Restoration of gut microbial diversity by donor FMT could resolve GI GvHD

There is an urgent medical need to evaluate new safe and effective treatment approaches for gastrointestinal graft-versus-host disease (GI GvHD) dependent or refractory to steroid treatment. To date, acute GI GvHD is still considered one of the major causes of mortality for patients who undergo an allogeneic hematopoietic stem cell transplant (allo-HSCT).¹

The critical role of gut microbiota in GI GvHD and the potential of donor fecal microbiota transplantation (FMT) has been previously suggested based on clinical data. However, the high risk of infection in these immunocompromised patients, and the possibility of transferring multidrug-resistant organisms to them, has obstructed advances in the field.

Yannouck F. van Lier and colleagues from Amsterdam UMC recently published their experience in the journal Science Translational Medicine, and here we provide a summary of their findings.¹

Study design

This is a single-center, single-arm trial to evaluate the efficacy and safety of a single administration of donor FMT via nasoduodenal infusion in adult patients who underwent allo-HSCT and experienced steroid-dependent or steroid-refractory intestinal GvHD (FARAH study, ISRCTN14530574). All patients received a bowel lavage 2 hours before donor FMT, in order to give the transplanted microbiota space to engraft.

Key selection criteria

• Histologically confirmed acute or late-onset acute GI GvHD with no alternative treatment options and
  • Steroid dependence, defined as the reappearance of GvHD symptoms (diarrhea and/or fecal incontinence) when steroid dose is reduced, or
  • Steroid refractoriness, defined as not responding to prednisolone 2 mg/kg/day, no improvement in GvHD grade after a 7-day treatment, or worsening ≥ 1 grade within 72 hours after starting high-dose prednisolone therapy.
• Patients who received other GvHD medication within 3 days after donor FMT were excluded from follow-up.

Sample collection and processing

• Fecal samples were collected before (baseline) and at 1, 4, 12, and 24 weeks after FMT.
• Donor samples were collected at baseline and throughout the study to evaluate the composition changes in their fecal microbiota.
• Bacterial composition was determined by 16S ribosomal RNA sequencing in all collected stool samples, and α-diversity, Shannon index, and phylogenetic diversity were assessed.

Study endpoints

• Primary efficacy endpoint: complete response (CR) to treatment, defined as complete resolution of GvHD symptoms on Week 4 after donor FMT.

• • Partial response (PR): improvement of at least one grade after 4 weeks of donor FMT.
  • Non-responders (NR): no improvement of GvHD or progression (increase by at least one grade) 4 weeks after FMT.
  • Complete responders with secondary failure (CRsf): patients who achieve a CR, initiate steroid tapering, and GvHD symptoms reappear.

• Safety endpoint: incidence of adverse events during the 4 weeks after donor FMT.

Results

A total of 15 patients were enrolled and evaluable for treatment response. Baseline characteristics (at the time of donor FMT) can be found in Table 1.

Table 1. Baseline characteristics¹

Characteristics at baseline	Value (N = 15)
*cyclosporin, mycophenolate mofetil, methotrexate, ruxolitinib, sirolimus, or tacrolimus. FMT, fecal microbiota transplantation; GI, gastrointestinal; GvHD, graft-versus-host disease; MF, myelofibrosis.
Median age (range), years	57 (20–72)
FMT indication
Steroid-refractory	6
Steroid dependent	9
Diagnosis
Acute myeloid leukemia	8
Myelodysplastic syndrome	3
Myeloproliferative disorder (MF)	2
Lymphoma	2
Conditioning regimen received
Reduced-intensity	12
Myeloablative	3
GI GvHD stage
I	7
II	3
III	3
IV	2
Additional treatment to systemic steroids
Local therapy (budesonide)	8
Systemic therapy*	12
GvHD involvement of other organs
Skin	3
Liver	1

Donor FMT was well tolerated

The procedure was considered safe for all participants due to a low incidence of adverse events, and all infections reported being considered not related to the FMT.

• During administration, patients experienced discomfort, transient abdominal distention, cramps, nausea, and regurgitation. All symptoms resolved without any clinical intervention.
• All patients were treated at baseline with prophylactic antibiotics, antivirals, and antifungal therapy in some cases.
• Bacterial infections were reported in five patients within 4 weeks after donor FMT, but these were considered not related to the procedure due to site of infection, the long period from FMT until infection manifestation, and absence of the pathogen in the stool samples.
  • Of note, four patients had active infections and received treatment at the time of FMT.
• No fungal or viral infection was reported in this timeframe.

Donor FMT was effective in two thirds of patients enrolled

Overall, ten of 15 patients were in CR at 4 weeks after donor FMT, showing improved GvHD symptoms during the first and second week after the procedure. Successful immunosuppressant tapering was possible in six patients. Find a summary of treatment responses in Table 2.

Table 2. Summary of donor FMT response¹

Response	Number of patients (N = 15)
*All patients received antibiotic therapy after donor FMT, in addition to the prophylactic treatment. †All patients were non-responders to donor FMT. CR, complete response; GI, gastrointestinal; GvHD, graft-versus-host disease.
CR at 4 weeks	10
CR of steroid-refractory GI GvHD	3
CR of steroid dependent GI GvHD	7
CR with secondary failure at 24 weeks	4*
Partial response	0
Non-responders	5
Died during follow-up	4†

 

• A higher probability of response was associated with a lower stage of GvHD at the time of FMT (p = 0.044).
• Achievement of complete response to FMT was associated with better outcome (p = 0.0068).

The response was associated with gut microbiota α-diversity and the presence of beneficial bacteria

The investigators reported that patients with higher gut microbial diversity at baseline were more likely to obtain a CR following FMT. Patients who responded to treatment showed a progressive improvement in microbiota composition which developed similarities to the donor sample. This change was directly associated with a decrease in diarrhea episodes.

Moreover, data suggest that higher diversity might not be the only factor influencing response, as also the presence of some specific bacterial species was linked with better outcomes.

• Precisely, a higher proportion of Blautia at baseline and rising numbers of Clostridiales and butyrate-producing bacteria after FMT were associated with CR and CRsf.
• On the contrary, non-responders showed stabilization or even a decline of these species.

Finally, the authors identified fucosyltransferase 2 (FUT2) secretor status as an additional prognostic factor for response. FUT2 is an enzyme found in the gastrointestinal epithelium crucial for the secretion of ABO blood group antigens into the gut lumen, where it constitutes a significant source of carbohydrates for the microbiota potentially improving engraftment of the donor fecal microbiota and maintaining its diversity. In this study, eight patients with a CR were considered FUT2 secretors (as detected by epithelial staining in gut biopsies); while two out of four non-responders presented with a non-secretory phenotype.

Conclusions

In line with previously published reports, the FARAH study results show the promising potential of donor FMT to improve acute GI GvHD safely and rapidly in a significant proportion of patients with steroid-refractory or dependent disease. However, these findings need to be confirmed in a larger clinical trial.

Additionally, the future implementation of donor FMT will also need to be accompanied by a better understanding of factors influencing engraftments, new approaches to enhance microbiota diversity, and tools to identify those patients who will benefit the most from this procedure.

Some questions remain unanswered and could also be addressed in future studies:

• How long will a response to donor FMT last?
• Would non-responders and patients with CRsf benefit from multiple administrations of donor FMT?²
• Does the use of antibiotics during the first week after FMT impact treatment response?
• Can close monitoring and guidance on patients’ diets improve the results of donor FMT?