Fecal microbiota transplantation in patients with acute and chronic graft-versus-host disease-spectrum of responses and safety profile. Results from a prospective, multicenter study.

Abstract

Allogeneic hematopoietic cell transplantation (alloHCT) is a potentially curative strategy for patients with selected blood diseases. Complications from the procedure comprise most of all infections and graft-versus-host disease (GvHD), which are the major causes of morbidity and mortality (45% of attributable deaths) apart from relapse.1 As the patients with gut GvHD are often colonized with antibiotic-resistant bacteria (ARB), we are presenting our experience performing fecal microbiota transplantation (FMT) in patients with acute (a) or chronic (c) GvHD co-colonized with ARB. Our group first published a prospective study assessing the efficacy of FMT in decolonization of ARB from the gastrointestinal (GI) tract.2 Our aim was to summarize and analyze the impact of FMT performed in patients with acute and chronic GvHD on the ARB decolonization rate and on the outcomes of GvHD treatment. This is, to date, one of two most populous cohorts published describing patients treated with FMT for GI GvHD and, to our knowledge, the first report on FMT in cGvHD. Recent data suggest very good efficacy of FMT performed to treat GI GvHD, as well as the safety of the procedure. In our work, apart from the main end-point being a decolonization rate, safety profile as well as overall response of GvHD were assessed.3 Patients were treated within a prospective, multicenter study according to protocol registered at clinicaltrials.gov (NCT02461199) and published previously2 that aimed to use FMT to decolonize ARB from the GI tract. Each FMT was carried out intraduodenally via a nasoduodenal tube between 2016 and 2020. All participants underwent FMT on two consecutive days. All detailed information about study design, definitions, microbiological, and GvHD assessment and diagnosis are published in the Appendix S1. The Medical University of Warsaw Ethics Committee approved the research protocol (KB/180/2014). A total of 16 FMTs were performed in 11 patients with aGvHD (repeated in 3 patients) and in 2 patients with cGvHD (13 patients overall), aged 23–66 years, suffering from a wide range of hematologic malignancies (see Table S1). Patients received myeloablative (N = 10) or reduced-intensity (N = 3) conditioning. All patients were given standard immunosuppressive prophylaxis with cyclosporine and methotrexate (in myeloablative regimens) or mycophenolate mofetil (in reduced-intensity regimens; see Table S2) with or without ATG (depending on type of donor and the source of cells). Median time from alloHCT to diagnosis of aGvHD was 54 days (range 19–125). All but one patient with aGvHD were steroid-resistant and had GI tract involvement (median grade: 3); additionally, eight had skin (median grade: 3) and five had liver (median grade: 1) GvHD symptoms. Patients with cGvHD presented progressive involvement of the skin, lungs, pericardium, and eyes. The median time from aGvHD diagnosis to the first FMT treatment was 45 days (25–489 days). Most patients were still on steroids during FMT (median dose 0.5 mg methylprednisolone/kg of body weight). All patients were colonized with ARB (from 1 to 3 ARB per patient; see Table S3). Three patients underwent second FMT, two due to relapse of aGvHD and one as a means to deepen the response (the patient achieved partial response after first FMT). Detailed characteristics of patients are provided in Table S1. In 11 out of 14 (71%; two FMT procedures were excluded from the analysis due to early death and failure to reach the first follow-up point), antibiotic-resistant bacteria decolonization or partial decolonization (decolonization of at least one ARB, but not all colonizing ARBs) were observed. New colonization after microbiologically confirmed decolonization was observed in 4 out of 11 individuals, and all of these patients received broad spectrum antibiotics within 7 days after FMT. One patient who did not get decolonized after FMT developed new colonization. Focusing on GvHD, overall response rate (ORR) in cases of aGvHD reached 57% (8/14 FMTs), including CR in 42% (6/14) procedures (Figure 1A). Median duration of response to relapse, death, or last follow-up was 153 days (17–757). Patients with PR were offered a second procedure; one of them agreed and reached CR, the other refused and remained in PR for 757 days until surgical removal of the affected length of small intestine, which resulted in CR. Interestingly, out of three patients who did not respond to ruxolitinib treatment (P7, P8, P11), two obtained CR after FMT (P8, P11). All non-responders died within 200 days after the procedure, with median overall survival (OS) of 66 days. In responders, median OS was 332 days (HR 0.18 95% CI 0.03–0.93, p <.005; Figure 1B). The majority of patients who responded to FMT (6/8; 75%) underwent the procedure in a better performance score (ECOG ≤2; see Figure S1). Details regarding patient outcomes are provided in Figure 1A, B. Both patients with cGvHD treated with FMT achieved stabilization or improvement of organ disease, and were alive at last follow-up. Severe adverse events (SAEs) were diagnosed in 8 (56%) patients (grade 3–5; see Table S4). Two SAEs were early deaths within 7 days from FMT. Most of the SAEs and AEs were classified as not related to FMT (ileus/subileus, colitis, bronchopneumonia, lobar pneumonia, and one septic shock); however, one episode of septic shock, one episode of sepsis, and one case of Norovirus-mediated gastrointestinal tract infection (FMT performed immediately before alloHCT and repeated after alloHCT)4 were possibly (one septic shock and one sepsis) or certainly (Norovirus infection) related to performed FMT. Sepsis and septic shock deemed related to FMT were caused by broadly susceptible pathogens. Both cases developed <6 hours after the procedure. None of these events resulted in death. Severe general condition (ECOG >2) seemed to be related with the presence of serious complications after FMT, but the difference between groups was not significant (ECOG ≤2; 33% vs. 83%, RR = 0.44; 95% CI 0.17–1.13; see Figure S2). All case series regarding the use of FMT to treat GvHD that have been published to date showed high effectiveness of the procedure, ranging between 73% and 100%.3 In our cohort, overall response rate was lower, at 57%. This may be associated with implementation of FMT in late stages of disease in our centers. The study is not large enough to perform valid statistical analyses looking for the predicting factors, but there is a tendency among our patients that the worse general performance status (measured with ECOG/WHO scale) correlated with no response to FMT and more frequent number of complications. High rate of SAE contributed to frequent antibiotics use within one week after FMT, which could decrease effectiveness of the procedure. This study also shows a high rate of partial or total gut decolonization from antibiotic-resistant bacteria (71% in all groups). These rates most likely are lowered by the high rate of antibiotic use within the first week after FMT. Our previously publish results showed that avoidance of antibiotics during the first 7 days after FMT increases the ratio of decolonization, even to 100%.2 Recently, DeFilipp et al documented a case of bacteremia caused by Escherichia coli that was trace back to material used during FMT,5 and the FDA has warned of infections caused by enteropathogenic E. coli (EPEC) and Shiga toxin-producing E. coli (STEC) which have occurred following investigational use of FMT due to transmission of these pathogenic organisms from FMT product.6 Apart from that, the use of FMT to treat GvHD showed an excellent safety profile.3 We observed a higher rate of adverse events, especially in patients with low performance status. Among the 16 FMT procedures performed, we recorded 8 SAEs, including 2 deaths, sepsis, septic shocks, and localized infections (Table S3). Although the majority of adverse reactions were not directly related to FMT, two episodes of sepsis and Norovirus infection seemed triggered by the FMT procedure. The sepsis and septic shock episodes deemed related to FMT were caused by broadly susceptible pathogens. In both cases, bacteremia developed within few hours (<6) after the procedure. Again, we observed a trend that most complications occurred in patients with late-decision of treatment with FMT and with worse performance status (ECOG >2). We believe that FMT to treat gut GvHD should be performed at earlier stages. Although the role of the microbiome in the pathophysiology of cGHvD remains unknown, history and severity of aGVHD is the strongest predictor of occurrence of cGvHD. Two patients in our cohort were diagnosed with extensive cGvHD. Depending on the organs involved, we observed either improvement (skin in both and pericarditis in one subject) or stabilization (ocular disease or lungs) of the symptoms. After the FMT procedure, immunosuppressive drugs were either reduced or completely tapered. In one case, after one year of stabilization, ocular disease progressed and the patient required intensification of the immunosuppressive treatment, but finally the disease caused blindness. Importantly, in both cases, we did not observe complications of the procedure, and both patients eradicated colonizing pathogen. In summary, our study confirms very good efficacy of FMT in the treatment of GvHD and in decolonization of the GI tract from ARB. However, in contrast to previous studies, it revealed a higher rate of adverse events, almost all in severely-ill patients. We also report encouraging data on the use of FMT in patients with chronic GvHD. We would like to thank all of the participating patients and their families, as well as the investigators, research nurses, study coordinators, and operations staff. Jaroslaw Bilinski and Grzegorz W. Basak are the owners of Human Biome Institute - a professional fecal donors bank and a commercial institution offering fecal microbiota products. Pawel Grzesiowski is an owner of a Foundation for the Infection Prevention Institute - a professional fecal donors bank and a non-profit institution offering fecal microbiota products. All other authors have no conflicts of interest to declare. The data that supports the findings of this study are available in the supplementary material of this article. Appendix S1. Supporting Information Figure S1. Response rate according to performance status measured by ECOG (Eastern Cooperative Oncology Group). Figure S2. Occurrence of SAE (serious adverse event) according to performance status measured by ECOG (Eastern Cooperative Oncology Group). Table S1. General characteristics of patients with GvHD treated with FMT. Table S2. Immunosuppressive therapy for patients with GvHD. Table S3. Outcome of decolonization of ARB after FMT in patients with GvHD. Table S4. Severe adverse events of the FMT procedure in patients with GvHD. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.