FDA's pathway for regulation of FMT: not so fraught

A recent article by Rachel Sachs and Carolyn Edelstein in the Journal of Law and the Biosciences¹ on regulation of fecal microbiota transplants (FMT) brought a welcome hard look at the difficulty of regulating an area that holds great promise but is often greeted by hoots and nervous laughter rather than comprehensive study. FMT has been given relatively little attention by legal scholars working in food and drug law. However, FMT and the broader field of the human microbiome² have enormous potential for the development of new therapeutics.³ Already, FMT has been shown to have greater efficacy for the treatment of recurrent Clostridium difficile infections (CDI) than traditional treatment with a broad spectrum antibiotic, vancomycin.⁴ Interventions that manipulate the human microbiota may also hold potential for the treatment of Inflammatory Bowel Disease⁵ and potentially a broader range of inflammatory, infectious, and metabolic diseases.

Sachs and Edelstein take issue with FDA's current approach to regulating FMT. FDA has announced that fecal microbiota will be regulated as a biological drug,⁶ but has opted to exercise enforcement discretion for FMT used for recurrent or fulminant CDI which fail to respond to standard therapy.⁷ While Sachs and Edelstein welcome FDA's decision to exercise enforcement discretion at this time, they note that ‘Responding to an application from a company called Rebiotix, FDA has recently designated “fecal microbiota” as an orphan product for the treatment of recurrent CDI.’ Further, they reason that, if Rebiotix clinical trials are successful, the Orphan Drug Act would prevent FDA from approving another manufacturer's application to process and distribute stool for FMT, barring a showing of clinical superiority, which could force stool banks out of the market. They liken the situation to the 2009 FDA approval of colchicine⁸ which gave its manufacturer market exclusivity for indications that had previously been treated off-label with cheap generics for decades prior.

Our view is that the above-mentioned FDA Orphan Drug designation would not apply to FMT at large, but specifically to Rebiotix's approach to process and distribute stool material for FMT, which presumably will follow a process within limitations reasonably agreed to by the sponsor and FDA. In support of this view, a second company, Seres Therapeutics, is also entering this market with their own approach to process and distribute stool material for FMT, and has also received Orphan Drug Designation for the same indication.⁹ The Orphan Drug Designation for colchicine applied to a well-characterized product with a defined composition. This is not the case for stool material for FMT, which is a minimally characterized product with an undefined composition, perhaps best described as a ‘process’. Each lot obtained from a different donor has a different composition, and even different lots obtained from the same donor on different days will have different compositions, making it difficult to conceive a ‘generic’ for FMT. Our view is that the commercial sponsors and FDA will agree on ways to produce their stool material for FMT by a certain, narrowly defined process, and that the Orphan Designation will apply to product obtained by that agreed upon process. This should leave room for FDA to continue to exercise enforcement discretion for FMT procedures conducted under an Investigational New Device Application (IND), which would give hospitals and OpenBiome, respectively, the ability to continue to perform FMT and distribute stool material for FMT. In this scenario, monopoly pricing of stool material for FMT by a commercial entity should be less of a concern. A non-profit or hospital-run stool bank operating under enforcement discretion might be an attractive alternative for payers if commercial competitors do not demonstrate clear benefits of their (presumably higher priced) stool material for FMT.

We agree with Sachs and Edelstein that blood and fecal microbiota share characteristics that distinguish them from traditional drugs and that FDA oversight should focus on controls for transmissible disease. There are a number of pathways FDA could take to regulate stool for FMT. At first blush, these treatments seem to fit squarely within FDA's definition of Live Biotherapeutic Products (LBPs).¹⁰ They involve biological products that (i) contain live organisms, such as bacteria; (ii) are applicable to the prevention, treatment, or cure of a disease or condition of human beings; (iii) are not a vaccine. However, as the authors note, serious challenges exist to characterizing the active ingredients and guaranteeing quality and consistency across batches given the variability in composition of human stool. The manufacturing technologies to produce biological products of uniform quality in use today, such as those applicable to monoclonal antibodies, would be insufficient to produce stool material for FMT at similar standards.

We also agree with the authors that regulation of stool should be tied to the process of preparation, rather than its variable contents. Future guidelines within the biological product pathway could lay out the requirements for key steps in the FMT process (eg patient screening, harvesting, preparation, and distribution of stool). However, the authors’ call for exploring alternative regulatory schemes also has merit.

FDA's regulation of blood transfusions with donor screening protocols and testing for transmissible disease provides another potential model pathway. Arguably, there could be less risk involved in FMT because there appears to be no need for immunological screening. But as Sachs and Edelstein note, that pathway was developed in historical contexts that are no longer present.

Like Sachs and Edelstein, we believe that HCT/P model developed for tissue and cell therapies¹¹ could be a reasonable pathway. Of course, this does not mean that these products can simply be integrated into the HCT/P rubric. As the authors note, FDA's Tissue Reference Group has recommended against regulating FTs as HCT/Ps: ‘Microbiota isolated from fecal matter of a donor is not an HCT/P, as defined under 21 CFR 1271.3(d)’¹². That determination probably makes sense; the cells in our microbiota are not human cells and they behave very differently—and the indications for such therapies are also very different. Sachs and Edelstein are pessimistic about FDA's potential for rulemaking in this field. However, FDA has sufficient statutory authority to regulate in this area. The statutory foundation of the HCT/P regulatory scheme is likely directly applicable to microbiota products and the tiered risk-based approach used for HCT/Ps fits the likely markets for these products. Because of the potential for contamination and transmissible disease, all such products should be subject to at least some controls. It is debatable whether the full process of regulatory approval for a new biological drug is appropriate for FMT; the HCT/P regulatory approach allows for more flexibility than an ‘all-or-nothing approach’. Indeed, one might argue that this is exactly what FDA is already doing without issuing formal procedural rules. By exercising enforcement discretion for the use of stool banks for CDI, it is lightly regulating the areas that it deems safe and reasonably effective, but requiring more where evidence of safety is still lacking. However, clear guidance would probably benefit all stakeholders. It would assure that stool banks operate in a safe manner, but avoid costly regulation, while at the same time preserving pre-approval for products with which FDA has less experience and where safety and efficacy are less clear.

If FDA were to adopt an HCT/P model for FMT products, all such products would be subject to Section 361 of the Public Health Service Act which gives FDA authority to regulate to prevent the transmission of communicable disease.¹³ FDA's authority to protect against the transmission of communicable disease under Section 361 is very broad¹⁴ and it seems axiomatic that that authority is directly applicable to FMT since the major risks of such treatments involve potential contaminants and transmission of disease. This would make stool banks subject to requirements similar to those currently imposed on tissue banks. They would be subject to registration and the equivalent of Current Good Tissue Practice (CGTP) requirements that would cover stool and other microbiota source recovery, donor screening and testing, processing controls, equipment and facility requirements, environmental controls, storage conditions, and require stool banks to meet labeling requirements.¹⁵ However, such products would not require premarket approval. At the other end of the tiered risk-based approach, stool that was extensively manipulated, isolated or combined with other components, or used for novel functions, would be regulated under Section 351 of the PHSA¹⁶ and the FDCA as biological drugs or devices requiring premarket approval.¹⁷

Since FDA already has statutory authority to pursue such a pathway, FDA could do this through guidance, which is currently its preferred method of rulemaking,¹⁸ or it could engage in so-called ‘informal rulemaking’ as it did with HCT/Ps, and issue formal regulations. There are a number of complexities that would need to be ironed out if the HCT/P paradigm were adopted. The first is definitional. It is not immediately clear which therapies should fit under the scheme and the overlap with the definition of LBPs would add additional confusion. In addition, some of the ambiguities that have plagued stem cell therapy under the HCT/P rubric might also exist here. With HCT/Ps, FDA requires that cells or tissue that is not used for homologous use or is ‘manipulated in a way that changes the biological characteristics of the cell population (eg by expansion, selection, encapsulation, activation or genetic modifications as part of gene therapy)’ requires premarket approval.¹⁹ The ‘more than minimally manipulated’ standard has been notoriously difficult to use, and arguably FDA has not used that standard with the flexibility that was anticipated when the HCT/P rubric was first developed.²⁰ The difficulty of using that standard may also be problematic in this area. For example, it is not clear what kind or how much manipulation of stool material would constitute a change in biological characteristics to the degree that FDA would require that a stool therapy be subject to Section 351 controls which would make it subject to the full new biological drug requirements.

Despite the obvious complexities of regulating FMT, we are perhaps a bit more optimistic than the authors about FDA's ability to find a reasonable regulatory scheme. It is at least clear that when FDA has been approached with questions about FMT, it has worked with clinicians and industry to develop solutions, as evidenced by its decision not to enforce the IND²¹ requirement for recurrent CDI. FDA likely would prefer to gain additional experience with the potential technologies before it wades into more formal rulemaking. When it is ready to do so, however, it has statutory authority to develop a regulatory pathway that both protects safety and has enough flexibility to support innovation.