Systematic combinatorial optimization of three-phage cocktails against multidrug-resistant Pseudomonas aeruginosa

Multidrug-resistant (MDR) Pseudomonas aeruginosa poses a significant global health threat, necessitating alternative therapeutic strategies. Bacteriophage therapy shows promise, but single-phage use is prone to resistance. Bacteriophage cocktails can address this issue, but they are typically formulated empirically without a systematic evaluation of constituent phage interactions. We employed a comprehensive combinatorial approach to optimize three-phage cocktails against clinical P. aeruginosa isolates. From 25 candidate bacteriophages, 5 were selected based on a broad host range (≥60% of 51 MDR P. aeruginosa clinical isolates). Ten possible three-phage cocktail combinations were systematically generated and evaluated using real-time Omnilog phenotypic microarray analysis. Phage interactions were quantified using the Highest Single Agent independence model to classify synergistic (Δ > +5%), neutral (-5% ≤ Δ ≤ +5%), or antagonistic (Δ < -5%) effects. Individual phage inhibition efficiencies ranged from 35.4% to 75.4%. Cocktail performance varied dramatically (16.1%-84.1% inhibition). Only one cocktail exhibited synergy, achieving 84.1% inhibition, an 8.7% improvement over the best individual phage. Four cocktails (40%) showed neutral interactions, while five (50%) demonstrated antagonistic effects. The optimal cocktail (Phages 1 + 3 + 4) demonstrated efficacy across six of eight diverse clinical isolates in biofilm inhibition assays (29%-63% biomass reduction, 2.7-3.6 log₁₀ colony-forming unit reduction) and achieved 80% survival in Galleria mellonella versus 10% in infected controls. These findings demonstrate that phage cocktail optimization requires rational validation, as antagonistic interactions can occur, establishing a critical validation step before committing resources to in vivo efficacy studies.

Bacteriophage cocktails are increasingly proposed as alternatives to antibiotics against multidrug-resistant bacteria, yet their development remains largely empirical. This study challenges the assumption that the combination of effective phages yields additive or synergistic benefits. Through systematic evaluation of all possible three-phage combinations from five phages, we discovered that antagonistic interactions (50%) predominated over synergistic effects (10%), with some cocktails performing worse than individual phages. Using the Highest Single Agent independence model, we established a quantitative framework for rational cocktail design that identified a therapeutically superior formulation. This approach is generalizable across bacterial pathogens and provides a critical validation step for phage therapy development. Our findings demonstrate that rational testing of phage interactions is essential before in vivo or clinical translation, potentially preventing costly failures and accelerating the development of effective phage-based therapeutics against multidrug-resistant infections.