Bacteriophage-Antibiotic Combinations in the Era of Antimicrobial Resistance

The continuing rise of antimicrobial resistance (AMR) has created an urgent need for therapeutic approaches that extend beyond the conventional development of new antibiotics. The challenge is particularly evident in multidrug-resistant Gram-negative bacteria, for which available treatment options are limited. The 2024 World Health Organization Bacterial Priority Pathogens List places carbapenem-resistant Enterobacterales among the critical-priority pathogens for research and development, together with other major resistant organisms that continue to threaten contemporary clinical practice [1]. Against this background, bacteriophage (or phage) therapy has re-emerged as a potential therapeutic option for difficult-to-treat bacterial infections, including those for which effective antibiotic options are limited, and as an adjunct to conventional therapy.

Unlike conventional antibiotics, phages kill susceptible bacteria through highly specific biological interactions and mechanisms that differ from those of conventional antibiotics. This makes it possible, at least in principle, to target bacterial strains that have accumulated resistance to multiple antibiotic classes. Phage therapy may be valuable when antibiotic options are limited, or conventional antibiotic therapy has failed. Because phages may also be used with antibiotics, understanding phage-antibiotic interactions is clinically important.

When phages are used in combination with antibiotics, the rationale is not limited to achieving greater bacterial killing by using two antibacterial agents. Because phages and antibiotics generally exert different selective pressures, resistance to one does not necessarily confer resistance to the other. Moreover, bacterial adaptation to phage selective pressure may come with fitness trade-offs that could be therapeutically beneficial. Phage resistance can involve alterations in bacterial receptors, capsules, lipopolysaccharides, outer-membrane components, or efflux systems [2,3]. When these structures also contribute to AMR or bacterial fitness, resistance emerging under phage selective pressure may come at the cost of increased antibiotic susceptibility, reduced virulence, impaired biofilm formation, or other unfavorable phenotypes [2,3]. The interaction between efflux pump-targeting phages and Pseudomonas aeruginosa is a well-described example of such an evolutionary trade-off [4].

Such evolutionary trade-offs could be clinically useful, but they are not consistently observed. Recent experimental studies have demonstrated that phage resistance may increase susceptibility to some antibiotics while increasing resistance to others, even within the same bacterial species [5]. Therefore, phage resistance can have different consequences for antibiotic susceptibility and bacterial fitness, and these consequences are not necessarily therapeutically favorable. The outcome depends on the bacterial strain, the receptor targeted by the phage, the antibiotic involved, and the underlying resistance mechanism. These observations support evaluating specific phage-antibiotic combinations on the basis of demonstrated biological interactions, rather than assuming that adding a phage will necessarily improve antibiotic activity.

Indeed, phage-antibiotic interactions can range from synergy to additivity, indifference and even antagonism [6]. Antibiotic class and concentration, bacterial physiological state, multiplicity of infection, phage replication characteristics, and the timing and sequence of administration can all affect the interaction. Therefore, phage-antibiotic synergy (PAS) should not be considered an intrinsic property of combination therapy, but rather a context-dependent interaction that needs to be experimentally defined for each clinically relevant phage-bacterium-antibiotic combination.

In this issue of Infection & Chemotherapy, Farmehr et al. provide a useful example of this principle by examining a newly isolated phage, vB_KpnS_Pey, alone and in combination with ciprofloxacin, meropenem, gentamicin, and colistin against clinical carbapenem-resistant Klebsiella pneumoniae (CRKP) isolates [7]. The study examines a clinically important pathogen and evaluates the interactions of a specific therapeutic phage with antibiotics from different classes.

Phage vB_KpnS_Pey exhibited activity against only 4 of 30 CRKP isolates, demonstrating the narrow host range commonly encountered with individual therapeutic phages. Among susceptible isolates, however, phage infection was generally productive.

When the phage was combined with antibiotics, the effect varied substantially according to both antibiotic and bacterial isolate. Reductions in minimum inhibitory concentrations (MICs) were observed for several antibiotics, but these changes were heterogeneous. In the time-kill experiments, a synergistic effect was demonstrated for the combination of vB_KpnS_Pey with subinhibitory concentrations of ciprofloxacin and meropenem against the Kp4 isolate, whereas comparable synergy was not observed with gentamicin or colistin [7].

However, in vitro synergy alone provides limited information about whether a combination will be clinically useful. A favorable PAS signal in an in vitro system does not necessarily translate into improved clinical outcomes. In the study by Farmehr et al., the time-kill synergy was demonstrated in a single phage-susceptible isolate, and bacterial growth was primarily assessed using optical density rather than direct enumeration of viable bacteria. The authors appropriately recognize that colony-forming unit-based confirmation, broader strain testing, and subsequent in vivo investigation are needed [7].

A further challenge is that the pharmacokinetics/pharmacodynamics (PK/PD) basis of phage-antibiotic combination therapy is still poorly defined. Conventional antibiotic PK/PD links drug exposure to bacterial killing through parameters such as the area under the concentration-time curve, peak concentration, or time above the MIC. Therapeutic phages introduce another level of complexity because they are self-replicating biological agents. Their concentration at the site of infection depends not only on administered dose and distribution, but also on bacterial density, adsorption kinetics, amplification in susceptible bacterial hosts, phage clearance, and host immunity [8]. The addition of antibiotics makes these dynamics even more complex. Antibiotic exposure can alter bacterial growth, morphology, receptor expression, and metabolism and may thereby enhance or inhibit phage adsorption and replication. Consequently, phage-antibiotic interactions may vary according to antibiotic concentration and the relative exposure to phage and antibiotic [9]. Treatment sequence may also influence these interactions, with simultaneous and sequential administration potentially yielding different antibacterial effects [10]. Antibiotic MIC and phage multiplicity of infection remain useful experimental parameters, but neither is sufficient to determine how a phage-antibiotic combination should be used clinically.

Future studies should therefore test phage-antibiotic combinations across clinically relevant drug exposures and treatment sequences, rather than relying on synergy demonstrated under a single in vitro condition. The emergence of resistance during combination treatment should also be assessed.

The therapeutic potential of phages in the AMR era is not limited to their use in combination with antibiotics. Their ability to target susceptible antibiotic-resistant bacteria may be valuable when effective antibiotic options are limited. At the same time, their interactions with antibiotics and effects on bacterial evolution create additional opportunities and challenges for combination therapy.

The study by Farmehr et al. contributes to this evolving field by showing that, even within a single phage-bacterium system, the interaction can vary according to the antibiotic used [7]. Their findings provide a useful example of the need to better understand phage-antibiotic interactions rather than evidence for a specific treatment regimen for CRKP.

In conclusion, a key clinical question for phage-antibiotic combination therapy is how to identify combinations that are reproducibly effective under relevant treatment conditions. Translating PAS from an in vitro phenomenon into a clinically useful strategy will require a better understanding of how the phage, bacterial strain, antibiotic, exposure conditions, and treatment sequence together determine the interaction.