Back

Long-term compositional stability of bacterial communities in frozen FMT capsules from a prospective donor cohort

Montenegro Borbolla, E.; Johner, N.; Moser, K.; Gerber, S.; Audry, M.; Ballif, A.; Chen, C.; Guery, B.; Bertelli, C.; Galperine, T.

2026-08-20 infectious diseases
10.64898/2026.08.18.26360593 medRxiv
Show abstract

Background: Faecal microbiota transplantation (FMT) is an effective treatment for recurrent Clostridioides difficile infections, yet the diversity of FMT formulations and delivery routes hampers comparisons across studies. Oral frozen capsules are widely used and current guidelines recommend storage at -80C for up to two years. Despite extensive use of FMT, data on the long-term persistence and maintenance of their microbial composition remains limited. Method: In this prospective study, we assessed the temporal stability of bacterial profiles in frozen FMT capsules derived from 48 donations of 10 healthy donors. Using metabarcoding, we longitudinally profiled one capsule per donation thawed within a month of production and after 3, 6, 12, and 24 months of storage. We used linear mixed effect models to evaluate changes in alpha diversity and community composition over time. Results: Species richness remained stable across all timepoints, whilst species evenness decreased slightly. Although changes in community composition were detectable, they were small and mostly affected low-abundance genera. Clinical efficacy, assessed in a subset of recipients, was not associated with storage duration. Conclusion: Our findings demonstrate that frozen FMT capsules preserve their bacterial community structure for at least two years of storage at -80C, supporting their suitability for long-term biobanking and standardised clinical or research use.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.