Antibiotics are losing ground to bacteria that shrug them off. Some doctors are turning to a much older idea for the hardest cases: viruses that infect and kill bacteria and leave human cells alone. They are called bacteriophages, or phages for short. They were first used as medicine more than a hundred years ago, then mostly forgotten in the West once penicillin arrived. Here is how they work, why they faded, and what the newest results say about bringing them back.
A virus that only eats bacteria
A phage is a virus that infects bacteria and nothing else. It lands on the outside of a bacterial cell, grips a specific spot on its surface, and injects its own genetic code. That code takes over the cell and turns it into a phage factory. Within a short time the cell bursts open, and dozens of new phages spill out to find the next victim.
Phages are everywhere bacteria are, which means almost everywhere: soil, seawater, rivers, sewage, and your own gut. By one common estimate, there are about ten phages on Earth for every bacterium.
The feature that makes them interesting as medicine is how picky they are. Most phages attack only one species of bacteria, and often only certain strains of it. An antibiotic is more like a broad spray that kills good bacteria along with the bad. A phage is closer to a key cut for one lock. That is a strength, because it leaves the helpful bacteria in your body alone. It is also the biggest headache, as later sections show.
Clear spots in a dish
Two people found phages on their own. The British bacteriologist Frederick Twort described a strange agent that dissolved bacteria in 1915, but he did not follow it up. The French microbiologist Félix d’Hérelle found it again and made it his life’s work.
D’Hérelle was largely self-taught. He had worked on fermentation in Canada, Guatemala, and Mexico, and spent years trying to wipe out locust swarms with bacteria. Along the way he kept seeing small clear spots in his bacterial cultures, as if something invisible were eating holes in them.
In 1917, while studying dysentery patients in Paris, he filtered their stool to remove every bacterium and added what was left to a cloudy culture of dysentery bacteria. He later described the next morning: “the broth culture, which the night before had been very turbid, was perfectly clear: all the bacteria had vanished, they had dissolved away like sugar in water.” He named the invisible agent the bacteriophage, which means bacteria eater. His note was presented to the French Academy of Sciences in September 1917.
From dysentery to cholera and plague
D’Hérelle moved quickly to people. In 1919 he cured an outbreak of a salmonella disease in chickens, and that August he treated his first human patient, a case of dysentery. In 1921, at a children’s hospital in Paris, he reported that young dysentery patients began to recover within about a day of swallowing phages.
He then took phages to the great epidemic diseases of the time. In the mid-1920s he gave phages to plague patients and reported that they recovered. In 1926 the British government in India asked for plague phages to test at the Haffkine Institute in Bombay, and d’Hérelle went there himself to fix problems with how they were being grown. If you want the background on the disease, our guide to the different types of plague explains how bubonic, septicemic, and pneumonic plague differ, and how antibiotics treat them today.
By then d’Hérelle was famous. His work helped inspire Sinclair Lewis’s 1925 novel Arrowsmith, about a young doctor who tests a phage during a plague outbreak. It won the Pulitzer Prize.
D’Hérelle’s best-known numbers came from cholera. During a 1927 outbreak in the Punjab, by his own account, 8 percent of the 74 patients he treated with phages died, compared with 63 percent of 124 patients who did not get them.
But d’Hérelle’s results were hard to repeat. Other labs in Germany, the United States, and Brazil tried his methods and got nothing. Few of his studies had proper comparison groups. Drug companies, including Eli Lilly in the United States, sold phage products in the 1930s and 1940s, and many did not work, probably because the phages were weak, damaged, or simply the wrong ones for the infection. D’Hérelle himself said none of the products on the market could cure anyone. He was nominated for the Nobel Prize several times but never won.
Penicillin wins, and phages move east
Then came antibiotics. Penicillin was purified in the early 1940s, mass-produced by 1944, and on pharmacy shelves in 1945. It was cheap, reliable, and worked against many kinds of bacteria at once. Western doctors had little reason to bother with picky viruses, and phage research there mostly stopped.
The story went differently in the Soviet Union. In 1923 a Georgian microbiologist named George Eliava, who had met d’Hérelle at the Pasteur Institute in Paris, founded an institute in Tbilisi devoted to phages. D’Hérelle worked there in the mid-1930s and started building a house on the grounds. In 1937 Eliava was arrested and executed in Stalin’s purges, and d’Hérelle never returned.
The institute survived and still bears Eliava’s name. Soviet doctors used phages to treat soldiers with dysentery and gangrene during World War II. Phage mixtures, called cocktails, are still sold in pharmacies in Georgia and Russia. In Poland, the Hirszfeld Institute in Wrocław has run a phage therapy unit since 2005. Much of that research was published only in Russian, Georgian, or Polish, so Western medicine barely noticed it for decades.
Why phages are back
Bacteria evolve. Every time an antibiotic is used, the few bacteria that happen to survive it can multiply and pass that toughness on. Over decades, that has produced superbugs that resist most or all available drugs. A large study in The Lancet estimated that drug-resistant bacterial infections directly caused about 1.27 million deaths worldwide in 2019. New antibiotics are slow and expensive to develop.
Phages have one trick antibiotics do not. They are alive, in a sense, so they can evolve too. If bacteria change, scientists can hunt for a new phage, or grow an existing one on the patient’s own bacteria until it gets better at killing them.
Sometimes the bacteria lose either way. A phage often grips a structure the bacterium needs, such as a pump it uses to push antibiotics back out of the cell. Bacteria that change that structure to dodge the phage can end up weaker, or vulnerable to antibiotics again.
Rescue cases that made headlines
Two cases did a lot to revive interest in the West. In 2016, a 68-year-old American man with diabetes lay critically ill at the University of California San Diego. He had picked up a drug-resistant strain of Acinetobacter baumannii, a bacterium that thrives in hospitals, during a severe bout of pancreatitis that began on a trip to Egypt. After four months, no antibiotic was working. His wife, the epidemiologist Steffanie Strathdee, went searching for phages. Labs at Texas A&M University and the U.S. Navy found nine that could kill his strain. Doctors gave them through his veins and straight into the infected pockets in his abdomen. His condition turned around, the infection cleared, and he went home. The man, Tom Patterson, later wrote a book about it with his wife.
In 2019, doctors at Great Ormond Street Hospital in London reported treating a British teenager named Isabelle Carnell-Holdaway. She had cystic fibrosis, and after a double lung transplant a relative of the tuberculosis germ, called Mycobacterium abscessus, spread through her body. Doctors gave her less than a 1 percent chance of survival. Graham Hatfull, a biologist at the University of Pittsburgh, searched a collection of about 15,000 phages, many found by students in a teaching program, and picked three. His team genetically modified two of them to make them better killers. It was the first reported treatment of a patient with engineered phages. Twice-daily infusions brought the infection under control, though they did not wipe it out completely.
What 100 patients showed
Single rescues make good stories but weak evidence, since unsuccessful attempts rarely get written up. In 2024, a team from the Queen Astrid Military Hospital in Brussels and partners published something closer to a fair picture. They reported on every one of their first 100 personalized phage treatments, given between 2008 and 2022 in 35 hospitals across 12 countries. Most patients had infections that standard antibiotics had already failed to clear, in the lungs, skin, or bone.
Doctors reported that 77 percent of the infections improved, and the targeted bacteria were wiped out in 61 percent. Combining phages with antibiotics mattered a lot. Without antibiotics, clearing the bacteria was about 70 percent less likely. There were seven mild side effects that may have been linked to the phages, and all of them went away.
The authors were frank about the limits. There was no comparison group and no blinding, and the doctors judged the results themselves. The study also showed how hard this is to scale. Out of 1,066 requests for phage treatment, only about one in ten led to actual treatment, often because the hospital had no phage that matched the patient’s bacteria.
The catch
The tighter tests have been disappointing so far. A European trial called PhagoBurn, which tested a fixed phage cocktail on infected burn wounds, was stopped early because the cocktail was not working well enough. Researchers think one-size-fits-all mixtures may fail because they are not matched to each patient’s strain.
That leads to the core problem. Drug rules in the United States and Europe were built for products made the same way in huge batches. Personalized phages are closer to a custom prescription. So far, no ready-made phage medicine has been approved for sale in the United States or the European Union. American doctors treat patients one at a time, under emergency permission from the Food and Drug Administration. Belgium found a workaround in 2019, letting hospital pharmacies prepare phages for named patients under its rules for custom-made medicines.
Bacteria can also become resistant to phages, sometimes within days, which is why doctors use mixtures and keep backups ready. And the body’s immune system can learn to neutralize phages during longer treatments.
More than 60 phage studies were registered as planned, active, or finished on ClinicalTrials.gov as of early 2025. Some test engineered phages, including ones armed with the gene-editing tool CRISPR, for common problems like urinary tract infections. If those trials go well, phages may end up as a regular backup to antibiotics rather than a last-ditch rescue. A century after d’Hérelle watched a cloudy flask turn clear, that would be a quiet kind of comeback.
Further reading
The Perfect Predator: A Scientist’s Race to Save Her Husband from a Deadly Superbug — Epidemiologist Steffanie Strathdee and Tom Patterson tell the story of the 2016 San Diego rescue from the inside, from the search for matching phages to the first infusions.
The Good Virus: The Amazing Story and Forgotten Promise of the Phage — Science journalist Tom Ireland’s readable history of phages, from d’Hérelle and the Eliava Institute to the labs trying to turn them into modern medicine.