Enterobacteriaceae

OVERVIEW: What every clinician needs to know

Pathogen name and classification

The family Enterobacteriaceae consists of a number of species that are gram-negative bacilli (GNB). Salmonella, Shigella, and Yersinia are not discussed here.

Most of the other Enterobacteriaceae cause a wide variety of extra-intestinal infections. Edwardsiella tardi can cause both extra-intestinal and intestinal infection. Klebsiella causes extra-intestinal infection, but a hemorrhagic colitis has been associated with Klebsiella oxytoca.

Members of this group are becoming highly resistant to antimicrobials, and some members are professional pathogens capable of infecting both healthy and compromised hosts.

A knowledge of their clinical presentations and treatment options is requisite for optimal outcome.

Escherichia coli

commensal

extra-intestinal pathogenic

intestinal pathogenic (not discussed here)
Klebsiella
Proteus
Enterobacter
Serratia
Citrobacter
Morganella
Providencia
Cronobacter
Edwardsiella

What is the best treatment?

General Treatment Principles

The following treatment principles should be followed:

Knowledge of local antimicrobial susceptibilities is critical for optimal care.

Antimicrobial resistance profiles vary with geographic location, hospital, local antimicrobial use, and site within the healthcare facility (e.g., ICU versus ward).

Local susceptibilities may differ from published data.

There may be lag between local “real-time” resistance rates and published data.
Evidence supports the concept that early use of appropriate empirical therapy improves outcome in serious infections due to GNB.

The “window” for initiation of early therapy remains incompletely defined

It is important to note that empirical use of certain agents should be reserved for appropriate patients at risk for multi-drug resistant (MDR) strains and/or those who are seriously ill, because over-use of these agents will limit their long-term usefulness (e.g., carbapenems, colistin, ceftazidime-avibactam).

Initial use of two potentially active agents for appropriately selected patients may be prudent to maximize the chances of a least one being active while waiting for susceptibilities.

Predictors of resistance include recent antimicrobial use, a healthcare association, or international travel.
When possible, the narrowest appropriate antimicrobial agent should be used and de-escalation should be implemented as soon as possible. This will:

Minimize selection of and possible superinfection with MDR-GNB

Maximize the longevity of the antimicrobial agents

Decrease the likelihood of developing Clostridium difficile infection

Break or at least slow down the ever-escalating cycle of selecting for increasingly resistant bacteria
Infected fluid collections (abscesses) should be drained, and infected foreign bodies should be removed if possible.
In addition, it is critical to not treat patients who are merely colonized but not infected. For example:

Asymptomatic bacteriuria (except for pregnant women and patients undergoing urologic instrumentation)

Sputum cultures positive for a GNB (or other pathogen) without evidence for pneumonia
The number of effective antimicrobial agents to treat certain Enterobacteriaceae is shrinking. True pan-resistant GNB exists. Recently ceftolozane/tazobactam and ceftazidime/avibactam were approved and active against MDR-GNB. The likelihood of additional new agents coming onto the market in the short- to mid-term is small. The antimicrobials currently available must be used judiciously.

Antimicrobial Resistance Mechanisms

An understanding of antimicrobial resistance mechanisms is important for the logical treatment of GNB infection, especially since the genes that encode antimicrobial resistance are often mobile and can move between strains and species. Further, there is a likelihood that species that have not acquired certain resistance genes will do so in the future, since resistance genes are often on mobile elements (e.g., plasmids). The recent emergence of the colistin resistance gene mcr-1 on a stable, transferrable plasmid is such an example and extremely concerning since polymyxins (polymyxin B and E [colistin]) are a last line of defense against strains that produce metallo-carbapenemases (e.g., NDM1).

Major mediators of antimicrobial resistance in the Enterobacteriaceae include:

broad-spectrum beta lactamases (BSBL)
extended-spectrum beta lactamases (ESBL)
AmpC beta lactamases
carbapenemases
decreased permeability
efflux pumps
modifying enzymes
alterations in target enzymes

For the Enterobacteriaceae, understanding the implications of a strain possessing the various beta lactamases and/or carbapenemases has the greatest implications for treatment, especially since many of the other resistance determinants are linked or associated with their presence.

BSBL

Confer resistance to many penicillins and first-generation cephalosporins (generation cephalo) via beta-lactam ring cleavage.
second- to fourth-generation cephalosporins, carbapenems, and aztreonam are usually resistant to hydrolysis by BSBL; (please note the cephamycins [e.g., cefoxitin] are considered second-generation cephalosphorins).
Examples include TEM and SHV.
These beta-lactamases are inhibited by beta-lactamase inhibitors (e.g., clavulanate, sulbactam, tazobactam [tazo], avibactam [avi]).

ESBL

ESBL are modifications of BSBL and include TEM-, SHV-, CTX-M-, and OXA-derived.
Confer resistance to the same antimicrobials as BSBL plus third- and sometimes fourth-generation cephalosporins and aztreonam.
These are often present on large, transferrable plasmids with linked or associated resistance to the fluoroquinolones levofloxacin and ciprofloxacin (hereafter referred to as FQ), trimethoprim-sulfamethoxazole (TMP-sulfa), aminoglycosides, and tetracyclines.
Activity of beta-lactam/beta-lactamase inhibitors are variable (tazo appears to be most potent inhibitor).
ESBL-containing strains often have porin mutations resulting in decreased permeability, making susceptibility of second-generation cephalosporins and beta-lactam/beta-lactamase inhibitors unpredictable.
Prevalence is increasing worldwide, particularly CTX-M types.
It is most prevalent in E. coli and K. pneumoniae and oxytoca.
Prevalence is also increasing in other Enterobacteriaceae.
16%, 38.6%, and 10.7% of Enterobacteriaceae strains from short-term care, long-term care, and inpatient rehab facilities demonstrated the ESBL phenotype respectively (US National Healthcare Safety Network [NHSN] data from 2008-2014).
Although ESBL containing strains are wide-spread in the United States, international travel, particularly to Asia or the Middle East, increases the likelihood of colonization with these strains.
The incidence of uncomplicated cystitis in healthy ambulatory women due to ESBL-containing E. coli is increasing worldwide, including within the United States.
Human acquisition of E. coli strains containing CTX-M is associated with the use of third- and fourth-generation cephalosporins and FQ in food animals.
Detection methods have improved.
Oral treatment options for ESBL-containing strains are limited to fosfomycin and nitrofurantoin as the best empirical choices for some types of urinary tract infections.
Carbapenems are the parenteral class of choice.
Clinical experience with second line agents is limited.

ceftazidime-avibactam

piperacillin-tazobactam (if MIC is <4 µg/mL) dosed at 4.5 grams q6h

ceftolozane/tazobactam

tigecycline (low urine & blood concentrations achieved; Proteus, Providencia, and Morganella inherently resistant)

polymyxins B and E (colistin)

FQ and aminoglycosides may be effective if sensitive, data are limited

AmpC beta lactamases

This can be considered as an ESBL but has some important fundamental differences.
Confer resistance to the same antimicrobials as ESBL plus second-generation cephalosporins.
These are primarily chromosomal enzymes present in nearly all strains of Enterobacter, Serratia, Citrobacter, Providencia, Morganella, and Proteus vulgaris at low constitutive (steady state) levels.
Occasionally, strains of E. coli, K. pneumoniae, and other Enterobacteriaceae possess plasmids expressing AmpC beta lactamases.
High-levels of AmpC production can be induced or stably de-repressed mutants by exposure to beta-lactams (e.g., third-generation cephalosporins), resulting in the development of resistance during therapy and treatment failure; risk greatest with E. cloacae and E. aerogenes, less with S. marcescens and C. freundii, and lowest with Providencia and M. morganii.
Carbapenems and ceftazidime-avibactam are active.

Cefepime (fourth-generation cephalosporin) may be an option with source control and exclusion of the concomitant presence of an ESBL.

FQs, piperacillin-tazobactam, TMP-SMX, tigecycline, and aminoglycosides, if isolates are susceptible in vitro, although clinical data are limited.

Ceftaz-avi & ceftolozane/tazo are active in vitro, but clinical data are limited.

Carbapenemases

The resistance genes bring us very close to Armageddon and include KPC, NDM-1, OXA-48, IMP, VIM.
Confer resistance to the same antimicrobials as ESBL plus second-generation cephalo and carbapenems.
These often present on large, transferrable plasmids with linked or associated resistance to FQ, TMP-sulfa, aminoglycosides, and tetracyclines.
The NDM-1 containing plasmid has a broad host range and contains fourteen, linked antimicrobial resistance genes.
As a result, some strains are nearly pan-resistant.
These genes are becoming increasing prevalent in Enterobacteriaceae on a worldwide basis, particularly in the Middle East and Asia.
The spread of NDM-1 from India around the world has generated significant concern.
KPC was first described in North Carolina in 1996, have been present in New York City since the late 1990s, and are increasing in California.
The highest prevalence is in K. pneumoniae and E. coli, but is described in nearly all Enterobacteriaceae.
2.8%, 12%, and 1.9% of Enterobacteriaceae strains from short-term care, long-term care, and inpatient rehab facilities are carbapenem resistant respectively (US National Healthcare Safety Network [NHSN] data from 2008-2014).
Automated susceptibility systems may fail to detect carbapenemases, and ertapenem resistance is the most sensitive marker for these systems.
A modified Hodge test or polymerase chain reaction (PCR) testing are more reliable detection systems.
Ceftazidime-avibactam is active in vitro against the serine carbapenemases (e.g., KPC, OXA-48), but not the metallo-carbapenemases (e.g., NDM1, VIM, IMP), but clinical data is limited.
Tigecycline and colistin are active agents but,

Tigecycline achieves low urine and blood concentrations, raising concerns for treating bacteremia and urinary tract infection (UTI).

Colistin is nephrotoxic (0-58%, dose dependent and potentially reversible).

Clinical studies on these agents are uncontrolled, retrospective, use variable regarding dosing and duration of treatment and often simultaneous administration of other antimicrobials.
Resistance is emerging to both of these agents.
Highly concerning is the recent emergence and dissemination of plasmid-mediated (colistin/polymyxin) resistance to the Enterobacteriaceae via modification of lipid A. Enterobacteriaceae-exhibiting plasmid-mediated resistance to polymixin and colistin have now been recovered on 5 continents and will likely result is more widespread infection over time.
Aminoglycosides may have treatment utility against some sensitive strains.
Fosfomycin was active against more than 90% carbapenemase-producing Enterobacteriaceae in an in vitro evaluation, including extensively-resistant isolates, but clinical data (for Enterobacteriaceae) is limited and concerns exist for resistance developing with monotherapy.

Specific Treatment Considerations for the Various Enterobacteriaceae

Extra-intestinal pathogenic E. coli (ExPEC) infections

A. Outpatient oral therapy of uncomplicated cystitis

most common E. coli infection
IDSA recommendations updated 2011

first-line agents

TMP-sulfa (DS tab BID x 3 d) if resistance is less than 20%

pivimethicillin (400mg BID x 5d)

nitrofurantoin (100mg BID x 5d)

fosfomycin (3 grams x single dose)

second-line agents

FQ x 3d

amoxicillin/clavulanate (amox/clav) x 3-7d

cefpodoxime or cefixime x 5-7d

Do not use ampicillin (amp) or amoxicillin (amox).

some issues with IDSA recommendations

TMP-sulfa, amox/clav, and even FQ resistance are greater than 20% in many locales.

Pivimethicillin is not available in United States; efficacy is an estimated 73%.

Consideration should be given to reserve fosfomycin use for ESBL-containing strains.

author’s recommendations

nitrofurantoin (but should not be used if pyelonephritis is a concern)

TMP-sulfa or FQ if local susceptibilities are known and greater than 80%

cefpodoxime or cefixime (expensive, clinical efficacy data limited).

Fosfomycin should be reserved to treat ESBL producing strains.

B. Inpatient parental treatment

Although the majority of E. coli strains are still relatively antimicrobial friendly in the United States, particularly in young, antibiotic naive, healthy individuals isolates are increasingly resistant (e.g., 17.1% FQ resistance in US outpatient isolates in 2010).
MDR strains are on the rise, largely because of the acquisition of ESBL and to a lesser degree carbapenemases, but NDM-1 may change this.
Prior antimicrobial use, recent or ongoing hospital care, and international travel, especially to the Middle East or Asia, increase the likelihood of MDR-strains.
NHSN data reported that 19% of central line-associated bacteremia isolates were resistant to third- and fourth-generation cephalosporins, 41.8% to FQ.
Outside of the United States, the International Nosocomial Infection Control Consortium (INICC) reported a significantly higher (66%) resistance rate for third-generation cephalosporins, 53.4% for FQ for ICU isolates from 2004 to 2009.
For stable, non-critically ill patients, empirical use of a third-generation cephalosporin or piperacillin/tazobactam (pip/tazo) is reasonable.
For critically-ill patients, amikacin, carbapenems, and and ceftaz-avibactam are the most reliable empirical agents.
Carbapenemase-producing strains are increasing (1-5% among health care-associated isolates in the US, higher in many other countries).

Tigecycline and the polymyxins, with or without a second agent, have been used most frequently for these extremely resistant isolates.

Do not use amp, amp/sulbactam, cephalexin, TMP-sulfa, or FQ empirically, unless local data demonstrates resistance rates less than 10%.

Klebsiella

Generally. Klebsiella are more antimicrobial resistant than E. coli.
NHSN data reported that 29% of central line-associated bacteremia isolates were resistant to third- and fourth-generation cephalosporins in 2009-20010.
INICC reported a 76% resistance rate in ICU isolates for third-generation cephalosporins from 2004 to 2009.
MDR strains are more prevalent because of the acquisition of ESBL and are more likely to express carbapenemases.
Carbapenem use has resulted in an increasing prevalence of carbapenemase-producing strains.
80-90% of carbapenem resistant strains in the United States are Klebsiella (primarily KPC, but recently NDM-1, OXA-48 and VIM as well).
For stable, non-critically ill patients, empirical use of piperacillin/tazobactam (pip/tazo) is reasonable.
For critically-ill patients, amikacin, carbapenems and ceftaz-avi are the most reliable empirical agents.
Carbapenemase-producing isolates should be treated with tigecycline and/or colistin or ceftazidime-avi bactyam (for KPC and OXA-48, but NOT NDM-1, VIM, or IMP). Combination therapy may be beneficial.
Fosfomycin is active in vitro against about 90% of carbapenemase-producing isolates, but clinical data are sparse and a parenteral formulation is not available in the United States.
Near pan-resistant strains (carbapenemase-producing, colistin, and/or tigecycline resistant) strains have been described in the United States and internationally.
Intrinsically resistant to amp, ticarcillin and nitrofurantoin have poor activity.
Do not use amp/sulbactam, cephalexin, TMP-sulfa, or FQ empirically, unless local data demonstrates resistance rates greater than 10%.

Proteus

The majority of strains are “antimicrobial friendly.”
MDR strains that produce ESBL are presently uncommon (about 5%), and carbapenemase producers are rare.
10-50% of P. mirabilis isolates are resistant to amp and cephalexin.
P. vulgaris and penneri are more resistant than P. mirabilis, and these former species possess AmpC beta-lactamases, resistance to amp and cephalexin is the rule, and 30-40% are resistant to FQ.
For stable, non-critically ill patients, empirical use of a third-generation cephalo or piperacillin/tazobactam (pip/tazo) is reasonable; however, one needs to be aware that the use of a third-generation cephalo for treating P. vulgaris, and penneri may induce the endogenous AmpC beta-lactamse or select out a stably de-repressed AmpC mutant.
For critically-ill patients, carbapenems, ceftaz-avi, amikacin, ceftolozane/tazo and cefepime are most reliable (>90% susceptibility).
Ceftaz-avi & ceftolozane/tazo are active in vitro, but clinical data are limited.
Do not use tetracyclines, tigecycline, nitofurantoin, or polymyxins.

Enterobacter and Cronobacter

These genera can be highly antimicrobial resistant.
NNIS reported that 37.4% of Enterobacter spp. CLABSI isolates were resistant to third-generation cephalosporins.
INICC reported a 57% resistance rate for third-generation cephalosporins from 2002 to 2007.
All strains possess AmpC beta-lactamases, which can be inducted by exposure to third-generation cephalosporins.
Third-generation cephalosporins should be avoided for the treatment of serious infections.
Cefepime is stable in the presence of AmpC beta-lactamases and is a suitable treatment option in the absence of ESBL.
Unfortunately, the prevalence of ESBLs in Enterobacter, especially E. cloacae, is increasing (5-50%).
Carbapenemase prevalence is low, but NDM-1-containing isolates are described.
For stable, non-critically ill patients, empirical use of cefepime, FQ, or pip/taz is reasonable.
For critically-ill patients, carbapenems, amikacin, and cefepime are most reliable.
Ceftazidime-avibactam & ceftolozane/tazobactam are active in vitro, but clinical data are limited.
Do not use amp, amp-sulbactam, and the first- and second-generation cephalosporins.

Serratia

The majority of strains are “antimicrobial friendly.”
MDR strains that produce ESBL are presently uncommon (<5%), but rates of 20-30% have been reported in Asia and Latin America .
Carbapenemase-producers are rare, but increasing.
FQ resistance is variable, ranging from 15 to 30%.
Although resistance to tigecycline remains uncommon, Serratia has a narrow susceptibility window and decreased permeability or efflux may confer resistance.
For stable, non-critically ill patients, empirical use of a third-generation cephalo, piperacillin/tazobactam (pip/tazo), aztreonam, or TMP-sulfa is reasonable; however, one needs to be aware that the use of a third-generation cephalo may induce the endogenous AmpC beta-lactamase or select out a stably de-repressed AmpC mutant.
For critically-ill patients, carbapenems, amikacin, and cefepime are most reliable (>90% susceptibility).
Ceftazidime-avibactam & ceftolozane/tazobactam are active in vitro, but clinical data are limited.
Do not use amp, amp/sulbactam, first-generation cephalosporins, cephamycins, tetracyclines, nitrofurantoin, and polymyxins.

Citrobacter

C. freundii is much more antimicrobial resistant than C. koseri.
MDR strains that produce ESBL are presently uncommon (<5%), and carbapenemase-producers are rare.
Amp/sulbactam, third-generation cephalosporin, aztreonam, FQ resistance is variable but increasing.
For stable, non-critically ill patients, empirical use of a third-generation cephalo, piperacillin/tazobactam (pip/tazo) is reasonable; however, one needs to be aware that the use of a third-generation cephalo may induce the endogenous AmpC beta-lactamase or select out a stably de-repressed AmpC mutant (but not for C. koseri).
For critically-ill patients, carbapenems, amikacin, and cefepime are most reliable (>90% susceptibility).
Ceftazidime-avibactam & ceftolozane/tazobactam are active in vitro, but clinical data are limited.
Do not use amp, first- and second-generation cephalosporins.

Morganella and Providencia

These species can be significantly antimicrobial resistant.
MDR strains that produce ESBL are presently uncommon (<5%), carbapenemase-producers are rare, but NDM-1-containing isolates have been described in Morganella.
Resistance to second- and third-generation cephalosporins, azreonam, TMP-sulfa is variable but increasing.
FQ resistance is significant (>30-40%).
For stable, non-critically ill patients, empirical use of a third-generation cephalo, piperacillin/tazobactam (pip/tazo) is reasonable; however, one needs to be aware that the use of a third-generation cephalo may induce the endogenous AmpC beta-lactamse or select out a stably de-repressed AmpC mutant.
For critically-ill patients, carbapenems, amikacin, and cefepime are most reliable (>90% susceptibility).
Ceftazidime-avibactam & ceftolozane/tazobactam are active in vitro, but clinical data are limited.
Do not use amp, amp/sulbactam, first-generation cephalosporins, tigecycline, tetracyclines, nitrofurantoin, fosfomycin, and polymyxins.

Edwardsiella

This is susceptible to most gram-negative active agents.
Gastroenteritis is often self-limiting, but FQ may hasten resolution.
For critically-ill patients, FQ, third- and fourth-generation cephalosporins (cefepime), carbapenems, and amikacin alone are most reliable (>90% susceptibility).

How do patients contract this infection, and how do I prevent spread to other patients?

Epidemiology

The Enterobacteriaceae are part of the normal animal and human intestinal flora and/or environmental loci including health-care facilities.
In healthy humans, commensal and extra-intestinal pathogenic E. coli (ExPEC) are the predominant GNB in the colonic microbiota; Klebsiella and Proteus are less prevalent.
In healthy humans, E. coli, Klebsiella, and Proteus only transiently colonize the oropharynx and skin.
By contrast, in hospitalized patients and, to a lesser degree, in residents of long-term-care facilities (LTCF), the Enterobacteriaceae emerge as the dominant flora of the oropharynx and skin, which, in turn, leads to an increased incidence of pneumonia and soft-tissue infections by these GNB in these settings.
LTCF are an important reservoir for resistant strains
Serratia and Enterobacter infection may rarely be acquired through a variety of infusates, including medications and blood products.
Edwardsiella infections are acquired through freshwater and marine environment exposures and are most common in Southeast Asia.
Antimicrobial use, co-morbidities, and extended length of hospital stay are associated with increased colonization with these GNB.
As a result of these epidemiologic features, they are global pathogens.
Generally, there are no seasonal differences in the incidence of infection.
The incidence of infection with these GNB is increasing due to an aging population and increasing antimicrobial resistances.

Infection Control

Antimicrobial stewardship programs should be implemented to minimize the development of antimicrobial resistance.
Inanimate objects (e.g., stethoscopes, blood pressure cuffs), the hands of health-care workers, and many other objects may be colonized with these GNB, theoretically enabling transmission to patients.
Therefore, it is critical for health-care personnel to be diligent about hand hygiene and appropriate cleaning/disinfection of any objects that come into contact with patients. Room dedicated instruments for patients on contact precautions are reasonable.
Contact precautions should be used for patients colonized or infected with carbapenem-resistant GNB and perhaps other multi-drug resistant (MDR)-GNB (e.g., those containing extended-spectrum beta-lactamases [ESBL]).
Avoiding or minimizing the duration of use of indwelling devices, such as intravascular and urinary catheters and endotracheal tubes, decreases the risk of infection.
When indwelling devices are required, an appropriate protocol should be used for placement, and certain infection control measures may decrease the incidence of subsequent infection and protocols for daily use, evaluation, and removal as soon as possible should be implemented.
Multi-use medication vials should be avoided if possible.
Presently, there are no vaccines available for the prevention of infection due to Enterobacteriaceae (except Salmonella typhi, which is not discussed here).
Prophylaxis should not routinely be used to prevent infection because of the Enterobacteriaceae, except for select patient groups (e.g., FQ prophylaxis) should be considered for high risk patients (anticipated absolute neutrophil count less than100 cell/mm³ for more than 7 days) undergoing chemotherapy.
Increasing data support the implementation of universal decolonization to prevent infection in ICU patients.

What host factors protect against this infection?

The innate immune system is the critical first line of defense in protecting against infections due to Enterobacteriaceae.

The Enterobacteriaceae discussed here are primarily extracellular pathogens.

Innate immunity (e.g., physical barriers, complement, antimicrobial peptides, professional phagocytes (neutrophils, monocytes, macrophages)) and humoral immunity (opsonic antibodies that activate, complement, and enhance phagocytosis) are the critical host defense systems.

The host’s innate immune system initially detects the presence of the invading pathogen by recognizing conserved bacterial motifs termed pathogen-associated molecular patterns (PAMP). Common PAMP for extracellular pathogens are lipidA, certain lipoproteins, and flagellin.

PAMP are recognized by host defense signaling systems, termed pattern or pathogen recognition receptors (PRR), which initiate the innate host defense response. Toll-like receptors (TLR) and C-type lectin receptors are the most important PRR for extracellular bacterial pathogens. The extracellular TLR are TLR 1, 2, 4, 5, and 6.

TLR4 lipopolysaccharide, mannans, and others

TLR1-2 triacyl lipopeptides and others

TLR2-6 diiacyl lipopeptides and others

TLR5 flagellin

Signaling initially results in the activation/production of inflammatory cytokines, which, in turn, mediate the initial inflammatory response that includes a wide range of host responses, including:

febrile response

inflammatory response proteins (e.g., C-reactive protein, procalcitonin, complement components, transferrin, and anticoagulants, such as proteins S and C)

neutrophil migration

antibody and cell-mediated responses

If host tissue damage ensues, these signaling systems are further activated by host-derived motifs (e.g., heat shock proteins, mitochondrial components), termed damage or danger-associated molecular patterns (DAMP).

Uncontrolled production of PAMP and DAMP may lead to an uncontrolled host immune response, leading to severe sepsis, septic shock, or death.

LipidA is an established mediator of severe sepsis and septic shock.

It is likely other PAMP and DAMP are contributory.

Pre-existing antibodies facilitate bacterial clearance and prevent or control infection. However, if the host is immunologically naive, in the absence of antimicrobial treatment, outcome is often decided by the effectiveness of initial innate response. Subsequent activation of the clonal, pathogen-specific adaptive response (e.g., antibody development) likely serves to “mop-up” residual bacteria and to protect against future bacterial challenges.

Therefore, deficiencies or dysfunction of the components of these systems increases the risk and severity of infection. Examples include:

Any cause of neutropenia or neutrophil dysfunction (e.g., steroids), particularly chemotherapy induced neutropenia and mucositis
Complement dysfunction or deficiency (particularly C3, C1qrs, C4, and C2 for encapsulated bacteria and C5,6,7 (attack complex)
Urinary catheter, endotracheal tube, and skin breaks (e.g., traumatic, surgical incisions, ulcers), which by-pass physical barriers
Splenic dysfunction or splenectomy (filters pathogens opsonized with antibodies or complement from the blood) has been associated with over-whelming infection due to various pathogens, particularly encapsulated bacterial pathogens. Although S. pneumoniae is most common, E. coli and other Enterobacteriaceae have also caused this syndrome.
Hepatic macrophage dysfunction, a decrease in complement production, and portal hypertension have been implicated as contributing causes of spontaneous bacterial peritonitis in cirrhotics.
TLR5 polymorphism results in susceptibility to recurrent cystitis (nearly all cases of which are due to E. coli).

By contrast to gram-positive pathogens, metastatic spread of the Enterobacteriaceae to bone, joints, and native heart valves is uncommon and, even then, usually requires a predisposing condition, such as:

Neutropenia may lead to endophthalmitis or ecthyma gangrenosum.
A physically abnormal heart value may enable the development of the rare circumstance of native valve endocarditis due to Enterobacteriaceae.

An exception to this general rule appears to be a new hypervirulent variant of K. pneumoniae that is highly metastatic.

The cell-mediated immune response is generally unimportant in directly protecting against extracellular pathogens (excepting its role in antibody development), such as the Enterobacteriaceae, since this host defense requires the recognition of HLA receptors, which are not present on bacteria.

Histopathology of infection

Given the extracellular location of these pathogen and the importance of the innate host response histologically:

A neutrophilic response is usually observed at the site of infection.
Tissue necrosis (due to an over-exuberant host response +/- microbial toxins) is seen with increasing severity of infection.
The Enterobacteriaceae are capable of producing gas with growth; therefore, gas may be seen at the site of infection, and this usually reflects a severe infection (e.g., emphysematous cystitis, pyelonephritis, cholecystitis).

What are the clinical manifestations of infection with this organism?

Clinical Syndromes and Manifestations Broadly Applicable to the Enterobacteriaceae

The Enterobacteriaceae are capable of infecting both healthy and compromised hosts.

The Enterobacteriaceae discussed here cause extra-intestinal infection (except Edwardsiella, which can cause extra- and intra-intestinal infection, and perhaps a toxin-producing variant of Klebsiella).

Infections involve nearly every extra-intestinal anatomic site and organ, and new infectious syndromes are emerging (e.g., those due to the new hypervirulent variant of K. pneumoniae).

The relative likelihood of infecting certain hosts and causing specific infectious syndromes is variable, depending on the specific genus/species.

ExPEC and Klebsiella are the most virulent, being able to cause infection in healthy hosts and cause the majority of infections.

Therefore, a knowledge of clinical presentations combined with predicted antimicrobial resistance will enable optimal early management decisions to be made.

Clinical Syndromes and Manifestations Specific to Members of Enterobacteriaceae

The more common clinical syndromes seen with each pathogen are listed below in decreasing order of likelihood along with a few clinical caveats.

E. coli

A. Commensal E. coli usually require an aggravating factor, a compromised host defense, or large challenge inoculum to cause infection; for example:

UTI in the presence of a urinary catheter
peritonitis after fecal contamination due to bowel disruption
cholangitis with biliary tract obstruction

B. ExPEC:

This is one of the most common bacterial agents to cause infection overall.

Since nearly everyone is colonized with ExPEC, entry, not acquisition is generally the limiting step for infection.
This is the most common cause of severe sepsis and septic shock. Bacteremia can arise from any site of infection. One-half of bacteremias arise in the community, and one-half arise in the hospital.
It is a professional pathogen that can readily infect a normal host.
The increasing number of ESBL and carbapenemase-producing ExPEC isolates is concerning, and, given their innate virulence, it is reasonable to anticipate that the incidence of nosocomial ExPEC infections may increase, since antimicrobial resistance is a critical enabler of infection in that setting.
The clonal group, ST131 (usually FQ-resistant & increasingly producing CTX-M ESBL) has undergone global dissemination.

UTI is the most common source for bacteremia, causing 50-67% of episodes.

cystitis, pyelonephritis, catheter-associated, prostatitis

accounts for 1% of ambulatory visits

second only to respiratory tract infections as an infectious cause for hospitalization

UTI in men is almost always preceded by urinary tract instrumentation.

An increasing number of cases of uncomplicated cystitis in ambulatory women is caused by ESBL-containing ExPEC, which makes oral treatment very challenging (fosfomycin and variably nitrofurantoin are often the only viable PO options).

abdominal and pelvic infection-source of 25% of bacteremias

peritonitis secondary to fecal contamination-often complicated by bacteremia

spontaneous bacterial peritonitis in the setting of cirrhosis

dialysis-associated peritonitis

diverticulitis

appendicitis

intra-peritoneal and visceral abscesses (hepatic, pancreatic, splenic)

infected pancreatic pseudocysts

septic cholangitis and cholecystitis-cholangitis with obstruction is often complicated by bacteremia

pneumonia

usually considered a rare cause but third or fourth most common GNB causing hospital-acquired pneumonia (5-8% of cases; due to increased oro-pharyngeal colonization in this setting)

Although community-acquired pneumonia due to Enterobacteriaceae is rare (2-5% of cases), ExPEC is common in this setting.

meningitis

One of the two leading causes of neonatal meningitis-neonatal sepsis (majority K1 serotype)

Rare settings after the neonatal period include disruption of the meninges due to trauma or neurosurgery or in the presence of cirrhosis.

cellulitis/musculoskeletal

primarily decubitii or ulcers of the lower extremity due to neurovascular compromise (e.g., diabetics); osteomyelitis occurs in this setting due to contiguous spread

cellulitis of surgical sites and infection of burn sites, especially when close to the perineum

discitis/vertebral osteomyelitis from hematogenous spread (more common than appreciated, up to 10% in some series)

occasionally, orthopedic-device related infection

occasionally, septic arthritis

rarely, hematogeous myositis (upper leg myositis or fasciitis should prompt an evaluation of an abdominal source with contiguous spread)

endovascular infection

native valve endocarditis rare, especially considering the frequency of ExPEC bacteremia, valves usually diseased

Prosthetic valve endocarditis, infection of vascular grafts (especially femoral), and aneurysms occasionally occur.

primary bacteremia

intravascular devices

transrectal prostatic biopsy

with increased intestinal mucosal permeability, such as occurs with neonates, severe burns, trauma, and chemotherapy induced mucositis and neutropenia

C. Intestinal pathogenic E. coli

This group of specific pathotypes causes gastroenteritis that presents with a variety of clinical manifestations that is somewhat pathotype-specific.
This pathogenic group of E. coli strains rarely, if ever, causes disease or invades outside of the gastrointestinal (GI) tract.

Klebsiella

K. pneumoniae is the most important species and can cause community, LTCF, and noscomial-acquired infection.

Two pathotypes of K. pneumoniae have emerged.

“Classical” K. pneumoniae, presently the most common cause of healthcare-associated infections, are becoming increasingly antimicrobial resistant, and appear to account for most infections in Western countries.
A new “hypervirulent” variant of K. pneumoniae is emerging, primarily in the Pacific Rim, but cases have been described worldwide.
K. pneumoniae, particularly the new hypervirulent variant, like ExPEC, is a professional pathogen that can infect healthy, ambulatory hosts.
Like ExPEC, the prevalence of MDR-Klebsiella strains is significantly increasing.
Clonal group ST258, many members of which produce the KPC carbapenemase is undergoing international dissemination; K. pneumoniae strains that contain the NDM-1 carbapenemase are also spreading worldwide.

A. “Classic” K. pneumoniae: Most infections occur in hospital or LTCF.

pneumonia- source of 15-30% of bacteremias

primarily occurs in healthcare setting

Mechanical ventilation is a major risk factor.

Community-acquired pneumonia (historically in alcoholics) is presently uncommon in the United States and Europe but is more common in Africa and Asia (albeit unclear if some of the causative strains may be the hypervirulent variant).
UTI- source of 15-30% of bacteremias

1-2% of episodes in healthy ambulatory hosts

5-17% of complicated UTI, particularly involving urinary catheters
Abdominal infection-source of 15-30% of bacteremias

similar to ExPEC but much less frequent
Other infections

similar to ExPEC but less frequent

B. “Hypervirulent” K. pneumoniae

These strains appear genetically distinct from classic K. pneumoniae, but the lack of an unequivocal genotypic/phenotypic marker(s) has precluded a comprehensive understanding of the prevalence and spectrum of disease; although a hypermucoviscous phenotype (+ string test) has been used as a surrogate.
Despite no clear surrogate marker, a hypermucoviscous phenotype (+ string test) has been associated with community-acquired hepatic abscess, a syndrome that has clinically defined the disease to date.
Community acquired infections, often associated with abscesses, have been primarily described up to now, but our understanding of the infectious spectrum of this pathogen is evolving, in part due to the lack of an optimal test for the clinical microbiology to differentiate cKP from hvKP.
Healthy hosts and often young hosts are commonly infected but also diabetics. Although cases have been described worldwide, the majority are from Asia. Therefore, it is unclear whether individuals of Asian descent are at higher risk or more commonly colonized.
A characteristic of hvKP is the propensity for metastatic spread from a site of infection (11-80% of cases); eyes, CNS and lungs are the most common, but virtually every site & organ has been described.

Hepatic abscess

This syndrome initially served as the distinguishing feature between “classic” and “hypervirulent” K. pneumonia.

community-acquired primary liver abscess

patients without a history of biliary disease

Pneumonia

Increasing incidence of K. pneumoniae-mediated community acquired pneumonia in Asia suggests the hypervirulent variant may be responsible.

Meningitis

Increasing incidence of K. pneumoniae-mediated community acquired meningitis in Asia and the fact that K. pneumoniae, up until now, almost never has been described as a cause of community acquired meningitis in adults suggests the hypervirulent variant may be responsible.

Non-hepatic abscess

community acquired

described in the absence of liver abscess

includes renal, deep neck, parotid gland, mycotic aneurysm, osteomyelitis, subdural empyema, and septic arthritis

Bacteremia

42% (83/200) in a series of community acquired bacteremia had a hypermucoviscous phenotype.

23% primary, source unknown

21% associated with pneumonia

11% associated with UTI

2% associated with spontaneous bacterial peritonitis

37% associated with pathognomic community-acquired primary liver abscess

C. K. oxytoca: K. oxytoca is primarily a pathogen in LTCF and hospital settings.

various extra-intestinal infections similar to K. pneumoniae that are primarily acquired in the healthcare setting

pneumonia, UTI, vascular-access device associated bacteremia most common
hemorrhagic colitis associated with cytotoxin-producing strains

D. K. pneumoniae subspecies rhinoscleromatis: genetically distinct from K. pneumoniae and causes disease in the tropics

rhinoscleroma

a granulomatous, slowly progressive (months-years) mucosal infection of the upper respiratory tract with resultant necrosis and occasional nasal obstruction.

E. K. pneumoniae subspecies ozaenae: genetically distinct from K. pneumoniae and causes disease in the tropics

chronic atrophic rhinitis
systemic infection-rarely described in immunocompromised hosts

Proteus

P. mirabilis causes 90% of infections, which are acquired in the community, LTCF, and hospital. P. vulgaris and P. penneri primarily cause health-care associated infection.
The generation of histamine from fish contaminated by these species can cause scromboid.

UTI- A high percentage of bacteremias originates from this site.

the site of the majority of Proteus infections

1-2% of uncomplicated cystitis in healthy women

5% of hospital-acquired UTIs

10-15% of complicated UTIs

20-45% of UTIs in chronically catheterized patients, often associated with biofilm formation

Proteus UTI can be complicated by infection stones, and stone removal is requisite for cure.

Recurrent Proteus UTI should prompt evaluation for urolithiasis.

Other infections

less common

types of infection similar to those described for ExPEC

Enterobacter and Cronobacter

E. cloacae (65-75%), E. aerogenes (15-25%), E. Cronobacter sakazakii (1%), and E. gergoviae (<1%) cause primarily health-care associated infections.
Widely prevalent in the environment, healthy humans are rarely colonized, but this likelihood increases in LTCF and hospital.
Colonization is usually requisite for infection, but, occasionally, infection results from direct infusion of contaminated fluids.
Co-morbid disease, prior antimicrobial therapy, and residence in an ICU increase the risk of infection.
The success of Enterobacter as pathogen is due, in part, to many strains being MDR. Its inherent pathogenicity is unclear but almost certainly less than ExPEC and Klebsiella.

Extra-intestinal infectious syndromes are similar to those already described for other Enterobacteriaceae, such as ExPEC. Most common are:

pneumonia

UTI, particularly catheter-related

intra-vascular device related

surgical site infection

post-operative abdominal infection or biliary stent related

meningitis s/p neurosurgical procedure, particularly intracranial pressure monitor-related

Bacteremia

usually originates from the site of infection

may be related to the infusion of contaminated fluids, blood components, and medications, and this possibility should be considered with outbreaks or an unclear source

rarely from neutropenia and mucositis

E. Cronobacter sakazakii is primarily a pathogen of neonates causing neonatal meningitis/sepsis/necrotizing enterocolitis.

particularly in preterms

contaminated formula epidemiologically linked as source in some cases

brain abscess or ventriculitis is a frequent complication

Serratia

S marcescens (>90%), S. liquefaciens, S. rubidaea, S. fonticola, S. grimesii, S. plymuthica, and S. odorifera primarily cause infection in the health-care setting (1-3% of hospital-acquired infections).
Some recent studies from Canada and Australia suggest community-acquired infections may be more common than appreciated.
Serratia is found widely in the environment (which includes hospitals), particularly in moist settings.
Similar to Enterobacter, colonization is rare in the healthy host but increases in the health-care setting.
Infection is usually sporadic, but epidemics may occur in ICUs related to MDR-strain or common-source outbreaks.
Infection results from either colonization or directly from reservoirs in hospital, which includes fingernails of healthcare personnel, food, milk in neonatal units, sinks, respiratory equipment, pressure monitors, IV solutions, parenteral medications (particularly those generated by compounding pharmacies), multiply accessed medication vials (e.g., heparin, saline), pre-filled syringes, hand soaps, blood products, irrigation solutions, and even disinfectants.

Extra-intestinal infectious syndromes are similar to those already described for other Enterobacteriaceae, such as ExPEC. Most common are:

pneumonia

UTI, particularly catheter-related

intra-vascular device related

surgical site infection

post-operative abdominal infection

soft-tissue (including myositis, fasciitis, mastitis)

septic arthritis, particularly related to intra-articular injection

contact-lens associated keratitis and other ocular infections

Bacteremia

from the site of infection

may be related to the infusion of contaminated fluids, blood components, and medications and, similar to Enterobacter, possibility should be considered with outbreaks or an unclear source

rarely from neutropenia and mucositis

Citrobacter

C. fruendii and C. koseri are responsible for the majority of Citrobacter infections, which occur primarily in LTCFs and hospitals (1-2% of hospital-acquired infections).
Citrobacter species are present in water, food, soil, and certain animals. Healthy humans are rarely colonized, but this likelihood increase in health-care settings.
Co-mordid disease, prior antimicrobial therapy, and residence in an ICU increases the risk of infection

Extra-intestinal infectious syndromes are similar to those already described for other Enterobacteriaceae, such as ExPEC. Most common are:

UTI, particularly complicated UTI or in the presence of an indwelling catheter (40-50% of cases)

biliary tree, especially with stones or obstruction

pneumonia

intra-vascular access device-related

surgical sites

soft-tissue infection, especially decubitii

Bacteremia

urinary tract, biliary tree or intra-vascular device-related infection are most common sources

rarely from neutropenia and mucositis

Citrobacter, particularly C. koseri is associated with neonatal meningitis/sepsis

particularly in preterms

brain abscess or ventriculitis is a frequent complication

Morganella and Providencia

M. morgannii, P. stuartii and less frequently P. rettgeri are the most common causes of human infection.
These pathogens are similar to Proteus in all respects except infection of healthy ambulatory hosts is uncommon.
They primarily infect residents of LTCF and less commonly hospitalized patients with indwelling urinary catheters.
Extensive use of polymyxins & tigecycline may select for these genera due to their inherent resistance to these agents.

Primarily urinary tract pathogens

most often associated with long-term catheterization (>30 days)

often results in biofilm formation and catheter encrustation, which may lead to obstruction

may lead to the development of “infection” (struvite) bladder or renal stones, which may lead to obstruction and renal damage

stones are foci for relapse and usually require removal for cure

Uncommon infections include:

surgical site

soft-tissue involving devitalized tissue or snake bites

ventilator-associated pneumonia

abdominal

intra-vascular access device

Bacteremia uncommon

majority from urinary tract

occasionally, the biliary tract, surgical site, or soft-tissue infection

Edwardsiella

E. tarda is the only member of the genus described to cause human infection.
Geographically and clinically, it is somewhat distinct from the other Enterobacteriaceae described here.
Found in freshwater and marine environments and the associated aquatic animal species and human infection, it occurs when interacting with these reservoirs.
Infections is rare in the United States. Most cases are described from Southeast Asia.
Clinical features consist of both intestinal infection and extra-intestinal infection, thereby sharing features more akin to Salmonella and Vibrio vulnificus.

Gastroenteritis

most common syndrome (50-80% of infections)

watery self-limiting diarrhea most common

occasionally more severe colitis

Wound infection

most common extra-intestinal infection

due to direct inoculation

associated with freshwater or marine injuries

associated with snake related injuries

primarily hepatic

occasionally intraperitoneal

Bacteremia

usually related to invasion of the GI tract

most individuals have co-morbidities, such as hepatobiliary disease, iron overload, diabetes mellitus

bacteremia may result in meningitis (40% mortality rate)

Miscellaneous Members of Enterobacteriaceae

Species of the genera Pantoea, Hafnia, Kluyvera, Cedecea, Leclercia, Photorhabdus, and Ewingella have been rarely isolated from diverse clinical specimens. These organism are capable of causing disease, and, although they have been occasionally described to cause infection in a normal host, most infections occur in a compromised host, in the presence of a foreign body (e.g., catheter), or after an invasive procedure.

Visceral abscesses

What common complications are associated with infection with this pathogen?

Complications Common to the Members of Enterobacteriaceae Discussed Here

Infection due to all of the Enterobacteriaceae discussed here has the potential to be complicated by sepsis, severe sepsis (sepsis plus organ failure distant from the site of infection), and septic shock.
Organ failure associated with severe sepsis includes:

renal failure

acute respiratory distress syndrome

disseminated intravascular coagulation

hepatic dysfunction

CNS dysfunction
The mortality rate is substantial when infections due to the Enterobacteriaceae are complicated by severe sepsis and/or septic shock (20-50%).
The mortality rate is also substantial from pneumonia due to Enterobacteriaceae (20-50%).
Mortality rates tend to correlate with severity of illness.
Further outcomes are generally worse when inappropriate therapy is given, and this occurs more frequently with the recent significant increase in prevalence of MDR-Enterobacteriaceae.
The Enterobacteriaceae are common causes of soft-tissue infection of compromised or devitalized tissue, such as decubitus or diabetic ulcers. Contiguous osteomyelitis is an all too common complication often not recognized. A delay is diagnosis may require surgery or more extensive surgery than originally needed for cure. MRI is the most specific and sensitive test for making this diagnosis.

Pathogen-specific Complications

Hypervirulent pathotype of K. pneumoniae: Infection due to the new hypervirulent variant of K. pneumoniae is often complicated by metastatic spread to various organs and sites, including endophthalmitis and CNS infection with catastrophic results (e.g., loss of vision, neurologic sequelae). This often occurs in healthy hosts. Metastatic spread is an unusual complication for GNB (other than the hypervirulent Klebsiella variant) in general, including the Enterobacteriaceae, unless the host is immunocompromised (e.g., prolonged neutropenia) and even then is unusual.
Proteus, Morganella, and Providencia: UTI can result in the development of “infection stones” and, if untreated (both the stones and bacteria), can lead to obstruction and renal failure.
Citrobacter koseri and E. Cronobacter sakazakii: Although brain abscess and/or ventriculitis are potential complications of neonatal meningitis/sepsis from ExPEC, Klebsiella and other Enterobacteriaceae, these complications occur at a much higher rate with Citrobacter.

How should I identify the organism?

Specimen Collection

Specimen collection for microscopic examination and culture should be performed prior to the initiation of antimicrobial therapy.
Blood cultures or cultures from potential sites of infection (e.g., urine, sputum, body fluids, and other potential sites of infection) should be obtained as appropriate.

Microscopic Examination of Specimen

On Gram-stain, they appear as non-spore-forming gram-negative bacilli; however, this morphology is not specific for the Enterobacteriaceae.

Microbiologic Isolation

Enterobacteriaceae are readily isolated on routine media (blood, chocolate, MacConkey plates) designed to sustain the growth of gram-negatives.
Enterobacteriaceae are relatively rapid growers and are usually isolated within 18-24 hours.
Enterobacteriaceae can grow aerobically and anaerobically.
Prior antimicrobial therapy may inhibit or slow growth.

Microbiologic Identification

Tentative identification by an experienced microbiologist is based on colony morphology, lactose fermentation, indole production, and sometimes hydrogen sulfide production, and these data may assist in guiding empirical therapy pending final identification.
Characteristics of the various genus/species are as follows:

E. coli

more than 90% are rapidly lactose + and indole +.

Klebsiella

K. pneumoniae: lactose +, indole –

Hypervirulent K. pneumoniae: its hyperviscous capsule (+ string test, formation of a viscous string >5mm in length when bacterial colonies on an agar plate are stretched by an inoculation loop) has been used to differentiate this pathotype from classical K. pneumoniae, but this test lacks optimal sensitivity/specificity; especially in low-prevalence areas

K. oxytoca: lactose +, indole +

K. rhinoscleromatis and K. ozaenae: lactose -, indole –

Proteus – characteristic swarming motility on agar plates

P. mirabilis: usually lactose -, indole -, hydrogen sulfide +

P. vulgaris and P. penneri: usually lactose -, indole +, hydrogen sulfide +

Enterobacter- usually lactose + (sometimes delayed), indole –

Serratia- some strains of S. marcescens and S. rubidaea are red-pigmented

S. marcesans: usually lactose -, indole –

S. liquefaciens and S. rubidaea: usually lactose + (sometimes delayed), indole –

Citrobacter

C. fruendii: variable lactose reaction, indole -, oxidase –

C. koseri: variable lactose reaction, indole +, oxidase –

Morganella and Providencia- lactose -, indole +

Edwardsiella- lactose -, indole +, hydrogen sulfide +
Enterobacteriaceae can reduce nitrates to nitrites, and this test is part of many urine dipsticks. However, a false-negative or positive-test can occur.
More rapid methods (e.g., PCR-based, MALDI-TOF MS) for the identification of the genus-species and antimicrobial resistance are under development but presently not in commercial use.

Significance of Microbiologic Isolation

Isolation of Enterobacteriaceae from an ordinarily sterile anatomic site (e.g., blood, CSF, pleural or peritoneal fluid) is almost always clinically significant.
By contrast, isolation from non-sterile sites, such as open wounds and the respiratory tract (sputum), requires clinical correlation to differentiate colonization from infection.

How does this organism cause disease?

Principles of Pathogenesis Common to Members of Enterobacteriaceae

Disease manifestations are a combination of the site (or sites) of infection, bacterial factors, and the host response to the following factors:

In an optimal situation, from the host’s point of view, entry of Enterobacteriaceae into an organ or site of infection results in an appropriate host response of sufficient magnitude to clear the inciting pathogen and is then down-regulated resulting in minimal collateral host damage.
In severe or fatal infection, some combination of unchecked bacterial growth, which can be due to an attenuated or appropriate host response, and/or an over-exuberant host response occurs. Tissue damage is due to histotoxic bacterial products or host factors designed to eradicate the infecting pathogen (e.g., reactive oxygen species and histotoxic compounds in the granules of neutrophils).
All of the members of the Enterobacteriaceae possess many different antigenic variants of their major surface components (capsule and the LPS).

This permits evasion of humoral immunity.

This enables recurrent infections.

This has impeded vaccine development.

The steps leading to bacterial infection can be organized into a series of steps outlined below. These steps may occur sequentially or in combination.

Acquisition

The human host is normally colonized with ExPEC and, less commonly, with P. mirabilis and K. pneumoniae.

Acquisition occurs sometime after birth, usually with the cessation of breast feeding. Strains are acquired from close contacts, environmental objects, and food, since Enterobacteriaceae also colonize (and infect) food animals, especially ExPEC.

Admission to a hospital, an LTCF, or treatment with antibiotics increases the likelihood of colonization with Enterobacteriaceae other than ExPEC, P. mirabilis, and K. pneumoniae.

Of the Enterobacteriaceae discussed here, Edwardsiella in an exception in that acquisition occurs from exposure to a freshwater or marine environment where it normally resides.
Entry

Entry into an organ is the next critical step in pathogenesis. Examples include:

Ascension into the bladder from the perineum; sexual activity or an indwelling catheter facilitate this process.

Disruption of the bowel enables entry of GI tract colonizers into the peritoneum.

Micro- or macro-aspiration of oropharyngeal colonizers into the lower respiratory tract; endotracheal intubation or loss of glottic control facilitate this process.

Disruption or breakdown or the skin barrier; neuropathy, poor vascular supply, and immobility facilitate this process.
Growth and Survival within the Host

Entry alone does not necessarily result in infection. Next, the bacteria must be able to survive and proliferate in the face of the innate host defenses.

This is achieved by some combination of the infecting inoculum, the inherent virulence of the bacterial strain, and the status of the host. For example, a high inoculum or compromised host may enable infection with a less virulent strain, and by contrast a virulent strain, such as ExPEC, and the new hypervirulent variant of K. pneumoniae may cause infection in a normal host.

The host and pathogens have been co-adapting throughout evolutionary history. The genetic rate of change is far more rapid for the bacteria. Recently, we have countered by our ability to develop “tools” (e.g., antimicrobials). However, most recently, the rate at which bacteria are countering these tools has exceeded our tool-making capacity.

Multiple virulence factors are required for infection. Generally, virulence factors that enable infection within the intestinal tract are distinct from those required for extra-intestinal infection because of the differences in the environment and host defenses.

Virulence factors may be genus-, species-, and even strain-specific; however, certain generalities exist to be able to cause extra-intestinal infection.

Adhesions

mediate attachment to host tissues

may dictate host range and site of infection

overcome host obstacles, such as urine flow, muco-ciliary blanket

examples include the type I and P pili of ExPEC, which mediate attachment to the bladder and kidney, respectively.

Nutrient acquisition

Required for growth

The human host is a relative minimal medium, and many nutrients, such as iron bound by hemoglobin, lactoferrin, transferrin, and ferritin, are unavailable to bacteria.

Enterobacteriaceae possess multiple means to acquire iron, such as siderophores, which are secreted, and “steal” iron from host proteins, transporting it back into the bacterial cytoplasm.

The host counters with lipocalin-2, which binds certain siderophores, preventing their re-uptake.

The bacteria counter by either modifying siderophores previously recognized by lipocalin-2 or producing other siderophores that lipocalin-2 is unable to bind.

Avoidance of initial host bactericidal activity

required to prevent antimicrobial peptide, complement-mediated bactericidal activity and phagocytosis by professional phagocytes

Enterobacteriaceae possess surface polysaccharides (e.g., capsule and lipopolysaccharide) that afford protection.

Bacterial biofilm formation on foreign bodies may also assist is avoiding host defenses; however, it remains poorly defined for the Enterobacteriaceae how often, when, and where this occurs on native tissues.

Toxins, such as the alpha-hemolysin produced by certain ExPEC strains, can lyse neutrophils.

Antimicrobial resistance

should be considered as a virulence factor

counters our “tool,” development of antimicrobials

Bacterial acquisition of antimicrobial resistance may decrease the inherent virulence of the strain.

The combination of a professional pathogen (e.g., ExPEC or the hypervirulent variant of K. pneumoniae) that becomes MDR or pan-resistant without loss of biological fitness is a sobering thought.

A harbinger of this scenario is the ST131 clone of ExPEC, which spread rapidly seemingly because of increased antimicrobial resistance (e.g., to fluoroquinolones) without an apparent loss in fitness.

Tissue destruction

mediated by toxins produced by some Enterobacteriaceae

Alpha-hemolysin is a common example capable of lysing a variety of host tissues.

An over exuberant host response can also cause host damage (e.g., histotoxic compounds and reactive oxygen species present or generated by neutrophils).

Tissue necrosis, which can also occur from surgery and trauma, generates an environment that facilitates bacterial growth and survival, since nutrients are more available and the access and/or function, especially in an anaerobic environment, and host defense components may be compromised.
Transmission

Transmission usually leads to colonization prior to infection, which may or may not develop.

Person-to-person transmission occurs in the hospital among health-care providers and patients.

Infection control measures are critical, especially for MDR and near pan-resistant strains.

The environment and inanimate objects serve as intermediates for transmission.

The ability of Enterobacteriaceae to survive within the environment facilitates transmission.

Biofilm formation may enhance survival outside of the host.

Person-to-person transmission occurs between close contacts, such as family members or sexual partners.

Animal-to-person transmission can occur between pets and their owners.

Pathogen-specific Mechanisms of Pathogenesis

The steps described above generally apply to extra-intestinal pathogenic members of Enterobacteriaceae.
Intestinal pathogenic E. coli are not extra-intestinal pathogens and cause disease by specific mechanism well defined and relatively pathotype specific.
Edwardsiella is capable of causing intestinal disease, as well as extra-intestinal disease. The mechanisms responsible for intestinal infection are less well understood.

WHAT’S THE EVIDENCE for specific management and treatment recommendations?

(Clinical paper on Serratia.)

(Discussed the evolving role and use of fosfomycin.)

(Clinical paper on Citrobacter.)

(Discussed the emergence of ESBL-producing E. coli.)

(A presentation of NHSN data from 2006-7.)

(The report on K. oxytoca as a cause of hemorrhagic colitis.)

(A report on food animals as a source of MDR-E. coli.)

(A report on Proteeae.)

(A review of Providencia bacteremia.)

(A review of Proteus, Providencia and Morganella.)

(A review of Enterobacter spp.)

(A report on ESBL-producing E. coli and community-acquired infection.)

(A report on E. sakazakii and its epidemiology.)

(A report describing the new hypervirulent variant of K. pneumoniae.)

(A report describing cases of the new hypervirulent variant of K. pneumoniae in the United States.)

(A report on ExPEC and its medical-economic importance.)

(A report on Edwardsiella.)

(A case report describing successful treatment of XDR K. pneumoniae.)

(Fosfomycin activity against ESBL-producing E. coli and the identification of resistance genes.)

(A perspective on the NDM-1 carbapenemase.)

(A report on the poor activity of amp-sulbactam against E. coli.)

(A report on carbapenemase-producing K. pneumoniae in Greece.)

(In vitro evaluation of treatment options for carbapenem resistant Enterobacteriaceae.)

(In vivo evaluation of pip-tazo against ESBL-producing E. coli and K. pneumoniae.)

(Review on PRR, signaling pathways and the consequences of genetic defects.)

(A discussion of treatment of XDR and MDR strain, particularly on the use of colistin and tigecycline.)

(Report on an outbreak of colistin and carbapenem-resistant K. pneumoniae in Detroit.)

(A report of the emergence of resistance to third-generation cephalosporins in Enterobacter during treatment.)

(A commentary on the NDM-1 carbapenemase.)

(A report on foreign travel with resultant colonization with ESBL-producing E. coli.)

(In vitro activity of tigecycline against MDR-Enterobacteriaceae.)

(A brief review of beta-lactamases and treatment approaches against ESBL and carbapenemase-producing Enterobacteriaceae.)

(A review of antimicrobial resistance in the Enterobacteriaceae.)

(CDC guidance on the infection control of carbapenemase-producing Enterobacteriaceae.)

(A review of the epidemiology and virulence factors in ExPEC, a paradigm for Enterobacteriaceae that cause extraintestinal infection.)

Hirai, Y., Asahata-Tago, S., Ainoda, Y., Fujita, T., Kikuchi, K.. “Edwardsiella tarda bacteremia. A rare but fatal water- and foodborne infection: Review of the literature and clinical cases from a single centre”. Can J Infect Dis Med Microbiol.. vol. 26. 2015. pp. 313-8. (A report on Edwardsiella.)

Liu, Y.Y., Wang, Y., Walsh, T.R., Yi, L.X., Zhang, R., Spencer, J., Doi, Y., Tian, G., Dong, B., Huang, X., Yu, L.F., Gu, D., Ren, H., Chen, X., Lv, L., He, D., Zhou, H., Liang, Z., Liu, J.H., Shen, J.. “Emergence of plasmid-mediated colistin resistance mechanism MCR-1 in animals and human beings in China: a microbiological and molecular biological study”. Lancet Infect Dis.. vol. 16. 2016. pp. 161-8. (A report on a new transferrable colistin resistance gene.)

Mazuski, J.E., Gasink, L.B., Armstrong, J., Broadhurst, H., Stone, G.G., Rank, D., Llorens, L., Newell, P., Pachl, J.. “Efficacy and Safety of Ceftazidime-Avibactam Plus Metronidazole Versus Meropenem in the Treatment of Complicated Intra-Abdominal Infection – Results from a Randomized, Controlled, Double-Blind, Phase 3 Program”. Clin Infect Dis.. 2016. (Clinical study on the use of ceftazidime-avibactam in intra-abdominal infection.)

Mendes, A.C., Rodrigues, C., Pires, J., Amorim, J., Ramos, M.H., Novais, A., Peixe, L.. “Importation of Fosfomycin Resistance fosA3 Gene to Europe”. Emerg Infect Dis.. vol. 22. 2016. pp. 346-8. (A report on a gene that confers resistance to fosfomycin.)

Scott, L.J.. “Ceftolozane/Tazobactam: A Review in Complicated Intra-Abdominal and Urinary Tract Infections”. Drugs. vol. 76. 2016. pp. 231-42. (Review on ceftolozane/tazobactam.)

Sharma, R., Eun Park, T., Moy, S.. “Ceftazidime-Avibactam: A Novel Cephalosporin/beta-Lactamase Inhibitor Combination for the Treatment of Resistant Gram-negative Organisms”. Clin Ther.. vol. 38. 2016. pp. 431-44. (Review on ceftazidime/avibactam.)

Solomkin, J., Hershberger, E., Miller, B., Popejoy, M., Friedland, I., Steenbergen, J., Yoon, M., Collins, S., Yuan, G., Barie, P.S., Eckmann, C.. “Ceftolozane/Tazobactam Plus Metronidazole for Complicated Intra-abdominal Infections in an Era of Multidrug Resistance: Results From a Randomized, Double-Blind, Phase 3 Trial (ASPECT-cIAI)”. Clin Infect Dis.. vol. 60. 2015. pp. 1462-71. (Clinical study on the use of ceftolozane/tazobactam in intra-abdominal infection.)

No sponsor or advertiser has participated in, approved or paid for the content provided by Decision Support in Medicine LLC. The Licensed Content is the property of and copyrighted by DSM.

Jump to Section

Infectious Diseases