1. U.S. Food and Drug Administration Notice: Archived Document The content in this document is provided on the FDA’s website for reference purposes only. It was current when produced, but is no longer maintained and may be outdated.

2. Pathogen Study - Design Considerations Jeffrey T. Gray USDA, ARS,ARRU Athens, GA

3. Study Design Considerations u Organism Characteristics Host-Range / Clinical Status u Study Design Challenge studies Natural inoculations u Measuring Effects u Confounding Factors

4. Most frequent Salmonella isolates u Swine choleraesuis-kC 1 typhimuriumB derbyB choleraesuisC 1 typhimurium-cB anatumE 1 heidelburgB brandenburgB worthingtonG 2 agonaB u Human typhimuriumB enteritidisD 1 heidelburgB newportC 2 hadarC 2 agonaB infantisC 1 thompsonC 1 braenderupC 1 javiannaD 1

5. Organism Characteristics u Salmonella Broad Host Range “ubiquitous” Clinical status questionable u Campylobacter Lack of clinical signs in food animals u E. coli O157:H7 Narrow host range No clinical signs in food animals

6. Organism Characteristics u Invasiveness Salmonella –“Shedding” –Carrier State Stress / Diet / Environment Strain differences among Salmonella isolates u Infectious Dose Shedding vs. Infectious dose

7. Study Design u Challenge Studies Challenge Dose Dose response curve Salmonella in swine 1 –10 3 cfu- no shedding, no deep tissue infection –10 6 cfu- 10 1 shedding peak, 8 weeks carriage –10 9 cfu- 10 3 shedding peak, long term carriage Controlled Natural Infections 2 –Seeder Animals 1. Gray et al AJVR Vol. 57, no 3, 1996 2. Gray et al. Appl. Environ. Microbiol. 62: 1996

8. Study Design u Controlled Natural Infections 10 3 shedding peak of seeder animals –All naïve animals infected –10 3 shedding peak in ‘naturally’ exposed animals –Long term carrier state u Swine infected with infected, desiccated feces 1 Dose = 10 5 cfu –All naïve animals infected / No shedding –10 3 /g deep tissue infections 1. Gray et al J. Food Prot. V 64 no 7, 2001.

9. Measuring Effects u Pathogen Detection Methods E. coli O157:H7 –1998-99’ Hancock et al. feedlot prevalence 5%-8% –2000-2001 Hancock et al. /Smith et al. feedlot prevalence 23-25% –Differences based on detection methods Are detection methods sensitive enough to detect treatment effects?

10. Measuring Effects u Temporal Measurements Salmonella high dose challenge shedding curve - cfu/g of feces At what point do we intervene?

11. Measuring Effects u Temporal Measurements Intervention points –Prior to infection – alterations in flora –During peak shedding - ? Clinical signs –Late in production cycle – Induction of carrier state?

12. Measuring Effects Measurement of effect –What measurements are used Method sensitivity Feces, Tissue What time point For how long –“No” Salmonella Not currently feasible Salmonella in many animals – clinical? Detectable in a few animals at low levels Likely the norm Food safety risk

13. Confounding Factors u Seasonal Changes E. coli O157:H7 –Seasonal prevalence 3% - 5% Salmonella –Species, serotype differences –Peak season prevalence 8-12% u Diet Composition of feedstuffs can effect prevalence and transmission of Foodborne zoonotic pathogens

14. Confounding Factors u ‘Stress’ Manner in which stress is measured Effect on pathogen shedding –Transportation effects Lairage exposure –Recent swine data indicate lairage may be important exposure point Deep tissue (lymph node) infections –Salmonella serotype differences

15. Confounding Factors u Environment Treatment effects may be environment specific Mechanism of action of treatment –Temporal relationship of shedding curve –Effect on flora or directly on pathogen Long term survival and infectivity of Salmonella –Years in desiccated state

16. Conclusions u Organism Characteristics Studies must adequetely account for specific characteristics of a foodborne zoonotic pathogen u Study Design Model a realistic infection cycle in target host species u Measuring Effects Must be robust enough to show differences under realistic circumstances u Confounding Factors Study must reasonably account for factors that exist in production systems and pathogen strain differences