Third-party fecal microbiota transplantation following allo-HCT reconstitutes microbiome diversity

Zachariah DeFilipp, Jonathan U. Peled, Shuli Li, Jasmin Mahabamunuge, Zeina Dagher, Ann E. Slingerland, Candice Del Rio, Betsy Valles, Maria E. Kempner, Melissa Smith, Jami Brown, Bimalangshu R. Dey, Areej El-Jawahri, Steven L. McAfee, Thomas R. Spitzer, Karen K. Ballen, Anthony D. Sung, Tara E. Dalton, Julia A. Messina, Katja Dettmer, Gerhard Liebisch, Peter Oefner, Ying Taur, Eric G. Pamer, Ernst Holler, Michael K. Mansour, Marcel R. M. van den Brink, Elizabeth Hohmann, Robert R. Jenq and Yi-Bin Chen

Data supplements

Article Figures & Data


  • Figure 2.

    Study flow diagram. IBD, inflammatory bowel disease.

  • Figure 3.

    Longitudinal changes in urinary 3-IS levels before allo-HCT and after FMT administration in the early posttransplant period.

  • Figure 4.

    Evaluation of the microbiome prior to and following administration of FMT. Longitudinal changes in (A) inverse Simpson index, (B) Clostridiales abundance, and (C) UniFrac distance before allo-HCT and after FMT administration in the early posttransplant period, as determined from 16S rRNA sequencing of stool specimens. (D) Longitudinal changes in the origin of operational taxonomic units. Four selected patients shown.

  • Figure 5.

    Microbiome assessment under conditions that highlight the potential benefit of FMT as part of an exploratory analysis in a subset of 8 patients. Eligibility criteria for inclusion in this subset analysis received microbiome-disrupting antibiotics (ceftazidime, cefepime, piperacillin-tazobactam, meropenem, oral vancomycin, or metronidazole) before FMT, and did not receive microbiome-disrupting antibiotics in the 60 days after FMT.

  • Figure 6.

    Microbiome diversity after FMT compared with a post hoc comparison cohort of patients with allo-HCT who did not receive FMT. The inverse Simpson scores of 8 FMT recipients from Figure 5 are compared with those of patients who underwent allo-HCT at 2 other institutions and had stool specimens collected and 16S sequenced in a manner identical to those from FMT recipients.


  • Table 1.

    Baseline characteristics

    Median age (range), y63 (26-71)
    Sex, female/male, n7/6
    Diagnosis, n
     Acute myeloid leukemia4
     Myelodysplastic syndrome3
     Non-Hodgkin lymphoma3
     Myeloproliferative disorder1
     Chronic lymphocytic leukemia1
    Graft source, n
     Peripheral blood stem cells12
     Bone marrow1
    Donor, n
     Matched unrelated9
     Matched related2
    Conditioning, n
     Reduced intensity10
    GVHD prophylaxis, n
     Posttransplant cyclophosphamide plus tacrolimus/mycophenolate mofetil2
  • Table 2.

    Clinical outcomes

    Patients receiving FMT, n13
    Median time from allo-HCT to FMT (range), d27 (19-45)
    Median follow-up (range), mo15 (13-20)
    C difficile colitis, n1
    Bacteremia, n1
    Grade 3-4 acute GVHD, n2
    Deaths, n2