Chapter 1

General Introduction:

Members of the genus Clostridium are anaerobic, motile, Gram-positive, spore-forming rod present in nature especially in the soil. Microscopically, they have a long drumstick like appearance with a bulge situated at their terminal ends. Gram-staining is one of the easiest methods employed to identify them as the cell simply incorporates the dye while the spore does not take up the stain.

Clostridium shows ideal growth when they are grown on blood agar at human body temperatures. Under unfavorable conditions for growth, however, the bacterium produces spores to tolerate the stressed environment where in, the active bacteria would not have survived. Clostridial species in their active forms secrete exotoxins responsible for specific and serious conditions such as tetanus, botulism and gas gangrene. The four clinically important species of Clostridium are C. botulinum, C. tetani, C. perfringens and C. difficile.

1.2 Types of Clostridium species important in humans

1.2.1 C. tetani

Clostridium tetani is the bacterium which causes tetanus (lockjaw) in human beings. The spores of C. tetani can be acquired from all kinds of skin trauma and they outgrow in deep, necrotic wounds. In the anaerobic environment, the spores germinate and lead to the formation of active C. tetani cells. If these cells are present at the tissue level, then they release an exotoxin named tetanospasmin which affects the nervous system specifically by transmission via the neurons, eventually to the brain. One of the major effects of the toxin involves constant contraction of the skeletal muscles which occurs as a result of blockage of inhibitory interneurons which controls the contraction of muscles. Prolonged contraction of the muscles eventually leads to respiratory failure which has a high mortality rate if not treated early. One of the best ways to avert infections caused by C. tetani is to immunize oneself.

1.2.2 C. botulinum

Clostridium botulinum is identified to generate one of the most powerful toxins till date and is the causative agent of the deadly botulism food poisoning. Due to the fact that Clostridium spores are ubiquitous, they sometimes find their way into foods placed in anaerobic storages such as cans and bottles. Once the cans are completely sealed, the spores begin to germinate and the bacteria then secretes their toxin which has an effect on the peripheral nerve cells (McLauchlin et al, 2006; McLauchlin, Grant et al.,2006). Patients suffer from muscular flaccid paralysis apart from blurred vision. Immediate administration of an anti-toxin to the patient is necessary to raise the probability of survival. Infantile botulism is also caused in a very similar way but is far milder than its adult counterpart. The most frequent source for the spores which germinate in the infant’s intestinal tract is however honey.

1.2.3 C. perfringens

This is a non-motile bacterium which is an invasive pathogen that can be contracted from dirt via large cuts or wounds. After spore germination takes place, C. perfringens cells proliferate and release their exotoxin which causes necrosis of the surrounding tissue (Clostridial myonecrosis destroys muscular tissues). The bacteria themselves produce gas that leads to a bubbly deformation of the infected tissues (Smedley et al.,2004) (Smedley, Fisher et al.,2004). In the United Kingdom and United States they are the third most common cause of food-borne illness, with poorly prepared meat and poultry being the main culprits in harboring the bacterium (Lin and Labbe 2003). The clostridial enterotoxin mediating the disease is often heat-resistant and can be detected in contaminated food and feces.

The bacteria are killed at cooking temperatures, but the heat-resistant spores they produce are able to survive and may actually be stimulated to germinate by the heat. If the food is not eaten at once but is allowed to cool slowly, the bacteria produced when the spores germinate multiply rapidly. Unless the food is reheated so that it is piping hot (at least to 60oC and preferably to 75oC), the bacteria will survive. After ingestion, if there are sufficient numbers present, the bacteria will produce toxins and the toxins will cause symptoms.

Infection with Clostridium perfringens normally causes diarrhoea and severe abdominal pain. It may occasionally cause nausea but it rarely causes vomiting or fever.

1.2.4 C. difficile

First described in the 1950s, pseudomembranous enterocolitis was thought to be due to either Staphylococcus aureus, an organism that had become prevalent in hospital in-house patients who had received antibiotics (Keidan and Sutherland 1954) or to Candida albicans. In 1974, a prospective study of 200 patients who were treated with clindamycin were detected with diarrhoea in 21% and pseudomembranous colitis in 10%. A toxin produced by a Clostridium species was proposed as the cause of clindamycin-induced ileocaecitis in hamsters in 1977 (Bartlett, Onderdonk et al.,1977); later this toxin was isolated from the samples of patients' stool, with evidence and counter-evidence presented for C difficile and Clostridium sordellii as causative organisms. However, by 1978, C difficile had been clearly identified as the causal agent of antibiotic-associated colitis (Chang, Bartlett et al.,1978).

1.3 Microbiology of C. difficile

C. difficile is a Gram-positive, motile bacterium, spore-forming rod, toxin-producing, obligate anaerobe that is present in nature. Colonies are relatively large (2–17 μm in length) rough, grey and fast growing; CCFA medium (consisting of cycloserine, cefoxitin, and fructose agar in an egg-yolk agar base) is highly selective for its growth (Aslam, Hamill et al.,2005). Clostridium also shows optimum growth when plated on blood agar at human body temperatures [Figure 1]. Over the past decade, it has become a very prominent nosocomial infection worldwide. It is notable that C. difficile infection caused ward closures in 5% of UK hospitals in 1993, and by 1996, this figure had risen to 16% (Popoola, Swann et al.,2000).

In 1935, Hall and O'Toole first isolated this organism, designated it to be Bacillus difficilis, from the meconium and faeces of newborn infants (Tabaqchali and Jumaa 1995). "The difficult clostridium" was resistant to early attempts at isolation and grew very slowly in culture. The organism was shown to produce a lethal toxin in experimental animals, but since it was commonly found in the stools of healthy neonates it was classified as commensal and subsequently attracted little attention until 1974, when a comprehensive study showed that C. difficile was widespread in nature and could be isolated from the stools of several animal species and from patients' faeces and genitourinary tracts (Hafiz and Oakley 1976; Bartlett 2007). Clostridium difficile is now the most frequent bacterial enteric pathogen in the developed world. This organism has been the recognized agent of 20% to 25% of cases of antibiotic-associated diarrhea and around 90% of serious pseudomemraneous colitis cases since its discovery in 1978.

Disease symptoms are due to the production of tow toxins (A and B). During the past 3 to 4 years there has been the recognition of a new strain designated the NAP-1/ ribotype 027 strain which has been associated with some unique features including epidemics in geographically defined areas, more serious forms of disease and relative refractoriness to standard therapy. This 027 strain is linked to several deadly hospital outbreaks of C. difficile-associated diarrhea (CDI) which are now found rather frequently in Canada, the United States, and in greater parts of Europe (Cloud and Kelly 2007). This strain was found to produce greater than 10 times as much of toxin A and toxin B, as historic isolates (Larson, Parry et al.,1977; Cloud and Kelly 2007).

The link between clindamycin associated colitis and C. difficile was not made until 1977. Stool filtrate from a patient with pseudomembranous colitis showed a cytotoxic effect on tissue culture cells, which suggested the presence of a toxin of unknown source (Larson, Parry et al.,1977). At the same time investigators in the United States showed that clindamycin and other antibiotics induced a fatal caecitis in hamsters; the caecal contents contained a filterable toxin that was cytopathic in a cell culture assay and would reproduce the typical lesions when injected intracaecally (Bartlett, Onderdonk et al.,1977).

An organism identified as C. difficile was isolated from the animals and was shown to be the source of the toxin. Soon after, C. difficile and its toxins were detected in the stools of patients with pseudomembranous colitis (Hopkin 1978; Larson, Price et al.,1978), and oral vancomycin was shown to be an effective treatment in animal models and in patients (Bartlett 1984). C difficile has since become established as a major cause of nosocomial diarrhoeal infection.

Figure 2: Coloured transmission electron micrograph of Clostridium difficle forming an endospore (red). Dr Kari Lounatmaa/Science Photo Library

1.4 Genome of C. difficile

The genome of C. difficile strain 630 which was known as a virulent and multidrug-resistant strain was completely sequenced by Sebaihia in 2006 (Sebaihia, Wren et al.,2006). A large proportion of the genome (11%) consists of mobile genetic elements, mainly in the form of transposons. These mobile elements are supposed to be responsible for the acquisition of an extensive numbers of genes which are involved in antimicrobial resistance, virulence, and host interaction. The metabolic abilities encoded in the genome show multiple adaptations which enable the bacteria to survive and grow within the gut in low acidic environment.

The genome consists of a circular chromosome of 4,290,252 bp with a G+C content of 29.06%, an additional circular plasmid of 7,881 bp with a G+C content of 27.9%, and 3971 genes. There are two separate genes, tcdA and tcdB encoded for enterotoxin A (308kDa) and cytotoxin B (270kDa) (Barroso, Wang et al.,1990; Dove, Wang et al.,1990; Hundsberger, Braun et al.,1997). Both toxins A and B, share 63% of amino acid sequence homology; enzymatic domain, a hydrophobic region believed to be involved in translocation through endocytic vesicles into the cytosol, and a carboxy-terminal domain which contain the so-called clostridial repetitive oligopeptides (CROPs); the latter mediate.

TcdA and TcdB are among the largest bacterial toxins reported to date and are joined by Clostridium sordellii lethal toxin (TcsL) and hemorrhagic toxin (TcsH) and Clostridium novyi alpha toxin (Tcnα) to form the group of large clostridial toxins (Table 1). TcdA and TcdB, located in a 19.6-kb pathogenicity locus (PaLoc), which is a short chromosomal segment carried by pathogenic strains of C. difficile which also encompasses three other small open reading frames [Figure. 2].

Nontoxigenic and nonpathogenic strains of C. difficile contain a 127-bp sequence (Hammond and Johnson 1995). The sequence similarity and its position suggests that the tcdA and tcdB genes are the result of gene duplication (von Eichel-Streiber, Laufenberg-Feldmann et al.,1992). The lack of toxin activity for nontoxigenic strains can be explained by the absence of at least part of the toxin A gene. The expression of these two genes is regulated by tcdC gene. The expression of the tcdC gene and weak transcription of the genes encoding toxin A (tcdA), toxin B (tcdB), a positive regulator (tcdD), and a holin-like protein (tcdE) (Hundsberger, Braun et al.,1997).

The inverse is seen during the stationary phase, suggesting that tcdC negatively regulates toxin expression (Hundsberger, Braun et al.,1997) (232 amino acid residues). This gene is believed to result in over expression of tcdA and tcdB and increased production of toxins A and B, which may be responsible for the apparent higher pathogenicity in certain ribotypes (i.e., PCR type 027). Some strains also have cdtA and cdtB which are encoded for binary toxin (Sebaihia, Wren et al.,2006).

1.5 Pathophysiology

The pathogenesis of CDI is complex and not fully understood but what is known is that important pathophysiological features of C. difficile include heat-resistance of the spore and toxin production. Also the precipitating event for C. difficile colitis is disruption of the normal colonic microflora which is usually caused by broad-spectrum antibiotics most commonly implicated (Figure 4), (Kyne, Hamel et al.,2002; Wilcox 2003) such as clindamycin, broad-spectrum penicillins, and cephalosporins.

There are a number of antibiotics with a reduced propensity to induce infection such as aminoglycosides, metronidazole, antipseudomonals, and vancomycin. The risk of developing antibiotic-associated diarrhea is twice more when antibiotic therapy is received for longer than three days (Wistrom, Norrby et al.,2001). After disruption of the colonic microflora, colonization of C. difficile generally occurs by ingestion of the heat-resistant spores, which in turn switch over to their vegetative forms in the colon.

Depending on the immunological status and the host factors, an asymptomatic carrier state or clinical manifestations of C. difficile colitis develop. Manifestation of the disease ranges from mild diarrhea to life-threatening- C. difficile colitis. C. difficile-associated diarrhea can occur up to eight weeks after the discontinuation of antibiotics. In most cases, C. difficile infection occurs on days 4 through 9 of antibiotic therapy (Cloud and Kelly 2007). As the leading cause of hospital-acquired diarrhoea, C. difficile colonizes the large bowel of patients receiving antibiotic therapy and produces two toxins, which are responsible for the disease pathologies. Toxin B is around 1000 times more cytotoxic than toxin A (Kabins and Spira 1975).

Toxin A is also an enterotoxin in that it loosens the tight junctions between the epithelial cells that line the colon, which in turn helps toxin B to enter into the epithelial cells. These two toxins, TcdA and TcdB, are encoded on a pathogenicity locus with both negative and positive regulators of their expression. Following expression and release from the bacterium, TcdA and TcdB translocate to the cytosol of target cells and inactivate small GTP-binding proteins, which include Rho, Rac, and Cdc42. Inactivation of these substrates occurs through monoglucosylation of a single reactive threonine, which lies within the effector-binding loop and coordinates a divalent cation critical to binding GTP. By glucosylating small GTPases, TcdA and TcdB cause actin condensation and cell rounding, which is followed by death of the cell. TcdA elicit effects primarily within the intestinal epithelium, while TcdB has a broader cell tropism (Farrell and LaMont 2000; Voth and Ballard 2005).

1.6 Host factors

The major host factors predisposing patients to the development of symptomatic C. difficile-associated Infections (CDI) include antibiotic therapy. A cohort study of Sherbrooke inpatients recorded that fluoroquinolone use (especially ciprofloxacin) has emerged as the major risk factor for CDI in the context of ongoing epidemic (Pepin, Saheb et al.,2005). Other risk factors include advanced age, especially people over 65 years; number and severity of underlying diseases and abnormal immune response to C. difficile toxins (Hundsberger, Braun et al.,1997).

Patients who recently received immunosuppressive therapy or recently underwent surgical procedures are at the highest risk for fulminant disease, and those with a previous history of CDI. The increased risk may be due partly to the debilitated patient's inability to mount an IgG antibody immune response against C. difficile toxin A. The ability to mount an immune response is not protective against C. difficile colonization, but is associated with decreased morbidity, mortality, and recurrence of CDI (Kyne, Hamel et al.,2002; Sebaihia, Wren et al.,2006).

1.7 Clinical presentation

The presentation of the infection can range from asymptomatic colonization or self-limiting diarrhea to severe diarrhea, pseudomembranous colitis [Figur7], megacolon, colonic perforation, and death (Larson, Price et al.,1978).
The incidence of diarrhea in hospitalized patients who receive antibiotics ranges from 3% to 29%. C difficile has been found as the causative agent in 10–25% of patients with antibiotic-associated diarrhea, 50–75% of those with antibiotic-associated colitis, and 90–100% of those with antibiotic-associated pseudomembranous colitis (Bartlett 1990). Mortality of CDI ranges from 6% to 30% when pseudomembranous colitis is shown to be present (Olson, Shanholtzer et al.,1994; Moshkowitz, Ben et al.,2004; Pepin, Valiquette et al.,2004), and is substantial even in the absence of colitis. Most patients present with passing of large amounts of watery stool which is well known by healthcare workers who can often recognize it from its unique characteristic foul smell (Brazier 1998; Wilcox 1998).

It was found in prospective case-controlled study that patients also present with paralytic ileus (21%), abdominal pain (22%), fever (28%) and a raised white cell count (50%) (Gerding, Johnson et al.,1995). Dehydration and electrolyte imbalance are often found due to passing large amount of diarrhea and, when disease is prolonged, significant malnutrition can develop (Brazier 1998).

The incubation period for disease after exposure or acquisition is probably <1 week. Infection with C. difficile can be diagnosed up to 4 weeks after discontinuing an implicated antibiotic. It can also be triggered by other drugs such as cytotoxic drugs, antacids, stool softeners and laxatives which may trigger CDI (Hundsberger, Braun et al.,1997). Certain procedures such as nasogastric intubation, enemas and other intensive care procedures may also predispose to the infection (Cunney, Magee et al.,1998).

1.8 Treatment

The initial treatment for CDI was oral vancomycin. In the early 1980s, metronidazole was also shown to be effective, perhaps equally so, and a strong preference to avoid the use of vancomycin in hospital inpatients, reinforced by several sets of therapeutic recommendations (Gerding, Johnson et al.,1995), has led to increasing reliance on metronidazole. In 1997, the American Gastroenterology Association published recommendations for treating CDI which include discontinuation of antibiotics to avoid tissue damage, supportive non-specific therapy, and addition of metronidazole for those who failed to respond within 2–3 days (Fekety 1997).

However, oral vancomycin was recommended for the following categories of patients: those who were critically ill, unable to tolerate metronidazole, pregnant women, or those under the age of 10 years, those who failed initial therapy with metronidazole, or those whose infecting organism proved to be metronidazole resistant. The past few years have witnessed an increase in the failure rate of antimicrobial therapy (Pepin, Valiquette et al.,2004). Some patients simply fail to respond to conventional therapy, and others relapse after discontinuation of treatment. The Cochrane database reports only nine well-designed randomized trials that have assessed treatments for CDI.

Importantly, antimicrobial susceptibility testing of contemporary and historic isolates of NAP1/027 indicates a substantial increase in resistance to all fluoroquinolones (McDonald, Killgore et al.,2005). Fluoroquinolones are now the most widely prescribed antibiotics in many developed countries, (Linder, Huang et al.,2005) and the acquisition of fluoroquinolone resistance has been thought to promote the emergence of NAP1 (McDonald, Killgore et al.,2005). A substantial increase in the proportion of patients who fail to respond to metronidazole and a doubling of the frequency of postmetronidazole relapses have been noted, which could also promote the dissemination of this strain.

1.9Recommendations for treatment

Treatment with the offending antibiotic has to be stopped, if possible. Fluids and electrolytes are given to compensate fluid loss during diarrhea. Antimotility agents should not be given. If specific treatment is required, metronidazole 500 mg is given orally every 6–8 hours for 7–10 days. Vancomycin at a dose of 125 mg orally every 6 hours is a second-line alternate agent. If the patient cannot tolerate the drug orally, intravenous metronidazole is used, but this should be switched to oral therapy once the patient is able to tolerate it. In the case of ileus or toxic megacolon, intravenous metronidazole is used, perhaps adding vancomycin retention enemas in a dose of 500 mg mixed in 100 mL normal saline.

Vancomycin is avoided unless metronidazole seems to be ineffective, the patient is pregnant or allergic to metronidazole, or true resistance is shown. In case of recurrence, the agent that had been used to treat the initial episode of CDI is re-used, usually metronidazole. In case of multiple recurrences or refractory disease, the use of probiotics, immunoglobulin, or steroid is considered. In all cases, strict contact isolation of the patient is essential in controlling the spread of the disease to other patients. Symptom-free carriers not to be treated. Emergency colectomy reduces mortality in patients with fulminant CDI. Like patient who aged 65 years or more, in those immunocompetent, those with a leukocytosis >or=20 x 10(9)/L or lactate between 2.2 and 4.9 mmol/L (Lamontagne, Labbe et al.,2007).

1.10 Epidemiology

CDI is increasingly recognised as one of the most important healthcare associated infections. A number of aspects classify CDI as a severe potential threat associated with receiving healthcare. The number of cases reported on a weekly basis has steadily increased in Scotland over the last 10 years. Increasing numbers of outbreaks in hospitals and other healthcare institutions have been observed in Scotland as well as the rest of the UK. Some of these outbreaks have included cases of severe disease and deaths. Mortality rates for all deaths mentioning CDI as underlying or direct cause of disease have more than doubled from 1999-2004 in England and Wales. Reports indicate that patients complicated with CDI spend 1-3 weeks longer in hospitals than control group patients. Frequent relapses of the disease are contributing to difficulties with the treatment and may cause adverse health effects. The increasing numbers of elderly is furthermore expected to increase the risk of epidemics in the future (Health Protection Scotland)

Toxin-producing strains of C. difficile are carried in the normal colonic microflora of only about 5% of healthy adults (Kelly, Pothoulakis et al.,1994). However, 15% to 70% of neonates are carriers of C. difficile (Riley 1998). This percentage varies as a result of the degree of hospital exposure, birth in an environment where C. difficile is abundant, or if the neonate obtained maternal antibodies through breast milk. Although neonates are more frequent carriers of C. difficile, they do not often develop pseudomembranous colitis unless gastrointestinal motility disorders or other conditions (eg, severe neutropenia with leukemia) are present to increase the risk. Neonatal resistance to C. difficile colitis is believed to be due to the inability of the toxins to attach to the mucosa of newborns, because of immature membrane toxin receptors, or the protection from the toxins by maternally-acquired antibodies. After the first year of life, the carrier rate gradually decreases, reaching adult levels by three years of age (Reinke and Messick 1994; Matsuki, Ozaki et al.,2005; Tonooka, Sakata et al.,2005; Trejo, Minnaard et al.,2006).

Clostridium difficile-associated diarrhea (CDI) has become an increasing clinical problem as a nosocomial disease affecting mainly the elderly, patients with serious underlying diseases, and surgical patients (Bignardi 1998; Brazier 1998; Karlstrom, Fryklund et al.,1998). C. difficile probably represents the most common current cause of bacterial diarrhea in developed countries and, besides caliciviruses, the most common nosocomial diarrheal pathogen (Samore, DeGirolami et al.,1994; Karlstrom, Fryklund et al.,1998). Based on laboratory reports, at least 5,000 cases of CDI occur every year in Sweden, corresponding to 60 cases per 100,000 inhabitants per year, and more than 70% of the cases are associated with a hospital stay (Karlstrom, Fryklund et al.,1998). Currently over 6,000 cases were reported in Scotland from October 2006 until September 2007 (Health Protection Scotland).

Clusters of nosocomial cases of CDI have been attributed to transmission of C. difficile between patients but also indirectly through the hands of health care workers or via contaminated surfaces or vomit (McFarland, Mulligan et al.,1989; Clabots, Johnson et al.,1992). Furthermore, some strains may be more transmissible and also more virulent than others and thus be associated with higher attack rates and a high local incidence of CDI (Johnson, Samore et al.,1999).
A study performed over 10 years ago reported a low frequency of CDI in the community (7.7 cases/100 000 person-years of observation) (Hirschhorn, Trnka et al.,1994). A more recent report from the Centers for Disease Control and Prevention estimated that the minimum annual incidence of community-acquired CDI in the Philadelphia area between July 2004 and June 2005 was 7.6 cases/100 000 population (2005). Nine percent of patients in the present study had no previous exposure to the healthcare system, and were considered to be genuine cases of community-acquired CDI (Price, Dao-Tran et al.,2007).

It was difficult to ascertain genuine cases of community-acquired CDI admitted to this tertiary care hospital.
C. difficile is now the first organism suspected by health care personnel when a hospitalized patient develops diarrhea. C. difficile infection is a nosocomial disease that spread primarily by the medical staff, and hospital epidemics are relatively common. Usually, patients acquire the organism from the hospital and not from their own flora. C. difficile-associated disease (CDI) is increasingly being reported in many regions throughout the world. Moreover, severe disease has been reported in non-traditional hosts (e.g. younger age, seemingly healthy, non-institutionalized individuals residing in the community, and some without apparent antimicrobial exposure). In addition to the sudden increase in frequency of CDI, an increased rate of relapse/recurrence, disease severity and resistance to traditional treatment have also been noted. Much of this change was due to the emergence of one toxigenic strain, classified according to PCR as ribotype 027/toxinotype III and pulsed-field gel electrophoresis (PFGE) as NAP1 (Warny, Pepin et al.,2005) (Figure 8). This epidemic strain represented 2–3% of hospital isolates of C difficile (Rupnik, Avesani et al.,1998; Rupnik, Brazier et al.,2001; Geric, Rupnik et al.,2004).

By using restriction endonuclease analysis, the same genotype as NAP1/027 (also known as type BI) was found in only 14 of more than 6000 US historic isolates obtained before 2001 (McDonald, Killgore et al.,2005). NAP1/027 was not reported to cause either severe disease or outbreaks until recently, when it was identified as the cause of several outbreaks in the USA (McDonald, Killgore et al.,2005).

In Sherbrooke, between 2003 and 2004, it was found that as many as a sixth of inpatients with health-care-associated C difficile as a direct or indirect consequence of this infection (Pepin, Valiquette et al.,2005).

On June 4, 2004 tow outbreaks of NAP1 were reported in Montreal, Quebec and Calgary, Alberta, in Canada (Eggertson 2004). Sources put the death count as low as 36 and high as 89, with approximately 1,400 cases in 2003 and within the first few months of 2004.C .difficile infection continues to be a problem in the Quebec health care system in the 2004. As of March 2005, it has spread into the Toronto, Ontario area, hospitalizing 10 people. One has died while the others have been discharged.

A dominant strain that was pulsed-field gel electrophoresis (PFGE) type NAP1, toxinotype III, and contained a tcdC deletion and ctdB was also discovered in samples from a Stoke Mandeville Hospital in the United Kingdom between 2003 and 2005 and from other outbreaks which were associated with increased morbidity, frequent need for colectomy, and mortality in the USA (McEllistrem, Carman et al.,2005). This strain has also been implicated in an epidemic at two Dutch hospitals (Harderwijk and Amersfoort, both 2005).

Moreover, retrospectively the strain has been identified in isolates from sporadic US cases obtained in the early 1980s (McDonald, Killgore et al.,2005). The finding of an association between NAP1/027 (or BI) and high toxin production in the context of an epidemic associated with a high case-fatality ratio confirms the suspicion that the epidemic in Quebec and UK is caused by a more virulent strain (Pepin, Valiquette et al.,2005).

In the UK, where the number of reported cases of CDI doubled over 3 years, NAP1/027 is the cause of ongoing outbreaks in at least three hospitals with a high case-fatality ratio (health protection agency). The appearance of this virulent strain, in association with certain environmental and antimicrobial exposure factors, may be combining to create the 'perfect storm' (Owens 2007).

In 2005 C. difficile spores was isolated from 12 (20%) of 60 retail ground meat samples purchased over a 10-months in Canada. Eleven isolates were toxigenic, and 8 (67%) were classified as toxinotype III. which suggests that C. difficile also could be responsible for food poisoning or at least be foodborne. Previously, a study investigating the role of psychrotrophic clostridia on "blown pack" spoilage of commercial packages of chilled vacuum-packed meats and dog rolls reported 2 incidental isolates of C. difficile (Moorhead and Bell 1999).

In Bacteriological evaluation of commercial canine and feline raw diets C. difficile was isolated on direct culture from 1 turkey-based food (Weese, Rousseau et al.,2005). In terms of cost and productivity, C. difficile is a major burden to our health care system. There are estimated to be 250,000 to 300,000 cases of C. difficile a year in U.S. hospitals, which cost hundreds of millions of dollars for hospital care. Hospital costs for this condition in the USA (Kyne, Hamel et al.,2002) and UK (Wilcox, Cunniffe et al.,1996) exceed US$4000 per case. A typical case results in 1 to 2 extra weeks of patient care costing roughly $10,000.

This price assumes that the patient responds to treatment and does not relapse or develops complications (Wilkins and Lyerly 2003). The incidence of CDI has increased in the past decade, with a 10-fold increase reported in Quebec (Pepin, Valiquette et al.,2004), as has the proportion of patients who have severe, refractory, or recurrent disease (Musher, Aslam et al.,2005). The successful control of C. difficile will require healthcare systems (including administrators, and leadership within several departments such as environmental services, infection control, infectious diseases, gastroenterology, surgery, microbiology and nursing), clinicians, long-term care and rehabilitation facilities, and patients themselves to be proactive in a collaborative effort (Owens 2007).

1.11 Reservoirs, sources, and transmission of C. difficile

The major reservoirs for C. difficile in the hospital and community infection are patients with CDI or asymptomatic carriers of C. difficile. Patients with symptomatic disease heavily contaminate their immediate hospital environment and the spores can persist for several months on surfaces. Shedding of C. difficile into the environment depends on the patient's status. In one study, they compared the rate of environmental contamination in rooms of patients with C. difficile-associated diarrhea to that of contamination in rooms of C. difficile asymptomatic carriers. They showed that contamination was significantly higher in rooms of patients with diarrhea compared to asymptomatic carriers. They also analyzed contamination in rooms without C. difficile-positive patients and found a contamination rate of 8%, showing that spores of C. difficile can persist, despite routine cleaning of rooms (McFarland, Surawicz et al.,1990).

C. difficile diarrhea was reported as a community acquired infection in Ireland in 1998 (Kyne, Merry et al.,1998). CDI has also been reported in the community as an emerging pathogen in animals. Early typing comparisons did not identify animals as an important source for human Infection. In 1983 C. difficile was isolated in household pets such as dogs and cats in the UK (Borriello, Honour et al.,1983).

In 2001 in Canada C. difficile was also associated with diarrhea in dogs and cats (Weese, Staempfli et al.,2001) but recent report in 2006 from Canada have shown a marked overlap between isolates from calves and humans, including two of the predominant outbreak types, 027 and 017 which suggested that C. difficile may be associated with calf diarrhea, and cattle may be reservoirs of C. difficile for humans (Rodriguez-Palacios, Stampfli et al.,2006). C. difficile has also been found in retail meat samples, suggesting that food could be involved in the transmission of C. difficile from animals to humans (Rupnik 2007).

Over the past 5 years, C. difficile has emerged as a major cause of neonatal enteritis in pigs. Piglets 1–7 days of age are affected (Songer and Anderson 2006; Yaeger, Kinyon et al.,2007), with gross lesions frequently including mesocolonic edema. Colonic contents may be pasty-to-watery and yellow, although some piglets are constipated or obstipated. In 2007 a study was carried out to assess the correlation between the presence of C. difficile toxins (TCd) in the colon contents of neonatal pigs and a number of parameters, including gross evidence of diarrhea mesocoloninc edema, typhlitis, and colitis. They found C. difficile may represent an important subclinical issue in neonatal swine (Yaeger, Funk et al.,2002). In 2006 study from Zimbabwe was to determine the prevalence of C. difficile in faeces of domestic animals (chickens, rabbits ,cattle, and goat), soil and drinking water in a rural community.

The results of the study have shown that 95% of the samples were positive for C. difficile and chickens are a major reservoir of C. difficile in rural communities in Zimbabwe, where it was isolated in 17.4% of chicken feces samples. Detection of C. difficile in well water and household-stored water demonstrates the potential of water as a source of infection. Some of the water contamination may have been through faecal material of some domestic animals such as chickens, cattle and goats, which are kept as free-range animals in the community studied. Also it was found that C. difficile was not-uncommon cause of enteric disease in mature horses, mostly when they are treated with antimicrobials and hospitalized (Baverud, Gustafsson et al.,2003; Baverud 2004).

In 2006 a first report of C. difficile as the main cause of fatal enterocolitis in elephant came from Denmakr when they retorted two cases of fatal enteritis caused by C. difficile in captive Asian elephants are reported from an outbreak affecting five females in the same zoo. It is speculated that the feeding of large quantities of broccoli, a rich source of sulforaphane, which has been shown to inhibit the growth of many intestinal microorganisms may have triggered a subsequent overgrowth by C. difficile (Bojesen, Olsen et al.,2006).

Transmission of C. difficile is thought to occur via the oro–fecal route. Outbreaks in hospitals and typing of strains suggested that transmission is probably via staff hands. A study documented positive hand cultures in 59% of hospital personnel caring for patients with positive culture (Samore, Venkataraman et al.,1996). Transmission can also occur by direct contact with contaminated surfaces. Some reports also suggested a transmission by direct inoculation into the bowel via contaminated materials such as thermometers (Jernigan, Siegman-Igra et al.,1998).

Factors that may explain the ease of transmission include resistance of the spores to the most commonly used disinfectants and antiseptics, the antibiotic pressure in hospitalised patients and the promiscuity of patients. Unrecognized patients with C. difficile, or re-admissions of patients with C. difficile, can contribute to the reintroduction and spread of C. difficile to other patients or the environment. By using restriction endonuclease as a typing method, showed that 84% of cases of nosocomial acquisition of C. difficile strains were preceded by a documented introduction of the strain to the ward by another asymptomatic admission (Clabots, Peterson et al.,1988).

1.12 C. difficile typing methods

1.12.1 Phenotyping method:

C. difficile has been typed by using different methods based on phenotypic properties; one of these methods was antibiograms, which was one of the early methods. Isolates resistant to three different antibiotics were found in one of the first documented outbreak investigations (Burdon 1982). Those isolates were found in the surgical ward and were distinct from isolates in the rest of the hospital. However, this method is at best only preliminary, and more detailed approach was tried by Wüst and co-workers who combined soluble protein polyacrylamide gel electrophoresis (PAGE), plasmid analysis, immunoelectrophoresis of extracellular antigens and antibiograms to a number of isolates from related cases of C. difficile infection (Wust, Sullivan et al.,1982).

By using these methods, they showed that 12 of the 16 strains were similar, showing strong evidence, that cross-infection had taken place. A combination of bacteriophage typing and bacteriocin methods has been used which showed a high percentage of non-typeable strains (Sell, Schaberg et al.,1983). In 1981 the immuno-chemical fingerprinting of EDTA-treated cell extracts of C. difficile was evaluated (Poxton and Byrne 1981). Nakamura and co-workers were the first investigators to use serum agglutination as a typing method by raising three antisera against C. difficile (Nakamura, Mikawa et al.,1981). This method could differentiate four marked serovars among 79 isolates from healthy carriers. In 1985 Delmée's group improved this method and developed a serotyping scheme that could recognise 19 distinct sero-groups (Delmee, Homel et al.,1985). This method is repeatedly used as the standard by which other typing methods are compared.

These early typing methods were developed to understand the epidemiology of C. difficile infection at a local level, which were adequate for local use; there was a need for typing schemes that could be implemented to further understanding of the epidemiology of C. difficile disease on a wider scale. To facilitate this, comparisons between typing schemes were performed, and in 1988 Mulligan and co-workers found good correlation between the types recognised by serotyping, PAGE, plasmid profiling of cell surface antigens and immunoblotting (Mulligan, Peterson et al.,1988). Sodium dodecyl sulfate (SDS)-PAGE of whole-cell proteins was applied to 79 isolates in an outbreak investigation, which yielded a maximum of approximately 40 bands ranging in size from 18 to 100 kilo-daltons (kDa).

This investigation showed 60 of the 79 isolates was similar. Serogrouping was compared to SDS-PAGE of EDTA-extracted cell surface antigens and 61 isolates were analyzed (Ogunsola, Ryley et al.,1995). This method showed bands of sizes between 30 and 67 kDa and distributed their 79 isolates into 17 groups, which were similar to the results of serogrouping. A method of whole-cell fingerprinting by pyrolysis mass spectrometry (PMS) has been successfully used in investigating C. difficile outbreaks which had the advantage that it could cope with a large number of strains and had a high degree of differentiation (Magee, Brazier et al.,1993). However, the disadvantages of this method were the high cost of the equipment and its inability to assign a permanent type to a strain.

1.12.2 Molecular typing methods:

In terms of the stability of marker expression and providing greater levels of typeability, molecular typing methods are generally regarded as superior to phenotypic methods and a number of molecular methods have been used in C. difficile. Due to the sparse distribution of the extra-chromosomal genetic elements within the species, plasmid profiling proved largely unsuccessful. However, in 1987 Kuijper’s group analyzed the chromosomal DNA of C. difficile by using whole cell DNA restriction endonuclease analysis (REA) in which HindIII was used in the investigation which showed cross-infection between two patients in the same room (Kuijper, Oudbier et al.,1987).

REA is a highly reproducible and discriminatory method; however, the disadvantages are it is very labor-intensive and a technically demanding procedure, especially for large numbers of isolates. An alternative genotypic method (Saiki, Scharf et al.,1985) called the restriction fragment length polymorphism (RFLP) which involves initial REA digestion followed by gel electrophoresis and Southern blotting was used to detect specific restriction site heterogeneity. RFLP, however, is also a very labor-intensive method and REA/RFLP methods have generally been replaced by methods based on the polymerase chain reaction (PCR).

Another genotypic method is called arbitrarily primed PCR (AP-PCR) which permits the detection of polymorphisms within the target genome without prior knowledge of the target nucleotide sequence. A closely related method is called random amplified polymorphic DNA (RAPD). This method commonly uses two oligonucleotide primers which are arbitrary sequences and short in length (c.10 bp). In 1994, a method was evaluated by Barbut and co-workers using two 10-bp primers in an investigation of antiobiotic-associated diarrhea (AAD) in AIDS patients (Barbut, Mario et al.,1994). The discriminating ability of t this method was virtually unlimited as it was always possible to use other random primers. It is also simpler and more rapid than the other molecular methods which came later on such as restriction enzyme analysis or pulsed field gel electrophoresis.

PCR ribotyping was first applied to C. difficile by using specific primers complementary to sites within the RNA operon this was carried out by Gurtler who targeted the amplification of the spacer region between the 16S and 23S rRNA regions (Gurtler 1993). C. difficile was shown to possess many copies of the rRNA genes, which differ in number and size between strains and also within the same genome. This approach was modified by Cartwright and co-workers who tested it in 102 isolates obtained from 73 symptomatic patients (Cartwright, Stock et al.,1995). The primers used in this study were the same primers used by Gurtler, and instead of using denaturing PAGE gels, their PCR fragments were separated by straightforward agarose gel electrophoresis.

Furthermore, they showed that the quantity of DNA used in the reaction does not affect the banding patterns which was a problem associated with AP-PCR and RAPD methods. This approach was modified for the routine use by O'Neill and co-workers who improved and simplified the DNA extraction method by using modified primers to the 16S−23S spacer region. This method was able to produce fragments ranging from 250 to 600 bp separated by agarose gel electrophoresis (Cartwright, Stock et al.,1995). Since 1995, this method has been used routinely by the UK Anaerobe Reference Unit in Cardiff, which provides a C. difficile typing service for the UK. A library consisting of 116 distinct ribotypes from over 3000 strains from all sources examined, has been constructed (Stubbs, Brazier et al.,1999).

Contrary to the other methods, the whole genome is analyzed by pulsed field gel electrophoresis (PFGE) after digestion with rare cutting restriction endonucleases, such as SmaI, KspI, SacII or NruI, which result in 10 fragment length polymorphisms per strain. PFGE has been applied successfully to investigate 30 epidemiologically unrelated isolates and 22 isolates of C. difficile from an outbreak in an elderly care centre (Talon, Bailly et al.,1995). PCR ribotyping was considered more distinctive than AP-PCR and PFGE methods in a study (Collier, Stock et al.,1996). The disadvantages of PFGE include the initial cost of the equipment, the slowness of the procedure and its complexity.

Bidet and co-workers compared all three methods and concluded that PCR ribotyping, is the best technique for C. difficile ribotyping (Bidet, Lalande et al.,2000). It was noted that some strains are untypeable by PFGE due to degradation of the extracted DNA and those strains belong to serogroup G, which corresponds to PCR ribotype 1 in the Stubb’s group library (Stubbs, Brazier et al.,1999). Rupnik has developed the toxinotyping method by describing 11 toxinotypes and has been compared to PCR ribotyping (Rupnik, Brazier et al.,2001). Good correlation has been noted between the two methods, whilst applying toxinotyping to each type in the PCR ribotype library, five novel toxinotypes were discovered (Brazier 2001).