Do the quinolones still constitute valid empirical therapy for community-acquired urinary tract infections in Spain?

Abstract

Urinary tract infections (UTIs) are among the most frequent bacterial infections in humans. Escherichia coli is, by far, the bacterium most commonly isolated in community-acquired UTI, varying according to series from 70% to 90%. Because of the high percentage of resistance to ampicillin and co-trimoxazole—approximately 60% and 30% respectively [1Alós JI Balas D Gómez-Garcés JL Grupo de Estudio de Infecciones en Atención Primaria Prevalencia de susceptibilidad a quinolonas y otros antibióticos en microorganismos aislados de bacteriurias extrahospitalarias en Madrid en 1995.Rev Clin Esp. 1997; 197: 167-171PubMed Google Scholar]—found in the strains of E. coli isolated in our area, the fluoroquinolones have become one of the first-choice therapies in the empirical treatment of UTI in Spain. Resistance to the quinolones in E. coli strains isolated in urine samples from out-patients has increased in Spain in recent years [1Alós JI Balas D Gómez-Garcés JL Grupo de Estudio de Infecciones en Atención Primaria Prevalencia de susceptibilidad a quinolonas y otros antibióticos en microorganismos aislados de bacteriurias extrahospitalarias en Madrid en 1995.Rev Clin Esp. 1997; 197: 167-171PubMed Google Scholar, 2Pérez-Trallero E Urbieta M Jimenez D García-Arenzana JM Cilla G Ten-year survey of quinolone resistance in Escherichia coli causing urinary tract infections.Eur J Clin Microbiol Infect Dis. 1993; 12: 349-351Crossref PubMed Scopus (60) Google Scholar, 3Aguiar JM Chacon J Cantón R Baquero F The emergence of highly fluoroquinolone-resistant Escherichia coli in community-acquired urinary tract infections.J Antimicrob Chemother. 1992; 29: 349-350Crossref PubMed Scopus (81) Google Scholar], as is beginning to occur in other countries [4Weber G Riesenberg K Schlaeffer F Peled N Bores A Yagupsky P Changing trends in frequency and antimicrobial resistance of urinary pathogens in outpatient clinics and a hospital in southern Israel, 1991–1995.Eur J Clin Microbiol Infect Dis. 1997; 16: 834-838Crossref PubMed Scopus (51) Google Scholar]. As an example, our data for 1997, similar to those of other laboratories in our country, show 27.3% resistance to nalidixic acid and 16.0% to the fluoroquinolones in 2581 strains tested. In outpatients, many UTIs are treated empirically [5Carlson KJ Mulley AG Management of acute dysuria: a decision analysis model of alternative strategies.Ann Intern Med. 1985; 102: 244-249Crossref PubMed Scopus (62) Google Scholar] and, with the latest data, many physicians would question the use of the quinolones as empirical therapy. Nevertheless, the published susceptibility data are provided by microbiology laboratories, and should be interpreted bearing in mind the fact that many of the uncomplicated UTIs in outpatients are treated empirically so that neither the bacterium implicated nor their antibiotic susceptibility are ever known. In addition, almost all of the strains isolated in patients with complicated UTI, with reinfections or relapses, are isolated and their antibiotic susceptibilities are tested, since in these cases it is usual to request a urine culture. These strains are not necessarily representative of those found in the community. In our opinion, this would result in a distortion of the real facts concerning quinolone resistance in E. coli. In the Emergency Service of our hospital, nearly all UTIs treated are community acquired, and, in contrast with what occurs in the community, a urine culture is performed on the slightest suspicion of UTI. Also, in the Emergency Service we find a greater number of community-acquired complicated UTIs than are found in the general practitioner's consulting rooms, which permits a statistical comparison of antibiotic susceptibility according to the type of UTI (complicated versus uncomplicated). E. coli accounts for the 78.3% of isolates from urine in this service during 1997. We studied the susceptibility of the E. coli isolated in patients with suspected UTI treated in the Emergency Service of our hospital and compared the pattern of resistance to quinolones and other antibiotics according to the type of UTI (complicated versus uncomplicated). All the isolates of E. coli obtained from urine cultures taken in the Emergency Service of our hospital on random days in 1997 and 1998 were studied. We obtained 125 strains from the same number of patients, all with community-acquired UTI. Clinical histories were reviewed and a protocol was completed for each patient. Reflected in the protocol, among other data, were: underlying disease, urinary tract pathology, prior episodes of UTI during the preceding year, signs and symptoms and the results of the urine analysis. According to the data obtained and following established criteria [6Stamm WE Hooton TM Management of urinary tract infection in adults.N Engl J Med. 1993; 329: 1328-1334Crossref PubMed Scopus (686) Google Scholar] it was decided in each case whether UTI was present or not and, if present, the form of presentation (acute uncomplicated cystitis, acute uncomplicated pyelonephritis, recurrent UTI or complicated UTI). When the patients had a functionally, metabolically or anatomically abnormal urinary tract, or they were immunocompromised, complicated UTI was considered [6Stamm WE Hooton TM Management of urinary tract infection in adults.N Engl J Med. 1993; 329: 1328-1334Crossref PubMed Scopus (686) Google Scholar]. The antibiotic susceptibility of the E. coli strains was ascertained by broth microdilution with PASCO (Difco, Detroit, MI, USA). The results were interpreted according to NCCLS criteria [7National Committee for Clinical Laboratory Standards Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically. Approved standard M7-A4. NCCLS, Villanova, Pa1997Google Scholar]. To compare antibiotic resistance in E. coli from complicated UTI and in E. coli from uncomplicated UTI, the two-tailed Fisher exact test was used. Of the 125 E. coli urine cultures, 69 were from women and 56 from men. All isolates obtained in men corresponded with complicated UTI. UTI could not be diagnosed in two of the women, and asymptomatic bacteriuria was considered. In the rest of the women, the UTI was classified as acute uncomplicated cystitis in 25 cases, acute uncomplicated pyelonephritis in 22, complicated UTI in 18 and recurrent UTI in two. Globally, there were 74 (59.2%) complicated UTIs, 49 (39.2%) uncomplicated UTIs and two asymptomatic cases of bacteriuria. Of the 74 patients with complicated UTI, 16 (26.1%) had an anatomically abnormal urinary tract, 14 (18.9%) had renouretheral stones, 10 (13.5%) were immunocompromised, six (8.1%) had diabetes mellitus, six (8.1%) had catheter-associated UTI, and six (8.1%) had been subjected to urinary tract surgery. In the remaining 16 (21.6%), no history of anatomic or structural anomalies was found and nor was there any underlying disease, but, since the patients were males, complicated UTI was considered. Of the 123 strains of E. coli causing UTI (excluding the asymptomatic bacteriuria), 20 (16.26%) were resistant to nalidixic acid and 16 (13%) to norfloxacin and ciprofloxacin. However, of the 74 E. coli strains causing complicated UTI, 18 (24.32%) were resistant to nalidixic acid, while only two (4.08%) of the 49 causing uncomplicated UTI were resistant (p=0.0025, Fisher exact test, two-tailed). Of the 74 E. coli strains isolated in complicated UTI, 16 (21.62%) were resistant to norfloxacin and ciprofloxacin, while none of those isolated in uncomplicated UTI was resistant to these antibiotics (p=0.00018, Fisher exact test, two-tailed). Resistances to ampicillin, cefazolin, gentamicin, co-trimoxazole and fosfomycin were also compared. There were no statistically significant differences. Table 1 shows the total resistance and the resistance for each type of UTI to each of the antibiotics tested.Table 1Number of strains resistant to each of the antibiotics tested, in total and according to the type of UTI (complicated or uncomplicated)AntibioticTotal (n=123)Complicated UTI (n=74)Uncomplicated UTI (n=49)Ampicillin78 (63.41%)49 (66.21%)29 (59.18%)Gentamicin6(4.88%)6(8.11%)0Fosfomycin1 (0.81%)01 (2.04%)Cefazolin4 (3.25%)3 (4.05%)1 (2.04%)Co‐trimoxazole25 (20.32%)19 (25.67%)6 (12.24%)Nalidixic acid20(16.26%)18(24.32%)ap<0.05.2(4.08%)Norfloxacin16(13%)16(21.62%)ap<0.05.0ap<0.05.Ciprofloxacin16(13%)16 (21.62%)ap<0.05.0ap<0.05.a p<0.05. Open table in a new tab In Spain, the annual consumption of quinolones, fundamentally fluoroquinolones, is high. In 1996, in a population of approximately 39 000 000 inhabitants, the consumption in outpatients was 4.87×106 units of fluoroquinolones (54% ciprofloxacin, 36% norfloxacin, 10% others) [8Information provided by the International Marketing Service, Spain, 1996.Google Scholar]. Published data show a high incidence of fluoroquinolone resistance in E. coli isolated in urine. These figures have increased in recent years [1Alós JI Balas D Gómez-Garcés JL Grupo de Estudio de Infecciones en Atención Primaria Prevalencia de susceptibilidad a quinolonas y otros antibióticos en microorganismos aislados de bacteriurias extrahospitalarias en Madrid en 1995.Rev Clin Esp. 1997; 197: 167-171PubMed Google Scholar, 2Pérez-Trallero E Urbieta M Jimenez D García-Arenzana JM Cilla G Ten-year survey of quinolone resistance in Escherichia coli causing urinary tract infections.Eur J Clin Microbiol Infect Dis. 1993; 12: 349-351Crossref PubMed Scopus (60) Google Scholar, 3Aguiar JM Chacon J Cantón R Baquero F The emergence of highly fluoroquinolone-resistant Escherichia coli in community-acquired urinary tract infections.J Antimicrob Chemother. 1992; 29: 349-350Crossref PubMed Scopus (81) Google Scholar]. Recent unpublished data, from both our laboratory and others, show up to 20% resistance to fluoroquinolone and up to 30% to nalidixic acid. Under these circumstances many physicians may feel that the quinolones are losing ground in their role as empirical therapy for community-acquired UTI. However, when comparing the susceptibility to quinolones according to the type of UTI, significant differences are found. The data obtained in this study seem to indicate that the overall increase in quinolone resistance in E. coli is caused largely by the strains implicated in complicated UTI. The probable cause of these differences is prior treatment with quinolones which has brought about a selection of resistant strains. Patients with complicated UTI, since they have an underlying disease or structural or functional abnormalities of the urinary tract, are treated more often and for longer periods with antibiotics. In Spain, since the late 1980s, these treatments have been carried out, in many cases, with fluoroquinolones. Patients with uncomplicated UTI have, in general, received fewer and shorter prior treatments, with a lesser risk of selection of quinolone-resistant strains. For the physician, it is easier to determine the type of UTI than to ascertain prior antibiotic treatment, since patients, very often, do not remember. According to the type of UTI, one may deduce possible resistance patterns which influence the choice of empirical therapy. The high frequency of resistance reported in some works may be due to the bias caused by the use of strains isolated in the microbiology laboratory. These strains are usually the cause of all complicated UTI and only a proportion of those implicated in uncomplicated UTI, since in the latter case a urine culture is not always requested and empirical treatment is often preferred. It may be deduced from our results that, in Spain, the quinolones still constitute valid therapy for community-acquired uncomplicated UTI. Judicious use of these antibiotics is advised to avoid the development of quinolone-resistant strains. However, in patients with complicated UTI, it is advisable to use the quinolone only when the result of the antibiogram is available.