FECAL MATTER CONTINUES TO SHOW EXTRAORDINARY PROMISE as a treatment for microbial imbalances. In the first randomized controlled trial, in 2013, a fecal microbiota transplant (FMT) from a healthy donor worked so effectively to counter Clostridium difficile—a drug-resistant and often deadly bacterium that thrives in a depleted microbiome—that the study was halted midway. FMT has also shown promise in treating Crohn’s disease and multiple sclerosis and in fighting other multi-drug-resistant bacteria.

But with FMT’s growing acceptance as a treatment comes a curious question: What is it? Donated fecal matter, rich in microorganisms that live in the human gut, isn’t easily classified under current regulatory categories. Is it a medicine? A kind of tissue? Or something else entirely? A designation governs rules for its safety and effectiveness, so the topic is both pressing and hotly debated.

In 2013, the Food and Drug Administration classified FMT as a biological product and drug. But drugs require consistency from batch to batch and dose to dose, a virtual impossibility for a substance that contains a microbial colony, explains Diane Hoffmann, director of the Law and Health Care Program at the University of Maryland Francis King Carey School of Law. The drug designation also requires that physicians file an investigational new drug (IND) application for each use outside of a clinical trial, a hurdle that many thought was unnecessarily cumbersome for a material so ubiquitous to the human experience.

Physicians and patients pushed back, and later that year the FDA announced that it would exercise “enforcement discretion”—a classification that permits more leeway as the particulars of a new treatment are worked out. In this case, the FDA would overlook the IND requirement for treating cases of recurrent C. difficile infection (RCDI). In 2014, it clarified that enforcement discretion only applies if the donated stool comes from a source known to the patient or the physician, and the physician must oversee the screening process.

This requirement posed another challenge, particularly in rural areas. Finding appropriate stool donors isn’t as easy as it might seem. A good donor must be in excellent health to reduce the chance of passing on infections, and should be screened for an ever-expanding list of conditions, as varied as obesity and depression, that seem to be tied to the many organisms that call the microbiome home.

Screening donors on an individual basis is not only expensive—it is also not for the squeamish. Carolyn Edelstein is the executive director of OpenBiome, the first stool bank, which operates in the Boston area and provides physicians and researchers with prescreened fecal matter. Before stool banks existed, Edelstein says, physicians had “shelves and shelves of blenders” in which, if they got as far as finding and screening a donor and acquiring a sample, they would need to puree the donation so it could be siphoned through a colonoscopy tube. The blenders had to be thrown away afterward, as they could not be properly sterilized. It was not a sideline most doctors found alluring.

The 2014 FDA guidelines provided no role for—or regulation of—such organizations. “Limiting donations to people the patient or physician knows would eliminate the possibility of using a stool bank,” says Hoffmann. The guidance was updated in 2016, allowing physicians to use a hospital stool bank, but even then the stool sample had to be obtained under the direction of the treating physician. That didn’t offer much of an improvement.

In a recent issue of Science, Hoffmann and other policy experts and researchers proposed a three-track FMT regulatory system. In the first track, treating a patient for a C. diff infection using stool from a known donor would be considered the “practice of medicine”—in other words, it wouldn’t be regulated by the FDA. FMT for other conditions would require more oversight, including filing an IND, with a special exemption for some patients in life-threatening situations.

Track two would lay out regulations for stool banks, which would be treated similarly to establishments that provide human cells and tissues. Registration, screening and testing rules would be put into place, and stool banks would need to report outcomes and other data to a national registry. Finally, track three would apply to “modified stool-based products,” such as a new generation of pills. These would be regulated in much the same way as other kinds of experimental drugs.

Elizabeth Hohmann, an infectious disease physician at Massachusetts General Hospital, has been researching and administering FMT since 2012, amid the shifting regulatory landscape. “We need to proceed with as much caution as possible,” she says. While some of the first, typically older patients to require FMT were facing life-threatening conditions, today more patients are younger and expected to live a very long time after the procedure: “Changing the microbiome might affect that patient over decades.”

And does she think the substance in question is closest to a tissue, a drug or another existing category of treatment? “None of the above,” Hohmann says. “The microbiome is different.”