Biofilms are an extremely common adaptation, allowing bacteria to colonize hostile environments. They present unique problems for antibiotics and biocides, both due to the nature of the extracellular matrix and to the presence within the biofilm of metabolically inactive persister cells. Such chemicals can be highly effective against planktonic bacterial cells, while being essentially ineffective against biofilms. By contrast, bacteriophages seem to have a greater ability to target this common form of bacterial growth. The high numbers of bacteria present within biofilms actually facilitate the action of bacteriophages by allowing rapid and efficient infection of the host and consequent amplification of the bacteriophage. Bacteriophages also have a number of properties that make biofilms susceptible to their action. They are known to produce (or to be able to induce) enzymes that degrade the extracellular matrix. They are also able to infect persister cells, remaining dormant within them, but re-activating when they become metabolically active. Some cultured biofilms also seem better able to support the replication of bacteriophages than comparable planktonic systems. It is perhaps unsurprising that bacteriophages, as the natural predators of bacteria, have the ability to target this common form of bacterial life.

Bacteriophages, viruses that only kill specific bacteria, are receiving substantial attention as nontraditional antibacterial agents that may help alleviate the growing antibiotic resistance problem in medicine. We describe the design and preclinical development of AB-SA01, a fixed-composition bacteriophage product intended to treat Staphylococcus aureus infections. AB-SA01 contains three naturally occurring, obligately lytic myoviruses related to Staphylococcus phage K. AB-SA01 component phages have been sequenced and contain no identifiable bacterial virulence or antibiotic resistance genes. In vitro, AB-SA01 killed 94.5% of 401 clinical Staphylococcus aureus isolates, including methicillin-resistant and vancomycin-intermediate ones for a total of 95% of the 205 known multidrug-resistant isolates. The spontaneous frequency of resistance to AB-SA01 was ≤3 × 10⁻⁹, and resistance emerging to one component phage could be complemented by the activity of another component phage. In both neutropenic and immunocompetent mouse models of acute pneumonia, AB-SA01 reduced lung S. aureus populations equivalently to vancomycin. Overall, the inherent characteristics of AB-SA01 component phages meet regulatory and generally accepted criteria for human use, and the preclinical data presented here have supported production under good manufacturing practices and phase 1 clinical studies with AB-SA01.