Journal of Antimicrobial Chemotherapy (2002) 49, 55 59 JAC Molecular epidemiology and evolution of resistance to quinolones in Escherichia coli after prolonged administration of ciprofloxacin in patients with prostatitis Juan P. Horcajada a *, Jordi Vila b, Antonio Moreno-Martínez a, Joaquim Ruiz b, Jose Antonio Martínez a, Miquel Sánchez c, Eladio Soriano a and Josep Mensa a Departments of a Infectious Diseases and b Microbiology, Institut Clínic d Infeccions i Immunologia (ICII) and c Department of Internal Medicine, Hospital Clínic Universitari, Villarroel 170, 08036 Barcelona, Spain The emergence and evolution of quinolone-resistant Escherichia coli in faeces of patients with prostatitis treated with high-dose oral ciprofloxacin for 1 month were studied. In 11 of 23 patients, from whom only quinolone-susceptible E. coli was isolated before treatment, quinolone-resistant strains, genetically distinct from the quinolone-susceptible ones, predominated during and just after therapy. Two months after treatment, these were completely displaced by quinolone-susceptible E. coli, genetically distinct from the E. coli isolated before and during therapy. Hence, during ciprofloxacin therapy, half of the patients were transiently colonized with new, quinolone-resistant strains of E. coli. Introduction The wide use of fluoroquinolones in both human and animal medicine has led to a significant increase in the level of resistance to these antimicrobial agents in clinical isolates of Escherichia coli from several geographical areas. 1 4 Patients receiving fluoroquinolones for long periods, either to prevent recurrent urinary tract infections or for selective intestinal decontamination, may be colonized by quinolone-resistant Enterobacteriaceae in the intestinal tract. 5 7 However, neither the origin of these resistant strains nor the evolution of intestinal colonization after the selective pressure exerted by the quinolones is removed has been studied. The resistant strains could be newly acquired exogenous resistant strains or resistant mutants arising from strains already in the intestinal tract. The selection of quinolone-resistant mutants correlates with ratios of the selective concentration of quinolone and MIC 8 mg/l. The use of fluoroquinolones with little intrinsic activity, the prescription of relatively low doses, such as in the course of prophylaxis of urinary tract infections, and the partial inactivation of fluoroquinolones in stools, are conditions that determine concentration/mic ratio and, therefore, favour the selection of resistant mutants. In contrast, selection of resistant mutants is likely to be less frequent if the patient is prescribed ciprofloxacin at therapeutic doses. 8 The main objectives of this study were: (i) to analyse the evolution of the susceptibility to fluoroquinolones of E. coli strains isolated from faecal flora during and after the administration of therapeutic doses of oral ciprofloxacin for 1 month; (ii) to investigate whether there is any relationship between the resistant strains isolated during treatment and susceptible strains in the gut before treatment. Materials and methods Patients The study involved 29 outpatients with acute prostatitis for whom the initial faecal sample was positive for E. coli and who attended the Infectious Disease Department of the Hospital Clínic (Barcelona, Spain), a 900 bed tertiary hospital. These patients were treated orally with ciprofloxacin 750 mg twice daily for 28 days. Compliance with prescribed treatment was monitored by means of a pill count of returned boxes of antibiotic and by interview. None of the patients had received antibiotics for at least 4 weeks before the study commenced. *Corresponding author. Tel: 34-93-2275586; Fax: 34-93-4514438; E-mail: jhorcaja@clinic.ub.es 2002 The British Society for Antimicrobial Chemotherapy 55
J. P. Horcajada et al. Microbiological methods Stool cultures were collected immediately before therapy commenced and every 2 weeks during and after treatment, until a predominant quinolone-susceptible E. coli strain was isolated. Fresh stools were collected in sterile containers and cultured immediately. One gram of faeces was homogenized in 9 ml of saline serum and plated by streaking 0.05 ml on to MacConkey agar. After inoculation, plates were incubated aerobically at 35 37 C for 2 days. To recover the predominant E. coli strains, all morphologically different colonies suspected to be E. coli were subcultured in the same medium and identified by conventional methods. E. coli isolates were then stored in skimmed milk at 70 C, until antibiotic susceptibility tests and molecular studies were performed. Antimicrobial susceptibility tests Antibiotic susceptibility testing was performed by the Etest method (AB Biodisk, Solna, Sweden) following the manufacturer s instructions. E. coli ATCC 25922 was used as the quality control organism. DNA sequencer (Applied Biosystems 377; Perkin-Elmer, Foster City, CA, USA). When the PCR product from the QRDR of the gyra gene was analysed by digestion with HinfI, 20 L of PCR mixture were incubated for 2.5 h with 10 U of restriction endonuclease. Digestion products were separated by electrophoresis in NuSieve (1.5% w/v) and agarose (1% w/v) gels. Results Before ciprofloxacin administration, the predominant E. coli strains in the faecal flora were susceptible to quinolones in 23 (79%) of 29 patients studied (patients 1 23 in Table). In the remaining six patients (21%), the initial strains were resistant to nalidixic acid (patients 24 29 in the Table and the Figure) and four of these were also resistant to ciprofloxacin (patients 24 27). For 12 (52%) of the 23 patients with quinolone-susceptible E. coli strains at the start of treatment, no growth of E. coli was observed from stool cultures made during treatment, whereas for each of the remaining 11 patients (48%) a quinolone-resistant strain was isolated during treatment Epidemiological analysis The epidemiological relationships among the E. coli strains isolated at different times from each patient were determined by repetitive extragenic palindromic PCR (REP- PCR). For REP-PCR, bacteria were grown on MacConkey agar overnight and one colony of each isolate was suspended in 25 L of reaction mixture containing 20 mm Tris HCl ph 8.8, 100 mm KCl, 3.0 mm MgCl 2, gelatin 0.1% (v/v) and 400 M dntps. Primer was added at 1 M, together with 2.5 U Taq polymerase. The reaction mixture was overlaid with mineral oil and DNA amplification was achieved with the following programme: 30 cycles of 94 C for 1 min, 40 C for 1 min and 65 C for 8 min, with a single final extension at 65 C for 16 min. Samples (10 L) of each PCR end-product were analysed on agarose (1.5% w/v) gels. The primer used was 5 -GCGCCGICATGCGGC- ATT-3. Detection of mutations in the gyra gene To study the mechanisms of quinolone resistance in the strains isolated, PCR amplification and DNA sequencing of the quinolone resistance determining region (QRDR) of the gyra gene were performed as described elsewhere. 9 PCRs were carried out in a DNA Thermal Cycler 480 (Perkin-Elmer Cetus, Emeryville, CA, USA). Primers and free nucleotides were removed with a QiaQuick spin PCR purification kit (Qiagen, Inc., Chatsworth, CA, USA), according to the manufacturer s instructions; the sample was directly processed for DNA sequencing with Thermo Sequenase II Dye Terminator cycle sequencing kit (Amersham, Cleveland, OH, USA) and analysed in an automatic Figure. REP-PCR patterns. Patient 24: lane 1, quinoloneresistant E. coli before treatment; lane 2, quinolone-resistant E. coli with the same REP-PCR pattern during treatment; lane 3, quinolone-susceptible E. coli with different REP-PCR pattern after treatment. Patient 13: lane 1, quinolone-susceptible E. coli before treatment; lane 2, quinolone-resistant E. coli with different REP-PCR pattern during treatment; lane 3, quinolone-susceptible E. coli with different REP-PCR pattern after treatment. Patient 1: lane 1, quinolone-susceptible E. coli before treatment; lane 2, quinolone-susceptible E. coli with different REP-PCR pattern after treatment. Lane M, DNA molecular weight markers (Gibco-BRL). 56
Epidemiology of E. coli quinolone resistance Table. Evolution of susceptibility to quinolones and molecular epidemiology of the E. coli isolates Before treatment During treatment a 1st month after treatment 2nd month after treatment Patient no. MIC NAL MIC CIP MIC NAL MIC CIP clonal identity b MIC NAL MIC CIP clonal identity b MIC NAL MIC CIP clonal identity b 1 2 0.008 NBG NBG NBG 2 0.004 no 2 1.5 0.003 NBG NBG NBG 1.5 0.003 no 3 2 0.008 NBG NBG NBG NBG NBG NBG 3 0.008 no 4 2 0.008 NBG NBG NBG NBG NBG NBG 2 0.008 no 5 3 0.008 NBG NBG NBG NBG NBG NBG 3 0.006 no 6 3 0.008 NBG NBG NBG NBG NBG NBG 3 0.008 no 7 1.5 0.006 NBG NBG NBG NBG NBG NBG 2 0.008 no 8 1.5 0.008 NBG NBG NBG NBG NBG NBG 2 0.008 no 9 3 0.008 NBG NBG NBG NBG NBG NBG 1 0.006 no 10 2 0.008 NBG NBG NBG NBG NBG NBG 2 0.012 no 11 2 0.006 NBG NBG NBG NBG NBG NBG 1.5 0.003 no 12 3 0.006 NBG NBG NBG NBG NBG NBG 1 0.003 no 13 3 0.006 >256 >32 no 3 0.006 no 14 3 0.008 >256 >32 no 2 0.006 no 15 3 0.006 >256 >32 no >256 3 no 3 0.008 no 16 2 0.002 >256 >32 no >256 >32 no 1.5 0.002 no 17 2 0.19 NBG NBG NBG >256 >32 no 3 0.012 no 18 2 0.008 NBG NBG NBG >256 >32 no 2 0.018 no 19 2 0.008 NBG NBG NBG >256 0.125 no 4 0.012 no 20 2 0.008 NBG NBG NBG >256 >32 no 1.5 0.008 no 21 3 0.006 NBG NBG NBG >256 >32 no 2 0.006 no 22 2 0.003 NBG NBG NBG >256 0.125 no 2 0.012 no 23 4 0.012 >256 8 no NBG NBG NBG 4 0.006 no 24 >256 >32 >256 >32 yes 3 0.004 no 25 >256 >32 >256 >32 yes 2 0.008 no 26 >256 >32 NBG NBG NBG >256 >32 yes 3 0.008 no 27 >256 >32 >256 >32 yes >256 >32 yes 2 0.008 no 28 >256 0.19 >256 >32 no NBG NBG NBG 2 0.006 no 29 >256 0.125 NBG NBG NBG >256 0.50 no 3 0.008 no MIC NAL, MIC of nalidixic acid (mg/l). MIC CIP, MIC of ciprofloxacin (mg/l). NBG, no bacterial growth. a Ciprofloxacin 750 mg bd po for 28 days. b Clonal identity with the strain isolated immediately before. 57
J. P. Horcajada et al. or from the first stool culture performed 15 days after treatment finished (patients 13 23 in the Table and the Figure). The MICs of nalidixic acid for these 11 strains were 256 mg/l and the MICs of ciprofloxacin were 32 mg/l in eight (72%) of the 11 strains. All the quinolone-resistant strains were genetically different from the quinolonesusceptible strains recovered before starting quinolone treatment. The four patients whose strains were initially resistant to nalidixic acid and ciprofloxacin harboured the same strains during treatment (patients 24 27 in the Table and the Figure). In the two patients whose strains at the start of the study were resistant only to nalidixc acid, quinoloneresistant but genetically different strains were isolated during or just after treatment (patients 28 and 29 in the Table). Within 2 months of the end of quinolone treatment, the predominant E. coli strains in the faecal flora of all patients were susceptible to quinolones (nalidixic acid MIC 4 mg/l, ciprofloxacin MIC 0.02 mg/l) and genetically different from both the quinolone-susceptible strains isolated before treatment and from the quinolone-resistant strains isolated during treatment (Table and Figure). All the quinolone-resistant strains showed at least one mutation, specifically in the triplet encoding amino acid residue Ser-83 of GyrA. Discussion The study illustrates the dynamics of the predominant E. coli faecal flora in 29 males with prostatitis given highdose oral ciprofloxacin therapy. Before treatment the predominant E. coli strains were susceptible to nalidixic acid and ciprofloxacin in 23 (79%) and 26 (90%) patients, respectively. During or shortly after therapy of these patients, the dominant quinolone-susceptible E. coli strains were no longer detected; in eight cases they were substituted by genetically unrelated strains resistant to ciprofloxacin; in three patients new strains, resistant to nalidixic acid but susceptible to ciprofloxacin, emerged. In the remaining four patients where the dominant pretherapy E. coli strains were resistant to ciprofloxacin, the same organisms persisted during therapy and for a short time thereafter. Finally, after ciprofloxacin therapy ceased, all patients acquired new quinolone-susceptible E. coli that were genetically unrelated to previous isolates, as the dominant strains. Since our method only detected the dominant faecal E. coli strain in each patient, it can only be asserted with certainty that high-dose oral ciprofloxacin did not select resistant mutants from the susceptible strain detected initially. The origin of the resistant E. coli that emerged during treatment cannot be determined from the current data. Three scenarios are conceivable: (i) unmasking of ciprofloxacin-resistant organisms present in low numbers before therapy but undetected by the methodology; (ii) acquisition of an exogenous resistant strain during treatment; and (iii) selection in the host of highly resistant mutants from pre-existing intermediately susceptible organisms. In the particular setting, either of the first two possibilities is likely, considering that the use of quinolones in the community is widespread and that resistant strains selected in food animals exposed to quinolones may have been introduced into the food chain. 2 4 We included in the study only patients who had not taken antibiotics during the month before taking the first sample, but we did not investigate the issue further. The presence of high-level quinolone-resistant E. coli strains in four of our patients at baseline could indicate previous exposure to quinolones. High rates of faecal carriage of quinolone-resistant E. coli strains in healthy people in our community and an extremely high prevalence of quinolone-resistant E. coli in the stools of farm animals in our geographical area have been reported. 4 These data are consistent with the hypothesis that the resistant E. coli strains found in our patients were either already carried in low numbers before therapy or were acquired during ciprofloxacin administration. The possibility that high-level resistance was selected by ciprofloxacin from pre-existing intermediate-susceptible strains seems less likely, since in two patients, where the predominant strain was one with a single Ser-83 mutation before treatment, high-dose oral ciprofloxacin therapy simply suppressed its growth without selecting mutants with increased resistance. The transient nature of the faecal carriage of high-level quinolone-resistant E. coli strains after ciprofloxacin therapy has been reported in neutropenic patients, but no information was given about the molecular epidemiology of the successive isolates. 10 It seems probable that susceptible E. coli strains, in the absence of selective antibiotic pressure, are more fit than resistant ones, and replace them. With our methodology we could not discern whether the resistant strains disappeared completely, or remained at low density together with the quinolone-susceptible strains. Studies are in progress to clarify this issue. Acknowledgements The competent assistance of Victoria Lasheras is gratefully acknowledged. This work was supported, in part, by grant SAF 97/0091 from Plan Nacional I D, Spain and Beca Agustí Pumarola from the Societat Catalana de Malalties Infeccioses. References 1. Cometta, A., Calandra, T., Bille, J. & Glauser, P. (1994). Escherichia coli resistant to fluoroquinolones in patients with cancer and neutropenia. New England Journal of Medicine 330, 1240 1. 2. Ena, J., Lopez-Perezagua, M. M., Martinez-Peinado, C., Cia- Barrio, M. A. & Ruiz-Lopez, I. (1998). Emergence of ciprofloxacin resistance in Escherichia coli isolates after widespread use of fluoro- 58
Epidemiology of E. coli quinolone resistance quinolones. Diagnostic Microbiology and Infectious Disease 30, 103 7. 3. Blanco, J. E., Blanco, M., Mora, A. & Blanco, J. (1997). Prevalence of bacterial resistance to quinolones and other antimicrobials among avian Escherichia coli strains isolated from septicemic and healthy chickens in Spain. Journal of Clinical Microbiology 35, 2184 5. 4. Garau, J., Xercavins, M., Rodríguez-Carballeira, M., Gómez- Vera, J. R., Coll, I., Vidal, D. et al. (1999). Emergence and dissemination of quinolone-resistant Escherichia coli in the community. Antimicrobial Agents and Chemotherapy 43, 2736 41. 5. Carratala, J., Fernandez-Sevilla, A., Tubau, F., Dominguez, M. A. & Gudiol, F. (1996). Emergence of fluoroquinolone-resistant Escherichia coli in faecal flora of cancer patients receiving norfloxacin prophylaxis. Antimicrobial Agents and Chemotherapy 40, 503 5. 6. Dupeyron, C., Mangeney, N., Sedrati, L., Campillo, B., Fouet, P. & Leluan, G. (1994). Rapid emergence of quinolone resistance in cirrhotic patients treated with norfloxacin to prevent spontaneous bacterial peritonitis. Antimicrobial Agents and Chemotherapy 38, 340 4. 7. Novella, M., Sola, R., Soriano, G., Andreu, M., Gana, J., Ortiz, J. et al. (1997). Continuous versus inpatient prophylaxis of the first episode of spontaneous bacterial peritonitis with norfloxacin. Hepatology 25, 532 6. 8. Daikos, G. L., Kathpalia, S. B., Sharifi, R., Lolans, V. T. & Jackson, G. G. (1997). Comparison of ciprofloxacin and beta-lactam antibiotics in the treatment of urinary tract infections and alteration of faecal flora. American Journal of Medicine 82, Suppl. 4A, 290 4. 9. Tavío, M. M., Vila, J., Ruiz, J., Ruiz, J., Martín-Sánchez, A. M. & Jiménez de Anta, M. T. (1999). Mechanisms involved in the development of resistance to fluoroquinolones in Escherichia coli strains. Journal of Antimicrobial Chemotherapy 44, 735 42. 10. Perea, S., Hidalgo, M., Arcediano A., Ramos, M. J., Gomez, C., Hornedo, J. et al. (1999). Incidence and clinical impact of fluoroquinolone-resistant Escherichia coli in the faecal flora of cancer patients treated with high dose chemotherapy and ciprofloxacin prophylaxis. Journal of Antimicrobial Chemotherapy 44, 117 20. Received 20 November 2000; returned 20 March 2001; revised 15 May 2001; accepted 4 September 2001 59