Ensuring the safe and effective FDA regulation of fecal microbiota transplantation

Abstract

Scientists, policymakers, and medical professionals alike have become increasingly worried about the rise of antibiotic resistance,
and the growing number of infections due to bacteria like Clostridium difficile, which cause a significant number of deaths and are imposing increasing costs on our health care system. However, in the
last few years, fecal microbiota transplantation (FMT), the transplantation of stool from a healthy donor into the bowel of
a patient, has emerged as a startlingly effective means to treat recurrent C. difficile infections. At present, the FDA is proposing to regulate FMT as a biologic drug. However, this proposed classification is
both underregulatory and overregulatory. The FDA's primary goal is to ensure that patients have access to safe, effective
treatments—and as such they should regulate some aspects of FMT more stringently than they propose to, and others less so.
This essay will examine the nature of the regulatory challenges the FDA will face in deciding to regulate FMT as a biologic
drug, and will then evaluate available policy alternatives for the FDA to pursue, ultimately concluding that the FDA ought
to consider adopting a hybrid regulatory model as it has done in the case of cord blood.

Full text

Journal of Law and the Biosciences, 396–415
doi:10.1093/jlb/lsv032
Original Article
Advance Access Publication 6 July 2015
Ensuring the safe and eective FDA
regulation of fecal microbiota
transplantation
Rachel E. Sachs
1,†
and Carolyn A. Edelstein
2,∗,‡
1. Petrie-Flom Center, Harvard Law School, 23 Evere Street, Room 325, Cambridge, MA 02138, USA
2. Global Partnerships, OpenBiome, Medford, MA, USA
∗
Corresponding author. E-mail. rsachs@law.harvard.edu
ABSTRACT
Scientists, policymakers, and medical professionals alike have become in-
creasingly worried about the rise of antibiotic resistance, and the growing
number of infections due to bacteria like Clostridium dicile, which cause
a signicant number of deaths and are imposing increasing costs on our
health care system. However, in the last few years, fecal microbiota trans-
plantation (FMT), the transplantation of stool from a healthy donor into
the bowel of a patient, has emerged as a startlingly eective means to treat
recurrent C. dicile infections. At present, the FDA is proposing to regulate
FMT as a biologic drug. However, this proposed classication is both un-
derregulatory and overregulatory. e FDA’s primary goal is to ensure that
patients have access to safe, eective treatments—and as such they should
regulate some aspects of FMT more stringently than they propose to, and
others lessso. is essay will examine the nature of the regulatory challenges
the FDA will face in deciding to regulate FMT as a biologic drug, and will
then evaluate available policy alternatives for the FDA to pursue, ultimately
concluding that the FDA ought to consider adopting a hybrid regulatory
model as it has done in the case of cord blood.
KEYWORDS: antibiotic resistance, FDA, fecal transplantation, micro-
biome
†
Rachel E. Sachs, JD, MPH, is an Academic Fellow at the Petrie-Flom Center for Health Law Policy, Biotech-
nology, and Bioethics at Harvard Law School.
‡
Carolyn A. Edelstein, MPA, is the Director of Global Partnerships at OpenBiome, a public, non-prot stool
bank.
C

e Author 2015. Published by Oxford University Press on behalf of Duke University School of Law,
Harvard Law School, Oxford University Press, and Stanford Law School. is is an Open Access arti-
cle distributed under the terms of the Creative Commons Aribution-NonCommercial-NoDerivs licence
(hp://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial reproduction and distri-
bution of the work, in any medium, provided the original work is not altered or transformed in any way, and that
the work is properly cited. For commercial re-use, please contact journals.permissions@oup.com
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
r
397
Scientists, policymakers, and medical professionals alike have become increasingly
worried about the rise of antibiotic resistance, particularly among hospital-acquired in-
fections. e number of infections due to the most common nosocomial pathogen,
1
Clostridium dicile, doubled between 2000 and 2005.
2
is aggressive intestinal su-
perbug causes more than 250,000 hospitalizations and 14,000 deaths in the USA each
year,
3
and it is estimated to cost the US health care system $4.8 billion annually.
4
Although the rst line of treatment for C. dicile remains antibiotic therapy, 25 per
cent of patientsexperience a recurrence aer their initial course, a number that has risen
in the last decade.
5
Failure rates aer treatment with metronidazole, which is typically
used to treat new cases of C. dicile, rose from 2.5 per cent to more than 18 per cent
between 2000 and 2008.
6
Failure rates aer treatments with vancomycin or dax-
omicin, standardtherapies for recurrences ofC. dicile infection, havealso risen, posing
a signicant challenge for patients with multiple recurrences.
7
Should a patient experi-
ence a second recurrence, his or her chanceof another relapserises to 35or 45 percent,
8
and reaches 60 per cent by the third recurrence.
9
e limited eectiveness of antibiotic
therapies for the treatment of recurrent C. dicile infection has prompted growing con-
cern,
10
asreected in the White House’snewly releasedNational ActionPlan to combat
antibiotic resistant bacteria, and which prioritizes addressing the C. dicile problem.
11
In the last few years, however, fecal microbiota transplantation (FMT) has emerged
as an eective alternative means to treat recurrent C. dicile infections.
12
FMT, also
called fecal bacteriotherapy, is the transfer of stool from a healthy donor into the bowel
of the patient. e cure rate for those who suer two or more recurrences of C. di-
cile infection and receive a fecal transplant is 90 per cent—dramatically above the 30–
40 per cent chance they face with antibiotic therapies.
13
1
Shelley S. Magill et al., Multistate Point-Prevalence Survey of Health Care-Associated Infections, 370 NEW ENG.J.
M
ED. 1198, 1198 (2014).
2
ere are now 11.2 cases of C. dicile infection per 10,000 inpatient hospitalizations. Marya D. Zilberberg,
Andrew F. Shorr & Marin H. Kollef, Increase in Adult Clostridium Dicile–Related Hospitalizations and Case-
Fatality Rate, United States, 2000–2005,14E
MERG.INFECT.DIS. 929, 929 (2008).
3
e White House, National Action Plan for Combating Antibiotic-Resistant Bacteria, 60, Mar. 2015,
hps://www.whitehouse.gov/sites/default/les/docs/national
action plan for combating antibotic-
resistant
bacteria.pdf (accessed June 18, 2015).
4
Erik R. Dubberke & Margaret A. Olsen, Burden of Clostridium Dicile on the Healthcare System,55CLIN.IN-
FECT.DIS. S88, S88 (2012).
5
Ciaran P. Kelly & J. omas LaMont, Clostridium Dicile—More Dicult an Ever, 359 NEW ENG.J.MED.
1932, 1936 (2008).
6
Id. at 1935.
7
Laurica A. Petrella et al., Decreased Cure and Increased Recurrence Rates for Clostridium Dicile Infection Caused
by the Epidemic C. Dicile BI Strain,55C
LIN.INFECT.DIS. 351, 352 (2012).
8
Kelly & LaMont, supra note 5, at 1936.
9
Id.; see also Gauree G. Konijeti et al., Cost-Eectiveness of Competing Strategies for Management of Recurrent
Clostridium Dicile Infection: A Decision Analysis,58C
LIN.INFECT.DIS. 1507, 1507 (2014).
10
SeeegJ.C.O’Horoetal.,Treatment of Recurrent Clostridium Dicile Infection: A Systematic Review, 42 INFEC-
TION 43, 43 (2014).
11
e White House, supra note 3, at 5, 60.
12
See generally eg Zain Kassam, Christine H. Lee & Richard H. Hunt, Review of the Emerging Treatment of
Clostridium dicile Infection with Fecal Miocrobiota Transplantation and Insights into Future Challenges,34
C
LIN.LAB.MED. 787 (2014). Although there is no scientic reason to believe that FMT would not be ef-
cacious in treating rst occurrences of Clostridium dicile, it has not yet been tested in such populations, and
given the unknown long-term safety prole of FMT, it may not be suitable for such purposes on balance.
13
Konijeti et al., supra note 9, at 1511.
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

398
r
Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
e US Food and Drug Administration (FDA) has responded to the emergence of
this therapy by deciding that fecal microbiota meets the statutory denitions for both
a drug and a biological product, and as such, warrants regulation by the agency as a
drug.
14
Drugs are dened in the Federal Food, Drug, and Cosmetic Act (FDCA) as
‘articles intended for use in the diagnosis, cure, mitigation, treatment, or prevention
of disease, and articles (other than foods) that are intended to aect the structure and
function of the body of man or other animals’.
15
e Public Health Service (PHS) Act
denesbiological product as ‘a virus, therapeutic serum, toxin, antitoxin, vaccine, blood,
blood component or derivative, allergenic product, protein, or analogous product ap-
plicable to the prevention, treatment, or cure of a disease or condition in human be-
ings’.
16
Biological products may be regulated as drugs under the FDCA, as biological
products under the PHS Act, or under both laws.
17
e FDA’s decision to regulate fecal microbiota as a drug has both positive and neg-
ative consequences. Broadly speaking, there is a clear need for oversight of the use of
fecal microbiota for transplantation, given the risks posed by stool-borne transmissible
illnesses.
18
e uniformity and expertise that the FDA can oer the industry are impor-
tant institutional advantages both for stool providers navigating their legal responsibil-
ities and for physicians seeking guidance on how to best help their patients.
However, the FDA’s decision has proven problematic for scientic as well as legal
and economic reasons. Not only does stool defy the typical scientic characterization
that the FDA has long applied to small molecule and biologic drugs, but the poten-
tial do-it-yourself nature of the treatment poses particular concerns in the context of a
regime involving periods of regulatory exclusivity. e goals of any regulation of FMT
for the treatment of recurrent C. dicile infection should be to optimize access to the
therapy while maintaining strict screening and reporting standards and oversight. is
Essay will rst examine how regulation under the drug paradigm fails to achieve these
objectives, and will then evaluate available policy alternatives for the FDA to pursue, ul-
timately concluding that the FDA should pursue a hybrid regulatory model much like
the one it has implemented in the case of cord blood and other human tissues.
I. AN INTRODUCTION TO FECAL MICROBIOTA TRANSPLANTATION
First reported in Western medical literature in 1958,
19
FMT remained a fringe medical
practice over the next 50 years. However, advances in molecular microbiology and the
14
Food & Drug Administration, Fecal Microbiota for Transplantation: Scientic and Regulatory Issues 309,
May 2, 2013, hp://www.fda.gov/downloads/BiologicsBloodVaccines/NewsEvents/WorkshopsMeetings
Conferences/UCM352903.pdf (accessed June 18, 2015).
15
21 U.S.C. § 321(g)(1) (2012).
16
42 U.S.C. § 262(i)(1) (2012).
17
Some therapies that might be colloquially referred to as ‘drugs’—for instance, some oncology treatments—
are also considered to be biological products and regulated through Biologics License Applications, rather
than New Drug Applications. For the purposes of this Article, however, we are concerned with the group of
‘biological products’ that includes substances that are generally viewed as parts of or sourced from the human
body—substances like blood, bone, and even semen.
18
See Mahew J. Hamilton et al., Standardized Frozen Preparation for Transplantation of Fecal Microbiota for
Recurrent Clostridium dicile Infection, 107 A
M.J.GASTROENTEROL 761, supplementary appendix 1 (2012).
19
See generallyBenEiseman et al., Fecal Enema as an Adjunct in the Treatmentof Pseudomembranous Enterocolitis,
44 S
URGERY 854 (1958).
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
r
399
rising incidence and severity of C. dicile infection prompted a resurgence of interest in
the therapy, especially as investigations consistently found that 90 per cent of recurrent
C. dicile cases treated with FMT achieve clinical resolution.
20
FMT refers to the practice of transplantinga ltered stoolpreparation from ahealthy
donor into the lower gastrointestinal tract, typically via colonoscopy or enema, or the
upper gastrointestinal tract, for instance by nasojejunal tube. Published protocols guide
the screening of stool donors for transmissible diseases including hepatitis and HIV,
21
and the preparation of stool either for immediate use or for cryopreservation for later
use.
22
ese protocols call for the stool to be mixed with a saline solution, ltered to
remove brous material, and either administered to the patient immediately or mixed
with glycerol and frozen until the time of administration. Protocols for encapsulating
stool have also been validated and published.
23
As mentioned above, systematic reviews and meta-analyses reviewing case series of
FMT for recurrent C. dicile infection found that 90 per cent of patients were cured by
the therapy.
24
By comparison, as noted above, rst-line antibiotics treatments cure C.
dicile infection 82 per cent of the time,
25
and aer two or more recurrences, the cure
rate of standard antibiotictherapies drops to below 40per cent.
26
In the rst clinicaltrial
evaluating the therapy, FMT proved so superior to standard antibiotics that the study’s
data and safety monitoring board stopped enrollment early, concluding that it was un-
ethical to withhold the treatment from the members of the control group.
27
Neither
the systematic reviews and randomized trial nor a one- to two-year follow-up investiga-
tion found a link between FMT and any adverse events, though colonoscopy and upper
routes of administration carry procedural risks.
28
However, even though studies con-
cluding a year or two post-FMT (a timeline which is typical of many clinical trials) have
reported no adverse events, diseases that have been linked to the microbiome may sur-
face years aer FMT. As such, there remains a need for more investigation of the safety
prole of FMT in the extreme long term.
29
e mechanism by which the transplanted microbiota out-compete the C. dicile
infection also requires further investigation. Analyses of microbiota communities be-
fore and aer transplantation indicate that FMT addresses the lack of bacterial diver-
sity in recurrent C. dicile patients and restores the composition of a normal microbiota
20
Zain Kassam et al., Fecal Microbiota for Clostridium dicile Infection: Systematic Review and Meta-Analysis, 108
A
M.J.GASTROENTEROL 500, 508 (2013); Sumei Sha et al., Systematic Review: Faecal Microbiota Transplanta-
tion erapy for Digestive and Nondigestive Disorders in Adults and Children,39A
LIMENT.PHARMACOL.THER.
1003, 1004 (2014).
21
Hamilton, supra note 18, at supplementary appendix 1.
22
See generally id.
23
See generally id.
24
See Kassam, supra note 20, at 504–5; Sha, supra note 20, at 1003.
25
See Kelly & LaMont, supra note 5, at 1935.
26
Id. at 1936.
27
Els van Nood et al., Duodenal Infusion of Donor Feces for Recurrent Clostridium dicile, 368 NEW ENG.J.MED.
407, 409 (2013).
28
See Kassam, supra note 20, at 505; Lawrence J. Brandt, et al., Long-Term Follow-up of Colonoscopic Fecal Mi-
crobiota Transplant for Recurrent Clostridium dicile Infection, 107 A
M.J.GASTROENTEROL. 1079, 1079 (2012)
(following patients for a mean of 17 months post-FMT).
29
See Kassam, supra note 21 at 506.
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

400
r
Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
community.
30
It was even demonstrated that the presence of a single bacterial strain of
C. scindens could enhance resistance to C. dicile infection in mice.
31
ese ndings
and others suggest the possibility of more targeted synthetic bacterial treatments of the
infection, which companies including Seres erapeutics are pursuing.
32
While there is a paucity of data on the scale at which FMT is practiced today, Open-
Biome, a public stool bank
33
that supplies fecal microbiota preparations for clinical use,
has provided material for more than 4,000 treatments of recurrent C. dicile infection
to 300 clinical sites across the USA since it rst began operations in 2013. OpenBiome
is also sponsoring a multicenter study of the long-term safety prole of FMT in recur-
rent C. dicile patients, under an Investigational New Drug (IND) application.
34
Beyond C. dicile, studies of the microbiome have revealed important relationships
between intestinal bacteria and human health.
35
FMT for the treatment of recurrent
C. dicile infection represents a rst foray into engineering the human microbiome to
yield positive clinical results. Investigations into other applications of this therapy are
underway, including testing FMT for the treatment of irritable bowel syndrome, in-
ammatory bowel disease, and a variety of metabolic diseases.
36
Companies such as
Vedanta Biosciences are also pursuing targeted synthetic bacterial treatments of sub-
sets of these indications.
37
II. THE DIFFICULTIES OF REGULATING FMT AS A DRUG
In May 2013, the FDA held a public workshop on FMT at which they rst conrmed
that fecal microbiota would be regulated as a drug.
38
e FDA had wrien that while
FMT‘may be’ an eective therapyfor themanagement ofrefractory C. dicile infection,
30
See eg Mahew J. Hamilton et al., High-roughput DNA Sequence Analysis Reveals Stable Engrament of Gut
Microbiota Following Transplantation of Previously Frozen Fecal Bacteria,4G
UT MICROBES 125 (2013).
31
See generally Charlie G. Bue et al., Precision Microbiome Reconstitution Restores Bile Acid Mediated Resistance
to Clostridium dicile, 517 N
ATURE 205 (2015).
32
Seres erapeutics, Seres Health Presents Final Data for Study of SER-109 in Recurrent Clostrid-
ium dicile Infection at ICAAC 2014 Conference, Sept. 10, 2014, hp://serestherapeutics.com/
news/newsroom/seres
health presents nal data for study of ser-109 in recurrent clostridium dicile
infection at icaac 2014 conference/ (accessed June 18, 2015)
33
In addition to OpenBiome, some hospitals and clinics also manage internal stool banking programs.
In late 2014, BloodSource, a California-based blood bank, opened AdvancingBio, a stool bank that has
processed and banked 47 fecal microbiota preparations as of Mar. 2015. Sammy Caiola, Sacramento
Nonprot Center’s Collection Helps Patients with Intestinal Disease, S
ACRAMENTO BEE, Mar. 13, 2015,
hp://www.sacbee.com/news/local/health-and-medicine/healthy-choices/article14072969.html (acces-
sed June 18, 2015).
34
OpenBiome, OpenBiome Receives Grant to Support FMT Patient Safety Study, Aug. 7, 2014, hp://www.
openbiome.org/press-releases/2014/8/8/openbiome-receives-grant-to-support-fmt-patient-safety-study
(accessed June 18, 2015).
35
Sha, supra note 20, at 1004.
36
See eg Paul Moayyedi et al., Canadian Association of Gastroenterology Position Statement: Fecal Microbiota
Transplant erapy, 28, C
AN.J.GASTROENTEROL.HEPATOL. 66, 67 (2014).
37
See Alex Lash, With Vedanta Deal, J&J Marks Big-Pharma Milestone in the Microbiome,XCONOMY,Jan.13,
2015, hp://www.xconomy.com/boston/2015/01/13/with-vedanta-deal-jj-marks-big-pharma-milestone-
in-the-microbiome/ (accessed June 18, 2015).
38
See supra note 14, at 309.
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
r
401
its ecacy ‘ha[d] not yet been demonstrated in controlled clinical trials’.
39
All uses of
FMT would therefore need to be part of an IND application. In other words, patients
who wished to be treated with a fecal transplant for recurrent C. dicile or other indi-
cations would need to participate in a clinical trial to do so.
40
Physicians and scientists, including representatives of the Centers for Disease
Control and members of professional medical societies, quickly responded with con-
cern to the FDA’s decision. ey argued that the available evidence supporting FMT’s
eectiveness as a therapy for refractory C. dicile infection was too compelling for reg-
ulators to restrict its availability to the treatment groups of clinical trials.
41
As such, just
two months later, the FDA revised its decision and announced that it would exercise
enforcement discretion when FMT was used to treat patients ‘with C. dicile infection
not responding to standard therapies’.
42
So long as the treating physician obtained ad-
equate informed consent, the FDA would not require recurrent C. dicile patients to
receive treatment through an FDA-reviewed clinical trial.
e FDA’s decision not to enforce the IND requirement for recurrent C. dicile in-
fection has signicantly expanded access to the therapy, in no small part by creating a
window in which public stool banks like OpenBiome could operate.
43
As a result, it has
undoubtedly led to improved patient outcomes and reduced health care costs. A recent
study of the cost-eectiveness of fecal transplantation has determined that it saves a
conservative estimate of $17,000 per patient.
44
Public stool banks have developed pro-
tocols for screening and processing stool that are even more rigorous than previously
published guidelines for directed donor fecal transplants,
45
and that can represent even
more cost savings by spreading the costs associated with stool screening and prepa-
ration over many treatments.
46
However, the long-term regulatory outcome for FMT
remains unclear, and as will be explained ina, if the FDA proceeds along its current
pathway, the most likely outcome is one in which a single provider receives a license
from the FDA to produce fecal microbiota for transplantation, at which time the FDA
will likely no longer exercise its enforcement discretion.
ere are several problems with this outcome, each of which we will address in
further detail in the sections to follow. First, to eectively regulate stool as a drug,
regulators will need to address the challenge of characterizing active ingredients
and guaranteeing safety and quality in spite of the variability of stool.
47
e FDA has
not yet publicly explained how it plans to address this challenge. Second, this outcome
39
Food & Drug Administration, Public Workshop: Fecal Microbiota for Transplantation, Announcement
of May 2013 Workshop, hp://www.fda.gov/BiologicsBloodVaccines/NewsEvents/WorkshopsMeetings
Conferences/ucm341643.htm (accessed June 18, 2015).
40
Patients would likely also be able to access FMT through an expanded access protocol.
41
Guidance for industry: enforcement policy regarding investigational new drug requirements for use of fe-
cal microbiota for transplantation to treat Clostridium dicile Infection not responsive to standard therapies;
availability, 78 Fed. Reg. 42,965, 42,965 (July 18, 2013).
42
Id.
43
Mark B. Smith, Colleen Kelly & Eric J. Alm, How to Regulate Faecal Transplants, 506 NATURE 290, 290 (2014).
44
Konijeti et al., supra note 9, at 1511.
45
Mark B. Smith et al., A Scalable Workow for Screening, Processing and Characterizing Donor Stool for Use in Fe-
cal Microbiota Transplantation; Poster presentation, 2014 James W. Freston Conference, 2014 A
M.GASTROEN-
TEROLOGICAL ASS’N (Aug. 16, 2014).
46
See Smith, supra note 42, at 291.
47
Id.
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

402
r
Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
will likely impose signicant burdens on the health care system, burdens that the FDA
has not yet faced with other therapies. As an initial maer, it is probable that a license
to produce fecal microbiota for transplantation will have the eect of granting market
exclusivity to a single provider for the provision of ubiquitous human stool. Yet, in this
case, like only a very few others, doing so would allow that provider to capture value
from knowledge that already exists in the public domain, and is already being used by
multiple providers to meet an urgent and growing public health need. is outcome
would work against the public interest by inhibiting treatment choices for patients and
raising prices across the health care system. When combined with the do-it-yourself po-
tential of FMT, the decision to regulate fecal microbiota as a drug may impose safety
risks that are unique to this therapy and do not arise in the context of other, more tra-
ditional drugs. ird, given the relative freedom with which clinicians may prescribe
drugs o-label, regulating stool under the traditional drug paradigm may also discour-
age investment into explorations of fecal therapies for other indications, if o-label uses
of fecal microbiota can meet demand from patients with those conditions to which re-
searchers have made tentative links to microbiome health. is concern is not unique
to FMT, but is worth aention because in this case the FDA has the ability to select an
alternative regulatory paradigm that would mitigate, if not eliminate, this result.
A. Scientic complexities
From a scientic perspective, the regulation of stool as a drug is complicated by the
material’s complexity and inconsistency across samples. e microbial and metabolic
contents ofhuman stool are known to vary enormouslyacross individuals and over time
within individuals.
48
Unless the active components are identied, puried, and tested,
it will not be possible to guarantee that the product is consistent across batches.
is characteristic suggests that the regulation of stool should be tied to the process
by which it is prepared for transplantation, rather than to the variable contents of the
product.
49
However, in the case of human stool, this ‘process’ not only includes prepa-
ration methods following stool collection; it is also driven by the complex and very spe-
cic life history of the individual donors, and might arguably be dened by the donors
themselves.
50
More practically, the ‘process’ associated with preparing fecal microbiota
for transplantation could be narrowly dened by the very particular methods used to
lter and prepare stool from donors who are selected in accordance with a consistent
protocol.
us, under the drug regulation paradigm, given the characteristic variance of stool,
the FDA would be underregulating stool from both a safety and an ecacy perspec-
tive should it simply provide traditional drug licenses to stool-based products. Stool
preparations have distinct compositions and resultant therapeutic properties, and they
also carry distinct pathogenic threats, and thus warrant ongoing safety monitoring and
evaluation beyond that which typically accompanies an approved drug. Perhaps more
48
See generally eg J. Gregory Caporaso et al., Moving Pictures of the Human Microbiome,12GENOME BIOL.R50
(2011); Lawrence A. David et al., Host Lifestyle Aects Human Microbiota on Daily Timescales,15G
ENOME
BIOL. R89 (2014); Human Microbiome Project Consortium, Structure, Function, and Diversity of the Healthy
Human Microbiome, 486 N
ATURE 207 (2012).
49
e FDA currently regulates many biologic drugs in this fashion, regulating not simply the compound itself
but the process by which the drug is produced.
50
David, supra note 48, at 1–2.
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
r
403
importantly, it would be inappropriate to provide licenses that bear lile relationship to
the specic donor screening or stool preparation methods used by the stool provider.
Instead, an approach that narrowly denes the process by which stool material is as-
sessed and prepared for fecal transplantation would be more consistent with our un-
derstanding of the extreme complexity of the microbial system. However, the standard
drug paradigm does not clearly allow the FDA to specify who may donate stool and
which pathogens must be screened for. It is important that these complicated issues,
many of which are still not well understood by science and medicine, be considered in
a public regulatory context, rather than in the FDA’s private negotiations with a single
company.
B. Exclusivity problems
Other problems with the FDA’s decision to regulate FMT as a drug stem from one
of the FDA’s oen-overlooked powers: its statutory requirement to award periods of
exclusivity for approved drugs under a variety of conditions. As one example, the Or-
phan Drug Act awards seven years of marketing exclusivity to FDA-designated orphan
drugs,
51
during which time the FDA may not approve a new or generic drug application
for the same product and indication.
52
e recently enacted Biologics Price Competi-
tion and Innovation Act goes further, providing 12 years of data exclusivity for refer-
ence biologics, during which the FDA may not approve an application for a biosimilar
product that relies on the innovator company’s clinical trial data.
53
Although the appropriate lengths of exclusivity periods like these have been hotly
contested,
54
there is general agreement that in many cases an exclusivity period helps
provide innovative drug manufacturers with sucient incentives to carry new prod-
ucts through the long, expensive, development process.
55
e Orphan Drug Act is the
paradigmatic example: where the market for a drug is by denition small, a long exclu-
sivity period helps assure pharmaceutical companies that they can recoup their invest-
ment into a drug for treating an orphan disease.
56
Exclusivity then functions to raise the
price of the drug above what it would be in an otherwise free market, but this is a social
bargain that has been made for purposes of allowing companies to capture more of the
51
Orphan drugs are those which treat a ‘rare disease or condition’, 21 U.S.C.A. § 360bb(a) (2006), which is
dened primarily by the number of people in the United States aicted with the disease. Id. at § 360cc(a)(2).
52
21 U.S.C.A. § 360cc(a) (2006). e FDA may, however, approve an application of a dierent drug for the
same indication. Joseph A. Levi & John V. Kelsey, e Orphan Drug Regulations and Related Issues,48F
OOD
&DRUG L.J. 525, 526–28 (1993). As such, the way in which the FDA denes the drug is key to the strength
of the Act’s exclusivity periods. Id.
53
42 U.S.C.A. § 262(k)(7)(A) (2011).
54
Compare eg Federal Trade Commission, Follow-on Biologic Drug Competition 5, 2009 (suggesting that having
no exclusivity period would be sucient to promote innovation), with Biotechnology Industry Organization,
A Follow-On Biologics Regime Without Strong Data Exclusivity Will Stie the Development of New Medicines 1, 4
2007 (arguingfor a minimum of 14 years of exclusivity). See also Yaniv Heled, Patents vs. Statutory Exclusivities
in Biological Pharmaceuticals–Do We Really Need Both?,18M
ICH.TELECOMM.&TECH.L.REV. 419, 423 n.5
(2012).
55
See eg Rebecca S. Eisenberg, e Shiing Functional Balance of Patents and Drug Regulation,19HEALTH AFF.
119, 122 (2001); Arti K. Rai, e Information Revolution Reaches Pharmaceuticals: Balancing Innovation Incen-
tives, Cost, and Access in the Post-Genomics Era, 2001 U. I
LL.L.REV. 173, 183 (2001).
56
See Richard A. Merrill, e Architecture of Government Regulation of Medical Products,82VA.L.REV. 1753,
1790–91 (1996).
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

404
r
Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
value that they generate through innovation, incentivizing them to engage in innovative
drug development activities that would not otherwise have occurred.
But sometimes this bargain breaks down. In some cases, the FDA has approved a
drug when that very same drug was already widely, cheaply available on the market.
e FDA’s approval and grant of an exclusivity period in such cases may function to
clear the market of existing therapies, permiing the new applicant to charge monopoly
prices without fullling their end of the social bargain and bringing a truly new prod-
uct to market. Scholars have been concerned about two such cases that have occurred
recently, with disparate outcomes.
e story of colchicine, a drug used to treat acute gout ares, has likely been the
most widely reported. Colchicinehad been widelyand cheaply available in genericform
since the 1800s,
57
long before the FDA began regulating the safetyand ecacy of drugs.
As such, even aer the FDA acquired such authority in 1962,
58
many drugs that were
already on the market remained publicly available, although they were never formally
evaluated by the FDA. A pharmaceutical company decided to invest in conducting a
series of trials involving colchicine, to demonstrate its safety and its ecacy for gout as
well as for an orphan disease known as familial Mediterranean fever.
59
In 2009, the FDA approved the use of colchicine for these indications, and under
the Orphan Drug Act the manufacturer received seven years of marketing exclusivity—
for a product that was already on the market.
60
e manufacturer and the FDA sub-
sequently took legal action to force all other makers of colchicine to exit the market,
61
and the manufacturer then raised the price per pill from about $0.09 to $4.85. e grant
of exclusivity resulted in signicantly increased costs not only for patients but also for
Medicare and Medicaid, common payers for the drug. Scholars have persuasively ar-
gued that the rewards granted to colchicine’s now monopoly supplier vastly exceed the
value of the information provided to the public, and that public health is likely to suer
as a result.
62
Another recent example involves a synthetic hormone used to reduce the risk of
preterm births in pregnant women with a history of preterm births. e hormone had
been available to women in compounded form for many years, and it was relatively in-
expensive, at $15 per injection or $300 per pregnancy.
63
But in February 2011, the FDA
approved Makena, a branded form of the hormone, for this orphan indication. e mar-
keting company rst set a price roughly 100 times higher than that of the compounded
therapy, for a total cost of almost $30,000 per pregnancy. Although they later cut that
price in half, the cost still vastly exceeded the cost of the compounded form.
64
57
Aaron S. Kesselheim & Daniel H. Solomon, Incentives for Drug Development—e Curious Case of Colchicine,
362 N
EW ENG.J.MED. 2045, 2045 (2010).
58
Kefauver-Harris Amendment to the Food, Drug & Cosmetic Act, Pub. L. No. 87–781, 76 Stat. 780 (1962).
59
Kesselheim & Solomon, supra note 57, at 2045.
60
Under the Hatch-Waxman Act, the manufacturer also received three years of exclusivity for the use of
colchicine to treat gout, its historical use, id., but since the periods run concurrently, the total exclusivity time
is just seven years.
61
Id. at 2046.
62
Id.
63
Yesha Patel & Martha M. Rumore, Hydroxyprogesterone Caproate Injection (Makena) One Year Later,37P&
T 405, 406 (2012).
64
Id.
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
r
405
In this case, though, Makena’s maker could not secure a monopoly over the distri-
bution of the hormone. Although the FDA could not have approved another manufac-
turer’s application under the Orphan Drug Act, the FDA stated that it would decline to
take enforcement action against pharmacies that compounded the drug for patients un-
less safety problems were reported.
65
is unusual decision was likely driven by several
motivations. Historically, the FDA has had a complicated relationship with compound-
ing pharmacies—its regulatory authority in this area is unclear, and the FDA’s exercise
of its enforcement discretion in this case avoided a legal bale with the compounding
pharmacies. However, the FDA was also likely motivated by concerns about patient
access, having observed the results of the colchicine case.
If the FDA continues in its eorts to regulate fecal microbiota as a drug, it may soon
be required to make a similar choice in the case of FMT. Responding to an application
from a company called Rebiotix, the FDA has recently designated ‘fecal microbiota’ as
an orphan product for the treatment of recurrent C. dicile infection.
66
Rebiotix is cur-
rently conductingclinical trials in an eort to gainFDA approval,
67
and ifit is successful,
the Orphan Drug Act would prevent the FDA from approving another manufacturer’s
application to process and distribute whole stool material for FMT, barring a showing
of clinical superiority.
68
e question then becomes whether other stool banks would be able to stay in busi-
ness, as in the case of Makena, or whether they would be forced out of the market, as
in the case of colchicine. In our view, the FDA ought to continue exercising its en-
forcement discretion and permit other stool banks to remain as market participants,
analogous to its behavior in the Makena case. However, the FDA certainly faces coun-
tervailing pressures. Perhaps most obviously, the FDA has prioritized the removal of
unapproved drugs from the market.
69
e FDA would also likely be sued for this
65
Food & Drug Administration, Questions and Answers on Updated FDA Statement on Compounded
Versions of Hydroxyprogesterone Caproate, 2012, hp://www.fda.gov/NewsEvents/Newsroom/
PressAnnouncements/ucm310215.htm (accessed June 18, 2015).
66
Food & Drug Administration, Orphan Drug Designation for Fecal Microbiota, Mar. 10, 2014,
hp://www.accessdata.fda.gov/scripts/opdlisting/oopd/OOPD
Results 2.cfm?Index Number=421013
(accessed June 18, 2015).
67
See eg Rebiotix, Rebiotix Announces First Ever Randomized Study of a Microbiota-Based Drug for Recurrent
Clostridium dicile Infection Under an FDA IND, Dec. 16, 2014, hp://www.rebiotix.com/index.php/
rst-randomized-clinical-trial-microbiota-drug-clostridium-dicile-infection (accessed June 18, 2015).
e product, RBX2660, is a fecal microbiota preparation for enema administration, derived from
human donors. As such, it is not compositionally consistent. See Alex Park, Should We Regulate
Poop as a Drug?,M
OTHER JONES, Aug. 18, 2014, hp://www.motherjones.com/environment/2014/
08/fecal-transplant-fda-openbiome-rebiotix (accessed June 18, 2015).
68
Depomed, Inc. v. United States Dep’t of Health & Human Servs., No. 12-cv-1592, 2014 WL 4457225, at 14
(D.D.C. Sept. 5, 2014).
69
Food & Drug Administration, Unapproved Drugs Initiative, 2014, hp://www.fda.gov/Drugs/Guidance-
ComplianceRegulatoryInformation/EnforcementActivitiesbyFDA/SelectedEnforcementActionson-
UnapprovedDrugs/ucm118990.htm (accessed June 18, 2015). Although it might be supposed that the
preparation of stool for FMT is more akin to compounding than it is to the production of colchicine, and
therefore that stool banks might be able to evade the exclusivity requirements in the way that compounding
pharmacies have, this is unlikely. Unlike compounding pharmacies, no stool banks have argued that their
activities lie entirely outside the FDA’s regulatory authority. Our argument here is simply that the FDA has
selected a suboptimal regulatory paradigm.
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

406
r
Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
behavior, as it was in the Makena case.
70
Also, given the FDA’s past aempts to cur-
tail its use of enforcement discretion in the FMT context, it is dicult to think that it
would be likely to continue exercising its enforcement discretion here.
Even if the FDA did continue to exercise its enforcement discretion aer granting
Orphan Drug exclusivity to one company, the award of exclusivity itself would likely
spell trouble for other stool banking operations. In the case of colchicine, recall that
before the FDA took action, the approved company independently brought suit against
other colchicine manufacturers, seeking to force them to exit the market.
71
Should the
approved FMT company do the same, even if it would not be likely to win such a suit,
72
the time and expense required by litigation would likely have an adverse eect on the
ability of existing stool banks to continue providing care, especially given their non-
prot approach.
Importantly, the possibility of having a single stool provider on the market in itself is
not a problem. e survival of any particular FMT company is incidental to the primary
goal: to ensure patients have access to safe, eective treatments. e problem, rather, is
two-fold. First, to the extent that the prospect of granting exclusivity in the colchicine
and Makena cases has troubled scholars because the reward received by companies
would be disproportionate to the social contribution made, exclusivity for FMT raises
the same concerns. But second, and even more troubling, is a factor unique to FMT:
the potential for do-it-yourself treatments. Since monopoly power leads to monopoly
prices, if Rebiotix chooses to sell patients stool for thousands or tens of thousands of
dollars,
73
depending on health insurance coverage,
74
patients may resort to essentially
free at-home transplantations, using friends or family members, screened at the pa-
tient’s discretion, as donors. e problem is a simple one of trading o access, cost,
and safety. At a low price, access can be assured. But at a high price, the prospect of a
monopoly FMT system poses signicant safety concerns.
C. Safety concerns
Aswe haveexplained, FMT as a therapy is unique for thediculty ofits characterization
and the simplicity ofits production, andeach of these characteristics raisesspecial safety
concerns. First, the complexity of the microbial community in stool and the variability
across stool samples makes it nearly impossible to guarantee the contents from batch to
70
Makena’s maker argued that the FDA had violated its own rules by not exercising its enforcement au-
thority over the compounding pharmacies. But before the courts could ocially decide the case, the par-
ties seled. Kurt R. Karst, KV Lawsuit Involving Makena and Compounded 17P Concludes in Sopranos
Style,FDA L
AW BLOG, July 7, 2014, hp://www.fdalawblog.net/fda law blog hyman phelps/2014/07/
kv-lawsuit-involving-makena-and-compounded-17p-concludes-in-sopranos-style.html (accessed June 18,
2015).
71
See generally Complaint, Mutual Pharmaceutical Co., Inc. v. Watson Pharmaceuticals, Inc., No. 2:09-cv-
05700-PA-RZ (C.D. Cal. Aug. 4, 2009).
72
Cf. Denial of Motion for a Preliminary Injunction (Doc. 139), Mutual Pharmaceutical Co., Inc. v. Watson
Pharmaceuticals, Inc., No. 2:09-cv-05700-PA-RZ (C.D. Cal. Oct. 19, 2009).
73
Rebiotix might benchmark the price of its therapy against the price of vancomycin, which at its least
expensive (short-course and low-dose) costs between $1000 and $1500. Andrew Pollack, e Hidden
Price of Drugs,N
EW YORK TIMES, May 31, 2011, hp://prescriptions.blogs.nytimes.com/2011/05/31/
the-hidden-price-of-drugs/ (accessed June 18, 2015). Fidaxomicin, another common antibiotic, costs twice
as much.
74
Today, to our knowledge no insurer covers the use of frozen, ready-to-use fecal microbiota preparations.
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
r
407
batch. As such, ongoing monitoring for the presence of possible pathogens is necessary
for maintaining a safe product and should either be considered part of the approved
manufacturing process or a condition imposed on manufacturers. Second, although
there is lile chance that patients will manufacture traditional small molecule therapies
in their bathrooms, processing stool for transplantation at a basic level requires very
lile training or equipment. Instructional guides produced by non-professionals and
posted on YouTube have received tens of thousands of views.
75
Patient support groups
and online forums include lengthy step-by-step instructions along with discussions of
best practices for mixing stool in a low-cost blender and administering it via enema. e
risk with any regulatory strategy that restricts access to this therapy, whether by raising
barriers to clinicalpractice or by supportingmonopolistic pricing, is that it will motivate
desperate patients to pursue self-treatment.
76
Unsupervised, do-it-yourself treatments carry considerable risk of the transmission
of pathogens from improperly screened and handled stool. Few healthy individuals
would be deemed eligible for donating stool for fecal transplantation. Only six per cent
of prospective donors to OpenBiome pass the full screening process, which includes
a 109-item clinical assessment administered by a nurse or physician, and 30 stool and
blood screens.
77
With a lack of long-term safety data on patient outcomes post-FMT, it
is prudent to be overly cautious about screening for diseases that are potentially me-
diated by the microbiome; for instance, researchers have notably linked the micro-
biome to obesity,
78
metabolic syndrome,
79
and behavior.
80
Similarly, it is just as im-
portant to collect longitudinal safety data to identify any conditions that may be trans-
mied via stool of which we are unaware.
81
us, given the known and unknown risks
that come with improper donor screening and inadequate patient follow-up, the ease
with which patients may prepare and administer fecal transplants themselves without
medical supervision, any regulatory outcome that results in restricted access by either
75
See eg HomeFMT, Fecal Transplant (FMT), hp://www.youtube.com/watch?v=xLIndT7fuGo (accessed
June 18, 2015).
76
Semen is another substance with the potential for at-home administration. Perhaps responding to pre-
cisely these kinds of do-it-yourself concerns, the FDA imposes fewer and less strict requirements on di-
rected sperm donors (where the recipient and donor know each other) than on anonymous donors. See
eg 21 C.F.R. § 1271.90(a)(2) (2001); Food & Drug Administration, Donor Eligibility Final Rule and Guid-
ance Questions and Answers, 2009, hp://www.fda.gov/BiologicsBloodVaccines/TissueTissueProducts/
QuestionsaboutTissues/ucm102842.htm (accessed June 18, 2015).
77
Many (55%) are excluded through the clinical interview, while a further 65% fail the stool screens. e most
common reason for failure at the stool screen stage is for asymptomatic rotavirus (20%), which has tradition-
ally not been included in screening protocols. Blood screen failure is extremely rare. Laura Burns et al., Donor
Recruitment and Eligibility for Fecal Microbiota Transplantation: Results om an International Public Stool Bank,
Digestive Diseases Week 2015 (May 19, 2015) (poster abstract).
78
Peter J. Turnbough et al., An Obesity-Associated Gut Microbiome with Increased Capacity for Energy Harvest,
444 N
ATURE 1027, 1027 (2006).
79
Anne Vrieze et al., Transfer of Intestinal Microbiota From Lean Donors Increases Insulin Sensitivity in Individuals
With Metabolic Syndrome, 143 G
ASTROENTEROLOGY 913, 913 (2012).
80
Stephen M. Collins et al., e Adoptive Transfer of Behavioral Phenotype Via the Intestinal Microbiota: Experi-
mental Evidence and Clinical Implications,16C
URR.OP.MICROBIOL. 240, 240 (2013).
81
Smith, supra note 43, at 290.
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

408
r
Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
limiting supply or signicantly increasing the cost of therapy should be adopted ex-
tremely cautiously.
82
III. TOWARD AN IMPROVED PARADIGM FOR FMT REGULATION
As we have explained, while FMT may fall within the broad statutory denition of
‘drug’,
83
there are reasons to be concerned about regulating it wholesale under that
paradigm. In the past, when similar concerns have arisen for other technologies in-
volving human cellular or tissue-based products of various types, the FDA has adopted
or developed alternative regulatory paradigms. It should also consider doing so here.
ere are several potential paradigms available, each with their own advantages and
disadvantages.
A. Blood
Although blood, like FMT, facially appears to qualify as a ‘drug’ under the FDCA, the
FDA has promulgated regulations establishing an entirely separate system for regu-
lating blood and blood products as biologic products under the PHS Act. is regu-
latory system is largely focused on ensuring the safety of the blood supply, with the
FDA restricting the possible donor pool
84
and requiring extensive testing and quar-
antine of donated blood before it can be transfused.
85
e FDA also regulates blood
banks themselves, with highly specied registration and facility requirements.
86
Yet,
generic licenses are issued to banks meeting requirements, rather than being awarded
exclusively to one or a small number of institutions. In many ways, the FDA regulates
blood more like a commodity, rather than a therapeutic compound.
87
e FDA’s regulation of whole blood and blood products is in some ways an artifact
of history, however. Blood transfusions were routinely performed in the 1800s, and the
discovery of blood typing in the early 1900s greatly improved the safety and ecacy of
the treatment.
88
By the time the FDA ocially acquired the authority to regulate blood
in 1970,
89
the safety and ecacy of transfusions was beyond question. A far greater
82
Some at-home experimentation with fecal transplantation is unavoidable given the hype that has surrounded
the potential for this treatment to help patients with chronic, dicult-to-treat conditions such as ulcerative
colitis or Crohn’s disease. See eg Emily Eakin, e Excrement Experiment,N
EW YORKER, Dec. 1, 2014, at 64,
64. Beyond its use for treating recurrent C. dicile, FMT’s ecacy remains largely unknown, and as such, it
is necessary to require that patients seeking FMT under medical supervision for all other indications enroll
in a clinical trial, despite concerns associated with do-it-yourself treatments. More must be known about the
risks and benets of FMT to drive eective medical practice, even though it will result in patients who cannot
participate in trials resorting to self-administered variations.
83
21 U.S.C. § 321(g)(1) (2012).
84
21 C.F.R. § 640.3 (2015); Food & Drug Administration, Keeping Blood Transfusions Safe: FDA’s
Multi-Layered Protections for Donated Blood, 2013, hp://www.fda.gov/BiologicsBloodVaccines/
SafetyAvailability/BloodSafety/ucm095522.htm (accessed June 18, 2015).
85
21 C.F.R.§ 640.5 (2015); Food & Drug Administration, Keeping Blood Transfusions Safe: FDA’s Multi-Layered
Protections for Donated Blood 2013, hp://www.fda.gov/BiologicsBloodVaccines/SafetyAvailability/
BloodSafety/ucm095522.htm (accessed June 18, 2015).
86
21 C.F.R. § 607 et seq. (2015).
87
PETER BARTON HUTT,RICHARD A. MERRILL &LEWIS A. GROSSMAN,FOOD AND DRUG LAW:CASES AND
MATERIALS 1154 (4th ed., Foundation Press 2014).
88
KARA W. SWANSON,BANKING ON THE BODY 24, 30 (2014).
89
Pub. L. No. 91–515, 84 Stat. 1297, 1308 (1970) [codied at 42 U.S.C. § 262(a) (2012)]; HUTT,MERRILL &
G
ROSSMAN, supra note 87, at 1153.
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
r
409
concern, in the FDA’s view,was the potentialfor scarcity. Unlike most traditional drugs,
production of blood cannot easily be scaled up. It relies on donors, and the possibil-
ity of granting exclusive licenses to certain banks would be unthinkable, in light of the
potential human cost. e FDA’s commitment of resources to create blood’s existing
regulatory structure must be understood in the context of this history.
Blood and fecal microbiota share many scientic characteristics that distinguish
them both from traditional drugs. Unlike traditional drugs, blood and fecal microbiota
pose safety concerns involving transmissible diseases. Further, in both cases we ought
to be concerned about the potential for scarcity, due to the need at present to pro-
cure both blood and stool from human donors. And just as the ecacy of blood and
blood products was known and appreciated prior to the FDA’s involvement in this area,
FMT’s ecacy for at least one indication—recurrent C. dicile infection—has already
been demonstrated.
Yet, blood and fecal microbiota are distinct in at least one important way: unlike
blood, scientists believe that FMT may be useful for a wide range of potential indica-
tions that implicate the human microbiome. Some of these, such as ulcerative colitis
or Crohn’s Disease, may naturally be suspected due to FMT’s locus of action. Yet oth-
ers, like Parkinson’s disease, multiple sclerosis, and childhood regressive autism,
90
are
also potential indications. Research into these indications but also into the microbiome
more broadly is desperately needed, and the current paradigm for regulating blood is
not set up to oversee such clinical examination.
Instead, then, we might look to another regulatory paradigm whose history more
closely resembles that of FMT. Specically, FMT’s development quite strongly resem-
bles the history of the use of cord blood, whose potential to treat a range of diseases was
not discovered until the 1980s.
91
Cord blood, which is regulated by the FDA under its
related oversight system for human cells and tissues, is a second potential model for
FMT.
B. Tissue
Blood is not theonly therapeutic thatthe FDA has carved outof its systemfor regulating
traditional drugs. e FDA has also issued special regulations for human cells, tissues,
and cellular and tissue-based products or HCT/Ps, which are ‘human cells or tissues
that are intended for implantation, transplantation, infusion, or transfer into a human
recipient’,
92
a group that includes bones, ligaments, skin, and cord blood.
93
Like blood,
these human tissues are regulated as biological products under the PHS Act, in which
the safety of the specimen is paramount.
94
But for HCT/Ps and specically for cord
blood, the FDA has chosen to regulate dierent uses in dierent ways.
90
See Sha, supra note 20,at1028.
91
Eliane Gluckman et al., Hematopoietic Reconstitution in a Patient with Fanconi’s Anemia by Means of Umbilical-
Cord Blood om an HLA-Identical Sibling, 321 N
EW ENG.J.MED. 1174, 1174 (1989); see also Jennifer
Kulynych, Blood As A Biological ‘Drug’: Scientic, Legal, and Policy Issues in the Regulation of Placental and Um-
bilical Cord Stem Cell Transplantation,32U.R
ICH.L.REV. 407, 408 (1998).
92
21 C.F.R. § 1271.3(d) (2014).
93
Id.
94
e FDA only began to regulate many of these products in earnest in the late 1970s, once the potential for
disease transmission became known. Richard A. Merrill, Human Tissues and Reproductive Cloning: New Tech-
nologies Challenge FDA,3H
OUS.J.HEALTH L. & POL’Y 1, 10 (2002).
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

410
r
Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
Specically, wherebanked cord blood is intended for use by itsdonor or that donor’s
close relatives, that cord blood is regulated like general whole blood, as a biological
product, in a way that prioritizes the safety of the process. Accordingly, private cord
blood banks must be registered and licensed much like typical blood banks, and they
must comply with all tissue regulations. Cord blood stored for personal use may be used
for any indication. Where cord blood is intended for use by a patient unrelated to the
donor, however, it is regulated both as a biological product and as a drug.
95
is means
that cord blood when sourced from an unrelated donor must be approved for use under
an IND or BLA for the requested indication, a process that might involve extensive clin-
ical trials. To put it succinctly, for cord blood, FDA regulation is concerned with both
the intended recipient and the intended use, toggling the type and level of regulation
along both dimensions.
A key virtue of adopting the HCT/P approach in the FMT context is that it would
permit the FDA to regulate dierent uses of FMT dierently. ere is already strong
evidence to suggest that FMT can be safe and eective for the treatment of recurrent C.
dicile infection.
96
However, whether FMT can be safe and eective for the treatment
of other indications is not known. As such, the FDA might use their hybrid approach to
regulating HCT/Ps like cord blood as a guide. In eect, choosing to regulate FMT as a
biological tissue product would allow the FDA to permit its use in treating recurrent C.
dicile infection, while still requiring additional proof before it could be prescribed for
other indications.
Adopting the HCT/P regulatory paradigm wholesale would, however, permit indi-
viduals to bank and use their own stool or stool from any rst- or second-degree rela-
tives for any purpose. As such, regulators ought to be concerned about the potential for
this carve-out to reduce the benets of adopting the HCT/P approach in terms of reg-
ulating dierent indications in dierent ways. Arguably, though, it is preferable for in-
dividuals who are determined to use FMT for unapproved indications to do so through
a process that at the very least minimizes the safety concerns associated with the ther-
apy. If no such exemption existed, the reality is that patients would undergo FMT in
an unregulated, do-it-yourself fashion, not that they would be unable to undergo FMT
at all. e FDA has faced this same tradeo in the context of semen, which has similar
potential for at-home administration. e FDA therefore imposes fewer and less strict
requirements on directed sperm donors than on anonymous donors.
97
In order to regulate the safety of FMT along the same lines as either blood or tis-
sue, though, the FDA would need to either interpret or amend its existing regulations.
At present, although fecal material qualies as a biological product, the FDA’s Tissue
Reference Group has recommended that it does not meet the statutory denition of
an HCT/P and therefore could not at present be subsumed within the relevant regula-
tions.
98
e Reference Group did not publicly explain their reasons, but a close reading
95
Food & Drug Administration, Cord Blood Banking: Information for Consumers, 2012, hp://www.
fda.gov/BiologicsBloodVaccines/ResourcesforYou/Consumers/ucm236044.htm (accessed June 18, 2015).
96
See generally van Nood, supra note 27.
97
See eg 21 C.F.R. § 1271.90(a)(2) (2001).
98
Food & Drug Administration, Tissue Reference Group: FY 2012 Update, 2012, hp://www.fda.gov/
biologicsbloodvaccines/tissuetissueproducts/regulationoissues/ucm152857.htm (accessed June 18,
2015).
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
r
411
of the regulations suggests two primary reasons the Reference Group may have reached
this conclusion, each of which can be surmounted.
First, the Group members may have reasoned that although fecal microbiota prepa-
rations are at least partly composed of ‘human cells or tissues that are intended for
implantation, transplantation, infusion, or transfer into a human recipient’, fecal mi-
crobiota falls within the explicit regulatory carve-out for ‘secreted or extracted human
products’, which include milk and collagen.
99
ese products are not considered to
be HCT/Ps, and therefore are not regulated under the tissue paradigm—their col-
lection and storage is not subject to the same rigorous safety requirements that apply
to cord blood products. Importantly, the FDA has made exceptions to this exception
before—the carve-out for secreted products goes onto state that ‘semen is considered
an HCT/P’, suggesting that otherwise semen would be considered a secreted product
as well.
100
If this limitation on what constitutes an HCT/P is the basis for the Reference
Group’s conclusion, the FDA would likely need to conduct an ocial notice-and-
comment rulemaking to clarify that fecal microbiota is also considered an HCT/P.
101
A simpler approach would be to aempt to issue an interpretive rule or guidance doc-
ument indicating that fecal maer simply does not qualify as a ‘secreted or extracted
human product’.
102
However, this approach might be viewed by courts to ‘purport[] to
impose legally binding obligations or prohibitions on regulated parties’,
103
and there-
fore might be rejected as an eort to circumvent the Administrative Procedure Act.
Second, the Reference Group may have reasoned that FMT does not require the
transplantation of ‘human cells or tissue’, but that it rather requires the transplanta-
tion of microbial cells or tissue.
104
is concern is easily surmounted through guidance
documents. e simplest option would be for the FDA to clarify that FMT as it is cur-
rently practiced simply does fall into the category of ‘articles containing ... humancells
or tissues’. Although it is possible that in the future the material for FMT will not need
to be sourced from humans and therefore will not necessarily contain human cells,
105
at present the raw material for FMT is inexorably sourced from humans and inevitably
contains human cells.
Yet, either of these approaches would require a signicant expenditure of resources,
which the FDA might not be willing to allocate without political pressure of the kind
99
21 C.F.R. § 1271.3(d)(3) (2014).
100
is is not precisely accurate because semen is also explicitly enumerated as an HCT/P, but in common par-
lance semen would otherwise be considered as something which is ‘secreted’ by the body.
101
5 U.S.C. § 553; 5 U.S.C. § 551(4).
102
5 U.S.C. § 553(b)(3)(A), (d)(2).
103
Nat’l Min. Ass’n v. McCarthy, 758 F.3d 243, 251 (D.C. Cir. 2014).
104
is is not the most natural reading of the regulation, which requires only that the article ‘contain[]’
human cells or tissues, and in fact many scientists have begun to refer to the microbiome as a human
organ. See eg Alexander Khoruts et al., Development of Fecal Microbiota Transplantation Suitable for
Mainstream Medicine,13C
LIN.GASTROENTEROL.&HEPATOL. 246, 247 (2015). However, in a May 2013
public meeting, Rebiotix stated that it had received a leer from the FDA reaching this interpretation. Food
& Drug Administration, Fecal Microbiota for Transplantation: Scientic and Regulatory Issues 229, May 2, 2013,
hp://www.fda.gov/downloads/BiologicsBloodVaccines/NewsEvents/WorkshopsMeetingsConferences/
UCM352902.pdf (accessed June 18, 2015). As such, we must take this argument seriously.
105
See ina text accompanying note 110.
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

412
r
Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
they faced in the cord blood context.
106
Given the potential for publicly reported
adverse events that may be aributed to the underregulation of processing and storage
of stool samples, the FDA might be willing to take such actions, but it is by no means
clear at this time. If they are not willing to devote the resources, a third option—simple
continued use of enforcement discretion—might be even more aractive.
C. Enforcement discretion
As discussed above, the FDA has currently chosen to use its enforcement discretion to
de facto permit physicians to provide FMT for the treatment of recurrent C. dicile in-
fections. If the FDA continued to exercise this discretion, it would be able to achieve
a similar outcome as it could achieve in the HCT/P context. at is, physicians would
be free to provide FMT to treat C. dicile infections not responding to standard ther-
apies, but a company hoping to market FMT for other indications would still need to
complete clinical trials.
Importantly, if the FDA chose to continue exercising its enforcement discretion for
FMT for recurrent C. dicile, health care providers would not be without expert advice
on the subject. Professional organizations like the American Gastroenterological Asso-
ciation have issued best practice guidelines for FMT, including restrictions on donor
selection, sample processing, and facilities management.
107
e FDA’s regulation of
blood products provides a helpful example. In that context, private standard-seing or-
ganizations like the American Association of Blood Banks had already begun to reg-
ulate the blood supply before the FDA was involved in the process, a role which the
professional association continues to fulll today. Although these guidelines lack the
legal force of the FDA’s regulatory scheme, they are not toothless. Physicians subject
to malpractice actions who have failed to adhere to guidelines that represent the stan-
dard of care may be subject to professional discipline. By extension, stool banks would
face pressure to demonstrate compliance with best practice guidelines from the clinical
sites and insurers using their services.
A key virtue of this approach is that it permits the FDA to postpone the o-label
prescription problems that might arise if it does approve Rebiotix’s FMT application.
Essentially, although manufacturers are not permied to advertise their products for
o-label uses, doctors are free toprescribe therapies and devices for o-label indications,
with one study suggesting that about 20 per cent of all drugs are prescribed o label.
108
Too oen there is no scientic support for o-label prescriptions, and with a grant of
exclusivity, the manufacturer has lile nancial incentive to conduct further validating
studies. Because approving FMT for the treatment of recurrent C. dicile infections
would permit physicians to prescribe it for all other unproven indications, it would per-
versely delay the FDA’s acquisition of the very information it wants most—data on the
106
See eg Kulynych, supra note 91, at 409, 410 (discussing various statutory eorts in this area).
107
Leer from the Presidents of the Infectious Disease Society of America, the American Gastroen-
terological Association, the American Society for Gastrointestinal Endoscopy et al., Current Consen-
sus Guidance on Donor Screening and Stool Testing for FMT, July 5, 2013, hp://www.gastro.org/
research/Joint
Society FMT˙Guidance.pdf (accessed June 18, 2015).
108
David C. Radley et al., O-label Prescribing Among Oce-Based Physicians, 166 ARCH.INTERN.MED. 1021,
1021 (2006).
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
r
413
use of FMT for other, non-C. dicile indications—by dampening the nancial incen-
tive to conduct those additional trials.
109
is approach also has benets from a regulatory eciency standpoint. When com-
pared with the possibility of constructing an entirely new regulatory system, as the FDA
has done in the context of blood, or even completing a notice-and-comment rulemak-
ing to bring FMT within the scope of the HCT/P regulatory scheme, the possibility
that the FDA could achieve a somewhat similar result by simply doing nothing is un-
doubtedly aractive. Further, it is useful to recall that the use of stool for FMT is only a
bridge technology, at least in the C. dicile context. Companies seeking to understand
the microbiome and use it directly to treat the disease
110
will (perhaps soon) obviate
the need to use donated stool for these purposes. As such, any regulatory action taken
by the FDA may be obsolete within a decade.
is approach is certainly not a perfect one. e possibility that enforcement discre-
tion may be withdrawn at any time places stool processing companies in a precarious,
uncertain position regarding their legality. It also does not avoid the o-label prescrib-
ing problem, but simply delays it until such time as the FDA approves FMT for a dif-
ferent indication. Still, the virtues of the status quo ought to be considered carefully.
D. Feasibility in light of path-dependency
e FDA cannot pursue this option on a blank slate. e possibilities above must be
considered in light of decisions that the agency has already made in this area that may
inhibit its ability to adopt a more exible paradigm in the near term. To briey summa-
rize whathas been described above,in 2012, the FDA’sTissue Reference Group recom-
mended that fecal microbiota did not meet the statutory denition of an HCT/P.
111
In
2013, the FDA explained that it considered fecal microbiota for transplantation a drug,
and would regulate it as such. A few months later, the agency permied Rebiotix to
pursue licensure for its fecal microbiota preparation under the approval program for
orphan drugs and subsequently awarded it an Orphan Drug designation for treating
recurrent C. dicile infection.
112
In light of the scientic, legal, and public health argu-
ments against regulating fecal microbiota preparations for C. dicile therapiesasdrugs,
what options do the rest of these decisions leave the FDA?
109
Although these concerns technically apply to all standard drugs, we might both be particularly worried about
o-label uses of FMT and particularly concerned about incentives for companies to perform clinical trials
here. O-label uses of FMT might be particularly concerning because we are just beginning to understand the
relationship between the human microbiome and disease. Scientists may have a relatively complete picture
of how a given small molecule drug interacts with the body, but the potential health eects of microbiome
manipulation require much greater study than they have received to date. Widespread o-label prescribing
has the potential to hamper eorts to do so. And given the diculty of protecting technologies like FMT
by traditional intellectual property means, companies will lack other incentives to pursue such testing. See
generally Amy Kapczynski & Talha Syed, e Continuum of Excludability and the Limits of Patents, 122 Y
ALE
L. J. 1900 (2013).
110
See eg Seres erapeutics, About Seres erapeutics, 2015, hp://serestherapeutics.com/about/about seres
therapeutics/ (accessed June 18, 2015).
111
Food & Drug Administration, Tissue Reference Group: FY 2012 Update, 2012, hp://www.fda.gov/
biologicsbloodvaccines/tissuetissueproducts/regulationoissues/ucm152857.htm (accessed June 18,
2015).
112
See supra note 66.
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

414
r
Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
Above, we considered theoption of surmounting the ReferenceGroup’s recommen-
dation through regulation, and indeed, the FDA could conceivably regulate FMT as an
HCT/P in addition to regulating it as a drug, though this route may prove costly. We
also considered the FDA’s ability to continue exercising its enforcement discretion in
an aempt to promote access to FMT for recurrent C. dicile, but this option would
indeed be jeopardized in light of its prior decisions in this area. Beyond regulation,
though, there are other ways in which the FDA could use various legal tools to minimize
the social harms that granting an exclusive license to provide stool for transplantation
would cause.
One option would be for the FDA to narrowly construe its Orphan Drug desig-
nation, dening the drug itself narrowly by the very specic processing methodology
to which it is subject. As a result, other companies seeking to provide stool for trans-
plantation would still need to complete clinical trials, but companies would be able to
invent around Rebiotix’s method. Or alternatively, the FDA might approve another
company’s FMT preparation on the grounds of clinical superiority.
113
As a third op-
tion, since the existing designation is limited to recurrent C. dicile infections, a prod-
uct could be tested and approved forinitial occurrences of C. dicile.
114
However, these
would be unusual outcomes.
ere are still other more promising ways for the FDA to regulate this new and
rapidly evolving eldwith exibility. Perhaps the most novel would be to carefully mon-
itor anyapproved company with exclusivityfor the presence ofa drug shortage. Dened
broadly as a situation in which ‘the demandor projected demand for the drug within the
United States exceeds the supply of the drug’,
115
in such an event the FDA is empow-
ered to take a range of actions, including obtaining the product from other sources. And
here it is again critical to remember the ways in which FMT is not a typical drug because
a company cannot simply ‘scale up’ production of stool by building more factories or
increasing production eciency. It must instead recruit additional healthy donors and
collect the product from them. OpenBiome, with which one of the authors of this Es-
say is aliated, has curated the raw material for just 7400 treatments in its two years of
existence. e FDA might reasonably decide that in the event of a shortage, whether
produced by an unanticipated surge in demand or barrier to supply, it would take
any single corporation too long to recruit enough healthy donors and de-risk them
through multiple rounds of screens for it to provide coverage alone. Indeed, any short-
age, whether now or in the future, would be made worse if a single company held a
monopoly over stool procurement and processing.
IV. CONCLUSION
e question of how the FDA should regulate FMT to prioritize the safety of the pro-
cedure while at the same time ensuring its eectiveness is a complicated one. As we
have explained, trying to shoehorn FMT into the traditional drug regulatory paradigm
is problematic, with the potential to both underregulate safety and overregulate access.
113
21 C.F.R. § 316.20(a) (2014).
114
See Levi & Kelsey, supra note 51, at 526, 528.
115
Food & Drug Administration, Drug Shortage Management, 12, 2014, hp://www.fda.gov/downloads/
AboutFDA/CentersOces/OceofMedicalProductsandTobacco/CDER/ManualofPoliciesProcedures/
UCM079936.pdf (accessed June 18, 2015).
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from

Ensuring the safe and eective FDA regulation of fecal microbiota transplantation
r
415
Regulating FMT as a tissue would more fully address our concerns, since so much is
still unknown about FMT’s potential ecacy across a wide range of indications.
A model much like the one the FDA has adopted in the case of HCT/Ps like cord
blood would seem to be the most appropriate solution. Under this scenario, the FDA
would enforce a suite of regulations that control for the safety of fecal microbiota prepa-
rations so long as the material is only being used to treat recurrent C. dicile infection.
For all other uses, the FDA would require demonstrations of safety and ecacy follow-
ing thepath for investigational newdrugs. is paradigm wouldgrant C.dicile patients
safe access to ubiquitous human stool, the data for which is available in the public do-
main, while at the same time encouraging scientists to study new applications for the
therapy.
FMT presents the FDA with a new regulatory challenge, and it provides the agency
with a chance for thinking critically and creatively about the most appropriate ways to
oversee this new technology. e FDA should seize the opportunity, as other emerg-
ing technologies like 3-D printing and mobile health may soon pose similar regulatory
challenges for allowing access under medical supervision while restricting that access
to evidence-based practices.
116
We hope the considerations presented in this Essay will
help the FDA in its endeavors to do so.
116
e FDA has undoubtedly faced numerous new challenges throughout its history, and it has had to consider
how it might adapt old statutes to new problems. Merrill, supra note 94, at 2, 3; see also generally Jody Freeman
& David B. Spence, Old Statutes, New Problems, 163 U. P
A.L.REV. 1 (2014). But the do-it-yourself nature of
FMT will soon be echoed by 3-D printing technology and mobile health developments.
by guest on November 5, 2015http://jlb.oxfordjournals.org/Downloaded from