BY last August, my 1-year-old son had taken five courses of antibiotics for recurrent ear infections. That was alarming. By age 10, the average American child has had about 10 courses, and some microbiologists argue that even one course a year is too many — that it might damage our native microbial ecosystem, with far-reaching consequences.
My son was off to a worrisome start. Why, I wondered, didn’t doctors work harder to prevent this collateral damage, not with store-bought probiotics, but with “microbial restoration”? Why didn’t we reinfuse patients with their own microbes after antibiotics?
The scientific term for this is “autologous fecal transplant.” In theory, it could work like a system reboot disk works for your computer. You’d freeze your feces, which are roughly half microbes, and when your microbiome became corrupted or was depleted with antimicrobials, you could “reinstall” it from a backup copy.
That damage from antibiotics may not be trivial. Studies have linked antibiotic use early in life with a modestly increased risk of asthma, inflammatory bowel disease, obesity and rheumatoid arthritis. These are associations, of course; they don’t prove that antibiotics cause disease.
Many microbiologists, however, take the possibility seriously that antibiotics could contribute to the development of those diseases. That’s because, in animal studies, depleting certain microbes early in life — microbes that may promote gut health and soothe the immune system — makes rodents more susceptible to inflammatory disease later.
The “self-transplant” isn’t a new idea. In the late 1950s, a medical technologist named Stanley Falkow practiced what he called “fecal reconstitution.” Gut troubles often plagued surgery patients during recovery. They’d received antibiotics prophylactically, depleting their native gut microbes. So Mr. Falkow, working with an internist, began giving these patients capsules containing their own feces, collected and frozen before treatment. It helped tremendously. But when the hospital administrator found out — patients didn’t know what they were swallowing — he fired Mr. Falkow. (Mr. Falkow, now Dr. Falkow, an emeritus microbiology professor at Stanford, was rehired soon thereafter, but had to abandon the project.)
Almost 60 years later, the “fecal transplant” is a cutting-edge treatment for the pathogen Clostridium difficile, a bug that kills 29,000 yearly and infects nearly half a million. “C. diff” tends to strike after antibiotics deplete the microbes that naturally inhabit the gut, leaving us vulnerable to invasion. So far, fecal transplants seem to be more than 90 percent effective at curing these infections.
As currently practiced, however, the transplant material usually comes from someone else. Even with careful screening, that presents some risk. It’s theoretically safer to receive one’s own microbes. North York General Hospital in Toronto recently completed a pilot study banking incoming patients’ own stool. Should any of these patients develop infections after antibiotics, their own microbes were on hand for reconstitution.
None fell sick in this case, so the transplants weren’t needed. But the project proved feasibility, and achieved a processing time — gathering, blending and freezing — of less than one hour.
Memorial Sloan Kettering Cancer Center in New York has also started a proactive stool-banking study. Most of the subjects are patients with leukemia. Before stem cell transplants, patients receive antibiotics and chemotherapy, often wiping out their microbiota.
Dr. Eric Pamer, a physician and scientist at Memorial Sloan Kettering, has discovered that the diversity of the microbiota just after the stem cell transplant predicts well-being and survival. After excluding death from leukemia recurrence, those with the least diverse microbiomes after surgery were five times less likely to remain alive three years later, when compared with those with the most diverse.
“We often wipe out the flora,” Dr. Pamer told me. “It should become a routine part of practice to restore the flora.” It’s “very reasonable,” he added, to apply this principle to everyone who receives broad-spectrum antibiotics, not just leukemia patients, although the idea needs testing.
In mice, simply caging an antibiotic-treated animal with a nontreated one will restore its microbiota. Mice are “coprophagic,” though. They eat one another’s feces. People don’t — or so we like to think. In reality, for much of our evolution, we shared more of our microbes. This pre-sanitary past was disease-ridden, but it may also have enabled acquisition of health-promoting microbes. One recent study in Papua New Guinea found that, in an environment with fewer sanitary amenities, people’s microbes seemed to recover naturally from antibiotics.
Shouldn’t we have a more precise microbiota-recovery protocol? I called up Mark Smith, founder of a nonprofit stool-banking organization called OpenBiome. Years ago, as a doctoral student at M.I.T., he’d watched a friend spend a year and a half sickened by C. difficile. Antibiotics failed to help and, at wits’ end, the friend finally gave himself a stool transplant, in his own apartment, with stool donated from his roommate. It worked.
Dr. Smith couldn’t understand why this highly effective treatment for a widely recognized scourge was relegated to D.I.Y. Lack of banked, screened fecal material, he discovered, was one hurdle. So he started OpenBiome, which screens donors and banks stool for use by medical professionals. Business is booming, he told me. This month, he started a pilot self-banking program called “PersonalBiome.” One complication: If he stores your stool, you can generally withdraw it only to treat C. difficile, not for preventive “reconstitution.” That’s because stool is regulated as a drug and not, as with embryos or blood, a tissue, which makes its use more complex.
Ultimately, my son got ear tubes, tiny chalice-shaped pipes of plastic through his eardrums that, by allowing drainage and aeration of the middle ear, prevent ear infections. The tubes have helped with the immediate problem. But I still fret about what we may have done to his microbes.
I didn’t give my son a self-transplant. The approach is unproven, and anyway, I hadn’t stored any of his microbes from his pre-antibiotic days. Even if I had, as the Stanford microbiologists Justin and Erica Sonnenburg reminded me, self-transplants aren’t as straightforward for children as they are for adults. The infant microbiota changes from month to month, moving erratically toward an adultlike state. If I were to store microbes from month six, and reintroduce them at month 12, the microbial community might no longer be appropriate for that stage of development.
But clearly, this idea needs testing. It may be one way to prevent infections and some percentage of lifelong, chronic diseases, not with fancy drugs or expensive procedures, but simply by restoring, after antimicrobial disturbances, the microbes we already carry.
An earlier version of this article imprecisely described the findings of a study. In a study of subjects who underwent stem cell transplants, those with the least diverse microbiomes after surgery were five times less likely to remain alive three years later, when compared with those with the most diverse microbiomes, when you exclude those who died from cancer relapse. The original article failed to specify that this finding excluded subjects who died of cancer and other non-transplant-related causes, implying that it was true for all subjects. When you include all causes of death, those who had less diverse microbiomes were two to three times more likely to die, not five.