Pseudomonas

TAXONOMY Back to top

Pseudomonas is a large and complex genus of gram-negative bacteria of importance, as it

includes species with both clinical and environmental implications. Many species are

saprophytic or pathogenic for plants. The metabolic versatility of species within this genus

allows many to degrade low-molecular-weight organic and aromatic compounds (83). The

genus Pseudomonas first proposed by Migula in 1894 (113) has undergone many taxonomic

revisions as methodologies of species identification continue to improve, and it was

comprised of five unrelated groups, as determined by rRNA-DNA hybridization studies in the

early 1970s by Palleroni et al. (129). Pseudomonas (sensu stricto) is rRNA homology group I

(27), in the gamma subclass of the Proteobacteria (161). The other rRNA homology groups

are II, Burkholderia and Ralstonia; III, Comamonas, Acidovorax,

Delftia, and Hydrogenophaga; IV, Brevundimonas; and

V, Stenotrophomonas and Xanthomonas (83).

Several of the clinically relevant Pseudomonas species demonstrate marked heterogeneity

and have been subdivided into biovars or genomovars. Genomovars are genetically distinct

groups that warrant species designation but lack phenotypically defining characteristics, and

they are determined by DNA-DNA reassociation experiments, 16S rRNA gene sequencing in

combination with chemotaxonomic total fatty acid analysis, and total protein pattern analysis

(56). Much work in genome sequencing is occurring; 3 strains ofPseudomonas

aeruginosa and 13 strains from other Pseudomonas species have been sequenced

(http://www.ncbi.nlm.nih.gov/genomes/genlist.cgi?taxid=2&type= 0&name=Complete%20

Bacteria).

The highest level of genetic diversity of any species known is found in Pseudomonas

stutzeri (140), as established by multilocus enzyme electrophoresis. P. stutzeri has at least

nine genomovars, with clinical isolates being found in genomovars 1 and 2. There are no

consistent phenotypic differences to justify splittingP. stutzeri into unique species (56).

Pseudomonas fluorescens was originally divided into biotypes A, B, C, D, E, F, and G

(biotypes A to E are also referred to as biovars I, II, III, IV, and V). Biotype B was

reclassified as Pseudomonas marginalis. Biotypes D and E (Pseudomonas

chlororaphis and Pseudomonas aureofaciens) have now been combined into the single

species P. chlororaphis, which is no longer considered a member of the fluorescens group.

Pseudomonas putida consists of biovars A and B. Biovar A should be regarded as the

“typical” P. putida (35), while biovar B may have a closer affinity with P. fluorescens. More

biovars of P. putida are warranted (35).

Great heterogeneity is found within the species of P. stutzeri, P. fluorescens, and P.

putida, and these species are of interest in plant, marine, soil, and biotechnical sciences.

They are of limited importance in clinical medicine. As polyphasic taxonomy continues to

advance, more changes will doubtlessly arise; the clinical laboratory must keep abreast of

such changes, in order to differentiate these isolates from the more clinically

important Pseudomonas species.

DESCRIPTION OF THE GENUS PSEUDOMONAS Back to top

Pseudomonas spp. are aerobic non-spore-forming, gram-negative rods which are straight or

slightly curved and are 0.5 to 1.0 by 1.5 to 5.0 μm (72). They are usually motile, with one or

several polar flagella. They possess a strictly aerobic respiratory metabolism with oxygen as

the terminal electron acceptor; in some cases nitrate can be used as an alternative electron

acceptor that allows anaerobic growth. Most species of clinical interest are oxidase positive

(except Pseudomonas luteola and Pseudomonas oryzihabitans).Pseudomonas spp. are

catalase positive and are chemolithotrophs.

EPIDEMIOLOGY AND TRANSMISSION Back to top

Autogenous versus Exogenous Infection

Autogenous infection can only occur in those whose colonization resistance has been

perturbed: bacteremia secondary to gastrointestinal colonization in neutropenic hosts (170)

and pneumonia in individuals who have required endotracheal intubation (ventilatorassociated

pneumonia) (8, 115). Exogenous infection likely occurs in patients with cystic

fibrosis (CF), as their initial infecting isolates usually resemble environmental morphotypes

(122), although patient-to-patient spread has occasionally been demonstrated (see below).

Most other infections caused by P. aeruginosa are probably acquired exogenously, such as in

burn wound sepsis, conjunctivitis, otitis externa, and osteochondritis.

Exposure to Inanimate Reservoirs

Although efforts to prevent colonization with P. aeruginosa have been made, none has

proven uniformly successful. Strict infection control procedures and the practice of

compulsory hand hygiene are most effective at preventing patient-to-patient spread,

particularly in hospitals.

Special Considerations for Patients with CF

P. aeruginosa is the predominant respiratory tract p athogen in patients with CF (54), but its

mode of acquisition is poorly understood (155). Several studies have each demonstrated a

common clone in particular groups of patients who have received their care at the same

center (5, 18, 79); the most likely explanation for this finding is patient-to-patient spread.

Since most patients each tend to carry a unique strain during the course of infection

(122, 158), one assumes that the infection was acquired from an environmental source.

Indeed, one large study performed in Vancouver, British Columbia, Canada, over more than

20 years failed to demonstrate patient-to-patient spread of P. aeruginosa except between

siblings who could have acquired it from a common environmental source (158). A recent

study has demonstrated that in chronic respiratory infections in CF, sequential isolates of P.

aeruginosa from 30 patients d isplayed a high prevalence of DNA mismatch repair systemdeficient

hypermutable strains on isogenic backgrounds for each patient (42). Infection

control policies with CF patients for transmission prevention should be determined by local

epidemiological experience (174).

Species other than P. aeruginosa

Pseudomonas species other than P. aeruginosa are usually acquired from the environment.

CLINICAL SIGNIFICANCE Back to top

Normal Host Defenses against P. aeruginosa

Individuals with intact host defenses are not at risk for serious infection with P.

aeruginosa, but those whose circulating neutrophil counts are profoundly depressed (such as

patients with cancer receiving chemotherapy) are at risk for invasive infection (156).

Significance of Recovery of P. aeruginosa from Clinical

Specimens

When P. aeruginosa is recovered from a normally sterile body site, such as blood, pleural

fluid, or joint space, it usually constitutes a true infection. However, pseudoinfection (36)

should be considered when there is a cluster of infections with the same strain

of Pseudomonas, especially when such infections had not been frequently seen previously,

and the patients are neither severely ill nor at enhanced risk of such infection. A search for

the source of the cluster should include culture of the antiseptic used for skin preparation for

venipuncture or similar procedures.

P. aeruginosa is able to colonize mucosal surfaces, such as the oropharynx of patients

receiving intensive care or the endotracheal tubes of patients receiving mechanical

ventilation. Under such circumstances, recovery ofP. aeruginosa from respiratory tract

cultures may not indicate a true infection, and the significance of its presence in the culture

should be interpreted with caution.

Infection in Patients with Neutropenia

Not all patients with neutropenia (neutrophil count of less than 0.5 × 109 per liter) are at risk

for invasive disease. Patients at greatest risk are adults undergoing cancer chemotherapy or

marrow ablation for bone marrow transplantation (9); children with similar conditions are at

lesser risk for P. aeruginosa bacteremia and are more often infected with gram-positive

bacteria.

Infection in Patients with CF

P. aeruginosa is the predominant respiratory tract pathogen in patients with CF, for reasons

which remain incompletely explained (54). The organism appears to have a particular

tropism for CF epithelial cells and can resist normal respiratory tract host defenses. Once

infection is established, it usually persists (48), and the bacteria undergo a transition to the

“CF phenotype,” consisting of the following: (i) a rough lipopolysaccharide (LPS) (60), in

which the O polysaccharide is incompletely expressed, rendering the bacteria susceptible to

the bactericidal effect of human serum; (ii) mucoid colonial morphology (95) resulting from

the exuberant production of a mucoid exopolysaccharide composed of O-acetylated guluronic

and mannuronic acids; (iii) nonmotility (104) in which the bacteria lack normal functional

flagellar function; and (iv) hypoexpression of various exotoxins and other exoproducts

(12, 185). Some of these changes may be under global regulation, but they can also be

expressed individually. Transition of P. aeruginosa from nonmucoid to mucoid in the CF lung

is usually associated with an accelerated decline in pulmonary function and an adverse

prognosis (37), perhaps because of the capacity of the mucoid exopolysaccharide to interfere

with normal host phagocytic defenses (54, 157) and to facilitate the formation of biofilms

(90). Biofilm formation may also be enhanced by another colonial form, small-colony

variants (previously known as dwarf colonies) (48, 64). Furthermore, CF patients receive

frequent courses of anti-Pseudomonas antimicrobial therapy, often rendering the bacterium

with which they are chronically infected resistant to a wide range of antimicrobial agents

(61, 86).

Ventilator-Associated and Nosocomial Pneumonia

The normal respiratory tract is well protected against infection by means of mucociliary

clearance of inhaled particles and potentially infectious agents. Placement of an endotracheal

tube for mechanical ventilation allows upper respiratory tract microbes to gain access to the

lower respiratory tract, where infection can be established. Adults receiving mechanical

ventilation are at high risk for developing P. aeruginosa ventilator-associated pneumonia

(17), particularly after or during treatment with broad-spectrum antimicrobial agents.

Nosocomial pneumonia most often occurs in neutropenic patients following broad-spectrum

antimicrobial therapy (143). Initial empirical therapy, until an etiologic agent is identified,

should include a drug effective against P. aeruginosa.

Burn Wound Infections

Thermal burns of the skin abrogate an essential component of the body’s defense against

infection, the physical barrier of the intact dermis (135). The resulting damaged tissue is a

rich culture medium and is at great risk for colonization and infection by P. aeruginosa; such

infections have been one of the leading causes of morbidity and mortality in victims of burns.

Topical therapy is designed to prevent P. aeruginosa and other pathogens from causing

infection. Infections of burn wounds with gram-negative bacteria (in particular P. aeruginosa)

typically occur about 1 week after the injury. The extent of the burn has a profound influence

on risk of infection and prognosis (135). Prevention of bacterial burn wound infection has

become so effective over the past decade that it is now very rare, and in many centers

fungal infections predominate.

Osteochondritis

P. aeruginosa is the most common cause of osteochondritis of the dorsum of the foot

following penetrating wounds (20). The typical scenario involves a child who has stepped on

a nail which pierces the foot after passing through the sole of a running shoe. The prevalence

of P. aeruginosa as the etiological agent in this situation may be due to its propensity to

survive in the rubber of old running shoes (38).

Folliculitis and Superficial Infections

Since P. aeruginosa is a hydrophilic bacterium and can survive at temperatures as high as

42°C, it has the propensity to cause infection in people who are exposed to heated water for

extended periods. Hot tub users are at risk of P. aeruginosa folliculitis (55), a condition which

is self-limited for healthy hosts and resolves rapidly. People who spend extended periods

swimming are at risk of external ear infections (“swimmer’s ear”), another self-limiting

condition in immunocompetent people which responds readily to therapy with topical

antimicrobial agents (7). The cornea is relatively resistant to infection except when its

integrity has been broken. Users of contact lenses are at risk of P. aeruginosa conjunctivitis,

especially if hygiene is poor or lenses are used for extended periods (164).

Other Infections with P. aeruginosa

P. aeruginosa can cause meningitis (usually following trauma or surgery) (39), malignant

otitis externa in diabetics (144), sepsis and meningitis in newborns (167), endocarditis or

osteomyelitis in users of intravenous drugs (148), community-acquired pneumonia

(especially in people with underlying lung disease such as bronchiectasis [47]), and urinary

tract infections in patients with complex urinary tract abnormalities (118,142). Each of these

presentations is unusual and is superimposed on some abrogation of normal host defenses.

Infections with Pseudomonas Species other than P.

aeruginosa

Healthy individuals are resistant to serious infections by all Pseudomonas species,

including P. aeruginosa. However, immunocompromised hosts are occasionally infected with

one of the many non-aeruginosa species, including (but not limited to) P. fluorescens, P.

putida, P. stutzeri, P. oryzihabitans, P. luteola, P. alcaligenes, P. mendocina, and P. veronii.

Several of these species have been recovered from the respiratory secretions of patients with

CF, but their role in pathogenesis of lung disease has not been determined. Some of these

species have the capacity, like P. aeruginosa, to grow in hostile environments, such as

antiseptic solutions; they can therefore be the cause of pseudobacteremia. Because of their

low virulence, infections due to these species are often iatrogenic and are associated with the

administration of contaminated solutions, medicines, and blood products or the presence of

indwelling catheters (94, 100, 107, 137, 150).

P. fluorescens and P. putida have the ability to grow at 4°C, and P. fluorescens can be

isolated from the skin of a small proportion of blood donors (150), resulting in occasional

transfusion-associated septicemia in the recipient. Various Pseudomonas species have been

implicated in outbreaks of pseudobacteremia (84, 153).

P. stutzeri is an unusual cause of human infection. It can cause bacteremia in

immunosuppressed persons (134), meningitis in human immunodeficiency virus-infected

individuals (141), pneumonia in alcoholics (16), and osteomyelitis (139). Iatrogenic

infections due to P. stutzeri include endophthalmitis following cataract surgery (77) and

bacteremia in hemodialysis patients as a result of contaminated dialysis fluid (52). P.

stutzeri has also been recovered from wounds, the respiratory tract of intubated patients,

and the urinary tract, although its pathogenic role in those settings is unclear (119).

P. oryzihabitans is being recognized increasingly as a cause of bacteremia in

immunocompromised patients with central venous access devices. Synthetic bath sponges

can be a source of bacteremia with this organism in patients with Hickman catheters (107).

This organism has also been reported to cause peritonitis in patients undergoing chronic

ambulatory peritoneal dialysis, cellulitis, abscesses, wound infections, and meningitis

following neurosurgical procedures (94).

P. luteola is a rare cause of infections in humans. There have been case reports of a variety

of different infections, including bacteremia, cellulitis, osteomyelitis, peritonitis, endocarditis,

and postsurgical meningitis (137, 138).

Other Pseudomonas species are found even less f requently in human infection. P.

alcaligenes has been associated with catheter-related endocarditis in a bone marrow

transplant recipient (111). P. mendocina has been isolated from two patients with

endocarditis (4, 78). P. veronii has been reported to be associated with an intestinal

inflammatory pseudotumor (19). P. monteilii has been recovered from stool, bile, placenta,

bronchial aspirates, pleural fluid, and urine, but its clinical significance is uncertain

(33, 34). P. mosselii has also been isolated from various specimens, but the clinical

significance is not known (25).

COLLECTION, TRANSPORT, AND STORAGE OF

SPECIMENS Back to top

Pseudomonas spp. are able to survive in diverse environments and through a wide

temperature range. Some species prefer incubation temperatures lower than 25°C, while P.

aeruginosa can grow at temperatures up to 42°C. These organisms are easily recovered from

clinical specimens using standard collection, transport, and storage techniques as outlined

in chapter 16. Samples of Pseudomonas spp. can be refrigerated at 2 to 8°C for up to 4

weeks. Organisms can be kept in long-term storage at -80°C using standard laboratory

freezing protocols.

DIRECT EXAMINATION Back to top

Microscopy

With regard to Gram stain morphology, Pseudomonas spp. are motile, gram-negative, nonspore-

forming, straight or slightly curved bacilli measuring 0.5 to 0.8 μm by 1.5 to 3.0 μm.

The Gram stain morphology cannot easily distinguish Pseudomonas spp. from other

nonfermenting bacilli, although they are usually thinner thanEnterobacteriaceae. Among the

pseudomonads, there is some variation in Gram stain morphology. Certain strains of P.

putida can appear elongated. Organisms from older cultures may appear slightly

pleomorphic. Flagellar stains reveal one or more polar flagella. P. aeruginosa has a single

polar flagellum.

Mucoid strains may be distinguished on direct examination by the presence of clusters or

long filaments of short gram-negative bacilli surrounded by darker pink-staining material

(alginate). It is important to note this on direct examination, as the organisms may grow

very slowly or not at all. The presence of these mucoid forms should be documented on

clinical reports. Because Pseudomonas spp. may be colonizers, their isolation does not

always link them to clinical disease. However, their presence intracellularly in

polymorphonuclear cells is clinically significant and should be documented and direct further

workup.

Nucleic Acid Detection

P. aeruginosa and other Pseudomonas species are detected ordinarily by culture techniques;

these methods are particularly important for determining antimicrobial susceptibility, as

these organisms have a high degree of intrinsic and acquired resistance (61, 98). However,

situations exist in which a more rapid method can be instituted, such as for screening

environmental niches or for rapidly evaluating the sputum of patients with CF

(22, 159, 186). Methods that have been used include PCR amplification of various genomic

regions, such as genes for rRNA (76), heat shock protein (22), or exotoxin A (85).

Conventional and real-time PCR have both proven to be useful, and the amplification of

multiple targets can be particularly valuable in identification of nonaeruginosa

Pseudomonas species (136). Nucleic acid probes have also been used to detect

these bacterial species, the specificity depending upon the unique nature of the genomic

region being detected. Probes directed at species-specific 16S rRNA have been most widely

used for this purpose (57, 163) and may have a role in the identification of clinically

relevant, biochemically inactive Pseudomonas species, including certain strains of P.

aeruginosa (136). PCR amplification of 16S ribosomal DNA (rDNA) followed by restriction

fragment length polymorphism (RFLP) analysis has been used to successfully identify and

characterize members of the fluorescent pseudomonad group (89). Since RFLP of the 16S

rRNA gene does not provide sufficient resolution among genomovars of a species, a more

discriminatory test may be used, such as sequencing the internally transcribed 16S-23S

rDNA spacer (internal transcribed spacer 1) regions, believed to have more genetic variability

among genomovars (56). Recently, peptide nucleic acid fluorescence in situ hybridization

(130) has been shown to be highly sensitive and specific for identification of P. aeruginosa.

Mechanisms of antimicrobial resistance, such as the identification of extended-spectrum β-

lactamases in P. aeruginosa, have also been detected using molecular techniques, by realtime

PCR detection (180).

ISOLATION PROCEDURES Back to top

Pseudomonas species have very simple nutritional requirements and grow well on standard

broth and solid laboratory media such as tryptic soy agar with 5% sheep blood, chocolate

agar, and MacConkey agar, which are recommended to isolate Pseudomonas spp. from

clinical specimens. MacConkey agar is also a differential medium helpful in identifying

different strains of Pseudomonas spp., including mucoid strains of P. aeruginosafrom CF

patients. Multiple selective media containing inhibitors such as acetamide, nitrofurantoin,

phenanthroline, 9-chloro-9-[4-(diethyamino)phenyl]-9,10-dihydro-10-phenylacridine

hydrochloride (C-390), and cetrimide (15, 66, 82, 87, 91, 109, 169, 171) have been used in

the past for the isolation and presumptive identification of P. aeruginosa from clinical and

environmental samples. Currently, cetrimide and a combination of phenanthroline with C-390

are the most commonly used selective agents. Inhibition of some strains of P.

aeruginosa from sputum specimens from CF patients has been reported using a selective

agar containing cetrimide (200 mg/liter) and nalidixic acid (15 mg/liter) (40), emphasizing

the need to use both selective and nonselective media for recovery of bacteria from these

patients. Some of the non-aeruginosa pseudomonads, like P. fluorescens, P. putida, and P.

oryzihabitans, may grow better at the lower temperatures of 28 to 30°C. Good growth is

usually achieved after 24 to 48 h of incubation. For cultures from CF patients, it is

recommended that solid medium plates be held at 35 to 37°C for 5 days.

IDENTIFICATION Back to top

Fluorescent Group

Members of the fluorescent pseudomonad group produce pyoverdin, a water-soluble yellowgreen

or yellow-brown pigment that fluoresces under short-wavelength UV light. Many

strains of P. aeruginosa can produce the blue pigment pyocyanin. When pyoverdin combines

with the blue water-soluble phenazine pigment pyocyanin, the bright green color

characteristic of P. aeruginosa is created. This organism may also produce other watersoluble

pigments such as pyorubrin (red) or pyomelanin (brown-black). Conditions of iron

limitation enhance pigment production, as these pigments act as siderophores in iron uptake

systems of the bacteria. Non-dye-containing media enhance visualization of pigments.

P. aeruginosa

Most P. aeruginosa organisms are easily recognizable on primary isolation media on the basis

of characteristic colonial morphology, production of diffusible pigments, and a grape-like

odor. Older cultures may exhibit a corn taco-like odor. Colonies are usually flat and

spreading and have a serrated edge and a metallic sheen that is often associated with

autolysis of the colonies (188). Other morphologies exist, including smooth, mucoid, and

dwarf (small-colony variants) (63, 64, 126, 175). Mucoid colonial variants are particularly

prevalent in respiratory tract specimens from CF patients (51).

P. aeruginosa is distinct from the rest of the clinically relevant fluorescent pseudomonads in

its ability to grow at 42°C. In addition to pigment production, other tests that confirm its

identification are positive oxidase and arginine tests and an alkaline over no-change reaction

in the triple sugar iron test.

Microbiologists must be aware of certain variations in the phenotypes of P. aeruginosa.

Isolates lacking oxidase activity have occasionally been reported, but they exhibit the other

characteristic features. Prior antibiotic therapy with agents that affect protein synthesis may

cause the aberrant phenotype (59, 112). Mucoid isolates of P. aeruginosa from CF patients

may undergo several phenotypic changes, including slow growth, loss of motility, and loss of

pigment production. Small-colony variants may require prolonged incubation, lack motility,

be hyperpiliated, adhere to agar surfaces, and show autoaggregative properties in liquid

medium (175).

P. fluorescens and P. putida

P. fluorescens and P. putida do not possess distinctive colony morphology or odor. Their

inability to reduce nitrates to nitrogen gas and their ability to produce acid from xylose

distinguish these two species from the other fluorescent pseudomonads. P. fluorescens can

be differentiated from P. putida by its ability to grow at 4°C and to hydrolyze gelatin; P.

putida can do neither. P. fluorescens isolates may require 4 to 7 days of incubation for

accurate detection of gelatin hydrolysis. According to the package insert for API 20NE

(version 7.0; bioMerieux, Inc., Durham, NC), only 39% of P. fluorescens isolates hydrolyze

gelatin in 24 to 48 h.

Nonfluorescent Group

P. stutzeri and P. mendocina

Most P. stutzeri isolates are easily recognized on primary isolation media by their distinctive

dry, wrinkled colony morphology, similar to the morphology of Burkholderia pseudomallei. P.

stutzeri can be distinguished from the latter species by its lack of arginine dihydrolase

activity and inability to produce acid from lactose. P. stutzeri colonies can pit or adhere to

the agar and are buff to brown. The adherence can make removal of colonies from agar

medium difficult. Because of the difficulty in making suspensions of specific turbidity,

commercial susceptibility systems may not work well with this organism. Not all isolates of P.

stutzeri produce wrinkled colonies; such strains can be distinguished from other

pseudomonads by their ability to hydrolyze starch, a unique reaction for this species.

P. mendocina colonies are smooth, nonwrinkled, and flat, producing a brownish yellow

pigment. Key biochemical characteristics of this species include the ability to reduce nitrates

to nitrogen gas, positive arginine dihydrolase activity, and inability to hydrolyze acetamide or

starch.

P. alcaligenes and P. pseudoalcaligenes

P. alcaligenes and P. pseudoalcaligenes have rarely been encountered in clinical samples

(111) and do not have a distinctive colony morphology. Compared to other pseudomonads,

they are biochemically inert. Characteristics that distinguish them from other biochemically

inert gram-negative rods are a positive oxidase reaction, motility due to a polar flagellum,

and growth on MacConkey agar. P. alcaligenes is distinguished fromP. pseudoalcaligenes by

its inability to oxidize fructose. Although growth at 42°C was thought to be a distinguishing

feature between them, further studies now indicate that growth at 41°C (and probably 42°C)

is also present in most strains of P. alcaligenes (N. Palleroni, personal communication).

These organisms are difficult to identify by many commercial systems, and for most clinical

situations they can simply be referred to as “Pseudomonas spp., not aeruginosa.” If the

clinical situation dictates a definitive identification, assistance from reference laboratories

should be sought.

P. luteola and P. oryzihabitans

P. luteola and P. oryzihabitans can be distinguished from other pseudomonads by their

negative oxidase reaction and production of an intracellular, nondiffusible yellow pigment.

Both organisms typically exhibit rough, wrinkled, adherent colonies or, more rarely, smooth

colonies. P. luteola can be differentiated from P. oryzihabitans on the basis of its ability to

hydrolyze o-nitrophenyl-β-D-galactopyranoside and esculin.

Use of Commercial Identification Systems

Commercial identification systems rather than conventional biochemical tests increasingly

are used in many laboratories to identify Pseudomonas spp. Commercial products can be

divided into manual and automated systems (123). The more frequently used manual

systems are the API 20NE (bioMerieux), Crystal E/NF (Becton Dickinson), and RapID NF Plus

(Innovative Diagnostic Systems). The manual systems usually provide accurate identification

of P. aeruginosa, including mucoid isolates as well as other Pseudomonas species, and are

preferred over automated systems for isolates from CF patients.

Automated systems are commonly used in many medium to large clinical laboratories. As P.

aeruginosa is easily identified by a few conventional biochemical tests, it is often not

necessary to use a more expensive commercial system. Several of the automated systems

are not very accurate and may require additional testing for non-P. aeruginosa species; thus,

their labor, cost, and time-saving benefits are lost. Automated systems can identify P.

aeruginosa from non-CF sources with 90 to 100% accuracy (44, 125), but some systems

may require additional tests to achieve these results (124, 168). Most reviews focus on the

evaluation of P. aeruginosa, with only a few, if any, other Pseudomonas species represented

in the organisms being tested. When other Pseudomonas species were included, the new

Vitek 2GN panel performed well (43,125), while other systems often relied on additional

testing to obtain an identification (29, 124, 162, 168). Hence, it is wise to consider carefully

the clinical significance, colonial morphology, and other key features before accepting results

from automated systems.

Identification of Pseudomonas species, especially those isolated from CF patients, is not

always optimal with rapid systems. The MicroScan (Dade International, Inc.) system

(Negative Combo 15), when incubated for 20 to 24 h according to the manufacturer’s

method, performed poorly for CF isolates, with only 57% of nonmucoid and 40% of

mucoid P. aeruginosa isolates correctly identified (147). Extended incubation for 48 h

improved accuracy to 86 and 83%, respectively. Misidentified species were most commonly

either Alcaligenes spp. or P. fluorescens/P. putida. For P. aeruginosa from non-CF samples,

the overall accuracy has been reported as 94% (147). Other automated systems have not

been evaluated to date specifically for the identification of CF isolates, so caution in

interpreting results is advised. The importance of non-aeruginosa Pseudomonas species as

the cause of significant infection has not been established in most cases. The need to pursue

species identification beyond the ruling out of P. aeruginosa will depend on the individual

institution’s requirements.

TYPING SYSTEMS Back to top

Phenotypic Typing Methods

Historically, typing of P. aeruginosa for epidemiological purposes has relied upon phenotypic

characteristics of the bacteria. The most widely used method was based upon differences in

LPS O polysaccharide (LPS serotyping). This method is good for most clinical isolates but can

only differentiate among the 17 different LPS types in the most widely used (Difco)

commercial typing set. Antimicrobial susceptibility profiling has been used, but the capacity

of P. aeruginosa to develop resistance under the pressure of antimicrobial therapy has

rendered this method unreliable in conditions of chronic infection such as CF. Each of these

methods has its shortcomings, as the phenotypic characteristics of the bacteria are highly

plastic. The LPS of CF isolates of P. aeruginosa often lack the O polysaccharide, against

which most serotyping reagents react. None of these phenotypic methods is reliable in CF,

and they have largely been replaced by genotypic methods.

Genotypic Typing Methods

Several genotypic methods have been developed over the past two decades for typing P.

aeruginosa for epidemiological purposes (155). These are briefly described in the order in

which they were developed. Each is useful, even for typing isolates from patients with CF,

but they are not available in most clinical diagnostic laboratories.

RFLP

RFLP relies upon the genetic diversity at a specific site within the bacterial genome. Such

diversity exists upstream of the gene for exotoxin A (exoA) in P. aeruginosa (122). In a

study of different typing methods, exoARFLP proved superior to all phenotypic methods for

typing P. aeruginosa (74). This method was also the first to demonstrate convincingly that

patients with CF were usually each infected with a unique strain that was usually present

durably (without eradication or replacement) for extended periods. Pilin gene RFLP has

demonstrated that individual CF patients are durably infected with the same strain despite

changes in pilin protein expression (122). The disadvantages of RFLP are its relatively weak

discriminatory power (compared to that of newer methods), its cumbersome nature, and its

predominant use of radioactive probes.

PFGE

Pulsed-field gel electrophoresis (PFGE) is often considered the “gold standard” for bacterial

typing, as it provides a view of the entire genome. The banding pattern is unique to each

strain (or clone) and can be used for any bacterial species.

PCR-Based Typing Methods

Several different PCR-based methods have been used for typing P. aeruginosa. They are

directed at known elements within the genome or against random but relatively frequently

encoded sequences. The latter, random amplified polymorphic DNA (RAPD) analysis, has

proved quite robust for typing P. aeruginosa (105), but it must be run consistently on the

same equipment to yield reproducible results. Data from RAPD analysis usually are highly

consistent with those from PFGE. PCR-amplified products can be digested with restriction

enzymes to yield more discriminatory data (3, 149).

MLST

Multilocus sequence typing (MLST) has only recently been employed for typing P.

aeruginosa. It is likely to be the most highly discriminatory among the genetic typing tools,

but it is extremely time-consuming and expensive to employ. The method entails PCR

amplification of specific genes and then sequencing of the gene products. This can be done

only in very specialized centers, but it has the power to provide highly reliable data on

relatedness among isolates. MLST is particularly useful in typing isolates from patients with

CF (176).

SEROLOGIC TESTS Back to top

Serologic tests are not recommended for patients with P. aeruginosa infections. However, a

commercial test system (Mediagnost, Reutlingen, Germany) detecting serum antibodies

against three P. aeruginosa antigens (alkaline phosphatase, elastase, and exotoxin A)

performed favorably for CF patients with negative or intermittent P. aeruginosa status, since

a rise in antibody titers indicated probable infection (81).

ANTIMICROBIAL SUSCEPTIBILITIES Back to top

P. aeruginosa possesses intrinsic resistance to many antibiotic classes and has the ability to

develop resistance by mutations in different chromosomal loci or by horizontal acquisition of

resistance genes carried on plasmids, transposons, or integrons. The various mechanisms of

resistance, substrate specificities, and geographic distributions are discussed below. The

frequent acquisition of antimicrobial resistance in P. aeruginosa limits the utility of

antimicrobial susceptibility patterns as a tool in epidemiological typing.

Mechanisms of Resistance

Intrinsic Resistance

Intrinsic resistance is mediated through multiple mechanisms. P. aeruginosa has an inducible

chromosomal AmpC β-lactamase that renders it resistant to ampicillin, amoxicillin,

amoxicillin-clavulanate, and first- and second-generation cephalosporins, as well as

cefotaxime and ceftriaxone (97). Although impermeability was originally thought to be

responsible for resistance to other antibiotic classes, efflux pump systems have been

identified as a more prevalent intrinsic mechanism of resistance.

Multiple efflux pumps exist in P. aeruginosa that can result in expulsion of β-lactams,

chloramphenicol, fluoroquinolones, macrolides, novobiocin, sulfonamides, tetracycline, and

trimethoprim. Sequencing of the P. aeruginosa genome indicates that a high proportion of

genes, including regulatory genes, are involved in the efflux of organic compounds,

accounting for this organism’s ability to adapt to diverse environments and to resist most

antimicrobial agents. Efflux systems also export virulence determinants in P.

aeruginosa, enhancing their toxicity to the host (69).

Acquired Resistance

Various antibiotics overcome the intrinsic resistance of P. aeruginosa and are active against

this organism. These include extended-spectrum penicillins (piperacillin and ticarcillin),

certain third- and fourth-generation cephaloporins (ceftazidime and cefipime), carbapenems

(imipenem and meropenem), monobactams (aztreonam), fluoroquinolones (ciprofloxacin and

levofloxacin), aminoglycosides (gentamicin, tobramycin, and amikacin), and colistin.

Unfortunately, mutational resistance to all the antipseudomonal antibiotics can develop.

Efflux Pumps

Although multidrug efflux pump systems play a significant role in the intrinsic resistance of P.

aeruginosa, they also are critical to the development of multidrug resistance. MexAB-OprM is

expressed constitutively in all strains of P. aeruginosa. Upregulation or a mutation in

the mexR repressor gene (nalB mutant) results in efflux pump overproduction and significant

increase in the MICs of multiple antibiotics, including quinolones, penicillins, cephalosporins,

aztreonam, and meropenem (low-level resistance MIC, 8 to 32 μg/ml), but not imipenem

(133).

Impermeability Mutations

Impermeability mutations may result in resistance to carbapenem, aminoglycosides, colistin,

and quinolones. They are important in carbapenem resistance and result from the loss of the

OprD porin, a protein that forms a narrow transmembrane channel permeable to

carbapenems but not β-lactams.

β-Lactamases

The acquisition of β-lactamases (177) is not as common for P. aeruginosa as it is

for Enterobacteriaceae (97). Nevertheless, β-lactamases are being recognized increasingly

and are very diverse in this organism. Genes for these enzymes are encoded in plasmids, on

transposons or integrons, making their further dissemination likely. They confer resistance

predominantly to antipseudomonal penicillins, ceftazidime, cefipime, and aztreonam but not

carbapenems. Their activity is inhibited poorly by clavulanic acid or tazobactam.

Carbapenemases

With the exception of GES-2, all carbapenemases in P. aeruginosa belong to Ambler class B,

commonly referred to as metalloenzymes. Metalloenzymes are not inhibited by clavulanic

acid but are susceptible to inhibition by divalent ion chelators such as EDTA. They hydrolyze

all β-lactam antibiotics, except aztreonam, and are associated with high-level (MIC > 32

μg/ml) carbapenem resistance. Underreporting of carbapenem resistance may occur, as

expression of the carbapenemases varies, resulting in a wide range of MICs (2 to 128 μg/ml)

that may go undetected in clinical laboratories that rely only on automated systems.

Genes for these enzymes are plasmid mediated and are located on mobile gene cassettes

inserted in variable regions of integrons, resulting in enhanced potential for expression and

dissemination. Of concern is the close proximity of these genes to those for aminoglycoside

resistance (98).

Carbapenemases are spreading throughout Asia, Europe, and the Americas

(24, 50, 88, 93, 131, 145, 172). The plasmid-mediated IMP family of enzymes was first

described in Japan (128, 179). The VIM family was first described in Italy (92). Enzymes

from both of these carbapenamase families have been found in Pseudomonasspp. (75, 120).

Aminoglycoside-Modifying Enzymes

Although impermeability mutations can result in aminoglycoside resistance, especially in CF

and intensive care patients, drug inactivation by plasmid-encoded or chromosomally encoded

enzymes is the most common mechanism for resistance worldwide to this class of

antimicrobials (132). Aminoglycoside-modifying enzymes have been detected in P.

aeruginosa for over 30 years; these result in various combinations of resistance to

gentamicin, tobramycin, and/or amikacin. P. aeruginosa isolates, especially those from

Europe and Latin America, increasingly carry multiple modifying enzymes resulting in broadspectrum

aminoglycoside resistance. These enzymes are often encoded on transposons

and/or integrons that carry resistance determinants for other classes of antibiotics such as

sulfonamides, β-lactams, and chloramphenicol. Multiresistance genes for both

aminoglycosides and extended-spectrum β-lactamases and metalloenzymes are of particular

concern (132). Aminoglycoside-modifying enzymes can occur together with impermeability

mutations (102, 114), resulting in broad-spectrum aminoglycoside resistance. Broadspectrum

aminoglycoside resistance due to a gene (rmtA) encoding a 16S rRNA methylase

has been described (187).

Other

The discovery of a plasmid-borne quinolone resistance determinant (qnr) in gram-negative

organisms (110,178) is of significance for several reasons: (i) it has been transferred by

conjugation to multiple organisms, including P. aeruginosa; (ii) it is associated with highlevel

quinolone resistance (up to 250-fold increase in MICs); (iii) it appears to be associated

with integrons that carry determinants for resistance to β-lactams and aminoglycosides; and

(iv) it expands the spectrum of high-level plasmid-mediated resistance to quinolones.

Antibiotic Tolerance

Biofilm-producing P. aeruginosa isolates appear to be protected from killing by antibiotics

(166). Although this is widely accepted to indicate antibiotic resistance, a more appropriate

term is antibiotic tolerance. Although slower or stationary growth phase has classically been

thought to account for relative antibiotic tolerance, many other mechanisms have been

proposed. These include quorum sensing (152), decreased diffusion of antibiotics through

the matrix polysaccharide alginate (67), synthesis of glucans that specifically bind antibiotics

(103), phenotypic variability (30, 41), the presence of persister cells (160), and anaerobic

growth of biofilm bacteria, which affects the activity of multiple antibiotics (11, 62).

Multidrug Resistance

Worldwide, despite some geographic variability, antimicrobial resistance, including multidrug

(three or more antimicrobial classes) resistance to P. aeruginosa, is widespread and

increasing (49, 96). In 2003, the European MYSTIC study group reported considerable

country-to-country variation in the proportion of multidrug-resistant P. aeruginosa isolates

within Europe, ranging from 50% to less than 3% (53). The SENTRY Antimicrobial

Surveillance Program confirmed geographic variation in Latin America but emphasized the

rapid increase in multidrug-resistant strains, with rates approaching 35% (45). From 1993 to

2002, in the United States, the rates of multidrug resistance increased from 4 to 14%, with

the highest rates of increase reported for ciprofloxacin, imipenem, tobramycin, and

aztreonam (121). Globally, multidrug resistance was found in 10% of P. aeruginosa strains

analyzed (45).

Antimicrobial Susceptibility Testing

It may be difficult to estimate the true prevalence of antimicrobial resistance in P.

aeruginosa, as detection of resistance by routine tests agrees poorly with MIC data

(2, 68, 73, 99). Worldwide, susceptibility methods vary in terms of choice of media,

inoculum preparation, antimicrobial disc content, breakpoints, and interpretation of those

breakpoints. Even when these variables are taken into consideration, susceptibility testing

of P. aeruginosa remains challenging given the multiple mechanisms of resistance, both

intrinsic and acquired, which are frequently expressed concurrently, often at low levels.

In clinical laboratories, susceptibility testing for Pseudomonas species may be performed by

disc diffusion, agar or broth dilution, Etest (bioMerieux), or automated susceptibility systems

using broth microdilution. Disc diffusion tests perform satisfactorily for most clinical isolates

of P. aeruginosa (23). Limitations to this method include the lack of a quantitative result

(MIC) and the potential to miss low-level resistance. Etest has been shown to correlate well

with agar dilution for isolates from CF (14, 108) and non-CF (28) patients. Breakpoint

interpretation for disc diffusion zones and MICs are standardized, by the Clinical and

Laboratory Standards Institute (http://www.clsi.org/) in North America and by the European

Committee on Antimicrobial Susceptibility Testing (http://www.eucast.org) in Europe. Due to

the differences between these two organizations, susceptibility results should be reported

according to an individual institution’s operating procedures.

Good correlation with reference methods has been reported for most automated systems

(65, 80, 106, 162) when testing Pseudomonas isolates from non-CF patients. Results

evaluating the performance of various automated systems must be interpreted with caution,

as the number of isolates tested is often limited, especially for non-P. aeruginosa strains.

Whereas most P. aeruginosa isolates grow well on agar media, growth of some isolates in

broth is variable and may pose difficulties for laboratories that rely solely on automated

systems. Alternatively, a liquid medium improves the detection of the efflux resistance

phenotype, which may not be detected using solid-medium-based testing (1, 29).This may

account for some of the discrepancies reported when comparing different susceptibility

testing methods.

Several antibiotics pose specific challenges to susceptibility testing. Carbapenem

susceptibility testing results are difficult to interpret due to several factors, including rapid

imipenem degradation (173, 183), variable levels of efflux pump expression, and unstable

impermeability mutations. Carbapenemase is especially challenging, as it is associated with a

wide range of MICs and lacks a simple test for detection. Susceptibility testing of imipenem

with and without EDTA (discs or Etest strips) may be used but has been associated with

falsely resistant results. Ceftazidime, with or without EDTA, may be a better substrate than

imipenem and increases the sensitivity of the test (D. Livermore, presented at the British

Society for Antimicrobial Chemotherapy, Standardized Disc Susceptibility Testing Method

User Group Meeting, Royal College of Physicians, London, United Kingdom, 25 November

2003). Reproducibility of carbapenem resistance results using various susceptibility test

methods is poor, and it is recommended that initial carbapenem resistance be confirmed by a

second antimicrobial susceptibility test method (165). Although still restricted to reference

laboratories, there are PCR-based methods for detection of carbapenemase production

(120, 181).

Colistin is being used more in the treatment of multidrug-resistant P. aeruginosa. Disc

diffusion testing does not correlate well with MIC results, and underreporting of resistance

has been found in 5% of strains (46). Susceptibility testing of colistin should be performed

by a MIC method such as agar dilution, Etest, or broth microdilution. Prolonged incubation

(for 48 h) is recommended for broth microdilution (70).

Isolates of P. aeruginosa from CF patients pose specific difficulties for microbiology

laboratories. Isolates from these patients often exhibit mixed morphotypes including mucoid

phenotypes, small-colony variants, and bacterial microcolonies in biofilms. Susceptibility

testing is complicated by several factors, including lack of correlation between susceptibility

results and clinical response (154), different susceptibility patterns within a morphotype

(41), lack of reproducibility of susceptibility tests, undercalling resistance, and presence of

hypermutable strains (58, 126). Mucoid and nonmucoid phenotypes of P. aeruginosa are

often coisolated from patients with CF. Mucoid isolates tend to be more susceptible and have

lower β-lactamase activity than nonmucoid isolates (21). One explanation may be that these

isolates are protected from selective antibiotic pressure. Selective antibiotic pressure,

notably from inhalational tobramycin or colistin therapy, gives rise to small-colony variants

of P. aeruginosa with properties of increased antimicrobial resistance, autoaggregative

growth behavior, and enhanced ability to form biofilms (63, 64). In turn, bacterial cells in

biofilms adapt into symbiotic bacterial communities in which the mucoid alginate-producing

bacterial cells provide physical protection to the biofilm, while the highly antibiotic-resistant

nonmucoid cells protect against antibiotic killing (21). Increased ability of biofilm bacteria to

acquire resistance phenotypes (26) and selection of hypermutable strains following

antimicrobial therapy (58, 126, 127) may further explain the lack of eradication of P.

aeruginosa from chronically infected CF patients. Since bacteria found in biofilms exhibit

MICs 100- to 1,000-fold higher than free-living, planktonic bacteria (71, 117), routine

susceptibility testing may underestimate resistance and may contribute to treatment failures.

In a study of 597 CF isolates (14), both disc diffusion and Etest were found to be generally

acceptable as routine susceptibility testing methods. However, poor correlation was found

with disc diffusion testing of mucoid isolates for piperacillin, piperacillin-tazobactam, and

meropenem. Underreporting of resistance was more frequent with disc diffusion than with

Etest, especially when testing ceftazidime, piperacillin, and piperacillin-tazobactam.

Hypermutable strains may be detected using either disc diffusion or Etest methods by the

presence of resistant mutant subpopulations within the inhibition zones of three or more

antibiotics (101).

Mucoid isolates pose a specific challenge for automated systems (6, 13, 29). Overestimating

susceptibility may occur, as mucoid isolates often demonstrate insufficient growth at 24 h.

Automated systems that allow for longer incubation may be preferable. On the other hand,

overcalling resistance may result from the presence of large amounts of exopolysaccharide,

resulting in turbidity without adequate bacterial growth. These limitations have led many

microbiologists who routinely work with mucoid isolates of P. aeruginosa to choose

alternative methods for susceptibility testing.

Isolating and individually testing all the morphotypes of P. aeruginosa is labor-intensive and

time-consuming and may not provide clinically relevant susceptibility results. Mixed

morphotype testing using phenotypically different colonies directly from sputum cultures, or

from subcultures of isolated colonies, has been shown to correlate well with disc diffusion

and MIC susceptibility methods (31, 184) and may provide clinically useful susceptibility data

with significant time and cost savings. However, the correlation appears to be better for

susceptible strains than for resistant strains (116). Direct sputum susceptibility testing using

the Etest method has been suggested as an alternative to morphotype testing in assessing

the in vivo situation by evaluating bacterial population susceptibility as well as potential

interactions with other organisms, including commensal microbes (151; M. Gallagher,

presented at the International Cystic Fibrosis Conference, Stockholm, Sweden, 2000).

Other methods have been recommended in an attempt to better predict susceptibility

results. Biofilm susceptibility assays have been developed which confirm that biofilm

inhibitory concentrations are much higher than conventionally determined MICs for multiple

antibiotics (117). Synergy testing using microtiter checkerboard, time-kill test, broth

macrodilution breakpoint combination sensitivity test, or Etest methods (10,151, 182) has

been used to assess the activities of antibiotic combinations in vitro in order to predict in vivo

synergistic activity. This testing is labor-intensive, time-consuming, and difficult to

reproduce, and it remains controversial, as very few clinical data exist demonstrating

correlation with prediction of outcomes.

Susceptibility testing of Pseudomonas species other than P. aeruginosa is rarely indicated,

and clinical correlation is required before susceptibility testing is performed. These organisms

are generally susceptible to most antipseudomonal antibiotics as well as to trimethoprimsulfamethoxazole

(except most P. fluorescens/putida isolates), a property that differentiates

them from P. aeruginosa. P. fluorescens, P. putida, andP. oryzihabitans may be more

resistant to aztreonam and ticarcillin-clavulanate. P. stutzeri is usually very susceptible to all

antipseudomonal agents (146).

EVALUATION, INTERPRETATION, AND REPORTING OF

RESULTS Back to top

P. aeruginosa may be associated with colonization or clinically significant infections.

Interpretation of the Gram stain often directs the further workup of this organism. The

presence of small clusters of gram-negative organisms surrounded by amorphous material is

indicative of biofilm formation compatible with a chronic infection. This finding should be

reported to physicians, and incubation should be prolonged, as these isolates usually exhibit

slower growth characteristics. The presence of these organisms intracellularly in

polymorphonuclear cells is a strong indication of true infection rather than colonization.

Isolation of P. aeruginosa from sterile body sites should always be interpreted as indicative of

probable infection. Isolation in mixed culture requires correlation with the direct smear, other

organisms isolated, and clinical history. Isolates from sites of chronic infection, such as CF

respiratory sites, often exhibit multiple morphotypes that can make identification difficult.

Molecular methods increasingly are finding a role in the identification of this organism,

especially for epidemiological studies. Susceptibility testing of this organism is difficult,

especially for mucoid isolates, due to increasing resistance, lack of reproducibility of results,

and lack of clinical correlation. Piperacillin and piperacillin-tazobactam results obtained from

automated systems may be unreliable forPseudomonas spp., and in particular for mucoid

isolates, and results should be confirmed by disc diffusion or Etest systems. A basic

understanding of the multiple mechanisms of resistance, both intrinsic and acquired, is

essential to interpret susceptibility testing results and give therapeutic recommendations to

physicians. OtherPseudomonas species are infrequently isolated in the laboratory and are

usually not clinically significant. Clinical correlation and correlation with the Gram stain are

essential before further workup is undertaken.