**Background:** Allogeneic hematopoietic stem-cell transplantation (allo-HSCT) is a key treatment for many hematological malignancies, but transplant-related mortality remains high, driven largely by graft-versus-host disease (GvHD) and infections. Disruption of the intestinal microbiota (IM) from chemotherapy, antibiotics, and artificial nutrition is implicated in GvHD pathogenesis. Fecal microbiota transplantation (FMT) can restore IM diversity and has shown safety in immunocompromised patients, but no randomized studies have evaluated FMT for GvHD prophylaxis. This trial aims to assess whether allogeneic FMT improves outcomes after myeloablative allo-HSCT.
**Methods:** This prospective, open-label, multi-centre, parallel-group, randomised phase-II trial will be conducted in French centers within the SFGM-TC network. A total of 120 patients (60 per arm) aged ≥18 years with hematological malignancies undergoing myeloablative allo-HSCT will be enrolled. Patients are randomised after engraftment (absolute neutrophil count >0.5×10^9/L for 3 consecutive days) to receive either FMT by enema within 4 weeks post-engraftment or no FMT (control). Randomisation is stratified by Disease Risk Index, use of antithymocyte globulin, and centre. FMT consists of 50 g of stool from a healthy unrelated donor diluted in 250 mL of 10% glycerol/NaCl 0.9%, delivered by enema after bowel cleansing. Antibiotics must be stopped ≥72 hours before and 48 hours after FMT. The primary endpoint is GvHD-free relapse-free survival (GRFS) at 1 year, a composite of grade III–IV acute GvHD, systemic therapy-requiring chronic GvHD, relapse, or death. Secondary endpoints include overall and progression-free survival at 1 and 2 years, cumulative incidence of acute and chronic GvHD, infectious complications, FMT safety/tolerance, microbiota composition (16S rRNA sequencing), quality of life (EORTC-QLQ-C30), and haematopoietic reconstitution. Sample size is based on a single-stage Fleming design: 60 evaluable patients per arm to reject a GRFS rate <40% and accept a rate >60% (α=0.05, β=0.10, one-sided). With this design, FMT will be considered acceptable if ≥31 of 60 patients are GRFS event-free at 1 year. Analyses will be intention-to-treat and per-protocol, using Kaplan-Meier, log-rank tests, and Cox models with centre as random effect.
**Key Results:** This is a study protocol; no results are reported. The trial received IRB approval on 27 January 2021 and French national authority approval on 15 April 2021. Recruitment of stool donors begins approximately 3 months before first patient inclusion. The study is registered at ClinicalTrials.gov (NCT04935684).
**Clinical Implications:** If FMT improves GRFS at 1 year, it would provide a novel, low-cost prophylactic strategy to reduce GvHD and other complications after allo-HSCT. The study's microbiota analyses may identify microbial signatures associated with favorable outcomes, potentially guiding donor selection or development of defined microbial consortia. The restrictive donor screening and frozen stool bank approach aim to ensure safety and reproducibility. Limitations include the open-label design (though the primary endpoint is objective) and restriction to myeloablative conditioning and peripheral blood grafts, which may limit generalizability to reduced-intensity regimens or other graft sources.