Multi-drug-resistant Gram-negative bacteria causing urinary tract infections: a review

Abstract

Urinary tract infections (UTIs) are among the most frequent infectious diseases affecting humans, and represent an important public health problem with a substantial economic burden. Due to the high empiric use of antibiotics for the treatment of UTI, antibacterial resistance of Enterobacteriaceae, specifically the main uropathogens Escherichia coli and Klebsiella pneumoniae, has significantly increased worldwide. In this article the worldwide epidemiology of resistant Gram-negative bacteria causing UTIs, with a special focus on extended spectrum beta lactamase (ESBL) positive pathogens, as well as new threats such as multi-drug-resistant (MDR) clones (e.g. E. coli 131 (ST131) and K. pneumoniae ST258), are reviewed. The increased prevalence of MDR Enterobacteriaceae, limiting available treatment options for infections caused by these organisms, and the lack of new antibiotics provide good rationale for using older antibiotics, such as fosfomycin, that have been shown to retain some activity against MDR bacteria.

Keywords:

• Gram-negative bacteria
• Epidemiology
• Urinary tract infections
• ESBL
• E. coli
• K. pneumoniae

Introduction

Urinary tract infections (UTIs) are among the most frequent infectious diseases affecting humans, and represent an important public health problem with a substantial economic burden. In the US, UTIs are responsible for >7 million physician visits annually and 15% of all community-prescribed antibiotics, with similar figures in some European countries.¹

UTIs are primarily caused by Gram-negative bacteria. The main pathogen responsible for uncomplicated cystitis and pyelonephritis is Escherichia coli followed by other species of Enterobacteriaceae, such as Proteus mirabilis and mostly Klebsiella pneumoniae, and by Gram-positive pathogens, such as Enterococcus faecalis and Staphylococcus saprophyticus.^2–3 International guidelines for the treatment of uncomplicated UTIs and pyelonephritis recommend various agents, such as nitrofurantoin monohydrate, trimethoprim-sulfamethoxazole, fosfomycin trometamol, pivmecillinam, fluoroquinolones and beta-lactams.^1,3 However, there is an alarming level of antimicrobial resistance developing in UTI pathogens as a result of indiscriminate and widespread use of antibiotics. Bacteria producing extended spectrum beta-lactamases (ESBLs), showing resistance to most antibiotics except for the carbapenem group, are steadily increasing in the population.^1,4

Clinicians treating these infections caused by drug-resistant pathogens have few effective antimicrobials to choose from and many are associated with significant adverse effects. There are now multi-, and often pan-resistant organisms that could be encountered in clinical practice. Therefore, it is crucial that the spread of resistance be accurately monitored and controlled.⁵

Extended spectrum beta-lactamases (ESBLs)

ESBLs are enzymes able to hydrolyze third-and fourth generation cephalosporins and monobactams, but not cephamycins or carbapenems. They are inhibited by clavulanic acid (sulbactam or tazobactam), the genetic determinants are generally associated with plasmids but often cause co-resistance (aminoglyclosides, quinolones). ESBLs represent a public health concern, due to the high prevalence in E. coli in the community, and have a variable prevalence in K. pneumoniae (higher in hospitals). Table 1 includes a classification of ‘classical’ ESBLs and Figure 1 shows the trend of variants of different beta-lactamase families.

Table 1 Classification of ‘classical’ extended spectrum beta-lactamases (ESBLs)

‘Old’ ESBLs	1983 SHV
Derived from plasmid broad spectrum β-lactamases
	1985 TEM
	1991 OXA
ESBLs of ‘current epidemiological importance’	1989 CTX-M
Derived from chromosomal β-lactamases
‘Unusual’ ESBLs	PER, VEB, TLA, GES, IBC, BES, SFO
Derived, in general, from chromosomal β-lactamases

Figure 1 Increased number of variants of different β-lactamase families.

In past years, CTX-M enzymes have become the most prevalent ESBLs in both nosocomial and community settings.⁶ These are molecular class-A ESBLs, inhibited by clavulanic acid, with higher hydrolytic activity against cefotaxime than ceftazidime (but new variants also hydrolyze ceftazidime) and genetically related to chromosomal β-lactamases from bacteria of the Kluyvera genus. At least five different phylogenetic groups (≤90%) are known. The first CTX-M was identified in 1989 in E. coli isolated from the ear exudate of a newborn, and was highly resistant to cefotaxime and the majority of beta-lactam antibiotics.⁷ Simultaneously, Salmonella typhimurium strains resistant to most beta-lactams, co-trimoxazole, tobramycin and gentamicin were isolated in Argentina, carrying a cefotaximase different from that isolated in E. coli, and named CTX-M-2.⁸

The epidemiology of ESBLs is changing (Table 2). It is noteworthy that in the two last decades of the past century, ESBL-positive strains were confined to the hospital setting, where they were responsible for nosocomial outbreaks, but these strains have become prevalent in infected outpatients since 2000. The prevalence of ESBLs, especially CTX-M, is increasing worldwide, with significant differences between geographic areas, again due to extensive and often indiscriminate use of antibiotics.

Table 2 Changing epidemiology of ESBLs

Characteristic	1980s–2000s	2000 onwards
Epidemiology	Nosocomial outbreaks	Complex population structure (polyclonal outbreaks, allodemic)
	Rare in the community	Mainly in outpatients (influx to hospitals)
Prevalence	Variable in different institutions	Increased worldwide, but different in diverse areas, units, …
	K. pneumonia (+++)	E. coli (+++)
	Enterobacter spp. (+)	K. pneumonia, Enterobacter spp. (++)
		Salmonella spp. (+)
	E. coli, Salmonella spp. (–/+)
Risk factors	ICU, long-term hospitalization	Long-term care facilities
	Urinary, blood catheter infections	Urinary tract infections
	β-lactam, aminoglycoside use	Fluouroquinolone use
Enzymes	TEM/SHV	TEM/SHV <<<< CTX-M
		Increased number of variants

ICU, intensive care unit.

Epidemiology of resistant Gram-negative bacteria causing UTIs

There is no general overview of ESBL in UTIs, but results from bloodstream and abdominal infections give an idea of the different prevalence of these enzymes throughout Europe.

According to the European Antimicrobial Resistance Surveillance Network (EARS-Net), the main European Union (EU) surveillance system for multi-drug-resistant pathogens isolated in bloodstream infections and collected in 28 EU member states plus two European Economic Area (EEA) countries (Iceland and Norway), the resistance of E. coli and K. pneumonia to most common antibacterial classes used in clinical practice is high particularly in some countries, where these drugs can no more be empirically used for the treatments of UTIs. For example, in Italy in 2014, rates of antibiotic resistance to aminopenicillins, aminoglycosides and fluoroquinolones in E. coli were 65, 19 and 44%, and rates of antibiotic resistance to aminoglycosides and fluoroquinolones in K. pneumonia were 49% and 56%.⁹

As far as ESBL-positive E. coli strains are concerned, national rates of resistant isolates in 2014 ranged from 3.3% (Iceland) to 40.4% (Bulgaria) (Figure 2, Table 3). It is noteworthy that most countries with resistance percentages of ≥25% were located in South and East Europe, with a clear gradient from North to South. In fact, rates of resistant isolates in Northern European countries was lower than 10%, but ranged from 12 to 30% in Mediterranean countries (with even higher rates in Eastern Europe). Between 2011 and 2014, increasing percentages of ESBL-positive E. coli were observed in 12 countries (Belgium, Bulgaria, the Czech Republic, France, Germany, Greece, Ireland, Italy, Norway, Portugal, Slovenia and Sweden). The same increasing trend was seen for the overall EU/EEA population rate (from 9.6% in 2011 to 12.0% in 2014).⁹

Figure 2 Escherichia coli. Percentage of invasive isolates with resistance to third-generation cephalosporins, by country, EU/EEA countries, 2014⁹.

Table 3 Percentage of ESBL positive E. coli and K. pneumoniae isolates in the EU/EEA Countries, 2014⁹

Country	ESBL-positive E. coli (%)	ESBL-positive K. pneumoniae (%)
Austria	9.4	8.2
Belgium	9.7	16.3
Bulgaria	40.4	74.8
Croatia	10.8	47.9
Cyprus	28.8	32.5
Czech Republic	14.0	52.9
Denmark	7.0	7.6
Estonia	9.3	20.7
Finland	5.5	2.4
France	9.9	29.6
Germany	10.5	12.7
Greece	21.0	72.5
Hungary	16.4	35.6
Iceland	3.3	0
Ireland	10.7	11.6
Italy	28.7	56.5
Latvia	10.9	52.9
Lithuania	8.1	52.6
Luxembourg	12.0	34.8
Malta	10.8	29.7
the Netherlands	5.7	5.5
Norway	5.8	5.9
Poland*	11	65
Portugal	16.4	40.9
Romania	29.4	73.8
Slovakia	31.8	69.4
Slovenia	12.7	26.6
Spain	12.3	18.0
Sweden	5.6	4.5
United Kingdom	10.3	9.3
EU/EEA population – Weighted mean)**	12.0	28.0

* year 2013

** The EU/EEA population-weighted mean excludes countries not reporting data for all four years.

Findings are similar for ESBL-positive K. pneumoniae strains. The percentage of resistant isolates in 2014 varied widely (from zero in Iceland to 74.8% in Bulgaria), with a significant increasing trend between 2011 and 2014 seen in 11 countries (the Czech Republic, France, Ireland, Italy, Malta, Norway, Portugal, Romania, Spain, Sweden and the United Kingdom) and overall (from 23.6% in 2011 to 28.0% in 2014) (Table 3, Figure 3). A North–South gradient was also present, because most Northern European countries had a rate <10% and most Mediterranean countries had a higher rate, ranging from 18% in Spain to 73% in Greece). Similarly to E. coli, the increase of combined resistance to more antimicrobial classes and the high frequency of ESBL-producing isolates may lead to increased use of carbapenems, thus, favouring the further dissemination of carbapenemase-producing bacteria.⁹

Figure 3 Klebsiella pneumoniae. Percentage of invasive isolates with resistance to third-generation cephalosporins, by country, EU/EEA countries, 2014⁹.

A study evaluated the ESBL rate in 3160 clinical strains of E. coli isolated from intra-abdominal infections from 13 European countries between 2008 and 2009. The average frequency of ESBL-producing strains was 11%, but this varied widely (from 1.4% in Estonia to 35% in Turkey) (Figure 4).¹⁰

Figure 4 Frequencies of ESBL-positive clinical isolates of E. coli in different countries¹⁰.

Outside Europe, ESBL diffusion is even higher, and prevalence is increasing everywhere, particularly in Asia, Latin America and the Middle East. Hawser et al. evaluated 2841 clinical isolates of K. pneumoniae from intra-abdominal infections worldwide collected in the Study for Monitoring Antimicrobial Resistance Trends (SMART) during 2008 and 2009. The overall rate of ESBL-positive strains was 22.4%, with the highest resistance rates observed in Latin America and Africa (Table 4).¹¹ SMART data confirmed the same global rate of ESBL positivity in E. coli (23%), with the highest values in Asia and Latin America.

Table 4 Percentages of ESBL-positive K. pneumoniae isolates worldwide, 2011¹¹

Country	ESBL-positive K. pneumoniae (%)
Global	22.4
Asia	25.6
Europe	19.7
North America	10
Latin America	34.6
South Pacific	19.5
Middle East	22.6
Africa	32.2

The SMART study also evaluated the prevalence of ESBLs in Enterobacteriaceae (E. coli, K. pneumoniae, K. oxytoca and P. mirabilis) isolated from patients with UTIs between 2009 and 2011 (Figure 5). In this case, Asia and the Middle East showed not only the highest ESBL prevalence, but also a significant increase from 2009 to 2011 (Middle East p < 0.0001, Asia p = 0.0002). In Europe, the average prevalence of ESBL-positive UTI pathogens was calculated to be 20%.¹² In all regions, with the partial exception of North America, molecular data show the absolute dominance of CTX-M, together with the emergence of carbapenemases.^11,13–15

Figure 5 Prevalence of ESBLs in E. coli, K. pneumonia, K. oxytoca and P. mirabilis from urinary tract infections by region from SMART 2009 to 2011¹².

Most ESBL-positive Enterobacteriaceae are resistant to several antibiotics, such as fluoroquinolones, trimethoprim-sulfamethoxazole and beta-lactams, which are no longer suitable for the empiric treatment of some infections, including UTIs. One of the most active antibiotics against ESBL-positive UTI pathogens is fosfomycin trometamol. Data from a systematic review of ESBL-producing E. coli susceptibility to fosfomycin between 2010 and 2015 showed rates of 81% to 100% (Table 5).¹⁶, ¹⁷

Table 5 Susceptibility to fosfomycin of ESBL-producing E. coli strains (2010–2017) Adapted from ^16,17

Author, year	Country, period	Method	No. of isolates	Susceptibility to fosfomycin (%)
Mezzatesta 2017^17	Italy, 2015–2016	AD, BMD, E test	106	100
Cho 2015^18	Korea, 2008–2013	MicroScan	217	94.9
Sultan 2015^19	India, NR	DD	330	98.9
Khan 2014^20	Pakistan, NR	DD	272	93
Cagan 2014^21	Turkey, 2011–2012	DD, E test	118	97
Sorlosano 2014^22	Spain, 2006–2012	Wider, MicroScan	17,467	88–93
Kresken 2014^23	Germany, 2009	BMD	499	97.5
Seo 2014^24	South Korea, 2010–2011	E tests	24	100
Villar 2014^25	Argentina, 2012–2013	DD	374	98.2
Gupta 2013^26	India, 2009	DD	150	100
Lai 2014^27	China, 2004–2012	AD	528	93.8
Qiao 2013^28	China, 2011	Double dilution	46	87
Karlowsky 2014^29	Canada, 2010–2013	AD	42	100
Demir 2013^30	Turkey, 2008–2012	DD, Vitek II	555	94.1
Sahni 2013^31	India, 2009–2010	DD	1150	81
Rodriguez 2013^32	Spain, 2005, 2009, 2011	NR	2005: 5053	2005: 100
			2009: 6324	2009: 90.7
			2011: 6644	2011: 85.6
Fournier 2013^33	France, 2009–2010	AD	100	99
Bonkat 2013^34	Switzerland, 2001–2010	BMD	196*	100
Araj 2012^35	Lebanon, NR	DD	374	86
Briongos 2012^36	Spain, 2009	Vitek II, E test	372	88.7
Meier 2011^37	Switzerland, 2007–2009	DD	123**	100
Liu 2011^38	Taiwan, 2008–2009	DD	134	95.5
Del Heijer 2010^39	Netherland, 2009	MD	785	100
Auer 2010^40	Austria, 2004–2008	Vitek II	100	97
Wachino 2010^41	Japan, 2002–2007	AD	192	96.4
Tena 2010^42	Spain, 2003–2007	Vitek II, MiscoScan, Wider	33,651	97.6
Chislet 2010^43	UK, 2009	DD	146	96.6

* 83 were tested for fosfomycin susceptibility

** Interpretation according to the CLSI is reported in the table; for EUCAST, 93.1% of E. coli were susceptible.

Notes: AD, agar dilution; BMD, broth microdilution; DD, disc diffusion; MD, microdilution; NA, not available; NR, not reported.

A new threat: multi-drug-resistant (MDR) clones

ESBL diffusion is typically driven by international MDR high-risk clones. E. coli sequence type 131 (ST131) and K. pneumoniae ST258 are the quintessential examples of these clones. They emerged in the 2000s as important human pathogens, have spread extensively throughout the world, and are the main clones responsible for the rapid increase in antimicrobial resistance among E. coli and K. pneumoniae strains, respectively. E. coli ST131 is often fluoroquinolone-resistant and associated with ESBL production, especially CTX-M-15. K. pneumoniae ST258 is associated with carbapenemases, most often KPC-2 and KPC-3. Epidemic plasmids with blaCTX-M and blaKPC have contributed significantly to the success of these clones.⁴⁴

Virulence determinants also appear to play a pivotal role in ST131 success among bacteria responsible for UTIs, as shown in animal models^45,46 and recently demonstrated in 16 CTX-M-15 producing uropathogenic ST131 E. coli isolates recovered in India. Several adhesins, toxins, siderophores and protectins are produced by these isolates.⁴⁷

ESBL-bearing strains seem particularly prone to harbouring other resistance genes, leading to multi-drug resistance. This is particularly the case for ST131,⁴⁸ for which an increase in fosfomycin resistance in urinary isolates has been reported in Spain to parallel fosfomycin consumption.⁴⁹ Another issue with ST131, and a further problem in UTI therapy, seems to be production of all types of carbapenemases by these strains.

Klebsiella pneumoniae from hospitals to outpatients

K. pneumoniae is a leading cause of nosocomial infections worldwide, causing pneumonia, bloodstream infections, UTIs, surgical site or wound infections and meningitis. It is among the most common multi-resistant bacteria causing healthcare-associated infections⁵⁰ and has significant epidemic potential, often being responsible for serious infections. In recent years, K. pneumoniae has shown its ability to acquire plasmids encoding for ESBLs, secondary to the abuse of wide-spectrum cephalosporins in hospitals, but infections caused by this resistant phenotype also spread in the community.⁵¹ The first plasmidic ESBL in K. pneumoniae was identified in Germany at beginning of 1980 and derived from SHV-1.⁵² Until the end of 1990s, K. pneumonia-producing ESBLs were found in hospital settings only and derived from TEM to SHV. From 2000 onwards, the epidemiology changed with the diffusion of CTX-M ESBLs in K. pneumoniae.⁵¹

Sporadic cases of K. pneumoniae with CTX-M-15 were initially reported from various European countries such as Great Britain, Italy and Spain,^53–55 but increasing percentages were soon reported worldwide, as shown by data presented in various multicentre studies.^56–58 The mobility of genetic elements, in particular those conferring antibiotic resistance, together with clonal expansion contribute to the persistence of these strains in hospitals and in the community.^56,58 There are a few reports of K. pneumoniae strains with chromosomally integrated CTX-M-15 at an undetermined locus, which were typed as ST1 in a Spanish study.⁵⁹ Recently, a site of integration of a complete CTX-M-15/ISEcp1 element into the chromosome of a K. pneumoniae ST101 strain⁶⁰ was identified in Germany.

Susceptibilities to fosfomycin between 2010 and 2015 were reported in a recent systematic review – ESBL-producing K. pneumoniae susceptibility to fosfomycin ranged from 40 to 97.5% (Table 6).

Table 6 Susceptibility to fosfomycin of ESBL-producing K. pneumoniae strains (2010–2015)¹⁶

Author, year	Country, period	Method	No. of isolates	Susceptibility to fosfomycin (%)
Cho 2015^18	Korea, 2008–2013	MicroScan	60	61.7
Alibi 2015^61	Tunisia, NR	DD	118	97.5
Khan 2014^20	Pakistan, NR	DD	56	64
Sorlosano 2014^22	Spain, 2006–2012	Wider, MicroScan	3,271	40–78
Villar 2014^25	Argentina, 2012–2013	DD	94	94.7
Kopacz 2013^62	USA, 2009	Phoenix	47	79
Demir 2013^30	Turkey, 2008–2012	DD, Vitek II	509	73.3
Morfin 2013^63	Mexico, 2010–2011	MD	21	94.4
Araj 2012^35	Lebanon, NR	DD	168	62
Briongos 2012^36	Spain, 2009	Vitek II, E test	28	46
Liu 2011^38	Taiwan, 2008–2009	DD	66	57.6

AD, agar dilution; BMD, broth microdilution; DD, disk diffusion; MD, microdilution; NA, not available; NR, not reported.

It was concluded that, on the grounds of local susceptibility data, fosfomycin trometamol could be considered for the treatment of patients with infections due to problematic MDR bacteria.¹⁶

Conclusions

The increasing prevalence of MDR Enterobacteriaceae limits available treatment options for infections caused by these organisms. Given the lack of new antibiotics, there is a valid rationale for testing older antibiotics that retain some activity against MDR bacteria.

Of these, fosfomycin trometamol appears to be a possible option because of its broad spectrum of activity against both Gram-positive and Gram-negative bacteria. Microbiological data appear to support this possibility, and susceptibility percentages are very high in K. pneumoniae and, mostly, in E. coli.

Conflict of interest

The authors state that they have no conflicts of interest to declare.

Funding

No external funding was obtained for this article; all authors take full responsibility for the content.

Notes on contributors

Annarita Mazzariol earned her PhD in Basic Microbiology at the University of Verona. She is currently an associate professor in the same University and is responsible for the MDR screening unit at the joint Clinical Microbiology Laboratories of Verona. Her main areas of interest is bacterial resistance mechanisms, with special focus on beta-lactamases.

Alda Bazaj has a masters in Bioinformatics and Medical Biotechnologies from the University of Verona, Italy. She attended the Department of Microbiology of the Peking University Medical Hospital, (with a scholarship from the European Society of Clinical Microbiology and Infectious Diseases) and the Unité des Rickettsies, Faculté de médecine, Marseille (with a grant from the Université de la Meditérranée), She has now started her PhD course in Applied Life and Health Sciences at the University of Verona.

Giuseppe Cornaglia obtained a degree in Medicine and Surgery from the University of Sassari, Italy. He is a full professor of Microbiology in the University of Verona and leads the joint Clinical Microbiology laboratories of Verona. He has been President of the European Society of Clinical Microbiology and Infectious Diseases, ESCMID (2007–2009 and 2010–2012) and is currently the ESCMID Director for International Affairs and the Chairman of the ESCMID Study Group for Antimicrobial Resistance Surveillance (ESGARS). A specialist in antimicrobial resistance epidemiology and related research, he received an Honorary Degree in Public Health from the St. Elizabeth University of Bratislava (Slovakia) for his ‘outstanding contribution to Public Health in the field of antimicrobial resistance research in Eastern Europe’.