Time-Kill Synergy Tests of Tigecycline Combined with Imipenem, Amikacin

Available online at www.annclinlabsci.org

Annals of Clinical & Laboratory Science, vol. 41, no. 1, 2011 39 Time-Kill Synergy Tests of Tigecycline Combined with Imipenem, Amikacin, and Ciprofloxacin against Clinical Isolates of Multidrug-Resistant Klebsiella pneumoniae and Escherichia coli

Haejun Yim,1 Heungjeong Woo,2 Wonkeun Song,3 Min-Jeong Park,3 Hyun Soo Kim,3 Kyu Man Lee,2 Jun Hur,1 and Man-Seung Park4 Departments of 1Burn Surgery, 2Internal Medicine, 3Laboratory Medicine, and 4Microbiology, Hallym University Medical Center and College of Medicine, Seoul and Chuncheon, Korea

Abstract. This study evaluated the activity of tigecycline combined with imipenem, amikacin, and ciprofloxacin against clinical isolates of multidrug-resistant Klebsiella pneumoniae and Escherichia coli co-producing extended- spectrum β-lactamases and acquired AmpC β-lactamases. Broth microdilution tests were performed for cefotaxime, ceftazidime, cefepime, imipenem, amikacin, ciprofloxacin, and tigecycline. Time-kill synergy studies were tested for tigecycline plus imipenem, tigecycline plus amikacin, and tigecycline plus ciprofloxacin. Imipenem (MIC90 = 1 µg/ ml for both K. pneumoniae and E. coli) and tigecycline (MIC90 = 2 µg/ml for K. pneumoniae and 1 µg/ml for E. coli) were the most potent agents. Combination studies with tigecycline plus imipenem resulted in synergy against 18 K. pneumoniae and 3 E. coli isolates; tigecycline plus amikacin yielded synergy against 8 K. pneumoniae and 3 E. coli isolates; tigecycline plus ciprofloxacin yielded synergy against 7K. pneumoniae and 2 E. coli isolates. No antagonism was observed with any combination. In the present study, imipenem, amikacin, and ciprofloxacin led to indifferent and some synergistic effects in combination with tigecycline, and none of them demonstrated antagonistic effects.

Introduction AmpCs, and isolates with reduced but may not always be effective as susceptibility to carbapenems con- monotherapy [9]. Therefore, a study b-Lactamase production is the most tinue to spread worldwide, new was designed to investigate the in common resistance mechanism to therapeutic options are needed [2-4]. vitro activities of combinations of b-lactams in Klebsiella pneumoniae Tigecycline, the first class of tigecycline and representative bac- and Escherichia coli. K. pneumoniae glycyclines, exhibits expanded- tericidal agents including imipenem, and E. coli isolates producing exten- spectrum of activity against a wide amikacin, and ciprofloxacin. Anti- ded-spectrum b-lactamases (ESBLs) variety of bacteria, including microbial combination therapy may and/or acquired AmpC b-lactamases multidrug-resistant strains such as provide clinicians an option in have been prevalent worldwide. ESBL-producing Enterobacteria- addition to imipenem for difficult ESBL-producing isolates are consid- ceae, multidrug-resistant Acineto- infections due to ESBL- or AmpC- ered resistant to all penicillins, bacter baumannii, methicillin- producing Enterobacteriaceae. cephalosporins, and aztreonam; resistant Staphylococcus aureus, and Although ESBL- and/or acquired carbapenems are the only b-lactams vancomycin-resistant enterococci AmpC-producing K. pneumoniae that remain consistently active for [5,6]. Tigecycline acts by binding to and E. coli routinely express cross- treatment in severe cases. The the 30S ribosomal subunit and resistance to other drug classes, therapeutic options for severe prevents the binding of aminoacyl synergy from combination therapy infections caused by acquired transfer RNA to the acceptor site on may achieve successful outcomes. AmpC-producing isolates are the messenger RNA-ribosome The purpose of this study was to almost restricted to carbapenems complex. Protein synthesis is evaluate the in vitro antimicrobial and cefepime [1]. Since K. pneu- inhibited, thereby exhibiting a activity of tigecycline combined moniae- and E. coli-resistant pheno- bacteriostatic effect [7]. Tigecycline with other antimicrobials against types, such as ESBLs, acquired has potentially useful activity [8] clinical isolates of multidrug-

Address correspondence to Wonkeun Song, M.D., Ph.D., Department of Laboratory Medicine, Kangnam Sacred Heart Hospital, 948-1 Daelim-dong, Youngdeungpo-gu, Seoul 150-950, Republic of Korea; e-mail [email protected]. 0091-7370/11/0039-0043. $2.50. © 2011 by the Association of Clinical Scientists, Inc. 40 Annals of Clinical & Laboratory Science, vol. 41, no. 1, 2011 resistant K. pneumoniae and E. coli 5 E. coli) with tigecycline MICs of each isolates, and killing was co-producing ESBLs and acquired ≥0.125 μg/ml were chosen for time- assessed at 0, 4, 8, and 24 hr. AmpC b-lactamases. kill experiments. The combination Aliquots (0.1 ml) were removed activities of tigecycline and either from 2-ml cultures at 0, 4, 8, and 24 Materials and Methods imipenem, amikacin, or ciproflox- hr and serially diluted in 0.85% Bacterial isolates. This study acin were further evaluated in time- sterile saline. Bacterial counts were included 35 clinical isolates of K. kill experiments with concentrations determined by plating 0.1 ml of pneumoniae and 8 clinical isolates of at equal and one-fourth the MICs. appropriate dilutions to enumerate E. coli known to co-produce ESBLs Mueller-Hinton broth cultures were CFU/ml. Plating was performed in 5 and AmpCs, which were obtained inoculated with ≥5 x 10 CFU/ml of duplicate and the blood agar plates from two teaching hospitals in Korea during 2009 (Table 1). The Table 1. Bacterial isolates co-producing extended-spectrum b-lactamases and isolates were identified with the acquired AmpC b-lactamases used in this study Vitek 2 system (bioMerieux Vitek, Hazelwood, MO). Searches for Organism No. of isolates β-Lactamase genes coding for the class A ESBLs K. pneumoniae (n = 35) 23 SHV-12 plus DHA-1 were performed by PCR amplifi- 1 SHV-12 plus CMY-2 cation as described previousely [10]. 1 SHV-12 and CTX-M-12 plus DHA-1 To identify isolates with ampC 1 SHV-12 and CTX-M-15 plus DHA-1 genes, AmpC multiplex PCR was 2 CTX-M-14 plus DHA-1 performed by the method of Perez- 7 CTX-M-15 plus DHA-1 Perez and Hanson [11]. Sequencing of ampC genes with primers DHA- E. coli (n = 8) 4 SHV-12 plus DHA-1 1, CMY-1, and CMY-2 was perform- 1 CTX-M-14 plus DHA-1 ed as described previously [12]. The 1 CTX-M-15 plus CMY-1 1 CTX-M-15 plus CMY-2 PCR products were subjected to 1 CTX-M-24 plus CMY-2 direct sequencing. Both strands of each PCR product were sequenced twice with an automatic sequencer (model 3730xl; Applied Biosystems, Table 2. MICs for 43 isolates of E. coli and K. pneumoniae co-producing extended- Weiterstadt, Germany). Sequence spectrum b-lactamases and acquired AmpC b-lactamases. alignment and analysis was performed online using the BLAST Organism and Agent MIC (μg/ml) of: program (National Center for Bio- Range 50% 90% technology Information (Bethesda, K. pneumoniae (n = 35) MD; www.ncbi.nlm.nih.gov). Ceftazidime 8 – ≥ 512 ≥ 512 ≥ 512 Antimicrobial susceptibility testing. Cefotaxime 8 – ≥ 512 64 ≥ 512 Broth microdilution MIC tests were Cefepime 0.25 – ≥ 512 8 256 performed for ceftazidime (Sigma- Imipenem 0.125 – 64 0.5 1 Aldrich, St. Louis, MO), cefotaxime Amikacin ≤ 0.125 – ≥ 512 ≥ 512 ≥ 512 (Sigma-Aldrich), cefepime (Bor- Ciprofloxacin ≤ 0.125 – ≥ 512 32 ≥ 512 yung, Seoul, Korea), imipenem Tigecycline ≤ 0.125 – 4 0.5 2 (LKT Laboratories, St. Paul, MN), amikacin (Sigma-Aldrich), cipro- E. coli (n = 8) floxacin (LKT Laboratories), and tigecycline (Wyeth, Pearl River, Ceftazidime 8 – ≥ 512 ≥ 512 ≥ 512 Cefotaxime 16 – ≥ 512 256 ≥ 512 NY), according to the CLSI M7-A7 Cefepime 1 – ≥ 512 16 ≥ 512 methods [13]. For quality control, Imipenem ≤ 0.125 – 1 0.25 1 E. coli ATCC 25922 and K. Amikacin 1 – ≥ 512 4 ≥ 512 pneumoniae ATCC 700603, were Ciprofloxacin 0.25 – ≥ 512 128 ≥ 512 included in each set of tests. Tigecycline ≤ 0.125 – 1 ≤ 0.125 1

Time-kill synergy study. Thirty- three strains (28 K. pneumoniae and Tigecycline against Enterobactericeae 41 Table 3. Results of time-kill synergy for 33 isolates of E. coli and K. pneumoniae co-producing extended-spectrum β-lactamases and acquired AmpC β-lactamases at equal to one-fourth the MICs for tigecycline, imipenem, amikacin, and ciprofloxacin.

Isolate MIC (μg/ml) Time-kill synergy (μg/ml) TIG IPM AMK CIP TIG+IPM TIG+AMK TIG+CIP

K. pneumoniae K19 0.5 0.5 0.125 2 S(0.5+0.5) I S(0.5+2) K21 0.125 0.5 >256 16 I NT S(0.125+16) K24 0.25 32 >256 64 S(0.25+32) NT I K35 2 <0.125 32 32 NT S(2+32, 2+8, 0.5+32, I 0.5+8) K36 0.25 1 >256 1 I NT 1 K50 0.5 1 >256 2 S(0.5+1, 0.5+0.125) NT S(0.5+2, 0.5+0.5) K55 0.5 1 16 0.5 S(0.5+1, 0.5+0.125) S(0.5+16) I K56 2 0.5 >256 128 S(2+0.5) NT S(2+128) K65 1 0.5 16 >256 S(1+0.5) I NT K72 0.5 0.5 >256 2 S(0.5+0.5) NT I K75 1 64 >256 >256 I NT NT K76 0.25 <0.125 2 64 NT S(0.25+2) I K77 4 1 8 >256 S(4+1, 4+0.125, 1+1) S(4+8, 4+2) NT K86 0.25 1 >256 4 I NT I K87 0.5 0.5 2 0.25 S(0.5+0.5) S(0.5+2, 0.5+0.5) S(0.5+0.25) K89 0.5 1 >256 64 S(0.5+1) NT I K91 0.25 0.25 0.5 32 I S(0.25+0.5) S(0.25+32) K92 1 0.5 16 >256 S(1+0.5) I NT K97 2 0.5 16 >256 I I NT K102 0.25 0.5 >256 2 S(0.25+0.5) NT I K124 0.5 1 >256 128 S(0.5+1) NT I K137 1 1 8 >256 S(1+1, 1+0.125) S(1+8) NT K138 4 1 >256 >256 S(4+1, 4+0.125) NT NT K149 1 0.5 8 >256 I I NT K153 4 1 >256 16 S(4+1) NT S(4+16, 4+4, 1+16) K155 0.25 0.25 1 0.125 I I I K188 0.25 1 >256 64 S(0.25+1, 0.25+0.125) NT I K223 2 0.5 16 4 S(2+0.5) S(2+16) I E. coli E11 0.125 0.125 1 128 S(0.125+0.125) S(0.125+1) S(0.125+128) E28 0.5 0.25 2 >256 I I NT E36 0.125 1 >256 32 S(0.125+1) NT I E96 1 1 64 64 S(1+1, 0.125+1, S(1+64, 0.125+64, S(1+64,0.125+64, 0.125+0.125) 0.125+16) 0.125+16) E173 0.125 0.25 64 128 I S(0.125+64) I

Abbreviations: TIG, tigecycline; IPM, imipenem; AMK, amikacin; CIP, ciprofloxacin; S, synergy; I, indifferent; NT, no tested.

were incubated 19 to 24 hr at 35°C. were defined at 24 hr as a ± 1 log10 μg/ml and 0.25 to 0.5 μg/ml, CFU/ml kill to <2-log CFU/ml Synergy was defined as a ≥2-log10 10 respectively, whereas MIC90 values CFU/ml diminution between the compared to the most active single ranged from 1 to 2 μg/ml and 1 μg/ combination and the most active agent and >1 log10 CFU/ml growth ml, respectively. Tigecycline MIC single agent at 24 hr. The number of compared to the least active single values of >2 μg/ml were detected in surviving organisms in the presence agent, respectively [14]. only three (8.6%) K. pneumoniae of the combination was ≥2-log10 Results isolates. Based on the MIC90s, CFU/ml and at least one of the imipenem (MIC90 = 1 μg/ml for drugs alone did not affect the MICs results are listed in Table 2. both K. pneumoniae and E. coli) and growth curve of the tested organism. The tigecycline and imipenem tigecycline (MIC90 = 2 μg/ml for K. Indifferent effect and antagonism MIC50s ranged from ≤0.125 to 0.5 pneumoniae and 1 μg/ml for E. coli) 42 Annals of Clinical & Laboratory Science, vol. 41, no. 1, 2011 were the most potent agents. with Enterobacteriaceae harboring ceae. Further studies in animal Ceftazidime, cefotaxime, cefepime, an ESBL and/or AmpC can be models are needed to better under- amikacin, and ciprofloxacin were treated with carbapenem mono- stand the potential utility of tigecyc- less active, exhibiting high MIC90 therapy. However, combination line combination therapy for multi- (≥256 μg/ml for both K. pneumoniae therapy may also reduce the drug-resistant Enterobacteriaceae. and E. coli). emergence of resistance and improve Acknowledgments Results of time-kill synergy the spectrum of activity [19]. Our studies are presented in Table 3. At finding that drug interations The authors are grateful to Tae-Jae levels equal the MIC, combination between tigecycline and other Lee for excellent technical assistance. studies with tigecycline plus compounds such as cefepime, This study was supported by a grant imipenem were synergistic against imipenem, and gentamicin were of the Korea Healthcare Technology 18 K. pneumoniae and 3 E. coli essentially indifferent or synergistic R & D Project, Ministry for Health, isolates; tigecycline plus amikacin is in accord with previous results Welfare & Family Affairs, Republic were synergistic against 8 K. pneu- [20]. The combination of one-fourth of Korea (A084589). moniae and 3 E. coli isolates; MIC of tigecycline and one-fourth tigecycline plus ciprofloxacin were MIC of imipenem, amikacin or References synergistic against 7 K. pneumoniae ciprofloxacin was rarely synergistic 1. Paterson DL. Resistance in gram- and 2 E. coli isolates. At one-fourth and no antagonism was observed negative bacteria: Enterobacter- the MIC, the combination studies with any combinations. Overall, the iaceae. Am J Infect Control 2006; with tigecycline plus imipenem present results indicate that the 34(Suppl 1):S20-S28. were synergistic against 6 K. pneu- interaction of tigecycline with 2. Baudry PJ, Nichol K, DeCorby M, moniae and 1 E. coli isolates; imipenem, amikacin, or ciproflox- Mataseje L, Mulvey MR, Hoban tigecycline plus amikacin were acin against K. pneumoniae and E. DJ, Zhanel GG. Comparison of synergistic against 3 K. pneumoniae coli co-producing ESBL and antimicrobial resistance profiles among extended-spectrum-b-lac- E. coli and 1 isolates; tigecycline acquired AmpC was essentially tamase-producing and acquired plus ciprofloxacin were synergistic indifferent or synergistic. Thus, AmpC b-lactamase-producing against 2 K. pneumoniae and 1 E. tigecycline could be used safely with Escherichia coli isolates from coli isolates. No antagonism was imipenem, amikacin or ciproflox- Canadian intensive care units. observed with any tigecycline acin. Antimicrob Agents Chemother combinations against all isolates. Two K. pneumoniae isolates 2008;52:1864-1869. showed high MICs of imipenem (32 3. Potz NA, Hope R, Warner M, Discussion and 64 μg/ml). The isolates showed Johnson AP, Livermore DM. negative results in a modified Hodge Prevalence and mechanisms of Tigecycline and imipenem showed cephalosporin resistance in Enter- excellent in vitro activity against test and EDTA-sodium mercapto- obacteriaceae in London and Enterobacteriaceae regardless of the acetic acid double-disk synergy test South-East England. J Antimicrob presence or absence of ESBLs and/or [21] for screening of carbapenemases Chemother 2006;58:32032-32036. AmpCs [15,16]. In this study, and metallo-b-lactamases, respec- 4. Woodford N, Reddy S, Fagan EJ, tigecycline and imipenem were also tively (data not shown). Porin loss Hill RL, Hopkins KL, Kaufmann shown to be the most active against may have reduced susceptibility to ME, Kistler J, Palepou MF, Pike K. pneumoniae and E. coli co-pro- carbapenems. However, no activity R, Ward ME, Cheesbrough J, Livermore DM. Wide geographic ducing ESBL and acquired AmpC. of tigecycline is affected by porin loss [15]. The two isolates were also spread of diverse acquired Inappropriate antimicrobial b-lactamases among Escherichia therapy for patients with isolates susceptible to tigecycline, with coli and Klebsiella spp. in the UK harboring an ESBL and/or AmpC is MICs of 0.25 and 1 μg/ml each. and Ireland. J Antimicrob Chemo- correlated with increased mortality In summary, in the present ther 2007;59:102-105. [17,18]. The complicated nature of study, imipenem, amikacin, and 5. Hoban DJ, Bouchillon SK, infections associated with ESBL- or ciprofloxacin led to indifferent and Dowzicky MJ. Antimicrobial AmpC-producing Enterobacteria- some synergistic effects in combin- susceptibility of extended- b ceae often leads us to consider ation with tigecycline, and none of spectrum -lactamase producers and multidrug-resistant Acineto- combination therapy. The use of them demonstrated antagonistic effects. Therefore, they may be used bacter baumannii throughout the antimicrobial combinations is one United States and comparative in of the best options available to treat often in clinical practice, especially vitro activity of tigecycline, a new infections caused by multidrug- treatment of infection due to glycycline antimicrobial. Diagn resistant bacteria. Most patients multidrug-resistant Enterobacteria- Tigecycline against Enterobactericeae 43 Microbiol Infect Dis 2007;57:423- PCR. J Clin Microbiol 2002;40: Proteus mirabilis isolates collected 428. 2153-2162. during tigecycline phase 3 clinical 6. Souli M, Kontopidou FV, 12. Song W, Kim JS, Kim HS, Yong trials. Antimicrob Agents Chemo- Koratzanis E, Antoniadou A, D, Jeong SH, Park MJ, Lee KM. ther 2009;53:465-475. Giannitsioti E, Evangelopoulou P, Increasing trend in the presence of 17. Kotapati S, Kuti JL, Nightingale Kannavaki S, Giamarellou H. In plasmid-mediated AmpC b-lacta- CH, Nicolau DP. Clinical vitro activity of tigecycline against mases in Enterobacteriaceae implications of extended-spectrum multi-drug-resistant, including lacking chromosomal ampC gene β-lactamase (ESBL) producing pan-resistant, gram-negative and at a Korean university hospital Klebsiella species and Escherichia gram-positive clinical isolates from from 2002 to 2004. Diagn Micro- coli on cefepime effectiveness. J Greek hospitals. Antimicrob Agents biol Infect Dis 2006;55:219-224. Infect 2005;51:211-217. Chemother 2006;50:3166-3169. 13. Clinical and Laboratory Standards 18. Pai H, Kang CI, Byeon JH, Lee 7. Mullangi PK, Pankey GA. Institute. Methods for dilution KD, Park WB, Kim HB, Kim EC, Tigecycline in critical care. Crit antimicrobial susceptibility tests Oh MD, Choe KW. Epidemiology Care Clin 2008;24:365-75. for bacteria grow aerobically: and clinical features of blood- 8. Milatovic D, Schmitz FJ, Verhoef approved standard M7-A7. CLSI stream infections caused by J, Fluit AC. Activities of the 2007; Wayne, PA. AmpC-type-β-lactamase prod- glycycline tigecycline (GAR-936) 14. Vidaillac C, Leonard SN, Sader ucing Klebsiella pneumoniae. against 1,924 recent European HS, Jones RN, Rybak MJ. In vitro Antimicrob Agents Chemother clinical bacterial isolates. Anti- activity of ceftaroline alone and in 2004;48:3720-3728. microb Agents Chemother 2003; combination against clinical 19. Bouza E, Munoz P. Monotherapy 47:400-404. isolates of resistant gram-negative versus combination therapy for 9. Nathisuwan S, Burgess DS, Lewis pathogens, including b-lactamase- bacterial infections. Med Clin JS II. Extended-spectrum b-lactam- producing Enterobacteriaceae and North Am 2000;84:1357-1389. ases: epidemiology, detection, and Pseudomonas aeruginosa. Anti- 20. Cha R. In vitro activity of treatment. Pharmacotherapy. 2001; microb Agents Chemother 2009; cefepime, imipenem, tigecycline, 21:920-928. 53:2360-2366. and gentamicin, alone and in 10. Ryoo NH, Kim EC, Hong SG, 15. Conejo MC, Hernandez JR, combination, against extended- Park YJ, Lee K, Bae IK, Song EH, Pascual A. Effect of porin loss on spectrum b-lactamase-producing Jeong SH. Dissemination of SHV- the activity of tigecycline against Klebsiella pneumoniae and 12 and CTX-M-type extended- Klebsiella pneumoniae producing Escherichia coli. Pharmacotherapy spectrum b-lactamases among extended-spectrum b-lactamases 2008;28:295-300. clinical isolates of Escherichia coli or plasmid-mediated AmpC-type 21. Lee K, Chong Y, Shin HB, Kim and Klebsiella pneumoniae and b-lactamases. Diagn Microbiol YA, Yong D, Yum JH. Modified emergence of GES-3 in Korea. J Infect Dis 2008;61:343-345. Hodge and EDTA-disk synergy Antimicrob Chemother 2005;56: 16. Jones CH, Tuckman M, Keeney tests to screen metallo-b- 698-702. D, Ruzin A, Bradford PA. Charac- lactamase-producing strains of 11. Perez-Perez FJ, Hanson ND. terization and sequence analysis of Pseudomonas and Acinetobacter Detection of plasmid-mediated extended-spectrum b-lactamase- species. Clin Microbiol Infect AmpC b-lactamase genes in encoding genes from Escherichia 2001;7:88-91. clinical isolates by using multiplex coli, Klebsiella pneumoniae, and