Assessment of Faecal Microbiota Transplant Stability in Deep‐Freeze Conditions: A 12‐Month Ex Vivo Viability Analysis

ABSTRACT

Background

Faecal microbiota transplantation (FMT) is an established treatment for Clostridioides difficile infection and is under investigation for other conditions. The availability of suitable donors and the logistics of fresh stool preparation present challenges, making frozen, biobanked stools an attractive alternative.

Aims

This study aimed to evaluate the long‐term viability of bacterial populations in faecal samples stored at −80°C for up to 12 months, supporting the feasibility of using frozen grafts for FMT.

Methods

Fifteen faecal samples from nine healthy donors were processed, mixed with cryoprotectants and stored at −80°C. Samples were assessed at baseline and after 3, 6 and 12 months using quantitative culturing methods to determine the concentration of live bacteria.

Results

Quantitative analysis showed no significant decrease in bacterial viability over the 12‐month period for both aerobic and anaerobic cultures (p = 0.09). At all timepoints, the coefficients of variability in colony‐forming unit (CFU) counts were greater between samples (102 ± 21% and 100 ± 13% for aerobic and anaerobic cultures, respectively) than the variability between measurements of the same sample (30 ± 22% and 30 ± 19%).

Conclusions

The study confirmed that faecal microbiota can be preserved with high viability in deep‐freeze storage for up to a year, making allogenic FMT from biobanked samples a viable and safer option for patients. However, a multidonor approach may be beneficial to mitigate the risk of viability loss in any single donor sample.

1. Introduction

Faecal bacterial transplantation (FBT) is a method involving the transfer of minimally modified donor stool into the digestive tract of the recipient. The goal is to modify intestinal dysbiosis and mitigate its consequences. This method is generally recognised as a therapeutic alternative for the treatment of recurrent or severe Clostridioides difficile colitis [1, 2, 3]. In other indications, it is used experimentally [4, 5, 6, 7], and the current research suggests that its range of applications may expand in the future. One of the limitations in conducting FBTs is the availability of a suitable related or unrelated donor who must undergo a lengthy and detailed examination, with not every candidate meeting the strict selection criteria and not every patient affording to wait until this happens. The logistics of administering a preprepared preparation are also significantly simpler than preparing fresh stool from an ad hoc donor. The use of biobanked frozen stool samples from vetted donors might be the way to make FBT feasible and immediately accessible to most patients who need it. In addition, faecal microbiota transplantation (FMT) safety is increased by quarantining the frozen preparation, that is, releasing it for use only after the donor's second examination (after the potential immunological window) instead of only after a one‐time examination is performed on fresh stool donors. In turn, quarantining the biobanked stool samples allows for the identification of any pathogens that may have been present but undetected at the time of the initial examination, thereby reducing the risk of transmitting infectious diseases through FMT.

Clinical outcomes seem to be noninferior to those of fresh stool (assuming adherence to the procedures described below) for donor stool stored deep frozen for up to 2 years [8, 9]. There is a degree of variability in standard operating procedures for processing and storing frozen samples across institutions [10, 11, 12, 13, 14] and legislative frameworks [15, 16, 17, 18], although these protocols generally agree on basic points, which include timely processing, the use of cryoprotectants and adherence to a storage temperature of −80°C [19]. Based on these principles, we developed a detailed standard operating procedure for the selection and examination of donors and their stool, the process of processing the donor's stool, the method of preservation, the process of thawing and preparation for administration [20]. The approval process of these standard operating procedures by the State Institute of Drug Control (SÚKL) unveiled the lack of data on the stability of the graft during cold storage. In this study, we aimed to use a quantitative culturing method and then compare the quantity of living bacteria in fresh stool and after 3, 6 and 12 months of storage in deep‐freezing conditions.

2. Methods

2.1. Donor Recruitment and Sample Processing

The stool donors (n = 9) were healthy human volunteers aged 34.9 ± 5.7 (range 27–42) years with a body mass index of 23.6 ± 2.9 kg/m² who were enrolled after advertisement on the Internet as part of the FEBATRICE project registered at www.clinicaltrials.gov (NCT05430269). All the subjects gave prospective, written informed consent, and The Ethics Committee for Multi‐Centric Clinical Trials (EC) of FNKV University Hospital approved the project on 2 June 2021, decision reference no. KH/40/00/2021. The conduct of the study including informed consent procedure is compliant with the latest Amendment of the Declaration of Helsinki (ver. 2013). The selection and assessment of donors were performed as previously described [20]. The donors were provided with a sterile faecal collector container (FECONTAINER; Excretas Ltd., The Netherlands) and were instructed to bring the stool as soon as possible but no later than within 3 h after defecation. The donor stool was processed according to published standard operating procedures [20]. Of the nine donors, two donated twice and two donated three times, yielding a total of 15 processed faecal samples. All chemicals were purchased from Sigma‐Aldrich, unless otherwise specified. In brief, the donor's stool was mixed with a titrated (vide infra) amount of sterile normal saline (0.9% NaCl), homogenised with a stick mixer, filtered and mixed with the cryoprotectant 80% glycerol (10% vol/vol). The density of the resulting homogenate was determined by a future method of administration (enema); therefore, we aimed to obtain homogenous, viscous semiliquid matter. At this stage, samples were taken from the homogenate for quantitative culture. The first sample was processed immediately after collection (at baseline, labelled Time 0 months), and the next four were frozen at −80°C and processed after 3, 6 and 12 months.

2.2. Viability Testing

In the preliminary experiments (data not shown), we established the optimal dilution and culture methodology. Our finalised protocol involved taking 300 μL (0.3 mL) of the homogenate, which was subsequently measured for weight to verify the accuracy of the pipette sampling, and diluting it in 3 mL of normal saline. This dilution process was repeated four additional times. Aliquots of 10 μL from the second to the fifth dilutions were inoculated on blood agar for aerobic cultivation and on Schaedler agar (Thermo Fisher Scientific, catalogue number CM0437B) for anaerobic cultivation, which was conducted using a Thermo Scientific anaerobic chamber system, with a gas composition of 85% nitrogen, 10% hydrogen and 5% carbon dioxide to create an oxygen‐free environment. The Schaedler agar plates were incubated at 37°C for 48 h within this system. This apparatus ensures a stable anaerobic condition, which is essential for the growth of strictly anaerobic bacteria from the faecal samples. The number of colony‐forming units (CFUs) that developed was then calculated relative to 1 mL of the original homogenate. When the initial sample was weighed, the CFU counts were also adjusted relative to 1 g of the original homogenate. To obtain data on the accuracy and reproducibility of the entire process, all the samples were processed three times from one sample of homogenate, three aliquots were taken, diluted and cultured in the abovementioned way at all the processing times (0, 3, 6 and 12 months). All tests were performed in triplicate.

2.3. Statistical Methods

CFU counts, which serve as a proxy for live bacterial quantity, were logged at each time point. To account for the nested data structure and potential nonindependence of the measurements, a mixed‐effects linear regression model was employed. This model included time as a fixed within‐subjects factor and culture condition and sample number as random between‐subjects factors. The model was adjusted for repeated measures by assigning a unique identifier to each sample, allowing for the assessment of within‐sample and between‐sample variability over the storage period. The statistical significance of the fixed effects was evaluated using Wald tests, and the variance components for random effects were estimated to assess the variability attributed to differences between samples. A significance level of α = 0.05 was used for all tests. All the statistical analyses were performed using Python (MixedLM package), and the convergence of the mixed‐effects model was confirmed. Coefficients of variability were calculated as the standard deviation/mean × 100%.

3. Results

Our quantitative analysis of faecal microbiota viability in frozen grafts over 12 months revealed a significant difference between aerobic and anaerobic conditions (p = 0.001) and between stable bacterial counts for both aerobic and anaerobic cultures, with no significant time effect (p = 0.09); see Table 1 and Figure 1.

TABLE 1.

Colony‐forming unit (CFU)/mL counts serving as a proxy for live bacterial quantity over time.

Time	Aerobic culture (106 CFU/mL)	Anaerobic culture (108 CFU/mL)
Baseline (fresh stool)	6.8 (95% CI 4.9–8.8)	3.1 (95% CI 2.3–3.9)
	Range 0.1–19.0	Range 0.1–9.6
3 months	7.4 (95% CI 5.5–9.4)	3.4 (95% CI 2.4–4.4)
	Range 0.1–36.5	Range 0.4–9.5
6 months	6.9 (95% CI 5.0–8.9)	2.4 (95% CI 1.6–3.1)
	Range 0.2–13.7	Range 0.1–9.5
12 months	5.8 (95% CI 3.5–8.2)	4.7 (95% CI 3.3–6.1)
	Range 0.2–10.9	Range 0.1–16.9

Note: Mean value of 15 samples with 95% confidence intervals.

FIGURE 1.

Quantity of alive bacteria in faecal sample stored deep frozen for up to 12 months. Raw data (black crosses) and medians (red horizontal lines). CFU, colony‐forming units.

The triplicate measurements from the same sample demonstrated coefficients of variability of 30 ± 22% and 30 ± 19% for the aerobic and anaerobic cultures, respectively. The variability between the different samples was greater, at 102 ± 21% for aerobic cultures and 100 ± 13% for anaerobic cultures.

4. Discussion

The stability of the faecal microbiota observed over a 12‐month period in deep‐frozen storage, as demonstrated in this study, confirms the viability of using frozen grafts for FMT [8, 9]. This sustained decrease in viability suggested that the therapeutic potential of FMT is not compromised by long‐term storage, lending support to the use of biobanked faecal matter.

This result is important because it makes allogenic FMT from an unrelated donor more feasible and safer, as it allows the graft to be quarantined until a second examination of the donor is performed.

However, the findings on intersample variability are particularly noteworthy. These authors suggested that a multidonor approach to FMT could be advantageous. By administering smaller aliquots from multiple donors, the variability in bacterial viability is less likely to adversely affect therapeutic outcomes. This strategy diversifies the microbial composition provided to the recipient, potentially increasing the chances of engraftment and providing a broader range of beneficial microbes [2]. The safety of the multidonor approach is enhanced by quarantining the sample [14], during which the donor is retested before releasing the graft.

There are several methodological limitations that should also be noted. The observed discrepancies in CFU counts within aliquots of the same sample can be attributed to several factors. A primary concern was the potential inaccuracy in sampling volume due to pipette clogging with the granular homogenate. This was substantiated by weighing the initial samples, which frequently revealed lesser amounts than the anticipated volume of 0.3 mL, assuming a density equivalent to 1 g/cm³ for simplification. Such volume discrepancies are likely to introduce variability in CFU counts. Moreover, when CFU determination was based on the weight of the input sample rather than on the volume, the variability was reduced, suggesting that factors such as the homogeneity of the sample and the precision of manual CFU counting also influence the variability in the results. Imperfect homogenisation could result in an uneven distribution of bacteria within the sample, affecting the reproducibility of the CFU counts. Additionally, manual counting is subject to individual error, which can contribute to the observed variability between aliquots. Finally, we have not attempted to look any deeper into the potential changes in microbial species present in the sample. These methodological considerations are important for refining the standard operating procedures of FMT preparation. It may be beneficial to implement the measures that ensure more accurate volumetric measurements and consider the potential benefits of standardising samples by weight to minimise variability. This approach could lead to more reliable preparation of faecal microbiota grafts and enhance the overall efficacy and safety profile of FMT. Finally, this study did not delve into the dynamics of microbial species diversity during the storage period, which is an important consideration for future research, as the differential survival rates of bacterial taxa could influence FMT outcomes [2, 14, 21].

An alternative to deep freezing is freeze‐drying of faeces, which maintains the viability of the microbial content [22, 23] and can be a practical option for FMT. The use of lyophilised faecal matter could offer several advantages, including ease of storage, reduced transportation costs due to decreased weight and volume, and the potential for long‐term storage without the need for ultra‐low temperatures. Furthermore, lyophilised faecal matter can be encapsulated, which might improve patient acceptance and the ease of administration. The effectiveness of this method, as shown in these studies, suggests that lyophilisation could be a feasible alternative to deep‐freeze storage, providing a stable and viable form of FMT material that is both patient‐friendly and logistically advantageous.

In conclusion, the long‐term stability of the faecal microbiota in frozen grafts allows for the enhancement of the accessibility and practicality of FMT in clinical practice. The minimal changes observed in bacterial viability over a 12‐month period support the use of deep‐freeze storage for faecal microbiota, potentially simplifying the logistics of FMT by enabling the use of biobanked samples and a multidonor approach.

Author Contributions

F. D. contributed to the conception/design of the work. H. S., V. R. and I. C. contributed to the acquisition, analysis or interpretation of data. F. D. drafted the manuscript, to which all authors contributed. All authors approved the version to be published and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Ethics Statement

This study was conducted in accordance with the Declaration of Helsinki (2013 Amendment), and the protocol was approved by the Ethics Committee of FNKV University Hospital, Decision Letter KH/40/00/2021.

Consent

Written informed consent was obtained from all participants.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The datasets generated and/or analysed during the current study are available from the corresponding author upon reasonable request.