The role of the microbiome in the pathogenesis and severity of GvHD

Graft-versus-host disease (GvHD) following allogeneic hematopoietic stem cell transplantation (allo-HSCT) is known to contribute to overall mortality, and is often caused by intestinal inflammation as a result of conditioning regimen and gut barrier defects, allowing bacteria and their metabolites to pass through damaged gut epithelium. Intestinal microbiota is thought to play a role as a modulator in the pathophysiology of GvHD, via stimulation and differentiation of donor T cells or consequent innate immune-inflammatory mediators. It may be critical to elucidate mechanisms to alter the microbiota to prevent GvHD or reduce its severity.

Sarah Lindner and Jonathan U. Peled have published a review on the most current clinical and mouse studies of the microbiome and allo-HSCT in Current Opinion in Hematology,¹ and here we are pleased to summarize the key findings.

Posttransplant intestinal microbiota

The composition of microbiota is considered to be affected by a number of factors including diet, antibiotics, conditioning therapy, mucositis, GvHD, and infections following allo-HSCT. Disruption of microbiota composition is measured by α-diversity and defined as the number of unique bacterial taxa, and their relative amount, in a fecal sample. An international collaborative study, analyzing 8,767 stool samples, demonstrated that loss of microbiota diversity and dominance of a single taxon (mostly Enterococcus) was common in patients who underwent transplant, and that a high-level of diversity in the peri-engraftment phase was associated with a lower risk of death. Diversity-based GvHD risk score assessments have shown that abundance of Lachnospiraceae represented the strongest factor to predict Grade II–IV acute GvHD (aGvHD), supporting previous findings where members of the Lachnospiraceae family were associated with less GvHD. Overall, these observations indicate that fecal microbiota may allow to predict outcomes after allo-HSCT.

The role of Enterococcus domination

Studies have shown that Enterococcus domination in the intestinal microbiota of patients at an early posttransplant stage was associated with unfavorable outcomes, such as lethal aGvHD and reduced survival. Although the use of antibiotics was considered the main cause for this domination, mouse models have also shown that members of the Enterococcus genus grow in the absence of antibiotics. Other preclinical transplant studies indicate that mice treated with Enterococcus have worse GvHD and die earlier. Lactose is thought to be an important source for growth of these bacteria, and a lactose-free diet was shown to reduce Enterococcus growth and the severity of GvHD in mice. Interestingly, some species of Enterococcus, including E. gallinarum and E. casseliflavus, were associated with longer survival. A lantibiotic-producing strain of Blautia producta has been associated with lower mortality related to GvHD and shown to inhibit the growth of vancomycin-resistant enterococci.

A recent study of whole-genome sequencing has demonstrated that the development of Enterococcus domination has a complex clonal dynamics, and an analysis of metagenomic DNA and RNA sequences revealed that posttransplant microbiota showed differences at strain-level. In addition, in both these studies, there were changes in genes conferring resistance to antibiotics over time.

Posttransplant infection risk and microbiota

Patients undergoing allo-HSCT are known to be at risk for bloodstream infections (BSI), particularly in the neutropenic phase following transplant, and intestinal bacteria have been thought to be the main cause of these infections. One study showed cases of BSI with Escherichia coli and Klebsiella pneumoniae in patients colonized by matching strains of these species in the gut, while another analysis of fecal samples demonstrated an association between Gram-negative intestinal domination and BSI.

Some studies have shown that prophylactic exposure to fluoroquinolones was associated with a lower risk of Gram-negative BSI, but with a higher proportion of intestinal domination and BSI by E. coli, and in particular, with a higher resistance to fluoroquinolones. These observations suggest that antibiotic prophylaxis has a role in shaping microbiota composition. However, these data are conflicting with other findings showing no difference in microbiota composition in fecal samples collected before and after levofloxacin exposure.

The evidence on fungi that are present in the gut (called mycobiota) and the risk of fungal infection is limited. However, there is evidence that the growth of pathogenic Candida spp. is observed before candidemia, and that Candida domination occurs when bacterial diversity is decreased. For example, when B. producta and Bacteroides thetaiotaomicron were used as markers for a healthy bacterial community, there was an inverse correlation with the abundance of fungi, suggesting that there may be an interaction between fungal and bacterial communities, which indicates that a healthy microbiome may protect against posttransplant fungal infections.

Microbiome-derived metabolites

Microbial metabolites take its source from diet, host molecules, and bacteria, and play a role in intestinal homeostasis, energy metabolism, and immunity. Previous investigations found significant differences in the microbial metabolic pathways related to amino acids, lipids, and complex lipid products between patients with and without aGvHD.

The short-chain fatty acid (SCFA) butyrate is a well-searched microbial metabolite, and an abundance of bacteria that produce butyrate has been associated with a lower GvHD incidence in mice, and less GvHD-related mortality and respiratory viral infections in transplanted patients. The comparison of stool samples collected from patients with GvHD at the time of onset and patients without GvHD demonstrated that the butyrate concentration was low at any stage of aGvHD, while the concentrations of other SCFAs, propionate and acetate, were low in severe aGvHD. This observation suggests butyrate concentration as a possible predictive marker for GvHD, and propionate and acetate concentrations at the time of GvHD onset as markers for severity. Butyrate concentrations were also studied in plasma of patients with steroid-refractory GvHD who received human chorionic gonadotropin. Butyrate concentration increased over time in patients who responded to gonadotropin, while there was no change in patients who did not respond. Also, low butyrate and propionate levels after transplant were associated with chronic GvHD (cGvHD). Another study produced somewhat conflicting data, showing that increased levels of at least one butyrate-producing bacterium at 21 days after aGvHD onset was associated with cGvHD and steroid-refractory GvHD. The authors concluded that butyrate may prevent colonic stem cells to form an undamaged epithelial layer during later stages of aGvHD development, while it has protective properties prior to aGvHD.

Transplantation is also known to disrupt oral microbiota which is associated with reduced polyamine levels and increased rates of oral mucositis.

Overall, these observations lay emphasis on the important role of microbiome-derived metabolites in GvHD pathogenesis; however, these mechanisms need further investigation.

Immune-related factors impacting GvHD

Recent mouse studies highlighted that antigen presentation by intestinal epithelial cells to donor T cells plays a role in the pathogenesis of GvHD via the MyD88/TRIF signaling pathway, which suggests an involvement of microbial signals. Another mouse study demonstrates that endogenous dysbiotic microbiota of mice deficient in IL-17RA increases GvHD. While there was no increase in GvHD when wild-type mice were cohabited with IL-17RA deficient mice before transplantation, there was, however, increased GvHD when they were cohoused in a posttransplant setting, with even greater effects when the cohousing period was longer. This may indicate a role of the microbiome in the severity of GvHD.

Clinical trials of microbiome and the pathogenesis of GvHD

Several clinical trials are currently investigating the possible mechanisms by which GvHD can be affected through gut microbiota (Table 1). Major treatment strategies include gut decontamination, the use of prebiotics and probiotics, and optimum use of antibiotics.

Table 1. Clinical trials currently investigating the role of microbiota in GvHD¹

Conclusion

There is an ongoing effort to elucidate the association between intestinal microbiome and posttransplant outcomes, and GvHD pathogenesis. To date, many mouse studies demonstrated a relationship between gut microbiota and GvHD severity, although the exact mechanism, and if it occurs in humans, is yet to be explained. Other areas to be investigated in the future include the role of intestinal microbiome in graft-versus-tumor activity, or the impact of microbial colonies at other body sites on posttransplant outcomes.

The authors also suggest that randomized, multicenter trials are needed to further investigate the role of microbiota-based approaches, as several factors influence intestinal microbiota and clinical outcomes. More mouse studies are warranted to increase knowledge in molecular and biological mechanisms.