Skip to main content
PMC home page
Journal of Animal Science logoLink to Journal of Animal Science
. 2021 Jul 1;99(7):skab157. doi: 10.1093/jas/skab157

Efficacy of phage therapy in pigs: systematic review and meta-analysis

Karina Desiree 1,#, Sabrina Mosimann 1,#, Paul Ebner 1,✉
PMCID: PMC8280926  PMID: 34196704

Abstract

Limits on the use and efficacy of various antibiotics coupled with negative consumer perception of the practice have together spurred substantial research into compounds that could reduce the use antibiotics to control bacterial diseases in pigs. Bacteriophages are often among such potential compounds, and various groups have examined the efficacy of bacteriophages or bacteriophage products in limiting transmission or colonization of targeted bacteria. The study presented here provides a systematic review of such studies followed by a meta-analysis of aggregated data produced by each study. The data set was limited to inputs (n = 19; 576 total observations) from studies where: 1) live pigs were inoculated with a known quantity of challenge bacteria; 2) challenged animals were treated with a known quantity of phages; 3) concentrations of the challenge bacteria were measured in different tissues/fluids following phage treatment; and 4) SD (or SE to allow calculation of SD) was reported. Concentrations of challenge bacteria were significantly lower in phage-treated pigs versus challenged but untreated pigs (P < 0.0001; effect size = −1.06 1log10 colony-forming units [CFU]/g). The effect size of phage treatment was significantly greater (P < 0.05) in samples collected 48 to 96 h following phage treatment versus those collected ≤ 24 h following phage treatment. Likewise, effect size of phage treatment was significantly greater in piglets versus market-weight pigs. Across observations, phage treatment effect sizes were greatest (P < 0.01) in fecal samples versus ileal or cecal samples. Taken together, these data indicate that phage treatment can significantly reduce the concentrations of targeted bacteria in pigs; scenarios exist, however, where phage treatment could predictably be more or less effective.

Keywords: meta-analysis, phage, systematic review, swine

Introduction

The control of bacterial diseases in both veterinary and human medicine is central to the One Health concept (CDC, 2021). Effective antibiotics have been and continue to be a cornerstone of bacterial disease prevention and management programs. By most metrics, however, the development of antibiotic resistance outpaces the discovery and availability of new antibiotic drugs (Årdal et al., 2020). At the same time, numerous countries including the United States, China, and EU member states, among many others, have implemented policies that effectively curtail the use of many antibiotics in livestock for performance (e.g., improved growth efficiency; European Commission, 2018; Hu and Cowling, 2020; US FDA, 2021). The prospect of limitations on both efficacy and availability of existing antibiotics for treating or preventing emerging bacterial pathogens has spurred research into the antibacterial properties of nonantibiotic compounds for the treatment or prevention of bacterial infections (Ghosh et al., 2019).

One potential alternative to antibiotics is bacteriophages. Bacteriophages, or simply “phages,” are viruses specific to bacteria. The traditional use of bacteriophages as antibacterials, termed “phage therapy,” takes advantage of the infectious cycle of lytic phages which results in lysis or death of the bacterial host. Recent phage research, however, has also investigated the therapeutic potential of isolated phage lysins (Fischetti, 2018; Vázquez et al., 2018) and lysogenic phages or temperate phages where infection leads to integration of phage DNA into the bacterial chromosome rather than immediate lysis (Yosef et al., 2015; Park et al., 2017; Monteiro et al., 2019). The use of bacteriophages as antibacterials predates the widespread introduction of antibiotics in both veterinary and human medicine (Sulakvelidze et al., 2001; Kutter et al., 2010), but the practice has regained attention in recent decades as a possible nonantibiotic means of bacterial disease control.

There is a growing body of research examining the efficacy of using phages as antibacterials specifically in food animal production. The majority of this research has involved monogastric animals, and this manuscript presents a systematic review of studies focused on the application of phage treatments to swine. The systematic review is accompanied by a meta-analysis aimed at aggregating results across studies to determine whether administering bacteriophages reduces concentrations of challenge bacteria in swine. The data used for meta-analysis were then disaggregated in an attempt to identify scenarios (e.g., phages used for prevention vs. treatment) or factors (e.g., effect of target pathogen, age of pigs, or sampling time, etc.) that may influence the impact of phage treatment on targeted bacterial concentrations.

Methods

A primary literature search for peer-reviewed published research assessing the efficacy of bacteriophage application in reducing or preventing bacterial colonization in pigs was conducted in the summer of 2020. The search utilized PubMed and Google Scholar databases and was restricted to research published between 1990 and present and identified using search terms including “swine + phage,” “pig + phage,” “swine + bacteriophage,” and “pig + bacteriophage.” The resulting articles were screened to remove review articles, non-peer-reviewed articles (e.g., theses/dissertations), other gray literature, articles utilizing species other than swine, and articles using ex vivo, in vitro, or in vivo models. This initial screen produced a pool of 25 articles describing experiments in which phages were delivered to live pigs to assess the antibacterial capacity of the phage treatment. References cited in remaining articles were screened to determine if other articles existed that fit selection criteria.

Descriptive statistics including the study author, year published, a brief results summary, study location, targeted bacteria, a description of experimental animal characteristics, and whether or not an animal trial was performed were recorded for each study. A qualitative assessment of each study was then conducted using the self-developed rubric described in Supplementary Table 1. Based on a review of this qualitative assessment, it was hypothesized that further meta-analysis with acceptable heterogeneity across studies could be conducted using studies that: 1) challenged live pigs with a known quantity of challenge bacteria (e.g., 108 colony-forming units [CFU] of Salmonella enterica Typhimurium per pig); 2) treated challenged animals with a known quantity of phages (e.g., 108 plaque-forming units [PFU] of Salmonella Phage FFH1); 3) measured concentrations of the challenge bacteria in different tissues/fluids following phage treatment; and 4) reported either SD or SE to allow calculation of SD. Thus, studies involving natural challenges or challenges with unknown quantities of bacteria, studies reporting only qualitative or semi-qualitative data (e.g., frequency of bacterial shedding), studies reporting total bacteria concentrations post-treatment but not challenge bacteria concentrations post-treatment specifically, or studies not reporting SD (or SE) were excluded. As a result, our final data set included 19 inputs (i.e., 19 distinct phage treatments/experiments) across five published studies with concentration of challenge bacteria following phage treatment as the principal outcome variable.

Statistical methods and packages (dmetar; Harrer et al., 2019a) were used following previously published guidelines (Harrer et al., 2019b) and processed using RStudio (RStudio Team, 2020). To construct the main data set for meta-analysis, the following data were extracted from each input (n = 19): 1) number of animals in treatment group; 2) number of animals in control group (i.e., animals receiving the challenge bacteria but not phage or any other antibacterial treatment); 3) mean concentration of the challenge bacteria in phage-treated animals with SD; and 4) mean concentration of challenge bacteria in control animals with SD. When concentration means, SD, or SE values were not included in the text, these values were estimated based on graphical data. These estimations are noted as such in the text of the systematic review below when they occurred. When multiple, independent experiments or treatments were reported within single research articles, each independent experiment or treatment was included as a separate input unless treatments only differed in terms of the titer of phage inocula, in which case input means were pooled. When a single experiment contained one control group but multiple phage treatment groups or inputs, the number of animals in the control group was divided equally across the number of treatment groups to avoid over-representation of individual studies in the data set. As a result, our data set included 360 treatment observations (challenge bacteria concentrations in challenged pigs treated with phages lytic against the challenge bacteria) and 240 control observations (challenge bacteria concentrations in challenged pigs not receiving phage or other antibacterial treatment).

Data were analyzed using the dmetar package of R (Harrer et al., 2019a) and the full data set was first screened for the presence of publication bias and “P-hacking”. Subsequently, data were analyzed as a random-effects model to account for variation between studies (i.e., differences in breed and age of pigs, differences in study methods and operators, etc.; Harrer et al., 2019b). Variance of distribution of effect size was estimated using Sidik-Jonkman and the random-effects model was adjusted using the Hartung-Knapp method (Harrer et al., 2019b). Experimental designs of those studies producing observations identified in R as outliers were re-examined for evidence justifying removal of the observations. Data were also disaggregated by sampling time post-phage treatment, sample site (ileum, cecum, feces, or rectum), delivery or application method (e.g., gavage vs. feed), delivery time (e.g., phage given before or after bacterial challenge), age of pigs, type of treatment (e.g., single vs. poly-phage), species of challenge bacteria, frequency of phage application, and phage preparation (e.g., microencapsulation or other form of protection). Heterogeneity across studies was measured using the inconsistency index, or percentage of heterogeneity (I2; Higgins and Thompson, 2002). Heterogeneity was considered low when I2 values were between 0% and 25%, moderate when I2 values were between 26% and 50%, considerable when I2 values were between 51% and 75%, and significant when I2 values were between 76% and 100%. As the identification of subgroups for data disaggregation was not random, comparisons between subgroups were made using a mixed-effects model that utilized a random-effects model within subgroups and a fixed-effects model between subgroups (Harrer et al., 2019a). Post-hoc power statistics (%) were calculated for the overall data set and for each disaggregated data set. Differences were considered statistically significant at P < 0.05.

Results and Discussion

Systematic review

Screening of the articles retrieved in the original primary literature search resulted in the identification of five articles for systematic review (Table 1) that: 1) challenged live animals with a known quantity of challenge bacteria (e.g., 108 CFU of Salmonella enterica Typhimurium per pig); 2) treated challenged animals with a known quantity of phages (e.g., 108 PFU of Salmonella Phage FFH1 per pig); 3) measured concentrations of the challenge bacteria in different tissues/fluids following phage treatment; and 4) reported a SD or SE. It is of note that the majority of studies included in this review measured other effects in addition to the principal outcome variable, such as performance (e.g., average daily gain, feed intake, etc.) and pathology (e.g., temperature, fecal scores, etc.) among others; however, those effects were outside the scope of this review and not discussed.

Table 1.

Descriptions of studies measuring challenge bacteria concentrations following application of phage in pigs included in systematic review and meta-analysis

Study code Reference Phage description Challenge organism Phage preparation Pig characteristics Delivery method Delivery schedule Sample sites Number of inputs Total number of observations
Swine_4 Albino et al., 2014 Six phages isolated with Salmonella enterica Typhimurium Salmonella enterica Typhimurium Suspended in SM buffer 1 90–100 kg Oral Phages delivered after bacterial challenge I2, C3, F4 4 90
Swine_14 Wall et al., 2010 15 phages isolated with Salmonella enterica Typhimurium Salmonella enterica Typhimurium Microencapsulated 15 kg (Exp. 1); 90 kg (Exp. 2) Oral gavage Phages delivered before bacterial challenge I, C, F 4 120
Swine_15 Callaway et al., 2011 Two phages isolated with Salmonella enterica Typhimurium Salmonella enterica Typhimurium Suspended in phosphate-buffered saline (PBS)5 10 kg Oral gavage Phages delivered after bacterial challenge C, F 4 192
Swine_16 Saez et al., 2011 15 phages isolated with Salmonella enterica Typhimurium Salmonella enterica Typhimurium Microencapsulated 15 kg Oral gavage (Exp. 1); In-feed (Exp. 2) Phages delivered before bacterial challenge I, C, F 4 126
Swine_22 Han et al., 2016 Multiple (number not stated) phages isolated with E. coli K88 or E. coli K99 E. coli K88 (Exp. 1) or E. coli K99 (Exp. 2) Suspended in SM buffer and freeze-dried 0.2 M Tris buffer (pH 7.5) containing 0.1 M NaCl, 1 mM MgSO4, and 0.01% gelatin, and freeze-dried ~28 d of age In-feed Phages delivered after bacterial challenge F 4 48
Open in a new tab

1SM buffer, 0.2 M Tris buffer (pH 7.5) containing 0.1 M NaCl, 1 mM MgSO4, and 0.01% gelatin; 2I, ileal contents; 3C, cecal contents; 4F, feces (or rectal contents); 5PBS, 0.137M NaCl, 2.7mM KCl, 0.01M Na2HPO4, 1.8mM KH2PO4.

Saez et al. (2011) administered a microencapsulated phage cocktail by gavage to young pigs challenged with Salmonella Typhimurium at 2, 4, and 6 h post-challenge. No significant differences (P > 0.05) in the incidence of the challenge organism in fecal samples were observed between phage-treated and untreated pigs at any sampling point (2, 4, and 6 h post-challenge). At necropsy (6 h post-challenge), cecal concentrations of Salmonella Typhimurium were not significantly different (P > 0.05) in phage-treated versus untreated pigs. However, ileal concentrations of Salmonella Typhimurium were significantly lower (P < 0.05) in phage-treated (1.0 log10 CFU/mL) versus untreated pigs (3.0 log10 CFU/mL). In addition, ileal concentrations of Salmonella Typhimurium were below the detectable limit (< 102 CFU/mL) in samples from phage-treated pigs (undetectable in 71.4% of ileal samples) significantly more often (P < 0.05) than in samples from untreated pigs (undetectable in 19.0% of ileal samples).

In the same paper, the authors administered the same phage cocktail via feed to pigs using a similar challenge and sampling model. In this experiment, phage-treated pigs had a significantly lower (P < 0.05) incidence of Salmonella Typhimurium in fecal samples at 2 (38.1% positive) and 4 h (42.9% positive) post-challenge compared to untreated pigs (2 h: 71.4% positive; 4 h: 85.7% positive). At 6 h post-challenge, all fecal samples were Salmonella Typhimurium-positive regardless of treatment. At necropsy, a numerical difference (P < 0.10) was observed in the cecal concentrations of Salmonella Typhimurium in phage-treated (2.7 log10 CFU/mL) versus untreated pigs (3.7 log10 CFU/mL). As in the previous experiment, ileal concentrations of challenge bacteria at necropsy were significantly lower (P < 0.05) in phage-treated (2.0 log10 CFU/mL) versus untreated pigs (3.0 log10 CFU/mL). It is not clear why ileal concentrations of the challenge bacteria significantly differed between phage-treated and untreated pigs in both trials while cecal concentrations did not. Average Salmonella Typhimurium concentrations were similar across cecal and ileal samples from untreated pigs, suggesting that phage-bacterium interactions and bacterial lysis were possible in both organs. As before, Salmonella Typhimurium concentrations in ileal samples from phage-treated pigs were more frequently (P < 0.05) below detectable levels (undetectable in 42.8% of ileal samples) than in samples from untreated pigs (undetectable in 19.0% of ileal samples). The authors noted that phage concentrations in ileal and cecal contents of pigs given phages via feed were 2 to 3 log10 PFU/mL higher at necropsy than those in ilea and ceca of pigs given phages via oral gavage, suggesting better rates of phage survival when phages were administered in feed.

Employing similar experimental methods, Wall et al. (2010) conducted two experiments in which a microencapsulated phage cocktail was administered to Salmonella Typhimurium-challenged young pigs. At necropsy (6 h post-challenge), phage-treated pigs had numerically lower (not statistically compared) concentrations of Salmonella Typhimurium in ileal (0.6 log10 CFU/mL), cecal (0.4 log10 CFU/mL), and cecal tonsil samples (0.4 log10 CFU/mL) versus untreated pigs (ileal: 2.6 log10 CFU/mL, cecal: 3.6 log10 CFU/mL; cecal tonsil: 3.6 log10 CFU/mL). Four of six lymph node samples from untreated pigs and three of six lymph nodes samples from phage-treated pigs were positive for the challenge organism; all fecal samples were positive for the challenge organism by 6 h post-challenge. Among phage-treated pigs, five of six ileal samples, five of six cecal samples, and five of six cecal tonsil samples had levels of Salmonella Typhimurium below the detectable limit (< 102 CFU/mL). In contrast, only three of five ileal samples, two of five cecal samples, and one of five cecal tonsil samples from untreated pigs had undetectable levels of Salmonella Typhimurium. These values, however, were not statistically compared.

In the second experiment of the same study, the authors inoculated market weight pigs with Salmonella Typhimurium, subsequently placing the pigs in a pen and allowing manure to accrue for 48 h. After 48 h, unchallenged pigs were administered the microencapsulated phage cocktail and comingled with the challenged pigs (mock treated pigs were included as controls). At necropsy (6 h post-mingling), cecal concentrations of Salmonella Typhimurium were significantly lower (P < 0.05) in phage-treated (1.5 log10 CFU/mL) versus mock-treated pigs (2.9 log10 CFU/mL). There was a statistical trend (P = 0.06) for lower ileal concentrations of Salmonella Typhimurium in phage-treated (1.7 log10 CFU/mL) versus mock-treated pigs (2.7 log10 CFU/mL). As the authors note, differences in concentrations of the challenge organism in phage-treated pigs in the first and second trials may have been due to differences in the challenge model and response to the bacterial challenge and phage treatment between older and younger pigs.

The incidence of Salmonella Typhimurium in lymph node and fecal samples was not significantly different (P > 0.05) between phage-treated and mock-treated pigs at any sampling point. All fecal samples were positive for the challenge organism at 6 h post-mingling regardless of treatment. The authors report that in both studies fecal and lymph node samples were enriched twice due to low initial counts of the challenge organism. For this reason, and because concentrations of the challenge organism were not enumerated for these sample types, the authors suggest that the quantitative data on cecal and ileal concentrations of Salmonella Typhimurium may be more meaningful than fecal and lymph node qualitative data when evaluating phage treatment efficacy.

Albino et al. (2014) administered a phage cocktail (at four different concentrations) suspended in sodium bicarbonate to Salmonella Typhimurium-challenged market-weight pigs (90 to 100 kg). At necropsy (18 ± 2 h post-treatment), concentrations of Salmonella Typhimurium in cecal and ileal samples were not significantly different (P > 0.05) between phage-treated and untreated pigs. The largest numerical differences in Salmonella Typhimurium concentrations between groups were observed between pigs treated with 1010 PFU/pig (ileal contents: ~4.2 log10 CFU/mL; cecal contents: ~5.1 log10 CFU/mL) and untreated pigs (ileal contents: ~5.5 log10 CFU/mL; cecal contents: ~6.3 log10 CFU/mL). Similarly, the incidence of Salmonella Typhimurium in fecal samples collected at necropsy was not significantly different (P > 0.05) in phage-treated versus untreated pigs, though numerical differences in incidence were observed (104 PFU/pig treatment group: 83.3% positive; 106 PFU/pig treatment group: 50% positive; 108 PFU/pig treatment group: 33.3% positive; 1010 PFU/pig treatment group: 33.3% positive; untreated pigs: 100% positive).

Callaway et al. (2011) orally administered a two-phage cocktail to weaned pigs (~10 kg) at 24 and 48 h after a Salmonella Typhimurium challenge. No significant differences (P > 0.05) in fecal concentrations of Salmonella Typhimurium were observed between phage-treated and untreated pigs at any sampling point (24, 48, 72, and 96 h post-challenge) during the experiment, though numerically lower fecal concentrations of the challenge organism were recorded for phage-treated pigs at 48 (phage-treated: ~1.25 log10 CFU/mL; untreated: ~2.25 log10 CFU/mL), 72 (phage-treated: ~0.70 log10 CFU/mL; untreated: ~1.70 log10 CFU/mL), and 96 h (phage-treated: ~0.38 log10 CFU/mL; untreated: ~1.25 log10 CFU/mL) post-challenge. At necropsy, cecal and rectal concentrations of Salmonella Typhimurium did not significantly differ (P > 0.05) between treatment groups, though concentrations were numerically lower in phage-treated pigs (cecal contents: ~1.40 log10 CFU/mL; rectal contents: ~0.38 log10 CFU/mL) versus untreated pigs (cecal contents: ~2.75 log10 CFU/mL; rectal contents: ~1.3 log10 CFU/mL). The incidence of Salmonella Typhimurium was significantly lower (P < 0.05) in rectal samples from phage-treated pigs (25% positive) compared to those of untreated pigs (~83.3% positive). No significant differences (P > 0.05) in the incidence of the challenge organism in ileal or lymph node samples were observed between any groups. To note, all concentrations and percentages stated here are approximations based on the authors’ graphical data.

Han et al. (2016) administered a phage cocktail via feed to 28-d-old pigs challenged with Escherichia coli K88 and K99. Phage treatment began on the day of challenge and continued until the conclusion of the study (7 d post-challenge). Over the course of sampling (1, 3, and 7 d post-challenge), fecal concentrations of E. coli K99 did not significantly differ (P > 0.05) between phage-treated and untreated pigs. However, overall fecal concentrations of E. coli K88 were significantly lower (P < 0.01) in phage-treated versus untreated pigs (quantitative data for overall fecal concentrations not shown). Adhesion of E. coli K88 to the ileum and cecum of phage-treated pigs (ileum: 5.57 ± 0.263 log10 CFU/g; cecum: 3.92 ± 0.800 log10 CFU/g) was significantly lower (P < 0.05) than adhesion to the same tissue types in untreated pigs (ileum: 7.24 ± 0.460 log10 CFU/g; cecum: 6.32 ± 0.504 log10 CFU/g). Adhesion of E. coli K88 to the duodenum, jejunum, colon, and mesenteric lymph node did not significantly differ (P > 0.05) between treatment groups. Adhesion of E. coli K99 did not significantly differ (P > 0.05) between phage-treated and untreated pigs for any sample type.

Meta-analysis

Data from the studies described in the above systematic review were aggregated, producing 19 inputs for meta-analysis and 576 total observations. Funnel plots did not show definite signs of asymmetry indicating no presence of publication bias (data not shown). Similarly, results from an Egger’s test were not significant (P = 0.34), indicating a low likelihood of small sample bias. Finally, P-curve analysis indicated the dataset had evidential value or effect size (Right Skewness Test: P < 0.05) and adequate power (Flatness Test: P = 0.99) with no evidence of “P-hacking”.

Comparisons of effects sizes across all groups and subgroups are presented in Tables 2 and 3. When all observations (n = 576) were analyzed together, concentrations of challenge bacteria were significantly lower (P < 0.0001) in pigs treated with phages compared to untreated pigs (P < 0.0001; effect size = −1.06 log10 CFU/g). While heterogeneity across all observations was considerable (I2 = 64.2%), post-hoc analysis of power indicated a 100% probability of detecting statistical differences in the data set.

Table 2.

Meta-analysis (effect sizes) of studies measuring challenge bacteria concentrations following application of phage in pigs

Parameter I2, % Effect size Effect significance, P-value Power, % No. observations, Treatment No. observations, Control Total observations
All observations together 64.2 −1.06 <0.0001 100 336 240 576
Sampling Time
 Samples collected ≤24 h post-phage treatment 57.4 −0.82 0.0007 96.96 252 156 408
 Samples collected 48–96 h post-phage treatment 36.6 −1.78 0.022 99.72 84 84 168
Age of Pig
 Piglets 62.8 −1.31 0.0001 100 216 174 390
  Piglets, samples collected ≤24 h post-phage treatment 65.8 −1.075 0.0037 99.97 132 90 222
  Piglets, samples collected 48–96 h post-phage treatment 36.6 −1.78 0.022 99.72 84 84 168
 Market Weight Pigs, samples collected ≤24 h post-phage treatment 0 −0.50 0.0138 99.76 120 66 186
Targeted Bacteria
 Salmonella 64.2 −0.95 <0.0001 100 312 216 528
  Salmonella, samples collected ≤24 h post-phage treatment 59.3 −0.77 0.002 100 240 144 384
  Salmonella, samples collected 48–96 h post-phage treatment1 0 −1.53 0.002 99.22 72 72 144
 E. coli 69.1 −1.89 0.1084 71.46 24 24 48
  E. coli, samples collected ≤24 h post-phage treatment1 66.5 −1.29 0.3743 43.15 12 12 24
  E. coli, samples collected 48–96 h post-phage treatment 82.7 −2.68 0.3622 431.5 12 12 24
Frequency of Phage Application
 Single Dose1 0.0 −0.25 0.0202 80.67 72 18 90
 Multiple Doses 63.4 −1.21 <0.0001 100 264 222 486
  Multiple Doses, samples collected ≤24 h post-phage treatment 60.9 −0.98 0.0011 100 180 138 318
  Multiple Doses, samples collected 48–96 h post-phage treatment 36.6 −1.78 0.022 99.72 84 84 168
Prophylaxis vs. Treatment
 Prophylaxis 0.0 −0.72 0.0017 99.99 144 102 270
  Prophylaxis, samples collected ≤24 h post-phage treatment 0.0 −0.24 0.0017 99.99 144 102 246
 Treatment 71.0 −1.31 0.0025 100 192 138 330
  Treatment, samples collected ≤24 h post-phage treatment 78.4 −0.93 0.0748 99.28 108 54 162
  Treatment, samples collected 48–96 h post-phage treatment 36.6 −1.78 0.022 99.72 84 84 168
Sample Site
 Ileum2 30.4 −0.40 0.0008 99.77 96 78 174
 Cecum 50.6 −0.92 0.0133 99.97 120 102 222
  Cecum, samples collected ≤24 h post-phage treatment 79.2 −0.60 0.2709 99.77 96 78 174
 Feces 69.4 −1.58 0.0062 99.97 120 102 222
  Feces, samples collected ≤24 h post-phage treatment 81.4 −1.28 0.1078 95.28 60 42 102
  Feces, samples collected 48–96 h post-phage treatment 50.6 −1.94 0.0617 97.91 60 60 120
Administration Route
 Gavage 68.9 −1.00 0.0002 100 270 216 486
  Gavage, samples collected ≤24 h post-phage treatment 67.6 −0.82 0.0061 100 198 144 342
  Gavage, samples collected 48–96 h post-phage treatment1 0 −1.53 0.002 99.22 72 72 144
 Feed 62.7 −1.35 0.0631 98.72 66 66 132
  Feed, Gavage, samples collected ≤24 h post-phage treatment 25.5 −0.81 0.0966 96.64 54 54 108
  Feed, samples collected 48–96 h post-phage treatment1 82.7 −2.68 0.3622 43.15 12 12 24
Phage Protection
 Protected 43.2 −1.027 0.0029 100 168 126 294
  Protected, samples collected ≤24 h post-phage treatment 10 −0.80 0.0011 100 156 114 270
  Protected, samples collected ≤24 h post-phage treatment1 82.7 −2.68 0.3622 43.15 12 12 24
 Unprotected 75.7 −1.13 0.0112 100 144 90 234
  Unprotected, samples collected ≤24 h post-phage treatment 85.0 −0.79 0.2307 97.82 96 42 138
  Unprotected, samples collected ≤24 h post-phage treatment1 0.0 −1.53 0.002 99.22 72 72 144
Open in a new tab

1Only one study; 2Data same as ileum <24 h.

Table 3.

Comparisons effect sizes across subgroups in studies measuring challenge bacteria concentrations following application of phage in pigs

Parameter
Subgroup 1		 Effect size Subgroup 2 Effect size Subgroup 3 Effect size P-value
All Observations
 Piglets −1.31 Market Wt. Pigs −0.50 0.002
 Feed Admin. −1.35 Gavage Admin. −1.01 0.57
 Poly-phage Treatment −1.21 Single-phage Treatment −0.25 <0.001
 Target: E. coli −1.89 Target: Salmonella −0.94 0.27
 Prophylaxis −0.72 Therapeutic −1.31 0.109
 Protected Phage −1.03 Unprotected Phage −1.13 0.80
 Cecal Samples −1.04 Fecal Samples −1.58 Ileal Samples −0.49 0.005
Samples collected ≤24 h post-phage treatment
 Market Wt. Pigs −0.50 Piglets −1.08 0.047
 Feed Admin −0.82 Gavage Admin. −0.82 0.99
 Poly-phage Treatment −0.98 Single-phage Treatment −0.25 0.0008
 Target: E. coli −1.29 Target: Salmonella −0.77 0.56
 Therapeutic −0.93 Prophylaxis −0.72 0.65
 Protected Phage −0.80 Unprotected Phage −0.79 0.99
 Cecal Samples −0.92 Fecal Samples −1.28 Ileal Samples −0.49 0.16
Samples collected 48–96 h post-phage treatment
 Feed Admin −2.68 Gavage Admin −1.53 0.50
 Target: E. coli −2.68 Target: Salmonella −1.53 0.50
 Protected Phage −2.68 Unprotected Phage −1.53 0.50
 Cecum −1.43 Feces −1.94 0.49
Piglets
 Samples collected ≤24 h post-phage treatment −1.08 Samples collected 48–96 h post-phage treatment −1.78 0.20
Feed Administrations
 Samples collected ≤24 h post-phage treatment −0.81 Samples collected 48–96 h post-phage treatment −2.68 0.29
Gavage Administrations
 Samples collected ≤24 h post-phage treatment −0.82 Samples collected 48–96 h post-phage treatment −1.53 0.0032
Poly-phage Treatment
 Samples collected ≤24 h post-phage treatment −0.98 Samples collected 48–96 h post-phage treatment −1.78 0.13
Target: E. coli
 Samples collected ≤24 h post-phage treatment −1.29 Samples collected 48–96 h post-phage treatment −2.68 0.47
Target: Salmonella
 Samples collected ≤24 h post-phage treatment −0.77 Samples collected 48–96 h post-phage treatment −1.53 0.00019
Therapeutic Application
 Samples collected ≤24 h post-phage treatment −0.93 Samples collected 48–96 h post-phage treatment −1.78 0.18
Protected Phage
 Samples collected ≤24 h post-phage treatment −0.80 Samples collected 48–96 h post-phage treatment −2.68 0.27
Unprotected Phage
 Samples collected ≤24 h post-phage treatment −0.79 Samples collected 48–96 h post-phage treatment −1.53 0.16
Cecal Samples
 Samples collected ≤24 h post-phage treatment −0.80 Samples collected 48–96 h post-phage treatment −1.43 0.14
Fecal Samples
 Samples collected ≤24 h post-phage treatment −1.28 Samples collected 48–96 h post-phage treatment −1.94 0.45
Open in a new tab

Phages are biologicals and their viability and subsequent lytic capacity can be affected by environmental conditions including heat, extreme pH, multiplicities of infection, and the presence of proteinases, among other factors (Iriarte et al., 2007; Huff et al., 2010; Knezevic et al., 2011; Hodyra-Stefaniak et al., 2015; Zhang et al., 2015; Colom et al., 2017; El-Dougdoug et al., 2019). Thus, our data set was disaggregated based on different factors in an effort to identify conditions under which effect size (i.e., reductions in challenge bacteria) was greatest.

As phages are self-replicating and lysis is often influenced by multiplicities of infection (Delbrück, 1940; Payne and Jansen, 2001; Kasman et al., 2002; Payne and Jansen, 2003; Huff et al., 2006; Callaway et al., 2008), we hypothesized that assessments of whether or not phage treatment reduced challenge bacteria could be influenced by sampling time, i.e., how soon bacterial concentrations were measured following phage application. To test this hypothesis, observations were disaggregated into two subgroups: 1) observations of bacterial concentrations measured ≤ 24 after the first application of bacteriophages; and 2) observations of bacterial concentrations measured 48 to 96 h after the first application of bacteriophages. Concentrations of challenge bacteria in phage-treated pigs were lower than in untreated pigs whether samples were collected ≤ 24 h following phage application (P < 0.001; effect size = −0.82 log10 CFU/mL; I2 = 57.4%) or 48 to 96 h following application of phages (P < 0.03; effect size = −1.78 log10 CFU/mL; I2 = 36.6%). Effect size, however, was larger (P < 0.05) with less heterogeneity in samples collected 48 to 96 h following application of phages. Thus, while phage application can significantly reduce concentrations of targeted bacteria within 24 h, greater reductions may be observed 2 to 4 d following phage application. Such results could have implications for the timing of phage administration for maximum bacterial reductions at desired time points.

Younger pigs are often more susceptible to infections than older pigs. This is true regarding gastrointestinal pathogens and, as our data set only included studies utilizing Salmonella or E. coli challenge models (Fairbrother et al., 2005; Wales et al., 2011), it was of interest to disaggregate the data set based on pigs’ age (i.e., piglets vs. market weight pigs). Treating piglets or market weight pigs with phages significantly (P < 0.05) reduced challenge bacteria concentrations (piglets: P < 0.0001; effect size = −1.31 log10 CFU/mL; I2 = 62.8%; market weight pigs: P < 0.015; effect size = −0.50 log10 CFU/mL; I2 = 0.0%). The piglet subgroup was further disaggregated by sample time. Concentrations of challenge bacteria in samples collected from piglets ≤ 24 h after phage application (P < 0.004; effect size = −1.076 log10 CFU/mL; I2 = 65.8%) or 48 to 96 h after phage application (effect size = −1.78 log10 CFU/mL; I2 = 36.6%) were both significantly (P = 0.05) lower than those of control pigs. As the data set did not contain any observations of samples collected from market weight pigs 48 to 96 h after phage application, the market weight pig subgroup was not further disaggregated by sample time. Taken together, phage treatment appears to be more effective at reducing challenge bacteria concentrations in younger pigs than in older pigs. The increased heterogeneity across studies utilizing market weight pigs and the absence of studies in the data set where challenge bacteria concentrations were measured > 24 h post-phage treatment in market weight pigs should temper these conclusions.

It is reasonable to hypothesize that efficacy of phage treatment depends on the bacterial organism targeted. As previously noted, however, the data set consisted only of studies utilizing either Salmonella or E. coli, which limited ability to make inferences on phage treatment efficacy to broader groups of bacteria. When the data set was disaggregated by challenge organism (i.e., Salmonella or E. coli), concentrations of Salmonella in phage-treated pigs were significantly lower in phage-treated pigs compared to untreated pig across sample times, but with considerable heterogeneity (P < 0.0001; effect size = −0.95log10 CFU/mL; I2 = 64.2%). As in other comparisons, effect size in studies targeting Salmonella was significantly (P < 0.001) greater when samples were collected 48 to 96 h (effect size = −1.53 log10 CFU/mL) compared to when samples were collected ≤ 24 h following phage application (effect size: −0.77 log10 CFU/mL). There was a trend for concentrations of E. coli in phage-treated pigs (P = 0.11; effect size: −1.90 CFU/mL) to be lower than those in untreated pigs. There were no significant differences in concentrations of E. coli between phage-treated and untreated pigs when those studies were further disaggregated by sample time. As disaggregation by sample time in this case reduced power statistics to 43.15%, inferences from these results should be tempered.

The studies included in our data set used a variety of dosing regimens. One variable across dosing regimens was the number of times phages were administered to the pigs. Only one study in the data set utilized a single-dose approach; in this instance, no differences in concentrations of the challenge bacteria between phage-treated and untreated pigs were reported. Administration of multiple doses of phages, however, significantly reduced challenge bacteria concentrations between phage-treated and untreated pigs (P = < 0.0001; effect size = −1.21 log10 CFU/mL; I2 = 63.4%). When studies using multiple doses of phage were disaggregated by sample time, there were no differences (P = 0.13) in effect size between samples collected ≤ 24 h after phage application (−0.98 CFU/mL) versus those collected 48 to 96 h following phage application (−1.78 CFU/mL).

Phage treatment could be used as a prophylaxis to prevent bacterial colonization and/or as a therapeutic to reduce existing bacterial infections. To assess the potential efficacy of phage treatments in these two scenarios, the data set was disaggregated based on whether phage treatment was applied before (prophylactic scenario) or after (therapeutic scenario) pigs were challenged with bacteria. Concentrations of challenge bacteria in phage-treated pigs were significantly lower in phage-treated pigs compared to control pigs whether the phage treatment was applied in prophylactic (P < 0.002; effect size = −0.72 log10 CFU/mL; I2 = 0.0%) or therapeutic manners (P < 0.003; effect size = −1.31 log10 CFU/mL; I2 = 71.0%). There were no statistical differences in targeted bacteria concentrations between pigs receiving phage prophylactically versus pigs receiving phage therapeutically regardless of sample time. When studies using a therapeutic approach were further disaggregated by sample time, there was a statistical trend (P = 0.08) for concentrations of challenge bacteria to be lower in phage-treated versus untreated pigs when samples were collected ≤ 24 h after phage application (effect size = −0.93; I2 = 78.4%). In contrast, concentrations of challenge bacteria were statistically lower in phage-treated pigs versus untreated pigs in therapeutic scenarios when samples were collected 48 to 96 h following phage application (P < 0.03; effect size = −1.78 log10 CFU/mL; I2 = 36.6) but not when samples were collected < 24 h following phage application. Finally, the efficacy of phages in a therapeutic scenario was not affected by sample time (no significant differences in ≤ 24 h vs. 48 to 96 h post-treatment samples). There were no studies in the data set that measured challenge bacteria concentrations at 48 to 96 h following phage application in a prophylaxis scenario making similar comparisons across sample time impossible.

As studies in the data set utilized gastrointestinal pathogens, bacterial concentrations were measured primarily in ileal, cecal, or fecal samples (including rectal samples). When the data set was disaggregated by sample site, significant differences were found in challenge bacteria concentrations in ileal (P < 0.001; effect size: −0.40 CFU/mL; I2 = 30.4%), cecal (P < 0.02; effect size: −0.92 CFU/mL; I2 = 50.6%) and fecal (P < 0.01; effect size: −0.1.58 CFU/mL; I2 = 69.4%) contents of phage-treated versus untreated pigs. When ileal, cecal, and fecal samples were disaggregated by sample time, no statistical differences in challenge bacteria concentrations between phage-treated and untreated pigs were detected across samples sites whether samples collected ≤ 24 after phage application or 48 to 96 h following phage application.

Phages have been applied orally, intranasally, and in-feed, among other routes (Wall et al., 2010; Saez et al., 2011; Verstappen et al., 2016). The studies included in the data set administered the phage treatments either by gavage (or similar oral application) or in feed. It is likely that the experiments in the data set utilized oral administration routes because the only challenge bacteria included in experiments in the data set, Salmonella and E. coli, are largely intestinal pathogens. Concentrations of challenge bacteria were significantly lower in phage-treated versus untreated pigs when phages were delivered by gavage (P < 0.001; effect size = −1.00 log10 CFU/mL; I2 = 68.9%) and there was a statistical trend for lower challenge organism concentrations in phage-treated versus untreated pigs when the phage was delivered in feed (P = 0.063; effect size = −1.35 log10 CFU/mL; I2 = 62.7%). The effect size in pigs administered phage by gavage was significantly (P < 0.004) greater when samples were collected 48 to 96 h following phage application (effect size: −1.53 CFU/mL) than when samples were collected ≤ 24 h after phage application (effect size: −0.82 log10 CFU/mL). In phage-treated pigs administered phages via feed, there were no statistical differences (P = 0.29) in effect size when samples were collected ≤ 24 h post-phage treatment (−0.81 log10 CFU/mL) versus samples collected 48 to 96 h post-phage treatment (−2.68 log10 CFU/mL). The smaller number of observations from studies using feed administration routes likely resulted in reduced statistical power (43.2%) and increased heterogeneity (I2 = 82.7%), which, when taken together, could have resulted in a scenario where substantially but not statistically different effect sizes were observed.

Numerous groups have shown that bacteriophage viability can be reduced due to digestive enzymes, extreme pH, high temperatures, host immunity, among other environmental factors (Vitiello et al., 2005; Capparelli et al., 2006; Iriarte et al., 2007; Huff et al., 2010; Knezevic et al., 2011; Hodyra-Stefaniak et al., 2015; Zhang et al., 2015; Colom et al., 2017; El-Dougdoug et al., 2019). As such, we disaggregated observations based on whether the phages themselves received any type of preparation in order to protect them from the gastric environment. Whether phages were unprotected or protected (microencapsulation and freeze drying were both used in different studies) prior to application, concentrations of challenge bacteria were significantly lower in phage-treated versus untreated pigs (unprotected: P < 0.02; effect size: −1.13 log10 CFU/mL; I2 = 75.70%; protected: P < 0.003; effect size: −1.03 log10 CFU/mL; I2 = 43.2%). When these data were disaggregated by sample time, concentrations of challenge bacteria in pigs treated with unprotected phages were significantly different from those of untreated pigs at when samples were collected 48 to 96 h following phage application (P < 0.01; effect size: −1.53 log10 CFU/mL; I2 = 0.0%) but not when collected ≤ 24 h after phage application (P = 0.24; effect size: −0.79 log10 CFU/mL; I2 = 85.0%). The opposite was seen among studies using protected phages where concentrations of target bacteria in phage-treated pigs were significantly lower than those of controls when samples were collected ≤ 24 h after phage application (P = 0.36; effect size: −2.6838; I2 = 82.7%). In both cases, the disappearance of statistical significance despite larger effect sizes may be due to the lower number of observations coupled with increased heterogeneity in the disaggregated data set. Likewise, no effect of sample time was detected between pigs treated with unprotected versus protected phages.

There are some potential limits to this study and to extrapolating its results to larger populations. In an effort to limit heterogeneity across inputs, the data set was limited to only studies that challenged pigs with known amounts of bacteria, treated those pigs with known amounts of phage, measured the concentration of the specific challenge bacteria post-phage treatment, and provided standard deviation values (or standard error values allowing standard deviation calculation). While using stringent criteria to select studies produced a data set of very methodologically similar experiments for comparison, doing so also reduced the number of inputs to the data set. Hippel et al. (2015) reported that I2, the statistic used in this study for gauging heterogeneity, can be biased in meta-analyses by small samples sizes (a case seen in many meta-analyses) as much as inter-study differences. Thus, while heterogeneity in meta-analyses is expected and heterogeneity in our data sets was usually moderate, heterogeneity was substantial in some cases when the data set was further disaggregated to form subgroups. In such cases, I2 values may have been influenced by small-sample bias rather than true heterogeneity. Furthermore, our post-hoc analysis revealed high power and significant effect between phage-treated and untreated group for most groups (Tables 2 and 3).

Additionally, it was of interest to understand the influence of phage concentration or titer on treatment efficacy. Creating sensible subgroups for comparison based on phage concentrations, however, was not possible as: 1) there was not a tremendous amount of variability in phage concentrations employed and few studies included low concentration (e.g., 105 PFU/mL) treatments; 2) different studies used different application rates and frequencies presenting confounding factors to accurately calculating the actual amount of phage administered; and 3) of the few studies that measured phages at sites of infection, none discriminated between applied phages and endogenous phages. In the end, subgroups for comparison were made based on the frequency of phage administration over the course of treatment (single vs. multiple applications); however, we recognize that such a comparison provides only limited insight into the impact of phage concentration on phage treatment efficacy.

Nevertheless, the data presented here do indicate that across independent studies, administering phage treatment to pigs results in significantly lower concentrations of the challenge bacteria in tissues of phage treated pigs compared to those of challenged but untreated pigs. As such, these findings support the use of phage therapy as a potential alternative to traditional antibiotics. These data also provide insight regarding scenarios in which phage therapy may be most effective in controlling bacterial transmission in pig production. For example, reductions in targeted bacteria may be time dependent with the greatest effect seen 48 h or more after phage treatment. As in the case of antibiotic treatment, multiple application of phage over the course of the treatments is likely to increase treatment efficacy.

Finally, this meta-analysis considers the impact of phage treatment on concentrations of targeted bacteria. In many cases, antibiotic treatment of animals as a management practice ultimately results in improved growth efficiency or performance. Thus, it will be of interest to conduct similar analyses as presented here to determine whether phage treatment improves key performance indicators such as average daily gain, feed intake, and/or feed:gain conversions.

Supplementary Material

skab157_suppl_Supplementary_Table_1
Click here for additional data file. (125.9KB, docx)

Acknowledgments

This research is possible through the financial support from the UK Government – Department of Health and Social Care (DHSC), the Global AMR Innovation Fund (GAMRIF) and the International Development Research Centre, Ottawa, Canada.

Glossary

Abbreviations

CFU

colony-forming units

DNA

deoxyribonucleic acid

I2

inconsistency index or percentage of heterogeneity

MOI

multiplicity of infection

PBS

phosphate-buffered saline

PFU

plaque-forming units

Conflict of interest statement

The authors declare they have no conflicts of interest.

Literature Cited

  1. Albino, L. A. A., Rostagno M. H., Húngaro H. M., and Mendonҫa R. C. S.. . 2014. Isolation, characterization, and application of bacteriophages for Salmonella spp. biocontrol in pigs. Foodborne Pathog Dis. 11(8):602–609. doi: 10.1089/fpd.2013.1600 [DOI] [PubMed] [Google Scholar]
  2. Årdal, C., Balasegaram M., Laxminarayan R., McAdams D., Outterson K., Rex J. H., and Sumpradit N.. . 2020. Antibiotic development—economic, regulatory and societal challenges. Nat. Rev. Microbiol. 18:267–274. doi: 10.1038/s41579-019-0293-3 [DOI] [PubMed] [Google Scholar]
  3. Callaway, T. R., Edrington T. S., Brabban A. D., Anderson R. C., Rossman M. L., Engler M. J., Carr M. A., Genovese K. J., Keen J. E., Looper M. L., Kutter E. M., and Nisbet D. J.. . 2008. Bacteriophage isolated from feedlot cattle can reduce Escherichia coli O157:H7 populations in ruminant gastrointestinal tracts. Foodborne Pathog. Dis. 5(2):183–191. doi: 10.1089/fpd.2007.0057 [DOI] [PubMed] [Google Scholar]
  4. Callaway, T. R., Edrington T. S., Brabban A., Kutter B., Karriker L., Stahl C., Wagstrom E., Anderson R., Poole T. L., Genovese K., . et al. 2011. Evaluation of phage treatment as a strategy to reduce Salmonella populations in growing swine. Foodborne Pathog Dis. 8(2):261–266. doi: 10.1089/fpd.2010.0671 [DOI] [PubMed] [Google Scholar]
  5. Capparelli, R., Ventimiglia I., Roperto S., Fenizia D., and Iannelli D.. . 2006. Selection of an Escherichia coli O157:H7 bacteriophage for persistence in the circulatory system of mice infected experimentally. Clin. Microbiol. Infect. 12: 248–253. doi: 10.1111/j.1469-0691.2005.01340.x [DOI] [PubMed] [Google Scholar]
  6. Centers for Disease Control and Prevention. 2021. One health [accessed May 25, 2021]. https://www.cdc.gov/onehealth/index.html.
  7. Colom, J., Cano-Sarabia M., Otero J., Aríñez-Soriano J., Cortés P., Maspoch D., and Llagostera M.. . 2017. Microencapsulation with alginate/CaCO3: a strategy for improved phage therapy. Sci. Rep. 7:41441. doi: 10.1038/srep41441 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Delbrück, M. 1940. The growth of bacteriophage and lysis of the host. J Gen Physiol. 23(5):643–660. doi: 10.1085/jgp.23.5.643 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. El-Dougdoug, N. K., Cucic S., Abdelhamid A. G., Brovko L., Kropinski A. M., Griffiths M. W., and Anany H.. . 2019. Control of Salmonella Newport on cherry tomato using a cocktail of lytic bacteriophages. Int. J. Food Microbiol. 293:60–71. doi: 10.1016/j.ijfoodmicro.2019.01.003 [DOI] [PubMed] [Google Scholar]
  10. European Commission. 2018. New EU rules on veterinary medicinal products and medicated feed [accessed May 25, 2021]. https://ec.europa.eu/food/sites/food/files/animals/docs/ah_vet-med_feed_factsheet-2018_en.pdf.
  11. Fairbrother, J. M., Nadeau E., and Gyles C. L.. . 2005. Escherichia coli in postweaning diarrhea in pigs: an update on bacterial types, pathogenesis, and prevention strategies. Anim. Health Res. Rev. 6:17–39. doi: 10.1079/ahr2005105 [DOI] [PubMed] [Google Scholar]
  12. Fischetti, V. A. 2018. Development of phage lysins as novel therapeutics: a historical perspective. Viruses. 10(6):310. doi: 10.3390/v10060310 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Ghosh, C., Sarkar P., Issa R., and Haldar J.. . 2019. Alternatives to conventional antibiotics in the era of antimicrobial resistance. Trends Microbiol. 27:323–338. doi: 10.1016/j.tim.2018.12.010 [DOI] [PubMed] [Google Scholar]
  14. Han, S. J., Oh Y., Lee C. Y., and Han J. H.. . 2016. Efficacy of dietary supplementation of bacteriophages in treatment of concurrent infections with enterotoxigenic Escherichia coli K88 and K99 in postweaning pigs. J Swine Health Prod. 24(5):259–263. [Google Scholar]
  15. Harrer, M., Cuijpers P., Furukawa T. A., and Ebert D. D.. . 2019a. Doing meta-analysis in R: a hands-on guide. https://bookdown.org/MathiasHarrer/Doing_Meta_Analysis_in_R/. doi: 10.5281/zenodo.2551803 [DOI]
  16. Harrer, M., Cuijpers P., Furukawa T., and Ebert D. D.. . 2019b. dmetar: Companion R package for the guide ‘Doing Meta-Analysis in R’. R package version 0.0.9000 [accessed May 25, 2021]. https://dmetar.protectlab.org/.
  17. Higgins, J. P., and Thompson S. G.. . 2002. Quantifying heterogeneity in a meta-analysis. Stat Med. 21(11):1539–58. doi: 10.1002/sim.1186.PMID:12111919 [DOI] [PubMed] [Google Scholar]
  18. Hippel, P. T. 2015. The heterogeneity statistic I2 can be biased in small meta-analyses. BMC Med Res Method. 15:35. doi: 10.1186/s12874-015-0024-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hodyra-Stefaniak, K., Miernikiewicz P., Drapała J., Drab M., Jończyk-Matysiak E., Lecion D., Kaźmierczak Z., Beta W., Majewska J., Harhala M., Bubak B., Kłopot A., Górski A., and Dąbrowska K.. . 2015. Mammalian host-versus-phage immune response determines phage fate in vivo. Sci. Rep., 5:14802. doi: 10.1038/srep14802 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hu, J. Y., and Cowling B. J.. . 2020. Reducing antibiotic use in livestock, China. Bulletin of the World Health Organization. 98:360–361. doi: 10.2471/BLT.19.243501 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Huff, W. E., Huff G. R., Rath N. C., and Donoghue A. M.. . 2006. Evaluation of the influence of bacteriophage titer on the treatment of colibacillosis in broiler chickens. Poult. Sci. 85(8):1373–1377. doi: 10.1093/ps/85.8.1373 [DOI] [PubMed] [Google Scholar]
  22. Huff, W. E., Huff G. R., Rath N. C., and Donoghue A. M.. . 2010. Immune interference of bacteriophage efficacy when treating colibacillosis in poultry. Poult Sci. 89(5):895–900. doi: 10.3382/ps.2009-00528 [DOI] [PubMed] [Google Scholar]
  23. Iriarte, F. B., Balogh B., Momol M. T., Smith L. M., Wilson M., and Jones J. B.. . 2007. Factors affecting survival of bacteriophage on tomato leaf surfaces. Appl. Environ. Microbiol. 73(6):1704–1711. doi: 10.1128/AEM.02118-06 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kasman, L. M., Kasman A., Westwater C., Dolan J., Schmidt M. G., and Norris J. S.. . 2002. Overcoming the phage replication threshold: a mathematical model with implications for phage therapy. J. Virol. 76:5557–5564. doi: 10.1128/jvi.76.11.5557-5564.2002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Knezevic, P., Obreht D., Curcin S., Petrusic M., Aleksic V., Kostanjsek R., and Petrovic O.. . 2011. Phages of Pseudomonas aeruginosa: response to environmental factors and in vitro ability to inhibit bacterial growth and biofilm formation. J. Appl. Microbiol. 111:245–254. doi: 10.1111/j.1365-2672.2011.05043.x [DOI] [PubMed] [Google Scholar]
  26. Kutter, E., De Vos D., Gvasalia G., Alavidze Z., Gogokhia L., Kuhl S., and Abedon S. T.. . 2010. Phage therapy in clinical practice: treatment of human infections. Curr. Pharm. Biotechnol. 11: 69–86. doi: 10.2174/138920110790725401 [DOI] [PubMed] [Google Scholar]
  27. Monteiro, R., Pires D. P., Costa A. R., and Azeredo J.. . 2019. Phage therapy: going temperate? Trends Microbiol. 27:368–378. doi: 10.1016/j.tim.2018.10.008 [DOI] [PubMed] [Google Scholar]
  28. Park, J. Y., Moon B. Y., Park J. W., Thornton J. A., Park Y. H., and Seo K. S.. . 2017. Genetic engineering of a temperate phage-based delivery system for CRISPR/Cas9 antimicrobials against Staphylococcus aureus. Sci Rep-UK. 7:44929. doi: 10.1038/srep44929 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Payne, R. J. H., and Jansen V. A. A.. . 2001. Understanding bacteriophage therapy as a density-dependent kinetic process. J Theor Biol. 208(1):37–48. doi: 10.1006/jtbi.2000.219 [DOI] [PubMed] [Google Scholar]
  30. Payne, R. J. H., and Jansen V. A. A.. . 2003. Pharmacokinetic principles of bacteriophage therapy. Clin Pharmacokinet. 42(4):315–325. doi: 10.2165/00003088-200342040-00002 [DOI] [PubMed] [Google Scholar]
  31. RStudio Team. 2020. RStudio: integrated development for R. Boston (MA): RStudio, PBC; [accessed May 25, 2021]. http://www.rstudio.com. [Google Scholar]
  32. Saez, A. C., Zhang J., Rostagno M. H., and Ebner P. D.. . 2011. Direct feeding of microencapsulated bacteriophages to reduce Salmonella colonization in pigs. Foodborne Pathog. Dis. 8: 1269–1274. doi: 10.1089/fpd.2011.0905 [DOI] [PubMed] [Google Scholar]
  33. Sulakvelidze, A., Alavidze Z., and J. G. Morris, Jr. 2001. Bacteriophage therapy. Antimicrob. Agents Chemother. 45: 649–659. doi: 10.1128/AAC.45.3.649-659.2001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. United States Food and Drug Administration [US FDA]. 2021. Veterinary Feed Directive (VFD) accessed May 25, 2021]. https://www.fda.gov/animal-veterinary/development-approval-process/veterinary-feed-directive-vfd.
  35. Vázquez, R., García E., and García P.. . 2018. Phage lysins for fighting bacterial respiratory infections: a new generation of antimicrobials. Front Immunol. 9:2252. doi: 10.3389/fimmu.2018.02252 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Verstappen, K. M., Tulinski P., Duim B., Fluit A. C., Carney J., van Nes A., and Wagenaar J. A.. . 2016. The effectiveness of bacteriophages against methicillin-resistant Staphylococcus aureus ST398 nasal colonization in pigs. PLoS One 11:e0160242. doi: 10.1371/journal.pone.0160242 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Vitiello, C. L., Merril C. R., and Adhya S.. . 2005. An amino acid substitution in a capsid protein enhances phage survival in mouse circulatory system more than a 1000-fold. Virus Res. 114:101–103. doi: 10.1016/j.virusres.2005.05.014. [DOI] [PubMed] [Google Scholar]
  38. Wales, A. D., Cook A. J., and Davies R. H.. . 2011. Producing Salmonella-free pigs: a review focusing on interventions at weaning. Vet. Rec. 168:267–276. doi: 10.1136/vr.d1125 [DOI] [PubMed] [Google Scholar]
  39. Wall, S. K., Zhang J., Rostagno M. H., and Ebner P. D.. . 2010. Phage therapy to reduce preprocessing Salmonella infections in market-weight swine. Appl. Environ. Microbiol. 76:48–53. doi: 10.1128/AEM.00785-09 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Yosef, I., Manor M., Kiro R., and Qimron U.. . 2015. Temperate and lytic bacteriophages programmed to sensitize and kill antibiotic-resistant bacteria. Proc. Natl. Acad. Sci. USA. 112:7267–7272. doi: 10.1073/pnas.1500107112 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Zhang, J., Hong Y., Fealey M., Singh A., Walton K., Martin C., Harman N. J., Mahlie J., and Ebner P. D.. . 2015. Physiological and molecular characterization of Salmonella Bacteriophages previously used in phage therapy. J. Food Prot. 78:2143–2149. doi: 10.4315/0362-028X.JFP-14-350 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

skab157_suppl_Supplementary_Table_1
Click here for additional data file. (125.9KB, docx)

Articles from Journal of Animal Science are provided here courtesy of Oxford University Press

RESOURCES