INTRODUCTION
Urinary tract infections (UTIs) are one of the most commonly diagnosed infections, both in community-based and hospitalized patients.1 In the United States, UTIs accounted for 10.5 million outpatient visits in 2007, of which about 21% corresponded to the emergency department.2 In Mexico, UTIs were the third leading cause of morbidity in 2014, with a total of 4,244,053 notified cases.3
The financial burden imposed by UTIs is remarkable, especially in Latin American countries, where antimicrobial prescription is less strictly controlled.4 A cross-sectional study carried out at a pediatric hospital in the Mexican state of Sonora, found that the average cost of medical care per episode of nosocomial UTI was $2062.00 USD.5
In general, women are more affected by UTIs than men. It is estimated that, by the age of 24, 33% of women will require antimicrobial treatment for at least one episode of UTI and more than 50% will experience at least one UTI during their lifetime.6 In fact, 75.6% of the cases of UTI reported in 2008 in Mexico were diagnosed in female patients.7
Escherichia coli (E. coli) is the most frequent causative agent of UTIs, accounting for 75%95% of infections.8 The remaining episodes are mostly caused by other Enterobacteriaceae, namely Proteus mirabilis (P. mirabilis) and Klebsiella pneumoniae (K. pneumoniae), followed by Gram positive cocci, such as Staphylococcus saprophyticus (S. saprophyticus).8 Fungal microorganisms are relatively common uropathogens in nosocomial UTIs.9
Treatment selection for UTI management is becoming increasingly challenging. Antimicrobial resistance, mostly due to antimicrobial misuse, is on the rise, thus limiting therapeutic options.10 The emergence of E. coli, as well as of other Enterobacteriaceae that produce extendedspectrum β-lactamases (ESBLs) is of particular concern, since these enzymes confer resistance to a number of antimicrobials.11
The aim of this study was to assess the prevalence of uropathogens identified in urine cultures analyzed at the laboratory of the Hospital Regional ISSSTE of Monterrey, as well as their antimicrobial susceptibility profiles, to help guide therapy selection at the local level.
MATERIALS AND METHODS
Study design
This was a non-interventional, retrospective cohort study based on urine sample analysis performed at the laboratory of the Hospital Regional ISSSTE of Monterrey (Nuevo León, Mexico) between January 2010 and September 2015. Information on microbial species and antimicrobial susceptibility was obtained from the laboratory’s electronic system.
The urine samples were obtained through sterile urine collection techniques (suprapubic aspiration or collection from catheters). Using a calibrated loop, these samples were inoculated on CLED and MacConkey agar media and incubated at 37º C.
Only urine cultures considered positive (microbial growth > 105 CFU/mL) and exhibiting growth of a single microbial strain were included in this analysis. Each urine culture corresponded to a single patient.
Uropathogen identification and antimicrobial sensitivity testing were performed using the Vitek® 2 system (bioMérieux, Inc., Durham, NC, USA) according to the manufacturer’s instructions. The Gram-negative (GN), Gram-positive (GP), and Yeast (YST) identification cards and the GP67, GN70, and XL05 antimicrobial susceptibility testing cards were used. Production of ESBLs was also assessed by the Vitek® 2 system. Quality control was performed using American Type Culture Collection strains.
The Ethics Committee of the Hospital Regional ISSSTE of Monterrey approved the study, and due to its retrospective design and in compliance with the provisions of Mexican law, written informed consent from the individuals whose urine samples were analyzed was not required.
RESULTS
Data on 4,394 positive urine cultures were obtained for analysis. The most frequently isolated uropathogen was E. coli, which was identified in 47.1% of the cultures (Table 1).
Microorganism | No. of isolates (%) (n = 4394) |
---|---|
Gram-negative bacteria | |
Escherichia coli | 2070 (47.1) |
Klebsiella pneumoniae | 410 (9.3) |
Pseudomonas aeruginosa | 450 (10.2) |
Enterobacter cloacae | 115 (2.6) |
Proteus mirabilis | 113 (2.6) |
Acinetobacter baumannii | 42 (1.0) |
Gram-positive bacteria | |
Enterococcus faecalis | 582 (13.2) |
Enterococcus faecium | 155 (3.5) |
Staphylococcus haemolyticus | 52 (1.2) |
Staphylococcus aureus | 44 (1.0) |
Staphylococcus epidermidis | 40 (0.9) |
Fungi | |
Candida albicans | 158 (3.6) |
Candida glabrata | 82 (1.9) |
Candida tropicalis | 81 (1.8) |
The most common uropathogens after E. coli were Enterococcus faecalis (E. faecalis) and Pseudomonas aeruginosa (P. aeruginosa), which were isolated in 13.2% and 10.2% of the cultures, respectively. K. pneumoniae, the second most frequently identified Gram-negative bacterium, was found in 410 (9.33%) urine cultures. Among the fungal species, Candida albicans (C. albicans) was the most common, identified in 3.6% of the urine cultures.
E. coli and K. pneumoniae were the ESBL-producing uropathogens found in the urine cultures and they presented similar prevalences: 59.4% and 59.6%, respectively (Table 2).
Microorganism | No. of ESBL-producing isolates (%) |
---|---|
Escherichia coli | 1229 (59.4) |
Klebsiella pneumoniae | 244 (59.6) |
The sensitivity of the isolated Gram-negative bacteria varied widely across antimicrobial classes (Table 3).
Antimicrobial classes | Gram-negative bacteria | |||||
---|---|---|---|---|---|---|
Escherichia coli, % (n = 2070) | Pseudomonas aeruginosa, % (n = 450) | Klebsiella pneumoniae, % (n = 410) | Enterobacter cloacae, % (n = 115) | Proteus mirabilis, % (n = 113) | Acinetobacter baumannii, % (n = 42) | |
Penicillins | ||||||
Ampicillin | 21.1 | 0.0 | 4.3 | 14.7 | 52.3 | 0.0 |
Penicillins/β-lactamase inhibitor | ||||||
Ampicillin/Sulbactam | 28.2 | 0.0 | 51.8 | 0.0 | 82.2 | 33.5 |
Piperacillin/Tazobactam | 83.9 | 71.9 | 81.8 | 25.7 | 92.3 | 17.4 |
Cephalosporins | ||||||
Cefazolin | 54.2 | 0.0 | 58.0 | 0.0 | 67.3 | 0.0 |
Cefepime | 46.7 | 25.6 | 60.8 | 42.8 | 68.3 | 18.9 |
Ceftriaxone | 57.3 | 0.2 | 59.4 | 25.3 | 69.1 | 7.1 |
Aminoglycosides | ||||||
Amikacin | 98.5 | 33.0 | 91.7 | 57.9 | 97.3 | 76.1 |
Gentamicin | 68.4 | 30.3 | 75.1 | 56.0 | 85.5 | 32.4 |
Tobramycin | 55.3 | 30.9 | 65.0 | 47.2 | 85.8 | 24.6 |
Streptomycin | ||||||
Fluoroquinolones | ||||||
Ciprofloxacin | 34.3 | 19.5 | 68.5 | 35.2 | 92.8 | 21.3 |
Moxifloxacin | 33.4 | 15.6 | 69.6 | 37.5 | 88.4 | 11.8 |
Carbapenems | ||||||
Ertapenem | 99.6 | 0.0 | 96.8 | 76.6 | 94.5 | 0.0 |
Imipenem | 100.0 | 22.6 | 97.3 | 94.1 | NA | 22.8 |
Meropenem | 100.0 | 28.9 | 98.1 | 93.1 | 95.4 | 23.8 |
Other antimicrobials | ||||||
Nitrofurantoin | 82.9 | 0.0 | 8.8 | 13.9 | 7.3 | 0.0 |
Tigecycline | 99.8 | 0.4 | 86.3 | 58.6 | 31.1 | 31.0 |
TMP/SMX | 42.5 | 0.8 | 62.3 | 28.8 | 56.8 | 11.0 |
Colistin | 0.0 | 30.0 | 0.0 | 0.0 | 0.0 | 40.6 |
Aztreonam | 57.7 | 14.6 | 45.9 | 32.8 | 63.0 | 2.3 |
NA: Not available
TMP/SMX: Trimethoprim/Sulfamethoxazole
Carbapenems were the most active antimicrobial agents against E. coli, with 100% of the isolates showing sensitivity to both imipenem and meropenem and 99.6% displaying sensitivity to ertapenem. The majority of E. coli isolates were also very sensitive to tigecycline (99.8%) and amikacin (98.5%).
K. pneumoniae presented a similar sensitivity pattern, also in its range of sensitivity to the various cephalosporins. However, more isolates of K. pneumoniae than E. coli were sensitive to ampicillin/sulbactam (51.8% vs. 28.2%, respectively).
As did E. coli and K. pneumoniae, P. mirabilis isolates showed a degree of sensitivity to all antimicrobials, with the exception of colistin.
A generally low proportion of P. aeruginosa isolates presented with sensitivity to the antimicrobials studied herein. Nevertheless, 71.9% of the isolates were sensitive to piperacillin/ tazobactam.
Gentamicin, ciprofloxacin, moxifloxacin, and vancomycin were the only antimicrobials to which a proportion of all species of Grampositive bacteria were sensitive (Table 4). One hundred percent of the isolates of all the species, except Staphylococcus epidermidis (S. epidermidis), were sensitive to tigecycline. Nitrofurantoin also showed good action against these bacteria, excluding Enterococcus faecium (E. faecium).
Antimicrobial classes | Gram-positive bacteria | ||||
---|---|---|---|---|---|
Enterococcus faecalis, % (n = 582) | Enterococcus faecium, % (n = 155) | Staphylococcus haemolyticus, % (n = 52) | Staphylococcus aureus, % (n = 44) | Staphylococcus epidermidis, % (n = 40) | |
Penicillins | |||||
Ampicillin | 62.1 | 13.3 | 0.0 | 0.0 | 0.0 |
Benzylpenicillin | 62.1 | 0.0 | 0.0 | 2.3 | 0.0 |
Aminoglycosides | |||||
Gentamicin | 33.4 | 22.6 | 47.9 | 82.6 | 44.9 |
Streptomycin | 47.0 | 43.5 | 0.0 | 0.0 | 0.0 |
Fluoroquinolones | |||||
Ciprofloxacin | 43.9 | 8.4 | 32.6 | 52.3 | 47.7 |
Moxifloxacin | 43.4 | 11.2 | 50.1 | 56.8 | 72.4 |
Levofloxacin | 44.8 | 0.0 | 5.8 | (54.5 | 47.7 |
Other antimicrobials | |||||
Nitrofurantoin | 96.3 | 0.0 | 100.0 | 100.0 | 100.0 |
Tigecycline | 100.0 | 100.0 | 100.0 | 100.0 | 0.0 |
TMP/SMX | 0.0 | 0.0 | 28.7 | 86.4 | 41.7 |
Vancomycin | 98.8 | 65.9 | 98.0 | 97.8 | 80.0 |
Linezolid | 96.8 | 88.2 | 100.0 | 0.0 | 100.0 |
Tetracycline | 35.1 | 71.7 | 0.0 | 95.5 | 87.6 |
Colistin | 0.0 | 0.0 | 9.6 | 0.0 | 0.0 |
Clindamycin | 0.0 | 0.0 | 51.9 | 54.7 | 72.5 |
Rifampicin | 0.0 | 0.0 | 94.1 | 95.5 | 100.0 |
Erythromycin | 15.5 | 0.0 | 23.2 | 50.2 | 67.7 |
In general, the isolated fungal species presented good sensitivity to the antifungals analyzed (Table 5). Candida tropicalis (C. tropicalis) isolates, in particular, were all sensitive to the four antifungals. One hundred percent of C. albicans isolates were sensitive to fluconazole and voriconazole. The majority of Candida glabrata (C. glabrata) isolates were sensitive to flucytosine (96.3%).
DISCUSSION
The most commonly isolated uropathogen in this sample of urine cultures was E. coli, accounting for 47.1% of all isolated microorganisms. This finding is consistent with numerous studies that have shown this bacterium to be the most important etiologic agent of UTI worldwide.8,12-14 In Mexico, this trend has been observed in both adult and pediatric populations, with local studies reporting frequencies of E. coli isolates ranging from 41% to 79%.15-18
After E. coli, the most frequently isolated microorganisms were E. faecalis (13.2%), P. aeruginosa (10.2%), and K. pneumoniae (9.3%), with prevalences similar to those reported elsewhere in Latin America.12,15,18
More than half of the E. coli isolates (59.4%), as well as of the K. pneumoniae isolates (59.6%), showed ESBL production. A similar prevalence of ESBL-producing K. pneumoniae (53.3%) was found by Flam et al. across 16 Latin American medical centers. However, the prevalence of ESBL-producing E. coli reported by those authors (37.6%) was lower than ours.19
The high rates of ESBL-producing bacteria identified in our sample are cause for concern. In the past, the majority of ESBL-producing bacteria were isolated from hospitalized patients, but recent data have shown that ESBL production by common, community-acquired uropathogens, such as E. coli, has markedly increased.20-21 This has also been reported in Latin America.4,14 ESBL-producing bacteria often show co-resistance to various antimicrobial drug classes (e.g. penicillins, cephalosporins, trimethoprim/sulfamethoxazole [TMP/SMX], among others).22-23 This issue is further complicated by the fact that these agents frequently cause serious invasive infections and are associated with a worse prognosis.24
In the present study, fungi were isolated in 321 urine cultures. Isolation frequency, specifically of C. albicans (3.6%), was lower than that reported in other studies,25-28 which could be explained by differences between study populations.
In our sample, almost half (42.5%) of the E.coli isolates were sensitive to TMP/SMX. The number of isolates sensitive to fluoroquinolones was lower (ciprofloxacin: 34.3%; moxifloxacin: 33.4%). Fluoroquinolones have been prescribed empirically when there is clinical suspicion of UTI, due to the known, widespread resistance to TMP/SMX. It appears, however, that such frequent use has compromised their efficacy.29
Overall, the number of E. coli and K. pneumoniae isolates sensitive to the antimicrobials tested was low. This could be explained by the high prevalence of ESBL-producing isolates of both species. Nevertheless, the sensitivity of these Enterobacteriaceae to carbapenems was high, with over 95% of the isolates of each species showing sensitivity to all the tested antimicrobials of this class. This is in accordance with the findings of the SENTRY and SMART surveillance programs.12,14 Indeed, carbapenems have been identified as the antimicrobials of choice for infections caused by ESBL-producing microorganisms.30 However, caution should be used, given that carbapenem resistance is reportedly increasing.24
Isolates of P. aeruginosa, which is a frequent etiologic agent of nosocomial infections,31 showed overall low sensitivity to the antimicrobials listed. Concerns over infections caused by drug-resistant P. aeruginosa are not new. A 5-year analysis of P. aeruginosa susceptibility rates in Latin American centers participating in the SENTRY surveillance program showed rapidly increasing resistance to common antipseudomonal agents, especially meropenem (from 83.0% in 1997 to 64.4% in 2001,p <0.001).32
The antimicrobials that exerted the most action on Gram-positive bacteria were nitrofurantoin and tigecycline, with 100%, or nearly 100%, of the isolates showing sensitivity to them (except those of E. faecium and S. epidermidis, respectively). Nitrofurantoin is still rarely used for the empirical treatment of UTIs. Its spectrum of susceptible organisms has remained virtually unchanged, with little evidence of resistance emergence, despite its being on the market for over 60 years.33 Nitrofurantoin might therefore be considered an interesting alternative for the treatment of UTIs caused by multi-drug resistant microorganisms.34
Tigecycline also presents strong antimicrobial activity against both Gram-positive and Gram-negative bacteria. Data on antimicrobial susceptibility of microorganisms collected in Mexico between 2005 and 2012 show 100% tigecycline susceptibility reported among isolates of Enterococcus sp. and S. aureus,35 which was similar to what we found. Jones et al. reported identical susceptibility rates in their 2011 analysis of Latin American isolates.36
Nearly 100% of the isolates of every fungal species identified in the present sample were sensitive to the antifungals tested. Nonetheless, it is worth noting that fungi in urine cultures are often asymptomatic and may not require pharmacologic treatment. Eliminating the elements that facilitate colonization, such as indwelling catheters, may suffice.37 Fluconazole has been the recommended treatment option, when applicable.38-39
Antimicrobial sensitivities show great geographic variability. Therefore, the present study is limited by the fact that our sample came from a single institution. Each institution should be aware of the relevance of collecting this type of information to better guide therapy selection for UTIs and implement control strategies for any antimicrobial resistance patterns identified.
Moreover, because the hospital laboratory system did not contain such information, it was not possible to assess relationships between uropathogen frequencies and the antimicrobial sensitivity of the isolates to factors such as sex, infection origin, type of UTI, or presence of risk factors, among others.
CONCLUSIONS
The results presented herein provide additional evidence in relation to the role of different microorganism species as etiologic agents of UTIs. Furthermore, they provide valuable antimicrobial sensitivity information that will help guide antimicrobial selection for the treatment of UTIs diagnosed at the Hospital Regional ISSSTE of Monterrey, in Mexico.
Implementation of the appropriate public health actions (e.g. antimicrobial susceptibility surveillance) is necessary to better understand the patterns of antimicrobial sensitivity in Mexico and thus aid in the appropriate decision-making regarding their prescription in clinical practice. This, in turn, is essential for preventing the spread of resistance mechanisms and for reducing direct healthcare costs associated with UTIs.