Bacterial keratitis is a major cause of corneal opacity and loss

Immunology and Microbiology An In Vitro Investigation of Synergy or Antagonism between Antimicrobial Combinations against Isolates from Bacterial Keratitis Henri Sueke, 1,2 Stephen B. Kaye, 1 Timothy Neal, 3 Amanda Hall, 2 Stephen Tuft, 4 and Christopher M. Parry 2 From the Departments of 1 Ophthalmology and 3 Medical Microbiology, Royal Liverpool University Hospital, Liverpool, United Kingdom; the 2 School of Infection and Host Defence, University of Liverpool, Liverpool, United Kingdom; and 4 Moorfields Eye Hospital, London, United Kingdom. Supported by a Research Grant from the Foundation for the Prevention of Blindness. Submitted for publication October 28, 2009; revised February 17, 2010; accepted March 8, 2010. Disclosure: H. Sueke, None; S.B. Kaye, None; T. Neal, None; A. Hall, None; S. Tuft, None; C.M. Parry, None Corresponding author: Henri Sueke, Department of Ophthalmology, Royal Liverpool University Hospital, Prescot Street, Liverpool L7 8XP, UK; hsueke@gmail.com. PURPOSE. To investigate antimicrobial combinations for synergy or antagonism against isolates of Staphylococcus aureus and Pseudomonas aeruginosa. METHODS. Isolates were collected from cases of microbial keratitis from six centers in the United Kingdom. Minimum inhibitory concentrations (MICs) were determined by using E-test strips for 16 antimicrobials, including both current and potentially available agents. E-test strips were used to test selected antimicrobials in combination against a representative set of 10 S. aureus and 10 P. aeruginosa isolates. E-tests of the two antimicrobials were placed on sensitivity agar at right angles intersecting at their respective MICs. Antimicrobial combinations were classified as synergistic, additive, indifferent, or antagonistic, according to their fractional inhibitory concentration (FIC). RESULTS. The combinations meropenem and ciprofloxacin, meropenem and teicoplanin, moxifloxacin and teicoplanin, and ciprofloxacin and teicoplanin, gave the lowest mean FICs for S. aureus, with synergy or additivity being seen in 60% to 80% of isolates. The meropenem/ciprofloxacin combination gave the lowest mean FIC for P. aeruginosa isolates, with 90% showing an additive or synergistic effect. The other combinations elicited a predominantly indifferent response. No consistent antagonistic effect was observed with the combinations used. CONCLUSIONS. The combination of meropenem and ciprofloxacin was predominantly additive or synergistic for both S. aureus and P. aeruginosa. Teicoplanin combined with meropenem, ciprofloxacin, or moxifloxacin was also predominantly additive or synergistic against S. aureus. (Invest Ophthalmol Vis Sci. 2010;51:4151 4155) DOI:10.1167/iovs.09-4839 Bacterial keratitis is a major cause of corneal opacity and loss of vision worldwide, and topical antimicrobial therapy is a critical component in its management. 1,2 Staphylococcus aureus and Pseudomonas aeruginosa are major causative pathogens in this condition. Before the introduction of the fluoroquinolones in the 1990s, combination therapy with a fortified aminoglycoside and a cephalosporin was the commonly used empiric therapy for suspected bacterial keratitis, providing inhibitory action against P. aeruginosa, S. aureus, and Streptococcus pneumoniae. The increase in use of the second- and third-generation fluoroquinolones in the 1990s 1 was subsequently accompanied by an increase in bacterial resistance in cases of bacterial keratitis. 2 5 As opposed to single therapy, an antimicrobial combination offers a broader spectrum of activity 6 and may reduce selective pressures. This finding is of particular importance for the fluoroquinolones, as increasing resistance has been reported in S. aureus 5,7 9 and P. aeruginosa 4,10 isolates from cases of bacterial keratitis. Combination therapy may also result in synergy as occurs, for example, with the combination of penicillin and gentamicin when used in the treatment of enterococcal endocarditis. 10 Conversely, combinations of antimicrobials may be antagonistic, as occurs with the combination of chloramphenicol and penicillin in the treatment of pneumococcal meningitis. 11 To our knowledge, there have been no in vitro studies in which researchers have assessed clinically relevant combinations of antimicrobials against isolates in bacterial keratitis. The Microbiology Ophthalmic Group (MOG), which comprises six ophthalmic centers in the United Kingdom (London, Birmingham, Newcastle, Bristol, Manchester, and Liverpool), was established in April 2003 to investigate the characteristics and antimicrobial susceptibility of bacterial isolates obtained from cases of microbial keratitis. In a previous study 12 the minimum inhibitory concentrations (MICs) of standard and newer antimicrobials were determined against these isolates. In the present study, we investigated the in vitro interaction of clinically relevant antimicrobial combinations against a selection of S. aureus and P. aeruginosa isolates collected from patients with bacterial keratitis. METHODS Bacterial isolates collected from the corneas of patients with bacterial keratitis at the six contributing centers were submitted at the time of isolation to the School of Infection and Host Defense, the University of Liverpool, where they were stored on beads (Protect; TSC Ltd., Heywood, UK) at 70 C. The isolates were subsequently subcultured from the beads, and their identity reconfirmed by standard methods. 13 A representative sample with a wide range of MICs to the individual antimicrobials previously tested 12,14 were chosen from the collection, comprising 10 S. aureus and P. aeruginosa isolates. Three of the S. aureus isolates were methicillin resistant (MRSA). MICs were determined using E-test strips (AB Biodisk, Solna, Sweden) according to British Society of Antimicrobial Chemotherapy (BSAC) methods, 15,16 the results of which have been published. 12 The quality controlled strains S. aureus ATCC (American Type Culture Collection, Manassas, Investigative Ophthalmology & Visual Science, August 2010, Vol. 51, No. 8 Copyright Association for Research in Vision and Ophthalmology 4151

4152 Sueke et al. IOVS, August 2010, Vol. 51, No. 8 measured in combination (FIC 0.125/1 0.063/0.5 0.25). Figure 2b shows additivity: The MICs of A and B were 1.0 and 0.5 mg/l and decreased to 0.5 and 0.063 mg/l, respectively, when measured in combination (FIC 0.5/1 0.063/0.5 0.62). It is apparent that a synergistic or additive effect can occur for the combination only if both FIC A and FIC B are each less than 1.0. Figure 2c shows indifference: MICs of A and B were 1.0 and 0.5 mg/l, with no change when measured in combination (FIC; 1/1 0.5/0.5 2). Figure 2d demonstrates antagonism: The MICs of A and B were 1 and 0.5 mg/l, respectively, and increased to 8 and 4 mg/l after combination (FIC 8/1 4/0.5 16). Figure 3 shows a photograph demonstrating the additivity of gentamicin and penicillin against fecal streptococci. The MICs of gentamicin and penicillin alone were 8.0 and 1.0 mg/l and, when measured in combination, were 2 and 0.38 mg/l, respectively (FIC 2.0/8.0 0.38/1 0.63). The mean FIC of triplicate experiments for each antimicrobial combination for a particular isolate was then used to determine whether the combination would demonstrate a synergistic, additive, indifferent, or antagonistic effect on that bacterium. The mean, standard deviation, minimum and maximum, and coefficient of variance of the FIC was calculated for each antimicrobial combination against all S. aureus and P. aeruginosa isolates. RESULTS FIGURE 1. The setup of an agar plate for E-test combination testing. The two E-tests are crossed at 90 at the point of their individual MICs. VA) 25923, P. aeruginosa ATCC 27853, and Escherichia coli ATCC 25922 were tested in the individual E-tests to ensure that the expected values were obtained, as previously described. 12 Once the MICs for individual antimicrobials against the S. aureus and P. aeruginosa isolates were measured, 12 the in vitro activity of each combination was determined by placing E-test strips of the two antimicrobials on the agar at a 90 angle with the intersection at the respective MICs for the organism (Fig. 1). The agar plates were incubated at 35 C to 37 C in air for 18 hours, and the MIC of each antimicrobial in the combination was read. Each bacterial isolate was tested three times with each antimicrobial combination. The antimicrobial combinations tested for S. aureus were teicoplanin and moxifloxacin, teicoplanin and ciprofloxacin, teicoplanin and meropenem, meropenem and linezolid, moxifloxacin and linezolid, meropenem and ciprofloxacin, and moxifloxacin and meropenem (n 210). The following combinations were tested for P. aeruginosa: meropenem and ciprofloxacin, gentamicin and moxifloxacin, moxifloxacin and meropenem, meropenem and levofloxacin, and gentamicin and ciprofloxacin (n 150). Using the results of MICs determined with the antimicrobial alone and in combination, the fractional inhibitory concentration (FIC) was calculated for each antimicrobial combination according to the following formulas 15 : MIC drug A when tested in combination with drug B FIC of drug A MIC of drug A alone MIC drug B when tested in combination with drug A FIC of drug B MIC of drug B alone FIC FIC A FIC B Our interpretation of the FIC results, according to accepted criteria, 15,17,18 were as follows: 0.5, synergy; 0.5 to 1.0, additivity; 1.0 to 4.0, indifference; and 4, antagonism. Examples are demonstrated in Figures 2 and 3. Figure 2a shows synergy: The MICs of A and B were 1.0 and 0.5 mg/l and decreased to 0.125 and 0.063 mg/l when The results for each antimicrobial combination against S. aureus and P. aeruginosa are shown in Tables 1 and 2, respectively. The coefficient of variance for the antimicrobial combinations was 10.9% against S. aureus and 21.7% against P. aeruginosa. The antimicrobial combinations with the greatest additive effect against S. aureus were meropenem and teicoplanin, meropenem and ciprofloxacin, and moxifloxacin and teicoplanin (Table 1). Synergy was demonstrated in 20% of isolates with the combination of meropenem and teicoplanin and in 10% of isolates with meropenem and ciprofloxacin. Synergy or additivity was seen in 80% of isolates with each of those two combinations. The remaining combinations including ciprofloxacin and teicoplanin, linezolid and moxifloxacin, moxifloxacin and meropenem, and linezolid and meropenem demonstrated a predominantly indifferent interaction. No antagonism was seen in any of the experiments. The results of the combinations against P. aeruginosa isolates are presented in Table 2. Meropenem and ciprofloxacin demonstrated the lowest mean FIC (0.7) with synergy observed in 10% and additivity in 80%. The interaction between gentamicin with moxifloxacin, meropenem and moxifloxacin, meropenem and levofloxacin, and gentamicin and ciprofloxacin were predominantly indifferent. No antagonistic effect was seen, although one isolate demonstrated a mean FIC of the three tests of 3.7 (3.5, 3.5, and 4.1) with the combination of meropenem and moxifloxacin. DISCUSSION The intention of antimicrobial sensitivity testing is to provide a prediction of success or failure when a particular antimicrobial is used to treat a specific infection. Although in bacterial keratitis the in vitro activity of an antimicrobial does not necessarily equate with the in vivo biological activity within the cornea, 19 results in previous work suggest a relationship between the MIC of an antimicrobial and clinical outcome in cases of keratitis due to S. aureus and P. aeruginosa treated with a single antimicrobial. 14 In particular, a lower MIC was associated with a faster healing response. It is reasonable to assume that the lower the MIC of an antimicrobial for a given isolate, the more likely it is that the infection will respond to treatment and that the MIC of the antimicrobial can be used to

IOVS, August 2010, Vol. 51, No. 8 Antimicrobial Combinations for Bacterial Keratitis 4153 FIGURE 2. E-test strip experiments on antimicrobial combinations labeled A and B. (a) Synergy: The MIC of antimicrobial A was 1.0 mg/l when tested alone, but was 0.125 mg/l when tested in combination with antimicrobial B; the MIC of antimicrobial B was 0.5 mg/l when tested alone, but was 0.063 mg/l when tested in combination with antimicrobial A (FIC 0.25). (b) Additivity: The MIC of antimicrobial A was 1.0 mg/l when tested alone, but was 0.5 mg/l when tested in combination with antimicrobial B; the MIC of antimicrobial B was 0.5 mg/l when tested alone, but was 0.063 mg/l when tested in combination with antimicrobial A (FIC 0.62). (c) Indifference: The MIC of antimicrobial A was 1.0 mg/l when tested alone or in combination with antimicrobial B; the MIC of antimicrobial B was 0.5 mg/l when tested alone or in combination with antimicrobial A (FIC 2). (d) Antagonism: The MIC of antimicrobial A was 1.0 mg/l when tested alone, but was 8.0 mg/l when tested in combination with antimicrobial B; the MIC of antimicrobial B was 0.5 mg/l when tested alone, but was 4.0 mg/l when tested in combination with antimicrobial A (FIC 16). Arrows: the changing MICs. evaluate the potential efficacy of a given agent for the treatment of microbial keratitis. It is not known whether the distribution of antimicrobial sensitivities of bacterial isolates tested in this study were affected by prior antimicrobial treatment before isolation of the bacteria from the corneal ulcer. However, the isolates were selected from a national collection with a distribution of isolates similar to that in previous studies, and we think they are therefore representative of the bacteria that cause keratitis. 20 22 If a combination of antimicrobials demonstrates a synergistic or additive effect as determined by MIC, this combination may prove more effective than monotherapy with the individ-

4154 Sueke et al. IOVS, August 2010, Vol. 51, No. 8 FIGURE 3. Photograph of an E-test combination experiment between gentamicin and penicillin against fecal streptococci, demonstrating additivity. The MICs of gentamicin and penicillin alone were 8 and 1.0 mg/l, respectively, and when measured in combination were 2 and 0.38 mg/l, respectively (FIC 0.63). ual agents. It should be noted that the definitions of effect, from synergy through to antagonism, 15,17,18 are definitions that relate to interaction in vitro, and it is unknown whether they translate into an improved outcome for topical combination therapy. If the extrapolation to an in vivo effect is valid, a synergistic or additive antimicrobial combination offers a broader spectrum of activity 6 that may reduce selective pressures and the emergence of resistance. The traditional approaches used to assess antimicrobial combinations in vitro are the checkerboard and time-kill methods. 15 These methods are costly in time and materials, however, and for that reason they are not used in routine clinical practice. The method used in this study with pairs of E-test strips is relatively new and has the advantage that it is easy and inexpensive to perform. 23 The degree of agreement between FIC results calculated by the checkerboard and the E-test method varies in the literature depending on the type of bacteria tested. 24 26 For example, a 55% agreement was found between the results of the two tests when used with Brucella melitensis isolates, 26 63% with Acinetobacter, 25 and 90% with P. aeruginosa. 24 A limitation of the E-test method is that it does not provide information about the bactericidal activity of the combination. We have shown that this method has a reasonably low coefficient of variance and is particularly useful in screening a large number of isolates against several combinations of antimicrobials. The combination of meropenem and teicoplanin gave the lowest mean FIC for S. aureus, with synergy or additivity in 80%. For P. aeruginosa the combination with the lowest mean FIC was meropenem and ciprofloxacin, with synergy or additivity in 90%. Against S. aureus the combinations of teicoplanin with meropenem, ciprofloxacin, or moxifloxacin also gave a low mean FIC with more than 50% of isolates demonstrating either an additive affect or synergy. Other combinations tested were predominantly indifferent. We did not find a combination of antimicrobials that was consistently antagonistic when used against S. aureus or P. aeruginosa. These results indicate that certain combinations of antimicrobials may act synergistically. It is necessary to conduct further investigations in which the checkerboard and time-kill methods are used to determine synergy with these combinations. TABLE 1. The Interactions of Seven Antimicrobial Combinations against S. aureus Isolates Determined by the E-test Method Combination Antagonism Indifferent Additive Synergy Mean (SD) Min/Max Meropenem and teicoplanin 0 20 60 20 0.8 (0.2) 0.3/1.3 Meropenem and ciprofloxacin 0 20 70 10 0.9 (0.3) 0.3/2.0 Meropenem and moxifloxacin 0 60 40 0 1.2 (0.2) 0.6/2.0 Meropenem and linezolid 0 100 0 0 1.9 (0.1) 1.6/2.0 Moxifloxacin and linezolid 0 100 0 0 2.0 (0.0) 1.8/2.0 Moxifloxacin and teicoplanin 0 40 60 0 0.9 (0.1) 0.5/1.3 Ciprofloxacin and teicoplanin 0 40 50 10 1.0 (0.1) 0.3/2.0 The FIC was determined in triplicate for each isolate. The percentage of isolates for which the respective combination produced an antagonistic, indifferent, additive, or synergistic effect is shown. The mean (SD), minimum and maximum FIC for each antimicrobial combination against all S. aureus isolates is given. TABLE 2. Results for the Five Antimicrobial Combinations against P. aeruginosa Isolates Combination Antagonism Indifferent Additive Synergy Mean (SD) Min/Max Meropenem and ciprofloxacin 0 10 80 10 0.7 (0.2) 0.3/1.7 Meropenem and moxifloxacin 0 50 30 20 1.3 (0.2) 0.4/3.7 Meropenem and levofloxacin 10 80 10 0 1.8 (0.5) 0.5/2.0 Gentamicin and ciprofloxacin 10 13 6 0 1.9 (0.4) 0.8/3.1 Gentamicin and moxifloxacin 0 80 0 20 1.3 (0.2) 0.3/2.0 The FIC was determined in triplicate for each isolate. The percentage of isolates for which the respective combination produced an antagonistic, indifferent, additive, or synergistic effect is shown. The mean (SD), minimum and maximum FIC for each antimicrobial combination against all P. aeruginosa isolates is given.

IOVS, August 2010, Vol. 51, No. 8 Antimicrobial Combinations for Bacterial Keratitis 4155 Acknowledgments The authors thank the members of the Microbiology Ophthalmic Group: Steven Tuft (Moorfields Eye Hospital), Stephen Kaye and Timothy Neal (Royal Liverpool University Hospital), Derek Tole and John Leeming (Bristol Eye Hospital), Peter McDonnell (Birmingham and Midlands Eye Hospital), Francisco Figueiredo and Steven Pedler (Royal Victoria Infirmary, Newcastle), and Andrew Tullo and Malcolm Armstrong (Manchester Royal Eye Hospital). The authors acknowledge the assistance of the late Timothy Weller (Birmingham and Midlands Eye Hospital). References 1. Bower KS, Kowalski RP, Gordon YJ. Fluoroquinolones in the treatment of bacterial keratitis. Am J Ophthalmol. 1996;121:712 715. 2. Moshirfar M, Mirzaian G, Feiz V, Kang PC. Fourth-generation fluoroquinolone-resistant bacterial keratitis after refractive surgery. J Cataract Refract Surg. 2006;32:515 518. 3. Park SH, Lim JA, Choi JS, Kim KA, Joo CK. The resistance patterns of normal ocular bacterial flora to 4 fluoroquinolone antibiotics. Cornea. 2009;28:68 72. 4. Garg P, Sharma S, Rao GN. Ciprofloxacin-resistant Pseudomonas keratitis. Ophthalmology. 1999;106:1319 1323. 5. Goldstein MH, Kowalski RP, Gordon YJ. Emerging fluoroquinolone resistance in bacterial keratitis: a 5-year review. Ophthalmology. 1999;106:1313 1318. 6. Allan BD, Dart JK. Strategies for the management of microbial keratitis. Br J Ophthalmol. 1995;79:777 786. 7. Betanzos-Cabrera G, Juarez-Verdayes MA, Gonzalez-Gonzalez G, Cancino-Diaz ME, Cancino-Diaz JC. Gatifloxacin, moxifloxacin, and balofloxacin resistance due to mutations in the gyra and parc genes of Staphylococcus epidermidis strains isolated from patients with endophthalmitis, corneal ulcers and conjunctivitis. Ophthalmic Res. 2009;42:43 48. 8. Iihara H, Suzuki T, Kawamura Y, et al. Emerging multiple mutations and high-level fluoroquinolone resistance in methicillin-resistant Staphylococcus aureus isolated from ocular infections. Diagn Microbiol Infect Dis. 2006;56:297 303. 9. Hoshi S, Kikuchi K, Sasaki T, Sotozono C, Kinoshita S, Hiramatsu K. Postantibiotic effects and bactericidal activities of levofloxacin and gatifloxacin at concentrations simulating those of topical ophthalmic administration against fluoroquinolone-resistant and fluoroquinolone-sensitive methicillin-resistant Staphylococcus aureus strains. Antimicrob Agents Chemother. 2008;52:2970 2973. 10. Moellering RC Jr, Wennersten C, Weinberg AN. Synergy of penicillin and gentamicin against Enterococci. J Infect Dis. 1971; 124(suppl):S207 S209. 11. Friedland IR, Klugman KP. Failure of chloramphenicol therapy in penicillin-resistant pneumococcal meningitis. Lancet. 1992;339: 405 408. 12. Sueke H, Kaye S, Neal T, et al. Minimum inhibitory concentrations of standard and novel antimicrobials for isolates from bacterial keratitis. Invest Ophthalmol Vis Sci. 2010;51:2519 2524. 13. Winn WC, Koneman EW. Koneman s Color Atlas and Textbook of Diagnostic Microbiology: Philadelphia: Lippincott Williams and Wilkins; 2006. 14. Kaye SB, Tuft S, Neal T, et al. Bacterial susceptibility to topical antimicrobials and clinical outcome in bacterial keratitis. Invest Ophthalmol Vis Sci. 2009. 15. Lorian V. Antibiotics in Laboratory Medicine. 4th ed. Baltimore: Williams & Wilkins; 1996:xvi, 1238. 16. Methods for Antimicrobial Susceptibility Testing. Birmingham, UK: British Society for Antimicrobial Chemotherapy; 2008. 17. Eliopoulos GM, Eliopoulos CT. Antibiotic combinations: should they be tested? Clin Microbiol Rev. 1988;1:139 156. 18. Moody JA, Gerding DN, Peterson LR. Evaluation of ciprofloxacin s synergism with other agents by multiple in vitro methods. Am J Med. 1987;82:44 54. 19. Kaye SB, Neal T, Nicholson S, et al. Concentration and bioavailability of ciprofloxacin and teicoplanin in the cornea. Invest Ophthalmol Vis Sci. 2009;50:3176 3184. 20. Green M, Apel A, Stapleton F. Risk factors and causative organisms in microbial keratitis. Cornea. 2008;27:22 27. 21. Tuft SJ, Matheson M. In vitro antibiotic resistance in bacterial keratitis in London. Br J Ophthalmol. 2000;84:687 691. 22. Bourcier T, Thomas F, Borderie V, Chaumeil C, Laroche L. Bacterial keratitis: predisposing factors, clinical and microbiological review of 300 cases. Br J Ophthalmol. 2003;87:834 838. 23. White RL, Burgess DS, Manduru M, Bosso JA. Comparison of three different in vitro methods of detecting synergy: time-kill, checkerboard, and E test. Antimicrob Agents Chemother. 1996;40:1914 1918. 24. Balke B, Hogardt M, Schmoldt S, Hoy L, Weissbrodt H, Haussler S. Evaluation of the E test for the assessment of synergy of antibiotic combinations against multiresistant Pseudomonas aeruginosa isolates from cystic fibrosis patients. Eur J Clin Microbiol Infect Dis. 2006;25:25 30. 25. Bonapace CR, White RL, Friedrich LV, Bosso JA. Evaluation of antibiotic synergy against Acinetobacter baumannii: a comparison with Etest, time-kill, and checkerboard methods. Diagn Microbiol Infect Dis. 2000;38:43 50. 26. Orhan G, Bayram A, Zer Y, Balci I. Synergy tests by E test and checkerboard methods of antimicrobial combinations against Brucella melitensis. J Clin Microbiol. 2005;43:140 143.