Efficacy of the Antimicrobial Compound U-82,127 as a Growth Promoter for Growing-Finishing Pigs1

G. L. Cromwell*,2, G. W. Davis, W. E. Morgan Morrow, R. A. Primo, D. W. Rozeboom,, M. D. Sims#, E. P. Stanisiewski**, and C. H. Ho**
*Animal Sciences Department, University of Kentucky, Lexington 40546; Greenbriar Veterinary Services, Delaware, OH; North Carolina State University, Raleigh; Ponderosa Research Co., French Village, MO; Michigan State University, East Lansing; #Virginia Diversified Research, Harrisonburg; and **Pharmacia & Upjohn, Kalamazoo, MI

ABSTRACT: The antimicrobial compound U82,127 (Pharmacia & Upjohn, Kalamazoo, MI) is a thiopeptide that belongs to a series of cyclic peptide antibiotics produced by Streptomyces arginensis. It is active mainly against Gram-positive organisms. A study involving 576 growing-finishing pigs was conducted at six locations to assess the efficacy of the growth-promoting compound from approximately 19 to 89 kg BW. The basal diet was an unmedicated cornsoybean meal diet fortified with vitamins and trace minerals and containing 16% CP (.80% lysine) during the growing stage (to 54 kg) followed by 13% CP (.60% lysine) during the finishing stage. Dietary dose concentrations of the antimicrobial compound were 0, 3.3, 6.6, and 9.9 mg/kg. At each location, there were six replications of four pigs (two barrows and two gilts) per pen. Diets and water were available for ad libitum consumption. The antimicrobial was provided in coded bags, and investigators were blind to the

treatments. The ADG during the growing stage was improved by all levels of the antimicrobial ( P < .04), but only the 6.6 mg/kg level improved ADG during the finishing stage ( P < .03). Feed:gain was improved by all concentrations of the antibiotic ( P < .01) during the growing stage and by the two lower levels of the drug ( P < .06) during the finishing stage. Over the entire study, the antimicrobial compound improved ADG (linear, P < .06) and feed:gain (quadratic, P < .01; minimum feed:gain was at 6.2 mg/kg). The lowest dose with a 90% confidence interval of its predicted value not overlapping with the predicted value of the control was 2.3 mg/kg; thus, the efficacious dose range for improving feed/gain was between 2.3 and 6.2 mg/ kg. Neither death loss nor pig removal from the experiment was affected by treatment. The results indicate that the antimicrobial compound U-82,127 is an effective growth-promoting agent for growingfinishing pigs.

Key Words: Pigs, Antibiotics J. Anim. Sci. 1996. 74:12841287

Introduction
The antibiotic U-82,127 is a thiopeptide that belongs to a series of cyclic peptide antibiotics produced by Streptomyces arginensis isolated from soil. The U-82,127 complex consists of a group of at least five antibacterial agents having physical properties similar to those reported for the sulfomycin group of antibiotics. The complex is active mainly against Gram-positive organisms. In a 3-wk experiment involving weanling pigs, U82,127 fed at a dietary concentration of 55 mg/kg

improved ADG by 21% and feed:gain ratio by 8.3% compared with untreated controls (Evans et al., 1987). In a subsequent study conducted at five locations (Ogilvie et al., 1989), growing-finishing pigs fed U-82,127 at 19.8 or 39.6 mg/kg gained faster (6.1 and 3.6%, respectively) and required less feed per unit of gain (2.7 and 2.4%, respectively) than controls. This research represents a coordinated study to assess further the efficacy of U-82,127 as a growth promoter for growing-finishing swine and to determine the most efficacious dose of the antibiotic.

1Paper no. 95-07-103 of the Kentucky Agric. Exp. Sta., Lexington

Experimental Procedures Locations. Experiments were conducted at six locations representative of major pork producing areas in the United States. Each station had 24 pens capable of

40546. 2To whom correspondence should be addressed. Received September 21, 1995. Accepted January 29, 1996.

1284

ANTIMICROBIAL COMPOUND FOR SWINE

1285

housing four pigs per pen up to market weight. A common protocol was used across the six stations. The stations were located at Lexington, KY (University of Kentucky); Columbus, OH (Greenbriar Veterinary Services); Raleigh, NC (North Carolina State University); French Village, MO (Ponderosa Research Co.); East Lansing, MI (Michigan State University); and Harrisonburg, VA (Virginia Diversified Research). Animals. A total of 576 Yorkshire or crossbred pigs were used in this study. The pigs originated from either closed farrow-to-finish operations or from assembled feeder pigs, and they initially averaged between 14 and 18 kg at the six locations when they arrived at the research units. The pigs were subjected to the routine health program regularly used by the respective investigator, including vaccination and parasite control. No additional feed medication was used during the study. Injectable medications were used when needed at the discretion of the investigator, but their usage was minimal. Housing. The pigs were housed in temperaturecontrolled buildings with total or partially slotted floors at five stations and in a modified-open-front unit at one station (MI). The experiments were started in November, January, or February, and the ambient temperature ranged from 0C ( M I ) to 21C (MO and NC) during the experimental periods. Solid pen dividers were maintained between each pen, which prevented feces and urine from a pen contaminating an adjacent pen. Feed was available on an ad libitum basis from self-feeders. Water was supplied from nipple or fountain waterers. Allotment and Records. Pigs were allotted in a randomized complete block design to four treatments (dietary concentrations of the antibiotic) from outcome groups based on initial weight, sex, and ancestry. Each pen consisted of two barrows and two gilts, and there were six pens (blocks) per treatment at each station. Thus, each station used a total of 24 pens and 96 pigs. Allocation of pigs to blocks occurred 7 d before the start of the experiment. During the 7-d adjustment period, a non-medicated diet (fortified corn-soybean meal diet, 16% CP) was fed to all treatment groups. During this preliminary period, any pigs that became ill or lame or that did not gain weight were replaced with a pig of the same sex and similar weight. The pigs were then reweighed on d 7 (average weight, 19.2 kg) and the experiment was started. Pigs were weighed and feed intake was determined at biweekly intervals. All pigs that died or that were removed because of weight loss were necropsied. Dietary Treatments. Corn-soybean meal-based diets, fortified with minerals and vitamins to meet or exceed NRC (1988) standards, were fed at each station. A 16% CP diet (.80% lysine) was fed from the beginning of the experiment until the pigs averaged 54 4.5 kg (growing stage), then the CP level was reduced to 13% (.60% lysine) for the remainder of the experi-

ment (finishing stage). All pens within a block were switched from the growing diet to the finishing diet at the same time. The four dose levels of U-82,127 (Pharmacia & Upjohn, Kalamazaoo, MI) added to the basal diet were 0, 3.3, 6.6, and 9.9 mg/kg. The drug was provided to the investigators in premix form in coded bags; thus, the investigators were blind to the treatments. Premixes were shipped on an as needed basis, and all diets were fed within 30 7 d after preparation. Samples of diets were collected and assayed for drug concentration by the manufacturer. Termination. The experiment was terminated on a block basis when the average weight of the pigs reached approximately 91 kg. The average final weight was 88.5 kg. After termination, the treated pigs were disposed of by burial or incineration. Statistical Analysis. The performance data (ADG, feed/gain) for the growing phase, finishing phase, and growing-finishing phase were pooled across stations and statistically analyzed as a randomized complete block design (Steel and Torrie, 1980) using the GLM procedure of SAS (1990). The pen was considered the experimental unit. The location effects were considered random, hence the interaction of location and dose was used as the error term for testing the dose effects. Statistical tests were conducted at a = .05. The three degrees of freedom for dose were partitioned into linear, quadratic, and remainder in order to describe the nature of the dose response for the drug. When the quadratic term was significant, the procedure of Walker and Carmer (1967) was used to identify where the plateau began, and the minimally effective dose was estimated based on the method of nonoverlapping confidence intervals. Contrasts also were made to compare individual doses of the antibiotic to the control. Homogeneity of variances among treatments also were tested (Shukla, 1982).

Results and Discussion

Assays of the diets mixed at each station indicated that the concentrations of U-82,127 were very close to targeted levels. The average concentrations were .1, 3.4, 6.4, and 9.5 mg/kg for the four diets targeted to contain 0, 3.3, 6.6, and 9.9 mg/kg, respectively. Residuals showed no violation of assumptions regarding normality and the constant variance of the error term for both ADG and feed:gain. Variances among treatment groups were found to be homogeneous by the Shukla test (Shukla, 1982) for both ADG ( P = .15) and feed:gain ( P = .31) at a 1% significance level. Health-related observations (Table 1 ) were few and not related to treatment. A total of 11 pigs died or were removed (1.9% of those that started on test) from the six locations during the course of the study,

1286

CROMWELL ET AL.

Table 1. Summary of health related incidences, removals, and deaths of pigs fed the antibiotic U-82,127a
U-82,127, mg/kg Item No. of pigs with Abscesses ( 2 ) b Lameness or injury ( 5 ) Respiratory ( 3 ) Diarrhea ( 2 ) Tail biting ( 1 ) Otherc ( 4 ) No. pigs removed or died ( 5 ) 0 6 1 6 9 1 1 2 3.3 5 1 2 8 1 4 4 6.6 7 2 1 8 1 1 2 9.9 7 4 4 7 2 2 3

a144 pigs per treatment. bNumber in parentheses indicates the number of stations with

this condition. cGaunt, staggering-vomiting, posterior paresis, nasal discharge, or rectal prolapse.

and there was no obvious relationship of treatment to pig removal or death (1.4, 2.8, 1.4, and 2.1% for the four treatments, respectively). Table 2 gives least squares means for the growing stage (19 to 54 kg), finishing stage (54 to 89 kg), and the entire growing-finishing period. During the growing stage, pigs fed the three levels of U-82,127 gained more quickly ( P < .01, P < .01, and P < .04 for the 3.3, 6.6, and 9.9 mg/kg doses, respectively) and more efficiently ( P < .01 for each dose) than pigs fed the nonmedicated control diet. The responses resulting from the three levels of the antibiotic were best

described as linear ( P < .01) for gain and quadratic ( P < .04) for feed:gain. During the finishing stage, only the 6.6 mg/kg dose level was different ( P < .03) from the control with respect to growth rate. Also during the finishing stage, feed:gain ratios of pigs fed the 3.3 and 6.6 dose levels were improved over those of controls, with the improvements approaching significance ( P < .06). Over the entire growing-finishing period (Table 2), the feeding of U-82,127 resulted in numeric improvements in growth rate, with the 6.6 and 9.9 doses being different from controls ( P < .01 and P < .06, respectively). The ADG response to increasing dose of the antibiotic tended to be linear ( P < .06). Although the pattern seems to be curvilinear, the quadratic component was not significant ( P = .11). Feed:gain of the pigs was improved by all three levels of the antibiotic ( P < .01, P < .01, and P < .03, respectively), and the response trend was both linear ( P < .04) and quadratic ( P < .01). Fitting a quadratic polynomial to the means resulted in a minimum feed:gain at 6.2 mg/ kg (Figure 1). With 95% confidence, the lower limit of this estimate was 5.2 mg/kg, which estimates the beginning of the plateau. Although the beginning of the plateau was estimated at 5.2 mg/kg, the 3.3 mg/kg dose, by pair-wise comparison, was different from the control ( P < .01) and was not different from the 6.6 ( P = .88) or 9.9 ( P = .58) dose groups. Because the quadratic curve changed rapidly, the beginning of the plateau, as determined by the method of Walker and Carmer

Table 2. Performance of pigs fed the antibiotic U-82,127 during the growing, finishing, and growing-finishing periods
U-82,127, mg/kg Item No. of pigs starting test No. of pigs finishing test Avg. weight, kg Initial End of growing stage End of finishing stage Growing stage Daily gain, kga Feed:gainb Finishing stage Daily gain, kgc Feed:gaind Growing-finishing stage Daily gain, kge Feed:gainf 0 144 142 19.2 53.6 87.2 .752 2.49 .870 3.30 .802 2.90 3.3 144 140 19.1 54.5 88.7 .776 2.42 .882 3.22 .817 2.82 6.6 144 142 19.1 54.9 89.4 .781 2.41 .898 3.23 .830 2.82 9.9 144 141 19.2 55.0 88.8 .783 2.40 .872 3.28 .820 2.84 CV 6.3 4.3 7.1 6.5 5.7 4.6

aLinear effect ( P < .01). Control vs 3.3 ( P < .04), 6.6 ( P < .01), and 9.9 ( P < .01) mg/kg doses. bQuadratic effect ( P < .04). Control vs 3.3 ( P < .01), 6.6 ( P < .01), and 9.9 ( P < .01) mg/kg doses. cControl vs 6.6 mg/kg dose ( P < .03). dControl vs 3.3 ( P < .06) and 6.6 ( P < .06) mg/kg doses. eLinear effect ( P < .06). Control vs 6.6 ( P < .01) and 9.9 ( P < .06) mg/kg doses. fQuadratic effect ( P < .01). Control vs 3.3 ( P < .01), 6.6 ( P < .01), and 9.9 ( P < .03) mg/kg doses.

ANTIMICROBIAL COMPOUND FOR SWINE

1287

and from 1.3 to 5.6% for feed:gain for the entire growing-finishing period. The improvements in growth rate and feed:gain from the feeding of U-82,127 to pigs are similar to the improvements of 3.6 to 6.1 for gain and 2.4 to 2.7% for feed:gain reported by Ogilvie et al. (1989). However, their improvements in gain and feed:gain were achieved with considerably higher dose levels of U82,127 (19.8 or 39.6 mg/kg).

Implications
Results of a coordinated experiment involving 576 growing-finishing pigs indicate that a new antibiotic (U-82,127; Pharmacia & Upjohn) is effective in improving growth rate and feed:gain of growingfinishing pigs. Overall improvements in growth rate and feed:gain ratio of 2.5% were achieved by feeding the antibiotic. The most effective dose for feed:gain was in the range of 2.3 to 6.2 milligrams per kilogram of diet.

Figure 1. Fitted dose-response curve and 90% confidence intervals for feed:gain of pigs fed U-82,127 during the growing-finishing stage. The minimum feed:gain is at 6.2 mg/kg, and the lowest effective dose that does not overlap with the confidence interval of the control is 2.3 mg/kg.

Literature Cited
(1967), was 5.2 mg/kg, which was close to 6.2 mg/kg (the dose level resulting in the lowest feed:gain ratio). Therefore, a method of non-overlapping confidence intervals (Condon and Lehmann, 1973) was used to determine the lower level of the efficacious dose range (Figure 1). The lowest dose with 90% confidence interval of its predicted value not overlapping with that of the predicted value of the control was 2.3 mg/ kg. This means that any dose level greater than or equal to 2.3 mg/kg was considered to result in a significantly lower feed:gain ratio when compared with the control. The dose of 6.2 was selected as the upper limit of the effective dose range because of the nature of the quadratic curve. Therefore, the efficacious dose range in the diet for improving feed:gain in pigs over the entire growing-finishing stage is between 2.3 and 6.2 mg/kg of U-82,127. The improvements over the control treatment in growth rate resulting from U-82,127 (averaged across all dose levels) were 3.7% during the growing stage, 1.6% during the finishing stage, and 2.5% for the entire experiment. Improvements in feed:gain from the antibiotic during the three stages were 3.2, 1.7, and 2.5%, respectively. These improvements are within the range of magnitude of those reported for other antibiotics for pigs of this general weight range (Hays, 1977; Zimmerman, 1986; Cromwell, 1991). The improvements were fairly consistent among the six locations, ranging from .4% to 5.9% for growth rate
Condon, R. J., and R. P. Lehmann. 1973. The responsibility of FDA to the livestock industry and the consumer. V. Statistical evaluation and interpretation of experimental data. J. Anim. Sci. 37: 221. Cromwell, G. L. 1991. Antimicrobial Agents. In: E. R. Miller, D. E. Ullrey, and A. J. Lewis (Ed.) Swine Nutrition. pp 297314. Butterworth Heinemann Publishing, Stoneham, MA. Evans, R. A., D. D. Kratzer, G. R. Bos, R. W. Simmons, and L. L. Dinvald. 1987. The effects of antibiotic 10381b on the growth performance of young pigs. Tech. Rep. No. 507-7922-87-002. The Upjohn Co., Kalamazoo, MI. Hays, V. W. 1977. Effectiveness of Feed Additive Usage of Antibacterial Agents in Swine and Poultry Production. Office of Technology Assessment, U.S. Congress, Washington, DC. [Edited version: Hays, V. W., 1981. The Hays Report. Rachelle Laboratories, Long Beach, CA.] NRC. 1988. Nutrient Requirements of Swine (9th Ed.). National Academy Press, Washington, DC. Ogilvie, M. L., R. A. Evans, K. R. Cox, D. D. Kratzer, and S. G. Carmer. 1989. Evaluation of the growth promoting efficacy and effect on feed conversion of U-82,127 in growing-finishing swine. Tech. Rep. No. 509-7922-89-004. The Upjohn Co., Kalamazoo, MI. SAS. 1990. SAS/STAT Users Guide (Version 6, 4th Ed.). SAS Inst. Inc., Cary, NC. Shukla, G. K. 1982. Testing the homogeneity of variances in a twoway classification. Biometrika 69:411. Steel, R.G.D., and J. H. Torrie. 1980. Principles and Procedures of Statistics: A Biometrical Approach (2nd Ed.). McGraw-Hill Publishing Co., New York. Walker, W. M., and S. G. Carmer. 1967. Determination of input levels for a selected probability of response on a curvilinear regression function. Agron. J. 59:161. Zimmerman, D. R. 1986. Role of subtherapeutic levels of antimicrobials in pig production. J. Anim. Sci. 62(Suppl. 3):6.