As multi-drug antibiotic resistance emerges as a potent public health challenge, medical science has placed renewed attention on the potential for bacteriophage therapy. Bacteriophage, or phage, are viruses that target, infect and replicate inside bacteria, destroying them in the process. To help maximize the success rate of this approach, researchers recently modeled the dynamics of bacteriophage therapy to explain the dynamics of particular therapies and optimize the composition of bacteriophage cocktails.
“Phage are the most prevalent organisms on the planet,” said Alan Perelson, Los Alamos National Laboratory scientist and co-author on the research. “They exist everywhere bacteria exist. However, each phage has evolved to narrowly target specific bacteria, and bacteria have evolved various mechanisms of resistance.”
Phage cocktails — combinations of particular phages for a patient facing a specific bacterial infection — represent a complicated form of personalized medicine; given the fast and complex dynamics of bacterial responses to the phages, it is not typically known why a particular phage therapy succeeded or failed. As described in the journal PLOS Computational Biology, the research team developed a mathematical model that could describe effective cocktails of phages that could optimize the diversity and timing of the cocktails.
Improving phage cocktail composition
The researchers developed their mathematical model by building on an existing model calibrated with data from phage therapy in a living mouse. The mathematical model was extended for the human situation, and it included multiple phages infecting multiple bacterial strains with varying phage resistance.
The model was able to predict success based on several key factors. The bacteria’s pretreatment resistance level was critical, as was the diversity of the phage cocktail and the timing of its delivery. Phage therapy is a complicated dynamic where the more infective the phages are, the faster they are able to wipe out the more sensitive (i.e., less resistant) of the bacteria present. That leaves the resistant bacteria to expand, and they can evolve better resistance quickly, mutating to avoid infection by the phage.
The team found that the therapy cocktail is best served with a diversity of phages, which overwhelm the bacteria’s ability to evolve resistance fast enough. The team also focused on the timing of phage delivery, determining that immediate treatment with the full suite of phage cocktail offered the most success. That creates a high genetic barrier to bacterial resistance, meaning that the bacteria would need to accumulate several genetic changes or mutations in order to survive the therapy.
“The rapid evolution of resistance is the main challenge to therapy,” Perelson said. “That capacity is why antibiotics may not work in the first place. For phage therapy to be effective, the cocktails should be diverse, sufficient and immediate. The approach amounts to ‘hit the bacteria hard and early.’”
Funding: The work was supported by the Laboratory Directed Research and Development program at Los Alamos National Laboratory.
The paper: “Towards modeling phage therapy.” PLOS Computational Biology. DOI: doi.org/10.1371/journal.pcbi.1014408
LA-UR-26-27527
Contact
Media Relations | media_relations@lanl.gov
