JAC

Journal of Antimicrobial Chemotherapy (1999) 43, 367371

Efficacy of -lactam antibiotics combined with gentamicin against

penicillin-resistant pneumococcal pneumonia in CBA/J mice
Kazuhiro Tateda*, Tetsuya Matsumoto, Shuichi Miyazaki and Keizo Yamaguchi
Department of Microbiology, Toho University School of Medicine, 5-21-16 Ohmori-nishi, Ohta-ku,
Tokyo 143, Japan
We examined the efficacy of gentamicin combined with -lactam antibiotics against penicillinresistant Streptococcus pneumoniae (PRSP) in a noncompromised mouse model of pneumonia. In the presence of 8 mg/L (0.25 3 MIC) and 16 mg/L (0.5 3 MIC) of gentamicin, MICs of
penicillin G against 23 strains of PRSP decreased from 14 mg/L to 0.03 mg/L in 14 (61%) and
23 strains (100%), respectively. A short-time killing study using strain 741 showed that 8 mg/L
of gentamicin (0.25 3 MIC) increased the killing activity of penicillin G, cefotaxime and
imipenem (at 0.25, 1 and 4 3 MIC) during a 6 h incubation period. Survival studies showed that
the combined treatment of penicillin-G (160 mg/kg) and gentamicin (10 mg/kg), which commenced 2 days after infection (twice a day for 5 days), provided complete protection, while no
animal survived when either antibiotic was used alone. A significant improvement in mortality
was observed when a small dose of imipenem (2.5 and 10 mg/kg) was used with gentamicin.
Our results suggest that gentamicin, when combined with -lactam antibiotics, especially
imipenem, may be potentially useful against PRSP pneumonia in noncompromised individuals.

Introduction
Streptococcus pneumoniae is still the leading cause of
bacterial pneumonia, which is associated with significant
morbidity and mortality. 1,2 Epidemiological surveys in the
past several years have demonstrated a dramatic increase
in the prevalence of penicillin-resistant S. pneumoniae
(PRSP) throughout the world.35 Moreover, the appearance of multiresistant strains, which also exhibit resistance
to extended-spectrum cephalosporins, such as cefotaxime
and ceftriaxone, limits the therapeutic choice in clinical
practice.68 Experience with clinical treatment failures in
infections caused by PRSP have prompted investigators to
re-examine the in-vitro and in-vivo therapeutic efficacies of
antibiotics.911 Combination therapy is an alternative form
of therapy available for the treatment of PRSP infections.
Several investigators have reported synergy between lactams and aminoglycosides against PRSP in killing curves
and chequerboard studies.1113 In addition to these in-vitro
studies, Darras-Joly et al.14 have clearly demonstrated the
usefulness of amoxycillin combined with gentamicin in a
leukopenic mouse model of pneumonia. Although this
combination has generally been accepted as a potential
alternative therapy, some points regarding the effective-

ness of this strategy still need confirming. These include the

exact type of -lactam that forms the best partner with gentamicin in vivo and whether the combined use of gentamicin is also effective in immunocompetent individuals, since
the majority of PRSP pneumonia occurs in otherwise
healthy individuals as community-acquired pneumonia.15,16
We have recently described a noncompromised CBA/J
mouse model of PRSP pneumonia.17,18 This model resembles human PRSP community-acquired pneumonia, particularly with regard to progression of the disease and
pathological findings. In the present study, we investigated
the efficacy of gentamicin combined with several -lactam
antibiotics, such as penicillin G, cefotaxime and imipenem,
using this newly developed PRSP pneumonia model.

Materials and methods

S. pneumoniae strains
Twenty-three strains of S. pneumoniae isolated at Toho
University School of Medicine were used in the present
study, including strain 741, reported previously.18 These
strains were isolated from different patients between 1993

*Corresponding author. Tel: 181-3-3762-4151; Fax: 181-3-5493-5415; E-mail: kazu@med.toho-u.ac.jp

367
1999 The British Society for Antimicrobial Chemotherapy

K. Tateda et al.
to 1995, with no two representing the same outbreak. All
strains were frozen at 280C in skimmed milk until used.

Animals
Five-week-old male CBA/J mice (body weight range,
1620 g) were purchased from Charles River Japan, Kanagawa, Japan. The experimental protocol was approved by
the Ethics Review Committee for Animal Experimentation of Toho University School of Medicine.

MuellerHinton broth supplemented with 5% lysed horse

blood and incubated at 35C until the culture appeared
turbid to the naked eye. This exponential-growth culture
was suspended in 0.9% saline to the desired concentration
(confirmed by plating serial ten-fold dilutions on to 5%
blood agar). Mice were anaesthetized lightly by intramuscular injection of a mixture of 60 mg/kg ketamine
(Sankyo Pharmaceutical, Tokyo, Japan) and 10 mg/kg
xylazine (Bayer Japan, Tokyo, Japan), after which each
mouse was challenged intranasally with approximately 10 5
logarithmic-phase organisms.

Antimicrobial agents
Penicillin G, cefotaxime, imipenem and gentamicin were
obtained from Meiji Seika (Tokyo, Japan), Hoechst Japan
(Tokyo, Japan), Banyu Pharmaceutical Co. (Tokyo,
Japan) and Schering-Plough Co. (Osaka, Japan), respectively. Imipenem was mixed with cilastatin (Banyu) at a
ratio of 1:1, and was used in an in-vivo pneumonia model.

Survival studies
To investigate the effects of gentamicin combined with
penicillin G or imipenem on survival, the indicated doses of
antibiotics were administered subcutaneously to a group of
ten animals twice a day for 5 days, commencing 2 days after
infection. Survival rates were recorded daily for 1214 days
after infection.

Antimicrobial susceptibility test

The MICs of penicillin G in the presence or absence of
gentamicin were determined by a broth dilution method
using MuellerHinton broth (Difco, Detroit, MI, USA)
supplemented with 5% lysed horse blood. Microtitre
plates, containing 5.0 3 104 cfu/well were incubated at 35C
for 18 h. The MIC was defined as the lowest concentration
of antimicrobial agent that prevented visible growth of
S. pneumoniae. In preliminary experiments, MICs of
penicillin G and gentamicin for 23 strains were examined.
These results showed that MIC of penicillin G ranged
between 1 and 4 mg/L while the MIC of gentamicin for all
strains was 32 mg/L. MICs of penicillin G, cefotaxime and
imipenem for strain 741 were 1, 0.5 and 0.25 mg/L, respectively.

In-vitro short-time killing study

Statistical analysis
The x2 test was used to compare the survival rates, and P
values of <0.05 were considered statistically significant.

Results
Effects of gentamicin combination on MIC of
penicillin G
In the absence of gentamicin, the MICs of penicillin G for
the 23 strains were 1, 2 and 4 mg/L for six, 16 and one
strains, respectively. Simultaneous presence of subinhibitory concentrations of gentamicin dramatically
reduced the MICs of penicillin G (Figure 1). When 8 mg/L
of gentamicin (0.25 3 MIC) was used, 14 strains (60.9%)

The in-vitro bactericidal effect of -lactams (penicillin,

cefotaxime and imipenem) was examined in the presence
of gentamicin. Strain 741 was inoculated at a final concentration of 107 cfu/mL into MuellerHinton broth supplemented with 5% lysed horse blood, in which a -lactam
antibiotic with or without gentamicin was added. The suspensions were incubated at 35C, and then a count of the
number of bacteria was performed after 1.5, 3 and 6 h by
plating serial ten-fold dilutions of the samples on to 5%
blood agar. Data were expressed as the average of two
experiments.

Pneumonia model
The noncompromised mouse model of penicillin-resistant
pneumococcal pneumonia was used as reported previously.18 For this purpose, strain 741 was inoculated into

Figure 1. Effects of gentamicin on MICs of penicillin G against

PRSP. MICs of penicillin G for 23 strains of PRSP were
examined in the presence or absence of various concentrations of
gentamicin.

368

Gentamicin-lactam combination against PRSP pneumonia

became highly sensitive to penicillin G. Moreover, the
addition of 16 mg/L of gentamicin (0.5 3 MIC) decreased
MICs of penicillin G for all strains tested to <0.03 mg/L.
These results indicated that a combination of subinhibitory concentrations of gentamicin generally sensitize
clinical isolates of PRSP to penicillin G in a gentamicin concentration-dependent manner.

Effects of gentamicin combination on short-time

killing of -lactams
Short-time killing (1.5, 3 and 6 h) by -lactam antibiotics
was examined in strain 741 in the presence or absence of
gentamicin (Figure 2). Without gentamicin, the best killing
activity was noted with imipenem at all tested concentrations, followed by cefotaxime and penicillin G. In the presence of 8 mg/L of gentamicin (0.25 3 MIC), these activities
were generally enhanced, although the reduction of bacterial numbers by gentamicin alone was ,1 log during the

6 h observation period. For example, in the presence of

0.25 3 MIC of imipenem at 3 h, a gentamicin combination
reduced the bacterial count from 6.7 to 3.3 log cfu/mL.

Effects of gentamicin combination on survival

(Figures 3 and 4)
Untreated control mice died between 6 and 10 days after
infection. Penicillin G (160 mg/kg) alone or gentamicin (10
mg/kg) alone did not improve the survival rate at 10 days
after infection although gentamicin-treated mice demonstrated slightly higher survival rates between 7 and 9 days. In
contrast, a dramatic improvement of survival was observed
when mice were treated simultaneously with penicillin G
(160 mg/kg) and gentamicin (100% survival at the end of
observation) A significant decrease in mortality was
observed using a considerably smaller doses of imipenem
(2.5 and 10 mg/kg), relative to penicillin G, when combined
with gentamicin.

Figure 2. Effects of gentamicin combination on short-time killing of penicillin G, cefotaxime and imipenem against PRSP strain 741.
Short-time killing activities of penicillin G (a), cefotaxime (b) and imipenem (c) in 0.25 3 MIC (circles), MIC (triangles) and 4 3 MIC
(squares) were examined in the absence (open symbols) or presence (solid symbols) of 8 mg/L of gentamicin. Horizontal dotted line
represents the detection limit of bacterial counts.

Figure 3. Effects of gentamicin combined with penicillin G on survival of mice infected with strain 741. Penicillin G at 40 mg/kg (m),
160 mg/kg (M) or saline (V) was subcutaneously administered without (a) or with (b) 10 mg/kg of gentamicin from 2 days after
infection twice a day for 5 days. *P , 0.05, compared with the corresponding control group.

369

K. Tateda et al.

Figure 4. Effects of gentamicin combined with imipenem on survival rates of mice infected with strain 741. Imipenem at 2.5 mg/kg
(m), 10 mg/kg (M) or saline (V) was subcutaneously administered without (a) or with (b) 10 mg/kg of gentamicin from 2 days after
infection twice a day for 5 days, * P , 0.05, compared with the corresponding control group.

Discussion
The major finding of the present study was the high efficacy
of gentamicin combined with -lactam antibiotics against
PRSP pneumonia in noncompromised mice. Our data were
consistent with previous results in neutropenic mouse
model,14 and suggest the clinical usefulness of such combination therapy in community-acquired PRSP pneumonia.
The use of gentamicin concentrations of 1/16 of the MIC
(2 mg/L) had little or no effect on MICs of penicillin G
while the use of a higher dose of gentamicin (0.25 3 MIC,
8 mg/L) dramatically sensitized PRSP to penicillin G in
a gentamicin concentration-dependent manner. In this
regard, Schlegel et al.,13 have reported synergy between
-lactams (cefotaxime and imipenem) and gentamicin
when 0.25 3 MIC (8 mg/L) of the latter and 0.5 3 MICs
of -lactams were used. In contrast, using a lower concentration of gentamicin (1 mg/L), Gross et al.19 were unable
to demonstrate a significant enhancement of the killing
activity of penicillin and cefotaxime. Our data, together
with those of previous reports suggest that 8 mg/L (0.25 3
MICs) of gentamicin may be the critical concentration
necessary to express its synergic activity in vitro.
In the pneumonia model of leukopenic mice, a combination of gentamicin 8 mg/kg with amoxycillin 100 mg/kg
yielded almost complete protection and survival of all
animals, which occurred at a maximum gentamicin serum
concentration of 12 mg/L.14 The results of our pulmonary
clearance and survival studies indicated the effectiveness of
10 mg/kg dose of gentamicin, although serum gentamicin
concentrations were not determined. Pharmacokinetic
studies of gentamicin in humans, conducted by Chung
and co-workers, 20 reported peak serum concentrations of
gentamicin of 5.76 and 11.0 mg/L following intramuscular
administration of 1.0 and 2.5 mg/kg, respectively.
Of the combinations examined in the present study, the

combination of imipenem and gentamicin exhibited potent

bactericidal activity in the lungs, which was associated
with higher survival rates than other combinations. In our
previous study using the CBA/J mice model, we reported
that in a single antibiotic regimen, imipenem was the
most active among antibiotics tested in clearing penicillinresistant pneumococci from the lungs.18 Our results are
also consistent with other previous reports showing excellent bactericidal activity of imipenem plus gentamicin
against PRSP in an in-vitro killing study12 and in a rabbit
model of penicillin-resistant S. sanguis endocarditis.21
These data suggest that imipenem may be one of the most
active antimicrobials when used in combination with gentamicin for the treatment of PRSP infection.
Although there are pharmacokinetic differences between
species, our mouse survival studies employed the usual
twice-a-day human dosage regimen. Whether this is the
most appropriate regimen, or whether different results
might be obtained with approaches such as frequent dosing
or constant infusion that address specific features of
mouse pharmacokinetics, certainly requires further study.
Although caution must be exercised in applying the results
of the present animal studies to the management of
patients, we believe that the combination of -lactams and
gentamicin should be considered in those patients in whom
infections caused by strains showing high-level penicillin
resistance is suspected, or in systemic infections in critically-ill patients.

Acknowledgements
We thank Professor Shogo Kuwahara for his critical reading of the manuscript and helpful suggestions. We also
thank Dr F. G. Issa for his expert editorial assistance.

370

Gentamicin-lactam combination against PRSP pneumonia

References
1. Macfarlane, J. T., Finch, R. G., Ward, M. J. & Macrae, A. D.
(1982). Hospital study of adult community-acquired pneumonia.
Lancet ii, 2558.
2. Austrian, R. (1986). Pneumococcal pneumonia. Diagnostic, epidemiologic, therapeutic and prophylactic considerations. Chest 90,
73843.
3. Spika, J. S., Facklam, R. R., Plikaytis, B. D., Oxtoby, M. J. & the
Pneumococcal Surveillance Working Group. (1991). Antimicrobial
resistance of Streptococcus pneumoniae in the United States,
19791987. Journal of Infectious Diseases 163, 127378.
4. Appelbaum, P. C. (1992). Antimicrobial resistance in Streptococcus pneumoniae: an overview. Clinical Infectious Diseases 15,
7783.
5. Caputo, G. M., Appelbaum, P. C. & Liu, H. H. (1993). Infections
due to penicillin-resistant pneumococci. Clinical, epidemiologic, and
microbiologic features. Archives of Internal Medicine 153, 130110.
6. Bradley, J. S. & Connor, J. D. (1991). Ceftriaxone failure in
meningitis caused by Streptococcus pneumoniae with reduced susceptibility to beta-lactam antibiotics. Pediatric Infectious Disease
Journal 10, 8713.
7. Canton, E. (1992). Cefotaxime breakpoint for Streptococcus
pneumoniae. Antimicrobial Agents and Chemotherapy 37, 6167.
8. John, C. C. (1994). Treatment failure with use of third-generation
cephalosporin for penicillin-resistant pneumococcal meningitis: case
report and review. Clinical Infectious Diseases 18, 18893.
9. McCracken, G. H. & Sakata, Y. (1985). Antimicrobial therapy of
experimental meningitis caused by Streptococcus pneumoniae
strains with different susceptibilities to penicillin. Antimicrobial
Agents and Chemotherapy 27, 1415.
10. Barakett, V., Lesage, D., Delisle, F., Richard, G. & Petit, J. C.
(1994). Killing kinetics of imipenem against penicillin-resistant
Streptococcus pneumoniae. Journal of Antimicrobial Chemotherapy
33, 10258.
11. Friedland, I. R., Paris, M., Shelton, S. & McCracken, G. H.
(1994). Timekill studies of antibiotic combinations against penicillin-resistant and -susceptible Streptococcus pneumoniae. Journal
of Antimicrobial Chemotherapy 34, 2317.
12. Doit, C. P., Bonacorsi, S. P., Fremaux, A. J., Sissia, G., Cohen,
R., Geslin, P. L. et al. (1994). In vitro killing activities of antibiotics at
clinically achievable concentrations in cerebrospinal fluid against

penicillin-resistant Streptococcus pneumoniae isolated form

children with meningitis. Antimicrobial Agents and Chemotherapy
38, 26559.
13. Schlegel, L., Sissia, G., Fremaux, A. & Geslin, P. (1997). In-vitro
killing activity of combinations of -lactam agents with aminoglyco sides against penicillin-resistant pneumococci. Journal of Antimicrobial Chemotherapy 39, 958.
14. Darras-Joly, C., Bedos, J. P., Sauve, C., Moine, P., Vallee, E.,
Carbon, C. et al . (1996). Synergy between amoxicillin and gentamicin in combination against a highly penicillin-resistant and -tolerant
strain of Streptococcus pneumoniae in a mouse pneumonia model.
Antimicrobial Agents and Chemotherapy 40, 214751.
15. Feldman, C., Kallenbach, J. M., Miller, S. D., Thorburn, J. R. &
Koornhof, H. J. (1985). Community-acquired pneumonia due to
penicillin-resistant pneumococci. New England Journal of Medicine
313, 6157.
16. Woodhead, M. A., Macfarlane, J. T., McCracken, J. S., Rose, D.
H. & Finch, R. G. (1987). Prospective study of the aetiology and outcome of pneumonia in the community. Lancet i, 6714.
17. Takashima, K., Tateda, K., Matsumoto, T., Iizawa, Y., Nakao,
M. & Yamaguchi, K. (1996). Establishment of a model of penicillinresistant Streptococcus pneumoniae pneumonia in healthy CBA/J
mice. Journal of Medical Microbiology 45, 31922.
18. Tateda, K., Takashima, K., Miyazaki, H., Matsumoto, T., Hatori,
T. & Yamaguchi, K. (1996). Noncompromised penicillin-resistant
pneumococcal pneumonia CBA/J mouse model and comparative
efficacies of antibiotics in this model. Antimicrobial Agents and
Chemotherapy 40, 15205.
19. Gross, M. E., Giron, K. P., Septimus, J. D., Mason, E. O. &
Musher, D. M. (1995). Antimicrobial activities of beta-lactam antibiotics and gentamicin against penicillin-susceptible and penicillinresistant pneumococci. Antimicrobial Agents and Chemotherapy 39,
11668.
20. Chung, M., Costello, R. & Symchowicz, S. (1980). Comparison
of netilmicin and gentamicin pharmacokinetics in humans. Antimicrobial Agents and Chemotherapy 17, 1847.
21. Martinez, F., Martin-Luengo, F., Garcia, A. & Valdes, M. (1994).
Treatment with imipenem of experimental endocarditis caused by
penicillin-resistant Streptococcus sanguis. Journal of Antimicrobial
Chemotherapy 33, 12017.
Received 11 May 1998; returned 20 June 1998; revised 28 July
1998; accepted 17 October 1998

371