Francisella and Brucella


Taxonomy

The family Francisellaceae, a member of the gamma subclass of proteobacteria, consists of

the single genus F rancisella. Francisella tularensis, Francisella novicida, Francisella

philomiragia, and Francisella noatunensis as well as unclassified Francisella spp. comprise the

genus. Several publications have reviewed the history and descriptions of members of the

genus Francisella (48, 165, 166).

F. tularensis is the causative agent of tularemia, a zoonosis affecting a wide range of animals

and humans. Three subspecies of F. tularensis, tularensis (type A), holarctica (type B),

and mediasiatica, displaying >99.8% identity in their 16S rRNA genes, have been described

(165). F. tularensis subsp. tularensis and F. tularensissubsp. holarctica have been further

subdivided into distinct subpopulations using molecular typing methods (117). F.

tularensis subsp. tularensis has been separated into three subclades, termed A1a, A1b, and

A2, and within F. tularensis subsp. holarctica, 10 separate subclades have been identified

(117).

F. novicida and F. philomiragia, originally isolated from salty water, are only infrequently

associated with human disease (74, 81, 92, 191). F. novicida was classifed as a species

distinct from F. tularensis on the basis of serologic, virulence, and biochemical comparisons

(92, 133). Subsequently, similarities between F. novicida andF. tularensis were identified

based on genome reassociation studies (75% DNA relatedness) and 16S rRNA gene

sequences (>99.8% identity) (74, 165). Recent whole-genome comparisons provide

evidence that the evolutions of F. novicida and F. tularensis occurred separately, consistent

with F. novicida and F. tularensisbeing two separate species (93). Between F.

philomiragia and F. tularensis, genome reassociation studies show 39% average DNA

relatedness and >98.3% identity in their 16S rRNA gene sequences (57, 74).Francisella

noatunensis has been linked only to environmental sources being recovered from a number

of saltwater fish species (27, 114, 131). DNA hybridization studies show a mean

reassociation value of 68% between F. noatunensis and F. philomiragia, with the two species

displaying 99.3% identity in their 16S rRNA genes (114).

Several additional Francisellaceae members, isolated in culture, including Wolbachia

persica (165) and Francisellaspp. recovered from seawater (143), have been described.

Additionally, a Francisella sp. was isolated from the blood and cerebrospinal fluid (CSF) of

two different patients in the United States in 2005 and 2006 (90), while

another Francisella sp. was isolated from the blood and urine of a patient in Spain (50).

Characterization data indicate that these Francisella spp. are distinct from F. tularensis, F. n

ovicida, and F. philomiragia.

Putative Francisellaceae family members have also been identified based on shared identities

among 16S rRNA gene sequences. These include Francisella-like endosymbionts (FLEs)

in Amblyomma maculatum and multipleDermacentor and Ornithodoros tick species

(160, 165). DNA from putative Francisellaceae family members has also been detected in soil

and water samples (22).

Description of the Genus

The genus Francisella comprises tiny gram-negative coccobacilli that can be distinguished

from similar genera by several features (Table 1). Members of the genus take up Gram’s

counterstain (safranin) poorly; they are strict aerobes, weakly catalase positive, urease

negative, nonmotile, and non-spore-forming, and they react with a limited number of

carbohydrates (Table 2). Only a few sugars (glucose, maltose, sucrose, and glycerol) are

utilized by most members of the genus. Acid is produced without gas. Unique cellular fatty

acids are associated with the genus (Table 1) (74, 165, 192). For most species, in vitro

growth is enhanced by sulfhydryl supplementation. Cultured members of the genus share

>97% identity within their 16S rRNA genes (90, 114,165).



A few key differences separate species of the Francisella genus (Table 2). F.

philomiragia and F. novicida are more biochemically reactive than F. tularensis. F.

philomiragia is oxidase positive by Kovacs’s modification, whereas F. novicida and F.

tularensis are oxidase negative. F. novicida and F. philomiragia differ from F. tularensis by

their ability to grow independently of cysteine supplementation and by their comparatively

large cell size. Additionally, levels of virulence in mice differ, with <10 cells of F.

tularensis being required to kill laboratory mice compared to <100 for F. novicida (133).

Nucleotide differences within 16S rRNA genes also discriminate F. tularensis, F.

novicida, and F. philomiragia (56).

Biochemical, molecular, and virulence differences also differentiate F. tularensis subspecies.

Glycerol fermentation and citrulline ureidase activity distinguish F.

tularensis subsp. tularensis from F. tularensis subsp.holarctica (Table 2) (127, 128). PCRbased

techniques as well as 16S rRNA gene sequencing can also differentiate F.

tularensis subsp. tularensis from F. tularensis subsp. holarctica (56, 197). Levels of virulence

in rabbits differ, with the 100% lethal dose of F. tularensis subsp. tularensis being <10 cells

when inoculated subcutaneously, compared to 109 cells for F.

tularensis subsp. holarctica (127). F. tularensis subsp. mediasiaticadiffers from both F.

tularensis subsp. tularensis and F. tularensis subsp. holarctica by its inability to utilize

glucose and its moderate pathogenicity in rabbits (128, 165).

Epidemiology and Transmission

Francisellaceae are widely distributed in the natural environment. Some species are obligate

pathogens of animals and humans (F. tularensis) (Table 3); others are present in the

environment (water) and are opportunistic pathogens of humans (F. novicida and F.

philomiragia) (Table 3), while others have been associated only with animals (F. n

oatunensis). F. tularensis isolates have been found only within the Northern Hemisphere

(Holarctic region), whereas F. novicida and F. noatunensis have been recovered from both

the Northern and Southern Hemispheres (27, 165, 166, 194).



The geographic distribution of F. tularensis varies throughout the Northern Hemisphere

(128, 165, 166). Infections caused by F. tularensis subsp. tularensis occur only in North

America, whereas F. tularensis subsp. holarctica has a much wider distribution, causing

disease in both the Old and New Worlds. F. tularensis subsp.mediasiatica has been found

only in regions of central Asia. The geographic distributions of F.

tularensis subsp.holarctica and subsp. tularensis subclades also vary. Subclades of F.

tularensis subsp. holarctica within North America are distinct from subclades throughout the

rest of the Northern Hemisphere, excepting Scandinavia (188). Within the United States, the

geographic distributions of F. tularensis subsp. tularensis subclades differ. A1a and A1b

predominate in the eastern half of the United States, whereas A2 appears restricted to the

western United States (91).

F. novicida has been isolated primarily in North America, with occurrences of F. novicida-like

organisms in Australia and Thailand (94, 194). Of the fewer than 20 known isolates of F.

philomiragia, most have been from North America, with three single incidences reported

from Central and Eastern Europe and Australia (60, 74,164, 191, 194). F. noatunensis has

been isolated from Scandinavia, Chile, and Japan (27, 114, 131).

F. tularensis is associated with a wide range of hosts; more than 100 species of wild animals,

birds, and arthropods have been found naturally infected (24, 75, 80). Habitats where the

Lagomorpha (Sylvilagus[rabbits], Lepus [hares], and Oryctolagus [Old World hares] genera)

and the Rodentia (water voles, muskrats, lemmings, voles, and beavers) thrive are believed

to be important in maintaining enzootic foci (75, 80). Biting arthropod vectors, primarily

tabanid flies (Chrysops and Tabanus), ticks (Dermacentor and Amblyomma), and mosquitoes

(Aedes and Ochlerotatus in Sweden and Russia), are implicated in the transmission of

tularemia (46, 75, 79, 144). Human infections with F. tularensis are acquired usually via

inhalation of infective aerosols, ingestion of contaminated water, handling of sick or dead

animals, or ingestion of infected meat or from the bite of an infected arthropod. In contrast

to F. tularensis, F. novicida and F. philomiragia do not appear to infect a wide range of hosts.

Their primary natural habitat appears to be saltwater. Only a single isolation of F.

philomiragia from an animal host (muskrat) has been documented (81). Neither F.

novicida nor F. philomiragiahas been shown to be transmitted to humans by the bite of an

infective arthropod, and arthropods infected with these species have not been identified.

Tularemia has been reported from many countries of the Northern Hemisphere. Foci of

endemicity have long been documented in Russia and Kazakhstan, as well as Finland and

Sweden (166). Annual cases are usually reported from most countries in Eastern Europe.

Cases of tularemia are reported less frequently from Western Europe; however, several

outbreaks of tularemia have occurred over the last decade in Spain (5, 108). Other areas

reporting outbreaks comprising many hundreds of cases in the last decade include Sweden

and Kosovo (139, 150).

In the United States, the incidence of tularemia has steadily declined to 100 to 200 cases

each year since the 1940s, when several thousand cases of tularemia were reported annually

(31). All states except Hawaii have reported cases. A total of 1,368 human cases from 44

states were reported to the Centers for Disease Control and Prevention (CDC) between 1990

and 2000, with four states, South Dakota, Arkansas, Missouri, and Oklahoma, accounting for

the majority (56%) of cases (31). Most cases in the United States are sporadic, with patients

acquiring tularemia from tick or deerfly bites or from contact with infected animals (e.g., by

skinning rabbits) (31, 91). Inhalation of an infective aerosol during landscaping and contact

with infected cats are also important modes of transmission (53, 54, 91). The majority of

tularemia cases occur in the summer months. Outbreaks of tularemia occur rarely in the

United States. An outbreak of pneumonic tularemia associated with landscaping activities

occurred on Martha ’s Vineyard in 2000 (54), and since that time, Martha ’a Vineyard has

recorded yearly cases (109). The most recent outbreak in the United States occurred in Utah

in 2007 and was attributed to deerfly bites (142). Cultures recovered from human cases in

the United States from 1996 to 2001 comprise equal proportions of F.

tularensis subsp. tularensis and F. tularensis subsp.holarctica (174).

Clinical Significance

F. tularensis

Tularemia is caused by F. tularensis. The disease has been known historically by a number of

synonyms, such as rabbit fever, deerfly fever, market men’s disease, the glandular type of

tick fever, Ohara ’s or yato-byo disease, and water rat trappers’ disease, attesting to the

variety of clinical presentations, the infectious agent’s ubiquitous presence in nature, and the

means by which humans may acquire the infection.

The clinical spectrum of tularemia depends on the mode of transmission, the virulence of the

infecting strain, the immune status of the host, and timely diagnosis and treatment (197).

Tularemia can be misdiagnosed since its symptoms are not unique; a sudden onset of chills,

fever, headache, and generalized malaise characterize the onset of illness. The differential

diagnosis includes a wide range of infectious diseases, such as cat scratch fever,

mycobacterial infections, anthrax, brucellosis, legionellosis, and plague (197). The incubation

period averages 3 to 5 days.

Patients may present with any one of the clinical forms of tularemia: ulceroglandular,

glandular, oculoglandular, oropharyngeal, typhoidal, and pneumonic (197). The most

common form is ulceroglandular disease (45 to 80% of the reported cases), where the portal

of entry is via an infective arthropod bite or other inoculation through the skin barrier.

Glandular tularemia is similar to ulceroglandular disease but lacks the ulcerated site of

infection. Oculoglandular tularemia occurs when the conjunctiva is the initial site of infection,

usually as a result of the mechanical transfer of organisms from an infectious source to the

eye by the fingers. Oropharyngeal tularemia occurs from ingestion of contaminated water or

food and is associated with pharyngeal lymphadenopathy. Pneumonic tularemia occurs by

direct inhalation of the organism and is considered the most severe form of the disease.

Typhoidal tularemia is the most difficult form to recognize because there is no identified

portal of entry and localized signs are absent. If untreated, bacterial dissemination from the

primary sites of infection can lead to secondary clinical presentations, such as sepsis and

meningitis.

The severity of infection can range from mild and self-limiting to fatal and is largely

dependent on the infecting strain. Little to no tularemia-related mortality is reported in

Europe and Asia, where only F. tularensis subsp.holarctica causes tularemia, while mortality

in the United States ranges between 2.3% (those diagnosed by culture and serology) and

9% (culture-confirmed cases only) (52, 174). Infections due to F.

tularensis subsp.tularensis are considered to be more severe than infections caused by F.

tularensis subsp. holarctica. Molecular epidemiologic studies have refined this picture and

demonstrate that among culture-confirmed cases in the United States, human infections due

to the A1b subclade of F. tularensis subsp. tularensis result in significantly higher mortality

(24%) than infections caused by F. tularensis subsp. tularensis subclades A1a (4%) and A2

(0%) or F. tularensis subsp. holarctica (7%) (91). Comparisons of levels of virulence in mice

corroborate the differences in virulence among F. tularensis subsp. tularensis clades and also

between F. tularensis subsp.tularensis clades and F. tularensis subsp. holarctica (118).

F. novicida and F. philomiragia

Human infections caused by F. novicida and F. philomiragia are rare and considered

opportunistic infections, infecting primarily patients with underlying immunocompromising

conditions. Fewer than 20 cases of F. philomiragia infection have been d escribed since the

discovery of this species in 1974 (60, 74, 164, 191). All but one case have involved a host

with an impaired physical barrier to infection (near d rowning) or an impaired immunologic

defense system (chronic granulomatous disease or myeloproliferative disease). In most

cases, F. philomiragia was isolated from normally sterile sites: blood, bone marrow,

cerebrospinal fluid, and pericardial fluid. The drowning and water exposure cases were

associated with saltwater and brackish water, in contrast to F. tularensis infections, which

are associated with freshwater sources. Fewer than 10 cases of F. novicida or F. novicida-like

infections have been reported (26, 35, 74, 94, 191, 194). Isolates were recovered from

blood, lymph node tissue, and wounds.

Collection, Handling, Storage, and Transport of Specimens

Personnel handling diagnostic cultures of F. tularensis are at considerable risk for infection.

Due to the extremely low infectious dose for F. tularensis, tularemia has been one of the

most commonly reported laboratory-associated bacterial infections (132, 145). Even though

the use of biological safety cabinets and prophylactic antibiotic therapy (as well as

vaccination, where available) provides safeguards for laboratory workers, these precautions

have not fully eliminated laboratory exposures or modified practices in the clinical laboratory

to minimize risks (155, 162). Manipulation of cultures presents the greatest risk to

laboratory workers, due to the high concentration of organisms. Clinical specimens should be

handled under biosafety level 2 (BSL-2) conditions with universal precautions, with work

being done using BSL-3 practices as soon as a suspect F. tularensis is isolated in culture

(44). The greatest risk of laboratory-acquired tularemia is by aerosol inhalation while

working with cultures. Any work with clinical samples that generates aerosols should be

performed using BSL-3 practices. Laboratories may want to consider developing policies that

encourage physicians to alert the laboratory if a diagnosis of tularemia is expected.

The choice of specimen for diagnostic testing is generally dependent on the form of clinical

illness: ulceroglandular, glandular, oculoglandular, oropharyngeal, pneumonic, or typhoidal

tularemia (197). Whole blood is an acceptable specimen for all clinical forms of tularemia,

although this sample may be negative, particularly if the disease is in the early stages of

progression. Serum for antibody detection is a standard specimen taken for diagnosis of all

forms of illness. A first specimen should be collected as early in the course of infection as

possible, followed by a second specimen taken in the convalescent period (at least 14 days

apart and preferably 3 to 4 weeks after onset of symptoms). Pharyngeal swabs,

bronchial/tracheal washes or aspirates, sputum, transthoracic lung aspirates, and pleural

fluid are appropriate specimens for pneumonic, typhoidal, or oropharyngeal tularemia. Swabs

of visible lesions or affected areas have been used most commonly for ulceroglandular and

oculoglandular tularemia. Aspirates from lymph nodes or lesions can be used for diagnosis of

ulceroglandular, glandular, and oropharyngeal tularemia. Necropsied materials from animals

that are appropriate for testing include samples from visible abscesses as well as samples

from lymph node, lung, liver, and spleen tissues and bone marrow.

For specimens to be tested by culture, it is important where possible to decontaminate the

surface area prior to specimen collection since contamination of the sample with normal flora

could interfere with the interpretation of culture results. To minimize loss in viability,

specimens should be delivered to the laboratory within 24 hours and preferably within 2

hours. In general, if transport is >24 hours, specimens should be stored chilled (2 to 8°C) in

an appropriate medium until processed in the laboratory. Freezing of samples, unless in a

preservative environment, such as tissue specimens or glycerol-containing solutions, is not

recommended because of the lysis of live bacteria upon thawing. Swabs should be placed in

Amies agar with charcoal, a commercial transport system designed for anaerobic and aerobic

pathogens (Becton, Dickinson and Company, Franklin Lakes, NJ). F. tularensis should remain

viable for 7 days at ambient room temperature when stored in Amies medium (82). Stuart

medium, designed for transporting gonococcal specimens, and saline are inadequate for

keeping F. tularensis viable during transport (82). For serum samples, separation from blood

should take place as soon as possible, preferably within 24 hours. Sera may be stored at 2 to

8°C for up to 10 days. If testing is delayed for a long period, serum samples may be frozen.

For PCR, specimens should be collected in guanidine isothiocyanate-containing buffer, which

preserves F. tularensis DNA for up to 1 month (82). Arthropods may be stored intact in 2%

NaCl for culture analysis or in ethanol for molecular testing.

Direct Detection in Specimens

Fresh clinical specimens (ulcer and wound swabs, tissues, and aspirates) where the

concentration of organisms might be expected to be high can be directly examined by

microscopy by Gram straining, direct fluorescent-antibody (DFA) binding, or

immunohistochemistry. Under microscopic examination of Gram-stained

specimens, Francisella cells (single and pleomorphic) appear tiny and counterstain so faintly

with safranin that they can easily be missed. Basic fuchsin counterstains F. tularensis better

than safranin. Due to the small size of F. tularensis, Gram staining of clinical specimens is

usually of little diagnostic value. DFA staining using a fluorescein isothiocyanate-labeled

hyperimmune rabbit polyclonal antibody directed against whole, killed F. tularensis cells can

be used to presumptively identify F. tularensis subsp. tularensis and F. tularensis subsp.

holarctica in clinical specimens (197). This DFA reagent does not react well or at all with F.

novicida or F. philomiragia. Immunohistochemical (IHC) staining using a monoclonal

antibody directed against the lipopolysaccharide (LPS) has been used successfully to

visualize F. tularensis in formalin-fixed tissues (71). Neither the DFA reagent nor the IHC

reagent is commercially available. DFA testing of specimens is provided by many reference

laboratories in the United States.

Because of the relative rarity of human tularemia, evaluation of molecular diagnostics with

clinical specimens has been challenging. PCR-based diagnostic methods have been used

most commonly for diagnosis of ulceroglandular tularemia, the most prevalent clinical form.

DNA detection by conventional PCR directed at thetul4 gene (unique to Francisella spp.) has

been successfully and widely applied for diagnosis of ulceroglandular tularemia

(82, 167, 197). The tul4 PCR assay displays a sensitivity of 75% when applied to wound

specimens from patients with ulceroglandular tularemia and was shown to be more sensitive

than culture (sensitivity of 62%) (82). More recently, real-time PCR assays using the TaqMan

5′ nuclease assay have been applied to detect F. tularensis DNA in a variety of clinical

specimens, including ulcer specimens, aspirates, throat swabs, bronchial washes, and pleural

fluid (30, 186, 195). These assays target multiple Francisella genes

(ISFtu2element, iglC, tul4, and fopA genes). A limitation of most PCR-based diagnostics

for F. tularensis is the inability to discriminate F. tularensis from F. novicida. While this may

not be significant for patient management, the correct identification of species has both

epidemiologic and public health value. Routine incorporation of real-time PCR into the

diagnosis of human tularemia remains in need of further evaluations and standardization.

Detection of F. tularensis in ectoparasites is based primarily on PCR methods (65, 203). Of

note, FLEs, present in a wide range of tick species, have been shown to cross-react with

molecular targets used for the detection of F. tularensis, leading to false-positive results

(89, 160, 173). If molecular assays are to be used for screening environmental samples

for F. tularensis, it is important to evaluate the assays for cross-reactivity with

other Francisellaceae members present in these sample types (22, 143). 16S rRNA gene

sequencing can also be helpful for discriminating F. tularensis from FLEs or

environmental Francisella spp. (22, 89).

Isolation Procedures

F. tularensis is slow growing and fastidious, requiring supplementation with sulfhydryl

compounds (cysteine, cystine, thiosulfate, and IsoVitaleX) to grow on artificial medium. F.

tularensis grows on several media common to clinical laboratories, including chocolate agar

(CA) (Fig. 1A), buffered charcoal-yeast extract agar (BCYE), and Thayer-Martin agar. It also

grows in thioglycolate broth. When supplemented with 1 to 2% IsoVitaleX, general

bacteriological media (tryptic soy broth and Mueller-Hinton broth) support the growth of F.

tularensis. The organisms grow slowly (60-min generation time); good growth, therefore, is

obtained by prolonged incubation (48 hours or longer). Suspect cultures should be incubated

at 35 to 37°C aerobically and observed daily for up to 14 days; CO2 does not impede growth.

A specialized medium, cysteine heart agar supplemented with 9% chocolatized sheep blood

(CHAB), is often used for the growth of F. tularensis in reference laboratories (197). F.

tularensis grows well on CHAB, as this is a high-nutrient medium. Additionally,F.

tularensis displays distinctive colony morphology on CHAB5 which can aid in identification

(Fig. 1). F. tularensis does not grow well or at all on general bacteriologic media, such as

sheep blood agar. Nutritionally enriched specimens, such as blood or tissue, provide an

intrinsic source of sulfhydryl compounds that may initially permit F. tularensis growth on

general bacteriological media. Upon subculture, the fastidious nature ofF. tularensis will

become evident as the exogenous compounds are depleted, leading to the loss of the

bacterium’s viability unless the subculture is propagated on cysteine-supplemented medium.



Clinical samples from normally sterile sites can be plated on nonselective agars that support

the growth of F. tularensis. Care should be taken not to permit laboratory contamination

of F. tularensis isolates by environmental bacteria such as Staphylococcus, as these bacteria

rapidly outcompete and inhibit the growth ofF. tularensis on nonselective media (141).

Clinical specimens obtained from nonsterile sites or autopsy and environmental sources

should be plated on antibiotic-containing media. Incorporation of an antibiotic supplement

(7.5-mg/liter colistin, 2.5-mg/liter amphotericin, 0.5-mg/liter lincomycin, 4.0-mg/liter

trimethoprim, and 10-mg/liter ampicillin) into the medium has been demonstrated to

prevent other organisms from overwhelming F. tularensis (141). Commercially available

antibiotic-containing media that support the growth of F. tularensis include improved Thayer-

Martin agar (Remel, Lenexa, KS), BCYE selective agar containing polymyxin B, anisomycin,

and vancomycin (Remel or Becton, Dickinson), and cysteine heart agar containing antibiotics

(polymyxin B and penicillin) (Remel).

Specimens for culture should be taken on the basis of clinical presentation and before

administration of antibiotics. Fresh clinical material that is likely to contain high

concentrations of F. tularensis organisms, such as ulcer and wound specimens and lymphoid

tissue (liver, spleen, or affected lymph node tissue), is inoculated directly onto an agar plate

by using a sample-laden swab or bacteriological loop. Larger inocula are necessary for the

recovery of F. tularensis from specimens that contain a lower concentration of the

organisms, such as aspirates of pharyngeal washes, bronchial wash fluids, pleural fluids, and

environmental samples. Whole blood should be directly inoculated into blood culture bottles

(197), with subsequent culture of blood culture-positive specimens on agar.

Identification

Because of the rarity as well as the sporadic nature of most tularemia cases, the organism is

often not easily identified when it is cultured. The isolation of a very tiny (individual cells may

be difficult to discern), poorly counterstaining (by safranin), gram-negative coccobacillus

(Fig. 2A) that produces 1- to 2-mm-diameter gray to gray-white colonies on CA after 48

hours (and scant to no growth on blood agar) should raise suspicion for F. tularensis. F.

tularensis is oxidase negative, weakly catalase positive, β-lactamase positive, X and V factor

negative, and urease negative. Oxidase, X/V factor, and urease testing can help

differentiate F. tularensisfrom other similar gram-negative organisms,

including Brucella spp., Haemophilus influenzae, and Acinetobacterspp. (Table 1). If further

testing according to the algorithms given in chapter 12 does not rule out F. tularensisand the

laboratory cannot do further confirmatory tests, the isolate should be sent to a reference

laboratory that can confirm (or rule out) its identification as F. tularensis. Proper biohazard

shipping procedures must be followed (see chapter 12 in this Manual). In the United States,

all states have at least one reference laboratory that is part of the Laboratory Response

Network (LRN). These LRN laboratories are able to confirm the identification of bacterial

select agents, including F. tularensis. See also http://www.bt.cdc.gov/lrn].



The colonial morphology of F. tularensis is most distinctive when it is grown on CHAB. On

CHAB, F. tularensisexhibits a prominent and unique opalescent sheen due to its production of

H2S; this sheen is less prominent in F. philomiragia than in F. tularensis and is absent in

cultures of other gram-negative organisms, such

asYersinia, Brucella, Haemophilus, and Pasteurella spp. On CA, BCYE, and Thayer-Martin

agar, F. tularensis colonies have an entire margin; they are gray, smooth, raised, and moist,

with a butyrous consistency. F. novicidagrows more robustly than F. tularensis and is

cysteine independent (Table 2). Colonies of F. philomiragia on CA are >5 mm in diameter,

with an entire margin; they are white, smooth, raised, mucoid, and cysteine independent.

Isolates can be identified as F. tularensis using antigen or molecular detection methods,

including slide agglutination, DFA staining (Fig. 3), PCR, or sequencing. The slide

agglutination test identifies a suspect culture by mixing commercially available (Becton,

Dickinson and Company) polyclonal rabbit anti-F. tularensisantibody with suspect cultures;

the polyclonal rabbit anti-F. tularensis antibody does not react well or at all

with F. novicida or F. philomiragia (197). DFA staining can also be used to identify F.

tularensis subsp. tularensisand F. tularensis subsp. holarctica isolates (197). Care should be

taken to ensure that prepared smears are not too thick, as this can interfere with the

performance and interpretation of the test. In general, antigen-based identification methods

work optimally when fresh cultures (24 hours) are tested. If a culture older than 24 hours is

to be tested by antigen detection methods, a fresh subculture should be prepared. PCR

methods targeting F. tularensis-specific genes can also be used for the identification of

suspect cultures (167, 186). As described earlier in this chapter, it is important to consider

that many F. tularensis PCR assays cross-react withF. novicida. Therefore, it may be

important to rule out F. novicida as the cause of infection, particularly in areas where

tularemia has not been previously reported. 16S rRNA gene sequencing can be useful if the

isolate is not easily identified as F. tularensis (5, 35). Universal 16S rRNA gene primers

or Francisella-specific 16S rRNA gene primers can be used (57).



Typing Systems

Once an isolate has been identified as a Francisella sp., supplemental tests can be used for

additional characterization, including typing of species, subspecies, and strain. Oxidase can

be used to differentiate F. philomiragia from F. novicida and F. tularensis (Table 2). Glycerol

fermentation and citrulline ureidase activity distinguish F. tularensis subsp. tularensis from F.

tularensis subsp. holarctica (Table 2); conventional assays for these biochemical tests have

been described (157). The 96-well automated MicroLog Micro Station system with GN2

microplates (Biolog Inc., Hayward, CA) can also be used to assess the glycerol fermentation

of F. tularensis (197). Genus, species, and subspecies can be typed by sequence analysis of

the 16S rRNA gene (56, 57). PCR methods (both conventional and real-time) can type

isolates at the level of genus, species, subspecies (F. tularensis subsp.

tularensis or F. tularensis subsp. holarctica), and subclades (F.

tularensissubsp. tularensis subclades A1 and A2) (116, 197). Pulsed-field gel electrophoresis

can differentiate the threeF. tularensis subsp. tularensis subclades, A1a, A1b, and A2 (91).

For discrimination of individual strains, a multilocus variable-number tandem-repeat assay

(MLVA) for F. tularensis, based on 25 different repeats in the genome (83), has been

utilized.

Serologic Tests

Antibodies may be detected as early as 1 week after the onset of symptoms (about 2 weeks

after infection). By 2 weeks after onset, antibodies may be detected in 89 to 95.4% of

samples. Antibodies can persist for more than 10 years (25, 190). Immunoglobulin M (IgM),

IgA, and IgG antibodies may appear simultaneously (178, 187). IgM antibodies can last for

many years; thus, their presence does not always indicate early or recent infection

(25, 178).

Agglutination testing, either by the tube agglutination (TA) or the microagglutination (MA)

method, is a standard serology test for determining the presence of antibodies in tularemia

(25, 29, 197). Formalin-killed antigen (prepared from F. tularensis subsp. tularensis strain

Schu S4) is commercially available from Becton, Dickinson. Formalin-killed F.

tularensis antigen is also prepared within reference laboratories worldwide. In the United

States, a single specimen with a TA titer of ≥1:160 or an MA titer of ≥1:128 is considered

positive. Formalin-killed F. tularensis whole-cell antigen may display low-level crossreactivity

with Brucella antibodies (23, 125). No cross-reactivity of F. novicida or F.

philomiragia sera has been observed with F. tularensis-killed cells (133). Enzyme-linked

immunosorbent assays (ELISAs) have been adopted for use in the parts of Europe where

tularemia is endemic (25, 146, 159, 197). The LPS and/or outer membrane fraction remains

the primary ELISA antigen used in test applications. Antigenic differences between F.

tularensis subsp. tularensis and subsp. holarctica have not been identified for use in serology

assays. Thus, serology assays do not distinguish the infecting subspecies. This is of most

importance in North America, where both F. tularensis subsp.tularensis and F.

tularensis subsp. holarctica cause tularemia.

F. tularensis organisms are intracellular bacteria and are capable of eliciting both humoral

and cell-mediated immunity (178). The latter response has been known to remain strong 25

years after infection (49). Host T cells retain proliferative responses to unique F.

tularensis membrane proteins, with concomitant increases in interferon and interleukin-2

levels (49, 175, 176, 178). Tests for measuring the cell-mediated immune response are

specialized and are not routinely used for diagnosis of tularemia (47).

Antimicrobial Susceptibilities

F. tularensis infections are treatable with narrow-spectrum antibiotics. All Francisella isolates

examined to date are β-lactamase positive, so penicillins and cephalosporins are not effective

and should not be used to treat tularemia. Antibiotics recommended for treatment and

prophylaxis include chloramphenicol, ciprofloxacin, gentamicin, streptomycin, and

tetracycline. Antimicrobial susceptibility testing (broth microdilution and Etest) of a large

collection of F. tularensis isolates worldwide has demonstrated no antimicrobial resistance to

drugs used for treating tularemia (77, 177, 181, 182, 184). Some isolates of F.

tularensis subsp. holarctica from Europe and Russia are erythromycin resistant.

Antimicrobial susceptibility testing of F. tularensis is not usually performed in clinical

microbiology laboratories because of safety concerns in working with this organism and

because resistance to antibiotics used for clinical treatment of tularemia has never been

reported (179). The Clinical and Laboratory Standards Institute (CLSI) has published

interpretative criteria and quality control limits for broth microdilution of F. tularensisusing

Mueller-Hinton medium supplemented with 2% IsoVitaleX (36, 37).

Evaluation, Interpretation, and Reporting of Results

Serology is generally the most common method for laboratory confirmation of F.

tularensis infection, due largely to the organism being slow growing and fastidious.

Nonetheless, culture provides a conclusive diagnosis of infection and whenever possible

should be attempted using appropriate biosafety measures. Isolation of a very tiny gramnegative

bacterium that shows fastidious growth characteristics and is oxidase negative,

weakly catalase positive, urease negative, X/V factor negative, and β-lactamase positive

should be strongly suspected as F. tularensis and referred to a reference laboratory.

Confirmation of F. tularensisinfections includes (i) identification of a culture as F.

tularensis and/or (ii) a fourfold difference in titers in acute- and convalescent-phase serum

samples, with one of the paired samples having a positive titer. A positive test result for a

primary clinical specimen using antigen or molecular detection methods, including DFA

staining, IHC staining, or PCR, provides only a presumptive diagnosis of F. tularensis. A

single positive serum sample is also considered presumptive for tularemia. For all cases

presumed to be tularemia, it is necessary to verify that the patients’ symptoms are

compatible with tularemia and, in the case of a single positive titer, that the patient has not

been previously vaccinated.

In the United States, F. tularensis is classified as a select agent. To transfer, receive, or

possess F. tularensis,laboratories must be registered both with the CDC and with the Animal

and Plant Health Inspection Service (APHIS) of the U.S. Department of Agriculture. The

registration process includes a U.S. Department of Justice investigation of all personnel

having access to select agents. Clinical laboratories are exempt from the registration

requirement provided that within 7 calendar days of identifying one of these agents, they

transfer it to a registered entity and/or destroy the agent on site. Laboratories identifying an

organism as F. tularensisare required to report this finding immediately to the CDC. Report

forms, contact information, laboratory registration information, and pertinent citations of the

U.S. Federal Code may be found athttp://www.cdc.gov/od/sap.

BRUCELLA Back to top

Brucella spp. are common zoonoses among domestic animals and among wildlife, including

novel species of marine mammals. Brucella spp. also cause infections in humans and can

mimic other infectious and noninfectious diseases, posing challenges to physicians in

reaching a diagnosis. The remittent/undulant fever of brucellosis was first confused with

other diseases, such as malaria and typhoid fever, and was called many synonyms pertaining

mainly to the geographic locations where the disease occurred: Mediterranean fever, Malta

fever, Gibraltar fever, and Cyprus fever (198). Over the last decade, there has been renewed

interest in this organism due to its inclusion in the potential biological weapons lists of most

authorities (8, 64, 100, 136,202;www.who.int/csr/resources/publications/Brucellosis.pdf).

Taxonomy and Genome

Brucellaceae is a family of phylogenetically closely related free-living soil organisms

composed of Brucella,Ochrobactrum, and Mycoplana spp. The Brucellaceae are part of the

order Rhizobiales, which includes other genera involved in human disease: Bartonella,

Afipia, Methylobacterium, and Roseomonas. (41, 61).

The taxonomy of Brucella spp. remains to be clarified. Studies indicate that

terrestrial Brucella spp. are homogeneous species harboring >90% interspecies homology by

DNA-DNA hybridization studies, identical 16S rRNA gene sequences, and >98% sequence

homology by comparative genomics. Because of these findings, a suggestion was made to

consider Brucella a monospecific genus and the different species as biovars ofBrucella

melitensis (73).

The average size of the genome is 2.37 × 109 Da, with a DNA G+C content of 58 to 59

mol%. Currently, the genus Brucella encompasses nine recognized species, six terrestrial

and three marine (58, 193). The six terrestrial Brucella species are B. melitensis (three

biovars) (preferred hosts are goats, sheep, and camels), B. abortus (seven biovars) (cattle,

bison, and buffalo), B. suis (five biovars) (swine and a range of wild animals),B.

canis (dogs), B. ovis (rams), and B. neotomae (desert and wood rats). The three identified

marine species, B. delphini, B. pinnipediae, and B. cetaceae, were recovered from marine

mammals (e.g., seals, whales, and dolphins) and were found to differ phenotypically from

the six terrestrial species by their patterns of substrate-mediated metabolic activity. Brucella

maris has been suggested as a name to encompass these marine isolates, but to date this

has not been accepted (38). Though preferred or predominant hosts are recognized

for Brucella spp., cross-infection of other mammalian species, including humans, may occur

(41).

Description of the Genus

Brucella spp. are facultative, intracellular, small (0.5- to 1.5-μm), gram-negative coccobacilli

that lack capsules, flagellae, endospores, or native plasmids. They are aerobic (some prefer

CO2 for their growth), do not ferment sugars, and are positive in a few oxidative metabolic

tests. Brucella spp. can grow on a wide range of culture media, and colonies appear after 24

to 48 hours of incubation as mostly smooth colonies, but rough variants can occur (4, 41).

Antigenic Components

Several antigenic determinants of Brucella, related mainly to LPS and protein antigens, have

been characterized. The LPS is the major antigen that dominates the antibody response. LPS

of rough strains is very similar to LPS of smooth strains. Based on their O side chain, smooth

strains were reported to be composed of two antigenic epitopes: A (B. abortus) and M (B.

melitensis). The smooth-strain LPS has been reported to be responsible for observed crossreactions

in both the agglutination and complement fixation tests between smooth species

of Brucella and Yersinia enterocolitica O:9, Escherichia hermannii, Escherichia

coliO:157, Salmonella enterica serovar O:30, Stenotrophomonas maltophilia, and Vibrio

cholerae O:1. Cross-reaction has been attributed to the similarities of the O-specific side

chains of the LPS molecules of these organisms (45).

The characterized protein antigens include outer and inner membrane, cytoplasmic, and

periplasmic antigens. Some are recognized by the immune system during infection and are

potentially useful in diagnostic tests (41,66, 120). For example, Omp25 is an outer

membrane structural protein that is highly conserved in all brucellae and is associated with

both LPS and peptidoglycan. In addition, some proteins, such as ribosomal proteins (e.g.,

L7/L12) and fusion proteins, demonstrate a protective effect against Brucella based on

antibody and cell-mediated responses (41, 126). These molecules may be useful in potential

vaccines.

Virulence Factors, Pathogenic Mechanisms, and Immune

Response

The incubation period is variable but generally is 1 to 4 weeks. The intracelluar location and

survival of the organism contribute to its virulence and pathogenesis. The exact

pathophysiologic aspects of infection remain to be defined (201). Briefly, once the brucellae

enter the body by various routes, they are encountered by polymorphonuclear and

mononuclear phagocytes, to which lectin facilitates a ttachment. In the process, several

factors are involved in enabling a brucella to enter a host, escape from phagocytic killing by

inhibiting the phagosome-lysosome fusion, and evade the immune system, and they aid in

its survival and propagation within macrophages and other cells. This is followed up by

brucellae being transferred through regional lymph nodes into the circulatory system and

subsequently being seeded throughout the body, with tropism for the reticuloendothelial

system, resulting in different clinical phases of disease (69). Virulence determinants include

urease to avoid stomach stress through oral passage (158) and Brucella-containing vacuoles

that enable escape from immune system recognition and provide an acidic environment to

hamper antibiotic activity. ABrucella LPS cell component (containing a poly N-formyl

perosamine O chain and a CuZn superoxide dismutase) and outer membrane protein 25

(OMP 25) were reported to help the bacteria survive within mononuclear phagocytes

(41, 55). Also, the uniqueness of Brucella LPS lies in its being a poor inducer of gamma

interferon and tumor necrosis factor alpha, both of which are essential for T-helper 1 (Th1)-

type-cell-mediated immunity for the elimination of the organism (67). The overall

inflammatory process results in a slow degradation ofBrucella cell wall components by the

polymorphonuclear leukocytes and can lead to granuloma formation, which is more often

associated with B. abortus than B. melitensis (69).

Protective immunity, though not long term, is conferred by antibodies to LPS and T-cellmediated

macrophage activation, triggered by protein antigens (41). A study showed a

significant increase in the levels of interleukin-12 and gamma interferon in patients

with Brucella infection compared to levels in controls, indicating that there is induction of

Th1-type cytokines during human brucellosis (1). The immune response

against Brucella involves antigen-specific T-cell activation, CD4+ Th lymphocytes, CD8+ T

cells, and humoral responses (67). Lymphocytes are the main stimulant of the immune

response. The Th1 response stimulates IgG2a production, which is involved mostly in

protection against intracellular pathogens through cell-mediated immunity, and is critical for

the clearance of Brucella infection. The Th2 response stimulates the production of IgG1 and

is mainly responsible for protection against extracellular pathogens through the humoral

immune response (67, 201). Recently, a study of B. abortus infection in rats showed that the

IgG2a response (indicative of a Th1 response) persisted and dominated over the IgG1

response (88). However, the exact nature of the immune response and protective factors

involved in this disease are still being investigated, and the pathogenic mechanisms of

reinfection remain unknown.

Epidemiology and Transmission

Although Brucella can be killed by pasteurization, exposure to UV light, acidity, or many

antiseptics and disinfectants, it can survive for long periods under various conditions, e.g.,

10 weeks in soil, 11 weeks in aborted fetuses, 17 weeks in bovine stool, around 3 weeks in

milk and ice cream, and several months in fresh cheese (41, 202). In terms of the total

numbers of infected cases, B. melitensis dominates the world arena (especially in the

Mediterranean and Arabian Gulf countries). However, B. abortus and B. suis supersede it in

certain geographic locations. B. canis has also been reported to cause human diseases,

while B. ovis and B.neotomae have not (41, 100, 137). Brucella spp. associated with marine

animals have been reported to cause disease in humans (28, 111, 170).

The epidemiologies of human brucellosis differ between areas of endemicity and

nonendemicity in terms of age, sex, season, and risk factors. In regions of endemicity, such

as the eastern Mediterranean basin, Middle East, the Arabian peninsula, Mexico, Central and

South America, the Balkan Peninsula, and the Indian subcontinent, the disease occurs

among the general population. In the general population, levels of infection are almost equal

among adults and children of both sexes and mostly due to ingestion of unpasteurized goat,

sheep, cow, and camel milk or its products (e.g., soft cheese) (59, 100, 137, 163, 202).

In areas where the disease is not endemic, infection is seen predominantly among adult

males, acquired occupationally by transmission through direct skin contact (e.g., through

cuts and abrasions) with infected animal parts, inhalation of aerosolized infected particles,

and accidental inoculation (e.g., sprays or aerosols inoculated into the eye, mouth, and

nose). These infections occur mostly among dairy industry professionals, veterinarians,

abattoir workers, and clinical and research microbiology staff (16, 200).

Very rare cases transmitted through blood and bone marrow transfusion, suspected sexual

intercourse, and banked human sperm have been reported (121, 138, 153, 180, 185). Also,

a few cases of neonatal brucellosis have been reported, and the isolation of Brucella from

human milk may explain it (99, 134).

Laboratory-acquired infection is an important source of transmission. Brucella has a very low

infectious dose (≤102 organisms), and personnel should adhere to strict safety precautions,

especially when handling cultures suspected of containing the organims in clinical, research,

and production laboratories (124, 200). Most cases of laboratory-acquired disease result

from mishandling and misidentification of the organism (63, 124). The frequent failure of

clinical laboratories to correctly identify isolates as Brucella species is particularly worrisome

from the perspectives of laboratory safety and potential use as a bioweapon. B. melitensis,

B. abortus, and B.suis are category B select agents (70, 136).

Clinical Categories of Human Brucellosis

The clinical categories of human brucellosis are based on arbitrary criteria. In 1956, Spink

based them on the duration of manifestations (acute, ≤2 months; subacute, 2 to 12 months;

chronic, ≥12 months) (172). Subsequently, others based them primarily on clinical

manifestations (e.g., subclinical, localized, chronic, and active, with or without localized

disease, including bacteremic and serological classifications) (102, 202). To date, no uniform

definition has been adopted.

The incubation period is variable but usually ranges between 1 and 4 weeks. The disease

onset is usually insidious, but its presentation encompasses a wide spectrum of nonspecific

clinical manifestations, such as fever, sweats, arthralgias, myalgia, fatigue, loss of appetite,

weight loss, hepatomegaly, and splenomegaly. Complications can involve many organs and

tissues with signs of focal disease. The routine hematology and biochemical profiles are

usually within normal limits, with some elevation in erythrocyte sedimentation rate and liver

function tests. Thus, to the unaware physician, the diagnosis of brucellosis can be a dilemma

and could protract for weeks and, in some complicated cases, for years

(59, 100, 106, 156, 202). Increased business and leisure travel to countries where the

disease is endemic has led to diagnostic challenges in areas where brucellosis is uncommon,

especially when the presentation is unusual (42, 100, 112, 113, 202). Overall, the mortality

is very low, but morbidity is high. Previously, brucellosis in childhood was thought to be

uncommon, but now it seems to be as prevalent as and presents in a manner similar to that

in adults in areas of endemicity (98, 100, 163).

Because of these nonspecific clinical features, human brucellosis was labeled the disease of

“mistakes.” It can be misdiagnosed and confused with other diseases, such as typhoid fever,

rheumatic fever, tuberculosis, malaria, infectious mononucleosis, endocarditis,

histoplasmosis, ankylosing spondylitis, pyelitis, cholecystitis, thrombophlebitis, chronic

fatigue syndrome, collagen vascular diseases, autoimmune diseases, and tumors

(100, 106, 135, 202).

Complications

The most commonly encountered focal complications are osteoarticular (10 to 70%) (mostly

joints), genital in both males (6 to 8%) and females (2 to 5%), neurological (3 to 5%),

cardiac (1 to 3%), pulmonary (1 to 2%), and renal (< 1%). Mortality is very low (<1%) and

is almost exclusively due to cardiac complications

(39, 42,100, 101, 102, 106, 135, 163, 202).

Osteoarticular complications occur mostly as arthritis (10 to 70%) and rarely as

osteomyelitis (<1%). The joints most frequently involved are, in descending order,

sacroiliac, knee, hip, vertebra, ankle, and multiple other joints. Generally, Brucella arthritis

can be misdiagnosed as rheumatoid arthritis, rheumatic fever, tuberculosis, and systemic

lupus erythematosis.

Neurobrucellosis (3 to 5% of cases) can affect both adults and children with diverse

presentations, including fever, headache, meningeal signs, coma, or paresis. Depression and

mental fatigue are not uncommon complaints (51, 97, 161). CSF analysis, of both adults and

children, is nonspecific and can overlap with other central nervous system diseases, such as

mycobacterial, viral, syphilitic or fungal infections, or with noninfectious diseases, such as

psychiatric problems, multiple sclerosis, and cancer (51, 97, 161). The yield

ofBrucella culture from CSF is low (5 to 30%). Therefore, the use of Brucella serology tests,

especially ELISA, on CSF specimens is essential to diagnose neurobrucellosis (9, 161). With

appropriate treatment, the prognosis is usually good for acute presentations and varies in

the setting of chronic disease.

Genital complications in males (6 to 8% of cases) are mostly orchitis or epididymoorchitis

(40, 76, 100). In females, abortion (2 to 5%) has been reported mostly in the first trimester

(103). Other rare complications reported for females include cervicitis, salpingitis, tuboovarian

abscess, and ovarian dermoid cyst (103, 183).

Relapse is considered one of the most important features of brucellosis and its complications

(19, 100, 102,171). Factors associated with relapse include the use of less effective

antibiotic therapy, a positive blood culture during the initial presentation, and ≤10 days’

duration of symptoms before initiation of treatment.

Collection, Handling, Storage, and Transport of Specimens

Specimens for the laboratory investigation of cases with brucellosis may be sent for culture,

serology, and/or molecular testing. Culture can be performed on a wide range of specimen

types, including blood (at least two sets), bone marrow, CSF, pleural and synovial fluids,

urine, abscess specimens, and tissue specimens. Adequate volumes should be secured prior

to initiation of antimicrobial therapy. Blood (serum) and, when relevant, CSF specimens are

used for serologic testing. Molecular testing, though usually for research purposes, can be

performed on blood (serum or whole blood, CSF, and bone marrow) specimens.

The guidelines for proper specimen collection, handling, transport, and processing are

generally similar to those reported for blood cultures and other specimens submitted for

bacterial culture (refer to chapter 16 in this Manual). If delay in delivery to the lab is

anticipated, specimens can be held in the refrigerator. To avoid/minimize laboratory

exposures to the pathogen, specimens from patients suspected of having the disease should

be labeled appropriately and referred to a reference laboratory, with the label specifying that

the laboratory should rule out brucellosis (44, 196).

Direct Detection

To circumvent the limitations of routine culture and serodiagnostic tests for human

brucellosis, in-house-developed conventional PCR and real-time (RT) PCR assays can directly

detect Brucella from clinical specimens. Several Brucella-specific gene targets have been

used, including BCS P31 (encodes a 31-kDa cell surface protein) and BP26 (encodes a 26-

kDa periplasmic protein), 16S rRNA, and the insertion sequence IS711. The sensitivities of

these assays are quite varied, ranging from 50% to 100%. This variation might be related to

different DNA extraction methods, detection formats, and different types of specimens

(43, 85, 110, 115, 119,122, 123, 147, 148, 189). The ribosomal 16S-23S ITS region

constitutes a suitable target in clinical specimens and formalin-fixed paraffin-embedded

archived tissue, as well as for identification of isolates from culture to the species level (85).

Molecular assays constitute a useful adjunct and have promising potential for the diagnosis

of human brucellosis in a clinical laboratory setting. Their routine incorporation in the

diagnosis of human brucellosis remains in need of further optimization, standardization, and

improvement (85, 149).

Culture

Culture is considered the gold standard in the laboratory diagnosis of brucellosis.

Conventional methods require long incubation times (6 weeks) and are generally of variable

yields, being higher among patients with acute brucellosis (40% to 90%) than in patients in

the chronic, focal, and complicated stages (5 to 20%) (10,41, 100, 199). When positive,

culture provides the definitive diagnosis. Bone marrow cultures result in a 15- to 20%-higher

yield than p eripheral blood cultures. The conventional standard m edium for the

nonautomated blood culture broth has been the biphasic Ruiz-Castaneda bottle. The growth

of the pathogen takes between 7 and 35 days to become positive, and the bottles should be

held for 6 weeks, with frequent visual inspection (every 3 days) and terminal subculture

before the specimen is discarded as negative (10, 199).

Automated continuously monitored blood culture systems such as Bactec (BD Diagnostics,

Sparks, MD) and BacTAlert (bioMerieux, Durham, NC) show higher yields than the

conventional culture method and expedite the detection of bacterial growth (majority

recovered within 1 week). There is no need to incubate bottles longer than 10 to 14 days

(9a; 199). The lysis centrifugation system showed improved and faster yields than

conventional methods in those labs that do not have automated blood culture systems (199).

However, due to the need for centrifugation and manipulation before direct plating, the

system may entail exposure and contamination hazards. Rarely, some patients with

brucellosis have a positive blood culture in the absence of positive serology (199, 202).

Recovery of Brucella from other clinical material, such as bone marrow, CSF, joint fluid,

homogenate of tissues, and bones, in addition to blood specimens, can be achieved by

inoculation of specimens into broth media (such as those used for blood cultures) in addition

to plated media (blood and CA). The latter medium is incubated at 37°C, preferably under 5

to 10% CO2, for up to 10 days prior to reporting as negative.

Identification

Clinical microbiology laboratories should report identifications of colonies suspected of

being Brucella spp. on the basis of a few morphologic, biochemical, and serologic tests. All

manipulations of Brucella cultures should be done in a biological safety cabinet. In these

setups, the colonies are generally recovered directly from inoculated clinical specimens or as

a result of subculture from broth media (e.g., biphasic Ruiz-Casta neda medium and blood

culture medium showing signs of growth) on blood (Fig. 1C) and CA. Colonies can grow on

other media as well (e.g., Mueller-Hinton agar [Fig. 1D] and MacConkey agar [can show

variable growth]). Thayer-Martin or Martin-Lewis medium can be used to

isolate Brucella spp. from contaminated specimens. Generally, colonies are visible after 24 to

48 hours of aerobic incubation or incubation with 5 to 10% CO2 at 37°C, and there is no

need to keep the plates more than 72 to 96 hours before discarding them as negative. The

colonies are 1 to 2 mm in diameter, entire, smooth, and glistening. Rough variants can occur

with B. caniscolonies. The presumptive identification of Brucella spp. from these colonies

entails demonstrating small gram-negative coccobacilli (0.5 to 0.7 μm in diameter and 0.6 to

1.5 μm in length) (Fig. 2B). Biochemical reactions show positive oxidase, catalase, and

urease tests, as well as a positive slide agglutination reaction with specific B.

abortus and/or B. melitensis antisera (4, 41). Once these tests are performed and

completed, the clinical laboratory may report the organism as presumptively Brucella spp.

Further characterization and speciation of the pathogen involves extensive testing not

routinely performed in most clinical laboratories (4,95). In the United States, LRN reference

laboratories are able to confirm and identify Brucella to the species level, and these

laboratories can provide guidance and additional pertinent information. See

alsohttp://www.bt.cdc.gov/lrn/.

When definitive identification is indicated, conventional and molecular characterizations can

be used. Conventional classification/identification to the species level of Brucella spp. can be

determined from results of certain reactions, such as dye inhibition (thionin, fuchsin,

safranin), CO2 requirement, Tbilisi phage lysis, oxidative metabolic tests (glutamic acid,

ornithine, ribose and lysine), and reaction to monospecific sera.Brucella is usually subtyped

into biovars using multilocus enzyme electrophoresis, pulsed-field gel electrophoresis,

randomly amplified polymorphic DNA analysis, enterobacterial repetitive intergenic

consensus sequence PCR, repetitive intergenic palindromic sequence PCR, amplified

fragment length polymorphism analysis, monolocus (such as omp2a and omp2b) sequence

analysis, or multilocus sequence typing (3, 86, 113,193).

Typing Systems

One of the most highly discriminatory methodologies for epidemiological subtyping of

isolates belonging to monomorphic bacterial species is MLVA (96). In Brucella, MLVA

schemes with 21 loci (MLVA-21) and MLVA-16, which use a combination of repeat markers

distributed across the Brucella genome, were able to distinguish isolates of Brucella spp. of

widespread temporal and geographical origins or of very close origins (3, 86, 193).

A Brucella MLVA database is hosted at http://mlva.u-psud.fr and contains data derived from

more than 500 animal and human Brucella isolates. Molecular subtyping methods, especially

the promising MLVA, may potentially be useful not only for epidemiological trace back

purposes or outbreaks but also for distinguishing relapses from reinfection, thereby

influencing clinical therapeutic decisions (86).

Serologic Tests

Serologic assays are the most commonly relied upon tests in the laboratory diagnosis of

brucellosis. Serological results are optimally interpreted in the context of the evolution of

antibody responses after infection with Brucella spp. IgM first appears, followed by the

appearance of IgG within 10 to 14 days. The general evolution of these and other

immunoglobulins depends on response to treatment: in recovery, a gradual and slow decline

in titers is observed, while persistent titers alert the clinician to a poor response to treatment

due to focal complications, chronic infections, or relapse (18, 62, 140). Persistence, (i.e.,

detection of antibodies, mostly IgG and some IgM, for a very prolonged time [months and

sometime years]) is observed in 15 to 20% of asymptomatic patients who have undergone

treatment and cure. The explanation for this remains elusive (100).

Several antigens are used for serologic diagnostic assays, generally obtained from B.

melitensis and B. abortus. Whole-cell preparations are the antigens used in the agglutination

and the indirect fluorescent-antibody (IFA) tests, while sonic extracts, purified LPS or protein

extracts of Brucella, are used mainly in ELISAs (4, 7, 10, 12, 13, 87). Detection of antibodies

against infections due to B. canis and B. ovis require using major outer membrane protein

antigens because these strains exist in a rough colony form and do not share cross-reacting

antigens with the other Brucella spp. (113). Since there is no standardized reference

antigen, it is important to note that the source of the antigen, commercial or otherwise, can

influence the test results (14).

A wide range of in-house serologic tests and formats have been used for investigating

patients with brucellosis (Table 4). These include direct agglutination tests in tubes, e.g., the

serum agglutination test (SAT), and on slides, e.g., the rose bengal test, indirect Coombs

test, Brucellacapt tests, IFA test, and ELISA for detection of immunoglobulin classes and

subclasses (6, 11, 13, 14, 15, 68, 154). Generally, agglutination-based tests cannot

differentiate the types of antibodies involved, while the enzyme immunoassay (EIA) and IFA

test can. Commercial EIAs detecting Brucella IgG and IgM with a high degree of sensitivity

and specificity have been available for a number of years (17) and are considered an

excellent method for screening sera forBrucella antibodies (13, 16).



Anti-Brucella Therapy

Appropriate antimicrobial therapy for treatment of human brucellosis reduces morbidity,

prevents complications, and minimizes relapses. Several anti-Brucella agents have been

used (e.g., doxycycline, rifampin, trimethoprim-sulfamethoxazole, streptomycin-gentamicin,

some quinolones, and cephalosporins) with various rates of success. Currently, the most

effective treatment regimen and optimal duration of treatment remain unclear

(20, 59, 84, 98, 168).

Fewer relapses with combined regimens than with monotherapy have been reported. For

adults with uncomplicated infection, the WHO recommends oral doxycycline and rifampin for

6 to 8 weeks. Triple regimens using doxycycline, rifampin, and an aminoglycoside for 2 to 3

months are recommended for patients with endocarditis and neurobrucellosis.

Treatment regimens with fluoroquinolones and broad-spectrum cephalosporins have been

used. Although these agents have good MICs in vitro against Brucella spp., patients treated

with these regimens have higher rates of relapses than patients on the standard regimen.

The use of fluoroquinolones in combination with rifampin for the treatment of bacteremia and

complicated brucellosis has yielded varied results (2, 168).

A recent systematic review and meta-analysis study covering 30 trials and 77 treatment

arms showed that among patients with bacteraemia and complicated brucellosis, higher

failure and relapse rates and shorter treatment durations (less than 6 weeks) were observed

with monotherapy than with multidrug therapy. The preferred treatment should be with dual

or triple regimens, including an aminoglycoside (168).

The question about postexposure prophylaxis (doxycycline and rifampin therapy for 3 to 6

weeks) after a high-risk exposure in the lab remains debatable. Guidelines for postexposure

management are empirical (32). Upon possible exposure, however, recommendations were

made to take a baseline blood sample, monitor for symptoms weekly for 6 months, and

perform serological surveillance at 0, 2, 4, 6, and 24 weeks (33).

Prevention

Vaccines have been successful in the control of livestock infections, which can subsequently

reduce infections in humans. Most veterinary vaccines focus on live, attenuated B.

abortus (strain S19) and a more stable rough mutant of B. abortus (strain RB51) for

cows, B. melitensis (strain Rev-1) for sheep and goats, and B. suis 2 for swine. However,

developed vaccines have had limited efficacy in humans and have been associated with

serious medical reactions. Heating of dairy products and related foods has also been effective

in preventing disease transmission. The most cost-effective approach to control and prevent

brucellosis relies on raising public awareness about the disease and greater cooperation

between human and animal health sectors (204).

Evaluation, Interpretation, and Reporting of Results

Interpretation of serologic test results in relation to exposure, diagnosis, and prognosis of

the disease necessitates an accurate assessment of the clinical history and current status of

patients and understanding the usefulness and pitfalls of the laboratory tests (7, 202).

Positive cutoff titers in the Brucella agglutination test for diagnosis have generally been

considered to be ≥160 in symptomatic patients. However, much lower titers with the SAT

have been reported for patients with active disease (100). Moreover, one has to be careful

when negative serology is encountered when brucellosis is suspected, since this could be

due, for example, to infection with B. canis, which can be missed by serologic assays

using B. abortus or B. melitensis antigen. In addition, this could be due to very early disease

presentation, and thus repeat testing after 1 to 2 weeks is warranted (18, 85).

In acute brucellosis, elevation in Brucella-specific IgG, IgM, IgA, IgE, IgG1, and IgG3 is

shown, while in those patients with chronic brucellosis, elevations in IgG, IgA, IgE, IgG1, and

IgG4 are usually seen (6, 10, 15, 62). Monitoring the treatment response requires a

sequential follow-up for patients with serologic titers. A decline indicates good prognosis,

persistently high titers necessitate continuous monitoring, and a resurgence in antibody

titers most likely indicates relapse or reinfection. Slide and TA titers fall faster than with the

EIA (18,21, 105, 140). Relapse has also been diagnosed by a detection of a resurgence

in Brucella-specific IgG and IgA antibodies, not IgM (18, 62, 140). Markers for differentiating

active from inactive disease are being sought. For example, anti-Brucella cytoplasmic or

periplasmic protein antibodies, as determined by ELISA, increased only in patients with

active brucellosis and were a better predictor of cure than antilipopolysaccharide antibodies

(21,66, 151). Also, some interleukins show a decrease posttherapy (104).

Though serologic tests are currently of high importance for the investigation of patients with

brucellosis, several limitations can be encountered, mainly lack of standardized antigen

preparations and assay methodologies, as well as the detection of sustained high antibody

titers in some patients, despite treatment and cure (7, 18, 62, 68, 140). False-positive

serologic results are rare. However, two cases were recently reported in the United States, a

finding which led to initiating not only unnecessary treatment but also a wide range of public

health investigations (34).

Based on the above, laboratories should use a combination of two agglutination tests,

namely, the SAT and indirect Coombs test, the SAT and Brucellacapt, or ELISAs for IgG and

IgM. In doing so, one would be able to detect antibodies in different stages of the disease,

since in the acute stage any test can be positive, while in chronic, complicated, or focal

disease cases, the SAT can be negative while the Coombs test, Brucellacapt, and ELISA using

IgG can be positive (Table 5). Again, one should keep in mind that any serologic test findings

need to be interpreted in the context of the patient’s clinical history.


No comments:

Post a Comment