Antibiotic resistance pattern among common bacterial uropathogens with a special reference to ciprofloxacin resistant Escherichia coli

Urinary tract infections (UTIs) are amongst the most common infections encountered in clinical practice1. The commonest bacterial agent involved in causation of UTIs is Escherichia coli, being the principal pathogen both in the community as well as in the hospital23.

The treatment of UTIs varies according to the age of the patient, sex, underlying disease, infecting agent and whether there is lower or upper urinary tract involvement. Trimethoprim/sulphamethoxazole is the recommended drug for the treatment of UTIs in settings where the prevalence of resistance is <10-20 per cent and ciprofloxacin is recommended where this resistance is >20 per cent, according to the Infectious Diseases Society of America (IDSA) guidelines45. The other agents used in the treatment of UTI include fluoroquinolones, cephalosporins and other β-lactams with or without β-lactamase inhibitors, nitrofurantoin45. Recently, several studies have revealed increasing trends of resistance to many antimicrobials including the fluoroquinolones6–8. The increase in bacterial resistance to fluoroquinolone is multifactorial9–13. With the increasing trend of antibiotic-resistance in E. coli, the management of urinary tract infections is likely to become complicated with limited therapeutic options.

The present study was undertaken to assess the current antibiotic resistance pattern in the common uropathogens isolated in a tertiary care hospital in south India with a special emphasis on ciprofloxacin. Since Escherichia coli is the predominant pathogen, the study was focused on it. Further, risk assessment was also performed to determine the factors responsible for the emergence of ciprofloxacin resistance in E. coli.

Material & Methods

Study site: The study was carried out in the department of Microbiology, Jawaharlal Institute of Postgraduate Institute Medical Education & Research (JIPMER), Puducherry, India, during March 2008 to April 2009. This was an analysis of data generated from the records of consecutive urine samples received in the laboratory during the study period. Only the initial sample of an individual received was included to avoid duplication. Analysis of the data was carried out focussing on the age, gender, whether admitted or not, whether received prior antibiotic therapy, any surgical or gynaecological intervention performed in the recent past, and any history of urinary tract infection in the past. The antibiotic susceptibility data of all isolates were also reviewed and analyzed.

Samples received included mid-stream clean catch urine, suprapubic aspirate, urine collected from Foley's catheter and from the nephrostomy tube under sterile precautions, in patients who had undergone percutaneous nephrostomy. Samples were processed and isolates were identified as per standard methods14. All urine samples were inoculated onto cysteine lactose electrolyte deficient (CLED) medium (Himedia, Mumbai, India) using a calibrated loop (volume-0.005 ml) and were incubated for 18-24 h at 37°C.Wet mount preparations were also made from all urine samples to look for pus cells and epithelial cells. Depending upon the number of the colonies grown on the CLED medium, the interpretations of urine culture were made as insignificant (<50 colonies), doubtful significance (>50 - <500 colonies) and significant (≥500 colonies) with due clinical correlation as per recommendations1415. The antibiotic susceptibility testing of the isolated bacteria was carried out by the Kirby Bauer method1516.

Determination of minimum inhibitory concentration (MIC): The MIC testing was performed as per guidelines1516. The MIC interpretive standards for the susceptibility categories were as per the recommended breakpoints by the Clinical Laboratory Standards Institute (CLSI) for ciprofloxacin16. Ciprofloxacin hydrochloride was obtained from Hi-media (Mumbai, India). The antibiotic was dissolved in sterile distilled water as per the recommendations. The antibiotic was used immediately after reconstitution. The different concentrations of the drug analyzed were 0.5 to 256 μg/ml1516. ATCC E. coli 25922 was inoculated on each plate as the growth control. The growth control was read first followed by the MICs of the test strains1516.

Statistical analysis: The chi-square test or Fisher's exact test was used to compare different groups. Relative risk and odds ratio were determined to compare the risk factors in the different groups of interest (UTI due to ciprofloxacin resistant E. coli and UTI due to E. coli). Statistical softwares GraphPadInStat3 software (GraphPadInc, San Diego, USA) and SPSS 16.0 (SPSS Inc, Chicago, Illinois, USA) were used to analyse the data.

Results

Data from a total of 19,050 consecutive urine samples were included in the study. Of these, 11,811 (62%) were sterile, 4956 (26.01%) showed significant growth, 438 (2.3%) showed insignificant growth and 1828 (9.6%) were found contaminated. Wet mount microscopy for presence of bacteria or pus cells in significant amount per field had sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of 83, 58, 44 and 89 per cent, respectively in detecting infections.

Of the 4956 culture positives, E. coli was the most common (59%) isolate. (Table I). The percentage of Klebsiella pneumoniae, Acinetobacter spp, Pseudomonas spp, Staphylococcus spp, Enterococcus faecalis was higher in patient females and patients with history of prior treatment with antibiotics, compared to the rest. In addition to these isolates, the percentage of Proteus spp was more in males, especially in cases with prior history of antibiotics administration (Table I).

Table I Organisms causing urinary tract infection (UTI) in relation to different host characteristics

Seventy three per cent of all E.coli isolates were found to be resistant to ciprofloxacin.Ciprofloxacin resistance was comparatively less among the other Gram-negative uropathogens like Pseudomonas spp, Proteus spp and K. pneumoniae as mentioned (Table II). Resistance to the aminoglycosides amikacin and gentamicin was also considerable especially among isolates of Acinetobacter spp with as many as 68.9 per cent of all Acinetobacter isolates showing resistance to gentamicin and 58.6 per cent to amikacin. The percentage of isolates of E. coli resistant to ampicillin was found to be as much as 80.6 per cent.

The rates of resistance among Gram-negative uropathogens to third generation cephalosporins like ceftriaxone and ceftazidime were high. Ceftriaxone resistance were seen in 60.5 and 59.3 per cent of all isolates of E. coli and K. pneumoniae. Resistance to ceftazidime among the Gram- negative non-fermenters was also considerable. 60 per cent of all Acinetobacter isolates and 44.5 per cent of all Pseudomonas spp isolates were found to be resistant to ceftazidime (Table II). Compared to the other Gram-negative uropathogens, resistance to the urinary antiseptic nitrofurantoin was comparatively less among isolates of E. coli.

Table II Antibiotic susceptibility pattern of the Gram-negative isolates (% resistance)

Amongst the Gram-positive isolates, Enterococcus faecalis was the most commonly isolated organism with 3.2 per cent resistance to vancomycin. Amongst the Streptococcus spp, 3 (15%) were identified as S. agalactiae, which were isolated from antenatal women and were sensitive to all the antimicrobial agents. Substantial number of Staphylococcus [109 (65.6%)] and Streptococcus [11 (55%)] isolates were resistant to ciprofloxacin. Resistance to nitrofurantoin was comparatively more amongst the Enterococcus spp (Table III).

Table III Antibiotic susceptibility pattern of the Gram-positive isolates

Our findings indicate that prior antibiotic therapy especially flouroquinolone therapy and post-operative patients of gynecological surgeries, were significant risk factors for the emergence of fluoroquinolone resistant E. coli (Table IV). The difference between the resistance patterns amongst the inpatients (IP) and the outpatients (OP) was very minimal and no significant difference (P=0.0981) was noted between the two groups (Tables V and VI). But, there was a significant difference (P=0.0058) noted between the IP and the OP groups particularly in the age group of 19 to 25 yr with UTI due to ciprofloxacin resistant E. coli. Though E. coli was the commonest organism associated with UTI in both the IP and OP patients, there were many other unusual isolates like Psuedomonas spp which were isolated both from the OP patients (Table VII).

Table IV Risk factors with respect to occurrence of UTI due to ciprofloxacin resistant E. coli

Table V Distribution of the cases with UTI due to ciprofloxacin resistant E. coli as a function of age, gender and inpatient/outpatient status

Table VI Distribution of the cases with UTI due to E. coli as a function of age, gender and inpatient/outpatient status

Table VII Distribution of the various organisms in the inpatients and the outpatient groups

A total of 150 consecutive isolates of E. coli were subjected to MIC determination. Of these, 137 were ciprofloxacin resistant while 23 were sensitive. MIC₅₀ was found to be between 32 and 64 μg/ml, while MIC₉₀ was determined to be at 128 μg/ml for ciprofloxacin. It was noted that 5 of 150 (3.33%) isolates had MIC at 256 μg/ml, 106 (70.67%) had MIC at 64 μg/ml, 58 (38.66%) at 32 μg/ml, 37 (24.66%) at 16 μg/ml, 26 (17.33%) had MIC at 8 μg/ml and 23 (15.33%) had MIC at 4 μg/ml. The resistance pattern varied in different age groups. In the 0-5 years group, it was noted that (10 out of 150) 6.6 per cent isolates of E. coli were ciprofloxacin resistant whereas the number of isolates resistant to ciprofloxacin in the 6-10 and 10-18 yr groups was equal i.e. (7 of 150) 4.6 per cent. In the adults, in the age group of 19-25 yr, the resistance was noted in (32 out of 150) 21.3 per cent. In the age group of >25 yr up to 40 yr, the resistance was found to be (39 out of 150) 26 per cent and in the age group of >40 up to 60 yr, the resistance was found to be 27 (18%). In the elderly (>60 yr), the resistance was found to be 18.6 per cent (28 out of 150).

Discussion

Ciprofloxacin and ofloxacin are the most extensively used fluoroquinolones for the treatment of UTIs.

This study showed that E. coli was the commonest pathogen causing complicated and uncomplicated UTI as described previously1–3. There are several organisms known to cause UTIs, including P. aeruginosa, S. saprophyticus, S.epidermidis, Enterococcus spp, P. mirabilis, Klebsiella spp., Citrobacter spp, etc. as reported by earlier workers1718. This study also demonstrates (Table VII) the emergence of E. faecalis and the non-fermenters Acinetobacter spp and Pseudomonas spp as major uropathogens especially in the patients admitted in the hospitals, more so in the intensive care units. Such findings have been documented elsewhere5–1619–22. The percentage of isolates of E.coli resistant to ampicillin was found to be as much as 80 per cent in our set up. Such high levels of resistance to ampicillin have been quoted by many other studies from different parts of India. Gupta et al23 in a study from the northern part of the country reported 76 per cent resistance to E.coli isolates for ampicillin. A more recent study from Karnataka reported a resistance rate of 80.6 per cent for ampicillin24.

Our MIC results showed that fluoroquinolone resistance increased significantly with patient's age. An MIC of 256 μg/ml was noted in the age group of >60 yr of age. There could be two explanations for this. Firstly, as a consequence of frequent exposure to fluoroquinolones resulting from the treatment of repeated infections in elderly leads to increase in MIC of fluoroquinolone19. Secondly, unlike urinary tract infections (UTIs) in females, UTIs in males are frequently complicated and are more likely to require prolonged antimicrobial therapy, especially in the elderly, potentially explaining the fluoroquinolone the higher MIC2526. Moreover, fluoroquinolones are used to treat chronic prostatitis, even though these do not all readily penetrate the prostate192728. Such doses of fluoroquinolones produces “sub-inhibitory concentration effects” which leads to the selection of mutants exhibiting resistance. All the elderly males, from whom the resistant isolates with high MIC were obtained, had prostatitis. In isolates from children below 18 yr, the MIC was 4 μg/ml. Amongst the cases with MIC >4 μg/ml, 14.6 per cent had obstructive uropathy, 6.6 per cent had undergone uro-surgery, 13.3 per cent had undergone gynaecological procedures, 22 per cent had indwelling catheter, and 28 per cent had history of prior antibiotic therapy.

The emergence of resistance for fluoroquinolones is multifactorial7171920. Resistance to ciprofloxacin has emerged in a variety of genera of the family Enterobacteriaceae2728. Apart from the notable resistance of E. coli to ciprofloxacin, other organisms were also found to be resistant to ciprofloxacin especially K. pneumoniae, Citrobacter spp, Pseudomonas spp, Acinetobacter spp, Proteus spp and Enterobacter spp, Staphylococcus spp, and E. faecalis. Also, fluoroquinolone resistance in E. coli has emerged particularly in patients with urinary tract infections who have received fluoroquinolone prophylaxis5–12. An association between the increase in quinolone prescriptions and an increase in bacterial resistance has been reported from several countries5–812. Usually, the prevalence of fluoroquinolone resistance is related to the intensity of antibiotic use5. Resistance rates for ciprofloxacin against uncomplicated UTI pathogens were reported as 0-14.7 per cent in the ECO-SENS Project, 2.5 per cent in the USA and 1.2 per cent in outpatients in Canada278.

In conclusion, the present results in increasing antibiotic resistance trends in UTI patients indicate that it is imperative to rationalize the use of antimicrobials and to use these conservatively.30