Personalized phage therapy at scale: why cost-effective infrastructure is the missing link
– by José C. Santos and José L. Bila

The views and opinions expressed in this article are solely those of the original author(s) and do not necessarily represent those of GARDP, their donors and partners, or other collaborators and contributors. GARDP is not responsible for the content of external sites.

AMR is outrunning antibiotics

Antimicrobial resistance (AMR) was associated with an estimated 4.95 million deaths worldwide in 2019, including 1.27 million attributed directly to resistant infections.¹ The response cannot rely on new antibiotics alone: World Health Organization (WHO) pipeline analyses, consistent with earlier independent tracking, find it insufficient against priority pathogens, so alternatives and adjuvant therapies are needed.^2,3

Phages offer precision advantages, and precision trade-offs

Bacteriophages (phages) are viruses that infect bacteria; for therapy, strictly lytic phages are preferred, because temperate phages can integrate into the bacterial genome and transfer virulence or resistance genes. The WHO Regional Office for Europe notes that phages can target harmful bacteria without harming human cells or beneficial microbiota and can be combined with antibiotics.^4,5 Reviews add further strengths, including biofilm disruption⁶  and localized amplification at infection sites.⁷

These strengths carry trade-offs. Many phages have narrow host ranges, so clinical use often depends on manually screening collections against a patient’s isolate.⁸ Bacteria and phages are locked in a co-evolutionary “arms race” in which resistance often emerges and so libraries need periodic updating.⁹ Immunity matters too: neutralizing antibodies against phages, documented in compassionate-use cases, may reduce efficacy with repeated exposure and contribute to non-response.^10,31 Activity also falls under conditions common in serious infection—biofilms, intracellular bacteria, low bacterial metabolism—and preclinical evidence remains limited relative to case reports. Many reported successes combine phages with antibiotics, obscuring their independent contribution; even in favorable series, responses occur in about half of patients.³¹

Genomic and geographic diversity disadvantages fixed cocktails – especially for some Gram-negatives

Strain-level diversity directly affects phage susceptibility, because the receptors phages recognize (capsule, lipopolysaccharide, pili) vary widely across strains—most starkly in Gram-negatives. Pan-genome studies quantify this: Escherichia coli has an open pan-genome exceeding 13,000 genes; in Klebsiella pneumoniae most accessory genes are rare (66% in ≤5% of genomes); and Pseudomonas aeruginosa (P. aeruginosa) has a small core genome relative to its pan-genome.^11-13

Phage therapy is still unfamiliar to many clinicians and patients and thus professional education and patient engagement matter too.

Geography compounds this: local lineages differ across hospitals and regions and persist for years, so phages effective in one setting may fail elsewhere unless libraries are locally curated and updated.^14,15 Diversity is not limited to Gram-negatives—Staphylococcus aureus, enterococci, and Mycobacterium vary similarly ^8,10,31 —so reviews favor regional phage banks and surveillance-informed selection over fixed cocktails, regardless of Gram status.^5,16

Evidence snapshot: heterogeneous trials, stronger tailored signals

Modern trial evidence remains mixed. Systematic reviews report few contemporary trials, more consistent safety than efficacy signals, and substantial heterogeneity in endpoints and designs.¹⁷ A European Commission Joint Research Centre report (February 2024) counted 72 phage studies on ClinicalTrials.gov and 28 in the WHO trials registry, with no product centrally authorised in the European Union (one nationally authorised in Czechia and Slovakia) and none licensed in the United States.¹⁸

The cocktail-versus-personalization tension is visible in specific studies. In the PhagoBurn randomized trial for P. aeruginosa burn-wound infections, a standardized cocktail underperformed, hampered by reduced active phage concentration during production and no isolate-matched selection.¹⁹ By contrast, a retrospective analysis of 100 consecutive personalized cases (12 countries; 2008–2022) reported clinical improvement and bacterial eradication in 77.2% and 61.3% of infections, using 26 phages in tailored cocktails—showing both the promise and complexity of personalization.²⁰

Biology works but scaling is hard

Tailored phage therapy becomes routine only if it fits clinical timelines and global access constraints. Diagnostics are the first bottleneck: even where culture and susceptibility testing exist, results typically take ~48 hours versus hours for molecular approaches²¹ , and the Lancet Commission estimated that 47% of the world’s population has little or no access to diagnostics.²² Without rapid pathogen characterization, personalized matching stays misaligned with acute-care decision windows.²³ A pragmatic approach tiers diagnostics to local capacity: rapid identification (culture plus molecular assays); phagograms against a curated local library; and sequencing to accelerate matching and monitor resistance—complementing culture, since viable isolates are still needed for phagograms and manufacturing.

Personalized phage therapy has strong biological promise, but its move to routine AMR care is limited mainly by infrastructure.

Manufacturing, regulation, and cost are the next obstacles. Manual isolate-by-isolate screening is labor-intensive, and tailored production must meet stringent quality controls (identity, purity, potency, stability) and manage endotoxin risks.^4,8,24,25 Implementation reports warn that the flexible small-scale production individualized therapy requires becomes disproportionately expensive under good manufacturing practice (GMP) without the right architecture.^16,18 Regulatory expectations are crystallizing: the European Medicines Agency (EMA) issued veterinary guidance²⁴ and, for human use, a 2023 concept paper³⁰ and a 2025 draft guideline³², while the European Pharmacopoeia chapter (5.31) has been in force since January 2025.²⁵

A scalable blueprint for personalized phage therapy

The goal is not phage discovery alone but an end-to-end clinical workflow—identification, rapid matching, quality-assured access, and outcome capture—making infrastructure the primary innovation frontier: scalable phage–bacterium matching (hybrid wet-lab plus computational triage); distributed, interoperable libraries aligned to local epidemiology; flexible, quality-assured manufacturing; and linked evidence systems (registries, harmonized endpoints, multicenter trials) for continual learning and safe updating.^{5,16,18,25,28} This shifts phage therapy from a reactive search after treatment failure to a preparedness model in which faster matching and standardized potency raise the chance an active phage reaches the patient in time.¹⁹

The response to AMR cannot rely on new antibiotics alone: World Health Organization (WHO) pipeline analyses, consistent with earlier independent tracking, find it insufficient against priority pathogens, so alternatives and adjuvant therapies are needed.

Machine learning is becoming a practical accelerator. Strain-level prediction has been demonstrated for Klebsiella phage–host specificity (ROC AUC up to ~0.82 with laboratory validation) and across hundreds of diverse Escherichia strains from genomic information alone (AUROC 0.86), supporting hybrid computational–experimental workflows that shorten isolate-to-phage time.^26,27 Coordination matters as much as algorithms: the American Society for Microbiology (ASM) Phage Therapy Coordination Network targets fragmented data, unclear regulation, scalable manufacturing, and multicenter trials and registries—the gaps preventing personalization from scaling.²⁸

Equity is essential, because diagnostic and manufacturing bottlenecks are greatest where AMR burden is highest and centralized bespoke production may be unrealistic^1,22,29; scalable models must therefore be modular—usable across diagnostic capacities, adaptable to local epidemiology, and compatible with distributed manufacturing and oversight. Phage therapy is still unfamiliar to many clinicians and patients and thus professional education and patient engagement matter too. Personalized phage therapy has strong biological promise, but its move to routine AMR care is limited mainly by infrastructure: the priority is clinically integrated systems—rapid diagnostics, matching platforms, locally relevant phage resources, cost-effective manufacturing, adaptive regulatory pathways, and coordinated evidence networks.^16,18,28,30 Infrastructure, not discovery alone, must be the next focus.

References

1. World Health Organization. (2023) Antimicrobial resistance. Accessed 18 August 2026.
2. World Health Organization. (2025) Analysis of antibacterial agents in clinical and preclinical development: overview and analysis. Accessed 18 August 2026.
3. The Pew Charitable Trusts. (2021) Tracking the global pipeline of antibiotics in development. Accessed 18 August 2026.
4. WHO Regional Office for Europe. (2025) Bacteriophages and their use in combating antimicrobial resistance. Accessed 18 August 2026.
5. WHO Regional Office for Europe. (2025) Building the evidence for the use of bacteriophage therapy Accessed 18 August 2026.
6. Mayorga-Ramos A, Carrera-Pacheco SE, Barba-Ostria C, Guamán LP. (2024) Bacteriophage-mediated approaches for biofilm control. Front Cell Infect Microbiol. 14:1428637.  
7. Chan BK, Stanley GL, Kortright KE, Vill AC, Modak M, Ott IM, et al. (2025) Personalized inhaled bacteriophage therapy for treatment of multidrug-resistant Pseudomonas aeruginosa in cystic fibrosis. Nat Med. 31(5):1494–1501.  
8. Doud MB, Robertson JM, Strathdee SA. (2025) Optimizing phage therapy with artificial intelligence: a perspective. Front Cell Infect Microbiol. 15:1611857.  
9. Wang Y, Fan H, Tong Y. (2023) Unveil the secret of the bacteria and phage arms race. Int J Mol Sci. 24(5):4363.  
10. Berkson JD, Wate CE, Allen GB, Schubert AM, Dunbar KE, Coryell MP, et al. (2024) Phage-specific immunity impairs efficacy of bacteriophage targeting Vancomycin Resistant Enterococcus in a murine model. Nat Commun. 15(1):2993.  
11. Rasko DA, Rosovitz MJ, Myers GSA, Mongodin EF, Fricke WF, Gajer P, et al. (2008) The pangenome structure of Escherichia coli: comparative genomic analysis of E. coli commensal and pathogenic isolates. J Bacteriol. 190(20):6881–93.  
12. Holt KE, Wertheim H, Zadoks RN, Baker S, Whitehouse CA, Dance D, et al. (2015) Genomic analysis of diversity, population structure, virulence, and antimicrobial resistance in Klebsiella pneumoniae, an urgent threat to public health. Proc Natl Acad Sci U S A. 112(27):E3574–81.  
13. Freschi L, Vincent AT, Jeukens J, Emond-Rheault JG, Kukavica-Ibrulj I, Dupont MJ, et al. (2019) The Pseudomonas aeruginosa pan-genome provides new insights on its population structure, horizontal gene transfer, and pathogenicity. Genome Biol Evol. 11(1):109–20.  
14. Connor TR, Barker CR, Baker KS, Weill FX, Talukder KA, Smith AM, et al. (2015) Species-wide whole genome sequencing reveals historical global spread and recent local persistence in Shigella flexneri. eLife. 4:e07335.  
15. Aranzamendi M, Xanthopoulou K, Sánchez-Urtaza S, Burgwinkel T, Arazo del Pino R, Lucaßen K, et al. (2024) Genomic surveillance uncovers a 10-year persistence of an OXA-24/40 Acinetobacter baumannii clone in a tertiary hospital in Northern Spain. Int J Mol Sci. 25(4):2333.  
16. Moon K, Coxon C, Årdal C, Botgros R, Djebara S, Durno L, et al. (2025) Considerations and perspectives on phage therapy from the transatlantic taskforce on antimicrobial resistance. Nat Commun. 16:10883.  
17. Stacey HJ, De Soir S, Jones JD. (2022) The safety and efficacy of phage therapy: a systematic review of clinical and safety trials. Antibiotics. 11(10):1340.  
18. European Commission, Joint Research Centre; Toussaint B, Muñoz Piñeiro A, Pirnay JP. (2024) Overview and outlook of phage therapy and phage biocontrol. Luxembourg: Publications Office of the European Union. EUR 31738 EN, JRC135367.  
19. Jault P, Leclerc T, Jennes S, Pirnay JP, Que YA, Resch G, et al. (2019) Efficacy and tolerability of a cocktail of bacteriophages to treat burn wounds infected by Pseudomonas aeruginosa (PhagoBurn): a randomised, controlled, double-blind phase 1/2 trial. Lancet Infect Dis. 19(1):35–45.  
20. Pirnay JP, Djebara S, Steurs G, Griselain J, Cochez C, De Soir S, et al. (2024) Personalized bacteriophage therapy outcomes for 100 consecutive cases: a multicentre, multinational, retrospective observational study. Nat Microbiol. 9(6):1434–53.  
21. Peeling RW, Heymann DL, Boeras D. (2024) The value of diagnostics in the fight against antimicrobial resistance. Accessed 18 August 2026.
22. Fleming KA, Horton S, Wilson ML, Atun R, DeStigter K, Flanigan J, et al. (2021) The Lancet Commission on diagnostics: transforming access to diagnostics. Lancet. 398(10315):1997–2050.  
23. Evans L, Rhodes A, Alhazzani W, Antonelli M, Coopersmith CM, French C, et al. (2021) Surviving Sepsis Campaign: international guidelines for management of sepsis and septic shock. Crit Care Med. 2021;49(11):e1063–e1143. 
24. European Medicines Agency. (2023) Guideline on the quality, safety and efficacy of veterinary medicinal products specifically designed for phage therapy. Accessed 7 May 2026
25. European Directorate for the Quality of Medicines & HealthCare. (2024) Phage therapy medicinal products. In: European Pharmacopoeia. General chapter 5.31. Supplement 11.6. Strasbourg: EDQM.  
26. Boeckaerts D, Stock M, Ferriol-González C, Oteo-Iglesias J, Sanjuán R, Domingo-Calap P, et al. (2024) Prediction of Klebsiella phage-host specificity at the strain level. Nat Commun. 15(1):4355.  
27. Gaborieau B, Vaysset H, Tesson F, Charachon I, Dib N, Bernier J, et al. (2024) Prediction of strain level phage-host interactions across the Escherichia genus using only genomic information. Nat Microbiol. 9(11):2847–61.  
28. American Society for Microbiology. (2026) Phage Therapy Coordination Network. Accessed 18 August 2026.  
29. World Health Assembly. (2024) Antimicrobial resistance: accelerating national and global responses — WHO strategic and operational priorities to address drug-resistant bacterial infections in the human health sector, 2025–2035.  
30. European Medicines Agency. (2023) Concept paper on the establishment of a guideline on the development and manufacture of human medicinal products specifically designed for phage therapy. Accessed 18 August 2026
31. Dedrick RM, Smith BE, Cristinziano M, Freeman KG, Jacobs-Sera D, Belessis Y, et al. (2023) Phage therapy of Mycobacterium infections: compassionate use of phages in 20 patients with drug-resistant mycobacterial disease. Clin Infect Dis. 76(1):103–12.  
32. European Medicines Agency. (2025) Draft guideline on quality aspects of phage therapy medicinal products.  Accessed 18 August 2026