thus sufficient for the
empirical treatment of prostatitis due to common causative path-
ogens such as E. coli, Klebsiella and Proteus species. The target-
attainment probability versus the MIC profiles (Fig. 4) should be
applicable for other potentially causative bacteria as well, regard-
less of the susceptibility pattern, in regions other than Japan. With
antimicrobial susceptibility testing, when the MIC for the patient's
causative pathogen is determined to be 0.5 mg/L (MIC90 of the
Proteus species isolates, Fig. 1), 1000 mg four times daily (4 g/day,
Fig. 4b) should be administered for the definitive treatment of
prostatitis, within the approved dosages of 1e4 g/day.
The Japanese Association for Infectious Disease (JAID)/Japanese
Society of Chemotherapy (JSC) guidelines for the clinical manage-
ment of infectious disease 2015 state the following regarding acute
bacterial prostatitis [6]: As an empiric therapy for moderate pros-
tatitis, antimicrobials are generally administered parenterally (first
choices are oral levofloxacin 500 mg/day, ciprofloxacin 600 mg/day,
tosufloxacin 450 mg/day or sitafloxacin 200 mg/day). For severe
prostatitis, the first choices are the intravenous administration of
cefotiam 1000 mg, two to four times daily (2e4 g/day); ceftazidime
1000 mg, two to four times daily (2e4 g/day); or flomoxef 1000 mg,
two to four times daily (2e4 g/day). Our dosing suggestions (flo-
moxef of 3e4 g/day) are consistent with this first-choice recom-
mendation and support the JAID/JSC guideline from a site-specific
pharmacodynamic perspective.
Flomoxef is also used as a prophylactic medication for prostatic
surgery or biopsy. For this reason, we employed a bacteriostatic
target of 40% T > MIC (for prophylactic use, assuming that the host's
immune status is good) in addition to the bactericidal target of 70%
T > MIC (for therapeutic use, assuming that the host's immune
status is poor). The results of the stochastic simulation based on the
prostate concentration (Table 3) indicate that 0.5-h infusion of
500 mg three times daily (1.5 g/day), with a prostatic
pharmacodynamic-based breakpoint of 0.5 mg/L (Fig. 4a), is
capable of bacteriostatic activity in the prostate against common
causative pathogens and thus appropriate for prophylactic
administration.
We believe that the results of this study help to fully charac-
terize the clinical pharmacokinetics of flomoxef in the prostate,
while also helping to rationalize and optimize the dosing regimens
for prostatitis, especially when the causative bacterial strains are
ESBL-producing, based on site-specific pharmacodynamic target
attainment. However, we should note that the present study sub-
jects were uninfected patients, a condition that may have affected
the penetration of flomoxef since prostatitis, a condition charac-
terized by inflammation of the prostate, can increase prostatic
vascular permeability. The study patients had a normal renal
function with a mean creatinine clearance (calculated using
Cockcroft-Gault formula) of 70.2 mL/min; however, infection often

Please cite this article as: Nakamura K et al., Clinical pharmacokinetics of flomoxef in prostate tissue and dosing considerations for prostatitis
based on site-specific pharmacodynamic target attainment, J Infect Chemother, https://doi.org/10.1016/j.jiac.2019.08.019
6 K. Nakamura et al. / J Infect Chemother xxx (xxxx) xxx
causes renal impairment or kidney failure, leading to the persis-
tence of higher flomoxef concentrations in plasma (T > MIC is
consequently longer in prostate tissue) because flomoxef is pre-
dominantly eliminated renally [1]. For these reasons, the pharma-
cokinetic and pharmacodynamic estimates for the tested regimens
in this study may be conservative, and lower flomoxef regimens
may be effective. Therefore, to confirm the therapeutic significance
of the current results and to optimize flomoxef regimens, clinical
studies should be conducted in infected patients to investigate the
relationship between the prostatic penetration and exposure of
flomoxef as well as its therapeutic efficacy.
5. Conclusions
Flomoxef rapidly penetrated into prostate tissue, with a pros-
tate/plasma ratio of around 0.5. Against E. coli, Klebsiella and Proteus
species isolates, 0.5-h infusion of 1000 mg three times daily (3 g/
day) was shown to be sufficiently bactericidal as an empiric therapy
for prostatitis. As a definitive therapy, flomoxef regimens of 500 mg
four times daily (2 g/day) and 1000 mg four times daily (4 g/day)
were shown to be sufficiently bactericidal against isolates with
susceptibility of 0.25 mg/L (MIC90 of the E. coli and Klebsiella spe-
cies) and 0.5 mg/L (MIC90 of the Proteus species), respectively.
Ethical approval
This study was approved by the Ethics Committee of Aichi
Medical University School of Medicine (Nagakute, Japan) (No. 13-
115).
Authorship statement
All authors have nothing to disclose based on ICMJE Form.
Funding
No funding sources.
Conflicts of interest
The authors declare no conflicts of interest associated with this
manuscript.
References
[1] Shionogi, Company Limited. Flumarin (flomoxef for injection) prescribing
information. 2019. http://www.info.pmda.go.jp/go/interview/1/340018_
6133401F1027_1_008_1F.pdf.
[2] Partina I, Kalinogorskaya O, Kojima S, Gostev V, Volkova M, Ageevets V, et al.
Surveillance of antimicrobial susceptibility of Enterobacteriaceae pathogens
isolated from intensive care units and surgical units in Russia. Jpn J Antibiot
2016;69:41e51.
[3] Santanirand P, Kojima S, Yamaguchi T, Wongnak K, Thokaew S,
Chiaranaicharoen S. Antimicrobial activity of flomoxef against Enterobacteri-
aceae including extended spectrum beta-lactamases-producing strains iso-
lated at Ramathibodi Hospital: a 1000-bed tertiary care hospital in Bangkok,
Thailand. J Infect Dis Ther 2018;6:378. https://doi.org/10.4172/2332-
0877.1000378.
Please cite this article as: Nakamura K et al., Clinical pharmacokinetics of flomoxef in prostate tissue and dosing considerations for prostatitis
based on site-specific pharmacodynamic target attainment, J Infect Chemother, https://doi.org/10.1016/j.jiac.2019.08.019
[4] Ito A, Tatsumi YM, Wajima T, Nakamura R, Tsuji M. Evaluation of antibacterial
activities of flomoxef against ESBL producing Enterobacteriaceae analyzed by
Monte Carlo simulation. Jpn J Antibiot 2013;66:71e86.
[5] Jung Y, Lee SS, Song W, Kim HS, Uh Y. In vitro activity of flomoxef against
extended-spectrum b-lactamase-producing Escherichia coli and Klebsiella
pneumoniae in Korea. Diagn Microbiol Infect Dis 2019;94:88e92.
[6] The Japanese Association for Infectious Disease/Japanese Society of Chemo-
therapy. The JAID/JSC guidelines for clinical management of infectious disease
2015durinary tract infection/male genital infection. J Infect Chemother
2017;23:733e51.
[7] Matsumoto T, Kiyota H, Matsukawa M, Yasuda M, Arakawa S, Monden K.
Japanese guidelines for prevention of perioperative infections in urological
field. Int J Urol 2007;14:890e909.
[8] Lee JK, Lee S, Hong SK, Byun SS, Lee SE. Clinical importance of the antibiotic
regimen in transrectal ultrasound-guided biopsy: quinolone versus cephalo-
sporin. BMC Urol 2016;16:51. https://doi.org/10.1186/s12894-016-0169-z.
[9] Hoshi S, Kuwahara M, Orikasa S, Nakano S, Onuma T, Chiba R, et al. Diffusion
into human prostate and clinical evaluation of 6315-S (flomoxef) in urinary
tract infection. Jpn J Chemother 1987;35(Supplment 1):968e78.
[10] Imaizumi M, Watanabe H, Ojika T, Fujita K, Sakakibara M, Kondo D, et al.
A clinical study on pulmonary tissue uptake of flomoxef. Jpn J Antibiot
1991;44:22e34.
[11] Hayashi I. Serum and sputum concentrations and clinical results of 6315-S
(flomoxef) in respiratory tract infection. Jpn J Chemother 1987;35(Sup-
plment 1):593e8.
[12] Craig WA. Pharmacokinetic/pharmacodynamic parameters: rationale for
antibacterial dosing of mice and men. Clin Infect Dis 1998;26:1e10.
[13] Craig WA. Does the dose matter? Clin Infect Dis 2001;33(Suppl. 3):S233e7.
[14] Konaka R, Hashimoto H, Nishimura R, Kuruma K, Hirauchi K. Assays of 6315-S
(flomoxef) and its metabolites in body fluids by high-performance liquid
chromatography. Jpn J Chemother 1987;35(Supplment 1):137e44.
[15] Yamaoka K, Tanigawara Y, Nakagawa T, Uno T. A pharmacokinetic analysis
program (multi) for microcomputer. J Pharmacobio-Dyn 1981;4:879e85.
[16] Nakamura K, Ikawa K, Yamada Y, Arakawa M, Zennami K, Nishikawa G, et al.
Determination of doripenem penetration into human prostate tissue and
assessment of dosing regimens for prostatitis based on site-specific phar-
macokinetic-pharmacodynamic evaluation. J Chemother 2012;24:32e7.
[17] Nishikawa G, Ikawa K, Nakamura K, Yamada Y, Zennami K, Mitsui K, et al.
Prostatic penetration of meropenem in humans, and dosage considerations
for prostatitis based on a site-specific pharmacokinetic/pharmacodynamic
evaluation. Int J Antimicrob Agents 2013;41:267e71.
[18] Kobayashi I, Ikawa K, Nakamura K, Nishikawa G, Kajikawa K, Yoshizawa T,
et al. Penetration of piperacillin-tazobactam into human prostate tissue and
dosing considerations for prostatitis based on site-specific pharmacokinetics
and pharmacodynamics. J Infect Chemother 2015;21:575e80.
[19] Nakamura K, Ikawa K, Nishikawa G, Kobayashi I, Narushima M, Muramatsu H,
et al. Clinical pharmacokinetics and pharmacodynamic target attainment of
pazufloxacin in prostate tissue: dosing considerations for prostatitis. J Infect
Chemother 2017;23:809e13.
[20] Lindbom L, Ribbing J, Jonsson EN. Perl-speaks-NONMEM (PsN)da Perl module
for NONMEM related programming. Comput Methods Progr Biomed 2004;75:
85e94.
[21] Drusano GL, Preston SL, Van Guilder M, North D, Gombert M, Oefelein M, et al.
A population pharmacokinetic analysis of the penetration of the prostate by
levofloxacin. Antimicrob Agents Chemother 2000;44:2046e51.
[22] Bulitta JB, Kinzig M, Naber CK, Wagenlehner FM, Sauber C, Landersdorfer CB,
et al. Population pharmacokinetics and penetration into prostatic, seminal,
and vaginal fluid for ciprofloxacin, levofloxacin, and their combination.
Chemotherapy 2011;57:402e16.
[23] Kimura Y, Nakashimizu H, Nakano M, Otsubo R, Matsubara H, Yoshida T.
Pharmacokinetic characterization of 6315-S (flomoxef) in experimental ani-
mals. Jpn J Chemother 1987;35(Supplement 1):161e75.
[24] Shinagawa N, Taniguchi M, Hirata K, Furuhata T, Mizuguchi T, Osanai H, et al.
Bacteria isolated from surgical infections and its susceptibilities to antimi-
crobial agentsdspecial references to bacteria isolated between April 2011 and
March 2012. Jpn J Antibiot 2014;67:339e83.
[25] Yamaguchi K, Ishii Y, Tateda K, Iwata M, Watanabe N, Shinagawa M, et al.
Nationwide surveillance of parenteral antibiotics containing meropenem ac-
tivities against clinically isolated strains in 2012. Jpn J Antibiot 2014;67:
73e107.