John Hicks, DVM, presented this information, during the Twelfth Annual Iowa State University Swine Disease Conference for Swine Practitioners last November in Ames, Iowa. Hicks is a veterinarian with Carroll Veterinary Clinic, Carroll, Iowa.

Introduction
Post‑weaning hemolytic E. coli infections in piglets have been a common traditional challenge in swine production for many years. Many of these cases were thought to involve nutritional and/or management problems. In most instances our recommendations for treating/controlling these problems would include:

1. Mild feed restriction.

2. Acidification of drinking water.

3. Strategic administration of antibi-otics in the drinking water according to a culture and sensitivity pattern.

These traditional protocols were effective until a re‑emergence of edema disease in the late 1990s. Nursery death losses were reported to exceed 25 percent in some systems in the southeastern United States.

Our first experiences with this dramatic disease occurred in November 2000. Two of our worst experiences were actually in newly placed finishing pigs.

Many other cases began to show up throughout the practice area over the next two years. We found, as did most other practitioners, that the traditional approach to post‑weaning hemolytic E. coli, most notably, antibiotic administration, was ineffective. Our clinic adopted the approach of preventive strategies, the cornerstone of which has been the administration of a non‑toxigenic strain of E. coli whose fimbria matched the isolate that was identified in the previous outbreak.

Agent
The etiological agents for edema disease are the B‑hemolytic strains of Escherichia coli that produce the enterotoxin known as STX‑2e, or Shiga‑like toxin. These strains may or may not also possess the genes to produce one or more of the other identified enterotoxins. These other toxins, known as LT (labile toxin), STa (stable toxin a) and STb (stable toxin b), are responsible for changes on the intestinal function that result in secretory diarrhea. Therefore, a case of edema disease may or may not be accompanied by diarrhea depending on the toxins produced by the causative strain.

Enterotoxigenic E. coli also are described by the type of surface fimbria that they possess. Historically, enterotoxigenic strains may express any of the four most common pilus types, e.g., K88, K99, 987p, F41. Recently, a newly described pilus type called FI8 has become increasingly common. According to a survey of isolates done at South Dakota State University (SDSU) on E. coli from porcine diarrhea cases during 1999‑2001, K88 was by far the most common (52 percent), followed by F18 (36 percent). The remaining isolates were 10 percent toxin only and 2 percent showed the gene for the K99 pilus.

Most, if not all, isolates that possess STX-2e toxin, are of the F18 group. This has been the case within our practice. According to David Francis, PhD, SDSU, K88 isolates also may have the edema disease toxin, but it is extremely rare.

Our clinic has made a practice of sending every significant hemolytic E. coli to SDSU for multiplex PCR. This test reveals what genes are present for the known virulence factors and pilus types. From October 2000 to present, we have categorized 78 strains of hemolytic E. coli taken from cases of porcine disease. Of these, 47 were K88 (60 percent), and 31 showed the gene for F18 (40 percent). Of the F18 isolates analyzed, 20 of 31, or 64.5 percent, also had the gene for the STX-2e toxin. Again, we did not find one isolate with the K88 pilus type that possessed the gene for making Shiga‑like toxin.

Pathogenesis/epidemiology
The pathogenesis of STX-2e producing E. coli involves colonization of the intestinal tract followed by proliferation. There is some disagreement as to the original source of the infection, but it has become evident that the problem is site-specific rather than pig-source specific. Expression of the toxin gene results in toxin production and systemic dissemination.

This toxin has been proposed to have two effects. First, it is a vasotoxin, capable of causing a microangiopathy. Leakage of fluid from the microvasculature leads to the tissue edema most notable in the mesocolon and the eyelids. It also has been reported that this toxin causes mean blood pressure to 200 mm Hg. At this pressure, hyperperfusion of brain tissue leads to edema and neurological signs.

In the field, pigs appear to start on feed well, and the first signs of a problem may be diarrhea at about eight days post-placement. Often, the first signs are large numbers of dead pigs on day 10 to 12 post-placement. This has been a remarkably consistent finding both in nursery and finish situations.

In cases where diarrhea is present, culture produces a pure growth of B‑hemolytic E. coli. Pigs with neurological signs appear from this point forward. These signs vary from ataxia to complete recumbency with lack of nystagmus. Some pigs in the early stages will go into a seizure upon excitation and return to standing after some period of rest.

Treatment of individual pigs is unrewarding. Mortality of affected pigs is very high (>90 percent). Morbidity and mortality in our experience has been 10 percent to 35 percent. The course of the disease is cited to be from four  to 14 days with an average of about eight days.

Our field experience would agree with this. However, occasionally “spinners” will show up for a few days to weeks thereafter. This is attributable to later exposure of some individual pigs and an 18- to 72‑hour delay between toxin absorption and clinical neurological signs.

CASE 1
Large system four-barn finishing site
I was called to a large system four‑barn-finishing site to help with continued death loss from “strep.” The field manager reported that despite oral penicillin and thorough individual treatment, death loss from pigs with CNS signs was getting to an alarming level.

Upon examination of several pigs, the CNS signs were slightly clinically different than those commonly observed with Streptococcus suis infection. Intermittent seizures and front leg paresis were among the strange presentations.

Submission of specimens to the Iowa State University Veterinary Diagnostic Laboratory (ISU-VDL) yielded hemolytic E. coli and lesions suggestive of edema disease. Death loss from this outbreak stopped at about 8 percent.

The very next turn through these barns, although no other pigs from this flow were affected, the situation deteriorated on day 10. This time profuse diarrhea was followed by the onset of the familiar CNS signs.

This isolate was resistant to nearly every antibiotic on test. Despite administration of large numbers of injections of ceftiofur, death loss attributable to E. coli surpassed the 700‑head mark on this one site. Interestingly, some pens originally stocked with 25 pigs would be down to only two or three pigs remaining.

In talking with the management, this was actually the third progressively worsening turn at this site. They were fearful that the next turn would continue to meet or exceed this level of mortality, and they would lose their contract with the pig company.

Communications with the company veterinarian led to formulation of a comprehensive plan including dietary manipulations (fiber), extreme sanitation measures and the inoculation of the incoming pigs with a live avirulent E. coli culture shown by multiplex PCR to contain the gene for F18 pilus but none of the other virulence factor genes. The resultant control was remarkable and death loss went back to baseline.

CASE 2
A 3,000-sow farrow-to-finish operation
A 3,000‑sow farrow‑to‑finish operation reported severe diarrhea and death loss in the newest of four nursery sites. The onset was day 9 to 11 post-weaning, and the pigs had not responded to any antibiotics. Cultures confirmed hemolytic E. coli. Death loss on this turn was 20.5 percent.

This was the second turn of pigs through this new contract nursery. After extreme measures were taken in cleaning and sanitation, all‑out pig flow and some other conventional management adjustments, the nursery was refilled. Again, on day 11, the formerly healthy‑appearing group began to scour and die.

This time death loss closed out at 33 percent. The next turn was treated with the toxin-negative culture for F18. Although some disease was noted, death loss was held to 5.95 percent. On the subsequent turn, also inoculated, death loss declined to 3.2 percent and very few deaths were related to E. coli. This nursery inoculated pigs for an additional two turns and then dropped the procedure. No signs of edema disease have been seen since early 2002.

Conclusions
In our practice, the use of toxin negative culture inoculation has been an important tool to help reduce the losses associated with edema disease producing Fl8 E. coli. Prior to this technique, the implementation of traditional intervention strategies had little benefit in the field against these organisms.

We have tried to develop a system to ensure that the culture delivered to the farm is pure. We also verify periodically that the culture has not lost the pilus gene or has obtained any virulence factors. The success of this technique has been so good, and the losses so substantial prior to its implementation, that many clients are reluctant to discontinue its use despite recommendations to do so.