A Monopoly on Poop

Dying from diarrhea? Get ready to pay $20,000 for a bag of literal shit

Correction: The $20k pricetag quoted in this piece was an estimate based on word of mouth from industry sources before Rebyota was actually available. As of May 2023, the price is $9487, according to Drugs.com.

This is not the outrageous part.

---

Two weeks ago, on November 30th, the FDA approved Rebyota—the first-ever microbiome “drug”. You’d think I would be over the moon, but the truth is that I’m furious.

I put the word “drug” in scare-quotes because Rebyota isn’t really a drug in the typical sense. It is, for all intents and purposes, a bag of shit. And it is expected to cost somewhere in the neighborhood of $20,000 before insurance.

This isn’t even the outrageous part. The outrageous part is what happened the day before, on November 29th.

Before we get to that, a little context. Rebyota is supposed to treat gut infections caused by Clostridioides difficile, or “C. diff”—basically a lethal case of diarrhea. It’s a pathogen that’s resistant to most antibiotics, which is how it becomes a problem and why it’s so hard to treat. Pretty much the only way to cure it is to burn a person’s entire microbiome to the ground with the heaviest of heavy-duty antibiotics, then import a whole new microbiome from a healthy person to cover over the ashes.

Fortunately, there’s no overall shortage of healthy human gut bacteria; your average person has them, quite literally, out the wazoo: Feces is roughly 60% bacteria by dry weight, and—provided it comes from a healthy person and is handled properly—it typically contains a representative sample of all the species necessary for human health.

So for the last few decades, this has been medicine’s dirtiest secret: when the “superbug” hits and all our best drugs have failed, your best shot at living to see another day is oftentimes a poop transplant (or FMT, for fecal microbiota transplant).

FMT is, to be frank, miraculously effective. People will go from “on death’s door” to “cheerily bustling around, never felt better” within a few hours. Last-resort antibiotics like vancomycin only work in about 60% of refractory C. diff cases, while an antibiotic+FMT regimen has a cure rate of around 80%. While a 20% difference might not sound like a miracle, the numbers put it in perspective: roughly 150,000 people are diagnosed with rCDI every year in the US, and 20% of that is 30,000—on par with annual fatalities from car crashes.

FMT is also unbelievably safe, for what it is. With the amount of emphasis we put on handwashing after using the bathroom for fear that an invisible smear on your hand might end up infecting someone, you’d think feeding a very ill person sixty pills full of shit would be liable to cause problems…but you’d be wrong. 999 times out of 1000, the worst side effects are some temporary gas, bloating, and probably some unpleasant burps. In a recent study put out by the nonprofit stool bank OpenBiome, which drew from their records of over 5,000 procedures, there were a total of 6 noteworthy adverse events (all in seriously immunocompormised people) which were considered “potentially” related to the FMT.

When All You Have is a Hammer…

Still, the FDA has not been happy about this state of affairs, which is understandable. Their job is making sure that when you buy a drug, the bottle contains exactly what the label says, the pills do what they’re supposed to, and if there’s a chance they’ll make your hair fall out, you’re at least given fair warning. Something as inherently variable and messy as FMT is a nightmare for a regulatory agency, especially given the mounting evidence that the microbiome influences your risk for all kinds of chronic diseases, and the potential for it to serve as a vector for emergent pathogens like COVID-19.

This is why, when C. diff started becoming a real problem over twenty years ago and American doctors began performing FMTs in substantial numbers,¹ the FDA’s first response was basically to say “cut that out”.

In May of 2013, they announced that FMT would be regulated like a drug, meaning it could only be performed as part of a registered clinical trial, under the auspices of an Investigational New Drug application—a lengthy, expensive process that involves a mountain of paperwork which only drug companies (and occasionally large institutions like universities or large hospital networks) typically have the funds and legal staff to go through.

This was met by immediate backlash from patients and physicians alike, who rightly pointed out that this decision was tantamount to homicide, in that it would restrict access to a uniquely effective therapy for a deadly disease. In response, the FDA didn’t quite reverse course, but at least pumped the brakes—that summer, they issued a notice saying they intended to exercise “discretion” in enforcing the Investigational New Drug requirement for use of FMT to treat C. diff—but only C. diff.

This “enforcement discretion” status has led to all kinds of complications: insurance won’t cover FMT, for one, meaning patients typically have to pay for the procedure out-of-pocket. Where an approved drug can be prescribed “off-label”, very few doctors are willing to go against the FDA’s edict and perform FMTs for other diseases, despite the ever-growing wealth of evidence suggesting it can help with everything from autism to alcoholism. So FMT has lingered in legal limbo: allowed, but not approved, and only in this narrow set of circumstances until we figure out something better.

You might think that Rebyota, having received the FDA’s stamp of approval, is that “something better”. Unfortunately, the only practical difference between Rebyota and FMT as it’s currently practiced is a 5x difference in price.

A War on Shit

OpenBiome, the nonprofit stool bank mentioned earlier, has been mailing out processed poop for FMTs to doctors all around the country for about a decade now. Their process is practically identical to the one Ferring pharmaceuticals is using to produce Rebyota, which boils down to:

• Find someone exceedingly healthy (like, top 1% kind of healthy; you’re paying them to shit, so you can afford to be picky.)

• Have them shit in a cup. Give them fifty bucks.

• Blend that turd up and skim off the ~40% that isn’t bacteria.  

• Test for various pathogens.

• Add in glycerol or something similar, so the cells don’t die when you freeze it.

• Freeze it. Portion it out (you can get six treatments’ worth from one good dump!)

• Send it where it needs to go, so that it can hopefully save some poor soul’s life.

At scale, this is a complicated process involving lots of donor screening and things like -80°C freezers and cold-chain shipping logistics, so OpenBiome ends up charging around $1-2k per treatment for their product.²

This brings us to the outrageous part. How does Ferring expect to charge $20k for a bag of shit, when OpenBiome is producing the same thing at a tenth of the cost? Many in the field, myself included, were hoping that their plan was to leverage the fact that insurance doesn’t cover FMT—so if you’ve got good insurance, you might pay a $500 copay on a $20k bill, rather than $2k out of pocket.

But, if that was their plan, the FDA did them one better: on November 29th, exactly one day before Rebyota was approved, the agency issued an update to their enforcement discretion policy, which effectively prohibits the use of FMT sourced from stool banks in treating C. diff, except as part of a clinical trial.

They are, in effect, shutting down OpenBiome to give Ferring a monopoly on shit.

The policy leaves room for practitioners and hospitals that are willing to do the whole procedure themselves, i.e. screening donors, collecting, testing, preparing, and administering the treatment. This is typically much more expensive than working with a stool bank, thanks to economies of scale—and if you can’t afford that, or can’t convince your doctor to go through all that, Rebyota is now your one option.

As a justification, the FDA cites “the risks that centralized manufacturing in stool banks presents to individuals receiving such products”, and goes on to state that this decision is about patient safety.

This is a $20,000 crock of shit, as is immediately apparent if you think for more than two seconds about the process, or the carve-outs the policy includes.

Citing the risks of “centralized manufacturing” makes it sound as if stool banks operate by throwing three hundred different donors’ shit into a big cauldron and stirring up a witch’s brew, which might contain pathogens from any of those donors. This would be bad practice, which is why they don’t do it that way. One sample goes to, at most, a few people.

As for the carve-outs: I mentioned earlier that FMT is remarkably safe, but it’s worth talking about the exceptions. Out of the tens of thousands of FMTs that have been performed in the US, there’s been exactly one death attributable to a bad one. It was a perfect storm of bad luck: A new opportunistic pathogen, which they weren’t testing for when the material was prepared, made its way into the bloodstream of a 73-year-old FMT recipient who had practically no immune system, because he was also undergoing a bone marrow transplant for a rare cancer at the same time.

The material for that FMT, however, wasn’t from a stool bank—it was prepared at the hospital where he was treated, meaning the new policy wouldn’t have prevented his death; it would fall under the carve-out for FMTs prepared “in-house”, as well as the one for clinical trials. Remember, this is the only death ever caused by an FMT on record.

The fact that the new policy isn’t about patient safety is something of an open secret. A few days ahead of the decision, I had the chance to speak with Dr. Peter Marks, the head of the FDA’s Center for Biologics Evaluation and Research, which is responsible for regulating FMT. In his words, their aim in crafting the new policy was to “incentivize innovation without creating an access crisis”.

My question is: What innovation?

The Problem With Randomized, Blinded, Placebo-Controlled Trials

There’s a reason that a full ten years have elapsed between the FDA announcing that it considers poop a drug and the first poop-based drug hitting the market. This is actually a pretty normal timeline for taking a drug from “bench to bedside”: the rule of thumb is “ten years and a billion dollars”. You’d think it would be a little shorter and cheaper though, if the drug was already in use at the bedside.

But imagine you’re a gastroenterologist who’s been asked to recruit patients for the Rebyota trial. Your conversation with a patient, who is probably sitting on the toilet and groaning because there’s a very real possibility that they’re going to shit themselves to death this week, goes something like this:

“Well, we can do a poop transplant. It costs a few thousand dollars and insurance won’t cover it, but it’s very safe and it’s got a cure rate in the 80-90% range…or we could enroll you in this trial for a drug which is basically the same thing. I can’t promise that it would work just as well, but I don’t see why it wouldn’t, and that one would be free.”
“So the drug is also just poop?”
“Yes, but there’s a 1 in 3 chance that you’d be sorted into the placebo group, meaning you’d just get more antibiotics and a fake poop transplant. Sham-poo, if you will, aha.”
*indistinct splashing noises, more groaning*
“I’ll take the poop.”
“Wait, there’s an open-label arm, so if you’re still not better two months from now and it turns out you were in the placebo group, you can get the real drug.”
“If I’m still alive.”
“If you’re still alive, that’s correct.”
-end scene-

This is the kind of issue you run into when there’s absolutely no need for your product, because it’s just a more expensive version of something that already exists.

But if, at the end of the day, both products are just shit in a bag (or in a pill; OpenBiome offers a choice) theoretically anyone could have done a series of clinical trials for FMT and submitted a Biologics License Application to the FDA, the way Ferring pharmaceuticals did for Rebyota. So why didn’t OpenBiome?

For one thing: you ostensibly do clinical trials to find out if a treatment is safe and effective—things that were already known for FMT in C. diff. When they invented blood transfusions, nobody asked for a randomized, double-blind, placebo-controlled trial to show that they were safer and more effective than saline in keeping people from bleeding to death, because fucking duh. If someone invents artificial blood, one will be warranted—but you would expect it to be benchmarked against actual blood transfusion, not saline.

If Ferring were trying to improve on industry-standard practices for FMT, it would be one thing, but they’re not. The cure rates in their clinical trials were actually substantially worse than OpenBiome’s average,³ even after some heroic Bayesian statistics to claw an extra half-percentage point of efficacy out of the data.

But ultimately, the reason OpenBiome hasn’t gone through the FDA approval process is that the kind of clinical trials required are hugely expensive. OpenBiome is a nonprofit, so they don’t have the cash reserves to fund them. They could have gotten those cash reserves by charging more for their product. But that likely would have meant fewer people having access to it, which would’ve meant more people dying of C. diff.

Implicit in the structure of the FDA’s regulatory process for drugs is the assumption that the primary motivation of its applicants is profit, which is why the laws have weird clauses like market exclusivity built in. The framework assumes that there will be investors to bankroll the clinical trials, that they’ll set prices to make their money back later, and that all patients will have health insurance to ensure that the pools which pay out these gamblers are sufficiently stocked.

It’s practically cliché to say at this point, but the reason that the FDA is trying to restrict access to FMT in favor of Rebyota is that the modern American healthcare system is built around making profit for corporations rather than getting people healthy—to the extent that the system is no longer capable of accommodating something that simply works. If liver transplants were invented today, they’d be granting market exclusivity to Merck’s “Hepaliiv”, which would still just be the liver of an otherwise-healthy person who died falling off a ladder. Fortunately, FMT isn’t like insulin or other “biologics”—in that, if your loved one needs it and can’t afford it, there’s nothing stopping you from brewing it up at home.

It’s widely said that when the government waged a war on drugs, the drugs won—largely because all it takes to grow pot is some seeds and some dirt. Any attempt to restrict access to FMT is going to face the even-thornier problem that, unlike drugs, you don’t even have to try to produce it.

You’d hope they’d have more sense this time around, but if not…the war on shit is going to be a really embarrassing one to lose.

-🖖🏼💩

1. In reality, FMT has been used in medicine since the 4th century, when Ge Hong provided the recipe for “Yellow Soup” to treat severe digestive diseases. Coincidentally, the same guy noted that a cold-water extract of Artemisia plants is good for malarial fever. It wasn’t until 1972 that Tu Youyo isolated the active compound artemisinin, and it took until 2006 for artemisinins to become the WHO treatment of choice. You could say he was ahead of his time.

2. “Selling shit to people with diarrhea” sounds like a folksy analogy for something exceedingly difficult or requiring lots of smooth-talking, doesn’t it?

3. There are two sides to this. Success rates tend to be lower in blinded, placebo-controlled studies versus open-label retrospectives for a variety of reasons, so it’s possible that Rebyota will be more successful than the trial suggests once it’s on the market. On the other hand, drug companies craft the enrollment criteria for a clinical trial based on what’s likely to give them the best odds of success—meaning that the sickest and most desperate people are typically not allowed to participate.