Clostridium

Phylogenetically, the genus Clostridium is heterogeneous, with many species intermixed with

other spore-forming and non-spore-forming genera. Traditionally, the different species have

been defined based on morphological, ultrastructural, and physiological features. During the

past 2 decades, analyses of 16S rRNA gene sequences indicated that the “clostridia” could be

divided into 19 clusters (57). Cluster I forms the basis of the genus Clostridium and is

analogous to group I proposed by Johnson and Francis over 30 years ago (98). The type

species, Clostridium butyricum, and most of the clinically relevant Clostridium species cluster

within rRNA homology group I (reviewed in reference 177). The heterogeneous non-group I

clostridia require reclassification; however, 16S rRNA gene sequences may not be adequate

alone in distinguishing genera, and it is necessary to find genetic and phenotypic characters

that enable rapid discrimination among genera within this group.

Two new species clustering within the C. coccoides rRNA group, C. hathewayi (179) and C.

bolteae (173), were described from human feces. Phenotypically, C. clostridioforme is a

relatively heterogeneous anaerobic species. Sequencing analyses of 16S rRNA genes from

107 strains that were previously identified phenotypically as C. clostridioforme in various

clinical laboratories revealed that C. clostridioforme in fact represents three distinct

species: C. bolteae, C. clostridioforme, and C. hathewayi (72). C. bartlettii is another

new Clostridium species described from human feces (174); the clinical significance of this

organism remains unknown. “C. neonatale”was proposed as a novel species recovered from

bacteremia in patients with necrotizing enterocolitis (NEC) (9). Anaerotruncus colihominis is

a new genus and species within the C. leptum rRNA cluster of organisms originally described

from human feces (121) and subsequently found in patients with bacteremia (119). Though

it was originally described as a non-spore-forming organism, further studies have revealed

that sporulation occurs under some conditions (119) and should therefore be considered

in Clostridiumidentification schemes. On the basis of biochemical properties, phylogenetic

position, DNA G+C content, and DNA-DNA hybridization, the unification of Clostridium

orbiscindens and Eubacterium plautii into the new genusFlavonifractor plautii has been

proposed (42).

DESCRIPTION OF THE GENUS Back to top

Clostridia belong to the phylum Firmicutes and comprise a heterogeneous (paraphyletic)

group consisting of at least 12 lineages. Clostridia have a wide range of G+C contents, from

22 to 55 mol%, while the toxigenic species have a much narrower range of G+C contents, 24

to 29 mol% (177). Morphological and phenotypic properties that have traditionally been used

to define the genus include (i) the formation of endospores, (ii) anaerobic energy

metabolism, (iii) an inability to reduce sulfate to sulfide, and (iv) a gram-positive cell wall

structure.

Vegetative cells of Clostridium species are pleomorphic, rod shaped, and arranged in pairs or

short chains; the cells have rounded or sometimes pointed ends (90). Rods may join to form

tight coils or spiral configurations in species such as C. cocleatum and C. spiroforme.

Clostridia stain gram positive in early stages of growth, although some species, such as C.

clostridioforme, C. hathewayi, C. innocuum, and C. ramosum, may appear gram negative.

Several species (e.g., C. tetani) appear gram negative by the time that spores have formed.

Endospores are often wider than the vegetative organisms, imparting characteristic spindle

shapes to clostridia. Most strains are motile by means of peritrichous flagella. Nonmotile

species include C. perfringens, C. ramosum, and C. innocuum (90).

Clostridium species are metabolically diverse. As currently designated (57), most species are

chemoorganotrophic; some species may be chemoautotrophic and chemolithotrophic. They

can be saccharolytic, proteolytic, neither, or both; they do not carry out dissimilatory sulfate

reduction. They usually produce mixtures of organic acids and alcohols from carbohydrates,

proteins and peptides, or purines and pyrimidines.

Most species are obligately anaerobic, although the tolerance to oxygen varies widely; some

species (e.g., C. tertium) grow but do not sporulate in the presence of air, and a few

aerotolerant species, such as C. carnis, C. histolyticum, and occasional strains of C.

perfringens, give scant growth on solid media incubated under 5 to 10% CO2. Aerotolerant

clostridia and certain Bacillus species may be distinguished by several means: (i) clostridia

usually form spores only under anaerobic conditions, (ii) they grow better anaerobically than

in air, (iii) they usually do not produce catalase, and (iv) they have straight-chain, saturated,

and monounsaturated cellular fatty acid (CFA) compositions, whereas Bacillus species have

branched-chained CFAs. AlthoughClostridium species are usually catalase and superoxide

dismutase negative, trace amounts of these enzyme activities may be detected in some

strains, such as C. perfringens. In addition, clostridia lack a cytochrome system and are thus

oxidase negative. Clostridia often occur in nature and in infections as consortia of mixed

species, wherein aerobic and facultative organisms utilize oxygen, provide nutrients or other

factors, and create an environment favorable for clostridial growth.

Clostridia produce more kinds of protein toxins than any other bacterial genus, and more

than 25 toxins lethal to mice have been identified (reviewed in reference 169). At least 15

species of cluster I Clostridium produce protein toxins, and new toxins and virulence proteins

have been discovered through traditional isolation techniques and genomic analyses

(36, 164). These proteins include neurotoxins, enterotoxins, cytotoxins, collagenases,

permeases, necrotizing toxins, lipases, lecithinases, hemolysins, proteinases, hyaluronidases,

DNases, ADP-ribosyltransferases, neuraminidases, and some others that are simply known

as lethal toxins. Botulinum neurotoxin and tetanus neurotoxin (BoNT and TeNT) are the most

potent toxins known, with lethal doses of 0.2 to 10 ng per kg of body weight for various

animals, including humans (22). Epsilon toxin is a 33-kDa protein produced by C.

perfringens types B and D strains, and in animals it causes edema and hemorrhage in the

brain, heart, spinal cord, and kidneys. It is among the most lethal of clostridial toxins and is

considered a potential bioterrorism agent (22, 167).

Recently, some genomic sequences of pathogenic clostridia have become available

(36, 164), which should facilitate a comprehensive approach for understanding virulence

factors involved in clostridial pathogenesis.

EPIDEMIOLOGY AND TRANSMISSION Back to top

Clostridium species are widespread in nature due to their ability to form resistant

endospores. They are commonly found in soil, feces, sewage, and marine sediments. The

ecology of C. perfringens in soil is greatly influenced by the degree and duration of animal

husbandry (reviewed in reference 171), and this has relevance to the incidence of gas

gangrene caused by contamination of war wounds with soil. For example, the incidence of

clostridial gas gangrene was higher in agricultural lands in Europe than in the Sahara Desert

of Africa (171). Similarly, the incidences of tetanus and foodborne botulism are also clearly

related to the presence of clostridial spores in soil, water, and many foods (171). Outbreaks

of hospital-acquired enteric C. difficile infections are often traceable to environmental

sources and other typical background factors for nosocomial infection (144). Clostridia are

present in large numbers in the indigenous microbiota of the intestinal tracts of humans and

animals, in the female genital tract, and in the oral mucosa as well.

CLINICAL SIGNIFICANCE Back to top

Although exogenous clostridial infections or intoxications, such as tetanus, foodborne

botulism, and gas gangrene, have been feared for centuries, severe cases of hospitalacquired

and community-acquired C. difficile colitis have recently emerged. Endogenous

clostridia, in association with non-spore-forming anaerobes and facultative or aerobic

organisms, also cause severe infections in diabetic patients and in patients in whom the

mucosal integrity of the bowel or respiratory system has been compromised. Head and neck

infections, brain abscesses, sinusitis, otitis, aspiration pneumonia, lung abscesses, pleural

empyemas, cholecystitis, intra-abdominal infections, gynecologic and obstetric infections,

soft tissue infections, myonecrosis, and septic arthritis and bone infections all may involve

clostridia (82). Common predisposing factors are surgical procedures, trauma, vascular

stasis, bowel obstruction, malignancy, immunosuppressive agents, diabetes mellitus, prior

aerobic infection, and use of antimicrobial agents with poor activity against clostridia (see

the section on C. difficile below).

Clostridial Bacteremia

Clostridium species are important causes of bloodstream infections (118, 122, 166). C.

septicum is isolated only rarely from the feces of healthy individuals but may be found in the

appendixes of normal individuals. Over 50% of patients whose blood cultures are positive for

this organism have some gastrointestinal anomaly, such as diverticular disease, or an

underlying malignancy, such as carcinoma of the colon. Another clinically important

association has been observed between C. septicum bacteremia and neutropenia of any

origin and, more specifically, neutropenic enterocolitis involving the terminal ileum or cecum

(112). Patients with diabetes mellitus, severe atherosclerotic cardiovascular disease, or

anaerobic myonecrosis (gas gangrene) may also develop C. septicum bacteremia (81). The

clinical importance of recognizing C. septicumbacteremia and starting appropriate treatment

immediately cannot be overemphasized. Patients with this condition are usually gravely ill

and may have metastatic spread to distant anatomic sites, resulting in spontaneous

myonecrosis. Mortality rates are very high. C. septicum has also been recovered from

cirrhotic patients with bacteremia, as have C. perfringens, C. bifermentans, and other

clostridia (50). Some of these patients have demonstrated septic shock.

Another clostridial species of importance in patients with serious underlying disease, such as

malignancy and acute pancreatitis, is C. tertium. This organism, as well as C.

septicum and C. perfringens, may be seen among the bacteria in the blood of such patients,

with or without neutropenic enterocolitis (124). C. tertium may present special problems in

terms of both identification and treatment. This organism may appear to be gram negative,

and it is aerotolerant and resistant to metronidazole, clindamycin, and

cephalosporins. Clostridium sordellii and C. perfringens have been associated with toxic

shock syndrome and abortion (7, 55).

Studies of anaerobic bacteremia by Woo et al. (202) and Simmon et al. (166) identified

clostridia based upon sequencing of genes encoding 16S rRNA. C. perfringens and C.

tertium were the two most frequently identified species, causing up to 79% and 5%,

respectively, of clostridial bacteremias. The mortality rate of clinically relevant clostridial

bacteremia ranged from 29 to 35%, and risk factors for mortality (200) were liver disease

and older age. The C. clostridioforme group (including C. clostridioforme, C.

hathewayi, and C. bolteae) has also caused bacteremia (72, 203).

Enteric Infections

Food Poisoning

C. perfringens is one of the most common bacterial causes of foodborne illness in the United

States and Canada (28, 95, 147), and virtually all cases have been due to type A strains

(28, 167). In C. perfringens type A foodborne disease, the food vehicle is typically

improperly cooked meat or a meat product, such as gravy, that has cooled slowly after being

cooked or may have been inadequately reheated. Spores surviving the initial cooking

germinate, and vegetative cells proliferate during slow cooling or insufficient reheating.

Illness results from the ingestion of food containing about 108 or more viable vegetative

cells, which sporulate in the alkaline environment of the small intestine, producing an

enterotoxin (C. perfringens enterotoxin [CPE]) in the process. Diarrhea develops within 7 to

30 h of ingestion of such food and is generally mild and self-limiting (167); however, in the

very young, the elderly, and the immunocompromised, symptoms are more severe,

occasionally resulting in death (29). Enterotoxin-producing C. perfringens has been

implicated as an etiologic agent of persistent diarrhea in elderly patients in nursing homes

and tertiary-care institutions and has been considered to play a role in antibiotic-associated

diarrhea (AAD) without pseudomembranous colitis.

C. perfringens strains associated with food poisoning produce the CPE, which generally acts

by forming pores in membranes of host cells (167). C. perfringens strains isolated from

nonfoodborne diseases, such as AAD and sporadic diarrhea, carry cpe on a plasmid (41, 73),

which may be transmitted to other strains.

Enteritis Necroticans (Pigbel and Darmbrand), Necrotizing Enteritis,

and NEC

Enteritis necroticans is caused by alpha-toxin- and beta-toxin-producing strains of C.

perfringens type C. Beta toxin is located on a plasmid (73) and is responsible mainly for

pathogenesis (157, 167, 175). Enteritis necroticans is a life-threatening infection causing

ischemic necrosis of the jejunum. In Papua New Guinea during the 1960s, it was found to be

the most frequent cause of death in children; it has been associated with pig feasts and

occurs both sporadically and in outbreaks. Immunization against the beta toxin decreased

the incidence of the disease in New Guinea (120). Enteritis necroticans has also been

recognized in the United States, the United Kingdom, Germany, and other developed

nations, especially involving adults who are malnourished or who have diabetes, alcoholic

liver disease (138, 151), or neutropenia (125). It should be noted that NEC, a disease

resembling enteritis necroticans but associated with C. perfringens type A, has been found in

North America in previously healthy adults (172).

NEC is a serious gastrointestinal disease affecting low-birth-weight (premature) infants

hospitalized in neonatal intensive care units. The etiology and pathogenesis of this disease

have remained an enigma for over 4 decades (146). Pathological similarities between NEC

and enteritis necroticans include their patterns of bowel necrosis and degrees of

inflammation (107). Both diseases may manifest intestinal gas cysts (107). The sources of

the gas, which contains hydrogen, methane, and carbon dioxide, are probably the

fermentative activities of intestinal bacteria, including clostridia. Epidemiological data support

an important role for C. perfringens or other gas-producing microorganisms (e.g., “C.

neonatale,” certain other clostridia, or Klebsiellaspp.) in the pathogenesis of NEC.

Clostridium difficile Infection (CDI)

Prevalence of CDIs

C. difficile, the major cause of antibiotic-associated pseudomembranous colitis, is also the

most frequently identified cause of hospital-acquired diarrhea and is responsible for more

than 250,000 cases of diarrheal disease per year in the United States, with a cost exceeding

$1 billion (114). C. difficile has been isolated from feces of 3 to 5% of the healthy

population, 30% of healthy neonates, and 20 to 30% of sedentary patients (185). McFarland

et al. (140) reported that 21% of 399 patients with negative cultures on admission to a

hospital with a high prevalence of C. difficile-associated disease (CDAD) acquired C.

difficile during hospitalization. Of these patients, 63% remained asymptomatic, while 37%

developed diarrhea.

Role of the PaLoc in CDI

Only strains that carry the pathogenicity locus (PaLoc) (32) possess the genetic information

for the C. difficileenterotoxin, TcdA, and the cytotoxin, TcdB (tcdA and tcdB, respectively).

Only strains producing TcdA and/or TcdB cause CDI. A limited number of cases of

pseudomembranous colitis are caused by TcdA2 TcdB1 strains (102, 130, 137, 153) or strains

that produce only TcdA (101, 130, 192). Recent results with a hamster model indicate that

TcdB may be more important for disease induction than TcdA (131). Strains that carry only

the genes for the binary toxin CdtA/B do not cause CDI or pseudomembranous colitis.

TcdA and TcdB, together with toxins from Clostridium sordellii, C. perfringens, and C.

novyi (191), belong to the family of large clostridial cytotoxins (LCC). The molecular masses

of TcdA and TcdB are 308 kDa and 270 kDa, respectively. Such LCC toxins glycosylate small

GTP-binding signal proteins of the Ras family, leading to a breakdown of the cell’s

cytoskeleton and thus causing apoptosis (104). Both TcdA and TcdB are auto-activated once

inside the cell. However, in contrast to A-B toxins of the diphtheria type, they are single

chained.

Two accessory proteins, TcdR and TcdC, of the PaLoc (32) are regulatory elements that

control toxin expression (93, 137). Recently, the tcdC gene has gained diagnostic attention

since it is shortened in endemic hypervirulent ribotype 027-NAP1 isolates (herein called

ribotype 027 isolates) (194). Such strains seem to overproduce toxin but surely lead to more

severe causes of CDI (129, 139).

Risk Factors and Course of CDI

Acquisition of C. difficile alone does not induce CDI. Several other risk factors, like age,

hospitalization, severe bowel surgery, treatment with proton pump inhibitors plus a change

in colonization resistance due to such treatments plus colonization with a TcdA/TcdBproducing

C. difficile strain, are necessary for development of CDI.

The spectrum of symptoms ranges from mild self-limiting diarrhea to bloody-slimy diarrhea

(called C. difficile-associated diarrhea) to the development of full-scale pseudomembranous

colitis (24). The onset of CDI may begin immediately following antibiotic treatment or as long

as 4 to 6 weeks after the course of antibiotics has been finished. Antibiotics most commonly

associated with CDI are clindamycin, expanded- and broad-spectrum cephalosporins, and

fluoroquinolones (189).

Bloody, mucus-filled stools generally indicate greater destruction of the colonic mucosa and

hence are associated with more severe disease. Clinical diagnosis may be established by

rectoscopy and the identification of pseudomembranes on the colonic mucosa. Severe cases

are typically observed among the elderly, in nursing home residents, and in

immunocompromised patients (24, 129, 139).

Epidemic Outbreaks

Hypervirulent strains (such as those of ribotype 027) have caused endemic outbreaks in

Canada, the United States, Europe, and even worldwide (113, 129, 139). These outbreaks

have occurred among younger age groups, in patients with no underlying diseases, and even

among outpatients. These cases are associated with megacolon and rupture of the large

bowel and are often lethal. There is evidence that use of fluoroquinolones may be an

essential trigger in the onset of such endemic outbreaks (176).

Particularly vexing complications of CDIs are relapses after antibiotic treatment caused by

the initial causative strain or by reinfection with a second C. difficile strain (99). Published

data report relapse rates of 20 to 50%. Even the first relapse should be treated with a

vancomycin step therapy (see below). Other forms of treatment, including the use of the

probiotic Saccharomyces boulardii (78) and stool transplants, have been suggested, but

results are not yet definitive. Eradication of C. difficile from the hospital environment is a

worthy objective but a difficult task for infection control practitioners. Commonly used

disinfectants are not sporicidal.

Other Etiologies of Antibiotic-Associated Diarrhea

C. difficile is responsible for ≤20% of cases of AAD (23, 198). Enterotoxin-producing C.

perfringens type A has been isolated from AAD patients who are negative for C. difficile and

who have no other apparent cause of the disease. Coinfection with C. difficile and

enterotoxigenic C. perfringens type A has also been reported for AAD patients (1). Though

the incidence of C. perfringens-associated AAD has been estimated to be 5 to 20% (15),

additional epidemiological studies are needed to accurately determine the role of this

organism in AAD.

Histotoxic Clostridial Skin and Soft Tissue Infections

Histotoxic clostridial species such as C. perfringens, C. histolyticum, C. septicum, C.

novyi, and C. sordellii cause aggressive necrotizing infections of the skin and soft tissues

attributable, in part, to the elaboration of bacterial proteases, phospholipases, and cytotoxins

(40). Necrotizing clostridial soft tissue infections (gas gangrene) are rapidly progressive and

characterized by marked tissue destruction, gas in the tissues, shock, and frequently death

(180).

Clostridial Myonecrosis

Traumatic Gas Gangrene due to C. perfringens

C. perfringens myonecrosis (gas gangrene) is one of the most fulminant gram-positive

infections of humans. Predisposing conditions include crush-type injury, laceration of largeor

medium-sized arteries, and open fractures of long bones which are contaminated with soil

containing the bacterial spores. Gas gangrene of the abdominal wall and flanks occurs after

penetrating injuries, such as knife or gunshot wounds, sufficient to compromise intestinal

integrity, with resultant leakage of bowel contents into the soft tissues. In the last few years,

cutaneous gas gangrene caused by C. perfringens, C. novyi type A, and C. sordellii have

been described in the United States and northern Europe among drug abusers injecting

“black-tar heroin” subcutaneously (18, 33, 45, 46, 106).

Clostridial gas gangrene is characterized by the sudden onset of excruciating pain at the

infection site (133) and rapid development of a foul-smelling wound containing a thin

serosanguinous discharge and gas bubbles. Brawny edema and induration develop and give

way to cutaneous blisters containing bluish-to-maroon fluid. Later, such tissue may become

liquefied and slough. The margin between healthy and necrotic tissue often advances several

inches per hour despite appropriate antibiotic therapy (133), and radical amputation remains

the single best life-saving treatment. Shock and organ failure frequently accompany gas

gangrene, and when patients become bacteremic, the mortality exceeds 50%.

Diagnosis is not difficult because the infection (i) always begins at the site of significant

trauma, (ii) is associated with gas in the tissue, and (iii) is rapidly progressive. A Gram stain

of drainage or a tissue biopsy specimen is usually definitive, demonstrating large grampositive

rods and an absence of inflammatory cells. Using experimental models, Bryant and

colleagues have recently demonstrated that the severe pain, rapid progression, marked

tissue destruction, and absence of neutrophils in C. perfringens gas gangrene is caused by

alpha-toxin-induced occlusion of blood vessels by platelets and neutrophils (38, 39).

Spontaneous, Nontraumatic Gas Gangrene due to C. septicum

The first symptom of spontaneous C. septicum gas gangrene may be confusion, followed by

the abrupt onset of excruciating pain and rapid progression of tissue destruction, with

demonstrable gas in the tissue (100,133, 171, 181). Swelling increases, and bullae appear

filled with clear, cloudy, hemorrhagic, or purplish fluid. The surrounding skin has a purple

hue, perhaps reflecting vascular compromise resulting from bacterial toxins diffusing into

surrounding tissues (181). The mortality of patients with spontaneous gangrene ranges from

67 to 100%, with the majority of deaths occurring within 24 h of onset. Predisposing host

factors include colonic carcinoma, diverticulitis, gastrointestinal surgery, leukemia,

lymphoproliferative disorders, cancer chemotherapy, radiation therapy, and, more recently,

AIDS (100, 181). Cyclic, congenital, or acquired neutropenia is also strongly associated with

an increased incidence of spontaneous gas gangrene due to C. septicum, and in such cases,

NEC, cecitis, or distal ileitis is commonly found. These gastrointestinal pathologies permit

bacterial access to the bloodstream; consequently, the aerotolerant C. septicum can

proliferate in normal tissues (171). Patients surviving bacteremia or spontaneous gangrene

due to C. septicum should have aggressive diagnostic studies to rule out gastrointestinal

pathology.

Gynecologic Infections due to C. sordellii

Gas gangrene of the uterus, especially that due to C. sordellii, has historically occurred as a

consequence of illegal or self-induced abortions but in modern times also follows

spontaneous abortion, normal vaginal delivery, and cesarean section (reviewed in

reference 7). Recently, C. sordellii has also been implicated in medically induced abortions

(7). Young, previously healthy women with fatal postpartum C. sordellii infections present

with a unique clinical picture of little or no fever, a lack of a purulent discharge, refractory

hypotension, extensive peripheral edema and effusions, hemoconcentration, and a markedly

elevated white blood cell count (7). Death in these cases ensues rapidly, and the infection is

almost uniformly fatal (7).

Other Clostridial Skin and Soft Tissue Infections

Crepitant cellulitis, also called anaerobic cellulitis, is seen principally in diabetic patients and

characteristically involves subcutaneous tissues or retroperitoneal tissues and can progress

to fulminant systemic disease; the muscle and fascia are not involved.

Cases of C. histolyticum infection with cellulitis, abscess formation, or endocarditis have also

been documented in injecting drug users (16). C. sordellii was responsible for

endophthalmitis after suture removal after a corneal transplant (205). C.

perfringens endophthalmitis due to penetrating injuries is a fulminant infection (92).

Exotoxins of the Histotoxic Clostridia

Our current understanding of the potent toxins produced by these clostridia is based upon

studies done between World Wars I and II, when gas gangrene was a major complication of

battlefield injuries. Investigators of this period designated the major lethal toxins of these

bacteria with Greek letters, with the letter “α” always used to designate the most potent or

most significant lethal factor. A marvelous review of these data can be found in the

monograph by Smith (171). Over the ensuing decades, modern technology has provided a

greater understanding of the mechanisms of action of some of these factors.

Major Extracellular Toxins of C. perfringens

The major C. perfringens extracellular toxins implicated in gas gangrene are alpha toxin and

theta toxin. Alpha toxin is a lethal lecithinase that has both phospholipase C and

sphingomyelinase activities and has been implicated as the major virulence factor based

upon the observation that immunization of mice with purified recombinant protein consisting

of the C-terminal alpha-toxin domain (amino acids 247 to 370) provided protection against

lethal challenge with C. perfringens (199). In addition, intravascular activation of platelets by

alpha toxin leads to platelet aggregation (38, 184) and formation of occlusive thrombi that

completely and irreversibly occlude capillaries, venules, and arterioles (38, 39). Without

adequate tissue perfusion, the anaerobic niche is extended and rapid destruction of viable

tissue, so characteristic of clostridial gas gangrene, ensues.

Theta toxin from C. perfringens (also known as perfringolysin) is a member of the thiolactivated

cytolysin family, now termed cholesterol-dependent cytolysins, that includes

streptolysin O from group A streptococci, pneumolysin from Streptococcus pneumoniae, and

several others. Upon contact with cholesterol in the host’s cell membranes, theta-toxin

monomers oligomerize and insert into the membrane, forming a pore and resulting in cell

lysis (161). Theta toxin contributes to the pathogenesis of gas gangrene, likely by its ability

to modulate the inflammatory response to infection (37, 182).

Major Extracellular Toxins of C. septicum

C. septicum produces four main toxins, alpha toxin (α, lethal, hemolytic, necrotizing

activity), beta toxin (β, DNase), gamma toxin (γ, hyaluronidase), and delta toxin (,

septicolysin, an oxygen-labile hemolysin), as well as a protease and a neuraminidase (171).

Unlike the alpha toxin from C. perfringens, the C. septicum alpha toxin does not possess

phospholipase activity. Active immunization against alpha toxin significantly protects against

challenge with viable C. septicum (17).

Major Extracellular Toxins of C. sordellii

Pathogenic strains of C. sordellii produce up to seven identified exotoxins. Of these, lethal

toxin (LT) and hemorrhagic toxin (HT) are regarded as the major virulence factors. LT and

HT are members of the LCC family, all having molecular masses between 250 and 308 kDa.

Other members include the C. difficile toxins A and B and C. novyi alpha toxin. All LCCs

possess remarkable amino acid similarity, with identities ranging between 26 and 76%. LT

and C. difficile toxin B have the highest homology, with amino acid sequences being 76%

identical and 90% homologous to one another. All LCCs possess glycosyltransferase activity

and modify signaling molecules that control the cell cycle, apoptosis, gene transcription, and

the structural functions of actin, such as cell morphology, migration, and polarity. Once

modified, these proteins become inoperative. Modification of actin cytoskeletal assembly and

organization presumably leads to the massive capillary leakage characteristic of C.

sordellii infection. The C. sordellii neuraminidase has been shown to contribute to the

leukemoid reaction, in part, by enhancing the proliferation of granulocyte progenitor cells

(6). Other exotoxins include an oxygen-labile hemolysin, DNase, collagenase, and

lysolecithinase; however, their roles in pathogenesis have not been extensively investigated.

Botulism

The Organism and Its Toxin

C. botulinum is the cause of the rare but frequently fatal illness known as botulism and which

is characterized by sudden flaccid paralysis. Spores of C. botulinum are widely distributed in

soil and aquatic habitats. C. botulinum, along with unique strains of C. butyricum, C.

baratii, and C. argentinense, can produce BoNT, the most lethal poison known. The

intravenous lethal dose for BoNT has been estimated as 0.1 to 0.5 ng per kg of body weight,

and BoNT is among the most potent protein toxins by oral ingestion, with an estimated oral

lethal dose of 0.2 to 1 μg per kg (13). There are seven antigenic serotypes of BoNT (A

through G) (115), which serve as useful clinical and epidemiological markers (132). Toxin

serotypes A, B, and E of C. botulinum are the principal causes of botulism in humans (88).

Neurotoxigenic strains of C. butyricum (70) and C. baratii (19, 86,149) that produce type E

and F neurotoxins, respectively, have been implicated mainly in infant botulism. Type E

botulinal-toxin-producing C. butyricum strains were confirmed by sequencing of the16S rRNA

gene (49), leading to the conclusion that neurotoxigenic C. butyricum must be regarded as

an emergent foodborne pathogen. C. argentinense, which produces type G neurotoxin (88),

has been isolated from soil in Argentina. Its reported isolation from autopsy materials from

five individuals who died suddenly has not been substantiated, and C. argentinense has not

been clearly implicated in botulism. C. botulinum types C and D are associated primarily with

botulism in birds and mammals (97, 168). Strains of C. botulinum that produce more than

one serotype of BoNTs, generally with one serotype being formed in much higher levels,

have been isolated from the environment and human and animal botulism cases (88, 96).

The BoNTs are coexpressed with nontoxic proteins of toxin gene clusters (31), and evidence

suggests that the complexes are much more stable than the labile BoNTs in the

gastrointestinal tract. The genes for BoNT complex formation are associated with unstable

genetic elements in certain serotypes, enabling toxin gene transfer to nontoxigenic clostridial

species that are closely related to C. botulinum, such as C. sporogenes and C.

subterminale (64).

There are four naturally occurring types of botulism: (i) classical foodborne botulism, an

intoxication caused by the ingestion of preformed botulinal toxin in contaminated food; (ii)

wound botulism, which results from elaboration of botulinal toxin in vivo after the growth

of C. botulinum in an infected wound; (iii) infant botulism, in which botulinal toxin is

elaborated in vivo in the gastrointestinal tract of an infant colonized with C.botulinum; and

(iv) botulism due to intestinal colonization in children and adults (12, 88). Intestinal

colonization in adults has been associated with surgery and administration of antibiotics

(88). C. botulinumhas been isolated from patients colonized with C. difficile (70), with viral

infections (69), or with Crohn’s disease (83). Recently, an international outbreak of botulism

caused by commercial carrot juice was reported by Sheth et al. (162).

Regardless of the category of botulism, the toxin enters the bloodstream at a peripheral site

(e.g., gut, wound, or lung) and is transferred to the neuromuscular junctions of motor

neurons, where it binds irreversibly to the presynaptic membranes. The site of action of all

serotypes of BoNT is the presynaptic terminal of motor neurons (51, 110, 117, 158).

Elucidation of the three-dimensional structures of botulinum and tetanus toxins and their

constituent domains has provided considerable insights into their mechanisms of action

(116, 117, 158, 186). BoNT penetrates the plasma membrane by receptor-mediated

endocytosis, and the light chain of 50 kDa (the catalytic domain) is internalized into the

nerve cell through a protein channel (117, 158). Once internalized, BoNT specifically cleaves

proteins involved in vesicle trafficking of neurotransmitters to the membrane (158).

Exocytosis of acetylcholine is prevented at the nerve terminal to the neuromuscular junction,

with consequent blockage of innervation of muscle activity (158). The clinical hallmark of

botulism is an acute flaccid paralysis, which begins with bilateral cranial nerve impairment

involving muscles of the eyes, face, head, and pharynx and then descends symmetrically to

involve muscles of the thorax and extremities. Botulinum toxin, unlike TeNT, probably does

not enter the central nervous system (CNS). In naturally occurring foodborne botulism,

gastrointestinal symptoms (e.g., abdominal cramps, nausea, vomiting, or diarrhea [more

often constipation or obstipation]) may precede the neurologic signs of descending flaccid

paralysis. Death results from respiratory failure caused by paralysis of the tongue or muscles

of the pharynx, leading to occlusion of the upper airway or from paralysis of the diaphragm

and intercostal muscles (13, 51). Generally, the patient’s hearing remains normal,

consciousness is not lost, and the victim is cognizant of the progression of the disease, which

of course can be a terrifying experience.

Wound Botulism in Intravenous Drug Users

An association between botulism and subcutaneous injection of Mexican black-tar heroin into

muscle or skin (skin popping) has been reported in the United States and in the United

Kingdom (34, 135). A study found 33 clinically diagnosed cases of wound botulism in the

United Kingdom and Ireland between 2000 and 2002 (34). The clinical diagnosis was

confirmed by laboratory tests in 20 of these cases; 18 cases were caused by type A toxin

and 2 by type B toxin. Wound botulism has also occurred after snorting of cocaine (111),

cosmetic injection of an unlicensed Botox preparation (52), and a tooth extraction (195).

Infant Botulism

Infant botulism is the most frequently recognized form of botulism in the United States (45%

of cases in California) and has been reported in at least 15 other countries

(12, 61, 75, 76, 148, 188). The geographic distribution of toxin types in infant botulism

cases has paralleled the spore distribution of C. botulinum toxin types in soils sampled from

different locations (12). Type A has been the most frequent BoNT type in cases of infant

botulism in states west of the Mississippi River, whereas type B cases have predominated

east of the Mississippi River (12, 170). Three cases have been caused by a strain(s) of C.

botulinum that produced toxins requiring both type B and F antitoxins for neutralization (88).

Type E infant botulism, caused by neurotoxigenic strains of C. butyricum, was initially

confirmed in two infants from Italy (76), and later in additional patients. Type F infant

botulism has been caused by neurotoxigenic C. baratii (76).

Most infants that contract botulism are 3 weeks to 6 months old (12), and the only clearly

defined risk factors have been exposure to soil, dust, and honey (12, 75). Since C.

botulinum spores have not been detected in any food or liquid ingested by these infants

other than honey (12), it is recommended that honey not be fed to infants less than 1 year

of age. Whatever the sources, the ingested spores of C. botulinum germinate within the

intestinal tract, and the vegetative cells multiply and produce the neurotoxin, which is then

absorbed into the bloodstream (12, 88). The first sign of illness is usually constipation, which

is often overlooked. Infants develop lethargy and mild weakness, with feeding difficulties,

pooled oral secretions, and an altered cry (12). They eventually lose head control and may

go on to develop ophthalmoplegia, ptosis, flaccid facial expression, dysphagia, other signs of

cranial nerve deficits, generalized muscular weakness, and finally respiratory insufficiency

and the inability to swallow. There is likely a spectrum of clinical features in infant botulism,

ranging from mild illness not requiring hospitalization to severe botulism requiring intensive

care. Human immune globulin that neutralizes BoNT (BabyBIG; intravenous BIG-IV) has

been licensed to the California Department of Public Health Infant Botulism Treatment&

Prevention Program [www.infantbotulism.org; 24-h/7-day phone, (510) 231-

7600 ] since 2003 (14, 75). Since 2007, it has been made available to physicians outside

the United States on a case-by-case basis. Early treatment has shortened hospital stays and

significantly reduced the associated costs of hospitalization (77).

Botulinum Toxin as a Bioterrorism Agent

Inhalational botulism, which results from aerosolization and inhalation of botulinum toxin,

has been considered a fifth category of botulism (13, 152, 165). Botulism could also result

from covert contamination of foods (13, 196). Inhalational botulism has been demonstrated

experimentally in monkeys (13, 152), accidentally in three veterinary personnel in Germany

who were exposed to reaerosolized BoNT from rabbits and guinea pigs with aerosolized BoNT

on their fur (13), and three researchers who were exposed to an aerosol during BoNT

manipulations (91). Terrorists have attempted to use aerosolized botulinum toxin as a

bioweapon in Japan but were not successful. Although inhalational botulism is possible, the

toxin is unstable in aerosols, and the more likely route of intentional intoxication is by food

contamination and oral ingestion.

Tetanus

Tetanus, caused by C. tetani, is often associated with puncture wounds that do not appear to

be infected. The organism and its spores can be isolated from a variety of sources, including

soil and the intestinal contents of numerous animal species. A potent neurotoxin (TeNT),

often referred to as tetanospasmin, is elaborated at the site of trauma and rapidly binds to

neural tissue, provoking a characteristic paralysis and tonic spasms (26). Tetanus is a totally

preventable infection with immunization with tetanus toxoid.

Tetanus is an intoxication analogous to botulism except that it occurs solely through wound

infection and production of tetanospasmin (TeNT). TeNT is synthesized as a single, inactive

polypeptide chain (150 kDa), which is cleaved by an intrinsic protease to produce an active

form, consisting of a heavy chain (100 kDa) and a light chain (50 kDa) linked by a disulfide

bond (158). The heavy chain binds to neuronal cells, and the three-dimensional structure of

this region has been elucidated (158). The light chain, a zinc endopeptidase, enters the cell

cytoplasm and traverses from the nerve terminal to the nerve cell body by retrograde axonal

transport (26, 158), eventually reaching neurons in the spinal cord and brain stem, where it

affects glycinergic and GABA (γ-amino-n-butyric acid)-ergic neurotransmission (26, 158).

Inhibitory impulses to CNS neurons are blocked, while uninhibited firing of motor nerve

transmission continues, resulting in prolonged muscle spasms of both flexor and extensor

muscles that can persist for weeks. The mechanism by which exocytosis of neurotransmitter

release is inhibited is analogous to that of BoNT; in fact, TeNT cleaves the vesicle-associated

membrane protein (VAMP) at the same peptide bond as BoNT B (117). Unlike with the

pathophysiology of botulism, TeNT is retrogradely transported in neurons to the CNS and its

site of action (26, 158).

The worldwide incidence of tetanus has been estimated to be as many as 500,000 cases per

year (26). Neonatal tetanus is endemic in developing countries due to a lack of vaccine

programs for infants or adult women. In developed countries, injection of drugs (i.e., skin

popping) has recently become an important risk factor (21, 85).

Additional Clostridial Species of Interest

C. innocuum is associated with bacteremia in immunocompromised hosts and has also been

recovered from patients with recurrent CDAD (3). It is often resistant to multiple drugs used

to treat anaerobic infections (3).C. ramosum was the second-mostcommon

Clostridium species (after C. perfringens) identified from clinical specimens from

children, including those with abscesses, peritonitis, bacteremia, and chronic otitis media

(35), and the third-most-common Clostridium species in adult cases of bacteremia (122).

This species may be resistant to clindamycin and multiple cephalosporins. As noted

earlier, C. tertium is often isolated from blood cultures from immunocompromised patients

and has been reported as a cause of neutropenic enterocolitis and meningitis

(56, 109, 124, 183). C. hathewayi and C. bolteae have been isolated from a variety of

human infections (72, 203), including a fatal case of sepsis (128). Phenotypically similar C.

clostridioforme is one of the clostridia most commonly isolated from human infections and

appears to be associated with human infections that are more serious or invasive than

infections with C. hathewayi or C. bolteae.

The emergence of 16S rRNA gene sequencing technology has provided a means of

identification of strains that may previously have been misidentified or classified

as Clostridium without species identification. Examples are from cases of bacteremia caused

by C. hathewayi (203), C. intestinale (66), and C. symbiosum(65); fatal sepsis due to C.

fallax in a previously healthy 16-year-old (89); and abscesses yielding C.

celerecrescens (80). Microarray analysis of DNA from fecal samples has also been useful in

the determination of predominant species in the large bowel (193). It is likely that in this era

of molecular identification techniques, a more accurate picture of clostridial infections will

emerge.

CLINICAL MICROBIOLOGY OF CLOSTRIDIAL DISEASES Back

to top

General Methods for Collection, Transport, and Storage of

Clinical Specimens

The proper selection, collection, and transport of clinical specimens are extremely important

for the laboratory diagnosis of clostridial infections. For recommended collection and

transport procedures in general, refer tochapter 16.

Specific Methods for Collection and Direct Examination of

Clinical Specimens

In addition to requiring aspirates and tissues, selected clostridial illnesses require special

specimens. The methods for collection and direct examination of these specimens are

described below.

Suspected Gas Gangrene or Necrotizing Fasciitis

Gas gangrene and necrotizing fasciitis represent extremely urgent situations requiring rapid

clinical diagnoses. Multiple tissue specimens should be sampled from the active sites of

infection when gas gangrene is suspected, because clostridia are often not distributed

uniformly in pathologic lesions. The direct examination of a Gram-stained smear of the

wound is of major importance for the early presumptive diagnosis of gas gangrene (10).

Characteristic findings in C. perfringens infections include the absence of leukocytic

infiltration and the presence of clostridia in smears prepared from central areas of the lesion.

Special note should be made of Gram-positive rods, with or without spores, because

sporulation in tissue is not common for the two species most frequently encountered in

wound and abscess materials, C. perfringens and C. ramosum. C. perfringens usually

appears as large, relatively short, fat, gram-positive rods with blunt ends and often in short

chains in tissue smears; the cells of C. ramosum are more slender and longer (Fig. 1). C.

perfringens may or may not be encapsulated in smears from wounds; capsules usually are

present in smears of endometrial specimens from postabortion C. perfringens infections.

Spore stains offer no advantage over Gram stains for demonstration of spores, but

examination with a phase-contrast or dark-field microscope may be helpful if the spores are

close to maturity. If spores are present, shapes (spherical or oval) and positions (terminal,

subterminal, or central) in the cells should be noted.

Suspected C. perfringens Foodborne Illness

A freshly passed fecal specimen and the suspected food are the preferred specimens for C.

perfringens culture and toxin assays. These specimens should be placed into sterile

containers, stored at 4°C, and shipped on cold packs as soon as possible. For optimal

recovery, stool specimens should be processed within 24 h of collection. Swab specimens are

inadequate for the toxin assay because the sample volume is insufficient.

Several methods have been described for the detection of CPE in feces, including cell culture

assays, enzyme-linked immunosorbent assay (ELISA), and reversed-phase latex

agglutination (141, 178). The cell culture assay using Vero cells is not as sensitive or as

reproducible as other methods (15, 74). The results of the RPLA kit (PET-RPLA; Oxoid,

Hampshire, United Kingdom, and Remel Inc., Lenexa, KS) are reproducible, and the test is

reasonably sensitive; however, nonspecific interference by fecal matter has been reported

(141). Similarly, the background bacterial DNA in stool has been reported to interfere with

PCR amplification of the enterotoxin gene (141). While an in-house ELISA system developed

by the Food Safety Microbiology Laboratory of the Central Public Health Laboratory, London,

United Kingdom, has been reported to be the most sensitive assay and is considered the gold

standard, the TechLab (Blacksburg, Virginia) CPE ELISA system has also provided a specific,

reliable, and practical tool for detecting CPE in fecal samples (15, 74).

Suspected Enteritis Necroticans (C. perfringens Type C)

If enteritis necroticans is suspected, the appropriate specimens include three blood cultures

from three different venipuncture sites, stool (at least 25 g, or 25 ml if liquid), and bowel

luminal contents or tissue from the involved bowel (e.g., surgical specimen or autopsy

material). Specimens should be transported in tightly sealed, leakproof containers for the

following: direct Gram staining, culture, isolation, identification, and typing of C. perfringens.

PCR assays for genotyping C. perfringens are being used in certain research or referral

laboratories to aid in diagnosis (175). Accordingly, DNA can be extracted for this purpose

from formalin-fixed intestinal tissue or culture and amplified by PCR using primers specific

for the cpa and cpb genes of C. perfringens type C.

Suspected CDI

Among the prerequisites for initiating a detailed microbiologic diagnosis of CDI are (i)

diarrhea as the lead symptom, (ii) the onset of diarrheal episodes 2 to 3 days after

hospitalization without exposure to other obvious inducing microorganisms, (iii) diarrhea for

more than 3 days without the causing organism being identified, (iv) a history of antibiotic

treatment of the patient, (v) belonging to a risk group (being >65 of age or

immunosuppressed or having severe gastrointestinal disease or another severe underlying

diseases), and (vi) frequent exposure to C. difficile, such as with exposure to nurses or other

medical personnel.

Confirming the diagnosis of C. difficile-associated enteric disease on the basis of both clinical

and laboratory criteria represents the ultimate gold standard. Different algorithms are

successfully applied in routine laboratory diagnoses. The different approaches are sometimes

governed by the number of stool samples to be processed in a laboratory unit. Generally,

laboratory results obtained with immunologic tests must be correlated and interpreted within

the context of the patient’s clinical presentation. The diagnosis of CDI has gained more

attention since the appearance of hypervirulent ribotype 027 strains. However, it needs to be

noted that more-severe cases also may arise from strains of other ribotypes.

Submission of Specimens

A single, freshly passed fecal specimen (ideally 10 to 20 ml of watery stool; minimum of 5.0

ml or 5 g) is the preferred specimen for C. difficile culture and toxin assay. To lessen the

chance of obtaining positive culture results from patients merely colonized with the

organism, only liquid or unformed stool specimens should be processed. Swab specimens of

stool are inadequate because the sample volume is insufficient for the toxin assay. Other

appropriate specimens include bowel luminal contents and surgical or autopsy samples of the

large bowel.

Specimens should be transported in tightly sealed, leakproof plastic or glass containers. For

optimal recovery, stool specimens should be cultured within 2 h of collection. Although

spores survive in refrigerated stool for several days, there will probably be a large decrease

in the number of viable vegetative cells of C. difficile in refrigerated specimens. Stools should

be placed in an anaerobic environment (anaerobic transport vial or bag) if culture must be

performed after storage. Adequate recovery of C. difficile organisms may be expected from

stools stored at 4°C for up to 2 days. Specimens for toxin assay may be stored at 4°C for up

to 3 days or should be frozen at −70°C if performance of the assay is delayed. Freezing at

−20°C results in a dramatic loss of cytotoxin activity, so detection limits may no longer be

reached.

Cell Culture-Based Methods of Diagnosis of CDI

Cultivation of C. difficile is not necessary for the molecularly based toxin assays described

below; however, cultivation is encouraged for subsequent molecular substrain typing and

epidemiologic studies. Cytotoxicity testing of cell cultures has long been called the gold

standard of C. difficile toxin testing due to its high sensitivity (94 to 100%) and high

specificity (99%) (60). However, normally only the activity of the TcdB cytotoxin is

monitored, since TcdA needs to be tested on special cells, like HT29 cells (187). Also, the

specificity of toxin-induced cytotoxicity is dependent upon neutralization of this effect by a

TcdB-specific antitoxin performed in parallel. The need for neutralization of this effect marks

a limitation of the test, and not every laboratory should perform neutralization due to time

and cost considerations.

Immunologic Methods for Diagnosis of CDI

Commercial tests that are available are listed in Table 2. Currently, the best approach is the

detection of TcdA and TcdB by enzyme immunoassay directly from stool specimens;

however, immunoassays generally show lower sensitivities and specificities (45 to 95% and

75 to 100%, respectively) than the tissue culture assay (136, 190). The result of toxin

testing is the declaration of the sample as being toxin positive or negative without any

differentiation of TcdA and TcdB. Immunological toxin detection should be done promptly

(within 24 h) following the collection of the sample. Due to the limitations of the specificity

and sensitivity of the enzyme immunoassay, the test should be repeated if initial results are

negative and if the clinical diagnosis is that of a CDI. The immunoassays that have been

introduced commercially all recognize TcdA effectively but recognize TcdB with a different

efficiency. TcdA is the antigen that is more easily detectable, and the tcdA gene is highly

conserved. In contrast, TcdB genes differ greatly between strains, and TcdB antibodies are

much more difficult to obtain. Single monoclonal antibodies detect only some of the TcdB

isoforms.

Testing of TcdA alone is no longer recommended, since some epidemic strains produce only

TcdB (198). Difficulties arise especially when C. difficile toxins are detected in strains like

ribotype 017 strains, which are TcdA2 and TcdB1, since sensitivity for TcdB is less than for

TcdA. Recent work with hamsters has attributed more importance to TcdB (131); in that

study, genetically manipulated strains deficient in TcdA production were still lethal, while

TcdA-positive strains in which the tcdB gene was interrupted were not (131). Attempts to

concentrate on the detection of TcdB (or its gene) alone are problematic since CDI may be

caused by strains that produce TcdA only (101, 130, 192).

Lateral-flow quick tests of different formats are being developed as “bedside tests.” In the

future, with such tests one might even differentiate between both toxins.

Antigen Detection (of GDH) for Diagnosis of CDI

Laboratories with high throughput are increasingly utilizing detection of glutamate

dehydrogenase (GDH), the method of choice for screening stool samples for C. difficile. GDH

is secreted by C. difficile into the stool and may be detected by commercial enzyme

immunoassays. Since GDH is not an enzyme exclusive to C. difficile, its detection is not

pathognomonic for CDI and the test does not inform us about the existence of the PaLoc or

its toxin profile. Lack of detection of GDH has a high negative predictive value, but false

positives occur. Accordingly, it is important to analyze GDH-positive stool samples for

TcdA/TcdB, preferably on the same day, to support the diagnosis of CDI. Stools that are GDH

positive, while toxin negative, should be routinely cultured, and the strains should be

submitted for toxin detection. Strains that remain GDH positive and toxin negative are

clearly atoxinogenic and thus apathogenic C. difficile isolates, and this represents a true

false-positive GDH test.

DNA-Based Methods for Diagnosis of CDI

Different commercial assays for the detection of the tcdA/tcdB genes have recently become

available (e.g., by BD, bioMerieux, and Seegene). A handicap of any genetic analysis is

that C. difficile colonization is not necessarily connected to CDI, since healthy carriers exist.

It is only when a high quantity of gene product (i.e., toxins) is expressed that CDI results.

Thus, if a PCR remains negative, CDI may be excluded to a high degree. The predictive value

of a negative PCR is, however, lost if testing is not reliable enough to definitively exclude the

presence of genetic variation within the gene of interest. This outcome is also the reason

why tests that detect only a single gene of the PaLoc, like tcdB, cannot be recommended for

routine use. The known variability of the tcdB gene classifies this gene as a poor candidate

for PCR diagnosis.

All tests are performed on DNA extracted from stool specimens by the use of a commercial

DNA preparation kit (e.g., those of Qiagen and MN-net). Among the prerequisites of a

reliable PCR assay, detection of a C. difficile-specific housekeeping gene (e.g., rrs, encoding

the 16S ribosomal subunit, or gluD, encoding the GDH) is necessary to be sure that C.

difficile is in the sample at all. In samples positive for the housekeeping gene but negative

for genes of the PaLoc, a nontoxinogenic C. difficile isolate is highly probable. Since it was

reported that such strains may protect against colonization with toxinogenic strains (68), this

may even argue against a cause of CDI. Thus, PCR-based approaches that detect only a

single gene should be considered unreliable for CDI diagnosis because of the adverse

consequences that a negative test result would have on treatment decisions.

Approaches that simultaneously detect a variety of species-specific genes and virulence

factors (e.g., tcdA andtcdB genes, cdtA and cdtB genes, and tcdC [Cepheid]) should be

developed and evaluated for their usefulness in routine diagnosis. Such assays would be of a

high predictive value since, for example, infection with a hypervirulent 027 strain could be

diagnosed. A potential dilemma that must be considered is that hypervirulence is not solely

associated with ribotype 027, and so a singular focus on this ribotype may be misleading.

Suspected Neutropenic Enterocolitis Involving C. septicum

The specimens of choice for suspected neutropenic enterocolitis involving C. septicum are (i)

three blood cultures collected from three different venipuncture sites, (ii) stool (at least 25 g,

or 25 ml if liquid), and (iii) luminal contents or tissue from the involved ileocecal area

collected at surgery or autopsy and transported in tightly sealed leakproof containers. In

addition, a biopsy sample of muscle (or an aspirate of fluid from the involved area, taken

with a needle and syringe) should be collected if the patient is also suspected of having

myonecrosis or another form of progressive infection.

Suspected C. botulinum or C. tetani Infection or Intoxication

Most hospital laboratories are not properly equipped to process specimens from patients

suspected of having botulism. Before collecting any specimens, the medical care providers

who suspect a diagnosis of botulism in a patient should immediately call their state health

department’s emergency 24-h telephone number or the CDC in Atlanta, GA [ (770)

488-7100 , 24-h/7-day emergency service] so that appropriate action can be taken to

establish the diagnosis, initiate treatment, and investigate the case. Acceptable specimens

include feces, enema fluid, gastric aspirates or vomitus, tissue or exudates, and postmortem

specimens. These specimens should be placed into sterile unbreakable containers. Serum

specimens (preferably >10 ml) should be collected as soon as possible after the onset of

symptoms. Clinical swabs should be collected in an anaerobic transport medium;

environmental swabs (from which spores may be isolated) may be sent in plastic containers

without any medium. Food specimens should be left in their original containers, if possible,

or placed in sterile unbreakable containers. All specimens should be stored at 4°C and

shipped on cold packs as soon as possible. Further information can be found at the CDC

botulism website (http://www.bt.cdc.gov/agent/botulism). Laboratories should have all of

the pertinent information and contact numbers handy. During investigations of possible

bioterrorism, sera, gastric aspirates, feces, and environmental or nasal swabs could be useful

for detecting aerosolized botulinum toxin that may have been inhaled (10, 204). All

specimens should be refrigerated until they can be transported to the laboratory for testing.

Certain clostridial toxins, particularly BoNT, TeNT, and iota toxin, are extremely toxic

molecules and are considered very potent poisons. The CDC recommends biosafety level 3

primary containment and personnel precautions for facilities producing BoNTs for study.

Ideally, personnel who work in laboratories should be immunized with a pentavalent (A to E)

toxoid; however, the vaccine is no longer available from the CDC. A biosafety manual should

be posted in the laboratory and should contain the proper emergency phone numbers and

procedures for emergency response. Regulations governing personnel safety for research

with select agents are outlined in the Code of Federal Regulations (67a) and the

manual Biosafety in Microbiological and Biomedical Laboratories (46a).

Direct Toxin Detection

Bioassays for BoNT and TeNT are currently the most important laboratory tests for the

diagnosis of botulism and tetanus (44, 63, 88). The definitive diagnosis of botulism is the

detection of BoNT (not the organism) (88). Currently, the only reliable assay for BoNT is the

mouse bioassay, together with neutralization of mouse toxicity with type-specific antitoxins

(88, 197). Detection of neurotoxins is usually performed on fecal specimens, blood (serum),

suspect foods in cases of foodborne botulism, and culture fluid following enrichment by

growth of the organism (44, 88). ELISAs, cell culture systems (87), and biosensor platforms

(62, 160) have also been used to detect BoNT (63, 160). Real-time PCR assays for detection

of C. botulinum BoNT gene fragments specific to BoNT A, B, and E have been developed as

alternatives to the mouse bioassay; this approach was found to demonstrate a sensitivity

and specificity similar to those of conventional approaches (5, 59). Potential problems with

PCR detection are strains that have the gene but do not produce toxin. Thus, the bioassay

remains the study of choice.

ISOLATION PROCEDURES Back to top

Isolation and Appearance on Plated Media

A summary of useful procedures for culture and isolation of clostridia is provided below.

Clostridia usually produce good growth on commercially available CDC anaerobe blood agar

and phenylethyl alcohol blood agar (PEA) after 1 to 2 days of incubation. Brucella agar with

5% sheep blood, Columbia agar, or brain-heart infusion agar supplemented with yeast

extract, vitamin K, and hemin may also be used as the nonselective blood agar medium.

Colony characteristics vary on different media. A few species, such as C. perfringens, form

colonies after overnight incubation or in as little as 6 h. When clostridia are suspected in

wound or abscess specimens (e.g., gas gangrene), egg yolk agar (modified McClung-Toabe

formula [see chapter 17]) should also be inoculated.

After incubation, the blood agar and PEA cultures should be examined under a dissecting

microscope, with attention being paid to the hemolysis pattern, colony structure, and

evidence of swarming or motile colonies. Egg yolk agar should be examined for evidence of

lecithinase (Fig. 2) or lipase production. Lecithinase activity is indicated by the development

of an insoluble, opaque, whitish precipitate within the agar. An iridescent sheen or oil-onwater

appearance (pearly layer) indicates lipase activity. Proteolysis, the third reaction that

can be seen on egg yolk agar, is indicated by a zone of translucent clearing in the medium

around the colonies. The same reactions can be visualized on the hemin-supplemented egg

yolk agar formulation recommended by Jousimies-Somer et al. (103) or on Lombard-Dowell

egg yolk agar (201), in addition to on the modified McClung-Toabe egg yolk agar

formulation.

Isolation of additional strains in the presence of swarming Proteus species or C.

septicum may require short incubation times (18 to 24 h), subculture onto PEA, or use of

anaerobe blood agar with 4% agar (“stiff blood agar”). When isolated colonies can be picked,

they should be subcultured to chopped-meat medium and incubated overnight for the

inoculation of differential media. Prereduced, anaerobically sterilized (PRAS) peptone-yeastglucose

media may be inoculated for detection of metabolic products by gas-liquid

chromatography (GLC) if the laboratory has that capability.

Spore Selection Techniques

Heat or ethanol treatment procedures can aid in detecting spores (103, 108). Ethanol may

be more effective than heat if the specimen contains relatively heat-sensitive clostridia

(e.g., C. botulinum type E and some strains of C. perfringens involved in foodborne

outbreaks). Heat treatment may be more effective than alcohol if homogenization is

incomplete and the specimen contains particulate matter that is not penetrated adequately

by the alcohol. For any spore selection technique, an untreated control subculture should be

prepared.

For alcohol treatment, an equal volume of absolute (or 95%) ethanol is added to a 1-ml

sample of a fecal suspension or homogenate of a wound or exudate in a sterile screw-cap

tube. The specimen is gently mixed at room temperature (22 to 25°C for 1 h). An Ames

aliquot mixer (Miles Laboratories, Inc., Elkhart, IN) is a convenient way to provide

continuous mixing. The treated material is used to inoculate chopped-meat–glucose or

thioglycolate medium, anaerobe blood agar, or egg yolk agar. The culture is incubated and

inspected for growth.

For heat treatment, a tube of chopped-meat–glucose or thioglycolate medium (5 ml) is

preheated in an 80°C water bath for 5 min, and 1 ml of sample suspension is added. The

culture is incubated for 10 min at 80°C, and the tube is removed and cooled in cold water.

The treated sample suspension is subcultured into an unheated tube of chopped-meat–

glucose or thioglycolate medium, anaerobe blood agar, or egg yolk agar. The cultures are

incubated anaerobically and examined for growth.

Isolation of C. difficile

Since C. difficile can be isolated from stool in asymptomatic patients, culture alone is not

sufficient to diagnose CDI and may misdiagnose AAD caused by other agents unless stool

samples are also assayed for the presence of C. difficile toxins. However, the recent

emergence of the epidemic, hypervirulent ribotype 027 strain has reinforced the need for

cultivation of C. difficile for subsequent typing, molecular studies, and determination of

antimicrobial susceptibility.

Currently, routine cultivation is done at 35 to 37°C on cycloserine-cefoxitin-fructose (CCF)

agar (see chapter 17) with or without blood. With prereduced medium, strains grow better.

Best results are achieved in CCF broth supplemented with pure taurocholate; however, 7

days of culture is recommended. Growth depends on strict anaerobic conditions. Typically,

the culturing time ranges from 3 to 7 days. Alcohol shock is a potential alternative to

improve C. difficile isolation (103, 108). Following incubation, plates should be examined

using a dissecting microscope. Colonies of C. difficile are yellowish to white, circular to

irregular, and flat, with a rhizoid or erose edge and a ground-glass appearance (Fig. 3). The

colonies have a distinctive odor like para-cresol (or horse manure). In addition, C.

difficile colonies on CCFA (cycloserine-cefoxitin-fructose agar) fluoresce chartreuse under UV

light (103).

Gram staining of C. difficile reveals rods that are gram positive to gram variable, thin, with

parallel sides, and 0.5 μm wide by 3- to 5-μm long. Isolation may be difficult due to the

presence of both vegetative and spore-forming bacteria. Presumptive identification of C.

difficile can be made by demonstrating typical colonies, Gram stain morphology, and

characteristic odor. Biochemical differentiation is easiest with detection of prolineaminopeptidase.

Definitive identification depends on demonstration of the unique pattern of

short-chain fatty acid metabolic products by GLC, by biochemical characterization of isolates,

or by 16S rRNA gene sequencing (11, 25, 79, 90) (Table 1).

IDENTIFICATION Back to top

Preliminary Identification

Identification of Clostridia in specimens from sites of infection due to mixed organisms can

be time-consuming and expensive. Use of selective and differential media for initial

processing can provide rapid and relevant information to the clinician. When isolated from

normally sterile sites and sites of serious infection, bacteria should always be completely

identified. Some of the organisms that warrant identification include C. septicum(associated

with gastrointestinal malignancy), C. ramosum, C. innocuum, and C. clostridioforme (which

are frequently resistant to antibiotics), and C. perfringens (53).

Clostridia are typically gram-positive rods by microscopic morphology. Some clostridia

appear to be gram negative, especially C. ramosum, C. innocuum, and the C.

clostridioforme group, but the special-potency antibiotic disk pattern (see below) verifies the

presence of gram-positive organisms. Second, it may be difficult to detect spores, so an

ethanol treatment, heat spore treatment, or malachite green stain may be necessary, and

phase-contrast or dark-field microscopy may be helpful. Third, the colonial morphology of

pure cultures may be variable, so the culture may appear to be mixed. Subcultures of single,

well-isolated colonies yield the same variable morphologies. Examination of colonies by

stereomicroscopy is helpful for noting colonial characteristics. Fourth, the aerotolerant

clostridia may be confused with Bacillus or Lactobacillus spp.Clostridium species sporulate

anaerobically only, grow much better anaerobically (larger colonies), and are almost always

catalase negative. Bacillus spp. sporulate aerobically only, usually grow better aerobically,

and are usually catalase positive. Aerobically grown C. tertium has colonial and cellular

morphologies similar to those of Lactobacillus spp. Certain clostridia can be identified with

relative ease by Gram staining, colony morphology determination, a positive indole reaction,

hemolysis on blood agar, and the tests described below (Table 3).

Special-Potency Disks

The isolate should be subcultured on blood agar with special-potency disks containing

vancomycin (5 μg), kanamycin (1 mg), or colistin (10 μg) and incubated anaerobically for 48

to 72 h at 35 to 37°C. Clostridia are colistin and kanamycin resistant and vancomycin

susceptible (Table 3), except for occasional C. innocuumisolates, which may be only

moderately susceptible to vancomycin (8).

Lecithinase and Lipase

The isolate should be subcultured on egg yolk agar and incubated anaerobically for 48 to 72

h at 35 to 37°C. Lecithinase activity is demonstrated by a white, opaque, diffuse zone

around the colonies that extends into the medium (Fig. 2). Lipase activity is indicated by an

iridescent sheen on the surface of bacterial growth and on the agar surface around the

colonies.

Spore Test

Media for the demonstration of spores include chopped-meat agar or broth and thioglycolate

medium. The culture should be incubated anaerobically at 5 to 7°C below the optimum

temperature (30°C) for the growth and sporulation of clostridia, except with C.

perfringens (should be induced at 37°C). Actively growing cultures may stand at room

temperature for several days to 1 week, and ethanol or heat spore treatments can be

performed as described above.

Definitive Identification of Clostridium Species

The traditional method for the phenotypic characterization and identification of clostridia is

the use of PRAS media for the determination of fermentation profiles and other

characteristics, combined with GLC analysis of metabolic end products (58, 150). However,

only a few laboratories have PRAS media or GLC available. Table 1 lists characteristics that

are useful for definitive identification of clinically relevant clostridia. The key reactions (bold

in Table 1) require minimal PRAS medium and can be used in conjunction with commercial

identification kits or individual preformed-enzyme tests, such as Wee Tabs (Key Scientific,

Round Rock, TX) or Rosco diagnostic tablets (Rosco, Taastrup, Denmark). Gelatin and esculin

hydrolysis, carbohydrate fermentation reactions, and metabolic end product analysis are

based on results obtained with PRAS media (Anaerobe Systems, Morgan Hill, CA).

PRAS Biochemical Inoculation

Actively growing broth cultures (without carbohydrate) or cell pastes suspended in broth

medium (e.g., peptone-yeast or thioglycolate) may be used to inoculate PRAS media.

Cultures are incubated for 48 to 72 h at 35 to 37°C, but overnight incubation is sufficient for

many clostridia.

Gelatin Hydrolysis

A PRAS gelatin tube with an actively growing culture is refrigerated along with an

uninoculated tube for at least 1 h. The tubes are removed to room temperature, inverted

immediately, and observed for liquefaction every 5 min. In a positive reaction, the gelatin is

hydrolyzed and thus fails to solidify, dropping to the top of the inverted tube immediately. In

a negative reaction, the medium fails to liquefy when it reaches room temperature (>30

min). A weakly positive reaction yields liquid medium at the time that it reaches room

temperature (<30 min).

Esculin Hydrolysis

Five drops of 1% ferric ammonium citrate are added to a tube of actively growing bacteria in

a PRAS esculin tube, and the tube is observed for a color change and fluorescence under UV

(366-nm) light. In a positive reaction, a black or dark-brown color develops, and there is no

fluorescence under UV light. In a negative reaction, no color develops, and the tube

fluoresces white-blue under UV light. Since many clostridia produce hydrogen sulfide (H2S),

which also reacts with the reagent to form a black complex, all tubes that darken after the

addition of reagent should be confirmed under UV light.

Carbohydrate Fermentation

The pH of actively growing organisms (>2+ turbidity) should be measured in a PRAS

carbohydrate tube. A positive reaction (“acid”) yields a pH below 5.5, and a negative reaction

results in a pH exceeding 5.9. “Weak acid” is indicated by a pH of 5.6 to 5.8. Details of GLC

procedures used for the analyses of metabolic end products listed in Table 1 are outlined

elsewhere (103).

Commercial kits, based on the detection of preformed enzymes with chromogenic or

fluorogenic substrates, have been marketed for the rapid identification of anaerobes. These

panels include RapID ANA II (Remel), Api 20A and RapID 32A (bioMerieux, Durham, NC),

Vitek ANI card and Vitek 2ANC card (bioMerieux), and the BBL Crystal anaerobe

identification system (BD, Franklin Lakes, NJ). The overall performances of these panels

vary, and the panels are not always satisfactory as the sole identification method for

clostridia (8, 43, 127, 134,142, 154, 159). In general, Gram stain reaction, cellular

morphologies, colonial characteristics, and aerotolerance of isolates (characteristics noted

above and in Tables 1 and 3) should be determined in conjunction with the use of

commercial microsystems. Supplementation of tests in these kits with individual tablets

(e.g., Wee tabs or Rosco tablets) can be helpful. Other useful supplemental tests for

clostridia include the tests outlined above, such as lipase and lecithinase production, the

reduction of nitrate, gelatin and esculin hydrolysis, carbohydrate fermentation, and metabolic

end product analysis using GLC.

Clostridial biochemical activity is quite variable, being saccharolytic/proteolytic and

saccharolytic/ nonproteolytic to asaccharolytic. The identification of asaccharolytic species

are most sophisticated. Liquid chromatography-mass spectroscopy (67) and molecular

biological methods such as 16S rRNA gene sequencing (166, 202) can be useful in these

cases. 16S rRNA gene sequencing is becoming more popular, though interpretation of results

must be done by those with special training. Other promising methods for the identification

of Clostridium species are fluorescent in situ hybridization (FISH) (27, 163) and matrixassisted

laser desorption–time of flight mass spectrometry (84).

Characteristics of Commonly Encountered Clostridia

Key characteristics that aid in the presumptive identification of the most common species are

listed below. See also Tables 1 and 3.

· C. bifermentans: colonies chalk-white on egg yolk agar, irregular, scalloped edge; many free

spores, often in chains; urease negative; indole and lecithinase positive. C. sordellii is similar

but is usually urease positive.

· C. bolteae: colonies usually have a slightly irregular edge; greening of agar around colonies;

gram negative; tapered ends; spores are rare; lactose negative and β-Nacetylglucosaminidase

(β-NAG) negative.

· C. butyricum: very large, irregular colonies with mottled-to-mosaic internal structure;

subterminal spores; ferments many carbohydrates.

· C. cadaveris: white-gray, entire or slightly irregular, raised to slightly convex; oval terminal

spores; spot indole positive.

· C. clostridioforme: same as for C. bolteae but lactose positive and β-NAG negative.

· C. difficile: colonies creamy yellow to gray-white (Fig. 3); irregular, coarse, mottled-tomosaic

internal structure; matte or dull surface; horse stable odor (para-cresol); subterminal

and free spores or spores infrequent; gelatin hydrolysis can be slow; mannitol and proline

positive; colonies fluoresce chartreuse on selective CCFA.

· C. glycolicum: colonies are gray-white with an entire or scalloped edge and convex;

subterminal and free spores.

· C. hathewayi: same as for C. bolteae but lactose and β-NAG positive.

· C. innocuum: gray-white to brilliant-greenish colonies, coarsely mottled-to-mosaic internal

structure, entire edge usually; terminal spores may be difficult to find; nonmotile; mannitol

positive; lactose, maltose, and proline negative.

· C. novyi type A: lecithinase and lipase positive, may swarm, strong beta-hemolysis.

· C. perfringens: double zone of beta-hemolysis around colonies (Fig. 4), boxcar-shaped rods,

spores rare, lecithinase positive (Fig. 2).

· C. ramosum: colonies resemble Bacteroides fragilis but usually have a slightly irregular

edge; Gram stain variable, palisading, slender rods; small round or oval terminal spores

(Fig.1); nonmotile; mannitol positive.

· C. septicum: swarms (Fig. 5); large, filamentous bacilli (Fig. 6); subterminal spores often in

“lemon” forms; DNase positive and sucrose negative.

· C. sporogenes: Medusa-head colonies, possible swarming, colonies adhere firmly to agar;

subterminal and many free spores; lipase positive.

· C. symbiosum: rods with tapered ends, football shaped, may form chains, often has spores.

· C. tertium: aerotolerant, terminal spores when anaerobically incubated. C. tetani: may form

a thin film of growth over entire agar plate, especially on moist media; drumstick spores.

Toxin tests are necessary for the identification of a few species. C. sporogenes cannot be

differentiated with certainty from the proteolytic group I strains of C. botulinum unless toxin

tests are used. A few strains of C. botulinum produce lecithinase as well as lipase and are

difficult to distinguish from C. novyi type A except by toxin tests. As a supplement to the

methods described, the various types of C. botulinum and other clostridia can be

presumptively identified on the basis of differences in their CFA profiles (11, 79, 90) and by

typing methods such as pulsed-field gel electrophoresis (PFGE) or other molecular analyses.

Finegold et al. (72) described a multiplex PCR procedure for rapid distinction of the three

species of the C. clostridioforme group.

TYPING SYSTEMS Back to top

In the event of severe cases of disease and lethal outcomes for patients, typing of strains is

recommended. Physicians should be aware of their local C. difficile situation and should be

sure that hypervirulent strains are not present. Monitoring of the local situation gives a hint

about the local standard of hygiene, the education of nurses (and other colleagues), and the

risks for patients. It is obvious from investigations of C. difficile strains that every region may

have its particular pattern of strains, with ribotypes differing between different countries in

Europe but also between different regions of a single state (20).

To resolve endemic-disease situations, to monitor the spread of infection, and to assess the

genetic relatedness of the associated strains, the successful cultivation of C. difficile is

required. Starting with the pure culture, several approaches have been used for such

analyses, including restriction endonuclease analysis, PFGE, PCR-ribotyping, multilocus

sequence typing, and others (105). For example, the recent hypervirulent isolates have been

typed as 027 by ribotyping, designated toxinotype III by typing the toxin A and B genes of

the PaLoc, and designated NAP1/2 by restriction fragment length polymorphism analysis

(North American PFGE type 1/2) (139). Recently, Bouvet and Popoff used triple-locus

sequence analysis of the toxin regulatory genestcdC, tcdR, and cdtR to assess the

evolutionary relatedness of strains isolated from humans and food animals (30). The variety

of methods used for this analysis already indicates the difficulties in setting a standard

procedure that may be used routinely and worldwide.

PCR-ribotyping could become such a reference method; however, type strains are not easy

enough to receive and are available only from a very few places. Therefore, only a limited

number of labs have enough standard strains available to qualify the ribotypes of the strains

that are posted. Normally the isolated strains should be sent to such labs; the use of Amies

transport medium is an appropriate means for doing so. PCR-ribotyping is done with two

specific primers that amplify the spacer region in between the 16S and the 23S ribosomal

RNAs. The spacer region is known for its heterogenous nature, as opposed to the highly

conserved rRNA genes themselves. C. difficile contains 10 rRNA copies, and variations in the

spacer regions are seen between different strains but also between different rRNA copies of a

single strain.

Multilocus sequence typing has been successfully used as a reproducible and discriminating

system for strain typing of Clostridium botulinum type A using clinical and food isolates

(94), C. perfringens isolates from necrotic enteritis outbreaks in broiler chicken populations

(47), and C. septicum isolates recovered from poultry flocks experiencing episodes of

gangrenous dermatitis (145). Leclair et al. described a modified PFGE protocol to be the

most useful method for typing epidemiologically related C. botulinum type E strains, in

comparison with randomly amplified polymorphic DNA analysis and automated ribotyping

using clinical and food isolates associated with four botulism outbreaks (123). Furthermore,

analysis of the variable numbers of tandem repeats (VNTRs) within the genome, called

multiple-locus VNTR analysis, has been described for C. perfringens(48, 156).

Epidemiologically related isolates previously typed by PFGE were also examined by multiplelocus

VNTR analysis, and the congruency of the two methods was found to be very high.

Macdonald et al. described VNTR regions in Clostridium botulinum strains, providing a rapid

and highly discriminatory tool to distinguish among C. botulinum BoNT/A1 strains for

investigations of botulism outbreaks (132).

SEROLOGIC TESTS Back to top

Serologic procedures are not practical for secure strain identification from colonies.

Furthermore, no standardized tests are available for the detection of antibodies

against Clostridium species in clinical specimens to confirm diagnoses of clostridial infections.

To evaluate the vaccination status, determination of immunoglobulin G antibodies against

tetanus toxin may be useful, but in cases of an unclear immunization status, preventable

vaccination should be done.

ANTIMICROBIAL SUSCEPTIBILITIES Back to top

Antimicrobial susceptibility studies with strains of a number of clostridial species are

summarized in Table 4. Now that more laboratories are identifying anaerobes by 16S rRNA

gene sequencing, more-accurate species identification will be available and more-reliable

susceptibility data can be generated. Drugs lacking antimicrobial activities against various

clostridia include trimethoprim-sulfamethoxazole, ampicillin, and clindamycin. No resistance

of clostridia to ampicillin-sulbactam or piperacillin-tazobactam has been noted, and

antimicrobial resistance is uncommon among clostridia with respect to imipenem,

metronidazole, and vancomycin. Five species (all with small numbers of strains) and C.

perfringens show little or no resistance to the antimicrobial agents under consideration

(Table 4). Organisms with some resistance to three drugs includeC. ramosum, C.

innocuum, and C. clostridioforme.

EVALUATION, INTERPRETATION, AND REPORTING OF

RESULTS Back to top

The isolation of a Clostridium species from a clinical specimen, even a blood culture, may or

may not be significant clinically, and culture results should be interpreted in relation to the

patient’s clinical findings. Clostridia of the patient’s own intestinal microbiota may be present

on the skin and may contaminate blood samples or other specimens. Bacteremia may be

transient or clinically insignificant. In addition, most clostridia currently encountered in

wounds, exudates, blood, and other normally sterile body fluids are opportunistic and may

not cause serious or progressive disease unless conditions are suitable in the host. As

discussed earlier in this chapter, one exception to this generalization is C. septicum, which is

rarely encountered in blood cultures except from patients who have an underlying

malignancy or neutropenic sepsis. C. septicum sepsis is an infectious disease emergency that

requires prompt and clear communication between the laboratory and the clinician in order

to institute early surgical measures and treatment with antimicrobial agents to improve

outcomes. C. tertium, C. perfringens, and other clostridia, to a lesser extent, may be

involved in serious infections that require emergency measures. The best approach for

preventing tragic consequences that may be avoidable is good communication between

microbiologists and clinicians.

The accurate and timely reporting of preliminary results (e.g., findings from direct

microscopic examinations of clinical specimens), as well as early culture results (after 24 and

48 h of incubation), can be extremely valuable to the physician. For smaller laboratories

without anaerobic chambers, incubation of the appropriate media in anaerobic jars provides

acceptable recovery for most clinically significant anaerobes, assuming that optimal

collection and transport of specimens are performed. The colony characteristics and

microscopic features of some clostridia (e.g., C. perfringens, C. sordellii, and C. sporogenes)

may be distinctive, so preliminary or presumptive reports may be released before

aerotolerance studies are completed. Accurate, definitive identification is needed to better

define the role of clostridia in disease, to aid the clinician in selecting optimal treatment, and

for public health purposes (e.g., hospital-acquired C. difficile disease).

Potentially life-threatening diseases due to Clostridium species or their toxins, such as

botulism, tetanus, or severe cases of C. difficile infection, should be carefully examined by

the physician and the microbiologist together to ensure optimal sample collection and

transport, immediate processing, and initiation of specific therapy. Furthermore, health care

institutions require accurate and rapid diagnosis for early detection of possible outbreaks and

to implement effective control measures.