Donal Minihan

E. coli O157 origin
E. coli is a bacterium, many strains of which live harmlessly in the gastro-intestine of both animals and humans (Anon., 1997d). E. coli O157 is a recently emerged pathogen in an evolutionary scale, with its first report as a cause of human illness in the United States in 1982 (Riley et al., 1983; Anon., 2001c). The main evolutionary tool of E. coli O157 is its enhanced ability to acquire useful foreign DNA (Whittam et al., 1998). It has been hypothesised that E. coli O157 evolved from an enteropathogenic E. coli ancestor that acquired virulent factors (e.g. genes encoding for shiga toxin production) from other organisms, both E. coli and non-E. coli species culminating in the zoonotic pathogen that exists (Law, 2000).

E. coli O157 Human Infection
E. coli O157 is a member of a group of zoonotic pathogens, verotoxigenic E. coli (VTEC) which were first detected in the US in 1982 (Riley et al., 1983). Non- O157 VTECs, for example E. coli O26, E. coli O111 and E. coli O113 have also been isolated from clinical cases worldwide (Anon., 1998d; Beutin et al., 1998; Keskimaki et al., 1998; Reilly, 1998; Anon., 2000a; Brooks et al., 2004). VTEC O157 has been detected worldwide including the United States and associated with several large and high profile outbreaks (Marsh et al., 1992; Kohli et al., 1994; Watanabe et al., 1999; Cowden et al., 2001).

In New York the annual incidence rate for E. coli O157 per 100,000 population has ranged from 1.6 in 1995 to 8.5 in 1999 (Chang et al., 2004). In 2006 a total of VTEC O157 (590), VTEC non-O157 (209) were recorded in 10 States by Foodnet in the US (Anon, 2007). VTEC O157 is the most commonly reported VTEC in Ireland, accounting for 96% of the recorded VTEC infections in man in Ireland since 2001 (Foley and McKeown, 2002a; Garvey et al., 2003).

The clinical manifestations of VTEC O157 infection in man includes freedom from symptoms; non-bloody diarrhoea; haemorrhagic colitis; haemolytic uraemic syndrome (HUS); thrombotic thrombocytopaenic purpura (TTP) (Swerdlow and Griffin, 1997; Todd and Dundas, 2001). The non-bloody diarrhoea can remain mild or become severe. However, in some cases the non-bloody diarrhoea may become haemorrhagic on the second or third day from the onset of diarrhoea (Mead and Griffin, 1998). The symptoms of haemorrhagic colitis include severe abdominal haemorrhage followed by bloody diarrhoea, oedema, erosion or haemorrhage of the mucosal lining of the colon. It has been estimated that 1% of haemorrhagic colitis cases die without progressing to HUS (Anon., 1999c).

In sporadic cases three to seven percent of cases progress to HUS. However, in outbreaks the percentage of cases progressing to HUS may increase to 20% or more (Mead and Griffin, 1998). In New York the annual incidence rate for diarrhea-associated HUS per 100,000 population has ranged from 0.1 in 1995 to 0.2 in 1999 (Anon, 2007). HUS is typically diagnosed six days after the onset of diarrhoea. Of the patients that develop HUS approximately three to five percent die, five percent develop chronic renal failure, strokes and other major sequelae, 30% develop proteinuria and other major sequelae and 60% resolve (Mead and Griffin, 1998).

TTP is a further potential complication of haemorrhagic colitis, which mainly affects adults rather than children. TTP is a serious illness involving haemorrhage due to a clotting deficiency. The clinical syndrome is predominantly neurological (Anon., 1999c). A significant facet of this disease is that vulnerable sectors of society, namely children and the elderly, are most at risk of developing severe VTEC O157 disease (Mead and Griffin, 1998; Duffy et al., 2000; Todd and Dundas, 2001).

Sources of E. coli O157 human infection
The first reported outbreak of VTEC O157 infection in man was linked to eating undercooked beef burgers in Oregon in 1982 (Riley et al., 1983). Since then, a proportion of VTEC O157 outbreaks have involved either hamburgers or ground (minced) beef products (Todd, 2000). For example 48 of 80 VTEC O157 outbreaks in the USA during 1998 and 1999 reported by the Centers for Disease Control and Prevention (CDC) were associated with food, and beef and beef products accounted for 21 of these food associated VTEC O157 outbreaks (Anon., 1999d; Anon., 2000b). In 2007 there were several large multi-State outbreaks of E. coli O157. Fresh spinach and beef were the sources of the E. coli O157 infection in the large outbreaks. In the September outbreak VTEC O157 was isolated from Topp's brand frozen ground beef patties. This lead the USDA to recall 21.7 million pounds of beef. Over the course of the outbreak 33 people were hospitalized and two patients developed HUS (Anon., 2007a). Studies of sporadic cases of VTEC O157 illness in the USA have identified ground beef as the substantial source of human exposure to the pathogen (MacDonald et al., 1988; Le Saux et al., 1993; Mead et al., 1997; Slutsker et al., 1998).

Many other foodborne sources of VTEC O157 infection have also been identified over the past 20 years. Foods such as cheese, milk and sprouts have also been associated with outbreaks worldwide (Chapman et al., 1993; Anon., 1998c; Durch et al., 1999; Breuer et al., 2001). In 2000 a large Canadian outbreak, in which 2,300 people became ill and seven people died was associated with E. coli O157 contamination of the water supply (Hrudey et al., 2003). Contamination of a water supply was associated with a local herd of cattle shedding E. coli O157 in their faeces. An increasingly reported and important source of infection for man is direct and indirect contact with ruminants (cattle, sheep, goats and deer) (Smith et al., 2001b; Warshawsky et al., 2002; Woodward et al., 2002; Payne et al., 2003).

Other species have also been implicated as a source of VTEC O157 infection. A report published in 2002 described a case in which rabbits were implicated in contaminating a children? playground (Bailey et al., 2002). Following investigation, it was found that rabbits and humans along with cattle from a neighboring herd all shed an indistinguishable VTEC O157 strain. In conclusion, there are a large variety of potential sources of VTEC O157 infection for man. Ruminants and in particular cattle are an important potential source of infection to man via direct and indirect contact and animal products such as ground beef.

Infection of cattle with E. coli O157
The gastrointestinal tract of clinically normal cattle appears to be transiently colonised by E. coli O157 (Besser et al., 1997). However, it has been reported that calves less than 36 hours old experimentally infected with E. coli O157 subsequently developed diarrhoea and enterocolitis (Dean-Nystrom et al., 1997). Faecal shedding of E. coli O157 varies widely among animals of the same age group.

The authors of an experimental infection study of calves and adult cattle reported that most calves and cattle only shed the organism in feces for seven and two weeks, respectively. However, the authors reported that some calves and adult cattle shed the organism for up to 20 and 14 weeks post-infection, respectively (Cray and Moon, 1995). A further field study that investigated E. coli O157 infection reported a pattern of prolonged periods of low prevalence shedding, interspersed by short periods of high prevalence shedding for groups of cattle (Mechie et al., 1997). A number of studies investigated the site in the alimentary tract where E. coli O157 persists. The colon was the area consistently highlighted as the site of persistence of the organism in ruminant animals (Brown et al., 1997; Dean-Nystrom et al., 1999; Buchko et al., 2000; Grauke et al., 2002). However, recent experimental evidence has shown the terminal rectum to be the principal site of colonisation. Detailed examination has demonstrated that in these experimentally infected individuals the majority of tissue-associated bacteria were adherent to mucosal epithelium within a defined region extending up to 5 cm proximally from the recto-anal junction (Naylor et al., 2003).

Prevalence of E. coli O157 in Cattle
The organism E. coli O157 has been isolated from cattle over the past three decades. The presence of the organism in the feces has been extensively used as an index of infection prevalence. However, it is difficult to compare and contrast the prevalence of E. coli O157 faecal shedding of cattle between different studies because of variations in sample size, sampling frequency, animal type, sampling method, recovery methodology, seasonal and geographical factors and local husbandry practices. Some of the earlier studies reported prevalence of E. coli O157 faecal shedding of 0.28% and 1% (Chapman et al., 1989; Hancock et al., 1994). The prevalence of cattle shedding of E. coli O157 in feces has varied considerably (1.3% to 41.5%) in later studies (Mechie et al., 1997; Bonardi et al., 2001; Chapman et al., 2001; Leung et al., 2001; Paiba et al., 2001; Sargeant et al., 2003; Minihan et al., 2003; Minihan et al., 2003a; Berg et al., 2004; Minihan et al., 2005).

Factors affecting E. coli O157 shedding by cattle
Many factors predisposing to high faecal shedding of E. coli O157 by cattle remain unclear. However, a seasonal variation in the faecal shedding of E. coli O157 with a summer peak has been widely described (Hancock et al., 1997b; Mechie et al., 1997; Heuvelink et al., 1998; Bonardi et al., 1999; Garber et al., 1999; Van Donkersgoed et al., 1999). A number of studies have also identified that younger cattle have a higher faecal E. coli O157 shedding prevalence than adult cattle (Dargatz et al., 1997; Bonardi et al., 1999; Van Donkersgoed et al., 1999). Not all studies have reported consistent observations regarding the prevalence of faecal shedding in different age groups. For example, a Norwegian study? results indicated that no difference in the faecal shedding between young and adult cattle (Vold et al., 1998).

Several reports have implicated diet as a risk factor associated with the shedding of E. coli O157. However, the observations of all these studies are not consistent. Cattle fed a barley based diet were associated with higher faecal shedding than cattle fed a corn based diet (Buchko et al., 2000). High fibre low protein diets (e.g. grass hay) were associated with higher faecal shedding than low fibre high protein diets (Kudva et al., 1997; Hovde et al., 1999). Interestingly, a number of reports have reported an increase in faecal shedding of E. coli O157 following a change in diet (Kudva et al., 1995; Dargatz et al., 1997; Kudva et al., 1997).

Certain husbandry practices have also been implicated as potential risk factors for shedding of E. coli O157. A study reported that farms on which alleyways were flushed with water to remove feces were 8.0 times more likely to have animals feces test positive for E. coli O157 than herds where other methods of manure removal were used (Garber et al., 1999).A longitudinal study in an Irish feedlot reported that the prevalence of E. coli O157 fecal shedding may be related to the pen and E. coli O157 contamination of the pen floor feces, water trough and feed.

The persistence of a small number of strains of E. coli O157 on farms over a period of time has highlighted the importance of the farm environment as a potential source of the organism, and of the on-farm transmission of these E. coli O157 strains (Heuvelink et al., 1998; LeJeune et al., 2004). A longitudinal feedlot study investigating cattle fecel and environmental E. coli O157:H7 isolates indicated that a single isolate of E. coli O157:H7 may have been passed rapidly through cattle pens, with the environment acting as a significant reservoir for transmission Scott et al., 2006).

Studies have demonstrated that transport and lairage do not cause an increase in fecel shedding of E. coli O157 (Barham et al., 2002; Minihan et al., 2003). This observation is potentially significant as it suggests that transport of cattle to the abattoir might not increase the E. coli O157 load entering the abattoir.

Prevalence of beef contamination with E. coli O157
It is widely accepted that cattle commonly shed E. coli O157 in their feces and are a potential source of the organism for man (Besser et al., 1997). Beef has been frequently associated with E. coli O157 infection in man, as discussed earlier. The prevalence of E. coli O157 on carcasses has been investigated. Studies have investigated the sequential contamination of beef carcasses at the pre-evisceration stage and at the post-processing stage (entering the chill-room). The results of these studies indicated a significant reduction in the level of carcass contamination between the pre-evisceration stage (26.7% and 43%) and the post-processing stage (1.2 % and 2%) (Elder et al., 2000; Barkocy-Gallagher et al., 2003). This highlights the importance of hygienic work practices by abattoir personnel during carcass dressing procedures on the subsequent carcass contamination levels. Post-carcass dressing E. coli O157 contamination levels have ranged from 0% to 3.2% in a number of studies conducted worldwide (Elder et al., 2000; Madden et al., 2001; Phillips et al., 2001; Arthur et al., 2002; Barkocy-Gallagher et al., 2003; McEvoy et al., 2003; Minihan et al., 2003). A number of studies have also investigated the prevalence of E. coli O157 in minced beef or beef products, with prevalences ranging from 0.0% to 2.8% (Anon., 2001c; Phillips et al., 2001; Vernozy-Rozand et al., 2002; Cagney et al., 2004). These figures confirm the potential of beef and beef products as a potential source for E. coli O157 infection for man.