Personalized phage therapy is moving from rescue idea to repeatable bedside practice
Rising antimicrobial resistance has pushed bacteriophage therapy back into clinical discussion, and this review captures what personalized use looks like in real patients rather than in theory. Across the peer-reviewed English-language literature from January 1, 2005, through August 31, 2025, investigators identified 323 patients across 83 studies with individual patient data available for synthesis. The picture that emerges is not one of a single standardized intervention. It is a field still being built one case at a time, with phage selection, dosing, and combination treatment varying widely from patient to patient.
Most patients received more than one phage, a practical reflection of how quickly bacteria can evade a single viral predator. In 221 of 323 patients, treatment used at least two phages in a cocktail, while 179 patients received bespoke preparations designed for their own infecting organism. Another 144 patients were treated with fixed, off-the-shelf phage preparations. Those two approaches point to a central tension in the field: speed and accessibility on one side, precision and matching on the other.
What the pooled case literature shows about bedside use
The review spans a broad swath of infections rather than one narrow syndrome. Skin and soft tissue infections, pulmonary disease, orthopedic and osteoarticular infections, and genitourinary infections were the most common indications. That distribution fits the basic biology of phage therapy, which depends on reaching the infected compartment, matching the pathogen, and keeping viable viral particles active long enough to exert pressure on the bacteria. In chronic, biofilm-heavy infections, the appeal is obvious. Antibiotics often struggle there because bacterial communities can shelter themselves on hardware, within bone, or inside dense tissue matrices.
Treatment almost always occurred alongside standard antimicrobial therapy. Antibiotics were used concurrently in 246 of 323 patients, reinforcing that phage therapy is being deployed as an adjunct rather than a clean replacement for conventional drugs. The median phage concentration was 10^8 PFU/mL, with an interquartile range of 10^7 to 10^9 PFU/mL, and the median treatment duration was 10 days, with an interquartile range of 5 to 21 days. Those values suggest a field where practice is still heterogeneous, but not directionless. Clinicians are converging on short courses, high-titer preparations, and combination treatment when they can obtain them.
A 10-day median course leaves little room for complacency.
Clinical response and pathogen clearance were common, but not universal
Outcome data were available for the full cohort of 323 patients, and 253 experienced clinical improvement, representing 78% of cases with reported outcomes. Microbiologic eradication of the targeted pathogen was documented in 177 of 290 cases, or 61%, where eradication data were available. Those two endpoints do not always move in lockstep, and the gap between symptomatic improvement and pathogen clearance matters clinically. A patient may feel better while residual bacterial burden persists, especially in infections involving prosthetic material, bone, or established biofilm. In that sense, phage therapy may sometimes function as a bridge to control rather than a complete sterilizing solution.
The data also show that treatment was not always smooth. Nine patients had therapy withdrawn because of tolerability concerns. The review does not frame those events as a dominant safety issue, but even a small withdrawal count is important in a therapy that is often considered for medically fragile patients with few alternatives. Personalized phage therapy is still delivered in a setting where access, manufacturing, susceptibility testing, and clinical urgency all collide. Each of those steps can shape whether a patient receives an optimal preparation or a pragmatic approximation.
Why personalization matters in resistant infection
The biological appeal of phages is straightforward. These viruses are selective for bacteria, replicate at the site of infection when susceptible organisms are present, and can be paired with antibiotics in ways that may help overcome resistance or reduce bacterial density. Personalization matters because phages are not a universal class in the way a broad-spectrum antibiotic is. A preparation effective against one strain may do little against a closely related isolate if receptor usage or resistance patterns differ. Bespoke therapy is therefore more than a branding term; it reflects a laboratory-guided attempt to align viral activity with the patient’s pathogen.
At the same time, fixed preparations have a practical advantage. They are easier to deploy rapidly, which is often essential when an infection is progressing and time for custom manufacturing is limited. The coexistence of these two models in the review indicates how early the field still is. Real-world use is being shaped by logistics as much as by microbiology.
The current evidence base is encouraging, but uneven by design
This literature synthesis is rich in clinical detail and thin in standardization. The cases come from 83 studies, which means the evidence base is broad but fragmented. Indications vary, treatment routes are not harmonized in the abstracted data, and the review necessarily reflects case-based practice rather than randomized comparison. That limits any claim of uniform efficacy. It also explains why the most useful takeaway is descriptive: personalized phage therapy is already being used across multiple infection types, most often with cocktails and concurrent antibiotics, and often with meaningful clinical response.
The review’s strength lies in showing how clinicians have actually deployed the intervention between 2005 and 2025. The field is not confined to isolated anecdotes anymore. It has accumulated enough patient-level experience to reveal recurring patterns: combination phage therapy is common, adjunctive antibiotic therapy remains the norm, and improvement is reported more often than not when outcomes are captured.
Standardization remains the next clinical obstacle
For phage therapy to move beyond individualized rescue use, several practical questions still need answers in routine care. Dosing, route of administration, duration, susceptibility testing, and the definition of response all need greater consistency if outcomes are to be compared across centers. The review also makes clear that the same clinical label can hide very different interventions, from fixed phage products to bespoke preparations assembled for a single isolate. Infections of skin, lung, bone, and genitourinary tract all appear in the same dataset, but their treatment demands are not interchangeable.
For now, the most defensible reading of the evidence is cautious but real. Personalized phage therapy can work, often alongside antibiotics, and the published case literature shows repeated use in infections where conventional options have narrowed. The remaining challenge is not whether phages can be used at the bedside. It is whether the field can turn a patchwork of successful cases into a reproducible therapeutic framework without losing the precision that makes the approach attractive in the first place.