The protective role of short-chain fatty acids from the development of GvHD

Perturbed microbiota has been associated with the establishment and progression of graft-versus-host disease (GvHD) following hematopoietic stem cell transplantation (HSCT). Furthermore, GvHD severity has been associated with broad-spectrum antibiotic use following HSCT, and reinstatement of healthy gut microbiota is being investigated as a therapeutic approach for treating GvHD.¹ These findings are undisputed, but there remain areas of uncertainty.

Short-chain fatty acids (SCFA), such as acetate, propionate, and butyrate are almost exclusively produced by the gut bacteria and have long been suspected to play a role in GvHD pathogenesis. Preclinical models have shown that butyrate supports intestinal epithelia and alters immune populations in the gastrointestinal tract. Mathilde Payen and colleagues aimed to determine whether changes in the microbiome composition with subsequent reduction of SCFAs contribute towards acute GvHD (aGvHD) onset, while a team led by Kate A. Markey focused on the potential protective role of SCFA from chronic GvHD (cGvHD) progression.^1,2

The microbiome and aGvHD onset¹

Study design

• Patients with leukemia, myelodysplastic syndrome, and lymphoma receiving allogeneic HSCT (allo-HSCT) at Saint-Louis Hospital, Paris, FR, between January 21, 2013 and October 31, 2016 were considered for enrollment
• Stool samples were taken weekly — patients who provided samples at the time of aGvHD onset (−2 /+7 days) were enrolled (n = 35)
• A further 35 patients were selected as a non-GvHD control cohort
• Assessment of patient microbiota was carried out using
  • next-generation sequencing (NGS) to assess fecal bacterial composition and diversity
  • real-time quantitative polymerase chain reaction (RT qPCR) to quantify microbiota composition
  • gas-liquid chromatography was used to determine the total concentrations of SCFAs: acetate, propionate, and butyrate

Results

Patient characteristics

• Patients with aGvHD were classified by gut involvement severity:
  • Mild aGvHD (Stage 1; n = 18)
  • Severe aGvHD (Stage 2/3; n = 17)
• Baseline patient characteristics are presented in Table 1
• Throughout stool collection, antibiotic exposure was similar between patients with and without GvHD

Table 1. Baseline characteristics of study participants¹

Patient characteristic	aGvHD (n = 35)	Control (n = 35)	Total (N = 70)
aGvHD, acute graft-versus-host disease; HLA, human leukocyte antigen
Mean age, years (range)	49.2 (16.4–69.0)	46.7 (18.2–68.9)	47.9 (16.4–69.0)
Male gender, n	22	19	41
Transplant characteristics, %
Myeloablative conditioning	22.8	45.7	34.3
Total body irradiation	20.0	25.7	22.9
Steam cell source, %
Peripheral blood	77.1	74.3	75.7
Bone marrow	14.3	20.0	17.2
Cord blood	8.6	5.7	7.1
HLA match, %
Identical sibling	31.4	48.6	40.0
Matched unrelated	45.7	17.1	31.4
Mismatched unrelated	20.0	34.3	27.2
Mismatched relative	2.9	0.0	1.4
Dead at 1 year, %	45.7	25.7	35.7
aGvHD cause of death, %	68.8	0.0	44.0

 

Microbial alterations

• When compared with patients with mild GvHD and controls, patients with severe GvHD demonstrated a significantly lower
  • bacterial count (p = 0.003)
  • microbiota α-diversity (p = 0.39)

Enteric microbiota

• Fewer patients with severe aGvHD had bacteria of the Blautia coccoides group, an important producer of SCFA
• Total numbers of B. coccoides (cluster XIVa) were significantly lower in patients with severe aGvHD compared with mild GvHD (p = 0.36) and the control arm (p = 0.007)

Microbiome signatures of severe aGvHD

• A significantly reduced number of two SCFA producing families of Firmicutes, Lachnospiraceae and Ruminococcaceae, was observed in patients with severe aGvHD vs the control arm (see Table 2 for details)

Table 2. Microbiome signatures of severe aGvHD¹

Depleted genera	P*	SCFA production
aGvHD, acute graft-versus-host disease; SCFA, short-chain fatty acids *vs control patients † Do not contribute, or contribute negatively, to SCFA production
Lachnospiraceae family Lachnoclostridium Blautia Sellimonas Anaerostipes	0.019 0.020 0.014 0.015	acetate, propionate, butyrate acetate acetate, butyrate †
Ruminococcaceae family Faecalibacterium UBA1819 Flavonifractor	0.011 0.012	propionate, butyrate acetate, propionate, butyrate
Erysipelotrichaceae family Erysipelatoclostridium	0.002	acetate
Streptococcaceae Lactococcus	0.005	†
Enriched genera
Prevotella 9	0.004	†
Stenotrophomonas	0.009	†

Metabolic signatures of aGvHD

• The total amount of the fecal SCFAs, butyrate and propionate, was significantly lower at the time of aGvHD onset and was associated with disease severity (Table 3)
• Butyrate was significantly depleted in patients with mild aGvHD and with severe aGvHD vs control (Table 3)
• Neither bacterial composition nor SCFA concentrations were associated with response to steroid treatment

Table 3. Metabolic signatures of aGvHD vs controls¹

Severe aGvHD vs control	Decrease, %	P
aGvHD, acute graft-versus-host disease; SCFAs, short-chain fatty acids
SCFAs	80.6	0.0003
Acetate	75.8	0.002
Butyrate	94.6	0.001
Propionate	95.8	0.0009
Mild aGvHD vs control	Decrease, %	P
SCFAs	N/A	< 0.05
Butyrate	86.0	0.021
Propionate	N/A	< 0.05

Study conclusions

• Lower bacterial biomass and species diversity was seen in severe aGvHD vs mild aGvHD and control patients
• Depleted levels of SCFAs, specifically acetate, butyrate, and propionate, were seen in patients with severe aGvHD compared with controls
• Reduced level of butyrate was identified as a potential marker of early aGvHD
• Limitations of the study:
  • Low patient numbers
  • The influence of antibiotic administration on microbiota function could not be determined, as all patients received antibiotics both before and during the study follow-up

 The microbiome and cGvHD²

Study design

• Adult patients receiving allo-HSCT at Memorial Sloan Kettering Cancer Center (MSKCC), New York, US, provided stool samples 30 days prior to, and 365 days following transplant
• Patients who developed cGvHD with banked stool samples available (n = 54) were included in this analysis
• A further 171 patients were selected as a control cohort
• 16S ribosomal-RNA amplicon sequencing was performed for stool microbiome sequencing
• Plasma samples were analyzed using a gas chromatography-quadrupole time of flight (GC-QTOF) for metabolomic analysis

Results

Patient characteristics

• Disease characterization is presented in Table 4

Table 4. Disease characterization²

Characteristic	Patients (n = 54; cGvHD, n = 45; control, n = 9)
cGvHD, chronic graft-versus-host disease *% of patients with cGvHD
Disease classification, % cGvHD -Mild* -Moderate* -Severe* Late onset cGvHD Overlap syndrome	42.5 53.7 16.6 14.8 14.8 42.5
Median time to cGvHD onset, days (range)	184 (55–451)

 Microbiota composition analysis

• Similarities were observed between the cGvHD and control arms, including
  • broad spectrum antibiotic exposure
  • microbiota alpha-diversity prior to, during, and 100 days following allo-HSCT
  • rates of preceding aGvHD (peri-d100)
  • microbiota domination by Firmicutes prior to allo-HSCT
  • expansion of the Enterococcus genus in the peri-engraftment period

Peri-d100 analysis

• Of the 225 patients enrolled in this study, 45 had samples available in the 100-day window (cGvHD, n = 9; control, n = 36)
• An analysis of 53 peri-d100 (cGvHD, n = 10; control, n = 43) uncovered that patients who developed cGvHD demonstrated
  • significantly lower concentrations of SCFAs butyrate, propionate, and hexanoate vs controls in plasma samples
  • high abundance of the genus Akkermansia in stool samples
  • low levels of the genus Faecalibacterium, and of butyrate-producing genera, Lachnoclostridium, Clostridium, and Faecalibacterium

Study conclusions

• Increased concentrations of the SCFAs, butyrate, and propionate, and high abundance of SCFA producing bacteria were associated with lower incidence of cGvHD

Overall conclusions

Data from these two studies suggest that depleted levels of certain microbe derived SCFAs are associated with both aGvHD establishment and cGvHD progression. These findings may help to refine novel microbiome restoration treatments, such as fecal transplantation, or SCFA supplementation to reduce the incidence of GvHD amongst patients receiving HSCT.