FDA Notion of Opportunity for Hearing re tetracyclines, 1977

NOTICES

[4110-03]DEPARTMENT OF HEALTH,

EDUCATION, AND WELFAREFood and Drug Administration

[Docket No. 77N-03161

PFIZER, INC., ET AL

Tetracycline (Chlortetracycline and Oxy-tetracycline)-Containing Premixes; Op-portunity for Hearing

AGENCY: Food and Drug Administra-tion.

ACTION: Notice.SUMMARY: This is a notice of opportu-nity for a hearing on the proposal by theDirector of the Bureau of VeterinaryMedicine to withdraw approval of newanimal drug applications (NADA's) fortetracycline (chlortetracycline and oxy-tetracycline) -containing premixes -in-tended for certain uses in animal feedon the grounds that (1) new evidenceshows that the tetracycline-containingproducts have not been shown to be safefor widespread subtherapentie use as re-quired by section 512(e) (1) (B) of the'Federal Food, Drug, and Cosmetic Act(21 U.S.C. 360b(e) (1) (B)) and § 558.15(21 CFR 558.15); (2) certain applicantshave failed to establish and maintainrecords and make reports as required bysection 512(e) (2) (A) of the act (21U.S.C. 360b(e) (2(A)) and § 558.15; and(3) new evidence shows that there Is -alack of substantial evidence that tetra-cycline-containing premixes are effectivefor certain subtreapeutic uses under sec-tion 512(e) (1) (C) of the act (21 U.S.C.360b(e) (1) (C)). "

DATES: Written appearances requestinga hearing must be submitted by Novem-ber 21, 1977; data and analysis uponwhich a request for a hearing repliesmust be submitted by January 19, 1978.

ADDRESS: Written appearances anddata and analysis to the Hearing Clerk(HFC-2O), Food and Drug Administra-tion, Rm. 4-65, 5600 Fishers Lane, Rock-ville, Md. 20857.

FOR FURTHER INFORMATION CON-TACT:

Gerald B. Guest, Bureau of VeterinaryMedicine (HFV-100), Food and DrugAdministration, Department of Health,Education, and Welfare, 5600 FishersLane, Rockville, Md. 20857, 301-443-4313.

SUPPLEMENTARY INFORMATION:

RELATED ACTIONS

In a notice published elsewhere in thisissue of the FEDERAL REGISTER, the Direc-tor of the Bureau of Veterinary Medi-cine is proposing to delete certain pro-visions that provide for the subtherapeu-tic use of tetracycline (chlortetracyclileand oxytetracycline) in animal feeds byamending § 510.515, Animal feeds bear-ing or containing new animal drugs sub-ject to the provisions of section 512(n)of the act (21 CFR 510.515); § 558.15iAntibiotic, nitrofuran, and sulfonamide

drugs in the feed of animals (21 CFR558.15); § 558.55, Amprolium (21 CFR558.55); § 558.58, Amprolium and etho-p'abate (21 CFR 558.58); § 558.105, Bu-quinolate (21 CFP. 558.105) ; § 558.128,Chlortetracycline (21 CFR 558.128);§ 558.145, Chlortetracycllne, procainepenicillin, and sulfamethazine (21 CFR558.145); § 558.155, Chloretracycline,procaine pencillin, and sulfathiazole (21CFR 558.155); § 558.175, Clopidol (21CFR. 558.175); § 558.195, Decoquinate(21 CFR 558.195); § 558.225, Diethylstil-bestrol (21 CFR 558.225); § 558.274, Hy-gromycin B (21 CFR 558.274); § 558.450,Oxytetracycline (21 CFR 558.450);§ 558.515, Robenidine hydrochloride (21CFR 558.515); and § 558.680, Zoalene(21 CFR 558.680).

DISCUSSION

Since the Director's discussion of theissues involved in this matter is neces-sarily detailed, he is setting forth, forthe reader's -convenience, an outline ofthe discussion as follows:

I. THE DUGSII. INTODUCTION

A. Regulatory Background.B. Safety Concerns.

m. SUMMARY OF THE AnGEuM=4

IV. STUDES RELEVAT TO HUMAN AND ANIMALHEALTH SFETY CRRIA

A. Transfer of Drug Resistance (Criterion1); The Pool of R-Plasmid-Bearlng Orga-nisms Is Increasing.-

1. Background.2. Criterion.3. Studies relevant to transfer of Drug

Resistance.(a) R-plasmid-bearing E. coli develop in

domestic animals that are fed subtherapeu-tic levels of antbiotics, including tetracy-cline.

(b) E. coli contribute their R-plasmids toman through several mechanisms.

(I) Direct contact with animals.(ii) Contact with E. cofl-contaminated

food.(Ill) Widespread presence In the environ-

ment.(c) R-plasmid-bearing human and animal

strains of bacteria overlap.(1) Epidemlological investigations--E. coli

serotyping.(ii) Direct ingestion evidence.(ill) In vivo studies show that R-plasmids

transfer from E. coli to pathogens.(iv) R-plasmid compatibility studies.(v) Hazards.4. Director's conclusions.B. Shedding and Resistance Characteristics

of Salmonella (Criterion 2).1. Background.2. Criterion.(a) Shedding.(b) Resistance characteristics.3. Industry studies in chickens on the ef-

fects of subtherapeutic tetracycline use Inanimal feed.

(a) American Cyanamid Co.* (I) Experimental design.

(i) Summary.(iii) Director's analysis.(b) Rachelle Laboratorias. Inc.(i) Experimental design.(it) Summary and the Director's analysis.(c) Pfizer, Inc.(1) Experimental design.(it) Summary.(iII) Director's analysis.

(d) Director's conclusions.4. Industry studies in swine on the offcota

of subtherapoutio tetracycline use in animalfeed.

(a) Tetracycline alono.(1) Experimental design.(ii) Summary and the Director's analyals.(b) TetracyCline in combination with sul-

fonamides and penicillin.(i) Experimental design.(11) Summary and the Director's analysis,(c) Director's conclusions.

5. Industry studies In cattle on the effectsof subthernpeutc tetracycline we in animaifeed.

(a) Studies of tetracyolino and tetracyclinecombinations in cattle and calves.

(I) Experimental design.(ii) Summary.(li) Director's analysis.(b) Director's conclusions.

6. Information from other studles relatingto Salmonella and E. colt antibiotic resist-ance.

(a) ,Surveys.(i) Neu,Chorubin, Longo, Fiouton, and

Winter studies.(ii) CDC reports.(ill) Amorican Cyanamid survey.(iv) Other surveys-of Salmonella resist-

ance.(b) Feeding studies.(I) Chickens.(it) Swine.(ill) Cattle.(c) Director's analysis.7. Director's conclusions,C. Compromise of Therapy (Criterion

2(r)).1. Background and criterion.2 Questions raised by FDA-funded re-

search and literature studies.(a) Experimental design.(b) Director's analysis.

3. Compromise of therapy studies Inchickens.

(a) Pfizer study.(b) American Cyanamid study.4. Compromise of therapy studies in Owine.(a) Diamond Shamrock Study No. 1,(b) Diamond Shamrock Study No, 2.(c) Pfizer study.(d) American Cyanamid study.5. Compromise of therapy study in cattle,(a) Diamond Shamrock study.(b) Pfizer study.(c) American Cyanamid study.

6. Director's conclusions.7. Optimal level of effoctivenes (Animal

Health Criterion 4).D. Pathogenicity (Criterion 3).1. Background and Criterion,2. Walton study.3. Falkow study.(a) In vitro transfer.(b) In vivo transfer.4. Questions raised by other studies.5. Director's conclusions.E. Tissue Residues (Criterion 4).1. The criterion.2. Background.3. American Cyanamid study.(a) Experimental design.(b) Summary.(c) Director's analysis.4. Literature survey.5. Director's conclusions.,

V. Er rcTvcNrSS

A. Oxytetracycline.B. Chlortetracycline.1. Roche Premixes.2. American Cyanamid and NaPpo pro-

mixes.3. American Cyanamid's chlortetracyoline

and vitamin products.4. Ralston Purina premix.C. Director's conclusions.

FEDERAL REGISTER, VOL. 42, NO. 204-FRIDAY, OCTOBER 21, 1977

HeinOnline -- 42 Fed. Reg. 56264 1977

Case 1:11-cv-03562-THK Document 33-2 Filed 10/06/11 Page 2 of 27

NOTICES

VI. CONCLUSION

L THE DRuGSThe generic names are chlortetracy-

cline as chlortetracycline hydrochloride,and oxytetracycline as the mono-alkyl-trimethyl-ammonium salt.

The dosage form is feed premix.The following companies hold or have

effective approvals for premixes whichcontain chlortetracycline or oxytetracy-cline and are subject to the provisionsof this notice:I.AD1-8-696; TM-5 Antibiotic Feed Supple-

ment (Oxytetracycline), Pfizer, Inc., 235M. 42d St., New York, N.Y. 10017.

NADA a-804; 7M-0; Terramycin AnimalMix Terramix-10 (oxytetracycline), Pfizer,Inc.

NADA 9-770; Stilbestrol-Oxytet Premix (di-ethylstilbestrol and oxytetracycline),Pfizer, Inc.

IT-ADA 11-661; ran-Q Plus Terramycin Pre-mix (oxytetracycline and hydroyzine hy-drocbloride), Pfizer, Inc.

NADA 13-470; TM-l0 Premix (oxytetracy-clme), Pfizer, Inc.

NADA 35-017; DES Premix (diethylstilbes-trol and chlortetracycline), Thompson-Hayward Chemical Co. P.O. Box 2383,Kansas- City, Mans. 66110.

NADA 35-4688; AUREO SP-250 (chlortetracy-cline, sulfamethazine, penicillin), Ameri-can Cyanamid Co., P.O. Box 400, Princeton,NJ. 08540.

NADA 36-361; AIPROL PLUS T CTC(amprollum, ethopabate, chlortetracy-cine), American Cyanamid Co.

NADA 36-554; Custom Beef Premix No. 6(diethylstilbestrol and oxytetracycline),Dale Alley Co., P.O. Box 444, 222 SylvanieSt., St. Joseph, Mo. 64502.

NADA 37-541; Falstaff Beef Fortifier B (di-ethylstilbestrol and chlortetracycline), Na-tional Oats Co., East St. Louis, Mo. 62205.

NADA 38-509; Vitality reedlot Premix (dl-et-hylstllbestrol and chlortetracycline),Texas Nutrition & Service Co., Fort Worth,Tex. 76108.

NADA- 39-077; CSP-250 (chlortetracycline,aulfathiazole, penicillin), Diamond Sham-rock Chemical Corp., Nutrition & AnimalHealth Div., 1100 Superior Ave., Cleveland,Ohio 44.114.

2NADA 44-795; Custom Beef Fortifier B (dl-ethylstilbestrol and chlortetracycline), Fal-staff Brewing Corp.

NADA 46-699; Nopco CTC 4/SS (chlortetra-cycline, sodium sulfate), Diamond Sham-rock. Chemical Co.; Nopoco CTC 6.66/SS(chlortetracycline, sodium sulfate), Dia-mond Shamrock Chemical Co.; Nopco CTC10, 25, 50, 100 (chlortetracycline), DiamondShamrock Chemical Co.

NADA 48-760; Deravet (chlortetracycline),American Cyanamid Co.

NADA 48-761; Aueromycin Feed Premixes(chlortetracycline), American CyananidCo.

N ADA 48-762; Aureomycin Crumbles with-Vitamins (chlortetracycline), AmericanCyanamid Co.

NIADA 48-763; Aureomycin Premix (chlortet-racycline), American Cyanamid Co.

ITADA 49-181; Spence Special Swine Premix;ARM-LA Special Swine Premix (chlortet-racycline), Hoffman-La Roche, Inc.. Nut-ley, N.J. 07110.

-NADA 49-287; CTC Premix (chlortetracyc-line, Rachelle Laboratories, Inc., 700Henry Ford Ave., P.O. Box 2029, Long-Beach, Calif. 90801.

NADA 65-005; Mortet 10; Klortet 50 (chlor-tetracycline),, Dawes Laboratories, Inc., 450State St., Chicago Heights, IMI. 60411.

NADA 65-20 Micro CTC 10a (chIortetra-eyeline), Diamond-Shamiock Chemical Co.

NADA 85-052; NOPCO CTC-60 (chlortetra-cycline), Diamond Shamrock Chemlal Co.

NADA 65-338; G"O Feed Grade (eblortetra-cycline), Cortex Chemicals S.P.A.

NADA 91-608: Chorachel 250; Super Chlc-rachel 250 (chloretracycline. sulfameth-azine, penicilln), nachello Laboratorlc3,Inc.

DESI 0-035; Purina Aurcomycln LtG Medl-cated (chlortetracycline), Ralston PurinaCo., Checerboard Square, St. LouL,, Mo.63188.Under section 108(b) 42) of the Animal

Drug Amendments of 1968 (Pub. I. 90-399), any approval of a new animal druggranted prior to the effective date of theamendments, whether through approvalof a new drug application, master file,antibiotic regulation, or food additiveregulation continues in effect until with-drawn in accordance with the provisionsof section 512 of the Federal Food, Drug,and Cosmetic Act (21 U.S.C. 360b). Manysuch approvals were Issued long ago, andsome may never have been used by theholder of the approval. Consequently, thecurrent files of the Food and Drug Ad-ministration (FDA) may be incompleteand may fail to reflect the existence ofsome approvals. Also, many approvalshave been withdrawn by other agencyactions, e.g., FDA's rule making proce-dure published in the FEDERAL REGSTERof February 25, 197G (41 FR 8282). Theburden of coming forward with docu-mentation of unrecorded approvals insuch circumstances is therefore properlyplaced on the person claiming to holdsuch approvals so as to permit definitiverevocation or amendment of the rezula-tions.

Tie Director of the Bureau of Veteri-nary Medicine Imows of no approvalsaffected by this notice other than thosenamed herein. Any person who intendsto assert or rely on such an approvalthat is not listed in this notice shallsubmit proof of Its existence within theperiod allowed by this notice for oppor-tunity to request a hearing. The failureof any person holding such an approvalto submit proof of its existence withinthat period shall constitute a waiver ofany right to assert ok rely on It. In theevent that proof of the exitenee of suchan approval is presented, this" noticeshall also constitute a notice of oppor-tunity for hearing with respect to thatapproval, based on the same grounds setforth In this notice.

31. Irrrnoaucrxoi;

A. REGULATORY oACIGROUMDAntibacterial drugs have been used at

subtherapeutic levels (lower levels thantherapeutic levels needed to cure dis-ease) in. animal feed for over 25 years.Growth benefits from this use were firstobserved when animals were fed the dis-card products from the fermentationprocess that was originally used in themanufacture of chloretetracycline. Theprecise mechanism of action, however,remains unclear.

Initially, certain antibiotics for use inanimal feed, eg., chlortetracycline, wereregulated under the provisions of section507 of the Federal Food, Drug, and Cos-

56265

metic Act (21 US.C. 357). Unlike thebasic private licensing system applicableto new drugs, the provisions of section50T of the act created a public regula-tion or monograph system for regulat-Ing these products, in part because ofthe complexities In manufacturing theproducts and the lack- of knowledge oftheir chemical structure. AntibioticresIdues In food from food-producinganimals were then regulated under theproviion5 of the act dealing with adul-teration and misbranding. After enact-ment of the Food Additives Amendmentof 1958 (Pub. L. 85-929), however, resi-dues were principally regulated by sec-tion 409 of the act (21 US.C. 348),which also established a public mona-graph system of premarket approvalUnder the antibiotic monograph proce-dure, the pioneer manufacturer gener-ated and submitted the basic safety andeffectiveness data In an FD Form 5 (now

D-1675). A regulation was subsequentlypublished setting forth the standards ofidentity, strength, quality, and purity,and the packaging and labeling require-ments that the product must meet. TheFood and Drug Administration approvalof the same product made by anothermanufacturer was then conditionedrolely upon a demonstration that it metthe requirements of the regulation, endthis is normally accomplished by batchcertification. Section 507(c) of the acto21 U.S.C. 357(c)), however, permits theagency to exempt by regulation any. drugor class of drugs from the certificationrequirement when he concludes that cer-tification is unnecessary for the manu-facture of the drugs. Antibiotics for useIn animal feeds as feed ingredients wereExempted from the certification require-ments In 1951 (,ee the FEzDRAL, PRxns=of April 28, 1951 (16 FR 3647)), andthose for use as drugs were exemptedin1953 (see the FtnrnAL Rrcsr of April22. 1953 (18 FR 2335)). These are nowset forth In 1 510.510 and 510.515 (21CIr 510.51(1 and 510.515).

Congress enacted thL Animal DrugAmendments of 1968 (Pub. L, 90-399)and consolidated the provisons of theact dealing with the premarket approvalof drugs Intended for use in animals(sections 409, 505, and 507) into one newsection, 512 (21 U.S.C. 360b), to regulatethese articles more efficiently and effec-tively (Senate Committee on Labor andPublic Welfare, Animal Drug Amend-ments of 1968, S. Rep. No. 1308, 90thCong., 2d Sess. 41968)). This legislationalso brought the manufacture of anti-biotics under the private license systemfor new drugs (Id: Hearing on S. 160and H.R. 3639 before the Subcommitteeon Health of the Senate Committee onLabor and Public Welfare. 90th Cong.2d Sess. $Q968)). To efficiently accom-plish this change, the amendments con-tained a transition clause (section108(b)) which provided that all priorapprovals continue in effect and be sub-ject to change In accordance with theprovislons of the basic act as amended.In summary, all perons- legally market-ing antibiotics under the provhfons ofsections 409. 505, and 507 of that act on

FEDERAL REGISTER, VOL 42, NO. 204--FRIDAY, OCTOBER 21, 1977



NOTICES

August 1, 1969, the effective date of theAnimal Drug Amendments of 1968, wereconsidered as holding the equivalent ofan approved new animal drug applica-tion; however, all holders of such ap-provals are also subject to all applicablerequirements of the act and regulations.

B. SAFETY CONCERNS

In the mid-1960's, FDA became con-cerned about the safety to man and ani-mals of subtherapeutic antibiotic use; itstudied the effects of low-level subthera-peutic feeding of antibiotics for someyears. The agency supported research,held symposia, and consulted with out-side experts to review these nonmedicaluses of antibiotics in animal feeds. Fol-lowing a report issued by the BritishGovernment Joint Committee (theSwann Committee) "On the Use of An-tibiotics in Animal Husbandry and Vet-erinary Medicine," the Commissioner ofFood and Drugs in April 1970 establisheda task force of scientists, with consplt-ants from government, universities, andIndustry, to review comprehensively theuse of antibiotic drugs in animal feeds.Its conclusions were published in a no-tice of proposed rulemaking publishedIn the FEDERAL REGISTER of February 1,1972 (37 FR 2444), -which initiated themandatory testing procedure to resolveconclusively the issues of safety sur-rounding the subtherapeutic use of an-tibiotics in animal feeds.

The principal conclusions of the taskforce were the following:

(1) The use of antibiotics and sulfon-amide drugs, especially in growth pro-motant and subtherapeutic amounts, fa-vors the selection and development ofsingle and multiple antibiotic-resistantand R-plasmid-bearing bacteria.

(2) Animals that have received eithersubtherapeutic and/or therapeuticamounts of antibiotic and sulfonamidedrugs in feeds may serve as a reservoirof antibiotic-resistant pathogens andnonpathogens. These reservoirs of path-ogens can produce human infections.

(3) The prevalance of multiresistantR-plasmid-bearing pathogenic and non-pathogenic bacteria in animals has in-creased and has been related to the useof antibiotics and sulfonamide drugs.

(4) Organisms resistant to antibae-terial agents have been found on meatand meat products.

(5) There has been an increase in theprevalence of antibiotic- and sulfona-mide-resistant bacteria in man.

In its report to the Commissioner, thetask force also identified three areas ofprimary concern: human health haz-ards, animal halth hazards, and anti-biotic effectiveness; guidelines were es-tablished to show whether use of anyantibiotic or antibacterial agent in ani-mal feed presents a hazard to human andanimal health.

The February 1972 proposal also an-nounced that all currently approved sub-therapeutic uses of antibiotics, nitro-furans, and sulfonamides in animal feedswould be revoked unless data were sub-mitted to resolve conclusively the issuesconcerning safety to man and animals in

accordance with thp task force guide-lines. That notice also proposed to es-tablish a time table for filing commit-ments, conducting studies, and submit-ting relevant data and information.Based on the guidelines, the agency thenbegan developing specific criteria bywhich the safety and effectiveness ofeach antibiotic product might be estab-lished. The notice further suggested thatprotocols be submitted to the agency forcomment. The criteria and studies toaddress them may be summarized asfollows:HUMIAN AND ANimAL HEALTH SAFETY CRTERIA

1. Transfer of drug resistance: (a) Anantibacterial drug fed at subtherapeuticlevels to animals must be shown not to pro-mote increased resistance to antibacterialsused in human medicine. Specifically, in-creased multiple resistance capable of beingtransferred to other bacteria In animals orman should not occur. (b) If increasedtransferable multiple -esistance is found incoliforms, studies may be done to showwhether this resistance Is transferable to.man.

2. The Salmonella reservoir: The use of an-tibacterial drugs at subtherapeutic levels inanimal feed must be shown not to result in(a) an increase in quantity, prevalence orduration of shedding of Salmonella in medi-cated animals as compared to nonmedicatedcontrols; (b) an increase In the number ofantibiotic resistant Salmonella or In thespectrum of antibiotic resistance; (c) disease(caused by Salmonella or other organisms)that is more difficult to treat with either thesame medicated or other drugs.

3. The use of subtherapeutic levels of anantibacterial drug should not enhance thepathogenicity of bacteria, e.g., by increasingenterotoxin production. The -association oftoxin production characteristics with trans-fer factors must be investigated in well-designed studies. (Final resolution of thisquestion was not expected within the 2-yearperiod. Drug sponsors were expected to showevidence of work underway which would leadtoward answers to this question.)

4. An antibacterial drug used at sub-therapeutical levels in the feed of animalsshall not result in residues in food ingestedwhich may cause either increased numbersof pathogenic bacteria or an increase in theresistance of pathogens to antibacterialagents used in human medicine. Hypersen-sitivity to residues was to.be addressed by aliterature survey.

The Commissioner promulgated a finalorder that was published in the FEDERALREGISTER of April 20, 1973 (38 FR 9811),and at that time the requirements im-posed by the regulation became legallybinding on all firms marketing anti-bacterial drugs used at subtherapeuticlevels in feed. In the FEDERAL REGISTERof August 6, 1974 (39 FR 2839), theCommissioner proposed withdrawal ofall approvals held by persons who hadnot complied with the initial require-ments, and all these approvals werewithdrawn by his order, published in theFEDERAL REGISTER of February 25, 1976(41 FR 8282). Therefore, only thoseproducts listed in Part 558 (21 CFR Part558) can be legally marketed at thistime.

By April 20, 1974, the Bureau of Vet-erinary Medicine had begun a review ofthe datd required by § 558.15 which wasapplicable to the principal antibiotics

used subtherapeutc.ally in animal feeds(penicillin and tetracycline), and byApril 20, 1975, data concerning the safetyand efficacy criteria for all antibioticand sulfonamide drug4 bad been re-ceived. To assist the Bureau, the Commi'-sioner asked the agency's National Ad-visory Food and Drug Committee(NAFDC) to review the data and Issuesinvolved and to make recommendationsto him on the future uses of subthera-peutic antibiotics In animal feeds, A sub-committee of three members, the Anti-biotics in Animal Feeds Subcommittee(AAFS), was appointed to work in con-junction with four expert consultantsfrom disciplines related to the issue.

The Bureau prepared 2 days' presen-tations concerning the tetracycline dur-ing which comments were heard from

'the drug industry, animal scientists, andother interested parties. (Chlortetra-cycline, oxytetracycline, and tetracy-cline have the same basic chemical struc-ture and mechanism of action. Histor-ically, FDA has treated these drugs simi-larly, and is treating them identicallyin this matter because there Is no scion-tifie basis for dealing with them other-wise.) The Bureau also prepared a com-prehensive summary report with tenta-tive recommendations for the subcom-mittee. (An Identical procedure was car-ried out for the penicillin.) Two addi-tional meetings were held during whichsubcommittee deliberations were con-ducted and other statements given.

In September 1976, the AAFS present-ed its preliminary recommendations con-cerning the continued subtherapeuticuse of the tetracyclines to the NAFDC,and in January 1977, the subcommittee'sfinal report was submitted to the NAFDC.For tetracyclines, the subcommittee rec-ommended that FDA (1) discontinuetheir use for growth pr6motion and/orfeed efficiency n all animal species forwhich effective substitutes are available,(2) permit their use for disease controlwhere effective alternate drugs are un-available (the approved use should belimited to the extent possible, to thosoperiods of time for which the presenceof the drug in the feed of a particularanimal species Is necessary due to thethreat of animal disease), and (3) con-trol the distribution of the tetracyclines(and penicillin) through FD Form 1800'sand a veterinarian's order to restricttheir use.

The NAFDC rejected the first two rec-ommendations. Instead, it recommendedthat FDA make no changes In the per-mitted uses of chlortetracycllno andoxytetracycline in animal feed. Thecomnittee did adopt the subcommittee'srecommendation that the addition of thetetracycline in feeds be restricted.

The Food and Drug Administrationcarefully considered the recommenda-tions of the NAFDC, the Subcommittee,and the Bureau of Veterinary Medicine,On the basis of this information, the Di-rector of the Bureau of Veterinary Medi-cine is proposing to withdraw approval ofthe subtherapeutic use of tetracyclinesin animal feeds except for those condi-tions of use for which there are no safe

FEDERAL REGISTER, VOL 42, NO. 204-FRIDAY, OCTOBER 21, 1977

56266



NOTICES

and effective substitutes. The Director isalso incorporating the conclusion of theNational Academy of Sciences/NationalResearch Council (NAS/NRC) Drug Ef-ficacy Study Group pertaining to the ef-fectiveness of the tetracycline for sub-therapeutic use; he accordingly is pro-posing to withdraw approval of all suchclaims for tetracycline use in animalfeed that he concludes lack substantialevidence of effectiveness. Therefore theDirector is proposing to withdraw ap-proval of all subtherapeutic tetracyclinein animal feed except for the following:

(1) Oxytetracycline, as an aid in thecontrol of fowl cholera caused by Pas-teurella multocida in chickens and infec-tious synovitis caused by Mycoylasmasinoviae in chickens and turkeys; (2)clortetracycline (a) as an aid in themaintenance of weight gains in thepresence of respiratory diseases, such asshipping fever, in combination with sul-famethazine in beef cattle, (b) as an aidin the control of infectious synovitiscaused by 31. pasteurella in chickens andturkeys, (c) for the control of active in-fections of anaplamosis in beef cattle(d) as an aid in reducing the incidenceof virbrionic abortion in breeding sheep.

IM SUAIJARY OF THE ARGUIIEMT

Soon after the discovery of penicillin,Sir Arthur Fleming noted that somebacterial organisms could become re-sistant to the antibiotic. As the use ofantibiotics has increased, the numberand types of bacterial resistance havealso multiplied. There is a serious con-cern that, in- time, this will lead to de-clining usefulness of antibiotics in thetreatment of both human and animaldiseases.

The Bureau's primary concern is withthat portion of increased bacterial anti-biotic resistance which may result fromthe widespread practice of using sub-therapeutic levels of the tetracyclinesand other antibiotics in animal feed forprolonged periods. This practice, whichsometimes produces increases in growth.promotion/feed efficiency, provides anideal environment for selective pressureto operate. When exposed to an anti-biotic, the organisms that are drug re-sistant survive while the growth of other(drug-sensitive) bacteria is inhibited.Eventually, the antibiotic-resistantorganisms predominate in the bacterialpopulation, and continuous antibioticpressure Perpetuates this abnormal situi-ation.

Bacterial antibiotic resistance is pri-marily determined by genetic elementstermed "R-plasmids" (R-factors, R+).The Bureau's specific concern, therefore,is with the health hazards that mayarise through an increase in the pool ofR-plasmids in the animal populationand the potential transfer of these R-plasmids and R-plasmid-bearing orga-nisms to the human population and sur-rounding environment.

R-plasmids are small lengths of DNAthat are separate from the bacterialchromosome. These R-plasmids carrytransferable genes for drug resistance

as well as the capacity to reproducethemselves. Plasmids may determineresistance to more than one antibiotic,and resistance to several antibiotics Iscommon. Moreover, plasmids can trans-fer from one bacteria to another andfrom nonpathogenic to pathogcnlcstrains. Transfer occurs, although withvarying frequency, among all membersof the enteric batterla and also to mem-hers of other families of bacteria. Thenormal Gram-negative bacterial intes-tinal flora (largely Escherichia colilserves as a reservoir of R-plasmids; theR-plasmld-bearing bacteria interchangeamong animals, man, and the environ-ment. The potential for harz incress-3as the R-plasmld reservoir increse be-cause the probability of R-plasmldtransfer to pathogens Increases. Whcnthe Commissloner required all holdersof approved NADA's for the subthera-peutic use of the tetracyclines in animalfeed to submit data to resolve the safetyquestions raised, he was principallyconcerned with the effect of the anti-biotics approved for subtherapeutic mueLn animal feed on the emergence oftransferable drug resistance in the Sal-monella reservoirs and the E. coll of ani-mals. In the Director's opinion, the re-sults of the studies submitted and thedata available are clear-the affectedparties have failed to show that exten-sive subtherapeutic use of the tetra-cyclines is safe.

Evidence demonstrates that the use ofsubtherapeutlc levels of the tetracyclinesand other antibiotics in animal feed con-tributes to the increase in antibiotic-re-sistant E. coli and in the subsequenttransfer of this resistance to Salmonella.Further, some strains of E. colt and Sal-moma infect both man and animals.

The holders of approved NADA's havesubmitted no evidence to demonstratethat the observed strains E. colt and Sal-monela in man and animals are mutu-ally exaluslve; in fact, there Is evidenceto the contrary. Furthermore, In somecases the R-plasmids as well as the re-sistance genes from humans and animalsources are indistinguishable. Thus, thepotential for harm exists, as illustratedby the studies submitted and verified byevidence from studies conducted by in-dependent scientists.

The holders of approved NADAs werealso required to submit studies demon-strating that the subtherapeutic use ofthe tetracycline in animal feed wouldnot compromise subsequent antibiotictherapy in man or animals, but animalstudies submitted to determine whethersubtherapeutic tetracycline use compro-mised subsequent therapy with relateddrugs were inconclusive because thestudies were inappropriate.

Additionally, the NADA holders wererequired to prove that the subtherapeu-tic use of the tetracyclines would not in-crease the pathogenicity of the infectingorganism. They have submitted no ade-quate studies on the issue, and other re-cent evidence now suggests that the ge-netic determinants for toxin productionmay become linked with drug resistancegenes.

Also the sponsors have failed to estah-lish tissue no-efecL levels for the derel-opment of transmissible R-Plasmid re-sistance, although heating may inacti-vate the residues.

irnally, the NAS/INRC Drug EfficacyStudy Group evaluated the effectivenessclaims for the tetracycline premies andconcluded that there was a lack of sub-stanti al evidence that the premixes wereeffective for many of their subthera~eu-tic labeling claims.

For all the foregoing reasons, the DV-rector Is proposing to withdraw approvalof certain N .D's for the subtheapeutisuse of tctracycline and tetracycline com-binatlon products (e.g., chlorotetracy-cline-sulfamethazlne-penicllin, in ani-mal feed), becauze they have not beenshovn to be safe or lack substantial evi-dence of effectivens.

IV. STUDxrs Rl=VN- TO HUMMA A-DA~riAL H=Tan SirET CRrIjA

A. Tr.A.SSZf or DrITO rESITAWZ (enI=-nro, 1); TH rooL OF R-PLA52l-DEMrnUlG OMZOAMS2ZS is rnlcrMESaN

1. Bac7:grovrd. One of themost impor-tant animal and human health safetycriteria (number I., set forth in I1B.above) concerns the role of subtherapeu-tic antibiotic uze in the selection for anincrease in the pool of microbial plas-mids determining multiple drug resist-ance, and in the transfer of these pTIa-

idz among bacteria in animals andman. Resistance to antibiotics has beenknown as long as the antibiotics them-selves have ben known. Until 1959, itwas believed that antibiotic resistanceva a result of chance mutation and nat-ural selection alone. However, in 1959,Japanese Investizators (Ref. 1) discov-ered that resistance to several commonantimIerobal agents could be transferredsimultaneously from one bacterium toanother by cel-to-cell contact (conju-gatlon). This was shown to bd due to thetransfer of extrachromcsomal resistancedeterminants called "R-plasmds," Le.,RZ-factors, or R-. Resistance producedby R-pla--mds frequently involves theproduction of enzymes that inactivatethe antibiotic. For example, R-plasmid-mediated penicillin resistance is due tothe production of azi enzyme, penicillin-are, that Inactivates the penicillinmole-cule. This same enzyme I. also activeagainst many semisynthetic penicillins,including ampicillin. R-plasmids are ex-trachromosomal genetic element- (DNAmolecule) that may carry as many asnine drug resistance genes. The plasmidsalso carry other genes that determinethe R-plasmld's replication, independ-ent of the host chrombsome, as well asinformation for transfer of the R-pIas-mids from one bacterium to another byconjugation. R-plasmids are transferredby conjugation to virtually all entero-bacterlaceae as well as to such unrelatedGram-negative bacteria as Vibrio, Pseu-domona., and Pasteurel. Thus, resist-ance may pass from strain to strain, spe-ies to species, and most importantly,

from nonpathogen to pathoge. R-plas-rids are now known to be the predomi-nant cause of antibiotic resistance 1A

FEDERAL REGISTER, VOL 42,, NO. 204-FRIDAY, OCTOBER 21, 1977

56-967



56268

Gram-negative organisms that cause hu-man disease, e.g, E. coli, Salmonella,Shigella, etc.

While the development of antibioticsrevolutionized the treatment of infec-tious disease In both man and animals,the magnitude of this achievement hasbeen diminished by the widespread emer-gence of antibiotic-resistant bacteria.R-plasmid-mediated resistance is par-ticularly ominous since selection of re-sistance to a single antibiotic may alsolead to the simultaneous selection of re-sistance to a wide spectrum of otherantibiotics. In recent years, antibiotic re-sistance has emerged in important path-ogens; for example, in Haemophilus in-fluenzae, Neisseria gonorrhoeae, Samo-nella typhi, and Shigella dysenteriae. R-plasmid-mediated resistance has beenidentified in epidemics around the world,e.g., Salmonella typhimurium. Some ofthese organisms have acquired bothampicillin and chloramphenicol resist-ance, resulting in disease that will nolonger respond to therapy. Hence, drug-resistant organisms have become animportant concern in both human andveterinary medicine (Refs. 2 and 3).

Because the use of antibiotics is exten-sive, an effort must be made to assurethe future utility of these lifesavingproducts. In 1960, the amlual productionof antibiotics in the United States was4.16 million pounds, of which 2.96 millionpounds were used for therapeutic pur-poses in human and veterinary medicineand 1.20 million pounds in animal feedadditives. By 1970, 9.6 million poundswere being used for human and veteri-nary medicine pharmaceuticals; while

'7.3 million pounds were being used foranimal feed additives. Moreover, accord-ing to "Synthetic Organic Chemicals,United, States Production and Sales(1971-1975)" (U.S. International TradeCommission Publication 804), the 5-yearaverage production for 1971 through1975 was 11.16 million pounds for me-dicinal uses and 7.68 million pounds fornomedicinal uses, including feed addi-tive uses. Over those 5 years, the aggre-gate average of the total production fornonmedicinal uses was 40.8 percent, but48.6 percent In 1975. Thus, theuse ofantibiotics in animal feeds is a consider-able element In the overall use of anti-biotics in. this country and conse-quently must be considered a potentiallysignificant contributor to the resistanceproblem.

REFERENCES

1. Watanabe, R., "Infective Heredity ofMultiple Drug Resistance in Bacteria," Bac-teriological reviews, 27:87-115, 1963.

2. Simmons, H. 6rd P. D. Stolley, "This isMedical Progress?" Journal of the'AmericanMedical Association, 222:1023-1028, 1974.

3. Linton, A: H., "Antibiotic Resistance:The Present Situation Reviewed," VeterinaryRecord, 100:354-360, 1977.

2. Criterion. The FDA task force con-cluded that a human health hazardexists if the subtherapeutic use of anti-biotics in animal feeds leads to an in-crease In R-plasmid-bearing organisms,if these antibiotics used subtherapeuti-cally are also used in human clinical

NOTICES

medicine, and if R-plasmids subsequentlyappear' In bacteria in man. It was theintent of the task force as well as theintent of § 558.15 to reduce the total loadof resistant organisms in the environ-ment and to insure the effectiveness ofantibiotics in the treatment of disease inman and animals. Accordingly, § 558.15required that an antibacterial drug fedto animals shall not proinote an increaseof coliforms that are resistant to anti-bacterial drugs used in human clinical"medicine and capable of transferringthis resistance to bacteria indigenous tothe intestinal tract of man. Studies mustbe undertaken to assess the occurrenceand significance of these events:

a. Controlled studies shall be under-taken to determine whether or not theadministration of an antibacterial drugat low and/or intermediate levels to tar-get animals results in an increase in thenumbers of coliforms bearing R-plas-mids present in the intestinal tract ofthe animal or a change in the resistancespectrum of these organisms comparedto those found, in controls receiving noantibacterial drug. The resistance spec-trum must be determined to ascertainwhether or'not there are determinantspresent for resistance to antibacterialdrugs used in human clinical medicine.

b. If the resistance determinants in-dicated in paragraph a above are found,a sponsor may elect to conduct addi-tional studies to determine if such multi-ple drug resistance is transferable to theindigenous coliforms in the intestinaltract of man.

3. Studies relevant to transfer of drugresistance-(a) R-plasmid-bearing E.coli develop in domestic animals Jed sub-therapeutic levels of antibiotics, includ-ing tetracycline. Many investigators havereported the presence of R-plasmid-bearing E. coli in domestic animals, andthe effect of antibiotic-supplementedfeed in increasing the number of anti-biotic-resistant organisms has been ex-tensively documented. Mercer et al.(Ref. 1) showed that 80 percent of thebacterial isolates from animals exposedto tetracycline and other antibiotics infeed were antibiotic resistant, while only21.9 percent of isolates obtained from un-exposed animals were resistant. Seigel etal. (Ref. 2) and Smith and Tucker (Ref.3) as well as others have also shownthat the addition of tetracyclines to feedat subtherapeutic levels causes an in-crease in the R-plasmit-bearing coliformpopulation of the intestinal flora. Datasubmitted by drug sponsors on the effectof subtherapeutic administration of tet-racyclines in animals also show an in-crease in drug-resistant E. coli In medi-cated animals, compared to nonmedi-cated 'controls. A review of data fromthe literature, from FDA control studies,and from drug sponsors' submissionsleads to the conclusion that subthera-peutic use of tetracyclines in animal feedproduces a high level of antibiotic-re-sistant E. coli in animals by selecting forR-plasmid-containing bacteria (HumanHealth Criteria No. la). These bacterialpopulations appear to be stable and per-

sistent, even in the absence of tetracy-cline pressure. Once the reservoir of R-plasmids develops (whether due to sub-therapeutic use of tetracycline or someother antibiotic), the plasmids cantransfer among bacteria Infecting ani-mals and man.

I REFE CES

1. Mercer, H. D., D. Pocurull, S. Gaines, S.Wilson, and J. V. Bennett, "Characteristicsof Antimicrobial Resistance of Escherichlacolt from Animals: Relationship to Veteri-nary and Management Uses of AntimicrobialAgents," Applied Microbiology, 22:700-705,1971.

2. Seigel, D., W. 0. Huber, and F. Enloe,"Continuous Nontherapeutic Use of Antibac-terial Drugs in Feed and Drug Resistance ofthe Gramnegativo Enteric Flora of Food-Producing Animals," Antimicrobial Agentsand Chemotherapy, 6:697-701, 1974.

3. Smith, H. W.. and J. V. Tucker, "TheEffect of Antibiotic Therapy on the PacalExcretion of Salmonella typhimurum byExperimentally Infected Chickens," Journalof Hygiene, 75:275-292, 1076.

(b) E. coli contribute their R-plas-mids to man through several mecha-nisms. Drug-resistant bacteria originat-ing in animals may reach man (1) by dl-rect contact with animals (2) throughthe food chain, and (3) because of theirwidespread occurrence in the environ-ment.

(i) Direct contact with animals. Anumber of studies have shown that hu-mans in contact with animals receivingmedicated feed have a higher Incidenceof drug-resistant organisms In their In-testinal flora than do control populationswithout this direct contact. Linton et al.(Ref. 1) found a higher Incidence ofdrug-resistant E. coli In adults employedwith livestock husbandry than in otherrural or urban adults. Wells and JamOs(Ref. 2) found, a higher incidence ofdrug-resistant E. coli in humans In con-tact with pigs given certain antibioticsthan in humans in contact with pigsthat had not been given antibiotics.

Siegel et al (Ref. 3) compared the pro-portion of resistant organisms In fecalsamples from: (a) Farm workers in con-tact with the resistant flora of animalsreceiving subtherapeutic levels of peni-cillin, (b) people residing on the samefarms with no direct exposure to thefarm animals; (c) nonfarm peopletreated with antibacterial drugs; (d) un-treated people residing with treated In-dividuals; (e) untreated people with noexposure to farm animals or treatedindividuals.

The data (Ref. 3) Indicate that theenteric flora of individuals who have notbeen treated with antibiotics can be af-fected by contact with animals; further-more, these individuals may be affectedby contact with people who have devel-oped a predominantly resistant flora asa result of their exposure to subthera-peutic levels of antibacterals In feeds.

A study sponsored by the AnimalHealth Institute, Levy et al. (Ref. 4), ex-amined the change in intestinal micro-flora of chickens, farm dwellers, andtheir neighbors before and after the In-troduction of a tetracycline-supplement-ed feed to the farm. Within 1 week after

FEDERAL REGISTER, VOL. 42,- NO. 204-FRIDAY, OCTOBER 21, 1977



NOTICES

introduction of this antibiotic in theirdiet, the E. coli of-the chickens werealmost entirely tetracycline-resistant.Subsequently, and at a slower rate, in-creased numbers of antibiotic-resistantbacteria appeared in the flora of thefarm dwellers. No such increase was ob-served in the farm neighbors, who werenot exposed to the animals fed subtheira-peutic antibiotics. Within 5 to 6 months,31.3 percent of weekly fecal samplesfrom farm dwellers contained greaterthan 80 percent tetracycline-resistantbacteria compared to 6.8 percent of thesamples from the neighbors. Using aspecially marked resistance gene to iden-tify a particular plasmid, Levy was alsoable to demonstrate the direct spread of

-resistant organisms from chickens tochickens and from chickens to man(Ref. 5).

The studies do not establish that theshift in the antibiotic-resistant E. collflora of rural human populations was aresult of contact with livestock, per se.since some shift could have also occur-red as a result of contact with the anti-biotic-supplemented feed used on thefarms. Nonetheless, it was demonstrated

-that the subtherapeutic use of certainantibiotics, including the tetracyclines,increases the pool of R-plasmid-bearingE. coil, and the studies define one routeby which antibiotic-resistant strains canenter the human poj~ulation. While thisroute is of great importance to farmdwellers, the majority of the populationhas no contact with live animals,. For thislatter group of individuals, a more Im-portant route of exposure by which re-sistant bacteria can pass to man is by thehandlifg and ingestion of meat and poul"try products contaminated with R-plas-mid-bearing E. coli of animal origin.

1. Linton, 3K. B., F. A. Lee, I. H. Richmond,W. A. Gillespie, V. T. Rowland, and V. N.Baker, "Antibiotic Resistance and Transmis-sible R-factors in the Intestinal CollformFlora, of Healthy Adults and Children In anUrban and Rural Community," Journal ofHygiene, 70:99-104. 1972.

2. Wells, D. 'AL, and 0. B. James, "Trans-mission of Infectious Drug Resistance fromAnimals to Man," Journal of Hygiene. 71:209-215, 1973.

3. Siegel, D., W. G. Huber, and S. Drysdale."Human Therapeutic and Agricultural Uses-of Antibacterial Drugs and Resistance of theEnteric 'lora of Humans," AntimicrobialAgents and Chemotherapy, 8:538-534, 1975.

4. Levy, S. B., G. B. FItzgerald. and B. S.Macone, "Change in Intestinal lora of FarmPersonnel After Introduction of a Tetracy-cline-Supplemented.I'eed on a Farm," NewEngland Journal of Medicine, 205:Z83-588,1976b.

5. Levy, S. B., G. B. Fitzgerald, and A. B.Iacone, "Spread of Antibiotic Resistant Plas-mids FTrom Chicken to Chicken and fromChicken to Man," Nature, 260:40-4-, 1976a.

(ii) Contact with E. coli-contaminatedfood. To assess adequately the signifi-cance of the problem of human food con-taminated with E. coil, Howe and Linton(Ref. 1) described four factors that mustbe measured: (a) The incidence of R-plasmid-bearing E. col in food-produc-ing animals; (b) the load and frequencyof excretion of E. coll from these ant-

mals; (c) the degree and source of con-tamination of carcasses at slaughter; and(d) the overlap of E. colt serotypes invarious host animals with those com-monly found in humans. A number ofsurveys have clearly documented thatpigs, calves, and poultry carry a-largereservoir of antibiotic-resistant E. colt(Anderson; Loker; Mercer; Smith; Howe,Linton and Osborne; Smith and Crabbe(Refs. 2 through 8, and 15)). The ani-mals excrete a large number of E. coilresistant to a wide range of clinicallyuseful antibiotics and constitute a reser-voir "rich" in R-plasmids. Moreover, theyexcrete a large variety of serotypes of E.coil.

During the slaughtering process, con-tamination of carcasses with intestinalmicroorganisms cannot be prevented.Meat and meat products are often con-taminated with antibiotic-resistant E.coli, and these often reach the consumer.Walton (Ref. 9) demonstrated that 52percent of the carcasses of cattle and 83percent of pig carcasses from commercialabattoirs were contaminated with E. col.Walton and Lewis (Ref. 10) Isolated re-sistant E. coli from 21 to 50 specimens offresh meat and from 4 of 50 specimens ofcooked meat. Babcock et al. (Ref. 11)Isolated multi-resistant E. coil from 80percent of 98 samples of dressed beef.Resistance in most cases was found to betransmissible.

Similar Incidents' of E. coll contamina-tion occur with the slaughter of chickens(Kim and Stephens (Ref. 12), Cooke etaL, and Shooter et al. (Refs. 14 and 18)).

The presence of antiblotic-resistant Z.coil in the animal intestinal tract andon. the carcass does not conclusivelyprove that the E. coil are Identical or-ganisms. However, recent studies usingserotyping methods have characterizedresistant and sensitive E. coil isolatedfrom the animal intestinal tract and car-cass (Refs. 13, 15, 16. and 17), and havefound that the resistant 0-serotypes onthe carcasses of pigs, calves, and poultryfrequently are Identical to those isolatedfrom the fecal contents of the same ani-mal. Moreover, Linton, Howe, et al. (Ref.17), showed that a large number of E.coil found on table-ready thawed chick-ens were resistant to therapeutically im-portant antibiotics. The organismsreaching the kitchen included a wide di-versity of 0-serotypes of antibiotic-re-sistant E. colf. Similarly, Shooter et al.(Ref. 13) described the distribution andserotype of strains of E. colt from a poul-try packing station and an abattoir andconcluded that "results In both the abat-toir and the poultry packing station in-dicate that there is transfer of strainsfrom the faeces of the animals to the en-vironment and that the strains of E. coUlfound on the carcasses of poultry, cattle,and beef will originate from the faeces ofthe animal and from the environmentand will reflect the history of the car-cass."

The epidemiology of Salmonella infec-tions also supports the conclusion thatthe reservoir of R-plasmil-bearing en-teric bacteria in animals is a significantsource of R-plasmds for humans. Food-

borne Salmonella infections In man area well-known and continuing problem.Animal meat products that serve as aprimary source of Salmonella infectionsin humans also serve as a source of otherbacteria for man. Including R-plamld-bearing enteric bacteria (Ref. 19).

Based on this evidence, the Directormust conclude that man is exposed to R-plasmid-bearing intestinal bacteriathrough contact with contaminated food.Because the drug resistance of these bac-teria is Increased by feeding the animalsubtherapeutlo levels of antibiotics, suchfeeding enhances the likelihood of trans-mitting R-plasmid-bearing bacteria toman through contact with contaminatedfood.

1. Howe, M., and A. H. Linton, "The DL-tri-butlon of O-Antigen Tpes of Ecscherlchi coliIn Normal Calves, Comp3red with Man, andTheir R-plaamld Carriage," Journal of Ap-plied Bacteriology, 40:317-330, 1976.

2. Andercon, E. S., "The Ecology of Trans-ferable Drug ResiJtance In the Enterobac-tera". Annual Reiew of Mfcrobiology, 22:131-180, 1958.

3. Howe. M, A. H. Linton, and A. D. Os-borne, "A Longitudinal Study of Escherichiacolt in Cows and Calves with Special Refar-enco to the Distribution of O-Antigen andAntibiotic Resistance." Journal of AppliedBacteriology. 40:331-340, 196a.

4. Loken. X. I. L. W. Wagner, and C. L.HenTe, '" Tanamitible Drug Resistance inEnterobacterace Iolated From CalvesGiven Antibiotics," American Journal ofVeterinary Research, 32:1207-1212, 19,71.

5. Mercer, K. D., D. Pocurull, S. Gaines,S. Wilson. and J. V. Bennett. "Characteristicsof Antimicrobial Resistance 6f Etcherichfacol from Animals: Relationship to Veteri-nary and Mana7ement Uses of AntimicrobialAgents," Applied Microbiology, 22:700-705,1071.

6. Smith H. W., "The Effect of the Use ofAntibacterial Drugs, Particularly as Food Ad-ditive, on the Emergence of Drug ResistantStrains of Bacteria in Animals," New ZealandVeterfntary Journal, 15:153-165, 1967.

7. Smith. K. X., "The Effect of the Use ofAntibacterial Drugs on the Emergence ofDrug-Resistant Bacteria in Animals," Ad-rancee in Veterinary Science and CorparatreMcdelne, 15:G7-160, 1971.

8. Smith. H. W., and W. K. Cabb, "The Effectof the Continuoua Administration of DietsContaining Low Levels of Tetracyclines onthe Incidence of Drug-Reistaant Bacteriumcolt in the Faeces of Pig, and Chickens: TheSensitivity of Bat. colt to Other Chemo-therapeutic Agents," Veterinary Record, 69:2-4-30, 1957.9. Walton, J. R, "Contamination of Meat

Carc=ze by Antiblotic-Resistant ColiformBIacterla," Lancet, 2:561-563,1970.

10. Walton, J. R., and L. K. Lewis, "Con-tamlnaton of Fresh and Cooked Meats byAntibiotic Resistant Coliform Bacteria,"Lancet, 2:215,1971.

11. Babccek, G. F., D. L. BerryhIll, andD. H. Marsh. "R-Factors of Ezcherichfa coltfrom Dreszed Beef and Humans " AppliedMicrobiology, 25:21-23,1973.

12. Kim, T. K. nd J. F. Stephens, "DrugResistance and Transferable Drug Resistanceof E. coIl Isolated from 'Ready-to-Cook'Broilers," Poultry Science, 51:1165-1170,1972.

13. Shooter, R. A. E. Mf. Cooke, S. Oarrell,K. A. Bettelhelm, M. E. Chandler, and F. ILBushrod, "The Isolation of Escherichia colifrom a Poultry Packing Station and an Ab-atoir," Journal of Hygiene. 73:245-217, 1974.

14. Cooke, E. 3., A. L. Braeden, R. A.ShootMer, and S. L. OFarrell, "Antibiotic Sen-


56269



NOTICES

sitivity of Escherichia colt Isolated .romAnimals, Food, Hospital Patients, and NormalPeople," Lancet, 2:8-10, 1971.

15. Howe, K., A. H. Linton, and A. D. Os-borne. "An Investigation of Calf Carcass Con-tamination from Escherichia colt from theGut Contents at Slaughter," Journal of Ap-plied Bacteriology, 41:37-45, 1976b.

16. Linton, A. H., B. Handley, A. D. Os-borne B. G. Shaw, T. A. Roberts, and W. R.Hudson, "Contamination of Pig Carcasses atTwo Abattoirs by Escherichia colt with Spe-cial Reference to O-Serotypes and AntibioticResistance," Journal of Applied Bacteriology,42:89-111, 1977.

17. Linton, A. H, K. Howe, C. L. Hartley,and H. M. Clements, "Antibiotic Resistantand Sensitive Escherichia colt 0-serotypes inthe Gut and on the Carcass by CommerciallySlaughtered Broiler Chickens and the Poten-tial Public Health Implications," in press1977.

18. Shooter, R. A., S. A. Rousseau, E. M.Cooke, and A. L. Braeden, "Animal Sourcesof Common Serotypes of Eschericfhtia coli inthe Food of Hospital Patients," Lancet, 2:226-228, 1970.

19. FIDC Docket No. TTN-0156, Environ-mental Impact Analysis and Assessment Re-ports (EIAR/EAR) for Chlortetracycline-Penicillin-Sulfonamides (CSP) and Penicil-lin Streptomycin Premix Combinations.

(li) Widespread presence in the en-vironment. Many studies (Refs. 1through 6) hdve shown that intestinalbacteria (e.g., E. colt and Salmonella)carrying R-plasmids are widespread inthe environment. Resistant strains reachthe environment from both raw andtreated municipal, hospital, and animalwastes. The number of resistant bacteriazeported in sewage and the effects ofsewage treatment vary. Most surveys in-dicate that hospital sewage containsmore drug-resistant coliforms, more R-plasmids, and a greater proportion ofR-plasmids carrying multiple resistancethan sewage from domestic and othersources. However, hospitals do not con-stitute a large proportion of total sew-age. Therefore, Linton et al. (Ref. 4)compared the contributions of hospitaland domestic sewage to the total pooledsewage output of the city of Bristol, andconcluded that sources such as indus-tries and homes, rather than the hospi-tals, appear to be by far the greatestcontributors to the reservoir of R--plas-mids in the community (Ref. 7).

R-plasmid-containing bacteria alsooccur in rivers and sea water, and someauthors have .urged stricter control ofdischarges to surface waters. Peary etal. (Ref. 2) examined the incidence ofantibiotic-resistant E. colt present atsites along a fresh water river systemand within the salt water bay into whichIt empties. Antibiotic-resistant colformswere detected in nearly all the freshwater sites sampled and in about 50 per-cent of the salt water sites. Feary foundthat 20 percent of the 194 strains testedcontained R-plasmids carrying multipleantibiotic resistance which could betransferred -to sensitive Salmonellatyphimurium (S. typhimurium), Shi-gella dysenteriae, and E. colt. They alsoIsolated coliforms containing R-plasmid-mediated resistance to chloramphenicol.Transferable chlotamphenicol resistanceIs a significant health concern since

chloramphenicol is often the antibioticof choice for the treatment of typhoidfever and for the treatment of systemicillness caused by other Salmonella spe-cies. In Peary's study, the incidence ofcoliform organisms appeared higheraround heavily populated areas, but coli-forms -were also recovered with ease fromrural areas. In one case where partic-ularly high counts were obtained, thesample was taken below a large cattlefeedlot.

The high levels of resistant coliformsmay be of more consequence in the saltwater gince certain sections are utilizedheavily by fishermen in harvesting fish,shrimp, clams, and oysters. Oysters andclams are of primary concern since theycontinuously filter water and concen-trate bacteria in their gut and are ofteneaten uncooked.

Recent reports by Cooke (Ref. 1) havealso described a high incidence of re-sistant conforms in marine shellfish andfreshwater mussels. These data are re-viewed in inore depth in the CSP EIAR/EAR (Docket 77N-0156).

Therefore, the Director must concludethat the environment is heavily conta-minated with bacteria containing trans-ferable antibiotic resistance. Man is ex-posed to the danger of acquiring resist-ant coliforms from the environment, andthe relative number of resistant bac-teria are increased both by the use ofantibiotics in animal husbandry and inhuman medicine. Antibiotic-resistantbacteria are now so widely distributed inthe general environment that it is diffi-cult to relate their appearance to a .par-ticular use, but any unnecessary practicewhich results in the ineffectiveness ofantibiotics for the treatment of diseaseshould be eliminated.

R-FERENCES

1. Cooke, L D., "Antibiotic Resistanceamong Coliform and Fecal Coliform BacteriaIsolated from Sewage, Seawater, and MarineShellfish," AntimicrobWal Agents and Chem-otherapy, 9:879-884, 1976.

.2. Peary, R. W., A. B. Sturvetant, and J.Lankford, "Antibiotic-Resistant Coliforms inFresh and Salt Water," Archives of En s-ronmental Health, 25:215-220, 1972.

3. Grabow, W. 0. K., andO. W. Prozesky,"Drug Resistance of Coliform Bacteria inHospital and City Sewage," AntimicrobialAgent and Chemotherapy, 3:175-180, 1973.

4. Linton, K. B., M. EL Richmond, R. Bevan,and W. A. Gillespie, "Antibiotic Resistanceand R Factors in Coliform Bacilli Isolatedfrom Hospital and Domestic Sewage" Journalof Medical Microbiology, 7:91-103, 1974.

5. Richmond, I. H., "R factors in lanand His Environment," in "Microblology-1974," edited by D. Schlessinger, AmericanSociety for Microbiology, Washington, D.C.,1975.

6. Smith, H. W., "Incidence in River Waterof Escherichia colt Containing 1-Factors,"Nature, 228:1286-1280.1970b.7. FDC Docket No. 77-0156, EIAR for CSP.

(c) R-plasmid-bearing human andanimal strains of bacteria overlap. Typ-ing of surface bacterial antigens is usedas a means of identifying bacteriastrains. Three types of specific surfaceantigens are associated with the E. coltcell: An "0" cell wall lipopolysaccharideantigen, "K" capsular or envelope anti-gen, and an "H" flagellar protein anti-

gen which occurs among mobile orga-nisms. The antigens are characteristicof a specific organism, and they servo toIdentify distinct bacterial types (sero-types) within species. Their presence i1detected by the ability of E. colt orga-nisms to interact with specific anti-serums.

(i) Epidemiologic l, investigations-Z,coli serotyping. (a) Despite the wide-spread occurrence of R-plasmids in theenvironment, some workers (Bettelhetmet al., Ref. 1) suggested that humanE. colt and animal E. colt were distinct.These workers argued that there weremarked differences in serotype distribu-tion in strains isolated from man andanimals; they also suggested that ani-mal strains of E. colt were not reachingthe human population or were failing toimplant in the bowel. More recently,however, this same group, Bettillelhm ctal. (Ref. 2), compared the serotypes of13,139 strains of E. colt isolated fromhumans with the serdtypes of 1,076 ani-mal strains of E. colt; 708 different O/Hserotype combinations were found. Ofthese, 520 were found in human strainsonly, 130 from animal strains only, and58 O/H serotype from humans and ani-mals. The'authors concluded:

At first glance the results described in thispaper would indeed support the view thathuman and animal strains of E. colt arelargely distinct. Second thoughts, however,suggest a little caution in accepting the opin-ion too lirmly.

However thoroughly human or animalstools .are examined, only a minute fractionof the total bacterial content is examined,and inevitably strains recorded as being Iso-lated tend to be those that predominate.It is always probable that if examifnation icontinued, further strains may be isolatedbut after an amount of work that Is im-practicable in any ordinary investigation.If this Is so, it is possible that many of thestrains recorded as coming from hlmansonly or fromannimals only might, with morediligent examination, be recorded as presentin both man and animals.

Rm-mnu CES1. Bettelhelm, K. A., A M. Busebrod, r.

Chandler, E. M. Cooke, S. OTarrell, and R. A.Shooter, "Escherichia coli Serotypo Dltrlbu-tion in Man and Animals," Journal o/ Hy-giene, 73:467-471, 1974.

2. Bettelhenir, K. A., w. Ismail, n, Shine-baum, R. A. Shooter, E. Moorhouse, andWendy Farrell, "The Distribution of Sero-types of Eschcrtchia colt in Cow-Pats andOther Animal Material Compared with Sero-types of E. colt, Isolated from HumanSources," Journal o/ Hygiene, 70:403-400,1976.

(b) Linton, Howe, Richmond, andtheir collaborators (Refs. 1 through 4)also conducted extensive epidemiologicalinvestigations. They found a wide rangeof resistant and sensitive O-serotypes ofE. col ;n calves, pigs, and poultry, andthey compared these serotypes withthose f6und in the human intestine. Theauthors found that many 0-serotypescommon to man were also common toone or more of the three animal speciesexamined. Thus, they concluded that Itis impossible to make a clear distinctionbetween "animal" and "human" intesti-nal strains of antiblotlc-resistant E. coltbased on 0-serotyping alone. More Ir-


56270



NOTICES

portantly, the studies suggest a consider-able overlap in the distribution of R-plasmid-bearing O-serotypes in man andin animals. Moreover, the same resistantserotypes, which predominate in the E.coi populations from healthy humanand animal fecal sources, were alsoprevalent among R-plasmid-bearingstrains from clinical material (Ref. 5).

Because the use of O-serotyping aloneas an epidemiological tool has been criti-cized on the grounds that it is incom-plete and inadequate, Howe and Itnton(Ref. 2) examined E. coi for the K andH antigens as well as the 0 antigen.They studied 90 strains, 17 chosen atrandom from human urinary tract in-fections, 17 from human feces, and 56from calf feces; all belonging to O-types8, 9, and 101. The authors found thesame K and H antigens in certain strainsof the same O-types from each of thethree E. coli sources. Additionally, K andH antigens associated with these O-sero-types were not specific to antigens as-sociated with these O-serotypes were notspecific to E. coi isolated from humansor from calves. Although further sub-division of the three O-serotypes waspossible by this means, the authors con-cluded that O-serotyping alone provideda very useful means of distinguishingstrains of E. col in a general survey.

These studies show that a similarrange of drug-resistant R-plasmid-bear-ing O-serotypes of E. coli have beenfound in man and the various animalspecies examined. Furthermore, thestudies show that the ratio of drug-re-sistant to drug-sensitive isolates wasmuch higher in animals than in man(Refs. 2 and 6). Thus the abundance anddiversity of drug-resistant R-plasmid-bearing O-serotypes in animals are muchgreater than that currently found inman, and the serotypes overlap.

REFEMREs1. Hartley, C. L., K. Howe, A. IL Linton,

M. B. Linton, and L H.r Richmond, "Distri-bution of R-Plasmids Among the O-AntigenTypes of Eseherichia coUl Isolated from Hu-man and Animal Sources," AntimicrobialAgents and Chemotherapy, 8:122-131, 1975.

2. Howe, M., and A. H. Linton, "The Distri-butlon of O-Antigen Types of Escherichiacoli in Normal Calves, Compared with Man,and Their R-Plasmid Carriage," Journal ofApplied Bacteriology, 40:317-330, 1976.-

3. Howe, K., A. H. Iinton, and A. D. Os-borne, "A Longitudinal Study of Escherichiacolt in Cows and Calves with Special Refer-ence to the Distribution of 0-Antigen andAntibiotic Resistance," Journal of AppliedBacteriology, 40:331-340, 1976a,

4. Linton, A. H.. B. Handey, A. D. Osborne,-3. G. Shaw, T. A. Roberts, and W. R. Hudson,"Contamination of Pig Carcasses at Two Ab-attoirs by Escherichia colt with Special Ref-erence to O-Serotypes and Antibiotic Resist-ance," Journal of Applied Bacteriology,42:89-111, 1977.

5. Petrocheilou, V., and M1. H. Richmond,'"istribution of R-Plasmids Among the 0-Antigen Types of Escherichia colt Isolatedfrom Various Clinical Sources," "Antimicro-bial Agents and Chemotherapy, 9:1-5, 1976.

6. Howe, H., A. H. Linton, and A. D. Os-borne, "The Effect of Tetracycline on theCollform Gut Flora of Broiler Chickens withSpecial Reference to Antibiotic Resistanceand O-Serotypes of Esaherichia colt," Jour-

nal of Applied Bacteriology, 41:453-464,1976c.

(li) Direct ingestion evidence. Directingestion experiments have also beenconducted to show that R-plasild-bear-Ing E. colt of farm origin can colonize thehuman intestinal tract. In 1969, Smith(Ref. 1) concluded that animal E. coltstrains were poorer at colonizing the in-testine of man than were human E. collstrains. However, his observations werebased on a single volunteer (himself)and a small number of E. coll strains.Cooke in 1972 (Ref. 2), on the otherhand, reported that it was relatively easyto produce temporary colonization of theintestine by E. colt strains from both hu-man and animal sources. She reportedthe persistence of an E. Coll infection ofanimal origin In a human volunteer for120 days following the ingestion of avery large dose.

Other experimental studies (Refs. 3and 4) confirm that temporary coloniza-tion occurs provided a large dose of theorganisms is taken, but there is a greatdeal of biological variation betweencolonization for different strains and fordifferent human individuals. In normalindividuals, the carriage of intestinal E.coli seems to follow a characteristic pat-tern. Each person carries one or two rest-dent strains that establish themselvesand multiply for months or years. Inaddition, four or more transient strainsare present for a few days or weeks.Strains disappear and are replaced byothers. Sometimes, under antibioticpressure, a new strain suddenly takesover, later disappearing. Strains of R.coli thus differ in their ability to colonizeman. Although some strains are not welladapted to colonizing man, others roas able to live in human as in animal In-testines. The greater the diversity ofR-plasmid-bearng O-serotypes thatreach the consumer, the greater theprobability that one more of these anti-biotic-resistant strains will be capableof colonizing man.

Recently, Linton, Hove, Bennett, et al(Ref. 5) demonstrated that antibiotic-resistant E. coli found on a commerciallyprepared chicken carcass colonized theintestinal tract of a human volunteer.Two strains present on the chicken car-cass handled and eaten by the humanvolunteer were subsequently excreted byher. Both strains were undetectable inthe human before contact with thechicken carcass. The strains were shownto be Identical in chicken And man bycomparing their serotypes (0, K, and Hantigens) and R-plasmds. The plasmidcomplements were determined to beidentical by electron microscopy and re-striction endonuclease patterns. Restric-tion endonucleases are enzymes thatDNA at specific sites. Physlochemicaltechniques then visualize these plasaidfragments. The identity of these plas-mids can be determined by a comparisonof the DNA fragments generated usingrestriction enzymes with different recog-nition sequences. The Linton study alsosuggested that the handling of the un-cooked carcass provided a greater oppor-tunity for transmisslon that does eating

56271

cooked meat. The strains persisted for10 days, and the process occurred with-out feeding any antibiotics to the volun-teer during the study. This is consistentwith reports of Salmonella infectionsfrom animal sources.

1. Smith. H. V., "ransfer of AntibioticRneaitanco from Farm Animals and HumanStrain of E.cher c ia coIl to Resident E.colt In the Alimentary Tract in IAhn," Lancet,1:1174-1176. 1969.

2. Coole. E. M., L G. T. Hetlarchy, and C. A.Buclk, "Fate of Ingrsted Echerchia coNl inNormn Peircona." Journal of Medical Micro-biology, 5:361-360. 1972.

3. Ander-on. J. D, W. A. Gillespie, andM. H. Richmond, "Chemotherapy and Anti-blotlc-Realatanco Transfer between Entero-bacteria In the Human Gastro-IntestionalTract," Journal of Medical Microbiology,6:461-473, 1973.

4. Anderson. J. D, L. C. Ingram, 2 H.Richmond. and B. Wledemann, "Studies onthe Nature of Pla-mrids Arising from Conju-gation In the Human Gastro-intestinalTract." Journal of Medical Microbiorogy,0:475-483. 1973.

5. Linton. A. MH E. Rowe, P. IL Bennett,I& H. Richmond. and F. . Whiteside, "TheColonization of the Human Gut by Anti-biotic R-_istant Escherichfa cof from Chick-ens," In prezz, 1977.

(il) In vivo studies sho thatR-plasmids transfer from E. colN topathogens. The ingestion of R-plasnid-containing bacteria can result in in vivoR-plasmtd transfer to the normal intes-tinal flora. When this occurs, the E. coliconstitute a reservoir of organisms ca-pable of transferring R-plasmids to in-testinal pathogens, e.g., Salmonella. TheIn vivo transfer of R-plasmids has beendemonstrated in sheep, mice, calves,pigs, chickens, turkeys, and in the hu-man allmentary tract (Refs. 1 through8). Generally, in vivo transfer is not asreadily detectable as in vitro transfer.In the absence of drug selection, the rateof in vivo R-factor transfer is generallylow, and large numbers of resistant do-nors may be required for transfer (Refs.1 and 6). Demonstrations of in vivotransfer have usually been achieved byfirst modifying the normal flora of thealimentary tract by feeding antibiotics,by starvation, or by using germ-freemice or newly hatched chicks, and theseprocedures probably counteract the in-hibitory effects of bile salts, fatty acids,acid pH, and anaerobic conditions of thenormal intestinal tract.

These experimental results may not bea true indication of the extent ofR-plasmd transfer in natural popula-tions since they often involve individualswho are exposed to restricted numbersand types of donor and recipient orga-nvla . In some instances the methodswere not suitable for the detection oflow level transfer. However, Smith andTucker (Ref. 9) studied the effect ofantibiotic administration on the fecalexcretion of Salmonella by experimen-tally infected chickens. The authorsfound that R-plasnmild resistance de-veloped in the indigenous E. coli andthat very similar resistance patternsthan developed in the Salmonella. Theseresults were duplicated in some of thestudies submitted by the NADA holders,

FEDERAL REGISTER, VOL 42, NO. 204-.RIDAY, OCTOBER 21, 1977



NOTICES

which are also discussed in depth underPart IV. B. below.

Regardless of the frequencp withwhich R-plasmid transfer occurs in theabsence of modifying influences, it hasoccurred and given 'rise to antibioticresistance in bacteria, including patho-gens. The conditions of the Smith andTucker studies -mimic those broughtabout by the practice of feeding sub-therapeutic levels of tetracycline andother antibiotics to animals. That prac-tice leads to an increase in and selectionfor R-plasmid-bearing organisms, andIt therefore increases the probability ofin vivo R-plasmid transfer to pathogens.

RMrE=cES

1. Anderson, J. D., W. A. Gillespie, and M.EL Richmond, "Chemotherapy and Anti-biotic-Reslstance Transfer between Entero-bacteria in the Human Gastro-intestinalTract," Journal of Medical Microbiology,6:461-473, 1973.

2. Guinee P. LT., '"Transfer of MultipleDrug Resistance from EsehericMa Coli toSalanonella typhkimurium in the Mouse In-testine," Antonic van Lecuwenli-elk; Journalof Microbiology and Serology, 31:314-22.1965.

3. BXasuya, AT., "Transfer -of Drug Re-sistance between Enteric Bacteria Inducedin the Mouse Intestine," Journal of Bac-teroloyy, 88:322-331, 1964.

4. Nivas, S. C., M. D. York, and B. S.Pomeroy, "In Vitro and In Vivo Transfer ofDrug Resistance from Salmonella andEscheri cha coli Strains in Turkeys." Ameri-can Journal of Veterinary Research, 37:433-437, 1976.

5. Salzman. T. C., and L. Elemm "Trans-fer of Antibiotic Resistance (R-factor) in theMouse Intestine," Proceedings of the Societyfor Experimental and Biological Medicine,128:392-394, 1968.

6. Smith, H. W., "Transfer of AntibioticResistance from Farm Animals and HumanStrains of Esehcriclia colt to Resident E.coll in the Alimentary Tract In Man," Lancet1:1171--1176,1969.

7. Smith, H. W., "The Transfer of Anti-biotic Resistance between Strains of Entero-bacteria In Chickens, Calves, and Pigs,"Jounral of Medical Microbiology, 3:165-180,1970a.

8. %Valton, J. R., "In Vivo Transfer of In-fective Drug Resistance," Nature, 211:312-313,1966.

9. Smith, H. W., and J. F. Tucker. "TheEffect of Antibiotic Therapy on the FaecalExcretion of Salmonella typhimurlum by Ex-porlmentafly Infected Chickens," Journal ofHygiene, 75:275-292 1975.

(iv) R-plasmid compatiblity studies.Another FDA sponsored study (Ref. 1)examined the compatibility properties ofmore than 100 R-plasmids from E. coliand Salmonella isolated form animals inorder to determine whether the plasmidsare related to those isolated from man.The usual method of genetically classify-ing plasmids is based on their ability toexist with each other in the same bac-terium. Genetically unrelated plasmidscan exist in the same host, and theyare called compatible. On the other hand,related pldsmids cannot coexist, and theyare called incompatible. Plasmids be-longing to the same incompatabilitygroup are presumed to be related.

The Food and Drug Administrationstudy showed that the R-plasmid Incom-

patibility groups seen in animal isolatesshow the same distribution as thosefound in human isolates. This-thereforesuggests that human and animal bac-terial populations overlap; there are notseparate and distinct human and animalR-plasmids.

A more direct approach for examiningthe relationships between plasmids is tomeasure the proportion of DNAsequences (that is, the number of similaror identical genes) that are common toany two plasmids (DNA-NA hybridiza-tion). R-plasmids belonging to the sameincompatibility groups of human andanimal origin are identical when ex-amined. by DNA-DNA hybridizationtechniques. (Refs. 2 and 3). Restrictionendonuclease activity has also confirmedthe similarity of R-plasmids isolatedfrom' enteric organisms of human andanimal sources (Ref. 4). Therefore, theDirector must conclude that R-plasmidsof human origin are indistinguishablefrom those of animal origin.

RErsENCES

1. FDA contract 223-73-7210.2. Anderson, E. S., G. 0. Humphreys, and

G. A. Willishaw, "The Molecular Relatednessof RLfactors in Enterobacteria of Humanand Animal Origin," Journal of GeneralMicrobiology, 91:376-382, 1975.

3. Grndlay, N. D. F., G.- 0. Humphreys,and E. S. Anderson, "Molecular Study ofR-factor Compatibility Groups," Journal ofBacteriology, 115:387-398, 1973.

4. Crosa, J., J. Olarte, L. Mlatry, L. Lut-tropp and IM. Penaranda, "Characterizationof an R-plasmid Associated with AmpicllinResistance in Shigella dysenteriae Type IIsolated from Epidemics,". AntimicrobialAgents and Chemotherapy, 11:553-558.

(v) Hqeards. While the presence ofantibiotic-resistant E. coli in the intesti-nal tract of humans may generaly causeno immediaete problems to an individ-ual, under certain circumstances it maylead to dangerous situations. For exam-.ple, E. coli is the most usual cause ofhuman urinary tract infections andcommonly arises from an individual'sown intestinal flora. Sulfonamides aregenerally the drug of choice for treat-ment of urinary infections; however, asignificant number of infections withsulfonamide-resistant strains are nowreported.

Antibiotic-resistant E. coil in thebowel of man also constitute a reser-voir of organisms capable of transferringresistance to intestinal pathogens. Per-haps the greatest hazard to humanhealth arising from the use and misuseof antibiotics is the large reservoir ofplasmid borne resistance genes in thenormal intestinal flora of animals andman and their presence in the environ-ment-resistance that can be tranferredfrom nonpathogenic to pathogenic orga-nisms..

In recent years the emergence of R-plasmid-mediated resistance in patho-gens has been identified in epidemicsaround the world. A strain of Salmonellatyphi carrying an R-plasmid-determin-ing resistance to chloramphenicol causedan epidemic of typhoid fever in Mexico.Transferable chloramphenicol resistance

has also become common in S. typhi iso-lated in Indian, Vietnam, and Thailand(Ref. 1). The recent epidemic of drug-resistant Shigella dysentcriae Infectionin Central America (Ref. 2) Is anotherexample of an epidemic disease whichwas no longer susceptible to treatmentby antibiotics that had previously beenuseful. Plasmid-mediated resistancehas been reported in strains of Borde-tella bronehiseptica (Ref. 3), and FDAscientists have demonstrated plasmid-mediated resistance to penicillin, tetra-cycline, streptomycin, and sulfonamidoin strains of Pasteurela multoclca andPasteurella haemolytica, both of whichcause serious diseases in animals (Refs.3 and4).,

Recent studies (Refs. 5 through 12)have also shown that the genes specify-ig resistance to ampicillin, tetracycline,kanamycin, chloramphenicol, trlmetho-prim, and streptomycin reside on DNAsequences that are able to tranlocateor move from plasmid to plasmid as adiscrete unit, or from a plasmid to thebacterial chromosome. Therefore, In ad-dition to movement of resistant bacteriafrom animals to man and the transfer ofR-plasmlds between bacteria, the genesthat reside on the plasmlds can them-selves migrate from plasmid to plasmidby translocation. Furthermore, an R-plasmid does not have to be stably main-tained within a cell to donate Its resistantgenes to a recipient chromosome or anindigenous plasmid.

Tetracyclines are the drug of choicefor most infections caused by myco-plasma, rickettsia and chlamydia. Someof these organisms (e.g., the causativeagents of Psittacosis, Ornithosis, andQ-fever) are known to spread fromanimals to man. Under antibiotic pres-sure, the development of tetracycline re-sistance has been shown in Coxiella bur-netif, the pathogenic rlckettsla causingQ-fever (Ref. 13). Mycoplasmas recentlyhave been shown to possess plasmids ofas yet unknown function (Ref. 14).Tetracycline-resistant mycoplasmas havebeen isolated from the urinogenitaltract of patients with various disorders(Refs. 15 and 16). It Is uncertain whetherthis resistance is chromosomal or pas-mid-mediated. However, there Is ce'-tainly a possibility of animals underantibiotic pressure acquiring tetracy-cline-resistant mycoplasmas, and of thetranslocation of chromosomal tetracy-cline resistance to R-plasmids. There arerecent data indicating that some Myco-plasma may be pathogenic for a widerspectrum of life than was originally bo-lieved (Ref. 17).

Most bacterial species possess Indige'nous plasmid gene pools. In fact, piss-mids have been found In all species ofbacteria which have been examined, Thefunction of these plasmids Is often un-known, but they could serve as effec-tive recipients for the insertion of trans-locatable genes. The recent emergenceof ampicillin-resistant strains of 'aemo-philus influenzae and penicillin-resistantstrains of Neisserita gonorrheac repro-sent alarming examples of the extensionof the R-plasmid gene pool (Refs. 18 and


56272



NOTICES

19). The resistance genes found in bothspecies are identical to those previouslyfound only in E. coli and other entericorganisms.

The -World Health Organization prophet-cally warned (Ref. 20) :

The point will ultimately be reached atwhich the transfer of resistance to pathogenabecomes inevitable and the larger the pool.the greater is this possibility. Moreover, thewider the distribution of R+ (R-factor)enterobacteria the greater the possibilitythat R-plasmids may emerge that can crossbiological barriers so that they can perhapsenter bacterial species and genera ap--parently widely different from their originalenterobacterial hosts.

RXFra=IcsS

1. Anderson, E. S, G. 0. Humphreys, andG. A. Willshaw, '"e Molecular Relatednessof R-factors in Enterobacterla of Human andAnimal Origin," Journal of General Micro-biology, 91:376-382, 1975.

2. Gangarosa, E. J., D. Perera, L. Mata, C,Morris, G. Guzman, L. Relier, "EpidemicShiga Bacillus Dysentery in Central America.M. Epidemiologle Studies in 1969." Journal ofInfectious Diseases, 122:181, 1970.

3. Terakado, N., H. Azechl, K. Ninomlya,and T. Shimizu, "Demonstration of P-factorsin BordeteUla bronchiseptica Isolated FromPigs," Antimicrobial Agents and Clemo-therapy, 3:555-558, 1973.

4- Silver, P. P., B. Leming. C. HJerpe,"R-plasmids in Pasteurella multoci-a " Ab-stracts of Bacteriological Proceedings of theAmerican Society for Microbiology. p. 141,1977.

5. Earth, P. T., N. Datta, n. W. Hedges, andN. J. Grinter. "'ransposition of a Deosyribo-nucleic Acid Sequence Encoding Trimethop-rim and Streptomycin Resistances fiom R483to Other eplicons," Journal of Bacteriology,125:800-810, 1976.

6. Berg, D. E., 1'. Davies, B. Ailet and J. D.Rochaiz, "Transportation of R-factor Genesto Bacteriophage," Proceedings of the Na-tional Academy of Sciences of the U.S.A.,72:3628-3632, 1975.

7. Foster, T. J., T. G. B. Howe, and AL H.Richmond, "Transocation of TetracyclineResistance Determinant from R100-! to theEscherichia coi IK-12 Chromosome," Journalof Bacteriology, 124:1153-1153.

8. Gottesman, AL M., and J. L. Rosner,"Acquisition of a Determinant for Chloram-phenicol Resistance by Coliphage Lambda,"Proceedings of the National Academy of Scf-ences of the U.S.A., 72:5041-5045, 1975.

9. Hedges, R. W., and A. E. Jacob, 'TErans-position of Ampicillin Resistance from RP4to Other Replicons." Molecular and GeneralGenetics, 132:31-40. 1974.

10. Heffron, R., R. Sublett. n. V/. Hedges,A. Jacob, and S. Falkow, "Origin of the TElUBeta-lactamase Gene Found on Plasmids,"Journal of Bacteriology, 122:250-256, 1975.

11. Kleclmer, N., R. K. Chan, B. K. Tye,and D. Bostein, "Mutagenesis by Insertionof a Drug Resistance Element Carrying anInverted Repetition," Journal of MolecularBiology, 97:561-575,1975.

12. Kopecko, D. J., and S. N. Cohen, "Site-Specific RecA Independent Recombinationbetween Bacterial Plasmids: Involvement ofPalinodromes at the Recombination Loci,"Proceedings of the National Academy of Sci-ences of the U.S.A., 72:1373-1377,1975.

13. Brezina, R., et al., "Selection of aChlortetracycine Resistant Strain of Coxielaburnetti," Acta Virologica, 19:496, 1975.

14. Liss, A. '"PIasmlds of Acholeplasmalaidlawit, I. Preliminary Char=cterization,"D53. Abstracts of 76th Annual Meeting of

the American Society for Microbiogy. At,lantlc City, NJ, 1970.

15. Ford, D. K., ot al.. "Non-spselflo Ureth-rits Assoclated with a Tetracycllne-RBeLstantT-mycoplama." British Journal of Ve--trcalDiseasc, 59:373-374,1974.

16. Spaepen, r.., at al., "Tetracyclino Ecst-ant T-mycoplssma From Patients vith i1:-.tory of Reproductive Falur,:* AntlmifrobfalAgents and Chcimothcrap-j, 9:1012-1010, 1970.

17. Tully, J.. R. Vjitacomb, H. Clara, aD. Vllilamson. "Pathozenec Liycoplama:Cultivation and Vertebrate Pathoenlclty" ofa New Sporoplasma," Science, 105:8)2-8511977.

18. Elvell, L. P., J. dcGraal, D. scllcrt,and S. Falkow, '"Pl -mid-Linked AmpicilinResistace In Ifacnwmlaflus infucnzac typob," Infcction and Irm'Jnity, 12:404-410,1975.

19. Elell, L. P., M. Roberts, L. V. M, yer,and S. Fall:ow. "Plasmld-mcdla!ed B-lacta-maze Production In Ncssetri ganorrhe c,"Antimicrobial Agents erl Clienofficrapy,11:528-533,1977.

20. W.H.O. Report, 1970. "Publc RcAlthAspect of Antibiotic Rec!stanca In Envlron-ment," Regional Ofic for ruroe. Cnpn-hagen.

4. Director's conclusions. The holdersof the approved NADA's for subthera-peutic tetracycline-containing productswere required to show that the sub-therapeutic use of tetracycline doesnot increase drug rezis:tant (i.e., increasethe pool of R-plasmld-bearing) orga-nisms in animals. If they were unable toshow that subtherapeutic tetracyclineuse does not increase the pool of R-pla3-mid-bearing organisms in animals, theholders were then required to show thatthe R-plasmlds are not transferablefrom animals to man. They failed to doany of this.

The evidence shows that the pool ofR-plasmid-bearlng organisms, particu-larly in E. coli, Is increasing, and thatthe increase Is due at least In part tothe subtherapeutlc use of the tetracy-clines in animal feed. Further evidenceshows that E. coi contribute their R-plasmids to man through his direct con-tact with animals, through his directcontact with E. coli-contaminated food.and by widespread presence of the R-plasmids In bacteria in the environment.Studies also show that there is no strictdistinction between the E. colt that colo-nize animals and those that Infect man.On the contrary, there is considerableoverlap in these strains, and there isalso an overlap in the enteric bacterialR-plasmid population in humans andanimals. This evidence is derived fromepidemiology studies, bacterial ingestionstudies, and compatibility studies of thenormal Intestinal flora of man and anl-mals. These bacteria may donate theirR-plasmid to pathogens n man andanimals even when transient, and theNADA holders have submitted no evi-dence on the degree of colonization, Ifany, that is necessary for thLs transferto occur. Accordingly, the Director con-cludes that the holders of the approvalsfor the subtherapeutic tetracycline-con-taining products for use in animal feedshave failed to show that extensive sub--therapeutic tetracycline use satlsfle3 the

requirements of 5 558.15 and criterion Iof this notice.

D. SHEDDIG AND r.EX, 313TCH CHAPACTZM3-TICS 0' 5M.!IONELLA (CIJTE'J0oZ 21

1. Eac,:ground. Under human and.nimal safety criterion number 2. theNADA holder.s must show that an anti-bacterial druj used in animal feed shallnot cause a significant increase In thequantity, prevalence, or duration of Sal-monella shedding or an increase in theantibiotic resistance characteristics ofsalmonellae. The Bureau of VeterinaryI'.edlcine emphasized this criterion be-cause (a) Independent studies indicatedthat use of an antibiotic had caused anincrease in Salmonella shedding in med-icated humans (Ref. 1) ; and (b) an epi-demic of a specific virulent (phage-type29) Salmonella typhimuritn had oc-curred in Great Britain after prophyIac-tic use of antibiotics in cattle feed. Thisresulted in human fatalities (Ref. 2).

Askeroff and Bennett (Ref. 1) pre-sented data on the effect of antibiotictherapy on the excretion of SalmonellaIn fecez of humans infected with acutesalmonellozis. After a large S. tzn'himu-numr epidemic caused by eating contam-inated turkey, the authors examined thefeces of untreated patients and patientstreated with tetracycline, ampiclin, andchloramphentcol for Salmonella, andthey determined the antibiotic susceptl-bility of the S. typhfrmLTiurs strains. Pa-tients generally received the recom-mended regimen of antibiotic therapy11 gram per day). Fecal samples fromG7 patient- not receiving medication and185 patients treated with antibioticswere examined. Of the patients treatedwith antibotics, 65 percent were shed-ding Salmonella 12 days after Infection,and 27 percent were positive (shedding)31 days after infection. In the untreatedpatients, however, Salmonella sheddingvmas ob:erved In only 42.5 percent at day12 and l1.5.percent at day 31.

Therapy also favored the acquLitionof antiblotic resistance by the infectingstrain isolated from poultry, which ini-tielly had been susceptible to antibiotics.Eighteeen of the 185 patients receivingantiblotlcs excreted resistant Salmonet,while none of the 87 untreated patientsexcreted resistant Salmonella (P<.05) .The antibiotic resistance acquired in theSalmonella strain was shown to betransferable.

Anderzon (Ref. 2) carefully docu-mented the buildup of a reservoir ofmultiply antibiotic-resistant Salmoner-lac in the outbreak of Salmonella typi-murium. phage-type 29 in calves inBritain from 1963 to 1959. Antibioticswere used both therapeutically and pro-phylactically in crowded feed lots. Aseach new antibiotic therapy va tried.a new antibiotic resistance emerged inthe pathogen, and eventually the S.typhimurium strain carried resistance toa wide range of antibiotics. In additionto disease and death In cattle, shedding(excretion) the multiply resistant S.typhimurium caued Infection and even

FEDERAL REGISTER, VOL 42, NO. 204-lXIDAY, OCTOSER 21, 1977

569-73



NOTICES

some deaths in humans in contact withthe animals.

RE ='RENCES

1. Askeroff, B. and J. V. Bennett, "Effect ofAntibiotic Therapy in Acite SalmonelIosison the Fecal Excretion of Salmonella," NewEngland Journal of Medicine, 281:631-640,1969.

2. Anderson, E. S., "Ecology and Epideml-ology of Transferable Drug Resistance," inCiba Foundation Symp. Bacterial EpIsomesand Plasmids, pp. 102-114, Churchill, Lon-don, 1969.

2. Criterion-(a) Shedding. Controlledstudies were to be designed to deter-mine whether the administration of anantibacterial drug at subtherapeuticlevels would result in an increase in therelative quantity, prevalence, or dura-tion of shedding of Salmonella whichare pathogens in animals. Salmonellaare often found in the intestinal tractof man and animals, and the small in-testine and colon are the primary sitesof multiplication. After penetrating theepithelial lining, they multiply and elicitan inflammatory response. Most Salmo-nella infections are limited to the gastro-intestinal tract, producing the clinicalsymptom termed "gastroenteritis." Oneof the more common strains, Salmonellatyphimurium, causes dseases in bothman and animals.

When an animal Is infected with thesebacteria, the live organisms are excretedin the feces ("shedding"). The-quantityof Salmonella in the feces can be deter-mined by a bacteriological proceduretermed a "standard plate count." A spe-cific amount of fecal material is dilutedand spread on a semi-solid bacterialgrowth medium which is selective forthe growth of Salmonella. After a suffi-cient time for growth, individual coloniesare counted and recorded as the number -of Salmonella per gram of wet feces.The proportion of antibiotic-resistantSalmonella in fecal specimens is inde-pendent of the quantity of Salmonellashed.

(b) Resistance characteristics. Con-trolled studies also were to be designedto determine whether the administration'of oxytetracycline and chlortetracyclineat subtherapeutic levels would result inan increase In. the proportion of anti-biotic-resistant Salmonella. Salmonellaisolated from feces can be tested fortheir susceptibility to various antibioticdrugs by standard procedures. Escherich-ia coli, a normal component of the intes-tinal flora, were also examined to deter-mine their resistance spectrum since oraladministration of certain antibiotics,whether at therapeutic or subtherapeu-tic levels, has been shown to result in anincreased proportion of indigenous E.Coli that contain R-plasmids which canbe transferable to other E. Coli or toSalmonella. Antibiotic resistance may bemeasured by use of an antibiotic incor-porated into the bacterial growth mei-um or by standardized antibiotic discs.

3. Industry studies in chickens on theeffects of subtherapeutie tetracycline usein animal feed-(a) American CyanamidCo.-(i) Experimental design. This studywas designed to determine whether there

are increases in the quantity, duration,and prevalence of -Salmonella shedding.in chickens, caused by subtherapeuticchlortetracycline in feed. Day-old chickswere fed either chortetracycline at 200grams/ton of feed or a nonmedicated dietfor 58 days after inoculation with achlortetracycline-sensitive strain of Sal-monella typhimurium (S. typhimurium).The chicks were divided into fourgroups: two environmental controlgroups, a nonmedicated control group,and a medicated (treatment) group.

(ii) Summary. While there were nosignificant differences in the quantity,prevalence, and duration of shedding be-tween the medicated (treatment) groupand nonmedicated groups, there werestatistically significant differences in theantibiotic-resistance of the Salmonellashed by these groups. Chlortetracyclineresistance of the Salmonella shed bybirds fed subtherapeutic levels of the an-tibiotic showed a statistically significantincrease when the ratio of antibiotic-re-sistant Salmonella shed to the number ofbirds excreting Salmonella is calculatedfor the nonmedicated and treatmentgroups. The ratio increased from 27percent drug-resistant on day 1 toapproxinjately 95 percent from day 22until the' end of the study. Further-more, when the total number of birdsexcreting antibiotic-resistant Salmo-nella is compared to the total numberof birds in the study, medicated birdsexcrete significantly higher percentagesof antibiotic-resistant Salmonella thannonmedicated birds (P<001), and theexcreted Salmonella predominantlyshowed one particular antibiotic resist-ance-tetracycline, streptomycin, kana-mycin, and neomycin.

(iii) Director's analysis. Comparingthe number of Salmonella-positive fecalsamples to the number of birds excretingSalmonella, the Director finds there areno signifi~ant differences between medi-cated groups and nonmedicated controlgroups when the tetracycline-sensitiveSalmonella strain was the Infectingagent. However, the percentage of anti-biotic-resistant Salmonella isolated frombirds given chlortetracycline increasedrapidly, remained at 93 to 95 percentfrom day 22 of the study until the-con-clusion; and the majority of the samplessimultaneously- developed resistance tostreptomycin, kanamycin, tetracycline,and neomycin in the samples of birdstreated with chlortetracycline.

(b) Rachelle Laboratories, Inc.-()Experimental design. This study was de-signed to measure the quantity, preva-lence, and duration of Salmonella shed-ding by chickens fed subtherapeuticchlortetracycline for 28 days postinocu-lation. The chickens were divided intotwo environmental control groups ofthree birds each, and two groups thatwere orally inoculated with a chlortetra-cycline-sensitive strain of S. typhimu-rium. The treatment roup received 100grams of chlortetracycline/ton of feed.

(ii) Summary and the Director's anal-ysis.'During the first 8 days of infection,chlortetracycline at the 100 grams/tonof feed reduced the quality and preva-

lence of Salmonella shedding; however,by the 10th day, Salmonella shedding Inboth the nonmedicated control and thetreatment groups was comparable. More-over, the study again showed that ad-ministration of subtherapeutic levels ofchiortetracycline to chickens resulted Inan, increase in the percentage of anti-biotic-resistant Salmonella isolated fromthe feces.

(c) Pfizer, Inc.-(i) Experimental de.sign. This study was designed to measurethe prevalence, quantity, and durationof Salmonella shedding In 8-day-oldbroilers fed 200 grams of oxytetracy-cline/ton of feed (subtherapeutlc) for 20days after inoculation with a tetracy-cline-sensitive strain of S. typhimurium,Ten birds were assigned to a treatmentgroup, 10 to a nonmedicated (active)control group, and 3 each to 2 envIron-mental control groups. Unlik ; the Ameri-can Cyanamid and Rachelle studies onthe effect of subtherapeutlc tetracyclineof Salmonella shedding by chickens,Pfizer measured the pretest level of anti-biotic resistance In the indigenouschicken coliforms (E. col). It found thetetracycline resistance level to be 25percent.

(ii) Summary. The two environmentalcontrol groups were Salmonella-froethroughout the study. In the nonmedi-cated group and the treatment group, theSalmonella population decreased withtime, although the decrease occurredmore rapidly in the medicated group. Theprevalence of the S. typhimurlum in thefeces of the medicated birds was lessthan the prevalence in the nonmedicatedbirds (17/170 (24 percent) v. 50/70 (85percent) (P<0.001)). But there was asignificantly higher percentage (P<0.01)of tetracycline resistance in Salmonellaisolated from medicated animals (21/32(66 percent)) than Isolated from thenonmedicated controls (0/263). The re-sistance in the Salmonella isolates waslimited to oxytetracycline and strepto-mycin with but one exception (ampi-dimlt).

(iii) Director's analjsls. The Directordoes not disagree with certain conclu-sions drawn from this study by Pfizer.Based on the information submitted, thestudy appears to show that subthera-peutic use of oxytetracycline did not In-crease the quantity of Salmonella shedin the feces of medicated birds. Nor didthe quantities found in the liver, spleen,or cecal tissues differ. Also, the durationof Salmonella shedding and the prev-alence of the infections were not greaterin the medicated chickens than In thenonmedicated control birds. Neverthe-less, the study fails to show that sub-therapeutic tetracycline is safe for usein feed for chickens since the percentageof resistant Salmonella Is increased Inmedicated chickens compared to non-medicated birds as in the AmericanCyanamid and Rachelle studies. NeitherAmerican Cyanamid nor Rachelle, how-ever, measured the prestudy levels ofantibiotic resistance in the indigenousE. coll, which was required by F DAguidelines. Since the Director and allothers in the area are concerned that


56274



NOTICES

indigenous E. col are a primary sourceof R-plasmids for the transfer of anti-biotic resistance to pathogens, FDAadded this point to its test guidelines.Failure to conduct the study properly isa glaring if not fatal omission andnegates its value.

(d) Director's conclusions. In all threestudies the percentage of antibiotic-resistant Salm6nella isolated from chick-ens fed subtherapeutic levpls of tetra-cycline was higher than that of chickensfed antibiotic-free feed. Moreover, thesponsors failed to use an enrichmentprocedure for culturing the bacteriawhich, as the Director explained in hisnotice for penicillin published in theFEDERAL REGISTER of August 30, 1977 (42FR 43782), may have biased the results.For these reabons, the Director concludesthat the studies have failed to demon-strate conclusively that the subthera-peutic use of tetracycline in chicken feedis safe.

4. Industry studies in swine on theeffects of subtherapeutic tetracycline usein animal feed-a) Tetracycline alone-(D Experimental design. Four holders ofapproved NADA's, American Cyanamid,Rachelle, Diamond Shamrock Corp., andPfizer, submitted four studies of similardesign to measure the effect of subthera-peutic tetracycline in feed on the quan-tity, prevalence, and duration of Salmo-nella shedding by swine. AmericanCyanamid and Rachelle studied theeffect of chlortetracycline at 200 grams/ton of feed, Diamond Shamrock studiedchlortetracycline at 100 gr-ms/ton offeed, and Pfizer studied oxytetracyclineat 150 grams/ton of feed. In each study,10 swine were assigned to a group givenantibiotics and 10 to a nonmedicatedgroup. Swine in both groups were inocu-lated with tetracycline-sensitive Salmo-nella. Only the Pfizer study lacked medi-cated and non-medicated environmentalcontrol groups not infected with Salmo-nella. The Pfizer study was conductedfor 37 days postinocuation, and theothers were for approximately 4 weeks.

(ii) Summary and the Director'sanalysis. When the Director comparedresults of Salmonella isolates from hemedicated and nonmedicated swine, hefound that the swine fed subtherapeutictetracycline showed no increase in Sal-monella colonization or shedding (preva-lence, duration, or quantity). But thestudies illustrate a general pattern-statistically significant increases in thepercentage of antibiotic-resistant Sal-monella isolated from medicated swinecompared to those isolated from the non-medicated controls (P<0.01 or 0.05).

(b) Tetracycline in combination witsulfonamides and penicilin-(t) Experi-mental design. American Cyanamid, Dia-mond Shamrock, and Rachelle each sub-mitted a study to measure the effect onSalmonella shedding of a widely usedcombination of subtherapefitic antibac-terials CSP, (chlortetracycline 100grams/ton, sulfonamide 100 grams/ton,and penicillin 50 grams/ton In swinefeed) on Salmonella shedding. The studyalso attempted to measure the change inpercentage of antibiotic resistance In In-

digenous E. coN and inoculated Salino-nelltr Again, the study designa werecomparable. In each study, 10 swine wereassigned to each group fed the combina-tion (no groups received the individualcomponents of the combination) and 10

-were assigned to a nonmedicated controlgroup. Swine In thee groups were theninfected with a tetracyclIne-censitivestrain of S. typhimuriumn; the swinewere monitored for 28 days postinfec-tion. Each study also had two environ-mental control groups, containing 3 to10 nonmedicated swine which wre notexperimentally infected.

(i) Summary and the DVrcctor'sanalysis. In no study did the anibioticcombination increase Salmonella -hd-ding in the swine. However, n each studyantibiotic resistance increased in theSalmonella Isolated from Swine fed theCSP combination compared to nonmedl-cated swine. American Cyanamid andRachelle failed to make prestudy deter-minations of the antibiotic resistance Inthe indigenous E. coli in any or all Swine,and n the Diamond Shamrock study.the background level of drug resistancen the E. coli was extremely high, 80 to

100 percent. Information on the E. coilresistance is cruclarto assessing the riskof harm associated with subtherapeutictetracycline. The E. colt may serve as areservoir of transmissible R-plasmids forpathogens. An initially very high back-ground level of resistance will make itdifficult to detect any further develop-ment of antibiotic resistance In the B.co/i during the course of exp oure to themedicated feed.

The studies were conducted for only28 days postinfection, until the swinewere aproximately 10 weeks old, whichdiffers from the conditions under whichswine are commercially grown for mar-keting. Normally, swine are fed antibi-otics until 16 weeks of age, and the Direc-tor has no basis for extrapolating theresults on shedding for more than the28 days that the study was actually con-ducted. In fact, an extrapolation basedon trends in some of the studies and theresults from similar studies In the liter-ature to be discussed below would s-gest that the prevalence, duration, andquantity of Salmonella shedding wouldincrease after a longer time period in theswine fed subtherapeutic leveL- of anti-biotics.

(c) Director's conclusfons. Based onthe results of these studies, the Directorconcludes that the subtherapeutic tetra-cycline has not been conclusively shownto be safe in swine. The ue of subthera-

eutic tetracycline in swine feed, in thepresence of R-plasmids. again causes anincrease n shedding of antibiotic-resist-ant Salmonella, although enrichmentprocedures were not used n culturingthe bacteria.

5. Industry studies in cattle on theeffects of subtherapeutie tetracycline usein animal fecd-(a) Studies of tetracy-cline and tetracycline combinations incattle and calves-a) Experimental de-sign. Five drug firms-American Cyana-mid. Diamond Shamrock, Rachelle, Vita-min Premixers of Omaha, and Pfizer-

conducted rix studies on SalmoneaShedding in calves fed subtherapeutictetracycline and a subtherapeutic com-bination of tstracycline and sulfametha-zinc. American Cyananid, DiamondShamrock, and Rachell studied chlorte-tracyulino at 350 mlligranis/bead/day,while Vitamin Premixers of Omaha(VPO) studied chlortetracycline at 200mulligranslcalf/day. Pfizer conducted astudy of oxytetracycline at 100 grams/ton of feed. American Cyanamid alsoperformed a study on the effect of chlor-tetracycline and sulfamethazine in com-bination each at 350 mIgrams/headday.Ia general, the experimental designs

were similar to the following plan:

fa t!s fzcd

4 No~

Th9 calves ranged in age from 6 to 8weekd, and they were housed in animalpens in a. variety of groups from one ani-mal per pen to -ll animals in a treatmentgroup per pen. Also, American Cyanamidused the same nonmedlcated control ani-maLs for both Its study of chlortetracy-cline alone and chiortetracycline plussulfamethazine. In three studies, thecalve- were infected with bovine S. typthi-murlum ATCC 14028 which has a well-characterized R-plasmid recipient abil-ity; In the Pfizer oxytetracycline study,another strain of S. typhimurium with awell-characterized recipient ability wasused. But In two studies the sponsorsprovided no details either of the bac-teria's ability to transfer or receive R-plasmids. Finally, the Salmonella orga-nisns ued In all the studies were sensi-tive to both the antibiotics used in theLtudy and to antibiotics in general.

(I) Summearg. The Racheile submis-zion contained no information on E. coNre.ostance, but the background level ofantibiotic resistance in E. col in theother studies generally ranged from 63to 100 percent (American Cyanamidmeasured only 1 calf per pen of 5 ani-mals). The prevalence of Salmonelrarhed In all the stdies was less in themedicated groups than in the nonmedi-cated control groups, and the medicatedgroups generally excreted fewer $al-monella. When the American Cyanamidchlortetracycline-alone study was termi-nated, however, more calves fed sub-therapeutic chlortetracycline than non-medicated calves were shedding Sal-monella. This was also observed In theDiamond Shamrock study.

(iWi) Director's analysis. In those casesIn which the initial drug resistance ofthe E. cof was determined, the Directorfound a correlation between the initiallyhigh antibiotic resistance in the E. coUand the development of antibiotic re-sIstance in Salmonella by transfer ofR-pla-,lds, whether or not the calves

FEDERAL REGIsTER, V6L 42, NO. 204-RIDAY, OCTOBER 21, 1977

562755



were exposed to antibiotics. For example, tions concerning the widespread use ofE. coli Isolated from two cattle in the subtheraputic tetracycline In animalAmerican Cyanamid study were 100 per- feed. Rather, the studies demonstratecent tetracycline-resistant, and both that subtherapeutic tetracycline use incattle developed drug-resistant sal- animal feed causes an increase ix anti-monellosis. Unfortunately, American biotic-resistant E. coli as well as an in-Cyanamid did not measure the back- crease in the percent of shed Salmonella.ground level of E. coli in all cattle, as that are antibiotic resistant. Studies alsorecommended in FDA guidelines; there-_' indicate that R-plasmid-bearing E. colifore, the Director cannot correlate the donate antibiotic resistance'plasmids todevelopment of antibiotic resistance in Salmonella. Investigations by independ-the E. coli with the development of anti- ent scientists have produced similarbiotic resistance in Salmonella. Despite findings (Refs. 1, 2). Patterns of drugthis, the Director identified the pervasiv4 resistance seen in E. coli and Salmonellapattern already observed in the chicken isolates from man and animals are sim-and swine studies when there is a high ilar and develop in a like manner. E. colilevel of antibiotic resistance in the E. coli first develops R-plasmid-mediated an-prestudy; antibiotic resistance (i.e. R- tibiotic resistance, and then the Salmo-plasmids) generally transfers to the Sal- nella develop a similar and frequentlymonella, either remaining high through- identical pattern of resistance. Studiesout the study or increasing in the medi- also show that the number of R-plasmid-cated animals, bearing strains of pathogenic Salmonella

In three experiments the percentage of are increasing. More importantly, theantibiotic-resistant coliforms was higher number of multiply resistant strains isin the calves'fed subtherapeutic levels of increasing.chlortetracycllne than in the nonmedi- (a) Surveys-(i) Neu, Cherubin,cated control, and in one study the dif- Longo, Flouton, and Winter studies. Re-ference was statistically significant at cently, Neu et al. (Ref. 3) examined theP<0.05. antimicrobial susceptibility of 718 Sal-

In the two studies where the sponsors monella isolates from humans collectedfollowed the changes in antibiotic resist- at a New York hospltal and 688 isolatesance in the eoliforms, they observed dif- from animals. They compared the cur-ferences between the nonmedicated and rent (1973) antibiotic resistance in hu-the medicated calves. The tetracycline man Salmonella isolates with data fromresistance in coliforms in the medicated a previous study which they had con-and unmedicated animals remained at ducted in 1968-1969. They also comparedapproximately 80 percent throughout the resistance patterns of animal Sal-those studies; nevertheless, differences in monella isolates from animals obtainedthe resistance to ampicillin, streptomy- -from the National Animal Disease Centercin, and sulfathiazole were observed. Al- during 1973.though the data are sparse, in every case Thirty percent of all human isolatesresistance in the nonmedicated control collected in 1973 were resistant to one orgroup decreased while the resistance In more antibiotics. S. typhimurium, a sero-the medicated group increased or re- type common to .man and animals, wasmained constant, e.g., the- resistan6e to the most frequent serotype isolated; 58ampicillin went 14 percent to 30 percent. percent were resistant to at least oneSimilarly in the Vitamin Premixers of __antibiotic. More than 50 percent of the S.Omaha study, the percentage of col- typhimurium were resistant to four toforms resistant to chlortetracycline, five antibacterials. Resistance to tetra-dihydrostreptomycin, and oxytetracy.- cycline in S. typhimurium had increasedcline declined in the nonmedicated con- from 12.5 percent in 1968-1969 to 44.8trol group, but it remained constant for percent in 1973, about.a 3.6-fold increase.the chlortetracycline and dihydrostrep- When these results were compared withtomycin, and increased slightly for oxy- a 1965 survey conducted in the Eastern,tetracycline in the cattle given subthera- United States by Gill andHook (Ref. 4),peutic tetracycline, the authors found that the percentage

(b) Director's conclusions. The studies of isolates of all serotypes resistant toof subtherapeutle chlortetracycline in tetracycline and streptomycin had ap-cattle pose and fail to resolve the similar proximately doubled. Antibiotic-resistantproblems raised in chicken and swine strains of S. typhimurium had increasedstudies. Subtherapeutic chliortetracycline from the 19 percent reported in the Gillcauses an increase in the percent of R- and Hook study to 58 percent in the 1973plasmid-bearing Salmonella shed. More- study of Neu et al., about a 3-fold in-over, these studies identify another criti- crease. Moreover, the resistance to am-cal problem associated with the use of picillin, tetracycline, chloramphenicol,subtherapeutic antibiotics in animal streptomycin, sulfisoxazole, and kana-feed. Indigenous E. coli, which have re- mycin was transferable among the vari-sistance plasmids, are selected for and ous Salmonella strains. -contribute their R-plasmds to the In animals S. typhimurium accountedpathogenic * S. typhimurium. Accord- for 70 percent of the isolates, and 80 per-ingly, the Director concludes subthera- cent were resistant to one or more anti-peutic chlortetracycline has not been microbial agents. R-plasmids were foundshown to be safe for use in cattle feed, in 86 percent of the S. typhimurium, and

resistance to ampicillin, tetracycline,6. Information from other studies re- chloramphenicol, streptomycin, sulfl-

lating to Salmonella and E. coli antibi- soxazole, and kanamycin was transfera-otic resistance. The studies submitted by - ble. Generally, the resistance patternsthe holders of the approved NADA's fail were similar to those encountered in theto answer conclusively the safety ques- Salmonella isolated from humans.

The authors concluded that the highincidence of transferable resistance inman and animals suggests that most re-sistant strains seen today contain com-plete R-plasmds, and that strains un-able to mobilize resistance determinantsare less common than was formerlythought. They further concluded thatcomparison of the resistance of salmo-nella isolates from humans with that ofSalmonella from animals shows that tot-racycline resistance is greater among thestrains from animals, as in the case withsulfonamide and streptomycin resistance,While the resistance to ampicillin ishigher in S. tYphimurium strains fromhumans than from animals, the reverseis true for other serotypes. This differ-ence may reflect the greater current useof tetracyclines, sulfonamides, and strep-tomycin in animals.

Finally, the authors conclude that thesurvey clearly demonstrates that resist-ance to antibiotics Is Increasing in Sal-monella isolated from both humans andanimals, and since there are great simi-larities in the resistance patterns ofhuman and animal Isolates, it would beuseful to know whether the R-plasmidsare of a similar nature since this WouldSuggest that animal strains have con-tributed to the human pool of resistantorganisms. This question has since thenbeen examined and certain R-plasmidshave been found to be similar In bothman and animals. (See Part IVA(3) (c)(iv) above.)

(ii) CDC reports. When the Center forDisease Control (Ref. 5) compared a1968 study on antibiotic resistance inSalmonella isolated from hospitalizedpatients with a more recent study (Ref.6), results similar to those seen by Neuet al. were found; the number of anti-biotic-resistant Salmonella showed amarked increase as can be seen from thetable below.

1057 1075400 stralns 75 strain

Resistance to one or moreantiblotls:

S. typhimurium ------ 41.1 peL..... 69,4 pet,Other scrotypes ---------15.8 ......... 43.9.

oAll strains .......... 22.2 ......... 49.7,Resistance to 2 or more an- 15.0 ......... 20.5,tibiotics. (60 strains) (200 Strains )Resistance to 6 ormoro an- 0.8 .......... 9,2.

tiblotics. (3 strain) (69 strains).

Nine antibiotics were used In commonin both studies-colistin, naldixic acid,sulfonamides, streptomycin, kanamycin,tetracycline, chloramphenicol, ampidll-lin, and cephalothin. In the 1976 studygentamycin and bactrim substituted forthe neomycin and nitrofurantoin in the1968 study. The substitutions fail to rx-plain the increase in antibiotic resistancesince frequency of resistance to the sub-stituted drugs was actually lower thanthe frequency of resistance to thosetested initially (gentamycin (1975), 0.1percent to neomycin (1967), 1.2 percent;bactrim (1975), 1.3 percent to nitro-furantoin (1967), 2.5 percent).

Between 1968 and 1975, overall anti-biotic resistance in Salmonella strainsmore than doubled, from 22.2 percentto 49.7 percent. Furthermore, although


56276 NOTICES



NOTICES

resistance in S. typhimurium increased1.7 times during this span, other sero-types of Salmonella showed a greaterincrease in antibiotic resistance-a 2.8-fold increase. Multiple antibiotic resist-ance increased significantly (from 15.0percent to 26.5 percent), and the numberof "super resistant strains," i.e., thosewith R-plasnids carrying resistance to6 or more antibiotics, increased drama-tically from 0.8 percent in 1967 to 9.2percent in 1975. Perhaps more impor-tantly, while the super-resistant strainsaccounted for only 0.5 percent of the to-tal number of pathogenic isolates in1968, they accounted for over one-thirdof all multiply resistant strains isolatedin 1975 (34.5 percent).

The 1968 study included both noso-comial- and community-acquired infec-tions; therefore, some isolates in thatsurvey may have been obtained afterpatients were treated with antibiotics.The 1975 figures, however, are based onisolates obtained only from untreatedcommunity-acquired infections and arethus particularly significant. These in-fections were likely to have occurred asa result of exposure to contaminatedanimal products rather than as a resultof unsuccessful or inappropriate ther-apeutic treatment of the patient. It isestimated that the United States has 2 Jmillion cases of salmonellosis per year,and about 30 percent of these cases aresevere enough to be seen by a physician.Approximately 1 percent of these devel-op life-threatening septicemia whereappropriate antibiotic therapy is critical.However, in 27 percent of the casestreated, the first antibiotic chosen fortreatment proves to be ineffective be-cause the disease is due to antibiotic-resistant Salmonella (Ref. 6a).

(ii) American Cyanamid survey.Langworth and Jarolmen in a study con-ducted for American Cyanamid (Ref.7), compared the antibiotic susceptibili-ty of bacterial isolates from patientsin a rural Iowa hospital with isolatesfrom patients in an urban Connecticuthospital. E. colt isolated from patients inthe Iowa hospital were significantly.more resistant to tetracycline and neo-mycin than were isolates from the Con-necticut hospital. There were no sig-nificant differences in antibiotic resist-ances in most species of bacteria studiedother than E. colt. However, when thepool of all bacterial isolates from theIowa hospital was compared with allisolates from the Connecticut hospital,the isolates from the Iowa hospital ex-hibited significantly greater resistanceto tretracycline, ampicillin, furazolidone,and kanamycin.

(iv) Other surveys of Salmonella re-sistance. Other surveys of antibiotic re-sistance in Salmonella in farm animalsshow a continuous increase in tetracy-cline resistance (Refs. 3, 8, 9, and 10).Also, in human infections, tetracyclineresistance of Salmonella has shown adramatic increase in the United States:

Tctraoycl(i, resistance in lunmanSalmonella lyphimurium isolaic3

Pcr-ntoNumtbcu of .!

Year LIahtc end surco tcetl odine ncl~zcmco

Pr'-199.. 1r, CDO ..... 1.0 11950 to 57. 11), CDC ..... 5.0 .. .. .

lS to . I 3, CDC...... 14.0 ........1062.... 213. CDC..... a.0 12192 to 63. &1. Now Ycrt.__ 010 131Q67___.... 49 New Y crk.._ 31.4 01nS to Cl;. 3 Northcst.... 113 141970..... 315. Xorthc.t.... 3.5 151970 to 72. 2.-0. Calfrrnb.._ 37.6 10197... 718, Northr-.... 44.8 3

(b) Feeding studies-M() ChicTces-(a) Reid et al. (Ref. 17) demonstratedthat feeding subtherapeutlc levels of tet-racycline to chickens resulted in astatls-tically significant increase In the chlor-tetracycline-resistant E. c li isolatedfrom the birds. This observation was firstmqde by Smith and Crabbe in 1957 (Ref,18). Gordon. Garside, and Tucker (Ref.19) also demonstrated that tetracycline-resistant B. colt emerge in chickens fedsubtherapeutic levels of tetracycline. Intheir study, chortetracyclire resistancein the E. coil Isolates dropped when anti-biotic use was discontinued, and It rosewhen chlortetracycline use was rein-stituted. Once R-plasmid-mediated tet-racycline resistance was established inan E. coil strain, the resistance remained'during the full course of the study, whichwas long after the investigators ceasedfeeding the birds subthempeutic tetra-cycline. Further, Harry (Ref. 20) foundthat coliform (E. col) isolates fromchicks fed subtherapeutic chlortetracy-cline (100 grams/ton of feed) were 100percent tetracycline-resist'int after 8weeks of treatment, while no resistancedeveloped in coliforms isolated from thenonmedicated control groups. However,when the birds in the control group weremixed with birds in the treatment group,56 percent of the E. coil isolated frombirds whose coliforms were previouslysensitive to tetracycline became tetracy-cline resistant, and the coliforms fromthe medicated group bacame more sen:-tive to the antibiotic.

(b) In a study sponsored by the AnimalHealth Institute, Levy et al. (Refs. 21-22)examined changes in the intestinal ml-roflora of chickens, farm dwellers, andtheir neighbors, before and after the in-troduction of subtherapeutic tetracy-cline in animal feed to farms. In the 300chickens studied, the initial resistance totetracycline in E. coil was less than 10percent. Within 48 hours after introduc-ing sub therapeutic tetracycline In thebirds' diets, almost all medicated birdscontained resistant collforms with R-plasmids bearing multiple transferableresistance to tetracycline, ampicillin,carbenicillin, streptomycin, and sulfona-mides in various combinations. After 1week, the E. coil isolated from the chick-ens were almost entirely tetracycline re-sistant. In contrast, E. coil from the non.medicated birds had not acquired any

56277

antibiotic resistance 2 months after theinvestigators terminated the use of sub-therapeutic tetracycline in their feed-Chickens in the treatment group werestill excreting tetracycline-resistant E.co~i, and cleaning the chicken cages didnot alter the excretion pattern.

After subtherapeutic tetracycline usewas introduced into the farm environ-ment, the number of antibiotic-resistantbacteria in the flora of the farm dwellersincreased, although at a slowver rate thanIn the animals, and no increase was ob-served in the flora of their neighbors,who were not expossd to the animals.WIthin 5 to 6 months, 31.3 percent ofweekly fecal samples from farm dwellerscontained greater than 80 percent tetra-cycline-resistant bac-ceria compared to6.8 percent of the samples from theneighbors. This is statistically significant(P < 0.001). Moreover, using a markedresistance gene, Levy was able to demon-strate the direct spread of R-plasmid-bearing E. coi among chickens and fromchickens to man.

(c) Further, Smith and Tucker dem-onstrated that E. colt donate their R-plasmids to pathogenic Salmonella undersubtherapeutlc tetracycline pressure(Ref. 23). They compared the resistancepatterns in Salmonella and E. coltisolated from unmedicated chickens andchickens fed tetracycline. No antibioticresistance appeared in the Salmonellaisolated from nonmedicated chicks, andlittle appeared in the E. coil. Althoughfeeding subtherapeutic oxytetracycline(100 milligrams of oxytetracycline/kilo-gram of body weight) to chickens for 46days did not produce a difference in thequantity and duration of Salmonella ex-cretion or the coliform number betweentreated and control groups, it did pro-duce a significant increase In the gnti-blotic-resistant organisms In the chick-ens. By the 35th day of the experiment,Pall E. coil and Salmonella isolates fromapproximately 30 percent of the chickensfed subtherapeutic oxytetracycline car-ried R-plasmids bearing multiple anti-biotic resistance. The transmissible pat-terns of resistance on the R-plasmidsincluded ampicillin, tetracycline, strepto-mycin, spectinomycin, sulfonamides, co-listimethate or combinations thereof.More importantly, any specific resistancepattern observed in the R-plasmids iso-lated from Salmonella was first observedin E. colt at least 1 week prior to theemergence of resistance pattern in theSalmonella. Long-term feeding of thera-peutic levels of oxytetracycline (500 mil-ligrams/kilogram of body weight) like-wise did not depress E. colt or Salmonel-la excretion by the chickens in this ex-periment; however, E. coil and Salmonel-la In the treatment group developed ahigher level of antibiotic resistance thandid the birds in the nomedicated controlgroup.

MacKenzie and Bains also showed thatquantities of S. typhimurium shed bychickens were not reduced by therapeu-tic levels of oxytetracycline or chlortetra-cycline (Ref. 24).




NOTICES

(d) In a study of the excretion ofSalmonella infantis, Rantala (Ref. 25)found that birds given subtherapeuticlevels of oxytetracycline had statistical-ly significant increases of Salmonella incrops and small intestines compared tononmedicated birds. Other investigationshave shown that the subtherapeutic useof antibiotics can increase Salmonellashedding and persistence (Garside andHobbs, Refs. 26, 27).

(e) Siegel (Ref. 28) conducted numer-ous trials on the effect of subtherapeutictetracycline on Salmonella in chickensusing Salmonella that were both drug-sensitive and drug-resistant. His resultsare similar to the previously discussedstudies. Subtherapeutic oxytetracyclineuse increased antibiotic resistance in for-merly drug-sensitive Salmonella, al-though shedding did not increase. In alltreatment groups inoculated with anti-biotic-resistant Salmonella, sheddingwas higher than in the nonmedicatedgroup.

(f) The literature studies on the use ofsubtherapeutic tetracycline in chickensdemonstrate that such use causes an in-crease in R-plasmid-bearing E. coli andR-plasmid-bearing Salmonella. The anti-biotic resistance patterns develop first inB. coli and then transfer to the pathogen,Salmonella. Antibiotic resistance, par-ticularly multiple antibiotic resistance, inSalmonella isolated from chickens is in-creasing as a result of subtherapeutictetracycline use in the feed.

(ii) Swine. (a) Mercer et al. (Ref. 29)studied the effect of tetracycline and thesubtherapeutic combination of chlortet-racycline-sulfamethazine-penicillin onthe resistance of E. coli isolated fromswine. They compared swine on farmsusing medicated feed with swine grownon other farms where there was no ex-posure to these antibiotics. On the treat-ment group farms, 79 percent of theE. coli isolated from swine fed subthera-peutic oxytetracycline were tetracyclineresistant, and 77 percent of the swine fedthe combination exhibited tetracycline

'resistance. The coliforms from the swinefed the combination were also 79 percentresistant to sulfonamides and 33 percentiresistant to ampicillin. No similar resist-ance patterns devloped on farms wherethe swine were fed other antibiotic com-binations.

In a study by McKay and Branion(Ref. 30), 6-week-old pigs were fed sub-therapeutic levels of oxytetracycline (20grams/ton of feed). Qver a 6-week pe-riod, E. coli and Aeobacter isolated fromthe treatment group developed tetra-cycline resistance, while bacteria fromswine in the nonmdlicated control herddid not. In the medicated group, tetra-cycline resistance also developed in Ba-cillus species.

(b) The Animal Health Institute (Ref.31) supported a study in Kentucky onthe effect of subtherapeutic tetracyclineuse on E. coli isolated from swine. Thestudy compared a herd in Coldstream,KY. fed subtherapeutic tetracycline con-tinuously from May 1972 until 1976 witha herd in Princeton, KY, that did not re-ceive antibiotics either therapeutically

or subtherapeutically after 1972. Al-though there was no difference In thetotal coliform counts between the twoherds, the percentage of chiortetracy-cline-resistant E. coli isolates from theantibiotic-free herd dropped from 81percent to 22 percent in 3 years. Duringthat time, chlortetracycline resistance inE. ccli from the swine that were contin-ually fed subtherapeutic levels of chlor-'tetracycline remained at 85 percent.Moreover, E. coli from the antibiotic-free swine whose resistance to tetracy-cline markedly decreased, showed a si-multaneous and related drop in resist-ance to ampicillin and streptomycin.This contrasts sharply with the resultsin the treated herd where resistance toampicillin and streptomycin remainedconstant and high. Finally, E. coli iso-lated ffom the soil and water surround-ings of the herd fTd subtherapeuticchlortetracycline contained a higherpercentage of tetracycline, penicillin,and sulfonamide resistance than did iso-lates from surroundings of the antibi-otic-free herd.

Although Farrington and Switzer(Ref. 32) suggest in a short-term studyon antibiotic resistance of coliforms inswine that tetracycline resistance willfluctuate even in animals not fed sub-therapeutic levels of the drugs, a Bureauof Veterinary Medicine analysis (Ref.33) of the allegedly nonmedicated feedused in the Iowa control herd for thisstudy found antibiotics in the feed. Inthe Director's opinion, this casts seriousdoubt on the results of that study, if itdoes not totally invalidate them.

(c) After observing antibiotic-resist-ant coliforms in swine fed subtherapeuticchlortetracycline, Bulling and Stephen(Ref. 34) infected the swine with Sal-monella. The swine were divided into twobasic groups. One group was infectedwith an antibiotic-sensitive Salmonellatyphimurium, and one was infected witha sensitive strain of Salmonella chlorae-suis. These groups were then subdividedinto antibiotic treatment and controlgroups. Although only 2 of the 8 pigs in-fected with the S. typhimurium excretedtetracycline-resistant Salmonella, 10 ofthe 12 pigs infected with S. choleraesuisdeveloped salmonellosis and excretedbacteria carrying antibiotic-resistant R-plasmids. Moreover, 3 of the 4 pigs fedsubtherapeutic levels of tefracycline de-veloped tetracycline-resistant S. chole-raesuis.

When Findiayson and Barnum (Ref.35) found that pigs fed subtherapeuticchlortetracycline excteted primarilycoliforms bearing multiple resistance R-pLasmids, they postulated that the anti-biotic-sensitive E. coli had been replacedunder antibiotic pressure with other an-tibiotic-resistant serotypes. They there-fore established a limited infection inswine fed antibiotic-sensitive S. typhi-murium, and found greater numbers ofantibiotic-resistant Salmonella in tissuesand feces of swine fed subtherapeuticchlortetracycline than in the controls(Ref. 36).

In a 1969 study, Sabo and Kromery(Ref. 37) reported that tetracycline-re-

sistant Salmonella did not transfer thetetracycline R-plasmid. However, In a1973 study, they found that 2 of 23monoresistant strains transferred anti-biotic resistance with "good" frequencyto an E. coli K12 recipient. Accordingly,Sabo and Kromery (Ref. 38) now believethat E. coli tetracycline R-plasmids canbe transferred from E. colt to all S.choleraesuis strains, including varlantsthat are fully virulent and can causofatal enteric disease In man. This, Intheir opinion, rebuts the earlier conceptof Jarolmen (Ref. 38) that virulentsmooth variants are poor recipients anddonors in- contrast to rough avirulentstrains.

(W) The Bureau of Veterinary Medi-cine conducted two studies (Ref. 1) de-signed to measure the effect of sub"therapeutic chlortetracycline in feed(100 grams/ton of feed) on swine in-fected with antibiotic-sensitIve or anti-biotic-resistant Salmonella. When swinefed subtherapeutic chlortetracyclinowere inoculated with drug-sensitive Sal-monella, they exhibited less sheddingover the duration of the study than didthe nonmedicated controls for that study,However, Salmonella Isolated from medi-cated swine developed more tetracyclineresistance than did those from nonmedi-cated swine. When the swine were Inoc-ulated with' tetracycline-resistant Sal-monella, the medicated animals shedSalmonella more persistently, preva-lently, and in higher quantities thannonmedicated swine.

(e) Epidemiological surveys demon-strate that Isolates of Salmonella aregenerally at least 10 to 20 percent R-plasmid-bearing. More importantly, theclinical Isolates, I.e., those that causedillness in man and animals and aretherefore the principal public healthconcern, have been reported to have atleast 60 percent R-plasmid-determinedantibiotic resistance. Some surveys showthe resistance as high as 90 percent(Refs. 10, 39 through 45).

In England, where tetracycline resist-ance in E. coli Isolated from swine wasubiquitous because of the widespread useof subtherapeutic tetracycline for 15years in swine feed, Smith (Ref. 46)determined that resistance decreasedonly slightly in the 4 years Immediatelyfollowing implementation of the SwannCormnittee's recommendations. He alsofound that the incidence of swine shed-ding tetracycline-resistant Salmonellahad not decreased. Smith, however, didnot measure the changes in the multiplyresistant bacteria that are documentedelsewhere in this notice, and he did findthat the proportion of tetracycline-re-sistant strains of E. coll with self-trans-missible R-plasmids had declined. Lin-ton (Ref. 47) more recently concludedthat there has been little adherence tothe recommendations of the Swann Com-mittee n England.

But in Denmark, where the use ofpenicillin and tetracycline has been re-stricted since 1972, Larsen and Neilsen(Ref. 48) found that coliforms isolatedfrom 17 swine herds have exhibited asharp drop in multiple antibiotic resist-


56278



NOTICES

ance (from 68 percent to 9.5 percent)and that there has been a simultaneousincrease in the number of tetracycline-sensitive strains (from 3 percent to 36percent). Changes in resistance weresomewhat less dramatic in herds withintermittent antibiotic use- or wheremedicated swine were added to the herd.

(W Again, this information on isolatesfrom swine corroborates the results seenin the swine studies submitted under 21CFR-558,15 for other animal species.Subtherapeutic tetracycline use causesan increase in R-plasmid-bearing R. coiland Salmonella; increasingly, the R-plasmids are carrying E. colt and Salmo-nella. Finally, overall antibotic resist-ance is increasing in Salmonella.

(iii) Cattle. (a) In 1958 before knowl-edge of R-plasmids was widespread, H.W1lliam Smith (Ref. 49) studied the ef-

fect of subtherapeutic tetracycline infeed on 750 calves. After 12 weeks of ex-posure, 84 percent of the E. col isolatedfrom the calves were tetracycline resist--ant, and the coliforms also were largelyresistant to steptomycin and the sulfon-amides. No tetracycline-resistant E. collwere ever isolated from the feces of the110 animals in the nonmedicated controlgroup. Further, 2 months after termina-tion of the experiment, half of the cattlein the treatment group were still shed-ding drug-resistant E. coll.

Mercer et al (Ref. 29) studied the ef-fect of subtherapeutic chlortetracyclineand sulfamethazine on the developmentof drug resistance in E. coi. The authorscompared isolates from farms usingmedicated feed with isolates from farmsusing antiblotic-fiee cattle feed. B. coltisolated from calves fed subtherapeutictetracycline acquired R-plasmid antibi-otic resistance while few did in the non-medicated groups.

In Edwards' study (Ref. 50) of sub-therapeutic tetracycline in calves, thenumber of E. co7i in the treatment groupswas not reduced by the antibiotic, and theresistance in the isolates remained highfor the duration of the 10-week study.Resistamce dropped-when the tetracy-cline was discontinued at the end of thestudy. While tetracycline resistance inB. colt from the untreated control groupwas initially high, the percentage f anti-biotic-resistant bacteria decreased tonearly zero in the 6th week of the experi-ment and, remained there until theconclusion.

(b) An FDA contract study with theUniversity of Missouri (Ref. 51) showed'that tetracycline resistance among E.col in calves fed subtherapeutic chlor-tetracycline went from 19 percent to 95percent during the study, while tetracy-cline resistance among E. coi in the con-trol group went from 34 percent to 74percent. Generally, resistance is higherin calves that are fed subtherapeutic an-tibiotics than in range cattle or dairycattle, which normally are not fed them.When Wyoming range cattle raised with-out antibiotics were compared with dairycows, the antibiotic resistance in the E.colt from the range cattle was 9 percentin that survey; the level of tetracyclineresistance in the dairy cattle was ap-proximately 50 percent. A study of tetra-

cycline resistance in E. colt isolated fromcalves fed subtherapeutlc levels of tetra-cycline with neomycin (Ref. 54) pro-duced striking results. Although tetracy-cline resistance in the E. coi from calvesin the treatment groups averaged 57percent, no tetracycline resistance wasfound in E. coil isolated from calves thatwere kept in a separate corral and hadnever been exposed to antibiotics. Onfive other ranges, where antibiotics hadnot been given for at least 1 year, lessthan 1 percent of the coliforms were tet-racycline resistant.

(c) Several studies have examined theeffect of subtherapeutlc tetracycline onthe development of tetracycline resist-ance in Salmonella isolated from calves.For example, Loken et a. (Ref. 59) ex-amined the n-plasmid resistance inSalmonella isolated from calves fed sub-therapeutic chlortetracycline. The au-thors compared R-plasmlds isolated fromE. coll and Salmonella. When the calveswere fed subtherapeutic chlortetracy-cline, the tetracycline resistance in theE. coli isolated increased to 100 percentafter 63 days of treatment, and a concur-rent increase in ampicillin, streptomycin,and neomycin resistance occurred. Theindigenous Salmonella developed thetetracycline resistance; they also becamemultiple resistant.

(d) Sato and Kodama (Ref. 60) exam-ined Salmonella typhlmurium isolatedfrom 36 calves fed subtherapeut chlor-tetracycline in a feedlot in Japan. Moststrains exhibited greatly increased levelsof antibiotic resistance after 20 days.

(c) Director's analysis. The Independ-ent studies in the literature on the sub-therapeutic use of tetracycline in cattlefeed show that this use causes an in-crease in R-plasmld-bearing E. colt andSalmonella. They also suggest that theR-plasmids In the Z. coli may be trans-ferred to the Salmonella.'7. Directors conclusions. The studies

submitted by the NADA holders and Inthe literature show that feeding sub-therapeutic tetracycline to chickens,swine, and calves results in an increasein antibiotic-resistant E. coil and Sal-monella, and resistant E. coil transfertheir R-plasmids to Salmonella givensufficient time. When the animals areinfected with resistant strains of Sal-monella, feeding subtherapeutic tetra-cycline leads to a prolongation of shed-ding which increases the R-plasmlds inthe Salmonella reservoir. Moreover, thepercentage of antibiotic-resistant Sal-monella, in particular the multiply re-sistant Salmonella, have Increased inboth man and animals as shown by re-cent epidemlological studies, and as aresult of this plasmid transfer, the pat-terns of resistance in man and animalsare similar. Accordingly, the Directorfinds that the holders of approvedNADA's for subtherapeutic tetracyclineuse have failed to show that widespreadsubtherapeutic tetracycline use in ani-mal feed is safe under 21 CFR 558.15.

RrrmsscEs1. Williams, n, I. Rollins, mI. Selwyn, D.

Pocurull, and H. D. Mercer, "The Effect ofFeeding Chlortetracycline on the Fecal Shedz

56279

ding of Salmonella ty phimurfumn. by Ex-perimentally Infected Swine," submitted toAntimicrobial Agents and Chemotherapy. inProccedln3 of Antiblotics in Animal FeedsSubcommittee, April 27-23, 1976.

2. Walton. 3., "1n riro Transfer of Dr-ug-RVez-Ltanco," ?Jature, 211:312, 1966.

3. l7eu. H. C., at al, "Antimicrobial Reost-ance and R-Factor Transfer Among Isolatesof Salmonella in the Northeastern UnitedStates: A Compsrizno o; Human and AnimalIolatea," Journal of Infectious DLecses,132:07-622.1975.

4 ill. F. A., F. W. Hoo., "SalmonellaStrains with Tranferable Antimicrobial Re-sistanco". Journal of the American MedicalAscociation, 193:123-131, 1905.

5. CDC Par-onal Communicatons from R_Ryder. June 18, 1971 and Personal Ccm-municatlon fromJ. Bennett, July26,1977.6. Schroeder, S. A., P. M. Terry, and J. V.

Bennett, "Antibiotic Resistance and Trans-fer Factor in SalmoneM6a United States10T, Journal of the American Medical As-,oclat ion, 205:87-00,1968.Ga. Bennett J Testimony Before the Sub-

committee on Oversight and Investigationof the House Commttee on Interstate andForeign Commerce. September 23, 1977.

7. langworth, B. and H. Jarolmen, "Anti-biotlo Defense: Transferable Drug Resistanceof Clinical Bacterial Isolates from a Hospitalin an Agricultural and in one Nonagricul-tural Area". Project 3-733, NADA 48-761, Ex-hIbit 8, Submitted to FDA May 20,1975.

8. Royal, W, R. Robinson, R. Hunter, ]LJacDarmld. "Multiple Drug Resistance"Hew Zealand Veterinary Journal, 16:20,1967.

9. Mars k, F. J. T. Parisi, et al, "Transmls-slble Drug Resistance of E. colf and Salmo-nella from Human,, Animals and Their RuralEnvironments", Journal of Infectious Dis-eases, 132: 290-302, 1975.9a. Manten, A., P. A. M. Gulnee, E. H.

Hampelmacher, and C. E. Voogd, "An Eleven-Year Study of Drug Resistance in Salmonellain the Netherlands: Bull. Or. Mond. Sante.,Bull. Wd. HlM. Org. 45:8,-93, 1971.

10. Pocurull, D. W., S. A. Gaines, and IL D.Mercer, "Survey of Infectious Multiple DrugRe3istance Salmonella Isolated from Animalsin the United States", Applied Microbfology,21:358-462, 1971.

11. Ramsey, 0. H.. and P. R. Edwards,"Resistance of Salmonellae Isolated in 1959and 1960 to Tetracyclines and Chlorampheni-col", Applied Microbiology, 9:389-391, 1961.

12. McWhorter, A. C., M. C. Murrell andP. IL Edwards. "Rezistance of SalmonellaeIsolated n 1962 to Chlortetracycline;" Ap-plied Mirobfo!oj, 11:368-370, 1963.

13. Kaye. D., J. C. Merseli, Jr, and E. W.Hook. "Su.'ceptibillty of Salmonella Speciesto Four Antibiotics," New England Journalof Medicine. 26:1084-1086, 1963.

14. Winshell, E. B., C. Cherubln, J. Winter,and H. C. Neu, "Antibiotic Resistance ofSalmonella in the Eastern United States,"Antimicrobial Agents and Chemotherapy,9:8G-8, 199.

15. Cherubin, C. E., IL Szinuness, and J.Winter, "Antibiotic Resistance of SalmonellaNlortheastern United State-1970," New YorkState Journal of MedIcine, 130:369--72, 1972.

16. Bl.sett, M. J., S. L. Abbott, and R. I.Wood, "Antimicrobial Resistance and R-Factors in Salmonella Isolated in California.1971-1972," Antimicrobial Agents andChemotheraj, 5:161-163, 1974.

17. Reid. B. L, J. F. Blam, 3. R. Couch, e.i., "The Eect of Oral and Parenteral Ad-

minLitration of Antibiotics on Growth andFecal Jicroflora In the Turkey Fault,"Poultry Science, 33:307--309, 1954.

18. Smith, H. W. and W. E. Crabb, '"TheEffect of the Continuous Administration ofDiets Containing Low Levels of Tetracyclineson the Incidence of Drug-resistant Bacteriumcol on the Faeces of Pigs and Chickens: The




56280

'Sensitivity of the Bacterium coli to OtherChemotherapeutic Agents," VeterinaryRecord, 69:24-30, 1957.

19. Gordon, R., J. S. Garside, and J. Tucker,"Emergence of Resistant Strains of BacteriaFolloving the Continuous Feeding of Anti-blotics to Poultry," Proc. XVIth mnt. Vet. Con-gress. Madrid, 2:347, 1959.

20. Harry, E. G., "The Ability of Low Con-centrations of Chemotherapeutic Substancesto Induce Resistance in E. colt," PoultryScience, 3:85-93, 1964.

21. Levy, S., G. Fitzgerald, A. Macone,"Changes in Intestinal Flora of Farm Per-ronnel after Introduction of a TetracyclineSupplemented Feed on a Farm," New EnglandJournal of Medicine, 295:553-588, 1976a.

22. Levy, S. B., G. B. Fitzgerald, A. B.Afacone, "Spread of Antibiotic ResistantPiasmids from Chicken to Chicken andfrom Chicken to Ilan," Letter to Nature,260:40-42, 1976b.

23. Smith, H., and J. F. Tucker, "The Effectof Antibiotic Terapy on the Faecal Excretionof Salmonella typhimurium by Experi-mentally Infected Chickens," Journal o1lygiene, 75:275-292, 1975.

24. MacKenzie, M. IT., and B. S. Baines,"The Effect of Antibacterials on Experimen-tally Induced Salmonella typhimurium In-fection in Chickens," Poultry Science, 53:307-310, 1974.

25. Rantala, M. "Nitrovin and Tetracycline:A Comparison of Their Effect on Salmonel-leis In Chicks," British Poultry Sciehce,15:299-303,1074.

20. Garside, J. S., R. F. Gordon, and J. F.Tucker, "The Emergence of Resistant Strainsof Salmonella tVyphtmurium in the Tissuesand Alimentary Tracts of Chickens followingtile Feeding of an Antibiotic," Research inVeterinary Science, 1-184-19D, 1960.

27. Hobbs, B. et al., "Antibiotic Treatment,of Poultry in Relation to S. typhfinurium,"Month. Bull. Ministry Health, G. B., 19:178-192, 1960.. 28. Siegel, D. "The Ecological Effects ofAntimicrobial Agents on Enteric Florae ofAnimals and Man," Final Technical ReportFDA Contract '71-269, University of Illinois,College of Veterinary Medicine, 1976.

29. Mercer, H. D., D. Pocurull, S. Gaines, S.Silson, and J. 0. Bennett, "Characteristics ofAntimicrobial Resistance of Escherichia coltfrom Animals. Relationships to Veterinaryend Management Uses of AntimicrobialAgents," Applied Microbiology, 22:4:'700-705,1971.

30. M.cKay, K. A. and H. D. Branion, "TheDevelopment of Resistance to Terramycin byIntestinal Bacteria of Swine," CanadianVeterinary Journal, 1:144-149, 1960.

31. Langlols, B., G. Cromwell, V. Hays, "In-fluence of Antibacterial Agents in Feed onthe Incidence and Persistence of Antibiotic-Resistant Members of the Family Entero-lbacteraccac E. colt Isolated From Swine,"Final report to Animal Health Inst., Submit-ted to FDA April 14, 1976.

32. Farrington and Switzer, "Determina-otion of the Ratio of Antibiotic-Resistant-Coliforms In Swine", Final Report to AHI.Submitted to FDA April 18 and April 25, 1975.

33. Antibiotic Residue Branch Memoran-dum. DVR Special Project 176, August 5,1976.

34. Bulling E., and R. Stephan, "Die Wir-kung der Antibiotihabeifutterung in nutriti-yen Dosen auf die- esistenzentwichlung dercoliformen darmflora bei Schweinen," Zen-Itralblatt fur Veterinarmedizin; B. 19:268-284, 1972.

35. Finlayson, M., and D. A. Barnum, "The'Effect of Chlortetracycline Feed Additives onthe Antibiotic Resistance of Fecal Coliformsof Weaned Pigs Subjected to Experimental,elmonelMa Infection," Canadian Journal of

L Comparative Medicine, 37:63-69, 1973.

NOTICES

36. Finlayson, M. and D. A. Barnum, "TheEffect of Chlortetracycline Feed Additive onExperimental Salmonella Infection of Swineand Antibiotic Resistance Transfer," Cana-dian Jouanal of -Comparative 1edicine, 37:139-146,1973.

37. Sabo, J., and V. Kremery, "TransferableTetracycline Resistance In Salmonella Chol.eraersuis var. Kunzendrof," Zentralblatt furBakteriologic Hygine; Erste Abtetlung, Orig-inale A, 229:421-422, 1974.

38. Jarolmen, HL, "Experimental and Clii-ical Aspects of Resistance Determinants. Ex-perimental Transfer of Antibiotic Resistancein Swine," Annals of the New York Academyof Sciences, 182:72-79, 1971.

39. Wilcock, 3. P., C. H. Armstrong, andH. J. Olamder, "The Significance of the Sero-type In the Clinical and Pathological Fea-tures of Naturally Occurring Porcine Sal-monellosls," Canadian Journal of Compara-tive Medicine, 40:80-88, 1976.

40. Groves, B. I., N. A. Fish, and D. A.Barnum, "An Epidemiological Study of Sal-monella Infection in swine In Ontario," Ca-nadian Journal of Public Health, 61:396-401,1970.

41. Voogd, C. E., P. A. M. Guinee, A. Mantenand J. J. Valkenburg, "Incidence of Resist-ance to Tetracycline, Chloramphenicol andAmpicllfln among Salmonella species isolatedin the Netherlands in 1969, 1970 and 1971,"Antonie van Leevwenhoek, 39:321-329, 1973.

42. Sojka, W. J., E. B. Hudson, 0. Stavin,"A Survey of Drug Resistance in Salmonellaisolated from Animals in England and WalesDuring 1971," BHitish Veterinary Journal,130:128-138,1974.

43. Sojka, W. J., G. Slavin, T. F. Brand,G. Davies, "A Survey of Drug Resistance inSalmonellae Isolated From Animals in Eng-land and Wales," British Veterinary Journal,128:189-198, 1972.

44. Read, R. B., "Survey of SalmonellaIsolates for Antibiotic Resistance," Memo toBureau of Veterinary Medicine, November 26,1973.

45. Gustafson, R., "Incidence and Anti-biotic Resistance of Salmonella in MarketSwine," Proceedings of Antibiotics in AnimalFeeds Subcommittee of National AdvisoryFood and Drug Committee, Rockvlle, MD,January 30, 1976.

46. Smith, H. W., "Persistence of Tetra-cycline Resistance in Pig E. .coli," Nature;258:5536:628-630,1975.

47. Linton, A., "Antibiotic Resistance: ThePresent Situation Reviewed," Veterinary Rec-ord, 100:354-360, 1977.

48. Larsen, J. and N. Neilsen, "Influenceof the Use of Antibiotics as Additives on theDevelopment of Drug-resistance in IntestinalEs herichia coli from pigs," Nordisk Veteri-inaermedicin, 27:353-364, 1975.

49. Smith, H. W., "Further Observationson the Effect of Chemotherapy on the Pres-ence of Drug Resistant Bacterium coli in theIntestinal Tract," Veterinary Record, 70:575-580,1958.

50. Edwards, S. J., "Effect of Antibioticson the Growth Rate and Intestinal Flora(Escherichia coli) of Calves," Journal. ofComparative Pathology, 72:420-432,1962.

51. FDA Contract 71-306, University of,Missouri, Final Technical Progress Report,July 1971-October 31,1974.

52. Huber, W. G., D. Korica, T. P. Neal,P. R. Schnurrenberger, and R. J. Martin,"Antibiotic Sensitivity Patterns and R-fac-tors in-Domestic and Wild Aialmals," Archivesof Environmental Health, 22:561-567, 1971.

53. Hariharan H., D. A. Barnum, IV. R.Mitchell, " Drug Resistance Among Patho-genic Bacteria rom Animals in Ontario,"Canadian Journal of Comparative Medicine,28:213-221, 1974.

54. FDA Contract 72-39, Colorado, FinalReport, 1974.

55. Siegel, D., W. G. Huber, F, Enloe, "Con-tinuous Non-Therapeutic Use of Antibac-terial Drugs In Feed and Drug Resistance ofthe Gram-Negativo enterlo Florae of Food-Producing Animals," Antimicrobial Agentsand Chemotherapy, 6:697-701, 1974.

56. Howe, K. and A. H. Linton, "The Dig-tribution of O-antigen Types of Escherichiacoli in Normal Calves, Compared vth Man,and their R-plasmid Carriage," Journal ofApplied Bacteriology, 40:317-330, 331-310,1976.

57. Babcock, 0. F., D. L. Berryhill, and D. 11,Marsh, "R-Factors of Eschericha coil fromDressed Beef and Humans," Applied Microbi-ology, 25:21-23, 1973.

58. Burton, G. C., D. C, Hirsh, D. C. Blen-den, et al., '"Te Effects of Tetracycline onthe Establishment of D. colt of Animal Ori-gin, and in vivo transfer of Antibiotic Re-sistance, In the Intestinal Tract of Man," So-ciety for Applied Bacteriology SymposiumSeries, London, 3(0):241-253, 1074.

59. Loken, K. I., L. W. Wagner, C. L. Henl'e,"Transmissible Drug Resistance In Entcro-bacteriaccac Isolated From Calves Given An-tibiotics," American Journal of VeterinaryResearch, 32:1207-1212, 1971.

60. Sato, G, and H. Kodama, "Appearanceof R-factor-mediated Drug Resistance inSalmonella typhimurium Excreted by Car-rier Calves on a Feedot," Japan Journal ofVeterinary Research 22:72-79, 1974.

C. COW'PROMISE OF THERAPY(CRITERION 2 (c))

1. Background and criterion. The 1972FDA task force was concerned that thecontinuous feeding of antibiotics to ani-mals might compromise the treatmentof certain animal diseases. It concludedthat additional Information was needed,and FDA accordingly determined thatepidemlological and controlled challengestudies should be carried out to deter-mine the relationship of the use of anti-biotics in animal feed to the effectivenesof subsequent treatment of animal dis-ease, which Is criterion 2(c) of this no-tice.

Controlled studies must be undertaken todetermine whether or not the administrationof an antibacterial drug at subtherapeutl0levels results In: disease that Is more dim-cult or Impossible to treat with therapeuticlevels of the same drug or If it is necessaryto resort to another drug for treatment.(Clinical disease must be present as a lbatu"ral or artificially Induced occurrence.)

As the Director explained earlier inthis notice, and in the previous noticeproposing to terminate approval of peni-cillin use in animal feed, the subtliera-peutic use of antibiotics, including tetra-cycline in animal feeds, causes an In-crease in R-plasmid-bearing (antibiotic-resistant) E. Colt and Salmonella. TheseR-plasmid-bearing bacteria have becomeubiquitous. Further, R-plasmids cantransfer among E. colt and Salmonella,and these antibiotic-resistant organismshave been causing Increased diseaseproblems in man and animals. Each stepin the process has been clearly and re-peatedly documented, and most havebeen illustrated by the submitted studiesconducted under 21 CFR 558.15.

2. Questions raised by FDA-funded re-search and literature studies. Neverthe-less, due to the complexity and impor-tance of the compromise of therapy Issue,FDA sponsored a study to develop a dis-




NOTICES

ease model with antibiotic-susceptibleorganisms in a manner that would pro-vide susceptible pathogenic E. coli withthe opportunity to Interact in the intes-tinal tract with R-plasmid-bearing or-ganisms and develop drug resistance(Ref. 1). A University of Missouri surveyfor tetracycline-susceptible pathogenicE. colf, howeier, failed to locate an anti-biotic-susceptible strain from swine, andtherefore a compromise of therapy ex-periment using tetracycline-resistantpathogenic E. coli was performed accord-ing to the following design.

(a) Experimental design. Swine werefed an unmedicated diet and two dietscontaining different subtherapeutic lev-els of the combination chlortetracycline,sulfamethazine, and penicillin; the in-vestigators then measured the effective-ness of therapeutic levels of chIoram-phenicol (a drug unrelated to chIortetra-cycline) and chlortetracycline.

Number Infection Oral theraputle agentGroup of with (per Mlogrrm of anima])

animals E. co

Diet l-Unmeddrated1 18 No..... None.2 23 Yes-. Do.3 23 Yes ---- Choramplcnicol--. ruffl-

4 39 Yes-- ClortetcyClno--50 mlt-ligrams.

Diet 2-Chlortetracycline (20 glton of feed), -uifArmctha-zina (20 gfton of feed), and penicllin (10 glton offeed)

Number Infection Oral thermp:uUo agentGroup of with (per kilogram of animni)

animals B.coi

1 17 Yes My.... None.2 21 'Yes ..... Chlorampheuic--59 ml-n~gram3.3 23 Yes---- Chlortetra cilao--50 ran-

Diet 3-Chlortetracyclino (103 gfton of fens.famethadne (100 gfton of feed), aud p.il&l2n (00glton of feed)

Number Infection Oral therapulc agentGroup of with (per klYrm of animal)

anImals B. co l

1 14 Yes- Nono.2 10 Yes- Chloramphenic.-5-3 ril-

3 23 Yes --- Chlortetracycinao--50 rail-

(b) Director's analysis. In each diet,chloramphenicol treatment was signifi-cantly more effective for the treatment ofthe disease than was treatment withchlortetracycline. The result, in fact,show that chlortetracycline treatmentwas ineffective both in the untreatedcontrol group and in the groups fed thecombination of subtherapeutic anti-biotics in the ration.

The Missouri study indicates than ani-mal therapy may be compromised wherethe pathogen is resistant to the anti-biotic used for treatment.

Mlackenzie and Baines (Ref. 2) in-fected broiler chickens witb tetracycline,neomycin-, and sulfonamide-resistantSalmonella typhimurium collected from

a field outbreak of salmonelIosis n broil-ers, and they then compared the resultsof subsequent tetracycline therapy withtherapy with furaltadone and chloram-phenicol. hile tetracycline therapy didnot produce a lower shedding rate thantherapy with the other antibiotics, thegroup given therapeutic tetracyclinetreatment exhibited a higher mortalityrate than the groups treated with fural-tadone and chloramphenicoL

Hierpe (Ref. 3) studied the effect ofchlortetracycline therapy on Pasteurellaisolated from feed lot cattle that hadbeen fed subtherapeutic chlortetracyc-line. He found that the use of subthera-peutic chlorteracycline caused an in-crease in Paste-arella that were resistantto chlortetracycline, penicillin, sulfona-mides, and other antiblotics; more im-portantly, subsequent chlortetraeyclinetherapy for the treatment of the Pas-teurcila infections in these animalsproved unsuccessful.

Therefore, the Director finds that thequestions posed by the FDA Task Forcehave been reinforced by compromise oftherapy studies in swine, chickens, andcattle conducted by other independentscientists.

The holders of the approved NADfAssubmitted nine studies in their attemptto resolve the compromise of therapyissue. After careful consideration of thesestudies, the Director has found them tobe inadequate for various reasons. Theyare of limited size and scope, and in lightof evidence generated from other sourcessince the regulations and guidelines wereestablished they are inconclusive. TheDirector believes that only careful long-term epldemiological field studies will beadequate to resolve the question of theextent to which therapy has been com-promised.

3. Compromise of therapy studies inchickens-(a) Pfizer Study. Pfizerstudied the effect of parenteral and oraloxytetracycline therapy in artificlallyinfected chickens that had been fed sub-therapeutic oxytetracycline. After a 25-day preexposume to subtherapeutlc oxy-tetracycline in the feed, chic's were in-feeted by intramuscular injection with apathogenic but tetracycline-sensitive E.coli. Although subsequent parenteral andoral therapeutic treatment with oxy-tetracycline (12.5 miilgrams/subeutane-ously/bird and 500 grams/ton of feed)reduced the mortality rate in the chick-ens, oral therapy did not produce a re-duced incidence of lesions. Moreover, theDirector finds the study design to befaulty because the nonintestinal route ofE. colt infection does not resemble thenormal route of infection, anU It there-fore bypasses the opportunity for the F.-plasmid transfer which can occur in theintestine. Pfizer also used a tetracycline-sensitive strain of E. colt. As recent evi-dence demonstrates, antibiotic rzzistanceIs now high in the animal population, andthis fact s important to the compromiseof therapy problem. For these reason3,the Director concludes that the study Isinadequate to resolve the compromise oftherapy Issue.

(b) American Cyanamid Study. Amer-Ican Cyanamid conducted a 2-phasestudy to measure the effects of chIor-tetracycline in water therapy (1 gram/gallon of water) when chickens wree in-fected with Salmonella isolated fromother birds that had been fed subthera-peutic chlortetracycline. Cyanamid usedthe 2-phase study because it had diffi-culty experimentally inducing fatal oralinfections in chicks more than 4 days old.The rate of fatal infection was con-sidered an indication of the adequacy ofthe exparimental infection.

In phase I, one group of chicks was fedsubtherapeutic chlortetracycline (200grams/ton of feed) for 2 weeks, and fecalcollforms were isolated. Then the chickzwere orally infected with a nalidixic acidmarked strain of a pathogen, Salmonellagallinarium. After 2 days fecal Salmo-

ella were Isolated. Coliforms and Sal-monella were Isolated from an untreatedbut Infected control group in the samemanner.

This phase of the study was designdto allow R-plaszmds from coliforms totransfer to Salmonella during the 2 daysin the chick. In Phase II a second groupof chicks wa, inoculated with bacteriaobtained from the first group accordingto the following design:

of b~r~I3 pt colr

1 10 laa- YCa(31).- CiL

Pknsi II

ar-p Namlr X"-Zu13MLa Aurcmydaoft trda tb'zry l-fs

1110~

A c

5~ L' NonOL... YrE i S2edC..... No.Zo 52 and CZ ---.-Y.

Chlortetracycline therapy was insti-tuted 48 hours after the Phase I inocu-lation. Although therapy proved to beequally succesful whether or not thebirds were infected with organis-ms iso-lated from chic!-s that had received sub-therapeutic tetracycline in Phase L thebasic experimental design did not trulyaddress- the compromise of therapy issue.Moreover, the experiment is defective inseveral other areas. Evidence from litera-ture shows that longer exposure to sub-therapeutic chiortetracycline in chickenfeed which is consitent with the actualconditions of the drug's use in the field.produces an increase in R-plasmid-bear-ng bacteria. Phase I of the study was

conducted for only 14 days, and the Di-rector finds this truncated aspect of thestudy inappropriate as a model for an ac-tual field Infection. In birds infected withboth Salmonella and E. coli, the orga-

FEDERAL REGISTER, VOL 42, NO. 204-FRIDAY, OCTOBFR 21, 1977

56281



NOTICES

nism had only 2 days to interact anddonate R-plasmids, which is inconsistentwith normal conditions and conditions "

in other studies reported in the litera-ture. Finally, the chickens in Phase IIwere never exposed by any subtherapeu-tic antibiotics in their feed, which wascontrary to -the guidelines. The agencydeveloped that aspect of the guidelinesto assess the element of concurrent con-tinuous antibiotic exposure, and the Di-rector believes that point is still relevant.For all these reasons, the Director con-cludes that this study has failed to re-solve the compromise of therapy issue.

(4) Compromise of therapy studies inswine-(a) Diamond Shamrock StudyNo. 1. Diamond Shamrock conducted acompromise of therapy study for swineusing the subtherapeutic combination ofchlortetracycline, sulfathiazole, penicillin(CSP-250) in feed and neomycin as thetherapeutic agent.

Forty pigs, 4 to 5 weeks of age, weredivided Into 4 groups of 10 pigs each.Groups A and B served as environmentalcontrols and did not receive CSP-250.Groups C and D were placed on CSP-250for the first 21 days of the trial. On the23d day, all four groups of pigs wereinoculated with S. choleraesuis. Approxi-mately 72 hours after inoculation, neo-mycin therapy (7 milligrams/pound/dayIn water) was initiated in groups B andD, and continued for 4Y2 days.

From the standpoint of growth,weight, and feed/gain, the neomycintreatment group (Group B) performedthe poorest of the four groups. Neomycinin the presence of CSP-250 (Group D)was better than neomycin without CSP-250 (Group B), but not significantly bet-,ter than the CSP-250 group alone or thenonmedicated controls. Because the neo-mycin shows no therapeutic value, theDirector concludes the study is immate-rial.

(b) Diamond Shamrock Study No. 2.The second Diamond Shamrock studyattempted to determine whether an E.coll infection of swine was more difficultto treat with ntrofurazone whenpigs hadbeen maintained for 3 weeks prior totherapy on chlortetracycline at 100grams/ton.

Forty 5-week-old pigs were dividedinto 4 groups of 10 animals each. Twogroups did not receive subtherapeuticantibiotics, while two groups were fedsubtherapeutic levels of chlortetracy-

cline (100 grams/ton). On the 21st day.pathogenic E. coli were added to the feedof all the pigs, and at the first sign ofdisease one group of pigs from both themedicated and nonmedicated groupswas treated with, therapeutic furazoli-done in water.

Feeding subtherapeutic chlortetracy-cline to the pigs did not interfere withfurazolidone treatment of the experi-mentally induced disease. However, thepigs were fed chlortetracycline for only3 weeks before infection, and therapywas initiated at the first signs ofdisease. These points minimized theopportunity for the transmission' ofR-plasmids. Also, Animal Health Cri-teria 1(c) states that the sponsors wereto assess the effect of subtherapeutic useof a drug on subsequent therapy by thatsame or a related drug. Chlotetracyclineand furazolidone are not chemicallyrelated, and plasmid-mediated .nitro-,furan resistance rarely occurs in a pat-tern of resistance with other drugs(Ref. 4). Moreover, because of questionsabout carcinogenicity, the Director pro-posed in a notice published in the FED-ERAL REGISTER of May 13, 1976 (41 FR19907) to withdraw approval of theNADA's for the use of furazolidone. Ac-cordingly, the Director concludes thestudy has failed to resolve the compro-mise of therapy issue.I (c) Pfizer Study. Pfizer carried out a

study to determine the therapeutic ef-ficacy of oxytetracycline (500 grams/ton) against induced 8almonelosis inpigs previously fed subtherapeutic oxy-tetracycline (150 grams/ton) for 21days. The infecting agent was Salmo-nella choleraesuis, given by oral inocu-lation.

Sixty pigs, 6 to 8 weeks'of age, weredivided into 2 groups (A and B), whichwere then further subdivided into groupsof 10 each. For 21 days, the 3 subgroupsin group A were maintained on a non-medicated diet while those in group Bwere fed a similar diet containing sub-therapeutic oxytetracycline. On days 22to 24 all animals were fed a nonmedi-cated ration. All feed was then with-drawn, and the pigs were infected withthe S. choleraesuis. One subgroup ofgroups. A and B was fed the treatmentration (oxytetracycline 500 grams/ton)at disease onset and continued for 14days. The table below summarizes theexperimental design.

Infection Frequency Average AveragePremedicated with Treatnent Mortality of diarrhea daily dally-

Salmonlla igain feedkilogram kilogram t

A TA Nonmedl- Noninfected. Nonmedicated ...... 0110, 0 pct___ 12 0.706 1.62cated.

A T2 - do --------- Infected ---------- do -------------------- 3/10, 10 pet.. 41 .031 .81A T - do -------------- do ----- Oxytetracycline 550 pm.. 0/10,0 pet... 3 633 1.64B T4 Modlcated .... Noninfeeted Nonnedleated --------- 0/10. 0 pct.. 6 .68 1.58B TS .-- do --------- Infected. ----- do ---------- - ,6-/10, 60 pct-- 42 -. 104 .86B TO- -.. do ----------- do-.....- Oxytetmcycline550p/n. 1110, 10 pet- 19- .289 1.06

Pigs that were given therapy after in- hours. Pathological finding at necropsyfection (T2 and T5) showed clinical were consistent with salmonellosis, andsigns of disease 24 hours postinoculation S. choleraesius was discovered from alland 100 percent morbidity by 48 to 96 animals that died.

Although oxytetracycline at 500grams/ton was efficacious In controllingmortality whether or not the animalshad been premedicated with oxytetra-cycline, 150 grams/ton, the results showa trend toward compromise of therapy.(For mortality compare T2 v. T5 and T3v. T6.) Group A, which was not fedthe subtherapeutic antibiotic-containingdiet before infection, exhibited a bettoroverall result against frequency ofdiarrhea and average daily gain. Despitethe fact that the differences In the re-sults are not statistically significant,there Is no basis for the Director to con-clude that the results are the same. Forthese reasons and the general problemsassociated with the study's design, theDirector concludes the study did not re-solve the compromise of therapy issue.

(d) American Cyanamid study. Ameri-can Cyanamid examined the use of thesubtherapeutic combination of chlor-tetracycline - sulfamethazine - penicillin(ASP-250) on the therapeutic effective-ness of sulfamethazine In pigs experi-mentally infected with Salmonella cho-leraesuis, variety konzendorf.

Sixty 4-week-old pigs were divided into6 groups. Half were fed ASP-250 for 2weeks, and half were fed plain swinegrower mash. One week later 40 of the60 pigs were inoculated via nonmedi-cated feed with S. choleraesuis. Feed wasremoved from all groups 18 hours beforeinfection. Sulfamethazine therapy wasinitiated in one infected group fed ASP-'250 and one that was only fed the un-medicated diet when 80 to 100 percent

-of the pigs in each group showed severediarrhea (3 days postinfection). Thedrug was given intraperitoneally at 100grams/pound of body weight, and dailymedication was continued at 50 milli-grams/pound until diarrhea had ceasedor 14 days postinfection.

Prior subtherapeutic treatment withASP-250 did not appear to reduce thetherapeutic effectiveness of sufametha-zine. Nevertheless, the study involvedshort-term exposure to the subthera-peutic drug. In addition, therapy wasadministered by an unusual method andnot geared to practical therapy. Forthese reasons, the Director rejects thestudy as inconclusive.

5. Compromise of therapy studies incattle-(a) Diamond Shamrock study.In this experiment, the effect of sub-therapeutic chlortetracycline in feed onthe oxytetracyclihe treatment of inducedsalmonellosis was measured. Twenty-eight calves were distributed Into 4groups of 7 each. Two groups receivedsubtheraputic chlortetracycline 70 milli-grams/calf/day), and two did not re-ceive any antibiotic in their feed.

On day 21, tetracycline-sensitive S.typhimurium were orally administeredto each calf. After fecal samples weretaken on day 2, parentral oxytetraoyclinotreatment (5 milligrams/pound bodyweight/day) was begun in one group ofpremedicaeted calves and in one groupof nonmedicated animals; treatment wascontinued for 3 days.


56282



NOTICES

Within 2 days of Salmonella innocu-lation, all calves had fevers of 105" F ormore; many animals had diarrhea, in-dicating that. disease had occurred.Three deaths occurred in the nontreatedgroup, but none of the group that wastreated with oxytetracycline.

Although injection with therapeuticoxytetracycline was successful in reduc-ing the febrile responses and diarrhea incalves inoculated orally with Salmo-nella, the study by no means resolvesthe compromise of therapy issue. Thecalves were fed subtherapeutic chlortet-racycline for only 21 days before infec-tion; the calves were infected with atetracycline-sensitive strain of S. typhi-murium; the Salmonella were never ex-posed to subtherapeutic antibiotics; andtherapy was initiated 2 days after in-troduction. For all of these reasons, theDirector finds the study inadequate toresolve the compromise of therapyproblem.. (b) Pfizer study. In this study of oxy-tetracycline, 20 calves were allotted ingroups of -5 to 4 pens. Two groups werefed a nonmediated basal ration for 21days, while the other two groups werefed subtherapeutic oxytetracycline (350milligrams/head/day). But the medi-cated diet was terminated after 21days, and normal ration was substituted.Three days later all calves were innocu-lated subcutaneously with a strain oftetracycline-sensitive Pasteurella mul-tocida. Parenteral oxYtetracycline ther-apy was initiated (5 milligrams/pound/day) immediately and continued for 2additional days.

The results illustrate that oxytetra-cycline injected at 5/milligrams/poundfollowing inoculation is effective in con-trolling tetracycline-sensitive Pasteur-ella that are never exposed to subthera-peutic antibiotics or R-plasmids in thegut.

But this study obviously does not re-solve the compromise of therapy issue.Indigenous E. coli were only briefly ex-posed to subtherapeutic oxytetracycline,'and the calves were placed on an anti-biotic-free diet before inoculation withthe Pasteurella. This was contrary tothe guidelines and sound science. more-over, the parenteral route of inoculationof the Pasteurella did not permit readyassociation of antibiotic-resistant en-teric coliforms and the infecting orga-nism, and after only 3 days' systemictherapy was initiated.

(c) American Cyanamid study. Thepurpose of this experiment was to de-termine the influence of a combinationof subtherapeutic combinations of anti-bacterials, chlortetracycline and sulfa-methazine, on the therapeutic effective-ness of sulfamethazifie in calves experi-mentally infected with S. typhimuriu7.Thirty-two 5- to 6-week-old male calveswere divided into 4 groups of 8 animalsper group. One group was premedicatedwith the combination for 2 weeks, whilethe others received an antibacterial-freediet. Then the premedicated group andthe three unmedicated groups were in-oculated orally with tetracycline andsulfonamide-sensitive S. typhimurium.

One day after infection. the premedi-cated group and an unmedicated Infectedgroup were treated with therapeutic sul-famethazlne (100 mlfligramrs/pound) for1 day followed by 50 mllligrans/poundfday for 4 additional days. The animalswere monitored for 14 days after In-fection. All of the chlortetracycline-resistant E. coUl isolated had multipleantibacterial-resistance patterns; themost common pattern was streptomycin,neomycin, kanamycin, triple sulfa, tet-racycline, and in some casez amplclllin.

American Cyanamid concludes thatfeeding subtherapeutlc chlortetracyclineand sulfamethazine does not interferewith the therapeutic activity of sulfa-methazine against Salmonella typ l-murlum, in calves. The Director dis-agrees. The coliforms were exposed tothe subthempeutc antibacterials foronly 2 weeks before infection and therewas no exposure after inoculation. Thus,the sensitive Salmonella were e.xpo-ed tocoliforms without therapeutic antibioticpressure for only 1 day. Based on thisanalysis, the Commlsoner concludes thestudy Is inadequate for resolving thecompromise of therapy Issue.

6. Director's conclusion. The Directorhas analyzed all the material submittedby the holders of NADA's submittedunder § 558.15 to addres the compro-mise of therapy Issue, and the informa-tion on this issue gathered from other,independent sources. In- his opinion, Itfails to resolve the questIons about thepotential for harm from compromise oftherapy that was first raised by the FDAtask force; rather, the questions raisedhave been reinforced by the Informationthat has been subsequently collected.

7. Optimal level o1 effectiveness (Ani-ma! Health Criterion 4). This was orig-inally stated as a separate criterion asfollows:

Tho optimum uzago level for each indL-cation of use of the antibacterial drug atsubtherapeutic levels shall not Increaerig-nificantly with continued use.

Once the optimum level Is es-tbllrhed, Lastudy shall continue over succeeding genera-tions or populations of animals to determineIf this e level continue. to yield the samomesurablo effect.To address this criterion, the AnimalHealth Institute submitted the results ofa study begun In 1972 which comparesthe effectiveness of four antibiotics(chlortetracycline, tylosin, bacitracin,and virginamycin) to a nonmedicatedgroup in swine (Ref. 5). The Directorconcludes that the study is inadequateto resolve the issue. However, this is inpart due to the inability to design studiesthat would produce meaningful resultswithin a 2-year period. This study wasconducted at only one location; tests atseveral locations are necessary to pro-vide any evidence that may have generalapplication to the swine industry. More-over, the antibiotics were not fed to theswine at graded dosage levels (dosagetitration), which is necessary to deter-mine the optimal level of the drug'seffectiveness. That Is the first step inattempting to address the concerns.Without that evidence, the Director can-not make any determination about the

role of R-plasmid-bearing organisms inthe continuing effectiveness and safetyof subtherapeutic use of any tested anti-blotie in animaI

1. FDA contract 71-05; University ofZIsso-uri.

2. LMc aenzie, f. LL and B.8. Balnes, '-Th-Effeflt of Antibaotarials on ExperimentallyInduced Salnonedit h'imurium infectionsIn C liozn," Poultry Sc ince, 53:357-310,15ZL

3. Hjsrpe. Dr. C. A., SChool of Veterinar?Aediclno. Univrsity of California, Davi,Letter to Saniey Fzikow. Scahol of Medizlne.University of Washington, November 18,

4. Anderson, E. &. "The Ecology of Teens-feeable Rezlatance in M-terobacterla'" Annul1eIWO= of r.Icrobiology, 22:131-180,. 19f8.

D. L.anglols B, G. Cromwell V. ]ays, -In-fluence of Antibacterial Agents In Feed inthe Incidence and Persistence of Antibiotic-]Pc-tstaut BMembers of the Family Entero-bscterlaceae R. col Isolated from Swine.-fins] report to the Animal Health InstituteApril 14, 1975.

1. PAT11OGr'nCn (crmTzot.T 3)

1. Bac7.ground and criterion. It is clearthat bacterial plasmids contribute sig-nificantly to a bacteria's capacity to pro-duce disease and to survive within thehoat organism (Ref- 1). The productionof enterotoxin.,Ior example Is an essen-tial factor in the pathogenicity of E. colstrains of porcine origin, and Smith andHalls (Ref. 2) demonstrated that thisproperty was governed by a plasmid,termed ENT. Similarly, the genetic de-terminants for enterotoxin production inV. call Isolated from calves and lambshave also been shown to be controlledby transmissible plasmlds (Ref. 3). Re-cent studies support the premise thatnterotoxin-producing strains of E. coil

are also responsible for a significant pro-portion of previously undiagnosed hu-man diarrheal disease (Refs. 4 through0). Researchers have now shown thatthe ability of human E. coli strain tomake an enteroto.dn is also mediated byaL tranmisible plasmid (Res. 7 and 8).

In addition to toxins, other plasmid-mediated virulence factors have been de-scribed. One of the characteristics of thediarrheal disease caused by enterotoxi-genic E. eCol in man or animals is theability of large numbers of the bacteriato colonize the small bowel. There is evi-dence that a surface associated antigen,

038. on E. coli increase pathogenicityfor pie'- since it facll-tes colonizationby helping to overcome intestinal mo-tility and other clearing mechanisms(Refs. 9 through 13). Further, Orskov etal. (Ref. 14) showed that K8 productionis governed by a transmissible plasmid.A similar antigen, K99, has been de-cecribed for calves (Refs. 15 through 17).Moreover, these K-antigens play a rolein the host specificity of these pathogens.The E83 antigen from precine isolatesis unable to produce adhesion to the calfintestine, and the K93 calf antigen is un-able to adhere to the pig intestine (ReL15). A similar plasmid-controlled sur-face antigen has recently been describedin a strain of E. coli, causing severe hu-man diarrheal disease (Ref. 18).

FEDERAL REGISTER, VOL 42, NO. 204--.I1DAY, OCTOBEI 21, 1977

56283



PROPOSED RULES

Another way plasmids can' contributeto virulence Is exemplified by the colicinV plasmid (Ref. 19). Colicin V is themost common colicin produced by E.col, and pathogenic E. coli containingthe colicin V plasmid have a greater abil-ity to resist the host species' defetisemechanism (Ref. 19). Such E. coli alsotend to be more refractory to the bac-tericidal effects of undefined componentsin serum. In addition, Smith's experi-ments in chickens and in humans revealthat the colicin V R-plasmid confers onorganisms an increased ability to sur-vive in the alimentary tract as well as-In the tissue (Ref. 20). On the basis ofthis evidence, the Director believes thatother plasmid-mediated factors thaten-hance pathogenicity may well be foundin the future.

Although pathogenicity is generallydetermined by more than one factor, theaddition of a single specific character toa nonvirulent organism can endow thatorganism with virulence, and the po-tential dangers of this character beingmediated by a transmissible element areapparent. Because R-plasmids and viru-lence plasmids can reside in the samebacterial cell, the possibility is increas-ing that plasmids that contribute topathogenicity may become more widely-disseminated among bacterial speciesdue to the selection of-the large reser-voir of R-plasmids within enteric or-ganisms.

For these reasons, FDA establishedHuman and Animal Health Safety Cri-terion 3: "The use of low and/or inter-mediate levels of an antibacterial drugshall not enhance the pathogenicity ofbacteria."

The Food and Drug Administration'sguidelines required a series of well de-signed studies to determine if the use ofantibacterial drugs in animal feeds en-hances pathogenicity of Gram-negativebacteria. First, the sponsors were to de-termine if plasmids coding for toxin pro-duction could become linked to an R-plasmid and be transferred in vitro.Finally, if this was demonstrated ingerm-free anmials, experiments were tobe conducted in conventional animals.

Due to the progressional nature of thestudies, the Director did not require thesponsors to complete the studies duringthe time alloted by § 558.15. The sponsorswere committed to conduct such studiesand to submit reports on the studies atregular intervals. The Animal HealthInstitute did submit a study conductedby Dr, John Walton to examine the asso-ciation of plasmid-mediated toxin pro-duction with R-plasmids, and data werealso obtained from FDA contracts withDr. Stanley Falkow and Dr. CarltonGyles.

2. Walton- study. The Walton study(Ref. 21) reported in vitro transfer ex-periments using a donor organism bear-ing both the enterotoxin pltsmid andR+ factors antibiotic resistance plas-mids and a recipient organism that lacksan R-plasmid. Walton concluded thatsubsequent selecti6n of R- transcon-jugants does not select for enterotoxinproduction.

The Director finds that the study con-tained major shortcomings in the pro-cedures used, and he rejects Walton'sconclusions as inadequately supported.The enterotoxin-producing strains (con-taining plasmids termed ENT) used inthe experiment were inadequately ex-amined for the frequency of transfer oftheir ENT plasmids, and the number ofR+ transconjugants tested for ENTtransfer (20) was insufficient since only afrequency of 5 percent or greater couldbe detected. From each mating, 20 trans-conjugant-colonles were pooled and sub-cultured into 100 milliliters of nutrientbroth; then they were grown overnightto' obtain cells and supernatant fluid totest for toxin production. However, nopositive control was included in the ex-periment to show that, in screening, atleast one known ENT positive colony, outof 20 colonies, would actually produce apositive reaction for toxin production.For these reasons, the Director concludesthat the study neither conclusively re-solves the issue nor provides adequateevidence to support the conclusion thatselection for R+ transconjugants doesnot select for enterotoxin production.

3. Falkow study-(a) In vitro transfer.On the other hand, Falkow (FDA Con-tract 73-7210) unequivocally demon-strated that ENT and R-plasmids do co-transfer and that drug selection for theR-plasmid and subsequent'clonal screen-ing for ENT was an adequate laboratorytool for detection of cotransfer.

In an in vitro mating, E. coli K12 (con-taining a bovine ENT plasmid, a K-anti-gen-determining plasmid (K99), and anR-plasmid coding for tetracycline andstreptomycin) was crossed to three drug-sensitive E. coli K12 recipient strains.The recipient strains were rifampicinresistant, and the donor was rifampicinsensitive. The rifampicin-resist-ant re-cipient that received the tetracycline-streptomycin plasmid were recovered onrifampicin-tetracycline drug plates;these recombinant clones were thenscored for coinheritance of ENT andK99. Of 225 clones tested (75 from eachof the 3 crosses), 2 clones (0.88 percent)received both ENT and K99+. Thus,totransfer of K99 and ENT plasmid forpathogenicity with the tetracycline-streptomycin drug resistance plasmidwas of a low but detectable incidence.

In another in vitro mating study, abovine enterotoxigenic nonlactose-fer-menting E. coli isolate (B44) (containingthe following plasmids: ENT, K99, andan R-plasmid (R1 ) containing genescoding for ampicillin, rhloramphenicol,kanamycin, and streptomycin resist-ance) was crossed with a lactose-fer-menting strain of E. colN, K92 strain1485. Lactose-fermenting and chloram-phenicol-resistant transconjugants werescored for K99 and ENT.

The incidence of K99 plasmid transferwas 3/37 (8 percent) and the incidenceof the ENT plasmid transfer was 9/37(24.3 percent). Furthermore, the inci-dence of K99, ENT, and R1 cotransferwas 3/37 (8 percent).

(b) In vivo transfer. Falkow fed B44E. coli bearing resistance (R1 ), ENT, and

K99 plasmids to baby calves, and in vivotransfer of the ( 1 ) plasmid to indi-genous microflora was monitored. In oneexperiment, ENT plasmid was cotrans-ferred at an incidence of 3/39 (7.7 per-cent); however, K99 was not transferred.In another in vivo transfer experiment,the ENT was cotransferred at an inol-dence of 1/88 (1.1 percent) and co-transfer of K99 did not occur. Further-more, detection of 99 cotransfer washampered by the autoagglutinaton of50 percent of the transconJugants whenslide agglutinations with KO antiserawere'performed.

From these experiments, Valkow con-cluded that possession of an R-plasmidby an enteropathogenc strain does notguarantee cotransfer of ENT or KO;nevertheless, the implications of co-transfer at even~a low Incidence in theintestinal tract of an animal, should theanimal be exposed to the same antibioticsto which the enteropathogen is resistant,

-has potent public health consequences,4. Questions raised by other studies.

(a) Naturally occurring toxigenic strainsof E. coli are often multiple resistant,and during a recent hospital outbreakof infantile diarrhea in Texas, Wachs-muth et al. (Ref. 23) reported thatplasmid-mediated toxin production andmultiple antibiotic resistance were dem-onstrated. Transfer of a 67 x 101 and30 x 100 dalton plasmid was associatedwith the transfer of resistances and en-terotoxin production, respectively. More-over, when antibiotics were used to selectE. coll K12 recipients from a one-stepbacterial cross, all the resistances wereconcurrently transferred, and 36 perceAtof these drug-resistant recipient orga-nisms also transferred their ENT plas-mids and produced enterotoxin. Clearly,the Director must conclude that R-plas-mid transfer can enhance the possibilityof ENT transfer and the production Ofenterotoxin.

(b) Translocation is believed to be theprimary mechanism for the dissemina-tion of resistance genes in vivo. UnderFDA Contract 223-73-7210, Falkow hasbeen able to show the translocaton ofantibiotic resistance genes to ENT pls-mids in vitro. He also demonstrated thatENT plasmids can acquire resistancegenes from R-plasmids if they inhabitthe same cell. Ampicillin, sulfonamide,and streptomycin plasmids constructedin vitro by translocatlon are indistin-guishable from such ampicillin plasmidsobtained from clinical isolates of E. coiland Salmonella (Ref. 24).

More recently, Gyles (FDA Contract223-73-7219) demonstrated the in vivotransfer of ENT plasmids in the intesti-nal tract of pigs, using the selection oftetracycline-resistant recipient orga-nisms as a basis 'for screening ENT+recipient colonies. All of the 35 tetracy-cline-resistant recipient colonies ob-tained were shown to bear the ENT plas-mid. Gyles also showed that tetracyclineresistance and enterotoxin biosynthesisreside on the same plasmid.

5. Director's conclusions. The evidencefrom both In vitro and In vivo expOrl-ments demonstrates that ENT plasmilds

FEDERAL REGISTER, VOL 42, NO. 204-FRIDAY, -OCTOBER 21, 1977

56284



and R-plasmids can become linked. OnlyDr. Walton's study describes data to thecontrary; however, his study is inade-quate for the reasons discussed. Accord-ingly, the Director concludes that theexisting evidence demonstrates that R-plasmids can increase the pathogenicityof organisms, and inadequate evidencehas been submitted to prove the con-trary.

REFERENWCES

1. Falkow, S, "Infectious Multiple DrugResistance," Pion, Ltd, London, 1975.

2. Smith, TL W. and S. Halls, "The Trans-missible Nature of the Genetic Factor inEscherichia colt That Controls EnterotoxinProduction," Journal of General Mlcroblol-ogy, 52:319-334, 1968.

3. Smith. H. W. and AL A. Linggood, 'TheTransmissible Nature of Enterotoxin Pro-duction in a Human EnteropathogenicStrain of Echerichia col" Journal of Medi-cal Microbiology. 4:301-305, 1971a.

4. Gorbach, S. L, "Acute diarrhea--'toxin'disease?" New England Journal of Medicine,283:44-45, 1970.

5. Gorbach, S. L., and C. Md. Khuran"Toxigenic Escherichia coil: Cause of Infan-tile Diarrhea in Chicago," New EnglandJournal of Medicine, 287:791-795, 1972

6. Dupont, H. L., S. B. Formal, R. B. Hor-nick, AL J. Snyder. J. P. Libonati, D. G.,Shea-ban, E. H. LeBrec, and J. P. Kalas, "Patho-genesis of Escherichia colt Diarrhea," NewEngland Journal of Medicine, 285:1-9, 1971.

7. Skerman F. J., S, B. Formal, and S.Falkow, Plasmid-Assoclation EnterotoxinProduction in a Strain of Eschericha colt Iso-lated from Humans," Infection and Immun-ity, 5:622-624, 1972.8. Smith, H. W., and IM. A. Linggood, "The

Transmissible Nature of Eenterotoxin Pro-duction in a Human Enteropathogenlc Strainof Escherichia coil," Journal of Medical Mi-croblology, 4:301-305, 1971a.

9. Orskov, L, F. Orskov, W. .3. Sojka, andJ. AL Leach, "Simult~neous Occurrence ofEscherichia coi B and L Antigens In Strainsfrom Diseased Swine," Acta Pathologica etMicroblologica Scandinavlca, 53:404-422,1961.

10. Orskov, I, M. Orskov, W. J. SoJka, andW. Wittig, "K Antigens K88ab(L) and K88ae(L) in E. CoIL A New 0 Antigen: 0141 anda New K Antigen IK89(B)," Acta Pathologicaet Microbiologica Scandianavica, 62:439-447,1954.

11. Smith, H. W., and AL A. Linggood, "Ob-servations on the 1.athogenio Properties ofthe NB8 Hly and Ent.Piasmids of Escheric acolt with Particular Reference to PorcineDiarrhea," Journal of Medical Microbiology,4:467-485, 1971b.

12. Jones, G. W. and J. M. Rutter. "Roleof the IK88 Antigen in the Pathogenesis ofNeonatal Disrrhea Caused by Escherlchiacoil in Piglets," Infection and Immunity,6:918-927, 1972.

13. Hobmann A., and I R. Wilson, "Ad-herence of Enteropathogenic Esc erlhia coilto Intestinal Epithelium in Vivo," Infectionand Immunity, 12:866-880, 1975.

14 Orskov, I., and F. Orskov, 'Tplsome-Carried Surface Antigen P88 of Escherichiacolt 1. Transmission of the Determinant ofthe E.88 Antigen and Influence on the Trans-fer of Chromosomal Markers," Journal ofBacteriology, 91:69-75, 1966.

15. Smith, H. W., and L. A. Linggood,'"urther Observations on Eschlerchia coltEnterotoxins with Particular Regard to ThoseProduced by Atypical Piglet Strains and byCalf and Lamb Strains: The TransmissibleNature of These Enterotoxins and of a PMAntigen Possessed by Calf and LambStrains," Journal -of Medical Mlcrobiology,5:243-250,1972. -

NOTICES

16. Orskov, 1, P. Orzkor H. W. Smith.and W. J. Sojka, "The Establi hment ofK99, a Thermolabile. Transmissible Es-cherichla colt Antigen, Previously Called'Kco.' Possessed by Calf and Lamb Enter-opathogenic Strains," Acta Pathologic etMicrobiologlca ScandinavIca, 83:31-30, 1975.

17. Burrows, IM. R. IL Sellwood and R.A. Gibbons, "HaemnagGlutinating and Ad-hesive Properties Associated with the K9AntiOn of Bovino Strains of Eschlerichfacoil," Journal of General MIcroblology,90:269-275, 1976.

18. Evans. D. G.. 1. P. Slver. D. J. Evans,Jr.. D. G. Chase, and S L Gorbach. "Placmid-Controlled Colonization Factor Associatedwith Virulence In Eschcrichia coil Entero-toxigenle for Humans," Infection and Im-munity, 12:150--67. 1975.

19. Smith. E w.. "A Search for Trans-missiblo Pathogenic Characters in InvasivoStrains of Escherdda colt: the Discovery ofPlasmld-Controlled Toxin and a Plasmld-Controlled Lethal Character Closely As-soclated, or Identical, with Collcln V,"Journal of General Microbiology. 83:95-111,1974.

20. Smith, H. W, and LT. B. Huggin, "Fur-ther Observations on the Asoclation of theColcin V Plasmld of Eserlchia colt withPathogenicity and with Survival In theAlimentary Tract." Journal of GeneralMicrobiology, 92:335-350. 1970.

21. Walton, J. and C. E. Snlth, "Transferof Ent Plasmnds" microbial Genetics Bul-letin, 38:10. 1975.

22. Meyers, J. A., L. U, Ehncs, K. Cot-tingham. S. Fallkow, 'Incidence of Cotrans-fer of Ent. E99 and R-Plasmld3," MicrobialGenetics Bulletin, 39:13, 1975.

23. Wachsmuth, I. K., S. Falkow, and R. W.Ryder, "Pl'amid-Medlated Properties of anEnterotoxigenle Escerlchfa coil Assoclatedwith Infantile Diarrhea," Infection and Im-munity. 14:403-407, 1976.

24. Heffron F., 0. Rubens, and S. Falkow,"Transportation of a Plasmld Deoxyribonu-cleIc Aclde Sequence that Mediates Amplcil-lin Resistance: Identity of Laboratory-Con-structed Plasmlds and Clinical Isolates,"Journal of Bacteriology,

. 129:530-533, 1977.

E. TISSUE RESxMUTS (caxrITMON 4)

1. The criterion. The FDA task forceexpressed concern about the effect of an-tibiotic residues In food ingested by manon the prevalence and resistance ofpathogenic bacteria in humans, and onpotential allergic or hypersensitivity re-actions. This resulted in Human HealthCriterion No. 4:

An antibacterial drug used at -ubthera-peutlc levels In the feed of animals shall notresult in residues of the parent compound.metabolites, or degradation products In thefood Ingested by man which are capable ofcausing (1) an Increase In the prevalence ofpathogenic bacteria, (2) an ncrease in theresistance of patogenlo bacteria to antibac-terial drugs used In human clinical medicine.

Controlled studies in appropriate test anl-mals shal be conducted to determinewhether the consumption of food producedby animals receiving antibacterial drugs willresult In:

(a) An Increase In the Intestinal flora ofthe prevalence of pathogenic bacteria;

(b) An increase In the degree and apec-trum of resistance of the intestinal flora todrugs used in human clinical medicine.

Experimental procedures shall Include ap-propriate consideration of maximum uselevel, minimum withdrawal time and estab-lished tolerances.

In addition, a literature survey shall beconducted to determine the Incidence of re-

56285

portz of hyperensItivity resulting from anti-bacterial drugs In food. The literature surveyhall Include Information regarding hyper-

consitivity reactions occurring as a result ofparenteral or topical exposure to antibac-terial drus as well as those Ingested In food.When hypersensitivity has been shown. ex-pcriments in appropriate laboratory animal--must be conducted to develop estimates ofvthat level of antibacterial drugs In food will,cause the production of hypersensitivity.

2. Bacl:ground. Mussman's 1975 reporton the United States Departmentof Agriculture's Drug Residue MonitoringProgram (Ref. 2). shows that tetracy-clines are among the antimicrobials con-sPtituting the bulk of violative residuesbecause they are used therapeuticallyand subtherapeutically. Violative oxytet-racycline and chlortetracycline residueswere also detected In 1975 and 1976.When Messersmith, et al., at AmericanCyanamid (Ref. 3) fed swine three tofive times the normal amount of chlortet-racycline, sulfamethazine, penicillincombination continuously for 14 weeks,they found residues of less than 1 partper million in alltissues sampled 0.5 and7 days after withdrawal The Food andDrug Administration conducted studiesin dogs, rats, and hamsters to find asuitable small animal model in which todetermine the no-effect level of anti-microbial drugs on the resistance char-acteristics of the enterie flora (Ref. 4). Indogs fed subtherapeutic oxytetracycline10 parts per million in their diet, thecoliform population shifted from pre-dominantly drug-sensitive to predomi-nantly drug-resistant coliforms. No suchchift in drug-resistance occurred in dozsfed oxytetracycline at 2 parts per millionor les. The study indicated a theoreticalpossibility for such a "no effect" level.

3. American Cyanamid study-(a) Ex-verimental design. American Cyanamidstudied the effect of tetracycline-con-taining chicken tissue on antimicrobialresistance In dogs. For this study, 450day-old chicks were divided into twogroups of 225 birds each. One group wasfed subtherapeutic chlortetracycline,while the other group was fed a non-medicated diet. The chickens were killedon days 55 and 56, and 200-gramn tissuesamples were prepared on days 58 and 59.

Two groups of 16 adult beagles were fedPurina Dog Chow for 20 days, and onthe 21st day the raw chicken was addedto this diet The dogs were fed until day40 according to the following design.

Tnafa- Daily ratmn

group Days 21 to 4 Das 41 to 9

A -M g. Purin Drg Chsr.. Purina Deg Chow2W x cblrken tiez ad ibltr.

(c5amedicated).B 23 9 Puri= Dcg Ch,-.. Do.

200 jr chicken tL-m-3( th chlsrtcr=-y-

Initially, the do- food and chickentissue were examined for Salmonelalactose-fermenting organisms (coil-forms). Additionally, raw and cookedchicken tissues from both groups ofbirds were assayed for chlortetracyclineresidues. Fresh fecal samples were ob-tained twice weekly from each dog-




NOTICES

and examined for Salmonella. ColiformsIn the feces were tested for sensitivity toampicillin, chloramphenical, chlortetra-cydline, and dihydrostreptomycin. Amer-ican Cyanamid also examined samplesof commercially purchased chicken forbacteria.

(b) Summary. Analyses of the dogfood and the raw chicken tissue revealedno Salmonella or coliforms. Salmonellawere isolated from the feces of only threedogs, and the isolations occurred on thesame day. None of the dogs exhibitedsigns of clinical salmonellosis.

The level of chlortetracycline residueIn the chicken tissue that was fed to thedogs varied from 0.025 part per millionin fat to 3.15 parts per million in kid-neys. The average concentration in thetissue samples was 0.26 part per million.

In dogs fed the raw chicken, the num-ber of chlortetracycline-resistant coli-forms shed increased significantly, as didthe number of coliforms resistant to di-hydrostreptomycin. Chicken tissue- con-taining chlortetracycline residues alsocarried two times as many coliforms astissue without chlortetracycline residuesdid. Further, chlortetracyclirie-contain-Ing tissue had four times more chortet-racycline-resistant organisms than didthe antibiotic-free tissue. Dihydrostrep-tomycin-resistant coliforms were presentat three times the number found in thecontrol tissues. Cyanamid also indicatesthat cooking tissues at 800 C for 20 min-utes may inactivate chlortetracyclineresidues. American Cyanamid also sur-veyed a few commercially purchasedpoultry specimens. The samples con-tained 1400o the number of 'coliformsfound in the raw tissue fed the animaLY(10 versus 1081).

(c) Director's analysis. The Directorfinds that the study has failed to estab-lislh conclusively a no-effect level for theselection of resistant organisms forchlortetracycline residues in raw chickentissue.

4. Literature survey. Some drug firmsconducted literature surveys on humanhypersensitivity to the tetracyclines andto the combination of tetracycline-sul-fonamide and penicillin. Anaphylactiereactions to penicilins are common; theymay occur as a. result of ingestion, con-tact, or occupational exposure. Derma-tological reactions to sulfonamides andto neomycins are frequent (Ref. 4 and5). The tetracyclines have producedphotoallergic and phototoxic reactions,and the hypersensitivity reactions rangefrom skin rashes to ingioedema andanaphylaxis. Moreover, cross-sensitiza-tion among the tetracyclines is com-monly observed. Although hypersensi-tivity reactions are rare, they are occa-sionally extremely severe (Ref. 6), andallergic reactions from a skin contactwith tetracyclines are common. For thisreason, hypersensitivity reactions- totetracycline and the tetracycline prod-ucts must be considered potentiallyharmful to man. However, there are noreported incidents of tetracycline hyper-sensitivity connected with, ingestion orhandling of tissue with tetracyclineresidues.

5. Director's conclusions. The Directorhas evaluated the literature and thestudies and concluded that the holders ofthe NADA's have failed to establish con-clusively a no-effect level for the tetra-cycline residues, although there is noevidence that below tolerance the resi-dues pose a public health problem inthese areas.

REERENCES

1. Rollins, L., S. Gaines, D. Pocurull, H. D.Mercer, "Animal Model for Determining theNo Effect Level of an Antimicrobial Drugon Drug Resistance in Lactose FermentingEnteric Flora," Antimicrobial Agents andChemotherapy, 7:661-665, 1975.

2. Mussman, H. C., "Drug and ChemicalResidues In Domestic Animals," FederationProceedings, 34: 2:197-201, 1975.

3. Messersmith, R. E., B. Sass, H. Berger,and.G. 0. Gale, "Safety and Tissue ResidueEvaluations in Swine Fed Rations Contain-Ing Chlortetracycline, Sulfamethazine and'Penicillin," Journal of American VeterinaryMedical Association, 151 (6):719-724, 1967.

4. Moller, H., "Eczematous Contact, Allergyto Oxytetracycline and Polymyxin B," Con-tact Dermatitis, 2:289-290, 1976. .

5. Carruthers, J. and E. Cronin, "Incidenceof Neomycin and Framycetn Sensitivity,"Contact Dermatitis, 2:269-270, 1976.

6. Schindel. L., "Clinical Side Effects of-the Tetracycliges,"' Antibiotics and Chemo-therapy, 13:300, 1965.

V. EFFECTIVENESS

In 1970-71 FDA issued a series of FED-ErAL REGISTER notices announcing theconclusions of the National Academy ofScience/National Research Council DrugEfficacy Study Group which evaluatedanimal feed premixes containing oxytet-racycline and chlortetracycline intendedfor subtherapeutc use. For mose of thoseproducts, the Director has previouslyissued notices either withdrawing ap-proval of the drugs or concluding thatthe the labeling claims were revised tocomport with the Academy's evaluation.The Director s proposing to completethe process in this notice in accordancewith the National Advisory Food andDrug Committee's recommendation thatFDA propose to limit subtherapeutictetracycline use in animal feed to unique,important claims. A condition precedentfor any claim is that it be supported bysubstantial evidence of effectiveness forthat claim.

A. OXYTETRACYCLINE

In the FEDERAL REGISTER Of May 5,1970 (35 FR 7089; DESI 8622V), FDAannounced the NAS/NRC evaluation ofPfizer's Terramycin TM- premixes, whichcontain-oxytetracycline quaternary salt.The NAS/NRC concluded that these pre-mixes were probably effective when usedfor the control and treatment of specificdiseases of livestock (swine, cattle, sheep,rabbits, and mink) and poultry (broilerchickens, laying chickens, and turkeys),and concluded that use may result infaster gains and improved feed efficiencyunder appropriate conditions. It also in-dicated that extensive labeling revisions,restrictions on the claims, and reword-ing of claims, for which further docu-mentation was required, were necessary.

The Food and Drug Administratunconcurred with the NAS/NRC's evalua-tion of the premixes and further con-cluded that:

(1) The claims lor hexmitlatis should beincluded under the susceptIblo host.

(2) Appropriate claims regarding fasterweight gains and Improved fecd efliclencyshould be stated as "For Increased rate ofweight gain and improved feed efficiency for(under appropriate conditions of use) ." (Id.)

E. CHLORTETRACYCLIJI.

I. Roche premixes. The Food and DrugAdministration announced the NAS/NRC's evaluation of Roche's Spence Spe-cial Premix (each pound contains 4grams chlortetracycline) and Ark-LaSpecial Swine Premix (each pound con-tains 2 grams chortetracycline hydro-chloride) in the FEDEnAL RrGIsTER of July9, 1970 (35 FR 11070; DESI 0173NV),

The Academy concluded that more In-foriaton was necessary to c-tabllsh theeffectiveness for faster gains and improv-ing feed efficiency In swine. It also dis-allowed claims for growth promotion orstimulation and Indicated that claimsfor faster gains and/or feed efflciencyshould be reworded. Finally, the NAS/NRC concluded that each active ingre-dient in a preparation containing morethan one drug must be effective or con-tribute to the effectiveness of the prep-aration to warrant acceptance as anactive ingredient.

The Food and Drug Administrationconcurred with this evaluation; however,the agency concluded that the appropri-ate claim for faster weight gains andimproved feed efficiency, if supported bysubstantial evidence, should be "For In-creased rate of weight gain and Improvedfee;d efficiency for (under appropriateconditions of use)." (Id.)

2. American Cyanamid and NoVco pre-mixes. In the FEDERAL REGISTER of July21, 1970 (35 FR 11646; DESI Ol3NV),the agency published the evaluation ofpremixes manufactured by AmericanCyanamid and Nopco containing chlor-tetracycline at levels ranging from 4 to50 grams per pound.

a. Aureomycln 50 Feed Premix; contains50 grams chlortetracyclino per pound.

b. Aureomycin MR Feed Premix; contains25 grams chlortetracyclino per pound.

c. Aureomycin 10 Feed Premix; contains10 grams chlortetraoyclne per pound.

d. Aurofac-DI contains grams chlortetra.cycline per pound.

e. Aureomycin Layer Brunch, contains 4grams chlortetracycline per pound.f. Deravet; contains 10 grams ablortetra-

cycline hydrochloride per pound.g. Aureomycin Soluble Powder; contalna

25 grams chlortetracyclino hydrochlorldoper pound.

h. Nopco CTC 4/SS; contains 4 grainschlortetracyclino per pound and 50 percentsodium sulfate.

i. Xopco CTC 6.66/SS; contains 6.6 grainschlortetracycllne per pound and 83,33 per-cent sodium sulfate.J. Nopco CTC 10, 25, 50, and 100; contain

10, 25, 50, and 100 grams of chlortetracyclInoper pound, respectively.

The NAS/NRC rated these productsas probably effective for growth promo-tion and feed efficiency and for the treat-ment of animal diseases caused by path-


56286



NOTICES

ogens sensitive to chlortetracycline. Italso reworded and restricted the claims.

The Food and Drug Administrationconcurred with these ratings, but itagain concluded that the appropriateclaim for faster weight gains and im-proved feed efficiency should be "For in-creased rate of weight gain and improvedfeed efficiency for (under appropriateconditions of use)" (Id.).

3. American Cyanamid's chlortetra-cycline and vitamin products. In theFEDERAL REGISTER of August 18, 1970 (35FR 13156; DESI 0115NV), FDA pub-lished the NAS/NRC evaluation ofAmerican Cyanamid's chlortetracyclineand. vitamin products:

a. Aureomycin Crumbles; each pound con-tains 2 grams of chlortbtracyclne, 250.000U.S.P. units of Vitamin A, and 25,000 U.S.P.units of vitamin D-3.

b. Aureomycin T.F.-5; each pound con-tains 5 grams of chlortetracycline and 0.5milligram of itamin B-12.c. Aureomycin TY.-15; each pound con-

tains 15 grams of chlortelacycline and 1.5milligrams of vitamin 13-12.

The NAS/NRC rated AureomycinCrumbles as probably not effective forprevention or treatment of bacterial in-fections or for increasing growth rate inswine, calves, beef cattle, sheep, andhorses. However, it concluded that Au-reomycin T.F.-5 and Aureomycin TF.-15 were probably effective for antibioticactivity in the control and treatment ofbacterial Infections in swine, calves,sheep, and poultry.

The NAS/NRC's reports indicate that(1) more information is necessary todocument the value of vitamins and theamounts of vitamins which are added tothe preparations, (2) substantial evi-dence was not presented to establish thateach ingredient designated as activemakes a contribution to the total effectclaimed for the drug combinations, and(3) the claims should be reworded andrestricted.

The Food and Drug Administrationagreed with the Acedemy's findings but itagain concluded that the standard word-ing for the faster weight gains and im-proved feed efficiency claims should beadopted if supported by evidence of effec-tiveness aId.).

4. Ralston Purina premix. The Foodand Drug Administration evaluated theNAS/NRC report on Purina AureomycinEtts Medicated (2 grams of chlortetra-cycline hydrochloride per pound), andpublished the results in the FEDERALREGISTER of July 22, 1970 (35 FR 11705;DESI 0035NV).

The Academy concluded that this vita-min-antibiotic preparation is probablynot effective for the therapeutic andnontherapeutic claims in hogs, cattle,and sheep. It found that the dose of thechlortetracycline to the animals is fre-quently low and inconsistent, and itquestioned the oral administration forseverely ill animals. The Academy alsoindicated that rewording and restrictionson the claims were necessary in additionto documentation of the value of vita-mins In this preparation.

The Food and Drug Administrationconcurred with the Academy's findings,but it concluded the agency's wording forthe faster weight gains and Improvedfeed efficiency claim where supported byevidence of effectiveness was more ap-propriate. (Id.)

C. DIRECTOR'S Co? cLUSIol.S

In accord with FDA'S conclusion toadopt the recommendation of the Anti-biotics in Animal Feeds Subcommittee ofthe National Advisory Food and DrugCommittee that the subtherapeutic useof tetracycline in animal feed be limitedto unique, essential claims, the Directorhas evaluated all of the Informationavailable concerning the effectiveness ofchlortetracycline and oxytetracyclinepremixes for subtherapeutic use. Basedon this review, the Director is proposingto restrict the use of chlortetracyclineand oxytetracycline in animal feed to thefollowing subtherapeutic conditions ofuse:

OXYTTRCYCLUM

(1) For chickens at 100 to 200 gramsper ton of feed as an aid In control offowl cholera caused by Pasteurella mul-tocida. At lQ0 to 200 grams per ton offeed as an aid in thecontrol of Infec-tious synovitis caused by yfycoplasmasynoviae susceptible to oxytetracycline.

(2) For turkeys at 200 grams per tonof food for the control of infectious syn-ovitis caused by Mycoplasma synoclaesusceptible to oxytetracycline.

CHLORTETRACYCLINE

(1) For chickens at 100 to 200 gramsper ton of feed as an aid in the controlof infectious synovitis caused by M.synoviae susceptible to chlortetracycline.

(2) For turkeys at 200 grams per tonof feed as an aid In the control of In-fectious synovitis caused by 11. inozdacsusceptible to chlortetracycline.

(3) For beef cattle at 0.5 mlligram/pound of body weght per day for con-trol of active infections of pnaplasmoss.

(4) For beef cattle at 350 milligramsper head per day In combination withsulfamethazine as an aid In the mainte-nance of weight gains In the presenceof respiratory disease such as shjppingfever.

(5) For breeding sheep at 80 millI-grams per head per day as an aid Inreducing the incidence of vibronle abor-tion.

The safe and effective new animalsubstitutes for the subthempeutic tetra-cycline uses that the Director Is pro-posing to withdraw are contained inSubpart B of 21 CFR Part 558. The drugsand their approved conditions of useare codified as follows:Arsanllate sodium -------------- 558. 60Arsanilie acid ------------------ 5538.62Bacitracin --------------- 558.76, 558.78Bambermycln ------------------ 558.95Carbadox ---------------------- 58.115Carbasone ------------------- 58. 120Erthromycln ------------------ 58.248Hygromycln B .................- 558.724Lincomyci -------------- 58.325afonensin ---------------------- s58.355

Olcandomycln --------..... --- 558.435P a.yrs ne ----------------------- 558.530Sulfadimethoinne-ormetoprim ---- 553.575Virgnlamyln -58.635

1. Relhey, E. J, VT. I. Brock, L 0. Ei;ewer,E. W. Jones, and R. G. White, "Low Levelsof Chlortetracycline for Anaplasmosis,"Journal of Science, 43:232.1976.

2. McCallon. B. n., Hyattsville, Md., Jour-nal of Dairy Science, 59:1171-74, 1976.

3. Franliin. T. E., R. V. Cow. D. J. Ander-con, and K. , Xuttler "Melium and LawLevel Fe d ing of Chlortetracycline withComparlon to Anaplasmozis Cr and CATeot ." Southwe tern Veterinarian, 20:101-101. 1067.

4. Bro2:. Vt. E., C. C. Pearson, and L 0.Ellewer. "Anaplmmasi Control By Test andSubcequent Treatment with Chlortetracy-cine." Prozeedinges of the Annual Meeting.U.S. Livestock San. A. pp. 66-70,1959.

5. Tviehau. 1.1. J, "Control of Anaplasmo-cis by Feeding An Antibiotic (Aureomycin) ,"Proceedings of the 4th National Anaplas-mnins Conference. Reno, Nevada, pp. 48-49,1062.

0. Jones. E. W. and W. E. Brock, "vineAnapasmosis: It3 Diagnosis, Treatment andControl." Journal of the American VeterinaryMedical Association. 149:1624--1633, 1968.

7. Ryfr, J. P, H. Breen, and G. J. Thomas."Control of Anaplasmosis Under WyomingConditions." Journal of the American Veter-iary Medical Association. 145:43-46,1964.

8. "Near-om's Sheep Disease." 3d Ed.. p.57. Hadleigh Marsh (editor), Wiliams & Wi.-hlin Co., 1905.

9. Fran F. W. r. H. Scrivner, T. W. Bailey,and W. A. Melnershagen. "Chlortetracyclineas a Preventative of Vibrionlc Abortion inSheep," Journal of the American VeterinaryAssoclation. 132:2t--26,1958.

10. Frank, F. W, W. A. Melnershagen, L. KLSCrlvner, and J. V. Bailey, "Antibiotics forthe Control of Vlbriosis in Ewes," AmericanJournal of Veterinary Research, 20:973-976,10S9.

11. Hulet, C. V., S. 0. Ercanbrack, D. A.Price, B. D. Humphrey. P. W. Frank, and W.A. Melnershagen, "Effects of Certain Antibi-otis in the Treatment of VlbrIosLs In Sheep."American Journal of Veterinary Research,21:441-444, 1960.

12. "Diseases of Poultry,' 6th Ed., p. 328,1. 0. Olson (editor), Iowa State University

Press, 1972.13. Snoeyenbcs, G. H., K L Basch, and 1L

Sovolan, 'Infectious Synovitts: I=, Drug Pro-phylawsi and Therapy," 1968.

14. Olson, N. 0, D. 0. Shelton, J. M. Blet-nr. and C. E. Weakley, "Infectious SynovitiControl. I1. A Comparison of Levels of Anti-biotics," American Journal of Veterinary Re-r.carch, 18:200-203, 1957.

15. FDA summary of confidential evidence,of effectiveness for control of fowl cholera byos-tetracycline.

VI. Co:NCLUSIoIq

Pursuant to §558.15, the holders ofapproved NADA's for tetracycline-con-taning drug products Intended for ad-dition to animal feeds at subtherapeuticlevels have the burden of establishingthat this extensive use is safe in accord-ance with the criteria andtguidelines es-tablished by that regulation, in additionto the basic requirements imposed by thegeneral safety provisions of the FederalFood, Drug, and Cosmetic Act. The Di-rector, in this notice, has set forth indetail the basis for the criteria andguldelines'implementing the regulationand this action. The holders of the ap-


56287



NOTICES

proved NADA's have failed to satisfy thelegal requirements imposed by the regu-lation; they have failed to resolve thebasic safety questions that underlie thewidespread subtherapeutic use of tetra-cycline in animal feed.

(a) Bacteria-bearing R-plasmidswhich confer resistance to multiple an-tibiotics have become widespread in theenvironment of man and animals. Anti-biotic resistance, medicated by transfer-able R-plasmids, is increasing in E. coli,and Salmonella, and other pathogens.The resistance patterns isolate from ani-mals are similar to those in their normalintestinal E. coli population, and thereis evidence that antibiotic resistance inpathogens can derive from the normalflora by means of R-plasmid transfer.There are well-established routes for thetransmission of bacteria between ani-mals and man. The R-plasmids found inbacteria isolated from man and animalsare indistinguishable, and commonserotypes of these organisms infect bothman and animals.

Studies in chickens, swine, and cattlesubmitted by the holders of approvedNADA's confirm that the subtherapeuticuse of the tetracyclines will cause an in-crease in the prevalence of R-plasmid-bearing organisms in animal intestinalflora.

(b) Antibiotic resistance in Salmonellacan lead to an increase in szhedding andtherefore contribute to an increase inthe Salmonella reservoir. The potentialfor harm arising from a compromise oftherapy is well documented. The studiessubmitted, however, are of insufficientscope and design to demonstrate con-clusively that the extensive use of. sub-therapeutic tetracycline is safe. Epide-miological studies assessing the long-term impact of the increase in R-plas-mids on the effectiveness of antibioticswould aid in assessing the extent of theproblem.

(c) Evidence demonstrates that R-plasmids controlling pathogenicity, drugresistance, and ability to colonize theintestines can and do cotransfer in vitroand in vivo.

(d) For tissue residues of tetraclines,FDA studies indicate that a theoreticalno-effect level exists for development of-transmissible antibiotic resistance (R-plasmid-mediated resistance). AmericanCyanamid's study and the literature sur-veys have failed to establish conclusivelythis no-effect level, although evidencefrom the Cyanamid study suggests thatheating the tissue may inactivate thetetracycline residues.

(e) Under 21 CER 558.15, the holdersof approved NADA's were required to filecommitments to conduct studies thatwould resolve conclusively the safety ofthe subtherapeutic use of antibiotics inanimal feeds and then to conduct thosestudies. To assure compliance with thelatter requirement, the regulation re-quired holders of the approved NADA'sto file periodic progess reports on the

studies. The Director Is proposing towithdraw approval of certain NADA's forwhich evidence was submitted pursuantto § 558.15 to resolve the safety issues,although he is unaware of any sponsorthat filed a commitment to conduct therequisite studies but submitted no evi-dence. Nevertheless, the Director con-cludes that the approval of any NADAfor which a commitment to conduct ap-propriate studies was filed but whoseholder filed no evidence should be with-drawn on the grounds that the holderof the NADA had failed to establish andmaifitain records and make reports asrequired by appropriate regulation.

(f) Finally, the NADAholders have theburden of demonstrating that their prod-ucts ar6 effective for the indications ofuse. The Director has evaluated theavailable evidence on all subtherapeuticclaims for effectiveness of tetracycline-containing premixes in conjunction withthe recommendation of the Antibiotics inAnimal Feed Subcommittee of the Na-tional Advisory Food and Drug Commit-tee that products be restricted to theclaims that are effective and unique andthe NAS/NRC's evaluation of thesepremixes.

On the basis of the foregoing analysis,the Director is unaware of evidence thatsatisfies the requirements for demon-strating the.safety of extensive use ofsubtherapeutic tetracycline-containingpremixes established by section 512 ofthe Federal Food, Drug, and CosmeticAct and § 558.15 of the agency's regula-tions. Accordingly, he concludes,, on thebasis of new information before him withrespect to these drug products, evaluatedtogether with the evidence available tohim when they were originally approved,that the drug products are safe only forthe limited conditions of use set forthabove.

Therefore, the Director announces heis proposing to withdraw all approvalsfor tetracycline-containing premix prod-ucts intended for subtherapeutic uses inanimal feed, other than those cited,whether granted under section 512 ofthe act or section 108(b) of the AnimalDrug Amendments of 1968 on thegrounds that they have not been shownto be safe and lack substantial evidenceof effectiveness for therapeutic use.Notice is hereby given to holders of theapprovals listed above and to all otherinterested parties. If a holder of an ap-proval or any other interested personelects to avail himself of an opportunityfor hearing pursuant to sections 512(e)(1) (B), 512(e) (1) (C), and 512(e) (2) (A)and § 514.200 (21 CPR 514.200Y, triepartymust file with the Hearing Clerk (HFC-20), Food and Drug Administration, Rm,4-65, 5600 Fishers Lane, Rockville, Md.20857, a written appearance requestingsuch a hearing by November 21,1977, andproviding a well-organized and full-factual analysis of the scientific andother investigational data that such per-son is prepared to prove by January 19,

1977, in support of Its opposition to the,Director's proposal. Such analysis shallinclude all protocols and underlying rawdata and should be submitted In accord-ance with the requirements of § 314.200(c) (2) and (d) (21 CFR 314.200 (c)(2),and (d)).

The failure of a holder of an approvalto file timely written appearance andrequest for hearing as required by 9 514.-200 constitutes an election not to availhimself of the opportunity for a hear-ing, and the Director of the Bureau ofVeterinary Medicine will summarilyenter a final order withdrawing theapprovals.

A request for a hearing may not restupon mere allegations of denials, but itmust set forth specific facts showing thatthere is a genuine and substantial issueof fact that requires a hearing. If It con-clusively appears from the face of thedata, information, and factual analysesin the request for hearing that there is nogenuine and substantial issue of fact thatprecludes the withdrawal of approval ofthe application, or when, a request forhearing is not made in the required for-mat or with the required analyses, theComnissioner will enter summary Judg-ment against the person who requests ahearing, making findings and conclu-sions, denying a hearing.

Four copies of all submissions pursu-ant to this notice must be filed with theHearing Clerk. Except for data and In-formation prohibited from public dis-closure pursuant to 21 U.S.C. 331(J) or18 U.S.C. 1905, responses to this noticeand copies of references cited In this no-tice not appearing in Journals designatedby 21 CFR 310.9 and 510.95 may be seenin the office of the Hearing Clerk, Foodand Drug Administration, between 9 ain.and 4 p.m., Monday through Friday.

If a hearing is requested and is Jus-tified by the applicant's response to thisnotice of opportunity for hearing, theissues will be defined, an administrativelaw judge will be assigned, and a writtennotice of the time and place at whichthe hearing will commence will be issuedas soon as practicable.

The Director has carefully consideredthe environmental effects of this action,and because it will not significantly affectthe quality of the human environment,he has concluded that an environmentalimpact statement is not required for thisnotice. A copy of the environmental im-pact assessment is on file with the Hear-ing Clerk. Moreover, in a iotice pub-lished in the FtDEDAL RECISTEn of May 27,1977 (42 FR 2739), the Commissioner ofFood and Drugs requested data concern-ing the potential environmental Impactof a series of regulatory actions, includ-ing this one, designed to restrict the sub-therapeutic use of antibacterials In ani-mal feeds. If the public discussion andinformation gathered warrant, a com-prehensive environmental impact state-ment will be prepared, evaluating the

FEDERAL REGISTER, VOL: 42, NO. 204-FRIDAY, OCTOBER 21, 1977'

56288



NOTICES

impact of all the actions as a singleprogram.

The Director has also carefully con-sidered the economic impact of the no-tice, and no major economic impact, asdefined in Executive Order 11821 (asamended by Executive Order 11949),OMB Circular A-107, and Guidelines Is-sued by the Department of Health, Edu-cation. and Welfare. has been found. Acopy of the FDA inflation impact as-sessment is on file with the HearingClerk, Food and Drug Administration.

This notice is issued under the Fed-

eral Food. Drug. and Co-metic Act (see.512,-82 Stat 343-351 121 U.S.C. 360b))and under authority delegated to theCommissioner of Foad and Drugs (21CFR 5.1) and redelegated to the Directorof the Bureau of Veterinary Medicine(21 CFP. 5.84).

Dated: October 14. 1977.

C. D. VAz Houwrx;o,Director, Bureau of Veterinary

ZMfedicine.IJ R Doc.77-30G8 Filed 10-17-'7;3:03 pm]


56289



Documents

FDA Notion of Opportunity for Hearing re tetracyclines, 1977