FDA logo -- links to the FDA home page
Center for Veterinary Medicine HHS logo -- links to the HHS home page

FDA Home Page | CVMHome Page | CVM A-Z Index | Contact CVM | SIte Map

horizonal rule

NOTE: See final version Guidance 80, November 21, 2002 | pdf | | doc |

Guideline No. 54

June 22 1994 DRAFT

DRAFT

Guideline for Utility Studies for Anti-salmonella Chemical Food Additives in Animal Feeds

Center for Veterinary Medicine
June. 1994
DOCKET 94D-0147

Table of Contents

April 22. 1994, DRAFT

Introduction

In September 1990, the Center for Veterinary Medicine (CVM or Center) announced a policy of zero salmonella in animal feeds (Reference 5). Zero salmonella in animal feeds may be achieved in several ways, including the addition of chemical substances to the feeds. Such substances are deemed to be unsafe food additives under section 409(a) of the Federal Food, Drug, and Cosmetic Act (the Act) (Reference 4), 21 SC 348(a), because none of them is currently approved or generally recognized as safe (GRAS) for use against salmonella present in feeds. According to the Act, a food additive is deemed to be unsafe unless it and its use or intended use conform to the terms of an exemption or existing regulation prescribing the conditions under which such additive may be safely used.

Companies or individuals (sponsors) intending to market an antimicrobial food additive for use against salmonella in feeds are thus required by section 409(b) of the Act to file a food additive petition to establish the safety of the food additive for such use. Regulations governing the filing of food additive petitions and setting out the conditions under which a food additive or food containing a food additive intended for investigational use shall be exempt from the requirements of section 409 of the Act can be found in title 21 of the Code of Federal Regulations (Reference 3) under part 571, section 1 (21 CFR 571.1), and part 570, section 17 (21 CFR 570.17), respectively. Briefly, a food additive petition must contain data establishing that the food additive is safe and will have the intended effect and identify the amount of additive necessary to accomplish this effect.

Our experience with food additive petitions indicates need for consistency in evaluating the ability of anti-salmonella food additives to achieve their intended effect of controlling salmonella in feeds. Thus, the objective of these guidelines is to help to assure that appropriate utility studies are conducted by sponsors and that uniform review and decision-making are accomplished by CVM. This would facilitate the approval process for such food additives.

General considerations

These guidelines pertain only to utility studies for chemical food additives intended for use against salmonella present in feeds. Our present policy is that such substances will be considered for approval only if they are intended for use in maintaining salmonella free feeds free of salmonella even after repeated exposure of the feeds to the microorganism. Experiments designed to establish the utility of an antimicrobial food additive for this purpose should be conducted in two main stages consisting of laboratory and field trials. These trials should be well planned and documented as protocols.

A protocol should contain a clearly expressed statement of purpose ,and details of the materials, methods, and statistical tools that will be used in the trial. The experiments should be designed to utilize finished feeds usually consumed by the proposed target animal(s), and/or ingredients used in manufacturing such feeds. Ingredients should be in compliance with definitions specified for such feeds by the Association of American Feed Control Officials (AAFCO). Finished feeds should be shown, by calculation and laboratory analysis, to meet the nutritional requirements established for the target animal(s) by the most current recommendations of the National Research Council of the United States National Academy of Sciences. The target animal(s) should be specified. For food additives containing two or more chemicals, the composition of the food additive and the amount of each chemical must be specified.

Sponsors are not required to submit protocols for review by CVM, and CVM review does not bind the Center to accept data generated using such protocols. All the same, sponsors are strongly encouraged to submit such protocols and do so well before the initiation of experiment. Submitting the protocols to CVM for review and comments before conducting the experiments generally facilitates the approval process for the food additive.

Data should be collected and recorded in ways that will enhance confidence in the integrity of data collected. Sponsors are referred to the Center's guideline entitled "Conduct of Clinical Investigations: Responsibilities of Clinical Investigators and Monitors for Investigational New Animal Drug Trials" for detailed information about record keeping and record retention. To facilitate the Center's review of statistical procedures, data used by the sponsor in statistical analyses should also be captured on a diskette, preferably in the form of an IBM compatable DOS ASCII file. The diskette should be included in the food additive petition.

Laboratory trials

General

The susceptibility of salmonella to insults has been shown to be dependent on several factors, including water activity, pH, temperature, oxygen content, type of medium, mode of contamination of medium, type of salmonella intensify of the insult, and the duration of contact between salmonella and the insult (Reference 11) .It is possible that manipulation of these factors could, in the absence of uniform testing methods, result in the overestimation of the anti-salmonella capabilities of food additives. To ensure that the data relied on by the FDA to approve anti-salmonella substances are not flawed because of these variables, the Center believes that the substances should be subjected to a uniform testing method.

The goal of the uniform testing method should be to determine, in a consistent manner, whether or not the anti-salmonella chemical food additives.are able to maintain salmonella-free feeds free of salmonella for a minimum of 14 days even if the feeds are repeatedly exposed to salmonella during the period. Sponsors should conduct two laboratory experiments: a "dose titration study" to determine the range of effective doses of the food additive, and a "prevention of recontamination" study to assess the ability of the minimum effective dose to maintain salmonella-free feeds free of salmonella despite repeated exposure of the feeds to the microorganism. The "prevention of recontamination" experiment should be duplicated in a different laboratory by a different investigator. The specific requirements for each of the two experiments follow: Dose titration (dose determination) study

The CVM considers the "dose titration" study to be pivotal. The main purpose of the study is to determine the range of doses over which the food additive is effective against salmonella in feeds. As much as it is possible, the range determined should include the minimum and maximum effective doses. Various concentrations of the food additive should be tested against a mixture of serotypes of salmonella inoculated into feed. The concentrations of food additive used should be multiples of the proposed effective dose. That dose should have been estimated earlier by the sponsor in preliminary studies using standard laboratory techniques. A minimum of six concentrations of the food additive in feed is recommended. One should be at zero, another at the proposed effective dose, and two each below and above the proposed effective dose (e.g. 0, 0.25X, 0.50X, 2.00X, and 4.00X; where X is the proposed effective dose). When using data from the experiments to plot the dose response curve, the dependent variable would be the concentration of salmonella in the sample of feed while the independent variable would be the concentration of food additive.

The feed used should be sterilized and efficacy of the techniques utilized in sterilizing assessed by determining the microbiological profile of representative samples collected from the feed before and after sterilization. Components of that profile should include:

The sterilized feed should be divided into a minimum of seven lots, including one control and one negative control. Each lot should be separately treated with an appropriate amount of the food additive and then subdivided into twelve or more units. The units should be randomly assigned to one of four groups containing equal number of units. There should be three or more units per group. Portions of each unit should be assayed for the food additive and the remainder inoculated with salmonella. Throughout the experiment, both the control containing salmonella but not the food additive and the negative control containing neither salmonella nor the food additive, should be handeled exactly like the other experimental lots of feed.

The method used in inoculating experimental feeds with salmonella can influence the outcome of the dose titration study. Three methods are currently available. Each involves inoculation with one of the following:

Contamination or recontamination of feeds with salmonella can occur through several avenues including infected or contaminated wild birds, rats, mice, and cockroaches (Reference 11) and it appears logical to utilize naturally contaminated materials (e.g. feces) from these sources as inoculi for experimental feeds. However, the use of this natural method of contamination would be incompatible with the goal of uniform testing, because the concentration of salmonella in such materials varies widely. Broth cultures can be manipulated to provide inoculi containing fairly uniform amounts of salmonella and their use in inoculation will significantly reduce the. variability in salmonella concentration associated with the natural method. However, the direct contamination of experimental feeds with such inoculi may result in an overestimation of the ability of the food additive to achieve its intended effect because salmonella present in broth culture-inoculated feeds has been shown to be less resistant to insults than that present in naturally contaminated feeds (Reference 10). The use of a method involving inoculation with simulated naturally-contaminated materials will overcome, to a large extent, the deficiencies (variability in concentration and reduction in resistance of salmonella) associated with the natural and broth culture methods. Therefore, this method involving the inoculation of feeds with simulated naturally-contaminated materials is recommended.

A technique for producing simulated naturally-contaminated meat and bone meal containing stable populations of salmonella has been described by Liu and his co-workers (Reference 9), who also showed that the salmonella present in feeds inoculated with the meal was more resistant to insults than that present in feeds directly inoculated with broth cultures of salmonella. Briefly, in the Liu et al technique, a sterile suspension of meat and bone meal inoculated with salmonella was incubated for 48 hr at 37 ° C. After incubation, the suspension was centrifuged and the sediment dried, ground to a fine powder and stored at 4 ° C pending its use as the inoculum for experimental feeds. The population of salmonella in the meal was reported to remain relatively stable over an 11-month period.

To produce simulated naturally-contaminated materials for use as inoculi for experimental feeds in the dose titration study, sponsors should utilize the Liu et al technique or any other similar one that is shown to equally or more effectively ensure the stability of salmonella even during cold storage.

One gram of inoculum should be deposited on a spot inside, and another on a spot on top of, each unit without mixing. The inoculum should contain equal amounts of ATCC cultures of Salmonella typhimurium, Salmonella senftenberg, Salmonella montevideo, and Salmonella enteritidis, S. typhimurium is being considered because of its universal nature, S. senftenberg. because it has been reported to be one of the more resistant serotypes (Reference 9), S. montivideo because it is reported to be the serotype most frequently isolated from feeds and is said to have a relatively high resistance (Reference 8), and S. enteritidis because it is a major public health concern. Other salmonella serotypes that are of animal or public health significance could be included by the sponsor, if desired.

Each inoculum should contain 10^8 colony-forming units (cfu) of salmonella. It has been reported that the concentration of salmonella in the feces of naturally infected mice can be as high as 2.3 x 10^5 cfu/fecal pellet or 2.6 x 10^7 cfu/g of feces (Reference 6).

Because the antimicrobial activity of organic acids and other chemicals appear to be time-dependent, the design of the dose titration study should permit elucidation of the nature of that time-dependency when the chemicals are used as antimicrobial food additives. The preferred method for achieving this is the generation of a series of dose titration curves using data obtained from analyses of the experimental units of feed at 1, 8, 24 and 48 hours after inoculation of the feeds with salmonella. The four groups of feed should be used for this purpose. The amount of inoculated feed constituting individual units in each group should be enough to ensure its adequacy for quantitative and qualitative analyses for salmonella. The number of samples to be collected from each unit for salmonella analysis should be determined by a statistical model that instills a 99% confidence that the two inoculi in the unit will be captured among the samples collected. The analysis of all of the inoculated feed in each unit would be compatible with this model.

Temperature, pH, water activity and other pertinent experimental conditions should be specified. The experiment should be repeated one or more times.Study of ability to maintain salmonella-free feed free of salmonella.

The purpose of the second laboratory experiment is to demonstrate the ability of the food additive to maintain salmonella-free feeds free of salmonella for a minimum of 14 days, even if the feeds are repeatedly exposed to salmonella during the period. !n this study, the level of food additive determined in the dose titration study to be the minimum effective dose should be added to sterilized feed, and challenged with salmonella contained in a simulated naturally-contaminated material as described earlier under the dose titration study. The challenge with salmonella should be conducted at the beginning of the experiment and again at day 14. A minimum of six representative samples each should be collected from the treated feed and its controls at two time intervals after each challenge with salmonella and analyzed qualitatively for salmonella. The first time interval should be that at which the minimum dose of the food additive was shown to be effective in the dose titration study, and the second one 24 hours afterwards. The quantity of feed used in the study should be enough to ensure its adequacy for the analysis for salmonella. All of the feed should be used in analysis. As with the dose titration study, temperature, pH, water activity, and other pertinent experimental conditions should be specified. The experiment should also have proper controls including a negative blank group and contain an adequate number of replicates. The duration of the experiment and the number of challenges with salmonella should be increased if a sponsor intends to claim a residual effect for an anti-salmonella food additive that is longer than 14 days. As stated earlier, this experiment should be duplicated in a different laboratory by a different investigator.

The feeds used in the laboratory trials should consist of specified quantities of compound feeds (e.g. broiler starter) and/or feeds intended for use in manufacturing compound feeds (e.g. meat and bone meal). The feed used should either be obtained commercially or self-produced. Details of the composition of compound feeds (including all added drugs and additives) and the techniques used for their manufacturing or processing should be specified. The amount of feed assigned to each experimental group should be enough to permit the collection of all required samples.

Field trials

The dose of food additive determined during the dose titration study to be the appropriate effective dose should be further evaluated in field trials. The purpose of a field trial is to determine if the food additive could be safely used to achieve its technical effect(s) under actual conditions of use. Such conditions of use would include those associated with the preparation, storage and consumption of food additive-treated feeds. The quantity of feeds used in the field trials and the conditions under which they are manufactured/produced, stored, transported, and consumed should be the same as those expected when the food additive is approved for use. Details of the composition of the feeds should be specified and include all added drugs, GRAS substances, and food additives. Samples should be collected from feeds at specified locations and analyzed for salmonella. For feeds like meat and bone meal, which are intended for use in manufacturing compound feeds, samples should be collected immediately before and after the addition of food additive and also immediately prior to usage of the food additive-treated feeds in compounding. For compound feeds like broiler or hog starter rations, samples should be collected immediately before and after the addition of the food additive, and also just before and during the entire period the food additive-treated feeds are made available for consumption by the target animals. The number of samples to be collected at each location should be determined by a statistical model that instills a 99% confidence that salmonella will be captured among samples collected even if the concentration of the microorganism in the feed was as low as 1 cfu/500 g. Additional criteria might need to be satisfied in field investigations involving anti-salmonella chemical food additives intended for use in the feeds of aquatic animals. Sponsors of such food additives are referred to the Center's "Guideline for the evaluation of the Utility of Food Additives in diets fed to Aquatic Animals."

Because food additives are approved for use on a nationwide basis, trials should be conducted in at least three geographically different locations and during two different seasons of the year, one of which should be the summer. Additional trials could be conducted during one or both of the other seasons of the year if so desired by the sponsor.

Collection and analysis of samples

The sampling techniques employed in this type of study are critical to the validity of results obtained. Samples should be aseptically collected without unduly disturbing the experimental feeds or target animals consuming the feeds. FDA's Bacteriological Analytical Manual (Reference 2) contains pertinent information about sample handling.

The time of sample collection should be recorded and scheduled so that storage of samples prior to analysis is avoided as much as possible. If it is necessary to store the samples, the conditions under which the samples are stored (e.g., temperature) should be specified and shown not to be inimical to either the survival of salmonella or the maintenance of the integrity of nutrients. The duration of storage should also be specified.

Laboratory analysis for salmonella should be initiated and the first stages completed preferably within one hour after sample collection. If this is not feasible, the activity of food additive in samples used for salmonella analysis should be arrested immediately after sample collection using a compound/technique that has been shown not to have any negative impact on the survival of salmonella. The length of time between the collection of samples and their analysis for salmonella should be specified.

Laboratory analyses should be conducted "blind." It is recommended that the experimental feeds should be prepared and coded by one person or team and samples collected from the feeds analyzed by a different person or team unaware of the concentration of the food additive in the samples. If it is necessary to transport samples, they should be packaged to ensure the maintenance of their microbiological integrity during transportation. and sent by the fastest means possible.

The concentration of salmonella present in feeds used in the dose titration study should be determined by standard dilution and pour-plate technique. In addition, because the sensitivity of this quantitative technique is low when used in isolating salmonella from feeds containing injured and/or very small numbers of salmonella, the technique should be supplemented by a qualitative procedure that includes preenrichment-enrichment. Samples of feed collected for salmonella analysis during field trials should be analyzed qualitatively to determine the presence or absence of the microorganism. The use of the quantitative and qualitative methods described in the FDA's Bacteriological Analytical Manual (Reference 1,7) is recommended. However, sponsors can use other analytical techniques provided those techniques are demonstrated to be at least as sensitive and specific for salmonella.

References

  1. Andrews, W. H.. V. R. Bruce. G. June, F. Satchell, and P. Sherrod. salmonella. FDA Bacteriological Analytical Manual, 7th ed.. p. 51 - 69, 1992. U.S. Food and Drug Administration, Washington, D.C.

  2. Andrews, W. H., G. June and V. R. Bruce. Food sampling plans and preparation of sample homogenates. FDA Bacteriological Analytical Manual, 7th ed.., p. 1 -9, !992. U.S. Food and Drug Administration, Washington, D.C.

  3. Code of Federal Regulations, title 21, April 1, 1993. Office of the Federal Register, U.S. National Archives and Records Administration, Washington, D.C.

  4. Federal Food, Drug, and Cosmetic Act, as amended, October 1989. Title 21 of the United States Code. The Superintendent of Documents, Washington, D.C.

  5. Graber,George. Sept, 1990. Remarks on salmonella in Animal Feeds. Symposium on Feed Quality Assurance. Arlington, Virginia

  6. Henzler, D.J., and H. M. Opitz. The role of mice in the epizootiology of salmonella enteritiisis infection on chicken layer farms. Avian Diseases 36:625-631, 1992

  7. Hitchins, F., P. Feng, W. D. Watkins, R. Rippey, and L. A. Changler. Escherichia coli and the coliform bacteria. FDA Bacteriological Analytical Manual, 7th ed., p.27-35,1992 U.S. Food and Drug Administration, Washington D. C.

  8. Juven, B. J., N. A. Cox, J. S. Bailey, J. E. Thomson. O. W. Charles, and J. V. Shutze, Survival of salmonella in dry food and feed. Journal of Food Protection 47:445-448, 1984

  9. Liu, T. S., G. H. Snoeyenbos, and V. L. Carlson. Thermal resistance of salmonella senftenberg 775W in dry animal feeds. Avian Diseases 13: 611-613, 1969

  10. Rasmussen, O. G., R. Hansen, N. J. Jacobs and O. H. M. Wilder. Dry heat resistance of salmonellae in rendered animal by-products. Poultry Science 43:1151-1157, 1964

  11. Williams, A. E. Paratyphoid infections. In: Deaseases of Poultry, 8th ed. Hofstad, M. S., H. J. Barnes, B. W. Calnek, W. M. Reid, and H. W. Yoder, Jr., eds. p. 91-129, 1984. Iowa State University Press, Ames Iowa, USA

horizonal rule

CVM Home Page | CVM A-Z Index | Contact CVM | SIte Map
FDA Home Page | Search FDA Site | FDA A-Z Index | Contact FDA | HHS Home Page

FDA/Center for Veterinary Medicine