Table 7.

a. Studies of microbiota changes associated with GVHD in murine models
Intervention/comparison	Outcome (Intervention vs. control)	References
Reduced-intensity allogeneic vs. syngeneic HCT	GVHD accompanied by higher Enterobacteriacea, Bacteroides and Enterococcus spp., and lower Lactobacilli, Clostridia, Bifidobacterium, and Bacillus spp. Higher Escherichia coli associated with GVHD severity and reduced survival Treatment with ciprofloxacin did not affect severity of GI GVHD	Heimesaat, 201084
MHC-disparate and MHC-matched/minor antigen-mismatched allo-HCT +/− donor T-cells	GVHD associated with: Loss of intestinal Paneth cells Reduction in α-defensin expression Lower levels of Firmicutes and Bacteroidetes Intestinal Escherichia coli expansion and dominance Lower microbiota diversity Oral administration of the antibiotic Polymyxin B, active against E. coli: Decreased GVHD severity Decreased GVHD-related mortality	Eriguchi, 201285
MHC-disparate allo-HCT +/− donor T-cells	GHVD associated with: Loss of intestinal microbiota diversity Higher Lactobacillus spp. and Enterobacteriales Lower Clostridiales Administration of ampicillin before HCT: Increased GVHD severity and lethality Higher Enterococcus spp. and Enterobacteriaceae Lower Blautia spp. abundance Administration of Lactobacillus johnsonii prevented ampicillin-induced effects	Jenq, 201276
MHC-disparate allo-HCT using T-cells from specific pathogen-free vs. germ-free donor	T-cell donor microbiota presence or absence did not alter: T-cell differentiation and proliferation GVHD severity GVHD-related mortality	Tawara, 201386
MHC-disparate allo-HCT +/− intestinal helminth infection	Infection with murine nematode Heligmosomoides polygyrus: Lower GVHD severity Preserved GVT effect Increased Treg abundance and improved immune regulation Increased survival of GVHD mice Protective effects of helminthic infection dependent on TGF-²	Li, 201587
MHC-matched/minor antigen-mismatched allo-HCT with anti-anaerobic antibiotics (imipenem-cilastin or piperacillin-tazobactam) vs. antibiotics lacking in anti-anaerobic activity (aztreonam)	Treatment with anti-anaerobic antibiotics associated with: Higher mortality Higher severity of GI GVHD Increased GI inflammatory infiltration Higher levels of IL-23 (mediator of GVHD) Greater abundance of mucin-degrading Akkermansia	Shono, 201688
MHC-disparate vs. syngeneic allo-HCT +/− intragastric lavage of 17 butyrate-producing strains of Clostridia spp. Anti-anaerobic antibiotics +/− Clostridia spp.	Gavage with Clostridia spp. resulted in lower GVHD severity and higher survival Anti-anaerobic antibiotics followed by gavage by Clostridia spp. replicated these findings	Mathewson, 201689
MHC-disparate allo-HCT + levofloxacin +/− clindamycin Clindamycin +/− anti-inflammatory Clostridia spp. (AIC)	Treatment with clindamycin decreased survival Clindamycin treatment + AIC increased survival	Simms-Waldrip, 201790

b. Microbiota changes associated with GVHD in clinical studies
Intervention or observational group	Control	Outcomes (Intervention v Control)	References
Randomized trials
Ciprofloxacin with metronidazole prophylaxis (n=68)	Ciprofloxacin prophylaxis (n=66)	Lower incidence and severity of acute GVHD No difference in chronic GVHD or OS	Beelen, 199922
Observational trials
Complete GI decontamination (n=40)	Selective GI decontamination (n=18)	Lower incidence of acute and chronic GVHD Lower rate of infection Lower combined TRM and chronic GVHD	Vossen, 199026
Sustained suppression of anaerobic intestinal flora (n=41)	Incomplete suppression of anaerobic intestinal flora (n=153)	Lower incidence of acute GVHD	Beelen, 199227
Acute GVHD (n=8)	No GVHD (n=10)	Lower intestinal microbiota diversity Higher Lactobacillales Lower Clostridiales	Jenq, 201276
Successful GI decontamination (n=57)	Unsuccessful GI decontamination (n=55)	Lower incidence of acute GVHD Lower infectious risk	Vossen, 201429
Lowest GI microbial diversity at engraftment (n=34)	Intermediate or high GI microbial diversity at engraftment (n=20 intermediate, n=26 high)	Lower OS Higher TRM, specifically mortality related to GVHD or infection	Taur, 20141
Colonized with Candida in the intestine (n=54)	Not colonized with Candida in the intestine (n=99)	Higher incidence of acute GVHD	van der Velden, 201391
Acute GI GVHD (n=8)	No GI GVHD (n=23)	On the day of transplant: Higher Enterococcus spp. Lower Clostridia spp. and Eubacterium rectale	Holler, 20143
Acute GVHD (n=5)	No GVHD (n=5)	Prior to transplant: Lower diversity Lower Bacteroides and Parabacteroides spp. Lower proprionate and SCFAs Between day 0 and day+35: Higher Enterococcus spp. Lower Faecalibacterium spp. Over all timepoints: Lower Bacteroidetes	Biagi, 201592
Lower microbial diversity (n=32), lower Blautia abundance (n=58)	Higher microbial diversity (n=32), higher Blautia abundance (n=57)	Lower microbiota diversity and lower abundance of Blautia spp. associated with higher GVHD-related mortality Higher Blautia abundance associated with higher OS	Jenq, 201548
Colonized with antibiotic resistant bacteria (ARB) pre-transplant (n=33)	Non-ARB colonized (n=74)	Higher incidence of acute GVHD and acute GI GVHD Higher rate of bacteremia, increased infection-related mortality Higher NRM Lower OS	Bilinksi, 201693
Treatment of febrile neutropenia with antibiotics effective against anaerobic bacteria: Imipenem-cilastin (n=148) or Piperacillin-tazobactam (n=300)	Treatment with antibiotics less effective or ineffective against anaerobic bacteria: Cefepime (n=152) Aztreonam (n=64)	Higher risk of 5-year GVHD-related mortality No difference in OS	Shono, 201688
Acute GVHD (n=6)	No GVHD (n=9)	GVHD associated with: Higher cumulative antibiotic exposure and anti-anaerobic antibiotic exposure, specifically clindamycin Higher Enterobacteriacea, Enterococcus spp., and Neisseriaceae Lower anti-inflammatory Clostridia, Bacteroidetes, and Actinobacteria Depletion of Ruminococcus and Blautia spp.	Simms-Waldrip, 201790
Pre-transplant antibiotic prophylaxis or treatment (n=239)	No pre-transplant antibiotics (n=261)	Higher incidence and severity of acute GVHD and GI GVHD Lower median and 10 year OS	Routy, 201794
Early pre-transplant antibiotics (n=236)	Late post-transplant antibiotics (n=297) or no antibiotics (n=88)	Lower urinary indoxyl sulfate Lower Clostridiales Higher acute and/or chronic GVHD-related mortality Higher TRM Higher TRM with ciprofloxacin/metronidazole vs. rifaximin Lower OS	Weber, 201754
Pre-conditioning low microbial diversity (n=18)	Pre-conditioning intermediate (n=48) and high diversity (n=41)	Before conditioning: No difference in incidence of acute GVHD or GI GVHD Higher Firmicutes, and a non significant trend toward lower Bacteroidetes in those who later developed aGVHD No difference in OS, relapse, or NRM	Doki, 201795
Severe acute GI GVHD (n=14)	Non-severe acute GI GVHD or no GVHD (n=52)	At the time of engraftment: Lower intestinal microbiota diversity Higher oral Actinobacteria and Firmicutes Lower Lachnospiraceae Higher gradient of positively to negatively correlated organisms	Golob, 201749
Acute GI GVHD (n=26)	No GI GVHD (n=18)	Longitudinal microbiome sampling: Picobirnaviruses predictive of severe enteric GVHD occurrence Higher picobirnaviruses before or within a week after transplant Increased rate of detection and number of sequences of persistent DNA viruses over time No difference in overall richness Reduced microbial phage richness over time	Legoff 201750
Acute GVHD (n=34) No GVHD (n=23)	HLA-matched sibling donors (n=22)	Before conditioning: Lower recipient intestinal microbiota diversity compared to HLA-matched sibling donors High donor intestinal microbiota diversity is associated with lower aGVHD incidence in recipients Low pre-conditioning intestinal microbiota diversity (in recipients) was not associated with higher risk of aGVHD Lower recipient diversity associated with lower OS	Liu, 201796