Nature Medicine Publishes First Detailed Consensus Guideline for Personalized Phage Therapy
Personalized bacteriophage therapy has reached an important milestone. On September 21, 2026, Nature Medicine published what its authors describe as the first detailed, consensus-based guideline covering the practical use of personalized phage therapy from the identification of a patient’s bacterial pathogen to phage selection, pharmaceutical preparation, clinical administration, monitoring and long-term follow-up. The document moves beyond general recommendations and attempts to define a workable clinical and pharmaceutical framework for treating difficult bacterial infections when conventional options are insufficient.
The consensus statement, entitled A consensus-based guideline for personalized bacteriophage therapy, was developed within the methodological framework of the Association of the Scientific Medical Societies in Germany (AWMF), under the leadership of the German Society for Infectious Diseases. It involved 20 professional societies, patient representatives, regulatory authorities and 18 international experts, ultimately producing more than 60 recommendations and statements covering clinical principles, infrastructure, phage preparation, quality control, administration and future research.
The author group brings together many prominent figures in contemporary phage research and clinical translation, including Silvia Würstle, Jean-Paul Pirnay, Shawna McCallin, Lorenz Leitner, Sebastian Leptihn, Paul E. Turner, Benjamin Chan, Tristan Ferry, Maria J. G. T. Vehreschild and Annika Y. Classen, alongside specialists in infectious diseases, microbiology, pharmacy, biotechnology, regulation, surgery and other disciplines. Institutions represented include Goethe University Frankfurt, the Paul-Ehrlich-Institut, Germany’s Federal Institute for Drugs and Medical Devices, Sciensano, Queen Astrid Military Hospital, the University of Zurich, Charité – Universitätsmedizin Berlin, Yale University and Yale School of Medicine, Hospices Civils de Lyon, the German Center for Infection Research and numerous other European academic and regulatory institutions.
The significance of the publication lies partly in the problem it is trying to solve. Personalized phage therapy does not resemble the development of a conventional drug with one fixed active ingredient administered to thousands of patients. A clinical bacterial isolate may need to be recovered from an individual patient, tested against available phages, matched to one or several active candidates, and potentially retested as the bacterial population evolves. The phage itself must then be propagated, purified, quality-controlled, formulated for an appropriate route of administration and delivered quickly enough to remain clinically relevant. Until now, different centers have often performed these steps using substantially different procedures.
The guideline therefore defines personalized phage therapy primarily as a case-by-case intervention. It states that phages may be considered particularly when standard guideline-based treatment has failed, is unavailable or cannot be applied. The bacterial target should normally be identified by culture, including species identification and conventional antimicrobial susceptibility testing, and phage susceptibility should be confirmed as close to the start of treatment as possible. In polymicrobial infections, the document recommends testing all bacterial species considered responsible for the infection rather than assuming that elimination of a single organism will be sufficient.
The guideline also recognizes that critically ill patients may not always be able to wait for complete phage susceptibility results. In life-threatening situations, the panel considers initiation with an available broad-spectrum phage cocktail before susceptibility is known as a possible option, with subsequent adaptation once microbiological information becomes available. This flexibility is characteristic of the document: rather than attempting to create one universal phage protocol, it proposes a controlled framework within which therapy can remain genuinely personalized.
One of the most consequential sections concerns pharmaceutical preparation. Conventional medicinal products generally follow manufacturing systems based on Good Manufacturing Practice, but producing a new GMP-grade phage for each individual infection can be too slow and expensive for personalized treatment. The guideline discusses magistral preparation as a pragmatic alternative. In the European Union, Article 3(1) of Directive 2001/83/EC provides an exemption for medicinal products prepared in a pharmacy according to a medical prescription for an individual patient. The authors argue that this framework can be used for personalized phage active pharmaceutical ingredients and finished phage therapy medicinal products, or PTMPs.
This should not be interpreted as permission to prepare phages without pharmaceutical controls. The document explicitly states that magistral preparations exempt from industrial GMP must still meet current scientific standards and be produced in controlled environments by qualified personnel. Quality control should address parameters including phage identity and potency, purity, microbiological quality or sterility, pyrogenicity, genomic characteristics, residual bacterial proteins and other process-related contaminants.
Endotoxin control is particularly important when phages are propagated in Gram-negative bacteria. The consensus describes downstream purification methods such as chromatography, endotoxin-binding materials, ultrafiltration and ultracentrifugation, and notes the conventional parenteral limit of approximately 5 endotoxin units per kilogram of body weight per hour, subject to route-specific considerations. Stability must also be experimentally established because phage infectivity can be affected by temperature, pH, ultraviolet exposure, ionic conditions, adsorption to surfaces, shear stress and phage concentration.
The proposed infrastructure extends well beyond a physician and a phage laboratory. The guideline describes an interconnected system involving phage collections or banks, clinical microbiology, a phage laboratory, an appropriate phage preparation facility, a pharmacy, an interdisciplinary clinical phage therapy board and the treating clinic. Importantly, these components do not necessarily have to exist within the same institution. Screening, manufacturing or specialist expertise may be provided externally, potentially making personalized treatment possible for hospitals that do not maintain their own phage production unit.
The guideline provides similar practical detail for the phages themselves. Therapeutic candidates may originate from existing collections, environmental or human samples, or synthetic and genetically engineered sources. The document accepts genetic modification and phage adaptation as potential strategies, including the removal of undesirable lysogenic determinants or adaptation of a phage against an individual patient isolate. Phage seed lots should be clonal and characterized, with information including genomic sequence, known host range and screening for genes associated with antimicrobial resistance, virulence or lysogeny.
Another major departure from a rigid drug-development model is that the route, dose and schedule are explicitly expected to vary between patients. Phages have already been administered topically, intravenously, orally, by inhalation, intravesically, intra-articularly and through other routes. The guideline says treatment should take into account infection location and severity, previous antibiotic and phage treatment, implants or prosthetic material, comorbidities, concomitant medications and overall clinical condition. Phage composition, formulation, timing, frequency and duration may then be adjusted dynamically during treatment.
The panel nevertheless proposes a practical starting range: in general, a phage titer between 10^6 and 10^9 PFU/mL is suggested, with adjustment when losses are expected because of the administration route. The authors explicitly acknowledge that the evidence remains insufficient to define a universally optimal concentration for every infection and route, making this a conditional recommendation rather than a fixed therapeutic rule.
Monitoring is treated as part of therapy rather than an afterthought. The document recommends assessment of vital signs, laboratory markers and microbiology before, during and after treatment. Where relevant, clinicians may also investigate whether phages actually reach the target site and evaluate host immune responses. Phage susceptibility can change during treatment as bacterial resistance evolves, meaning that repeated microbiological testing may become necessary. The document suggests that at least one year of follow-up may be considered and recommends documenting suspected recurrences and contributing cases to centralized phage registries when available.
This emphasis on documentation is particularly important because personalized therapy generates evidence differently from conventional large-scale drug development. The guideline encourages national and international registries, standardized outcome reporting and approaches such as n-of-1 trials. It also identifies artificial intelligence and mathematical modeling as potentially useful tools for predicting phage host range and improving pharmacokinetic and pharmacodynamic optimization.
The authors are careful not to present the guideline as proof that phage therapy is already established as an effective treatment across indications. The clinical evidence remains heterogeneous. Modern studies have generally produced reassuring safety data, but randomized trials have not yet demonstrated consistent efficacy, and previous studies have encountered problems including inadequate phage delivery, stability, dosing, patient selection and phage–bacterium matching. The guideline is therefore primarily a framework for making existing and future clinical use safer, more reproducible and scientifically interpretable.
Indeed, one of the strongest features of the consensus is the amount of space devoted to what remains unknown. The panel calls for better models of biofilm and polymicrobial infections, systematic investigation of phage–antibiotic and phage–phage interactions, research on resistance evolution, studies of interactions between phages and the innate and adaptive immune systems, improved delivery systems, broader-host-range engineered phages and faster methods for phage purification and endotoxin removal. Nearly all of these research priorities received strong consensus.
The immune system receives particular attention because circulating or locally administered phages do not behave like conventional small-molecule antibiotics. Endocytosis, pre-existing antibodies, treatment-induced antibodies and local physiological conditions can alter their persistence and activity. The guideline therefore identifies formulation and delivery — including encapsulation and nanoparticle-based approaches — as major areas for future development rather than treating the phage particle alone as the complete therapeutic product.
The consensus process itself was unusually broad for the field. Recommendations were developed through six specialist working groups and discussed during six formally moderated consensus conferences. Recommendations receiving more than 95% approval were classified as having strong consensus. Conflicts of interest were formally assessed, and individuals with moderate conflicts were excluded from leadership positions; repeating the consensus analysis without their votes did not change the results. The guideline is scheduled for annual review and is formally valid through June 2030.
What this publication does not do is create a new regulatory authorization for phage therapy, legalize treatment automatically in every country or establish a universal standard of care. National laws still apply, and the authors explicitly state that local regulatory adaptation will remain necessary. Nor does the consensus eliminate the need for randomized clinical trials. Its importance is different: it provides clinicians, pharmacists, microbiologists and regulators with a common technical framework for treatments that are already being requested and, in selected settings, already being performed.
For personalized phage therapy, that may represent a significant transition. Much of the field has historically depended on individual expert centers, bespoke procedures and case-by-case regulatory negotiations. A consensus document in Nature Medicine now describes how a patient isolate can move through phage matching, preparation, quality control, formulation, treatment, microbiological monitoring and follow-up within a defined pharmaceutical and clinical system. Whether different countries adopt these recommendations remains to be seen, but the publication creates a reference point against which future personalized phage programs can increasingly be designed, compared and inspected.
Sources :
Würstle S, Lieberknecht-Jouy SC, Düchting A, et al. A consensus-based guideline for personalized bacteriophage therapy. Nature Medicine. Published September 21, 2026.
https://doi.org/10.1038/s41591-026-04654-6
Nature Medicine — official article page: https://www.nature.com/articles/s41591-026-04654-6
.png){kind=logo}
Comments
Post a Comment