Christiane Gaudreau* and Huguette Gilbert

Journal of Antimicrobial Chemotherapy (1997) 39, 707 712 JAC Comparison of disc diffusion and agar dilution methods for antibiotic susceptibility testing of Campylobacter jejuni subsp. jejuni and Campylobacter coli Christiane Gaudreau* and Huguette Gilbert Service de Microbiologie Médicale et Maladies Infectieuses, Hôpital Saint-Luc and Université de Montréal, Montréal, Québec, Canada H2X 3J4 The correlation between disc diffusion and agar dilution susceptibility testing of five antibiotics was studied against 145 Campylobacter strains: 99 Campylobacter jejuni subsp. jejuni and 46 Campylobacter coli. The percentages of true results and 95% CI for disc diffusion for resistant strains were 100% (93.2 100%) for tetracycline (53 strains tested), 100% (77.2 100%) for ciprofloxacin (13 strains tested), 86.7% (62.1 96.3%) for nalidixic acid (15 strains tested), 100% (56.6 100%) for erythromycin (five strains tested) and 68.8% (44.4 85.8%) for ampicillin (16 strains tested). The percentages of true results and 95% CI were 97.6 100% and 93.2 100% respectively for 89 140 susceptible strains to the five antibiotics tested. There was a 1.4% major error for nalidixic acid, 0.7% very major error for erythromycin, 5.5% and 1.4% minor and major errors respectively for ampicillin. There was complete agreement for ciprofloxacin and tetracycline. Results of ampicillin susceptibility are not expected to be useful in a clinical setting. The nalidixic acid disc is a marker of ciprofloxacin susceptibility as the nalidixic acid-susceptible strains were susceptible to ciprofloxacin while most of the resistant ones were resistant to ciprofloxacin. Overall, our results suggest that disc diffusion is a reliable, easy and inexpensive susceptibility testing method for C. jejuni and C. coli for erythromycin, ciprofloxacin and tetracycline. Until more erythromycin- and ciprofloxacin-resistant strains are tested to confirm the reliability of this test, the resistance to these drugs needs to be confirmed using the Etest or the agar dilution method. Introduction Campylobacter jejuni subsp. jejuni (C. jejuni) is a major human enteropathogen, while Campylobacter coli causes 5 10% of human Campylobacter spp. infections. 1 Resistance of C. jejuni and C. coli to quinolones is increasing in many countries. 2 4 Resistance to erythromycin is more frequent in C. coli than in C. jejuni and varies between countries for both species. 3 5 An easy method for antimicrobial susceptibility testing of Campylobacter spp. might be useful. The aim of this study was to evaluate the disc diffusion method in comparison with the agar dilution method for antimicrobial susceptibility testing of C. jejuni and C. coli. Materials and methods Bacterial strains A total of 145 human strains of Campylobacter spp. were tested: 10 out of the 99 C. jejuni, nalidixic acid-resistant strains and 37 out of the 46 C. coli were obtained from the Laboratoire de Santé Publique du Québec. The other strains were mostly Campylobacter spp. recently isolated at St-Luc Hospital in Montreal. The strains were identified by standardized methods 1 and were preserved at 70 C in Trypticase Soy Broth (BBL Microbiology Systems, Cockeysville, MD, USA) supplemented with 15% (v/v) glycerol. *Corresponding address: Service de Microbiologie Médicale et Maladies Infectieuses, Hôpital Saint-Luc, 1058 rue Saint-Denis, Montréal, Québec, Canada H2X 3J4; Tel: 1-514-281-2100; Fax: 1-514-281-2443. 1997 The British Society for Antimicrobial Chemotherapy 707

C. Gaudreau and H. Gilbert Antimicrobial agents For the agar dilution method, the following antibiotics were tested: ampicillin, ciprofloxacin, erythromycin and tetracycline (Sigma Chemical Co., St Louis, MO, USA). For the disc diffusion testing, nalidixic acid (30 g) (Oxoid, Unipath Ltd, Basingstoke, UK), ampicillin (10 g), ciprofloxacin (5 g), erythromycin (15 g) and tetracycline (30 g) (BBL, Becton Dickinson and Co., Cockeysville, MD, USA) discs were used. Antibiotic susceptibility testing The agar dilution method was performed as described previously, 6 the concentrations of the antibiotics tested were 0.06 256 mg/l and a control plate without antibiotic was inoculated at the end of the procedure. Inocula were prepared in Mueller Hinton broth (BBL Microbiology Systems) at a density adjusted to a 0.5 McFarland turbidity standard for the disc diffusion and diluted 1:10 for the agar dilution. With a Steers Cathra 3 mm, a final inoculum of about 10 4 cfu was placed on Mueller Hinton agar plates (Difco Laboratories, Detroit, MI, USA). The same medium was used for the disc diffusion testing. The inoculated plates were incubated at 35 C under a microaerophilic atmosphere obtained with a gas generator envelope (Difco), for 48 h. For the agar dilution, the endpoint was taken as the complete inhibition of macroscopic growth; for the disc diffusion, the zone diameters were measured with slipping callipers. The following control strains were used: Staphylococcus aureus ATCC 25923, S. aureus ATCC 29213 and Escherichia coli ATCC 25922. Susceptibility criteria for the agar dilution method and the drug concentrations in the discs were those of the National Committee for Clinical Laboratory Standards. 7 The breakpoints used for the disc diffusion method are those mentioned in the Results section. Results The results of susceptibility testing using the disc diffusion and the agar dilution methods for each antibiotic are reported in the Table. The percentages of true results and the 95% CI for disc diffusion were 97.6 100% and 93.2 100% respectively for susceptible strains to the five antibiotics tested. The percentages of true results and 95% CI for resistant Campylobacter spp. were 100% (77.2 100%) for ciprofloxacin, 86.7% (62.1 96.3%) for nalidixic acid, 100% (93.2 100%) for tetracycline, 100% (56.6 100%) for erythromycin and 68.8% (44.4 85.8%) for ampicillin. One strain of C. jejuni, highly resistant to erythromycin, was also resistant to ampicillin and tetracycline. Seven of the 13 strains (53.8%) resistant to and 46 of 129 strains (35.7%) susceptible to ciprofloxacin were resistant to tetracycline (P 0.23; Fisher s exact two-tailed test). There was complete agreement between disc diffusion and the MICs of ciprofloxacin; this correlation is shown in Figure 1. The 132 ciprofloxacin-susceptible strains (MIC 1 mg/l) had zone diameters of 28.2 mm, while the 13 resistant strains (MIC 4 mg/l) had zones of 15.2 mm. The correlation coefficient was 0.61. The correlation of disc diffusion of nalidixic acid and the MICs of ciprofloxacin is shown in Figure 2. Of the 132 ciprofloxacin-susceptible strains, 129 had zone diameters of 20.8 mm and one of 12 mm around the nalidixic acid disc. The 13 ciprofloxacin-resistant strains had no zone around this disc. Two susceptible strains had no zone for a major error of 1.4% to nalidixic acid disc. The correlation coefficient was 0.52. The correlation of the two susceptibility methods for tetracycline is shown in Figure 3. The 89 tetracycline-susceptible and 53 tetracycline-resistant Campylobacter spp. had zone diameters of 29 mm and 17 mm for complete agreement between both methods. The correlation coefficient was 0.78. Table. Results of susceptibility testing of 145 strains of C. jejuni and C. coli by disc diffusion and agar dilution methods for five antibiotics Ciprofloxacin Tetracycline a Erythromycin Ampicillin Agar dilution 132 S (MIC 1 mg/l) 89 S (MIC 4 mg/l) 139 NR (MIC 4 mg/l) 126 S (MIC 8 mg/l) method 13 R (MIC 4 mg/l) 53 R (MIC 16 mg/l) 6 R (MIC 8 mg/l) 8 I (MIC 16 mg/l) 11 R (MIC 32 mg/l) Disc diffusion 132 S ( 27 mm) b 89 S ( 29 mm) f 140 S ( 20 mm) h 125 S ( 14 mm) j method 13 R ( 16 mm) c 53 R ( 17 mm) g 5 R ( 20 mm) i 4 I ( 6 14 mm) 16 R (6 mm) k Disc diffusion 130 S ( 6 mm) d (nalidixic acid) 15 R (6 mm) e S, susceptible; I, intermediate; R, resistant; NR, non-resistant strains. a 142 strains tested. Percentages of true result and 95% Cl: b 100% (97.2 100%); c 100% (77.2 100%); d 100% (97.1 100%); e 86.7% (62.1 96.3%); f 100% (95.9 100%); g 100% (93.2 100%); h 99.3% (96.1 99.9%); i 100% (56.6 100%); j 97.6% (93.2 99.2%); k 68.8% (44.4 85.8%). 708

Disc diffusion of Campylobacter spp. Figure 1. Correlation between disc diffusion and MICs for ciprofloxacin (r 0.61). 0.25 0.12 0.06 Figure 2. Correlation between disc diffusion for nalidixic acid and MICs of ciprofloxacin (r 0.52). 709

C. Gaudreau and H. Gilbert 0.25 0.12 0.06 Figure 3. Correlation between disc diffusion and MICs for tetracycline (r 0.78). Figure 4. Correlation between disc diffusion and MICs for erythromycin (r 0.66). 710

Disc diffusion of Campylobacter spp. 0.25 0.12 0.06 Figure 5. Correlation between disc diffusion and MICs for ampicillin (r 0.57). The correlation of disc diffusion and agar dilution for erythromycin is shown in Figure 4. The non-resistant strains (MIC 4 mg/l) had zone diameters of 20 mm. Four strains with high-level resistance (MIC 128 mg/l) had no zone around the erythromycin disc. Two strains with an MIC of 8 mg/l had zone diameters of 19.5 and 22.4 mm. There was a very major error of 0.7% if 20 mm was used as the criterion of non-resistance and the correlation coefficient was 0.66. The correlation of the two susceptibility methods for ampicillin is shown in Figure 5. Out of the 126 susceptible, 122 strains and three with an intermediate MIC had zone diameters of 14 mm around the ampicillin disc. Two intermediate and two susceptible Campylobacter spp. had zones of 6 but 14 mm. The 11 resistant, three intermediate and two susceptible strains had no zone around this disc. There were 5.5% minor and 1.4% major errors for disc diffusion for this antibiotic. The correlation coefficient was 0.57. Discussion In C. jejuni and C. coli, resistance to ciprofloxacin has increased from 0% to 2 50% in the last decade. 2,4,8 The MICs for susceptible C. jejuni and C. coli strains are 10 30 times higher than the MICs for susceptible Enterobacteriaceae and the DNA gyrase of these bacteria is intrinsically more resistant. 9 The MICs for resistant C. jejuni and C. coli strains are 8 256 times higher than those for susceptible strains. 8,9 Resistance to the new quinolones is secondary to a single mutation causing a change in the A subunit of DNA gyrase. 9,10 Microbiological failure with or without clinical failure as a consequence of the emergence of a resistant C. jejuni in patients treated with ciprofloxacin has been reported. 8 Nalidixic acid susceptibility is a marker for C. jejuni and C. coli ciprofloxacin susceptibility: the nalidixic acid-susceptible strains are susceptible to ciprofloxacin while most of the resistant ones are resistant to ciprofloxacin. 2,3 The correlation between MICs and disc diffusion for ciprofloxacin was studied by Endtz et al.: 2 21 susceptible and 21 resistant C. jejuni or C. coli strains (MIC 1 mg/l and 4 mg/l respectively) had zone diameters of 26 and 17 mm respectively. If these strains were added to ours, the 95% CI for ciprofloxacin-resistant strains would be 89.8 100%. Between 0 and 60% of C. jejuni and C. coli strains are reported to be resistant to tetracycline. 4,11,12 Resistance to this antibiotic in these bacteria is mainly plasmid-mediated, 4 a fact that might explain the complete agreement between disc diffusion and MICs for this antibiotic with 95% CI of 93% in our study. The plasmid-carrying strains are resistant to tetracycline and those without plasmids are susceptible. Tetracycline is considered an alternative 711

C. Gaudreau and H. Gilbert treatment for diarrhoea caused by these Campylobacter spp. Resistance to erythromycin has been reported in 0 12.6% of C. jejuni 3,4,11 and in 10 91% of C. coli. 3 5 As reported by others, 3,5,13 more than 40% of C. jejuni and C. coli strains have intermediate MICs if the criteria of NCCLS are used, namely susceptibility, MIC 0.5 mg/l; resistance, MIC 8 mg/l. The amount of oxygen in the atmosphere might have an adverse effect on erythromycin and falsely increase the MICs. 3,12 The breakpoint for resistance to erythromycin might have to be changed. 12 The strains with an MIC of 8 mg/l might not be truly resistant. 3 The efficacy of erythromycin in the treatment of C. jejuni enterocolitis has been demonstrated in clinical studies. 1,14 -Lactamase is found in 83 92% of C. jejuni 6 and in 68 93.5% of C. coli. 5 The -lactamase-positive strains of Campylobacter spp. are significantly less susceptible to ampicillin, amoxycillin and ticarcillin even if the MICs produced are not high enough to be regarded as resistant. 5,6 Ampicillin is not considered useful for the treatment of C. jejuni and C. coli infections. 1,14 Vanhoof et al. 15 reported good correlation between disc diffusion and agar dilution for susceptibility testing of C. jejuni to many antibiotics. However, they used discs with drug concentrations not usually recommended by NCCLS. 12 In this study, we found the disc diffusion to be a reliable, easy and inexpensive method for testing the susceptibility of C. jejuni and C. coli to erythromycin, tetracycline and ciprofloxacin. The price for each disc is about US$0.03 compared with US$2.10 for one Etest antimicrobial strip. Until more erythromycin- and ciprofloxacinresistant strains are tested to confirm the reliability of this test, the resistance to these two drugs needs to be confirmed using the Etest 13 or the agar dilution method. 1 Acknowledgements We thank Brigite Chevrier and Angela Gurd for secretarial services and Jean Vincelette for assistance in statistical data analysis. References 1. Nachamkin, I. (1995). Campylobacter and Arcobacter. In Manual of Clinical Microbiology, 6th edn (Murray, P. R., Baron, E. J., Pfaller, M. A., Tenover, F. C. & Yolken, R. H., Eds), pp. 483 91. American Society for Microbiology, Washington, DC. 2. Endtz, H. P., Ruijs, G. J., van Klingeren, B., Jansen, W. H., van der Reyden, T. & Mouton, R. P. (1991). Quinolone resistance in Campylobacter isolated from man and poultry following the introduction of fluoroquinolones in veterinary medicine. Journal of Antimicrobial Chemotherapy 27, 199 208. 3. Rautelin, H., Renkonen, O. V. & Kosunen, T. U. (1991). Emergence of fluoroquinolone resistance in Campylobacter jejuni and Campylobacter coli in subjects from Finland. Antimicrobial Agents and Chemotherapy 35, 2065 9. 4. Reina, J., Ros, M. J. & Serra, A. (1994). Susceptibilities to 10 antimicrobial agents of 1,220 Campylobacter strains isolated from 1987 to 1993 from feces of pediatric patients. Antimicrobial Agents and Chemotherapy 38, 2917 20. 5. Lachance, N., Gaudreau, C., Lamothe, F. & Turgeon, F. (1993). Susceptibilities of -lactamase-positive and -negative strains of Campylobacter coli to -lactam agents. Antimicrobial Agents and Chemotherapy 37, 1174 6. 6. Lachance, N., Gaudreau, C., Lamothe, F. & Larivière, L. A. (1991). Role of the -lactamase of Campylobacter jejuni in resistance to -lactam agents. Antimicrobial Agents and Chemotherapy 35, 813 8. 7. National Committee for Clinical Laboratory Standards. (1993). Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically Third Edition; Approved Standard M7-A3. NCCLS, Villanova, PA. 8. Segreti, J., Gootz, T. D., Goodman, L. J., Parkhurst, G. W., Quinn, J. P., Martin, B. A. et al. (1992). High-level quinolone resistance in clinical isolates of Campylobacter jejuni. Journal of Infectious Diseases 165, 667 70. 9. Gootz, T. D. & Martin, B. A. (1991). Characterization of highlevel quinolone resistance in Campylobacter jejuni. Antimicrobial Agents and Chemotherapy 35, 840 5. 10. Wang, Y., Huang, W. M. & Taylor, D. E. (1993). Cloning and nucleotide sequence of the Campylobacter jejuni gyra gene and characterization of quinolone resistance mutations. Antimicrobial Agents and Chemotherapy 37, 457 63. 11. Larivière, L. A., Gaudreau, C. L. & Turgeon, F. F. (1986). Susceptibility of clinical isolates of Campylobacter jejuni to twenty-five antimicrobial agents. Journal of Antimicrobial Chemotherapy 18, 681 5. 12. Tenover, F. C., Baker, C. N., Fennell, C. L. & Ryan, C. A. (1992). Antimicrobial resistance in Campylobacter species. In Campylobacter jejuni: Current Status and Future Trends (Nachamkin, I., Blaser, M. J. & Tompkins, L. S., Eds), pp. 66 73. American Society for Microbiology, Washington, DC. 13. Huang, M. B., Baker, C. N., Banerjee, S. & Tenover, F. C. (1992). Accuracy of the E test for determining antimicrobial susceptibilities of staphylococci, enterococci, Campylobacter jejuni, and Gram-negative bacteria resistant to antimicrobial agents. Jour - nal of Clinical Microbiology 30, 3243 8. 14. Blaser, M. J. (1995). Campylobacter and related species. In Mandell, Douglas and Bennett s Principles and Practice in Infec - tious Diseases, 4th edn (Mandell, G. L., Bennett, J. E. & Dolin, R., Eds), pp. 1948 56. Churchill Livingstone, New York, NY. 15. Vanhoof, R., Dierickx, R., Coignau, H., Hubrechts, J. M. H., Kaufman, L. & Butzler, J. P. (1985). Disk sensitivity testing for Campylobacter jejuni. European Journal of Clinical Microbiology 3, 160 2. Received 28 August 1996; returned 21 October 1996; revised 7 November 1996; accepted 14 January 1997 712