![]() |
|||||||||||||||||
|
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
|||||||||||
|
|
![]() |
![]() Perspective Competitive Exclusion of Salmonella Enteritidis by Salmonella Gallinarum in PoultryWolfgang Rabsch,* Billy M. Hargis,† Renée M. Tsolis,† Robert A. Kingsley,† Karl-Heinz
Hinz,‡ Helmut Tschäpe,* and Andreas J. Bäumler†
The avian-adapted serovar Salmonella Gallinarum, which includes two biovars, Gallinarum and Pullorum, was endemic in poultry flocks in Europe and the Americas in the early 20th century (1). To reduce economic losses to the poultry industry, national surveillance programs were established in the United States (National Poultry Improvement Plan, 1935) and England and Wales (Poultry Stock Improvement Plan, 1939). Since S. Gallinarum (antigen formula O9,12:-:-) has no animal reservoir other than domestic and aquatic fowl, the test-and-slaughter method of disease control under these surveillance programs led to its eradication from commercial poultry flocks in the United States, England, and Wales by the 1970s (1,2). At that time, the number of human cases of infection with serovar S. Enteritidis (antigen formula O9,12:g,m:1,7) began to increase in these countries (3,4). By the 1980s, S. Enteritidis had emerged as a major concern for food safety in Europe and the Americas (5); by 1990 it was the most frequently reported Salmonella serovar in the United States (6). Most S. Enteritidis outbreaks in Europe and the United States are associated with foods containing undercooked eggs (7-10). Eggs can become contaminated with S. Enteritidis through cracks in the shell after contact with chicken feces or by transovarian infection (11). Thus, laying hens were the likely source of the S. Enteritidis epidemic in Europe and the Americas. The inverse relationship between the incidence of S. Gallinarum infection in chickens and egg-associated S. Enteritidis infections in humans prompted the hypothesis that S. Enteritidis filled the ecologic niche vacated by eradication of S. Gallinarum from domestic fowl (12). The hypothesis suggests that the epidemic increase in human S. Enteritidis cases in several geographic areas can be traced to the same origin, accounting for the simultaneous emergence of S. Enteritidis as a major egg-associated pathogen on three continents (5). A connection between the epidemics in Western Europe and the United States was not apparent from analysis of epidemic isolates. Although most human cases from England and Wales result from infection with S. Enteritidis phage type 4 (PT4), most cases in the United States are due to infections with PT8 and PT13a (13,14). The PT4 clone is genetically distinct from PT8 and 13a, as shown by IS200 profiling, ribotyping, and restriction length polymorphism of genomic DNA fragments separated by pulsed field gel electrophoresis (15). The reasons for the differing clonal isolates in the United States and Western Europe are unknown. S. Enteritidis was likely introduced into poultry flocks from its rodent reservoir (12). The geographic differences in predominant phage types may reflect the fact that at the time of introduction into poultry flocks, different S. Enteritidis strains were endemic in rodent populations in Europe and the United States. Subsequently, S. Enteritidis strains with the highest transmissibility may have become predominant in poultry flocks on each continent. An alternative explanation for the predominance of PT4 in England and Wales is its introduction into poultry breeding lines in the early 1980s (16), which may have accelerated the epidemic spread of PT4 in laying hens and resulted in its dominance in human isolates from England and Wales. However, factors responsible for the beginning of the S. Enteritidis epidemic should be considered separately from those important for its subsequent spread within the poultry industry. These factors were not specific to PT4 but rather allowed different phage types to emerge as egg-associated pathogens on different continents at the same time (5). One such factor could be the eradication of S. Gallinarum from poultry, which would facilitate circulation of S. Enteritidis strains within this animal reservoir regardless of phage type. Experimental evidence indicates that immunization with one Salmonella serovar can generate cross-immunity against a second serovar if both organisms have the same immunodominant O-antigen on their cell surface (17-19). The immunodominant epitope of the lipopolysaccharide of S. Gallinarum and S. Enteritidis is the O9-antigen, a tyvelose residue of the O-antigen repeat (20). Immunization of chickens with S. Gallinarum protects against colonization with S. Enteritidis (21,22) but not S. Typhimurium, a serovar expressing a different immunodominant determinant, the O4-antigen (23). Theory indicates that coexistence of S. Gallinarum and S. Enteritidis in an animal population prompts competition as a result of the shared immunodominant O9-antigen, which generates cross-immunity. Mathematical models predict that the most likely outcome of this competition between serovars is that the serovar with the higher transmission success will competitively exclude the other from the host population (24-26). S. Gallinarum may have generated populationwide immunity (flock immunity) against the O9-antigen at the beginning of the 20th century, thereby excluding S. Enteritidis strains from circulation in poultry flocks (12). This proposal is based on analysis of epidemiologic data from the United States, England, and Wales. To formally test this hypothesis, we analyzed epidemiologic data from Germany to determine whether the numbers of human S. Enteritidis cases are inversely related to those of S. Gallinarum cases reported in poultry. We used mathematical models to determine whether our hypothesis is consistent with theoretical considerations regarding transmissibility and flock immunity. Inverse relationship of S. Enteritidis and S. Gallinarum isolations in Germany
In West Germany, the number of human S. Enteritidis cases was monitored by a national surveillance program (Figure) (Zentrales Überwachungsprogram Salmonella, ZÜPSALM)from 1973 to 1982. In 1975, the number of human infections began to increase, indicating the beginning of the S. Enteritidis epidemic in West Germany. In 1983 the ZÜPSALM program was replaced by a national program for surveillance of foodborne disease outbreaks (Zentrale Erfassung von Ausbrüchen lebensmittelbedingter Infektionen, ZEVALI), implemented by the Department of Public Health (Bundesgesundheitsamt). In the first year of this program, S. Enteritidis was responsible for 62 outbreaks, most of which were traced to raw eggs. By 1988, the number of disease outbreaks caused by S. Enteritidis had increased to 1,365. In 1967 in England and Wales, poultry, particularly chickens, became the main human food source of S. Enteritidis (3). Before that date, the organism had only sporadically been isolated from poultry (3). A continuous increase in human S. Enteritidis cases was recorded from 1968 until the epidemic peaked in 1994 (12,16). Thus, the human S. Enteritidis epidemic in England and Wales probably began in 1968 after this organism became associated with a human food source, chickens. The rapid increase in the number of human cases from 1982 to 1988 was probably due to the introduction of PT4 into poultry breeding lines in England and Wales (16). Comparison of data from England and Wales (3,27) showed that S. Enteritidis emerged somewhat later in West Germany (Figure). Eradication of S. Gallinarum was among the factors contributing to the emergence of S. Enteritidis as a foodborne pathogen (12). To determine whether delayed elimination of avian-adapted Salmonella serovars from commercial flocks contributed to the late start of the human epidemic in Germany, we compared the results of surveys performed in poultry flocks in Germany with those from the United Kingdom and the United States. Control programs in the 1930s triggered a steady decline in the incidence of S. Gallinarum in poultry flocks in the United States, England, and Wales (1,2,12). By the early 70s, only a few cases of S. Gallinarum were reported each year to veterinary investigation centers in England and Wales (28). In Germany, the first national survey performed by the Department of Public Health (Reichsgesundheitsamt) in 1929 showed that 16.3% of birds were seropositive for S. Gallinarum (29). Blood-testing performed 20 years later with 6,313 birds in a province (Südbaden) of West Germany still detected 19.5% reactors (30). This high prevalence of S. Gallinarum in 1949 likely reflects the fact that after World War II available resources were directed toward rebuilding the poultry industry rather than improving disease control. The comparatively slow decline in the prevalence of S. Gallinarum in West Germany is illustrated further by data for cases of disease reported from poultry. The number of S. Gallinarum isolations from chicken carcasses received by veterinary laboratories in West Germany was reported by a surveillance program from 1963 to 1981 (31). During this period, the rate of decrease in numbers of S. Gallinarum cases in England and Wales was considerably higher than that reported from West Germany (Figure). In each country the numbers of S. Gallinarum cases were inversely related to the numbers of human S. Enteritidis cases. These data are consistent with the concept that the relative delay in eradicating S. Gallinarum from poultry may have contributed to delayed onset of the S. Enteritidis epidemic in West Germany. Competitive exclusion of S. Enteritidis by S. GallinarumTo calculate whether the prevalence of S. Gallinarum in chickens was high enough to generate flock immunity against S. Enteritidis, we analyzed epidemiologic data by mathematical models combining epidemiology with population biology (24-26). The transmission success of a pathogen is measured by the basic case-reproductive number, R0, which is defined as the average number of secondary cases of infection from a primary case in a susceptible host population (32). In direct transmission, the basic case-reproductive number of a pathogen is directly proportional to the duration, D, for which an infected host can transmit the disease before it is either killed or clear of infection; the probability, ß, by which the disease is transmitted from an infected animal to a susceptible host; and the density of susceptible hosts, X (24). |
![]()
|
![]() |
|
|||
![]() |
![]() |
Home | Top of Page | Current Issue | Expedited | Upcoming Issue | Past Issue | EID Search | Contact Us CDC Home | Search | Health Topics A-Z This page last reviewed May 26, 2001 Emerging Infectious Diseases Journal |