Diagnostic Microbiology and Infectious Disease 88 (2017) 322–329

Contents lists available at ScienceDirect

Diagnostic Microbiology and Infectious Disease

journal homepage: www.elsevier.com/locate/diagmicrobio

Clinical Studies

Is frozen fecal microbiota transplantation as effective as fresh fecal

microbiota transplantation in patients with recurrent or refractory
Clostridium difficile infection: A meta-analysis?
Guihua Tang a,1, Wen Yin b,1, Wenen Liu a,⁎
a
Department of Clinical Laboratory, Xiangya Hospital of Central South University, 410008, Changsha, China
b
Department of Neurosurgery, Xiangya Hospital of Central South University, 410008, Changsha, China

a r t i c l e i n f o a b s t r a c t

Article history: Fecal microbiota transplantation (FMT) is a remarkably efficacious therapy for recurrent or refractory Clostridium difficile
Received 31 October 2016 infection (CDI), but not standardized. This work is to determine whether frozen FMT is as effective as fresh FMT. Meta-
Received in revised form 7 November 2016 analysis showed that frozen FMT was as effective as fresh FMT, both pooled first effective rate (65.0% (95% CI 57.0–
Accepted 14 May 2017
73.0%) vs. 65.0% (95% CI 57.0–73.0%), P = 0.962) and pooled second effective rate (95.0% (95% CI 91.0–99.0%) vs. 95.0%
Available online 18 May 2017
(95% CI 92.0–99.0%), P = 0.880). In conclusion, among patients with recurrent or refractory CDI, frozen FMT is as effective
Keywords:
as fresh FMT. Considering potential advantages of performing frozen FMT, it is a reasonable option to select frozen FMT.
Clostridium difficile © 2017 Elsevier Inc. All rights reserved.
Frozen fecal microbiota transplantation
Fresh fecal microbiota transplantation

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
2. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
2.1. Search methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
2.2. Selection criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
2.3. Study selection and data collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
2.4. Methodological quality appraisal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
2.5. Statistical methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
3. Main results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
3.1. Methodological quality of included studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
3.2. Demographic and fecal microbiota transplant information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
3.3. Treatment efficacy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
3.4. Adverse event and follow-up date . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
4.1. Donor screening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
4.2. Adverse events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
4.3. Delivery modalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
4.4. Limitations of this review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
Funding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
Conflicts of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328

⁎ Corresponding author. Tel./fax: +86-731-84327437.

E-mail addresses: [email protected] (G. Tang), [email protected] (W. Yin), [email protected] (W. Liu).
1
Guihua Tang and Wen Yin contributed equally to this manuscript.

http://dx.doi.org/10.1016/j.diagmicrobio.2017.05.007
0732-8893/© 2017 Elsevier Inc. All rights reserved.
 G. Tang et al. / Diagnostic Microbiology and Infectious Disease 88 (2017) 322–329 323

1. Introduction (defined as ≥3 unformed stools in 24 hours) for a minimum of 48 hours

and laboratory confirmed (toxin assay or culture, or colonoscopic/endo-
Clostridium difficile infections (CDI) is a major public health scopic findings demonstrating pseudomembranous colitis (Bagdasarian
challenge, and the incidence and mortality rate of CDI has increased et al., 2015; Dupont et al., 2016; Lubbert et al., 2014)) CDI. (2) These pa-
tremendously in both hospitals and community settings in the last tients developed into recurrent CDI (diagnosed with diarrhea the same
two decades (Goudarzi et al., 2014). As the first-line basic therapy, as CDI with or without stool tests after successful antibiotic therapy) or
metronidazole and vancomycin are associated with unacceptable high refractory CDI (persistent or aggravating of diarrhea despite treatment
CDI recurrence rate (Kociolek and Gerding, 2016). Besides, the use of with standard antibacterial therapy for at least 5 days). (3) Therapies in-
antibiotics may disturb the intestinal microbiota (Mikkelsen et al., cluded fecal transplantation with clear fresh or frozen stool samples.
2015). Therefore, the first step of treating C. difficile related diseases is (4) Only case series with no less than 10 patients were included to min-
cessation of all antibiotics being used if possible (To and Napolitano, imize bias inevitably associated with individual small cases series.
2014), and novel therapeutic strategies are ideal alternatives to current (5) Studies were only included if they took clinical diarrhea resolution
treatment methods (Kociolek and Gerding, 2016). as the primary outcome but not CDI toxin assay or culture, which is con-
Fecal microbiota transplantation (FMT) is a remarkably efficacious sistent with current guidelines (Surawicz et al., 2013).
therapy for recurrent or refractory CDI, which was demonstrated by a Exclusion criteria included: (1) Study, which was a review or
large amount of clinical studies with unlimited delivery methods, such researched on animal model, was excluded in this meta-analysis.
as orally administered capsules (Hirsch et al., 2015), nasogastric tube (2) Publications without original data (abstract data, note, comment,
(Rubin et al., 2013), colonoscopy (Allegretti et al., 2014; Brandt et al., etc.) were excluded and only completed studies that were included.
2012; Patel et al., 2013; Pathak et al., 2013; Ray et al., 2014), enteroscopy (3) It was an inappropriate study that only included a particular
(Ganc et al., 2015) and retention enema (Kassam et al., 2012; Lee et al., crowd such as the child, the aged et al. (4) Non-English article was
2014). The clinical solution rate kept at a higher level varying from 73% eliminated as well.
to 100%. Even more important, FMT is able to restore damaged intestinal
flora (Khoruts et al., 2010). Nowadays, FMT has been recommended to
2.3. Study selection and data collection
management of recurrent CDI, which coincides with current guidelines
(Lubbert et al., 2014; Surawicz et al., 2013). But there exists no standard
Title, abstract or full text were screened and eligible studies was
for preparation because it's complex and highly variable composition
determined by two independent reviewers (Guihua Tang and
cannot be easily characterized, nor do its quality.
Wen Yin) using predefined inclusion and exclusion criteria. Disagreements
Clearly, international consensus on standard protocols can only be
were resolved by discussion and compromise settlement, involvement of
achieved with further experience to support the production of safeness
another investigator, if necessary.
and effectiveness. One of the important obstacles is that logistical
Data extraction for qualified studies was also carried out by the
challenge in terms of obtaining and processing fecal materials. However,
above two reviewers respectively with a standardized, predefined
these problems are being overcome by the emergence of frozen stool
data extraction form. Extracted data was compared in the end and any
banks (Goldenberg, 2016). This frozen protocol is quite striking that the
discrepancies were resolved by consensus (Wenen Liu). Information
same sample can be used for several FMTs. It could avoid the duplicate
collected included demographic information; CDI type (recurrent or
laboratory examinations, reduce cost with less number and frequency of
refractory); pre-FMT therapy; FMT delivery modality; if colonic lavage
donor screenings and make rapid treatment possible in case of emergency
or not; FMT dose; FMT donor (patient selected or not); clinical resolution
that could not in on-site laboratory facilities. However, it is indecisive that
rate; adverse events and follow-up date.
if frozen stool has the same efficacy with fresh stool in clinical treatment.
Hence, we performed a meta-analysis to determine whether frozen fecal
transplantation is as effective as fresh fecal transplantation for patients 2.4. Methodological quality appraisal
with recurrent or refractory CDI.
Methodological quality of included studies was assessed by a
2. Methods reviewer (Guihua Tang). The Cochrane risk of bias tool (Higgins and
Green, n.d.) was used to measure randomized controlled trials (RCTs),
2.1. Search methods assessing risks of bias for sequence generation, allocation sequence
concealment, blinding, dealing with incomplete outcome data, selective
A comprehensive literature search of the Pubmed, Embase and outcome reporting and other potential threats to validity. These items
Cochrane Central Register of Controlled Trials was conducted in June were rated as high (e.g. an ‘intention-to-treat’ analysis included),
8, 2016. The search strategy was not limited by date and language. unclear (e.g. reasons for missing data were not adequately described),
Our search terms included medical subject headings (MeSH) and free or low risk of bias (e.g. all patients completed the study and there
words found in title or abstract. Fecal microbiota transplantation was were no losses to follow up). While non-RCTs (e.g. cohort studies or
the first MeSH, studies identified were combined using the set operator case series) were appraised using the 8 quality criterions from the
“or “with its free words: microbiota transplantation, fecal/transplantation, Centre for Reviews and Dissemination checklist (Chambers et al., 2009)
fecal microbiota/intestinal microbiota transfer/microbiota transfer,
intestinal/transfer, intestinal microbiota/donor feces infusion/fecal Table 1
transplant/transplantation, fecal/fecal transplantation/transplant, fecal/ Quality criteria for assessing non-RCTs.
feces infusion, donor/infusion, donor feces. The second MeSH: clostridium Quality rating good, if the answer is “yes” to all 8 criteria; satisfactory, if the answer
difficile, which were combined with its free words: C. difficile/clostridium/ is “yes” to criteria 1–5; poor, if the answer is not “yes” to one or more of the
CDI/CDAD in the same way. Afterwards, we combined the two groups of criteria listed for “satisfactory”.
terms with the operational character “and” to complete our literature 1. Were patients recruited prospectively?
2. Was the selected population representative of that seen in normal practice?
query. No limits were applied for study design.
3. Were patients recruited consecutively?
4. Were at least 90% of those included at baseline followed up?
2.2. Selection criteria 5. Was loss to follow-up reported or explained?
6. Were selection/eligibility criteria adequately reported?
Eligibility criteria was established in advance by the study authors 7. Was an appropriate measure of variability reported?
8. Did the study report relevant prognostic factors?
and listed as follows: (1) Patients with a history of presence of diarrhea
324 G. Tang et al. / Diagnostic Microbiology and Infectious Disease 88 (2017) 322–329

(Table 1). Results of methodological quality assessment did not decide the Records identified through
eligibility of the studies. database searching (n = 848)

2.5. Statistical methods Reserved records after duplicates

removed (n = 609)
The efficacy of FMT treatment was calculated and we compared fro-
zen FMT with fresh FMT across studies as well as other predefined sub- Studies excluded for inappropriate
groups. Clinical effective rates were judged by the resolution rate of title or abstract (n = 564)
clinical diarrhea at the observation timing in original text, including
first effective rate (defined as clinical resolution rate after only once Full-text articles assessed for
FMT without the need for other therapy) and second effective rate (de- eligibility (n = 45)
fined the same as the former but after multiple FMT). In addition, it is di-
rectly deemed to a failure case if treated with other therapy methods
Full-text articles excluded (n = 39)
instead of FMT. The data was collected using our interpretation of clin-
ical resolution after assessing cases within each study, when available.
All clinical effective rates were calculated with corresponding pooled Studies included in qualitative assessment (n = 6)
rates and 95% confidence intervals (CI) for overall studies. The resolu-
tion rates were analyzed using the fixed effects model (fixed, inverse
Fig. 1. Flow diagram of study selection.
variance). Heterogeneity among studies was evaluated using the
Cochran Q test and the I 2 statistic, while Cochrane Q b 0.10 or I 2 N 50%
was considered to significant heterogeneity. Clinical effective rates be-
tween subgroups were compared by pooled first effective rate and 3.3. Treatment efficacy
pooled second effective rate. Statistical analyses were performed by
using statistical software stata/SE (version 12.0). In this meta-analysis, the pooled first effective rate of FMT was 65.2%
(95% CI 59.5–70.9%) with no statistically significant heterogeneity
between studies (Cochran Q test P = 0.53, I 2 = 0.0%). Similarly, the
3. Main results
meta-analysis plot of pooled second effective rate of FMT was also
finished, reaching up to 95.2% (95% CI 92.7–97.8%), without significant
Our search strategy yielded 848 potentially relevant publications,
heterogeneity across studies too (Cochran Q test P = 0.561, I 2 = 0.0%).
which were reserved 609 records after duplicates removed. Afterwards,
Priori subgroups analysis sorted out with clinical plausibility were
564 articles were excluded based on titles and abstracts for not fulfilling
fulfilled when data was available. It was our first selection that compar-
inclusion criteria. Ultimately, 45 papers underwent full-length article
ing the effective rate (both pooled first effective rate and pooled second
review and 6 studies (Cammarota et al., 2015; Lee et al., 2016; Van
effective rate) of frozen FMT with fresh FMT, which was in accordance
Nood et al., 2013; Youngster et al., 2014a, 2014b; Zainah et al., 2015)
with our objective. If frozen FMT and fresh FMT simultaneously being
met our eligible inclusion criteria. The excluded articles contained 17 in-
used in a study, we performed the analysis with the original data sepa-
complete publications (abstracts, comments, letter et al.) without orig-
rately. As displayed in Figs. 2 and 3, the pooled effective rates both had
inal data (Arkkila et al., 2010, 2014; Berg and Farraye, 2013; Ganc and
no statistical difference between groups.
Ganc, 2014; Ganc et al., 2013; Ho and Prasad, 2011; Kassam et al.,
Our second subgroup analysis was comparing the efficacy based on
2010; Khoruts et al., 2012; Listed, 2013; Malani and Rao, 2016; Mellow
the colonic lavage or not before FMT. Three studies carried out colonic
and Kanatzar, 2010; Obi et al., 2014; Patel et al., 2015; Ramsauer et al.,
lavage (Cammarota et al., 2015; Van Nood et al., 2013; Youngster
2016; Tian et al., 2015; Vigvari et al., 2014; Wise, 2016), 5 studies
et al., 2014b) prior to FMT and three studies did not (Lee et al., 2016;
were without defined frozen or fresh FMT (Aroniadis et al., 2016; Brandt
Youngster et al., 2014a; Zainah et al., 2015). The pooled first effective
et al., 2012; Lee et al., 2014; Ray et al., 2014; Rohlke et al., 2010), 3
rate of colonic lavage group was 63.0% (95% CI 56.0–69.0%) with no sig-
studies were insufficient of predefined sample size (Costello et al.,
nificant heterogeneity between studies (Cochran Q test P = 0.706, I2 =
2015a; Stollman et al., 2015; Weingarden et al., 2013), and 14 studies
0.0%) and pooled first effective of the other group was 73.0% (95% CI
in which patient populations didn't match our inclusion criteria
61.0–84.0%) with no significant heterogeneity between studies
(Allegretti et al., 2014; Costello et al., 2015b; Hamilton et al., 2012;
(Cochran Q test P = 0.504, I 2 = 0.0%). Though there were differences
Hirsch et al., 2015; Kassam et al., 2012; Kelly et al., 2012; MacConnachie
of pooled first effective rate but without statistical significance (P =
et al., 2009; Mattila et al., 2012; Patel et al., 2013; Pathak et al., 2013;
0.530). Similarly, pooled second effective rates were reaching 96.0%
Rubin et al., 2013; Satokari et al., 2015; Vigvári et al., 2015; Yoon and
(95% CI 93.0–98.0%) and 92.0% (95% CI 84.0–99.0%) respectively. The
Brandt, 2010) (Fig. 1).
proportion difference was not statistical significance (P = 0.561).

3.1. Methodological quality of included studies

3.4. Adverse event and follow-up date
Quality assessment of four included RCTs is shown in Table 2. They
were all rated as “high” methodological quality. However, methodological As showed in all eligible studies that FMT was safe. Although adverse
quality of two included non-RCTs (Youngster et al., 2014a; Zainah et al., events related to FMT were reported but were little aggressive. No any
2015) both rated as “poor”, because patients recruited in the two studies severe adverse events were reported. The mentioned adverse events in-
were not prospectively. cluded symptoms of digestive tract (diarrhea, bloating, abdominal
cramping and belching) and a transient fever in a child treated via colo-
3.2. Demographic and fecal microbiota transplant information noscopy (Youngster et al., 2014b). These problems all resolved within
12–72 h. Serious cases needed another hospitalization even death
The demographic and fecal microbiota transplant information cases also reported, but none were due to the FMT procedure. However,
collected was as detailed in Table 3. The modified intention-to-treat such cases generally occurred in the critically ill patients with unre-
populations in Lee et al. (2016) study were excluded in our analysis. solved CDI, malignant tumor or other severe comorbidities. The
Thus, there were a total of 268 patients included in this review. follow-up date was diverse and varied from 10 weeks to 6 months.
 G. Tang et al. / Diagnostic Microbiology and Infectious Disease 88 (2017) 322–329 325

Table 2
Methodological quality of included RCTs.

references Random sequence generation Allocation sequence concealment Blinding Incomplete outcome Selective reporting

Lee et al. Low Low Low High3 Low

Cammarota et al. Low Low High Low Low
Youngster, Sauk et al. Low Unclear2 High Low Low
Van Nood et al. Low1 Unclear2 High Low Low

Note: 1 = Blocked randomization or a computer random number generator was not in the paper, but the first and last two authors vouched for the fidelity of the study protocol.
2 = Not described details of allocation concealment, rated as “unclear” for this item.
3 = A number of incomplete outcome data, “as-treated” (per-protocol) analysis is an inappropriate approach.

4. Discussion the FMT procedure, and further studies about adverse events are re-
quired to registry.
Based on the 6 studies in our meta-analysis, the treatment efficacy of
frozen FMT and fresh FMT had no statistical significance. Some previous 4.3. Delivery modalities
studies demonstrated the efficacy of frozen FMT for treating recurrent
or refractory CDI (Youngster et al., 2014a, 2014b). Hamilton and col- As showed in this review, several delivery modalities have been de-
leagues (Hamilton et al., 2013) used high-throughput DNA sequence scribed in spite of frozen FMT or fresh FMT, no assessment achieved
analysis of gut microbiota following frozen fecal bacteria and Costello comparing outcomes of different delivery methods. But there were
et al. (2015b) evaluated the viability of bacteria in stool frozen both re- some other studies that assessed delivery modalities (Kassam et al.,
vealed clinical efficacy of frozen fecal bacteria. In addition, the study of 2013; Postigo and Kim, 2012). Although those studies suggested that
Lee et al. (2016) suggested frozen FMT did not lead to worse proportion lower gastrointestinal FMT delivery may be preferable compared with
of clinical resolution of diarrhea compared with fresh FMT. However, upper gastrointestinal FMT delivery; however, there is no widespread
this was the first meta-analysis concentrated on this issue. Meanwhile, consensus on the optimal FMT route of administration. More relevant
this meta-analysis compared the pooled first effective rate and pooled randomized controlled trials are urgently needed to determine the
second effective rate, reflecting the first therapeutic efficacy and overall best modality.
therapeutic efficacy respectively. Several advantages or disadvantages of different delivery methods
FMT is a novel colonization with donor intestinal flora instead of briefly illustrated here. Using a NGT provides more exposure of donor
pathogenic bacteria. The viability of some bacteria reduced at fecal bacterial flora in recipients' digestive tract. It allows transplanted
6 months when stool specimens stored in normal saline but not in flora to sufficient survival and results in overcoming the possible
10% glycerol or in normal saline at 2 months (Costello et al., 2015b). It growth of pathogens. In addition, NGT was tolerated well and with
suggested frozen FMT was an ideal treatment as long as stool samples less significant insertion-related complications than colonoscopy
stored in appropriate conditions. Our analysis suggests that frozen (Postigo and Kim, 2012). Colonoscopy-guided FMT allows physician to
FMT is most probably a promising intervention as effective as fresh directly evaluate the severity of inflammation, select preferential sites
FMT for recurrent or refractory CDI, when fails to conventional for sufficient amount of feces (Cammarota et al., 2015). And the incor-
treatments. However, there are still many barriers to overcome before porated bowel lavage can reduce the existing pathogenic content and
being a standard therapy. facilitate colonization of healthy donor microbiota (Van Nood et al.,
2013). However, our analysis suggested bowel lavage prior to FMT did
4.1. Donor screening not contribute to the efficacy of recurrent or refractory CDI, which re-
quires more high quality RCTs to determine the optimal conclusion.
Although patients are willing to consider the inherently unappealing The use of capsule or enema is less invasive, does not require
nature of FMT as a treatment alternative recommended by a physician specialized, costly devices and can be performed outside a care facility
(Zipursky et al., 2012), it remains a major issue that the known and (Lee et al., 2016; Youngster et al., 2014a). Last but not least, no matter
unknown risks for the procedure. It is inevitable to completely avoid what delivery method is used, frozen feces can be prepared (e.g. donor
communicable disease transmission despite rigorous screening screening, feces testing) in advance to be in favor of emergency usage. It
procedure for contagious disease is widely performed. Although can also reduce frequency of donor screenings and testing, and may
prospective donors are needed to screen risk factors for infectious help decrease potential transmission of pathogens. It is mainly based
diseases; however, the risks are still less clear, especially screening on patient's state of an illness, clinicians' experience as well as patient
protocols are not standardized (e.g. how often needs to carry out a preference to select a suitable FMT delivery currently.
screening tests and which items may not need to repeat for the same
donor next time). It may be helpful for reducing some infectious 4.4. Limitations of this review
diseases to freeze stool samples, which still requires more high quality
research to support as yet. One of the limitations of this review is the rigorous criteria for the
patients. Diagnosis of CDI must contain presence of diarrhea and
4.2. Adverse events laboratory examination mentioned in original article. So this may result
in fewer probably appropriate studies included. However, it is
According to our review of included studies, only some slight recommended that only patient with diarrhea should be tested for
adverse events attributed to frozen FMT and fresh FMT were reported, Clostridium difficile (Surawicz et al., 2013). Symptomatic diarrhea is
which suggested that FMT is a promising intervention for recurrent or necessary for recurrent CDI with or without stool test, for it often
refractory CDI. However, serious adverse events (such as bacteremia, takes treated patient several weeks to months to shed spores (Dupont
perforations and even death) cannot be ignored, although it was et al., 2016). Another limitation is the restriction to English-language
uncommon (Baxter and Colville, 2016). Moreover, the literature is articles, which may decrease suitable studies. Only an appropriate RCT
short of long-term (for example, five years, ten years or longer) were located that directly comparing the efficacy of frozen FMT and
follow-up adverse events, which have been spontaneously reported fresh FMT, besides, the rest five included studies with a smaller sample
instead of actively sought. It is undoubtedly that this can underestimate size provided a smaller weighted proportion. They both suggested that
the risks related. So it should be took into consideration when performs more high quality RCTs are needed for further confirmation.
 326
Table 3

G. Tang et al. / Diagnostic Microbiology and Infectious Disease 88 (2017) 322–329

Study characteristics.

Author (reference) Year Country Sample Age Female Strain type CDI type Pre-FMT therapy Donor FMT style delivery method colonic
size (mean ± SD/range) gender BI/027 No. (recurrent, (patient selected, (frozen or lavage
No. (%) (%) refractory, unrelated donors, fresh) (yes or no)
both) both)

Lee et al. 2016 Canada 91a 87b 72.2 ± 15.9a 58 (63.7%)a 12 (41.4%)a both suppressive antibiotics unrelated donors both enema no
72.9 ± 15.4b 54 11 (39.3%)b were discontinued 24 to
(62.1%)b 48 hours prior to FMT
Cammarota et al. 2015 Rome 20 71(29–89) 12 (60%) not reported recurrent vancomycin (125 mg both fresh colonoscopy yes
four
times a day for 3d until
2 days before FMT
Youngster, Russell 2014 U.S. 20 64.5(11–89) 9 (45%) not reported both antibiotic regimen for unrelated donors frozen Oral capsule no
et al. CDI was discontinued
48 h prior to IMT
Zainah et al. 2015 U.S. 14 73.4 ± 11.9 9 (64.3%) not reported refractory antibiotic regimen for both fresh NGT, no
CDI was discontinued but one via
24 h prior to IMT colonoscopy
Youngster, Sauk et al. 2014 U.S. 20 54.5 ± 24.6 11 (55%) not reported both discontinue all Unrelated donors frozen colonoscopy yes
antibiotics and NGT
at least 48 hours prior
to the procedure
Van Nood et al. 2013 U.S. 16 73 ± 13 8 (50%) 3 (23%) recurrent 500 mg vancomycin Unrelated donors fresh Nasoduodenal yes
4 times a day for 4 or tube
5 days until the day
before FMT
a
Frozen FMT group; b Fresh FMT group.
NGT: nasogastric tube.
 G. Tang et al. / Diagnostic Microbiology and Infectious Disease 88 (2017) 322–329 327

Fig. 2. Meta-analysis plot of pooled first effective rate comparing frozen FMT with fresh FMT.

Fig. 3. Meta-analysis plot of pooled second effective rate comparing frozen FMT with fresh FMT.

5. Conclusions studies are still needed to standardize this hopeful procedure and more
attentions should be paid to the rare or long-term adverse events.
Among patients with recurrent or refractory CDI, frozen FMT is as
effective as fresh FMT in clinical resolution of diarrhea. In consideration Funding
of the potential advantages of performing frozen FMT, it is a reasonable
option to select frozen FMT. Additionally, colonic lavage before FMT may This research did not receive any specific grant from funding
not contribute to clinical resolution of diarrhea. At the same time, more agencies in the public, commercial, or not-for-profit sectors.
328 G. Tang et al. / Diagnostic Microbiology and Infectious Disease 88 (2017) 322–329

Conflicts of interest Kelly CR, de Leon L, Jasutkar N. Fecal microbiota transplantation for relapsing Clostridium
difficile infection in 26 patients: methodology and results. J Clin Gastroenterol 2012;
46:145–9. http://dx.doi.org/10.1097/MCG.0b013e318234570b.
None. Khoruts A, Dicksved J, Jansson JK, Sadowsky MJ. Changes in the composition of the human
fecal microbiome after bacteriotherapy for recurrent clostridium difficile-associated
diarrhea. J Clin Gastroenterol 2010;44:354–60. http://dx.doi.org/10.1097/MCG.
References 0b013e3181c87e02.
Khoruts A, Hamilton MJ, Weingarden A, Sadowsky MJ. Treatment of C. difficile by fecal
Allegretti JR, Korzenik JR, Hamilton MJ. Fecal microbiota transplantation via colonoscopy transplantation. Gastrointest Endosc 2012;75:AB329.
for recurrent C difficile infection. J Vis Exp 2014;8. http://dx.doi.org/10.3791/52154. Kociolek LK, Gerding DN. Breakthroughs in the treatment and prevention of Clostridium
Arkkila PE, Uusitalo-Seppälä R, Lehtola L, Moilanen V, Ristikankare M, Mattila EJ. Fecal difficile infection. Nat Rev Gastroenterol Hepatol 2016;13:150–60. http://dx.doi.org/
bacteriotherapy for recurrent clostridium difficile infection. Gastroenterology 2010; 10.1038/nrgastro.2015.220.
138:S5. Lee CH, Belanger JE, Kassam Z, Smieja M, Higgins D, Broukhanski G, et al. The outcome
Arkkila PE, Mattila E, Kainulainen V, Satokari R. Simple and practical frozen preparation and long-term follow-up of 94 patients with recurrent and refractory Clostridium dif-
for transplantation of fecal microbiota for recurrent clostridium difficile infection. ficile infection using single to multiple fecal microbiota transplantation via retention
Gastroenterology 2014;146:S193–4. enema. Eur J Clin Microbiol Infect Dis 2014;33:1425–8. http://dx.doi.org/10.1007/
Aroniadis OC, Brandt LJ, Greenberg A, Borody T, Kelly CR, Mellow M, et al. Long-term fol- s10096-014-2088-9.
low-up study of fecal microbiota transplantation for severe and/or complicated clos- Lee CH, Steiner T, Petrof EO, Smieja M, Roscoe D, Nematallah A, et al. Frozen vs fresh fecal
tridium difficile infection: a multicenter experience. J Clin Gastroenterol 2016;50: Microbiota transplantation and clinical resolution of diarrhea in patients with recur-
398–402. http://dx.doi.org/10.1097/MCG.0000000000000374. rent Clostridium difficile infection: a randomized clinical trial. JAMA 2016;315:142–9.
Bagdasarian N, Rao K, Malani PN. Diagnosis and treatment of clostridium difficile in http://dx.doi.org/10.1001/jama.2015.18098.
adults: a systematic review. JAMA 2015;313:398–408. http://dx.doi.org/10.1001/ Listed NA. Fecal microbiota transplantation for treating recurrent Clostridium difficile in-
jama.2014.17103. fection. Manag Care 2013;22:18–9.
Baxter M, Colville A. Adverse events in faecal microbiota transplant: a review of the liter- Lubbert C, John E, von Muller L. Clostridium difficile infection: guideline-based diagnosis
ature. J Hosp Infect 2016;92:117–27. http://dx.doi.org/10.1016/j.jhin.2015.10.024. and treatment. Dtsch Arztebl Int 2014;111:723–31. http://dx.doi.org/10.3238/
Berg AM, Farraye FA. Duodenal infusion of stool is more effective than vancomycin in pa- arztebl.2014.0723.
tients with recurrent Clostridium difficile. Evid Based Med 2013;18:220–1. http://dx. MacConnachie AA, Fox R, Kennedy DR, Seaton RA. Faecal transplant for recurrent Clostrid-
doi.org/10.1136/eb-2013-101308. ium difficile-associated diarrhoea: a UK case series. QJM 2009;102:781–4. http://dx.
Brandt LJ, Aroniadis OC, Mellow M, Kanatzar A, Kelly C, Park T, et al. Long-term follow-up doi.org/10.1093/qjmed/hcp118.
of colonoscopic fecal microbiota transplant for recurrent clostridium difficile infec- Malani PN, Rao K. Expanded evidence for frozen fecal microbiota transplantation for clos-
tion. Am J Gastroenterol 2012;107:1079–87. http://dx.doi.org/10.1038/ajg.2012.60. tridium difficileinfection a fresh take. JAMA 2016;315:137–8. http://dx.doi.org/10.
Cammarota G, Masucci L, Ianiro G, Bibbò S, Dinoi G, Costamagna G, et al. Randomised clin- 1001/jama.2015.18100.
ical trial: faecal microbiota transplantation by colonoscopy vs. vancomycin for the Mattila E, Uusitalo-Seppala R, Wuorela M, Lehtola L, Nurmi H, Ristikankare M, et al. Fecal
treatment of recurrent Clostridium difficile infection. Aliment Pharmacol Ther 2015; transplantation, through colonoscopy, is effective therapy for recurrent Clostridium
41:835–43. http://dx.doi.org/10.1111/apt.13144. difficile infection. Gastroenterology 2012;142:490–6. http://dx.doi.org/10.1053/j.
Chambers D, R M., Woolacott N. Not only randomized controlled trials, but also case series gastro.2011.11.037.
should be considered in systematic reviews of rapidly developing technologies. J Clin Mellow M, Kanatzar A. Colonoscopic fecal bacteriotherapy in the treatment of recurrent
Epidemiol 2009;62:1253–1260.e1254. http://dx.doi.org/10.1016/j.jclinepi.2008.12.010. clostridium difficile infection-results and follow-up. Am J Gastroenterol 2010;105:
Costello SP, Chung A, Andrews JM, Fraser RJ. Fecal Microbiota transplant for Clostridium S135.
difficile colitis-induced toxic Megacolon. Case Rep Gastrointest Med 2015a;110: Mikkelsen KH, Frost M, Bahl MI, Licht TR, Jensen US, Rosenberg J, et al. Effect of antibiotics
775–7. http://dx.doi.org/10.1155/2015/35629310.1038/ajg.2015.70. on gut Microbiota, gut hormones and glucose metabolism. PLoS One 2015;10,
Costello SP, Conlon MA, Vuaran MS, Roberts-Thomson IC, Andrews JM. Faecal microbiota e0142352. http://dx.doi.org/10.1371/journal.pone.0142352.
transplant for recurrent Clostridium difficile infection using long-term frozen stool is Obi O, Hampton D, Anderson T, Leung P, Abdul MKM, Chandra G, et al. Fecal micro-
effective: clinical efficacy and bacterial viability data. Aliment Pharmacol Ther biota transplant for treatment of resistant C. difficile infection using a standard-
2015b;42:1011–8. http://dx.doi.org/10.1111/apt.13366. ized protocol: a community hospital experience. Am J Gastroenterol 2014;
Dupont AW, Avila MB, Avila NP. Recent advances in the diagnosis and treatment of clos- 109(S629).
tridium Difficile infection, F1000Res 5; 2016. http://dx.doi.org/10.12688/f1000re- Patel NC, Griesbach CL, DiBaise JK, Orenstein R. Fecal microbiota transplant for recurrent
search.7109.1 [eCollection 2016]. Clostridium difficile infection: Mayo Clinic in Arizona experience. Mayo Clin Proc
Ganc AJ, Ganc RL. Fecal microbiota transplantation, by means of push enteroscopy. A 2013;88:799–805. http://dx.doi.org/10.1016/j.mayocp.2013.04.022.
novel endoscopic technique, for the treatment of chronic diarrhea associated with Patel KV, Digby-Bell JL, Goel RM, Henry N, Sanderson JD, Irving PM, et al. Faecal microbi-
clostridium difficile-a pilot study. Gastrointest Endosc 2014;79:AB380–1. ota transplantation: implementing a new treatment for recurrent/refractory clostrid-
Ganc A, Ganc R, Frisoli Jr A, Pasternak J. Fecal transplantation-an original per-oral endo- ium difficile infection using banked stool in a tertiary UK Centre. Gut 2015;64:A23.
scopic technique with a pediatric colonoscope. J Gastroenterol Hepatol 2013;28:115. [http://onlinelibrary.wiley.com/o/cochrane/clcentral/articles/461/CN-01098461/
Ganc AJ, Ganc RL, Reimao SM, Frisoli Junior A, Pasternak J. Fecal microbiota transplant by frame.html].
push enteroscopy to treat diarrhea caused by Clostridium difficile. Einstein (Sao Paulo, Pathak R, Enuh HA, Patel A, Wickremesinghe P. Treatment of relapsing Clostridium difficile
Brazil) 2015;13:338–9. http://dx.doi.org/10.1590/s1679-45082015md3106. infection using fecal microbiota transplantation. Clin Exp Gastroenterol 2013;7:1–6.
Goldenberg SD. Faecal microbiota transplantation for recurrent Clostridium difficile infec- http://dx.doi.org/10.2147/ceg.s53410.
tion and beyond: risks and regulation. J Hosp Infect 2016;92:115–6. http://dx.doi.org/ Postigo R, Kim JH. Colonoscopic versus nasogastric fecal transplantation for the treatment
10.1016/j.jhin.2015.12.004. of Clostridium difficile infection: a review and pooled analysis. Infection 2012;40:
Goudarzi M, Seyedjavadi SS, Goudarzi H, Mehdizadeh Aghdam E, Nazeri S. Clostridium dif- 643–8. http://dx.doi.org/10.1007/s15010-012-0307-9.
ficile infection: epidemiology, pathogenesis, risk factors, and therapeutic options. Fo- Ramsauer B, Konig C, Sabelhaus T, Ockenga J, Otte JM. Fecal microbiota transplantation in
rensic Sci 2014;2014:916826. http://dx.doi.org/10.1155/2014/916826. relapsing clostridium difficile colitis. MMW Fortschr Med 2016;158(Suppl. 4):17–20.
Hamilton MJ, Weingarden AR, Sadowsky MJ, Khoruts A. Standardized frozen preparation http://dx.doi.org/10.1007/s15006-016-8305-y.
for transplantation of fecal microbiota for recurrent Clostridium difficile infection. Am J Ray A, Smith R, Breaux J. Fecal microbiota transplantation for clostridium difficile infec-
Gastroenterol 2012;107:761–7. http://dx.doi.org/10.1038/ajg.2011.482. tion: the ochsner experience. Ochsner J 2014;14:538–44.
Hamilton MJ, Weingarden AR, Unno T, Khoruts A, Sadowsky MJ. High-throughput DNA Rohlke F, Surawicz CM, Stollman N. Fecal flora reconstitution for recurrent Clostridium dif-
sequence analysis reveals stable engraftment of gut microbiota following transplan- ficile infection: results and methodology. J Clin Gastroenterol 2010;44:567–70. http://
tation of previously frozen fecal bacteria. Gut Microbes 2013;4:125–35. http://dx. dx.doi.org/10.1097/MCG.0b013e3181dadb10.
doi.org/10.4161/gmic.23571. Rubin TA, Gessert CE, Aas J, Bakken JS. Fecal microbiome transplantation for recurrent
Higgins JP, Green S. Assessing risk of bias in included studies. http://handbook.cochrane. Clostridium difficile infection: report on a case series. Anaerobe 2013;19:22–6.
org/. [updated March 2011]. http://dx.doi.org/10.1016/j.anaerobe.2012.11.004.
Hirsch BE, Saraiya N, Poeth K, Schwartz RM, Epstein ME, Honig G. Effectiveness of fecal- Satokari R, Mattila E, Kainulainen V, Arkkila PE. Simple faecal preparation and efficacy of
derived microbiota transfer using orally administered capsules for recurrent Clostrid- frozen inoculum in faecal microbiota transplantation for recurrent Clostridium difficile
ium difficile infection. BMC Infect Dis 2015;15:191. http://dx.doi.org/10.1186/s12879- infection – an observational cohort study. Aliment Pharmacol Ther 2015;41:46–53.
015-0930-z. http://dx.doi.org/10.1111/apt.13009.
Ho N, Prasad V. Clostridium difficile diarrhea and fecal transplantation. J Clin Gastroenterol Stollman N, Smith M, Giovanelli A, Mendolia G, Burns L, Didyk E, et al. Frozen encapsulat-
2011;45:742–3. ed stool in recurrent clostridium difficile: exploring the role of pills in the treatment
Kassam Z, Hundal R, Marshall JK, Lee CH. Fecal transplantation via retention enema is ef- hierarchy of fecal microbiota transplant nonresponders. Am J Gastroenterol 2015;
fective for recurrent or refractory clostridium difficile-associated diarrhea. Gastroen- 110:600–1. http://dx.doi.org/10.1038/ajg.2015.81.
terology 2010;138:S207–8. Surawicz CM, Brandt LJ, Binion DG, Ananthakrishnan AN, Curry SR, Gilligan PH,
Kassam Z, Hundal R, Marshall JK, Lee CH. Fecal transplant via retention enema for refrac- et al. Guidelines for diagnosis, treatment, and prevention of Clostridium diffi-
tory or recurrent Clostridium difficile infection. Arch Intern Med 2012;172:191–3. cile infections. Am J Gastroenterol 2013;108:478–98. http://dx.doi.org/10.
http://dx.doi.org/10.1001/archinte.172.2.191. 1038/ajg.2013.4.
Kassam Z, Lee CH, Yuan Y, Hunt RH. Fecal microbiota transplantation for clostridium dif- Tian H, Ding C, Gong J, Wei Y, McFarland LV, Li N. Freeze-dried, capsulized fecal microbi-
ficile infection: systematic review and meta-analysis. Am J Gastroenterol 2013;108: ota transplantation for relapsing clostridium difficile infection. J Clin Gastroenterol
500–8. http://dx.doi.org/10.1038/ajg.2013.59. 2015;49:537–8. http://dx.doi.org/10.1097/MCG.0000000000000330.
 G. Tang et al. / Diagnostic Microbiology and Infectious Disease 88 (2017) 322–329 329

To KB, Napolitano LM. Clostridium difficile infection: update on diagnosis, epidemiology, Yoon SS, Brandt LJ. Treatment of refractory/recurrent C. difficile-associated disease by do-
and treatment strategies. Surg Infect 2014;15:490–502. http://dx.doi.org/10.1089/ nated stool transplanted via colonoscopy: a case series of 12 patients. J Clin
sur.2013.186. Gastroenterol 2010;44:562–6. http://dx.doi.org/10.1097/MCG.0b013e3181dac035.
Van Nood E, Vrieze A, Nieuwdorp M, Fuentes S, Zoetendal EG, De Vos WM, et al. Duodenal Youngster I, Russell GH, Pindar C, Ziv-Baran T, Sauk J, Hohmann EL. Oral, capsulized, frozen
infusion of donor feces for recurrent clostridium difficile. N Engl J Med 2013;368: fecal Microbiota transplantation for relapsing Clostridium difficile infection. JAMA 2014a;
407–15. http://dx.doi.org/10.1056/NEJMoa1205037. 312:1772–8. http://dx.doi.org/10.1001/jama.2014.13875.
Vigvari S, Nemes Z, Vincze A, Solt J, Sipos D, Feiszt Z, et al. Experience with fecal transplan- Youngster I, Sauk J, Pindar C, Wilson RG, Kaplan JL, Smith MB, et al. Fecal microbiota trans-
tation in the treatment of Clostridium difficile infection. Orv Hetil 2014;155:1758–62. plant for relapsing Clostridium difficile infection using a frozen inoculum from unrelat-
http://dx.doi.org/10.1556/oh.2014.30020. ed donors: a randomized, open-label, controlled pilot study. Clin Infect Dis 2014b;58:
Vigvári S, Nemes Z, Vincze Á, Solt J, Sipos D, Feiszt Z, et al. Faecal microbiota transplanta- 1515–22. http://dx.doi.org/10.1093/cid/ciu135.
tion in Clostridium difficile infections. Infect Dis 2015;47:114–6. http://dx.doi.org/10. Zainah H, Hassan M, Shiekh-Sroujieh L, Hassan S, Alangaden G, Ramesh M. Intestinal mi-
3109/00365548.2014.969305. crobiota transplantation, a simple and effective treatment for severe and refractory
Weingarden AR, Hamilton MJ, Sadowsky MJ, Khoruts A. Resolution of severe Clostridium Clostridium difficile infection. Dig Dis Sci 2015;60:181–5. http://dx.doi.org/10.1007/
difficile infection following sequential fecal microbiota transplantation. J Clin s10620-014-3296-y.
Gastroenterol 2013;47:735–7. http://dx.doi.org/10.1097/MCG.0b013e31829004ae. Zipursky JS, Sidorsky TI, Freedman CA, Sidorsky MN, Kirkland KB. Patient attitudes toward
Wise J. Frozen faecal matter works as well as fresh for transplantation in C difficile pa- the use of fecal microbiota transplantation in the treatment of recurrent Clostridium dif-
tients. BMJ 2016;352. http://dx.doi.org/10.1136/bmj.i138. ficile infection. Clin Infect Dis 2012;55:1652–8. http://dx.doi.org/10.1093/cid/cis809.