Journal of Antimicrobial Chemotherapy (2006) 57, 573576

doi:10.1093/jac/dki477
Advance Access publication 23 January 2006

In vitro antibacterial activities of tigecycline in combination

with other antimicrobial agents determined by chequerboard and
time-kill kinetic analysis

Peter J. Petersen, Ponpen Labthavikul, C. Hal Jones* and Patricia A. Bradford

Infectious Disease Discovery Research, Wyeth Research, Pearl River, NY 10965, USA

Received 7 October 2005; returned 16 November 2005; revised 30 November 2005; accepted 12 December 2005

Objectives: This study was undertaken to determine the interaction of tigecycline with 13 select
antimicrobial agents against a wide variety of Gram-negative and Gram-positive bacterial isolates.

Downloaded from http://jac.oxfordjournals.org/ by guest on June 2, 2013

Methods: Antibiotic interactions were assayed using the chequerboard MIC format and selected syner-
gistic combinations were confirmed using time-kill kinetic analysis.
Results: Microdilution chequerboard analysis of tigecycline in combination with amikacin, ampicillin/
sulbactam, azithromycin, ciprofloxacin, colistin, imipenem, levofloxacin, piperacillin, piperacillin/
tazobactam, polymyxin B, rifampicin, minocycline and vancomycin resulted in an interpretation of either
no interaction or synergy. Time-kill kinetic analysis resulted in an interpretation of no interaction for all but
one of the drug combinations that resulted in an interpretation of synergy by the chequerboard analysis.
Antagonism was not observed for any combination when assayed by either method.
Conclusions: The lack of antagonism seen with tigecycline combinations in both chequerboard and time-
kill kinetic studies is an encouraging outcome, suggesting that tigecycline may prove to be effective in
combination therapy as well as in monotherapy.

Keywords: antibiotics, synergy, antagonism, susceptibility

Introduction by both preclinical and clinical studies, it is important to

characterize tigecycline in combination with other antibiotics
Tigecycline, the 9-glycylamido derivative of minocycline, is the in order to identify synergistic and/or antagonistic combinations
first member of the glycylcycline class of antibiotics to enter providing guidance for empirical use as well as for treatment of
the clinic. Tigecycline acts by preventing translation through a poly-microbial infections where combination therapy is warranted.3
reversible binding interaction that blocks the association of Due to the emergence of multidrug-resistant pathogens,
charged tRNA with the ribosome. Tigecycline has a distinct treatment with combination therapy, using two or more anti-
advantage over tetracycline and minocycline in that it is not subject bacterials, has become commonplace.3 Two of the most widely
to either the efflux or ribosomal protection mechanisms of used in vitro methodologies to assess drugdrug interactions are
tetracycline resistance.1,2 Preclinical studies have demonstrated the chequerboard MIC technique, yielding the fractional inhi-
the potent in vitro activity of tigecycline against a broad spectrum bitory concentration index (FICI), and time-kill kinetics.3,4 The
of Gram-positive, Gram-negative, anaerobic and atypical patho- chequerboard MIC method is prone to error5 and, by necessity,
gens, including those organisms expressing tetracycline resistance results from the chequerboard MIC are often confirmed with the
determinants.1,2 Moreover, tigecycline is active against methicillin- more dynamic interaction provided by the time-kill kinetic study
resistant Staphylococcus aureus (MRSA), vancomycin-resistant format.68
Enterococcus spp. (VRE), penicillin-resistant Streptococcus This study was undertaken to determine the interaction of
pneumoniae (PRSP) and extended spectrum b-lactamase (ESBL) tigecycline with other antimicrobial agents against a variety of
producing Klebsiella pneumoniae and Escherichia coli.1 Despite bacterial isolates collected during clinical trials in the United
the potent broad spectrum of activity of tigecycline, supported States and Canada between 1990 and 2000.
.............................................................................................................................................................................................................................................................................................................................................................................................................................

*Corresponding author. Tel: +1-845-602-4612; Fax: +1-845-602-5671; E-mail: jonesh3@wyeth.com

.............................................................................................................................................................................................................................................................................................................................................................................................................................

573

The Author 2006. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved.
For Permissions, please e-mail: journals.permissions@oxfordjournals.org
 Petersen et al.

Materials and methods
Bacterial strains
Representative isolates of clinically relevant species, collected during
clinical trials, from various medical centres in the United States
and Canada between 1990 and 2000, were used in this study. The
Gram-negative organisms used were chosen from the collection at
random and do not represent any specific resistance mechanism.
The Gram-positive organisms were also chosen from the clinical
collection; however, these were chosen to represent important
resistance mechanisms (MRSA, PRSP and VRE).
Antimicrobial agents
This study was designed to evaluate tigecycline in combination with
a wide variety of antimicrobial agents in support of the use of
tigecycline in combination therapy for a compassionate use clinical
protocol. The antimicrobial agents used in the study were: tigecycline,
piperacillin, tazobactam (Wyeth Research, Pearl River, NY, USA),
ampicillin, minocycline, amikacin, ciprofloxacin, vancomycin,
rifampicin, polymyxin B, colistin (Sigma-Aldrich Co., St Louis,
MO, USA), azithromycin, sulbactam, imipenem (USP, Rockville,
any combination with tigecycline, against any of the strains tested
(Table 1). A higher percentage of synergistic combinations with
tigecycline were observed with amikacin (56%), ampicillin/sulbac-
tam (33%), piperacillin/tazobactam (50%) and rifampicin (33%).
Interestingly, 73% of the Proteus spp. showed synergy when tige-
cycline was tested in combination with minocycline. No other clear
trend could be established for synergy occurring with any other
bacterial species and drug combinations.
With the Gram-positive isolates, rifampicin displayed a
synergistic effect with tigecycline for 66% of the isolates tested
(Table 2). The majority of these strains showing synergy were
Enterococcus spp. including vancomycin-resistant (VRE) strains
and penicillin-resistant Streptococcus pneumoniae (PRSP). Con-
versely, the combination of vancomycin and tigecycline resulted
in a larger percentage of no interaction (71%) than synergistic
effects (29%) against the Gram-positive isolates. Antagonism
was not observed in this analysis.
In order to confirm a result of synergy (FICI 0.5) by the
chequerboard MIC method, time-kill kinetic studies were per-
formed with tigecycline combinations against selected bacterial
species.9 The antibiotics were tested at concentrations based

Downloaded from http://jac.oxfordjournals.org/ by guest on June 2, 2013

MD, USA) and levofloxacin (R. W. Johnson, Princeton, NJ, USA).
Chequerboard MIC
Antibiotic interactions were determined using the chequerboard MIC
assay as previously described.5 MuellerHinton II broth (MHB) was
used for the Enterobacteriaceae, staphylococci and enterococci and
was supplemented with 5% lysed horse blood for streptococci.
Seven doubling dilutions of tigecycline and 11 doubling dilutions
of the test antimicrobial agent were tested. After drug dilution, micro-
broth dilution plates were inoculated with each organism to yield
the appropriate density (105 cfu/mL) in a 100 mL final volume and
incubated for 1822 h at 35 C in ambient air.
The FICI was calculated for each combination using the following
formula: FICA + FICB = FICI, where FICA = MIC of drug A in
combination/MIC of drug A alone, and FICB = MIC of drug B in
combination/MIC of drug B alone. The FICI was interpreted as
follows: synergy = FICI 0.5; no interaction = FICI >0.54;
antagonism = FICI > 4.
Time-kill assays
Flasks containing MHB and drug were inoculated with test organism
to a density of 106 cfu/mL in a final volume of 100 mL and incuba-
ted in a shaking water bath at 35 C in ambient air. Aliquots were
removed at time 0 and 3, 6 and 24 h post-inoculation and serially
diluted in 0.85% sodium chloride solution for determination of
viable counts. Diluted samples, 0.05 mL, were plated in duplicate
on trypticase soy agar plates using a spiral plater (Don Whitley
Scientific Ltd). Total bacterial cfu/mL (log10cfu/mL) were determined
after 18 h of incubation at 35 C.
Results and discussion
The in vitro interactive effects of the antibiotics were determined
by the broth microdilution chequerboard method as previously
described.5 The range of drug concentrations used in the chequer-
board analysis was such that the dilution range encompassed
the MIC of each drug used in the analysis.
The combination of tigecycline and another antibiotic demon-
strated either synergy (24%) or no interaction (76%) against the
panel of Gram-negative bacteria; antagonism was not observed for
on the MIC determined from microbroth chequerboard testing:
alone at 1 and 0.5 the MIC and in combination at 0.5 the
MIC. The concentrations of the antibiotics used in the time-kill
assays were, at a minimum, one dilution higher than the synergistic
combination shown by chequerboard MIC analysis.
As determined by Eliopoulos and Moellering10 an interpretation
of synergy required a 2 log10 decrease in cfu/mL by the drug
combination when compared with its most active constituent
after 24 h and a 2 log10 decrease in the cfu/mL below the starting
inoculum. Likewise, the drug combination was considered to be
antagonistic if there was a 2 log10 increase in cfu/mL and no
interaction was the interpretation of a <2 log10 change in cfu/mL.
The results of time-kill kinetic studies confirmed the chequerboard
data in that none of the tigecycline combinations resulted in antago-
nism. However, synergy results by FICI were confirmed by time-kill
kinetics for only one of the seventeen combinations examined, which
was tigecycline, tested at 2 mg/L, combined with amikacin, tested at 8
mg/L against one strain of Acinetobacter baumannii (PT 9158).
Indicative of the majority of strains tested was the finding that
tigecycline, tested at 0.5 mg/L, in combination with azithromycin,
tested at 16 mg/L, was synergistic by FICI against a strain of K.
pneumoniae (PT 9266); however, when assayed by time-kill kinet-
ics this combination failed to meet the criteria for synergy. In
approximately half of the time-kill kinetics studies, the combina-
tions demonstrated better killing at the 6 h time point than either drug
alone; however, there were no changes in interpretations regarding
synergy when measured at the earlier time point (data not shown).
Antimicrobial combinations are used frequently in the clinic to
provide broad-spectrum coverage until the causative pathogens
are isolated and identified.3 In the clinical setting, combination
therapy is most often given empirically without the use of
in vitro synergy data, as there is a lack of clinical data to correlate
the results of in vitro synergy studies with patient outcome.3
Clearly, from a clinical viewpoint, antagonism is the least
desirable outcome possible with an antimicrobial combination.
Recent in vitro studies have demonstrated various antibiotic com-
binations that resulted in no interaction and or synergy.6,7 Although
there is no consensus in the field as to the best methodology
for measuring synergy, the most commonly used assay is the
chequerboard MIC test; however, this is most often used only as

574
 Tigecycline synergy

Table 1. Results of chequerboard testing of tigecycline and a second Table 1. (continued)

antibacterial agent against Gram-negative bacteria
No. of strains
No. of strains showing synergy/total
showing synergy/total Organism Second agent no. of strainsa
Organism Second agent no. of strainsa
Proteus vulgaris imipenem 0/4
Acinetobacter amikacin 4/9 levofloxacin 1/4
baumannii ampicillin/sulbactam 1/9 minocycline 3/4
azithromycin 0/9 piperacillin 3/4
ciprofloxacin 0/9 Providencia imipenem 1/3
colistin 1/9 rettgeri levofloxacin 0/3
piperacillin/tazobactam 2/9 minocycline 2/3
polymyxin B 2/9 piperacillin 1/3
rifampicin 1/9
Pseudomonas amikacin 0/3
Acinetobacter spp.b imipenem 3/11 aeruginosa ampicillin/sulbactam 2/3
levofloxacin 0/11 azithromycin 0/3
piperacillin 3/11 ciprofloxacin 0/3
Enterobacter amikacin 1/1 colistin 0/3
aerogenes ampicillin/sulbactam 0/1 imipenem 1/11

Downloaded from http://jac.oxfordjournals.org/ by guest on June 2, 2013

azithromycin 0/1 levofloxacin 0/11
ciprofloxacin 0/1 piperacillin 3/11
colistin 1/1 piperacillin/tazobactam 2/3
imipenem 1/5 polymyxin B 0/3
levofloxacin 4/5 rifampicin 0/3
piperacillin 3/5 Stenotrophomonas imipenem 0/10
piperacillin/tazobactam 1/1 maltophilia levofloxacin 0/10
polymyxin B 0/1 piperacillin 0/10
rifampicin 1/1
a
Enterobacter amikacin 2/2 None of the strains tested showed antagonism.
b
Acinetobacter spp. includes: A. baumannii (13), A. anitratus (4), A. lwoffi (3).
cloacae ampicillin/sulbactam 2/2
azithromycin 1/2
ciprofloxacin 0/2
colistin 0/2
imipenem 1/5
levofloxacin 1/5 Table 2. Results of chequerboard testing of tigecycline and a second
piperacillin 3/5 antibacterial agent against Gram-positive bacteria
piperacillin/tazobactam 1/2
polymyxin B 0/2 No. of strains
rifampicin 2/2 showing
Escherichia coli imipenem 1/11 synergy/total
levofloxacin 0/11 Organism Second agent no. of strainsa
piperacillin 2/11
Staphylococcus vancomycin 1/10
Klebsiella amikacin 3/3
aureus (MRSA) rifampicin 2/10
pneumoniae ampicillin/sulbactam 2/3
azithromycin 0/3 Enterococcus vancomycin 3/5
ciprofloxacin 0/3 faecium (VRE) rifampicin 4/5
colistin 0/3 Enterococcus vancomycin 1/5
imipenem 3/10 faecium rifampicin 4/5
levofloxacin 3/10
piperacillin 1/10 Enterococcus vancomycin 0/3
piperacillin/tazobactam 3/3 faecalis (VRE) rifampicin 2/3
polymyxin B 0/3 Enterococcus vancomycin 0/8
rifampicin 2/3 faecalis rifampicin 5/8
Proteus mirabilis imipenem 2/4 Streptococcus vancomycin 7/10
levofloxacin 0/4 pneumoniae (PRSP) rifampicin 10/10
minocycline 3/4
a
piperacillin 2/4 None of the strains tested showed antagonism.

575
 Petersen et al.
a screening test.38 The chequerboard MIC test suffers due to
lack of reproducibility and only measures bacteriostatic effects.5
Variability in the test as well as testing a bacteriostatic agent in
combination with mostly bactericidal agents may be the cause for
the overestimate of synergy experienced with the chequerboard
test. Accordingly, synergy testing performed by time-kill
kinetics was used to confirm the results of chequerboard MIC
testing, as is standard protocol in many laboratories.7,8
The drug concentrations used in the time-kill kinetic studies,
based on the MIC determined in the FICI analysis, were expected
to have an effect in the growth assay. The fact that these concen-
trations do not result in synergy in the more constrained experi-
mental format suggests again that the chequerboard analysis is
overestimating synergy, possibly due to the reproducibility issue
inherent in the test.5
Although synergy detected by in vitro chequerboard studies
could not in the majority of cases be confirmed by time-kill kinetic
analysis, the lack of antagonism seen with tigecycline com-
binations in both studies is an encouraging outcome suggesting
that tigecycline may prove to be effective in combination therapy
as well as in monotherapy.
Transparency declarations
None to declare.
References
1. Bradford PA, Weaver-Sands DT Petersen PJ. In vitro activity of
tigecycline against isolates from patients enrolled in phase 3 clinical trials
for complicated skin and skin structure infections and complicated
intra-abdominal infections. Clin Infect Dis 2005; 41 Suppl 5: S31532.
576
The author has requested enhancement of the downloaded file. All in-text references underlined in blue are linked to publications on ResearchGate.
2. Petersen PJ, Bradford PA, Weiss WJ et al. In vitro and in vivo
activities of tigecycline (GAR-936), daptomycin, and comparative
antimicrobial agents against glycopeptide-intermediate Staphylococcus
aureus and other resistant Gram-positive pathogens. Antimicrob Agents
Chemother 2002; 46: 2595601.
3. Rybak MJ McGrath BJ. Combination antimicrobial therapy
for bacterial infections. Guidelines for the clinician. Drugs 1996; 52:
390405.
4. White R, Burgess D, Manduru M et al. Comparison of three different
in vitro methods of detecting synergy: time-kill, checkerboard, and E test.
Antimicrob Agents Chemother 1996; 40: 191418.
5. Rand KH, Houck HJ, Brown P et al. Reproducibility of the micro-
dilution checkerboard method for antibiotic synergy. Antimicrob Agents
Chemother 1993; 37: 613615.
6. Alou L, Cafini F, Sevillano D et al. In vitro activity of mupirocin and
amoxicillin-clavulanate alone and in combination against staphylococci
including those resistant to methicillin. Int J Antimicrob Agents 2004;
23: 5136.
7. Jacqueline C, Navas D, Batard E et al. In vitro and in vivo
synergistic activities of linezolid combined with subinhibitory concentra-
tions of imipenem against methicillin-resistant Staphylococcus aureus.
Antimicrob Agents Chemother 2005; 49: 4551.
Downloaded from http://jac.oxfordjournals.org/ by guest on June 2, 2013
8. Cappelletty DM Rybak MJ. Comparison of methodologies for
synergism testing of drug combinations against resistant strains of
Pseudomonas aeruginosa. Antimicrob Agents Chemother 1996; 40:
67783.
9. National Committee for Clinical Laboratory Standards. Methods
for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow
AerobicallySixth Edition: Approved Standard M7-A6. NCCLS,
Wayne, PA, USA, 2003.
10. Eliopoulos GM Moellering RC Antimicrobial combinations. In:
Lorian V, ed. Antibiotic in Laboratory Medicine, 4th edn. Baltimore, MD:
Williams and Wilkins, 1996; 330-96.