Prophylactic administration of a combined prebiotic and probiotic, or therapeutic administration of enrofloxacin, to reduce the incidence of bacterial chondronecrosis with osteomyelitis in broilers

INTRODUCTION

Tight junctional complexes comprise a key component of the intestinal barrier by sealing the apical surfaces of adjacent epithelial cells. “Leaky” tight junctions provide paracellular portals through which pathogenic bacteria can cross the gastrointestinal epithelium and ultimately enter the systemic arterial circulation. This process of bacterial leakage across the intestinal epithelial barrier, known as bacterial translocation, can lead to the hematogenous distribution of pathogenic bacteria throughout an animal’s body. Factors known to modulate the integrity of existing tight junctions and influence the dynamic synthesis of new tight junction proteins include physiological stress and “crosstalk” (direct cell to cell signaling) between gastrointestinal epithelial cells and commensal or pathogenic bacteria of the intestinal microbiome (Saunders et al., 1994; Ando et al., 2000; Steinwender et al., 2001; Ulluwishewa et al., 2011; Pastorelli et al., 2013). Recent studies demonstrated that heat stress and enhanced intestinal microbial challenges can impair the integrity of tight junctions and facilitate bacterial translocation across the epithelium of the small intestine in broilers (Quinteiro-Filho et al., 2010, 2012a,b; Murugesan et al., 2014). It also has been demonstrated that direct fed microbial (DFM) probiotics alone or in combination with a mannanoligosaccharide (MOS) prebiotic derived from yeast cell walls can attenuate intestinal barrier dysfunction in broilers challenged by heat stress or pathogenic bacteria (Sohail et al., 2010, 2012; Murugesan et al., 2014; Song et al., 2014). Commensal and probiotic bacterial species that promote intestinal barrier function by enhancing tight junction protein expression and the formation of occlusive tight junctional complexes also are effective in preventing bacterial translocation (Ulluwishewa et al., 2011; Pastorelli et al., 2013).

Bacterial translocation and bacteremia are key components in the pathogenesis of a form of lameness known as bacterial chondronecrosis with osteomyelitis (BCO) in broilers. Pathognomonic BCO lesions develop in the epiphysis (growth plate) and metaphysis at the proximal ends of the femora and tibiae where broad columns of poorly mineralized and structurally immature chondrocytes are susceptible to mechanical damage (osteochondrosis). Osteochondrotic microfractures and clefts expose a collagenous matrix that facilitates adhesion and infection by opportunistic bacteria. Fenestrated capillaries supplying the epiphysis and metaphysis provide a direct route through which hematogenously distributed bacteria can exit the bloodstream to colonize the microfractures and clefts. Bacterial proliferation and the ensuing immunological response create the necrotic abscesses and voids in the proximal femora and tibiae that are characteristic of BCO (Nairn and Watson, 1972; Howlett, 1980; Thorp et al., 1993; McNamee et al., 1998, 1999; McNamee and Smyth, 2000; Smeltzer and Gillaspy, 2000; Wideman et al., 2012; Wideman and Prisby, 2013). BCO has been diagnosed worldwide, is considered the most common cause of lameness in broilers, and is estimated to affect approximately 1.5% of all broilers grown to yield weights (past 42 d of age) in the United States (Pattison, 1992; McNamee et al., 1998; Butterworth, 1999; McNamee and Smyth, 2000; Bradshaw et al., 2002; Dinev, 2009; Wideman et al., 2012). Multiple opportunistic organisms in mixed cultures have been isolated from BCO lesions, including predominately Staphylococcus spp., Escherichia coli, and Enterococcus spp. (Wise, 1971; Butterworth, 1999; McNamee and Smyth, 2000; Martin et al., 2011; Wideman et al., 2012).

Previous studies demonstrated that prophylactically feeding a probiotic such as PoultryStar that contains Enterococcus faecium, Bifidobacterium animalis, Pediococcus acidilactici, and Lactobacillus reuteri consistently reduced the incidence of BCO in broilers reared on wire flooring (Wideman et al., 2012). Wire flooring provides a reliable experimental model for triggering significantly elevated incidences of BCO in research flocks. The footing instability created by wire flooring appears to amplify the mechanical stresses exerted on the growth plates and accelerates the formation of osteochondrotic microfractures and clefts in the proximal epiphyseal-physeal cartilage of the femora and tibiae. Wire flooring (or the lack of access to litter) also triggers physiological stress that can lead to immunosuppression and enhanced bacterial translocation from the gastrointestinal tract into the systemic circulation (El-Lethey et al., 2003; Wideman and Pevzner, 2012; Wideman et al., 2012, 2013, 2014; Wideman and Prisby, 2013; Gilley et al., 2014). Probiotics were thought to reduce the incidence of BCO in broilers reared on wire flooring by improving intestinal epithelial barrier function and thereby at least partially delaying or attenuating the process of bacterial translocation (Wideman et al., 2012; Wideman and Prisby, 2013).

For the present study, experiment 1 was conducted to determine if adding BacPack 2X to the feed prophylactically (prior to the onset of lameness) would reduce the incidence of BCO in broilers reared on wire flooring. BacPack 2X is a combination of the MOS beta-glucan yeast cell wall prebiotic IMW50 plus the DFM probiotic Calsporin Bacillus subtilis C-3102. Wire flooring imposes a rigorous challenge that leads to high incidences of BCO that can be difficult to suppress. Accordingly, experiment 2 was conducted to determine the extent to which administering the potent fluoroquinolone antimicrobial enrofloxacin therapeutically (after the initiation of BCO) via the drinking water could reduce the incidence of BCO in broilers reared on wire flooring.

MATERIALS AND METHODS

Animal procedures were approved by the University of Arkansas Institutional Animal Care and Use Committee (Protocols 11002 and 14005). Environmental chambers (3.7 by 2.5 by 2.5 m) within the Poultry Environmental Research Lab at the University of Arkansas Poultry Research Farm were used to conduct 2 independent experiments. The chambers use single-pass ventilation at a constant rate of 6 m³/min. Each chamber contained a rectangular pen with dimensions of 3 by 1.5 m with flooring consisting of clean wood shavings litter or flat wire panels. The wire panels were constructed from 5 by 5 cm lumber and were 3 by 1.5 m, with 5 by 5 cm cross-members added for support. Hardware cloth (1.3 by 2.54 cm mesh, 0.063 gauge, galvanized welded wire cloth; Direct Metals, Kennesaw, GA) was fastened to the top of the frame and cross-members. The panels were elevated on 30 cm high masonry blocks to permit manure to pass through and accumulate underneath the wire surface. Tube feeders were positioned at the front, and nipple waterers were positioned at the rear of the pen, thereby forcing birds to traverse the length of the floor to eat and drink. Between experiments, the pens were disassembled, and the chambers and wire flooring panels were thoroughly cleaned using a pressure washer and detergent.

Unvaccinated broiler chicks from 2 different commercial hatcheries were placed at 61 (experiment 1; all males) or 69 (experiment 2; straight run) per chamber at 1 d of age. At 14 d of age, the population was reduced to 50 (experiment 1) or 59 (experiment 2) of the largest, healthiest chicks per pen. The early culling protocol was instituted because necropsies of runts and culls during the first 2 wk often reveal macroscopic evidence of systemic bacterial infection, including osteomyelitis and BCO lesions (Wideman et al., 2012). Representative groups of the chicks culled on d 14 in experiment 1 were necropsied to assess the prevalence of early lesions associated with BCO. The photoperiod was set for 23 h light to 1 h dark (23L:1D) throughout the experiment. Thermoneutral temperatures were maintained throughout: 32°C for d 1 to 3, 31°C for d 4 to 6, 29°C for d 7 to 10, 26°C for d 11 to 14, and 24°C thereafter. Chicks reared on wire flooring are exposed to circulating air on their ventral surfaces and therefore require somewhat elevated brooding temperatures. Feed (crumbles through wk 3 and pellets thereafter) and water were provided ad libitum. The control diets were corn and soybean meal–based broiler starter (crumbles) and finisher (pellets) feeds formulated to meet minimum NRC (1994) standards for all ingredients. In experiment 1, the experimental diet consisted of control feed mixed prior to pelleting with 1 lb/ton of Quality Technology International Incorporated’s BacPack 2X. Feed specifications were determined by QTI’s technical staff, and the feed was manufactured by the University of Arkansas Poultry Research Feed Mill. Control diets were used in all chambers throughout experiment 2.

Beginning on d 15. all birds were observed daily to detect the onset of lameness. Affected broilers had difficulty standing, exhibited an obvious limping gait while dipping one or both wing tips, and, if not removed, became completely immobilized within 48 h. Birds were humanely euthanized as soon as the onset of lameness was noticed and were necropsied within 20 min postmortem. Lame birds with BCO can die quickly because they have difficulty accessing food and water, and they can be trampled by their flockmates. Therefore, birds found dead also were necropsied to ascertain the cause of death and assess leg lesions. Body weights were not recorded during the course of experiment to avoid imposing additional stress on birds that already were under significant stress due to the wire flooring (Wideman and Pevzner, 2012; Wideman and Prisby, 2013).

Experiment 1

Environmental chambers 1 through 8 were used for experiment 1. Pens in chambers 1 and 2 had clean wood shavings litter flooring, and pens in chambers 3 through 8 had flat wire flooring. Beginning on d 1 and continuing through d 56, the control feed was provided in chambers 1, 3, 5, and 7, and the experimental feed containing BacPack 2X was provided in chambers 2, 4, 6, and 8. On d 56, representative survivors that appeared to be clinically healthy were euthanized, weighed, and necropsied to assess subclinical lesion incidences (n = 20 each for the litter flooring chambers; 10 each for the wire flooring chambers).

Experiment 2

Environmental chambers 1 through 8 were used for experiment 2. Pens in all chambers had flat wire flooring. The nipple waterers in chambers 1, 3, 5, and 7 were supplied with tap water (Fayetteville, AR, municipal water) throughout the experiment. The nipple waterers in chambers 2, 4, 6 and 8 were supplied from 20 L water carboys suspended at a height of 2 m to create a hydrostatic pressure head sufficient to operate the pressure regulators for the nipple water lines. The birds in all chambers received tap water through d 34. On d 35 and continuing through d 54, enrofloxacin was added to the carboys. Enrofloxacin (MP Biomedicals, LLC, Santa Ana, CA, 92707, 199019; CAS 93106-60- 6) was dissolved at 60 mg/mL in 50% ETOH acidified with concentrated acetic acid (Seedher and Agarwal, 2009). One mL of this stock solution was added per L of water in the carboy (0.06 mg/mL) to provide each bird in the even-numbered chambers an estimated 10 mg enrofloxacin/kg BW/d based on typical water consumption rates (per Bayer HealthCare, Leverkusen, Germany, AN:00740/2007 recommendations for Baytril 10% Oral Solution). Tap water was restored to the even-numbered chambers on d 55 through 62. On d 62, representative clinically healthy survivors were euthanized and necropsied to assess subclinical lesion incidences (n = 30 per chamber). Survivors were not weighed in experiment 2.

Necropsy Procedures

All birds that died or developed clinical lameness were recorded by date and pen number, and then they were necropsied and assigned to one of the following categories: Cull (runts and moribund individuals that failed to thrive); UNK (unknown cause of death); SDS (sudden death syndrome, flipover, heart attacks); PHS (pulmonary hypertension syndrome, ascites); KB (kinky back, spondylolisthesis, or vertebral BCO; diagnosed based on the characteristic posterior paraparesis and hock-resting posture, and the absence of macroscopic severe BCO lesions of the femora and tibiae (Martin et al., 2011); TW (twisted leg or slipped tendon, perosis, chondrodystrophy); Normal F (no macroscopic abnormalities of the proximal femur); FHS (proximal femoral head separation or epiphyseolysis); FHT (proximal femoral head transitional degeneration); FHN (proximal femoral head necrosis); Normal T (no macroscopic abnormalities of the proximal tibia); THN (mild proximal tibial head necrosis, a subcategory of BCO in the tibiotarsus); THNs (“severe” THN in which the growth plate was imminently threatened or damaged); THNc (“caseous” THN in which caseous exudates or bacterial sequestrae were macroscopically evident); TD (tibial dyschondroplasia); and, Lame-UNK (lameness for unknown/undetermined reasons). Previously published photographs illustrate typical BCO lesions of the proximal femora and tibiae (Wideman et al., 2012, 2014; Wideman and Prisby, 2013). Proximal femoral head lesions (FHS, FHT, FHN) and tibial head lesions (THN, THNs, THNc) were categorized separately to emphasize the progressive development of BCO in the proximal ends of both long bones of the legs (Mutalib et al., 1983; Thorp and Waddington, 1997; McNamee et al., 1998, 1999; Butterworth, 1999; McNamee and Smyth, 2000; Dinev, 2009; Durairaj et al., 2009; Wideman et al., 2012, 2013; Wideman and Pevzner, 2012). Proximal femora and tibiae that appeared to be normal macroscopically were not routinely evaluated microscopically.

Statistical Analysis

The total incidence of femoral BCO lesions was calculated as: All femur = FHS + FHT + FHN. The total incidence of tibial BCO lesions was calculated as: All Tibia = THN + THNs + THNc. The total incidence of lameness was calculated as: Total Lame = KB + TW + TD + Lame UNK + All Femur + All Tibia. For comparisons of lameness incidences, the individual bird was used as the experimental unit, and the SigmaStat Z-test procedure was used to compare proportions (Jandel Scientific, 1994). For comparisons of lesion incidences, the number of legs evaluated was used as the sample size, and the percentage of affected legs was used as the proportion for all Z-tests. The SigmaStat ANOVA package was used to compare BW among diet treatments and flooring types in experiment 1.

RESULTS

Experiment 1

Chick mortality plus runts culled during the first 2 wk totaled 2 in each of the litter flooring chambers 1 and 2. There were 10 in chambers 3, 5, and 7 combined (control group), and 5 in chambers 4, 6, and 8 combined (BacPack 2X group). Several of the chicks that were culled to reduce bird densities on d 14 exhibited severe BCO lesions of the proximal femora (FHN) and tibiae (THNs) (Table 1). These cull chicks tended to be smaller than their hatchmates, but none of the culls exhibited overt symptoms of lameness. Between d 15 and 56, there were 4 birds in chamber 1 and 1 in chamber 2 that died from SDS, whereas as 1 bird in chamber 1 and 2 in chamber 2 developed BCO lameness (2 and 4% incidences, respectively). Necropsy results for birds that died or became clinically lame on wire flooring between d 15 and 56 are shown in Table 2. The incidences were calculated as percentages of 150 birds per diet treatment on wire flooring after culling on d 14. PHS and SDS were the primary causes of nonlame mortality, and their incidences did not differ between diet treatments. No mortality was attributed to unknown causes. Lameness due to KB, TW, TD, and LAME-UNK were minimal, and these non-BCO causes of lameness did not differ between diet treatments. Lameness overwhelmingly was attributable to BCO lesions of the proximal femora and tibiae. Incidences of BCO lameness and total lameness were significantly lower (P = 0.003) in the BacPack 2X group than in the control group (Table 2).

The time course for cumulative BCO lameness on wire flooring is shown in Figure 1. Broilers in the control group had higher incidences of BCO on d 36 through 43 (P = 0.036), and on d 46 through 56 (P = 0.001) when compared contemporaneously with broilers in the BacPack 2X group. The groups did not differ on d 44 and 45 (ns, P = 0.090). The onset of BCO occurred earlier in broilers in the control group, and the overwhelming majority of BCO lameness developing during wk 7 and 8 in both groups (Figure 1). Variation in total BCO incidences among the individual environmental chambers is illustrated in Figure 2. The lowest incidence among the wire flooring chambers assigned to the control group (36%, CW chamber 3) numerically exceeded the highest incidence for the wire flooring chambers assigned to the BacPack 2X group (32%, BPW chamber 4). This range of pen-to-pen variability is typical based on extensive prior experience with the wire flooring model. Lesion incidences for the proximal femora and proximal tibiae from broilers that developed BCO lameness on wire flooring are shown in Table 3. The groups did not differ within any of the lesion categories. Very few proximal tibiae remained macroscopically “normal” in appearance in both groups, and both groups exhibited similarly elevated incidences of severe caseous tibial head necrosis (Table 3).

Prophylactic administration of a combined prebiotic and probiotic, or therapeutic administration of enrofloxacin, to reduce the incidence of bacterial chondronecrosis with osteomyelitis in broilers - Image 1

Lesion incidences for the proximal femora and proximal tibiae from clinically healthy survivors reared on litter or wire flooring are shown in Table 4. Within each flooring category, broilers fed the control and BacPack diets differed minimally, with the exception that survivors on litter flooring had higher incidences of normal proximal femora and lower incidences of FHT when they had been reared on the control diet. Independent of diet treatment, broilers that survived 8 wk on wire flooring had significantly higher incidences of proximal tibial BCO lesions (THN + THNs + THNc) than the survivors reared on litter (Table 4). Body weights of the clinically healthy survivors are shown in Table 5. Broilers reared on litter and fed the control diet were the heaviest, whereas broilers reared on wire and fed the control diet were the lightest. Clinically healthy survivors that had been reared on wire flooring for 56 d had higher BW when they had been fed the BacPack 2X diet instead of the control diet (4.03 ± 0.05 vs. 3.81 ± 0.07 kg, respectively; P = 0.012). Pooling these data by diet treatment regardless of floor type eliminated the differences in 8 wk BW (3.99 ± 0.06 for control vs. 4.02 ± 0.04 kg for BacPack 2X; P = 0.645).

Experiment 2

Chick mortality plus runts culled during the first 2 wk totaled 7 (2.5%) in chambers 1, 3, 5, and 7 combined and 8 (2.9%) in chambers 2, 4, 6, and 8 combined. Necropsy results for birds that died or became clinically lame on wire flooring between d 15 and 62 are shown in Table 6. The incidences were calculated as percentages of 59 birds per chamber (236 per treatment group) after culling on d 14. The primary causes of nonlame mortality were PHS and SDS, and their incidences did not differ between groups. No mortality was attributed to unknown causes. Lameness due to KB, TW, TD, and LAME-UNK were minimal, and these non-BCO causes of lameness did not differ between groups. Lameness overwhelmingly was attributable to BCO lesions of the proximal femora and tibiae. Incidences of BCO lameness and total lameness were significantly lower in broilers in the enrofloxacin group than in those in the tap water group (P = 0.003) (Table 6).

The time course for cumulative BCO lameness is shown in Figure 3. Broilers receiving tap water throughout had higher incidences of BCO on d 51 through 62 when compared contemporaneously with broilers in the enrofloxacin group (P = 0.001). During the 11 d after the start of enrofloxacin administration, the cumulative BCO incidence increased by 0.4% in the enrofloxacin group and by 4.2% in the tap water group (P = 0.025).

The overwhelming majority of BCO lameness developed during wk 7 through 9 in both groups (Figure 3). As shown in Figure 4, by 34 d of age, the incidence of BCO was 1.3% in the odd-numbered chambers (tap water group) and 3.0% in the even-numbered chambers (presumptive enrofloxacin group). Enrofloxacin was added to the drinking water in the even-numbered chambers beginning on d 35. Over the entire interval of enrofloxacin administration (d 35 to 54), twice as many birds developed lameness attributable to BCO in the tap water group when compared with the enrofloxacin group (19.5 vs. 8.1%, respectively; P = 0.001). Subsequent to enrofloxacin withdrawal on d 54, similar numbers of birds developed lameness in both groups (17.2 tap water vs. 13.8% enrofloxacin, d 55 through 62) (Figure 4). Total BCO lameness over the d 14 through d 62 course of the experiment was higher in the tap water group than in the enrofloxacin group (36.9 vs. 24.2%, respectively; P = 0.003) (Figures 3 and 4). Variation in total BCO incidences among the individual environmental chambers is illustrated in Figure 5. The lowest incidence among the wire flooring chambers receiving only tap water (30.5%, TW chamber 1) was not different from the highest incidence for the chambers receiving enrofloxacin (32.2%, enrofloxacin chamber 8; P = 1.0). This range of pen-to-pen variability is typical for the wire flooring model. Lesion incidences for the proximal femora and proximal tibiae from broilers that developed BCO lameness on wire flooring are shown in Table 7. During the necropsies for experiment 2, the proximal tibiae were categorized as being macroscopically normal or as exhibiting tibial head necrosis, without discriminating between THN, THNs, or THNc. Broilers that became lame in the enrofloxacin group had more normal femora, fewer FHT lesions, fewer total femoral lesions, and no differences in normal tibiae or tibiae with BCO lesions when compared with broilers in the tap water group (Table 7). Lesion incidences for the proximal femora and proximal tibiae from clinically healthy survivors are shown in Table 8. There were no differences between the groups within any of the lesion categories.

DISCUSSION

Wire flooring continues to provide a reliable experimental challenge for triggering high incidences of lameness caused by BCO. Wire flooring is believed to create footing instability that amplifies mechanical stresses exerted on the structurally immature and poorly aligned columns of chondrocytes in the proximal growth plates (epiphysis and metaphysis) of the femora and tibiae in fast-growing broilers. The resulting osteochondrotic microfractures and exposed collagenous matrix are conducive for bacterial colonization. Hematogenously distributed bacteria gain access to these wound sites by exiting the bloodstream through fenestrations (windows or gaps) in the capillaries supplying the epiphysis and metaphysis. Accordingly, osteochondrosis per se is not the direct cause of clinical lameness; instead, it is the subsequent infection by opportunistic pathogenic bacteria that triggers the formation of grossly destructive BCO lesions (Thorp et al., 1993; McNamee and Smyth, 2000; Smeltzer and Gillaspy, 2000; Wideman and Prisby, 2013). This hypothesis for the pathogenesis of BCO led to the prediction that clinical lameness should be minimized if bacterial translocation into the blood stream could be reduced to levels that do not overwhelm the immune system. Indeed, several recent studies demonstrated that prebiotics and probiotics can improve intestinal barrier function and counteract stress-mediated increases in intestinal permeability in broilers (Sohail et al., 2010, 2012; Murugesan et al., 2014; Song et al., 2014). Probiotics clearly are capable of reducing the incidence of BCO in broilers reared on wire flooring (Figures 1 and 2; Wideman et al., 2012).

In the present study, the objective of experiment 1 was to determine if prophylactically providing BacPack 2X would reduce the incidence of BCO. Prophylactic probiotic administration beginning at 1 d of age previously was found to be effective, whereas therapeutic administration beginning on d 28 was not, presumably because bacterial translocation and osteomyelitis can be initiated in the hatchery or immediately posthatch (Wideman et al., 2012). With regard to the possibility that bacterial translocation can occur very early pre- or posthatch, it is noteworthy that in experiment 1, twice as many chicks were culled for failing to thrive in the control group when compared with the prophylactically treated BacPack 2X group, and subclinical BCO lesions were detected in 2-wk-old chicks culled from both groups. Furthermore, clinically lame broilers with classic BCO lesions began to accumulate in the control group during wk 4 and 5, whereas prophylactic administration of BacPack 2X delayed the onset of BCO until the wk 6. Hatchery management and sanitation previously have been implicated as factors that potentially can influence the incidence of BCO (Wideman et al., 2013). The cumulative incidence of BCO lameness was significantly lower in the BacPack 2X group than in the control group beginning on d 35 and essentially continuing through d 56. The efficacy of BacPack 2X cannot be attributed to reduced BW gain because in pens with wire flooring, the survivors in the BacPack 2X group were heavier than the survivors in the control group. Nor can the efficacy of BacPack 2X be attributed to a differential reduction in the severity of femoral and tibial BCO lesion incidences. The BCO lesion distributions for broilers reared on wire flooring did not differ between diet treatments, either for birds that developed BCO lameness or for survivors on d 56. Accordingly, although the specific biological mechanism remains to be determined, the significant response in experiment 1 to prophylactic administration of BacPack 2X provides additional evidence that prebiotics and probiotics can significantly interrupt the pathogenesis of lameness attributable to BCO.

Wire flooring imposes a rigorous challenge that can be difficult for fast-growing broilers to survive regardless of the efficacies of prophylactic or therapeutic treatment regimens. In addition to the mechanical stress exerted on the proximal femoral and tibial growth plates, wire flooring (or the lack of access to litter) also induces physiological stress that facilitates bacterial translocation and suppresses the immune response. The objective of experiment 2 was to determine the extent to which administering enrofloxacin therapeutically (after the initiation of BCO) could reduce the incidence of BCO lameness in broilers reared on wire flooring. Enrofloxacin is a potent fluoroquinolone antimicrobial that is used routinely outside of the United States to treat bacterial or mycoplasmal diseases of the respiratory and alimentary tracts in poultry. Enrofloxacin can be administered via the drinking water at therapeutic levels (10 mg/kg BW/d) that presumably should acutely reduce or eliminate bacterial infections. Based on our hypothesis for the pathogenesis of BCO, lameness incidences should be reduced or even eliminated if therapeutic levels of enrofloxacin are capable of attenuating the bacteremia and femoral and tibial abscess development in broilers reared on wire flooring. Enrofloxacin was added to the drinking water on d 35 after the cumulative incidence of BCO had reached an average of 3% in the even-numbered environmental chambers. Thereafter, during the ensuing d 35 through 45 interval, the cumulative incidence of lameness in the enrofloxacin chambers increased by only 0.4%, indicating the therapeutic antibiotic administration had transiently halted the further development of clinical BCO. However, after d 45, the efficacy of ongoing enrofloxacin administration appeared to diminish, and by d 54, the cumulative BCO incidence in the enrofloxacin group was approximately half the incidence in the tap water control group. As expected, BCO incidences accelerated rapidly when tap water was restored to the even-numbered (enrofloxacin) chambers after d 54 (Figure 3). These time-course trends suggest enrofloxacin began to lose efficacy after approximately 2 wk of therapeutic treatment. This apparent diminution of efficacy may reflect depletion of the commensal (beneficial) microbial population combined with an overgrowth of antibiotic-resistant pathogenic species. It also is probable that enrofloxacin was unable to reach areas of active infection in the more secluded osteochondrotic microfractures, particularly after protective biofilms had surrounded the bacterial abscesses (Wideman et al., 2012). The collateral necrosis in BCO lesions also destroys the epiphyseal and metaphyseal vasculature that otherwise would carry antibiotics and responding immune cells to foci of infection (Wideman and Prisby, 2013). The necropsy observations for clinically lame birds suggest enrofloxacin did tend to reduce the incidence of severe proximal femoral lesions (Table 7). Accordingly, enrofloxacin may help prevent or attenuate the progressive deterioration of early femoral lesions into the more severe degenerative lesions that cause terminal lameness in broilers.

In summary, wire flooring provides an experimental challenge that reliably induces high incidences of BCO. Consistently elevated incidences of BCO in experimental flocks are conducive to discovering the basis for innate susceptibility as well for evaluating potential strategies for preventing or treating commercial outbreaks. The mechanical and physiological stresses that are imposed chronically by wire flooring can be very difficult for fast-growing broilers to survive, regardless of the efficacies of prophylactic or therapeutic treatment regimens. The consensus hypothesis for the pathogenesis of BCO led to our predictions that clinical lameness should be minimized if probiotics can successfully attenuate bacterial translocation and if antibiotics can acutely retard or eliminate bacterial infections. The results of the present and previous studies are summarized in Figure 6, which illustrates the cumulative BCO incidences in control and treated groups of broilers reared on wire flooring in 4 independent experiments conducted within the Poultry Environmental Research Lab. This meta-analysis indicates that 2 out of 3 prophylactically administered probiotics successfully reduced the incidence of BCO lameness by approximately half, and this magnitude of response was similar to the reduction achieved therapeutically with the potent antibiotic enrofloxacin. The biological mechanisms involved in these responses clearly differ substantially, and the experiments were conducted independently rather than contemporaneously. It also is clear that placing wire flooring pens in environmental chambers creates an ideal experimental paradigm for revealing beneficial responses to probiotics, including excellent air quality (respiratory challenges are minimized) and negligible contact with fecal material (necrotic enteritis and coccidiosis are eliminated) in combination with continuous probiotic reinoculation via the feed. Under these circumstances, the ability of prebiotics and probiotics to delay the onset and significantly reduce the cumulative incidences of BCO strongly implicate the gastrointestinal epithelium as the primary portal for bacterial translocation, rather than the respiratory tract or integument. The data summarized in Figure 6 demonstrate that probiotics and antibiotics do partially surmount the severe, chronic mechanical and physiological stresses imposed by rearing broilers on wire flooring. The further implication is that probiotics have the potential to provide effective alternatives to antibiotics for reducing BCO lameness attributable to bacterial translocation and hematogenous distribution.

This article was originally published in 2015 Poultry Science 94:25–36. http://dx.doi.org/10.3382/ps/peu025. This is an Open Access article distributed under a Creative Commons Attribution License.