A cocktail of in vitro efficient phages is not a guarantee for in vivo therapeutic results against avian colibacillosis

Avian pathogenic Escherichia coli (APEC) causes colibacillosis in poultry, leading to important economic losses worldwide. To cure APEC-infected chickens, a cocktail of four different APEC-specific bacteriophages (phages) was composed and tested. Specific phages were selected from a collection of phages isolated in Belgium. The selection was based on their obligate lytic infection cycle, a broad host range, low cross-resistance and low frequency of development of resistant APEC mutants. Genome analysis of the phages indicated they were close relatives of T4 and N4, considered to be safe in vivo. Chickens were intratracheally infected with APEC strain CH2 (serogroup O78), causing a mortality of about 50% during the seven days following the infection. The phage cocktail was administered 2 h after the infection, via three different ways: intratracheally, intra-esophageally or via the drinking water. Treated groups did not show a significant decrease in mortality, lesion scores or weight loss compared to untreated groups, although the APEC-specific phages could be re-isolated from the lung and heart of chickens that were euthanized. Moreover, the re-isolated bacteria from infected chickens had remained sensitive to the phage cocktail. Our results indicate that the efficiency of the phage cocktail used in treating CH2-infected chickens in vivo is negligible, even though in vitro, the phages in the cocktail were able to efficiently lyse the APEC strain CH2. Our results emphasize that the 'traditional' pathway of isolation, followed by phenotypical and genotypical characterization of phages composing the cocktail, does not lead to success in phage therapy in all cases.