For decades, phage therapy was a footnote in Western medicine — a curiosity from Soviet-era Georgia, eclipsed by the antibiotic revolution. Now, as antibiotics fail against resistant superbugs, the oldest antimicrobial strategy on Earth is being reborn. Not as it was, but as something far more powerful: CRISPR-enhanced, AI-designed, and clinically validated.
Bacteriophages — viruses that infect and kill bacteria — have been co-evolving with their hosts for billions of years. They are the most abundant biological entities on the planet, outnumbering bacteria ten to one. Every environment harbors them: soil, oceans, the human gut. They are specialists. Each phage typically targets a narrow range of bacterial strains, binding to specific receptors on the cell surface, injecting its genome, hijacking the bacterium's machinery, and ultimately lysing (bursting) the cell to release dozens of progeny phages that repeat the cycle.
This natural precision is exactly what makes them attractive as medicine. Unlike broad-spectrum antibiotics that carpet-bomb the microbiome, phages can be selected to kill only the pathogen. And unlike antibiotics, phages replicate at the site of infection — their numbers increase precisely where they are needed.
But phages are not new. What is new is the convergence of technologies that is transforming them from a niche curiosity into a clinical reality.
The Clinical Vanguard
Two companies are leading the charge with the most advanced clinical programs in phage therapy, and their approaches represent two distinct philosophies: personalized phage banks versus engineered precision cocktails.
BiomX: Phages Against Diabetic Foot Osteomyelitis
BiomX (which absorbed Adaptive Phage Therapeutics in 2024) is running the DANCE trial — a randomized, double-blind, placebo-controlled Phase 2 study of BX211, a phage cocktail targeting Staphylococcus aureus in diabetic foot osteomyelitis (DFO). This is arguably the highest-quality phage therapy trial ever conducted.
The numbers tell the story. At Week 13, BX211 showed statistically significant ulcer size reduction compared to placebo (p = 0.046 at week 12), with the separation from placebo beginning at week 7 and exceeding 40% by week 10. The phages worked equally well against both methicillin-susceptible and methicillin-resistant S. aureus (MRSA) — they don't care about antibiotic resistance mechanisms because they kill through an entirely different pathway. They also destroyed biofilms, the sticky bacterial communities that make DFO so difficult to treat.
BX211 Phase 2 — Week 13 Results
Why does this matter beyond the numbers? Today, 30–40% of DFO cases lead to lower extremity amputations. Those amputations carry a five-year mortality rate of roughly 50%. If BX211 can reduce amputations — the Week 52 readout on amputation rates and osteomyelitis resolution was expected in Q1 2026 and may be imminent — it would represent the most impactful phage therapy result in modern medicine.
The program has been supported by approximately $40 million in non-dilutive funding from the U.S. Defense Health Agency, which recognizes antibiotic-resistant wound infections as a military readiness issue.
Locus Biosciences: CRISPR-Armed Phages
If BiomX represents the power of natural phage therapy refined, Locus Biosciences represents something more radical: phages engineered to be 100 to 1,000 times more lethal than their natural counterparts.
Locus uses CRISPR-Cas3 — a different CRISPR system than the Cas9 most people know. While Cas9 makes precise cuts (useful for gene editing), Cas3 is a processive nuclease: once activated, it chews through DNA like a molecular shredder. Locus engineers this system into their phages so that when the phage injects its genome into a bacterium, the CRISPR-Cas3 machinery activates and systematically destroys the bacterial chromosome, targeting conserved essential genes. The bacterium cannot survive.
Locus Biosciences Pipeline
| Program | Pathogen | Indication | Stage |
|---|---|---|---|
| LBP-EC01 | E. coli | Uncomplicated UTIs | Phase 2 |
| LBP-PA01 | P. aeruginosa | HAP/VAP (pneumonia) | Phase 1b |
| LBP-KP01 | K. pneumoniae | Complicated UTIs | Preclinical |
| LBP-AIEC01 | AIEC (E. coli) | Crohn's disease | Preclinical |
Their lead program, LBP-EC01, targets E. coli urinary tract infections. Part 1 of the ELIMINATE Phase 2 trial was published in The Lancet Infectious Diseases (August 2024), showing the treatment was safe and well-tolerated with no genetic resistance observed. Part 2 is now enrolling 288 patients in a randomized, controlled, blinded study — backed by up to $93 million from BARDA, part of a $152 million program to support through FDA approval.
Perhaps even more exciting is LBP-PA01, an AI-designed phage cocktail targeting Pseudomonas aeruginosa in hospital-acquired and ventilator-associated pneumonia (HAP/VAP). In January 2026, Locus received $3.3 million from NIAID to fund a Phase 1b clinical trial. The cocktail was designed using Locus's AI discovery engine, which experimentally measures millions of phage-bacteria interactions and simulates over a quadrillion potential combinations in silico. The result: a six-phage cocktail covering 97.3% of 570 diverse clinical P. aeruginosa isolates, with demonstrated efficacy against biofilms, in the presence of mucin, and superiority to standard-of-care antibiotics in animal models.
The Evolutionary Chess Match
Every antimicrobial strategy faces the same question: what happens when bacteria resist? With antibiotics, resistance mutations often come at little fitness cost — bacteria can become resistant and stay virulent. The resistance spreads through populations, and we lose another drug.
Phage resistance works differently, and this difference is arguably the most strategically important insight in the field.
Bacteria have evolved over 250 distinct anti-phage defense systems — a staggering arsenal that includes restriction-modification systems, CRISPR-Cas immunity, abortive infection (programmed cell suicide), cyclic nucleotide signaling (CBASS), and many recently discovered systems still being characterized. When a phage attacks, bacteria can mutate or downregulate the surface receptors the phage uses for entry, modify their DNA to evade restriction enzymes, or trigger suicide programs that sacrifice the infected cell to protect the colony.
But here is the crucial insight: many of these resistance mechanisms come at a steep fitness cost.
The Trade-Off Principle
When bacteria mutate to resist phages, they often lose the very mechanisms that made them antibiotic-resistant.
The most striking example involves efflux pumps — protein complexes that bacteria use to pump antibiotics out of the cell. These pumps are a major driver of multidrug resistance. But some phages use these same pumps as entry points. When bacteria mutate to resist these phages, they lose or degrade their efflux pumps — and suddenly become sensitive to antibiotics again.
This phenomenon, called "import-export pleiotropy," was demonstrated clinically with a phage called OMKO1 that targets the MexAB and MexXY-OprM efflux systems of Pseudomonas aeruginosa. Combined with ceftazidime, the phage drove the bacteria into an evolutionary corner: resist the phage and become treatable with antibiotics, or remain susceptible to the phage. Either way, the patient wins.
Researchers call this strategy "phage steering" — using phages not just as killers, but as evolutionary pressure to force bacteria back toward drug sensitivity.
But there is a caveat. Not all phage resistance creates beneficial trade-offs. In some cases, phage resistance can drive what researchers call "trade-ups" — where resistance to the phage coincides with enhanced biofilm formation or even increased antibiotic resistance. This is why rational phage selection matters: you need to choose phages that target receptors whose loss genuinely costs the bacterium something important.
Phage-Antibiotic Synergy
Beyond evolutionary trade-offs, phages and antibiotics can directly enhance each other's killing power through a phenomenon called phage-antibiotic synergy (PAS).
The mechanisms are multiple. Subinhibitory concentrations of certain antibiotics (particularly those that disrupt cell division, like beta-lactams) cause bacteria to elongate and produce more cellular machinery — which phages exploit to produce more progeny per infected cell. Phage-derived lysins (enzymes that degrade the bacterial cell wall) can increase the penetration of antibiotics like daptomycin and vancomycin by exposing their targets. And when phages lyse bacteria within biofilms, they physically open channels for antibiotics to reach bacteria that were previously shielded.
A multicenter cohort study of 100 patients with diverse infections (pulmonary, integumentary, soft tissue, and osteoarticular) demonstrated 70% superior eradication rates with combination phage-antibiotic therapy compared to phage monotherapy.
The evidence is growing, but important nuances remain. Some antibiotic classes — particularly those targeting ribosomes (aminoglycosides, macrolides) — can suppress phage replication by inhibiting the protein synthesis phages depend on for replication. The timing, dose, and sequence of combination therapy all matter. PAS is not a universal phenomenon but a relationship that must be carefully optimized for each pathogen-phage-antibiotic combination.
The Regulatory Frontier
No phage therapy product has received FDA approval. Currently, phage products are classified as biological products under CBER and must follow the standard IND pathway. The FDA has granted expanded access (compassionate use) for phage therapy on a case-by-case basis, allowing >99% of individual applications to proceed.
The regulatory landscape is evolving. In December 2025, the Transatlantic Task Force on AMR (TATFAR) published a landmark perspective in Nature Communications integrating regulatory viewpoints from the FDA, EMA, Health Canada, UK MHRA, and other agencies — the first coordinated international framework for phage therapy regulation. The European Medicines Agency has drafted phage-specific quality guidelines, and the European Pharmacopeia is developing a standardized phage potency assay (chapter 2.7.38, expected 2026).
Countries like Belgium, Georgia, and Poland have already established frameworks for compassionate use of personalized phage therapy. The United States is catching up: the FDA's approval of Adaptive Phage Therapeutics' PhageBank IND was a milestone, and the substantial BARDA and NIAID funding behind Locus Biosciences signals serious institutional commitment.
The most likely path to the first FDA-approved phage product runs through BiomX's BX211 or Locus's LBP-EC01 — both in Phase 2/3 territory with robust institutional backing. Commercial phage products may be available within 2–5 years.
The Bigger Picture
The Phage Therapy Landscape — By the Numbers
What we are witnessing is not just the return of an old idea. It is the emergence of a fundamentally new class of antimicrobial — one that can be programmed (CRISPR-Cas3), designed by artificial intelligence (LBP-PA01's quadrillion-combination optimization), personalized to individual patients (PhageBank), and strategically combined with antibiotics to suppress resistance evolution.
The phage renaissance is being powered by the convergence of three forces:
- Desperation — With the antibiotic pipeline shrinking and resistance accelerating, the medical establishment needs alternatives. The 160,000 amputations per year in diabetic patients alone represent a massive unmet need.
- Technology — CRISPR engineering, AI-driven cocktail design, high-throughput phage discovery platforms, and computational modeling have transformed phage therapy from artisanal medicine into precision biotechnology.
- Funding — Institutional money is finally flowing. BARDA's $152M commitment to Locus, the Defense Health Agency's $40M for BiomX, NIAID's CAPT-CEP program, and CARB-X support for LBP-KP01 signal that governments are betting on phages.
The old objection was always: "Bacteria will just evolve resistance to phages too." They will. But unlike antibiotics, phages co-evolve back. And as we now understand, the evolutionary trade-offs bacteria face when resisting phages can be strategically exploited to make them vulnerable again — to phages, to antibiotics, or to the immune system.
The arms race between bacteria and phages has been running for billions of years. We are just learning to take sides.
Sources: BiomX Phase 2 DANCE trial topline results (March 2025); Locus Biosciences LBP-EC01 ELIMINATE trial, Lancet Infectious Diseases (Aug 2024); Locus Biosciences NIH award for LBP-PA01 (Jan 2026); TATFAR phage therapy perspectives, Nature Communications (Dec 2025); Hasan et al., "Phage Therapy as a Novel Alternative," Antibiotics (Oct 2025); Fujiki et al., "Biocontrol of Phage Resistance in Pseudomonas," PMC (2025); PAS landscape review, Antibiotics (May 2025); Science Advances, "Accumulation of defense systems in phage-resistant P. aeruginosa"; Nature Microbiology, "Personalized bacteriophage therapy outcomes for 100 consecutive cases" (2024).