Modeling multiphage-bacteria kinetics to predict phage therapy potency and longevity

Pseudomonas aeruginosa is a frequent cause of life-threatening opportunistic infections in the critically ill and immunocompromised. Its treatment is challenging due to the increasing prevalence of resistance to most conventional antibiotics. Although numerous alternative therapies are currently under investigation, bacteriophage (phage) cocktail therapy appears poised for long-term success. Here, we investigate potency and longevity of individual Pseudomonas phages in cocktail to determine viral co-factors that promote optimal treatment efficacy. We combined in vitro and in silico models to predict sixty-eight treatment permutations with three phages that adsorb symmetrically and asymmetrically when administered singly, double simultaneously, or double sequentially. We showed that simultaneously administering two asymmetrically binding phages with high cell lysis efficiencies improved cocktail potency. Use of a higher-potency cocktail, along with a reduction in the net probability of independent gene mutations was associated with prolonged bacterial suppression. Nevertheless, in vitro we almost always observed evolution of multiphage resistance. Simulations also predict that when combining phages with polar potencies, susceptible host cells are monopolized by the more efficiently replicating phage. Thus, further perpetuating the growth demise of the weaker phage in cocktail. Our mathematical model was used to explore and predict changes in phage and bacterial populations that were difficult to measure experimentally. This framework has many inferential and exploratory uses for clinical investigation such as identifying the most sensitive parameters for phage selection and exploring different treatment regimens. Collectively, our findings attempt to dissect the mechanisms of phage cocktails combating P. aeruginosa infections and highlight the viral co-factors necessary for treatment efficacy.