Uneven phage killing emerged inside chronic Pseudomonas aeruginosa lung infections
Phage therapy is gaining renewed attention as antibiotic resistance erodes treatment options, but this preprint shows why bacterial precision matters so much in practice. In chronic Pseudomonas aeruginosa lung infections from people with cystic fibrosis, susceptibility to bacteriophages was not uniform across isolates taken from the same patient. Instead, the investigators found substantial variation within individual lung-derived populations, a finding with direct implications for any therapy that depends on a phage matching its bacterial target closely enough to lyse it.
The study examined four chronic lung populations and tested 12 virulent phages against 40 isolates from each patient. Most of the variation centered on four lipopolysaccharide, or LPS, targeting phages. Two of the four lung-derived populations showed substantial shifts in susceptibility across the isolates sampled from those patients. In a chronic airway infection, where bacteria persist under immune pressure, antibiotics, and phage exposure, that kind of diversity is not a side issue. It is the biological obstacle sitting inside the infection itself.
One sentence captures the clinical problem cleanly: a single patient’s lung can contain bacterial subpopulations that do not answer to the same phage.
LPS variation, not phage defense cargo, appeared to drive the split
The genomic analysis did not point to a simple explanation based on classical phage defense systems. Carriage of phage defense genes did not strongly vary within the lung-derived populations, making gain or loss of those systems a weak fit for the observed heterogeneity. In other words, the isolates were not obviously diverging because one subcluster had imported a protective molecular shield while another had not.
Instead, the most informative signal came from one lung-derived population in which phylogenetic analysis linked some evolutionary transitions in phage susceptibility to mutations in genes involved in LPS biosynthesis. That matters because LPS sits on the bacterial surface and often serves as the first molecular landing zone for phages. Alterations in this structure can change whether a phage can attach, inject its genome, and begin the lytic cycle. In chronic infection, surface remodeling is a familiar survival strategy: bacteria under long-term pressure often trade away receptor structures or modify them to escape predators, but those same changes can reshape phage susceptibility in uneven ways across the population.
The pattern supports a straightforward biological interpretation. Within-patient diversification in cell-surface features may have produced the susceptible and resistant subgroups observed in the same infection. Rather than a uniform population drifting together, the lung appears to function as a selective environment where different surface states can persist side by side.
Antibiotic resistance and phage susceptibility aligned only in one population
The phenotypic analysis added another layer. In one population, phage susceptibility tracked with favorable clinical associations to antibiotic resistance, suggesting a possible route by which phage pressure could push bacteria toward a less clinically troublesome phenotype. But that relationship was not consistent. Associations were largely absent in the other populations, limiting confidence that phage use can reliably drive predictable phenotypic shifts once heterogeneity is already established.
That inconsistency is clinically important. Phage therapy is often discussed as if the treatment simply needs the right viral match, yet this preprint shows that matching may need to happen repeatedly inside the same patient, against multiple coexisting bacterial lineages. If only a subset of isolates within a chronic infection are susceptible, then a phage cocktail chosen from one sample may miss the organisms most relevant to persistence. The result could be partial killing, local ecological replacement, and survival of resistant clones that were already present at baseline.
The airway ecology of cystic fibrosis helps explain why this variation could emerge. Chronic infection creates a compartmentalized environment with persistent inflammation, mucus retention, and repeated antimicrobial exposure. Under those conditions, bacterial lineages can diversify in surface architecture, especially in traits tied to host interaction and phage receptor use. The LPS-linked findings fit that framework well, because small biosynthetic changes can have outsized effects on receptor availability without requiring wholesale rewiring of bacterial defenses.
One practical consequence stands out: phage susceptibility cannot be assumed from a single colony pick when chronic infection is genetically heterogeneous.
What this means for phage therapy design in cystic fibrosis
The preprint argues for a more granular approach to phage selection in chronic Pseudomonas aeruginosa infection. Sampling depth matters. So does awareness that within-patient heterogeneity may arise from standing variation in cell-surface traits rather than from obvious changes in canonical defense systems. If phage therapy is to work in this setting, treatment design may need to account for multiple bacterial subtypes within the same lung, not just the most easily cultured isolate.
The study also places a practical ceiling on a common hope: that phages will automatically steer bacterial populations toward favorable trade-offs, such as increased antibiotic sensitivity. That outcome may happen in some infections, but these data suggest it is not dependable when baseline variation in susceptibility is already broad. Phage therapy in that context may require combination strategies, repeated susceptibility profiling, or phage banks broad enough to cover the receptor diversity present inside a single host.
Because this is a preprint, the findings remain provisional and unreviewed. Even so, the message is strong and clinically legible: chronic cystic fibrosis lung infections can harbor substantial intrapatient diversity in phage susceptibility, and the variation appears tied in part to changes in LPS-related surface biology. For phage treatment to become dependable in this setting, future work will need to solve the heterogeneity problem at the level where it actually exists — within the patient’s own bacterial population.