Fecal microbiota transplantation (FMT) is an emerging therapy for recurrent or refractory Clostridioides difficile infection and is being actively investigated for other conditions. We describe two patients in whom extended-spectrum beta-lactamase (ESBL)–producing Escherichia coli bacteremia occurred after they had undergone FMT in two independent clinical trials; both cases were linked to the same stool donor by means of genomic sequencing. One of the patients died. Enhanced donor screening to limit the transmission of microorganisms that could lead to adverse infectious events and continued vigilance to define the benefits and risks of FMT across different patient populations are warranted.
An altered intestinal microbiome has been implicated in the pathogenesis of many disorders. Fecal microbiota transplantation (FMT) refers to the administration of intestinal microbes from a healthy donor into a recipient with the intent of modifying the recipient’s intestinal microbiome.1 FMT is an emerging treatment for recurrent or refractory Clostridioides difficile infection, and randomized, controlled trials and meta-analyses have supported the safety and efficacy of this procedure.2-7 FMT is being used experimentally in the treatment of other disease states linked to dysbiosis of the intestinal microbiota. ClinicalTrials.gov currently lists more than 300 studies evaluating FMT for various indications, primarily gastrointestinal, but also for neurologic, behavioral, and metabolic conditions. The investigational use of FMT in oncology, specifically for recipients of an allogeneic hematopoietic-cell transplant, is also of interest to prevent or treat post-transplantation complications such as acute graft-versus-host disease8,9 and to enhance the efficacy of newer immunotherapies.
FMT is associated with rare inflammatory, infectious, and procedural complications.10,11 Placebo-controlled trials have not reported serious adverse events or transmission of infection, and a systematic review of FMT showed similar rates of adverse events in immunocompromised and immunocompetent recipients, including infectious events.12 Gram-negative bacteremia has been reported in four patients after they had undergone FMT. In three of these cases, there was a plausible alternative explanation for the bacteremia, including ventilator-associated pneumonia, toxic megacolon, and Crohn’s disease.13-15 The fourth case involved aspiration of feculent material during upper endoscopic FMT delivery, followed by pneumonia, Escherichia coli bacteremia, septic shock, and death.16 Here, we describe two patients in whom extended-spectrum beta-lactamase (ESBL)–producing E. coli bacteremia occurred after they had undergone FMT in two separate clinical trials; both cases were linked to the same stool donor.
Donor Screening and Capsule Preparation
Donor screening was performed according to the protocols approved by our local institutional review board and the Food and Drug Administration (FDA) (Table 1). Donors were adults 18 to 50 years of age with a body-mass index (the weight in kilograms divided by the square of the height in meters) of 18.5 to 25.0. Medical history, physical examination findings, and laboratory test results for each donor must have met the standards outlined in the protocols (see Section 1 in the Supplementary Appendix, available with the full text of this article at NEJM.org). During the donation period, volunteers underwent an interim health query for febrile, systemic, and gastrointestinal symptoms and were deferred for any change in health status. All donations were stored and not used for an additional 4 weeks to allow retesting for certain infections that have a longer incubation period.
The donated stool was liquefied in a blender, sieved, centrifuged, and resuspended in concentrated form in sterile saline with 40% glycerol. The suspension was double-encapsulated in hydroxypropyl methylcellulose acid-resistant capsules (Capsugel) and stored at –80°C. The frozen FMT capsules were stored for up to 9 months in accordance with approved protocols based on stability test data.6
In January 2019, in response to a regulatory review by the FDA of a separate trial, we prospectively expanded donor-stool screening to include tests for ESBL-producing organisms (HardyChrom ESBL selective agar, Hardy Diagnostics), norovirus and adenovirus (polymerase-chain-reaction assay of stool specimens), and human T-lymphotropic virus type 1 and type 2 antibodies. We did not retrospectively test or destroy the existing capsule inventory that was generated before the expansion of donor screening. We were not directed by regulators to test stored materials or apply these new criteria to any other study. The FMT capsules that were administered to the two patients described below were manufactured in November 2018.
A 69-year-old man with cirrhosis of the liver attributed to hepatitis C virus infection (Model for End-Stage Liver Disease score of 18, on a scale from 6 to 40, with higher scores indicating more advanced liver disease) was enrolled in an open-label trial of FMT oral capsules to treat refractory hepatic encephalopathy (ClinicalTrials.gov number, NCT03420482). In March and April 2019, the patient received 15 FMT capsules five times over 3 weeks. He was also receiving rifaximin prophylaxis for hepatic encephalopathy before, during, and after the 3 weeks of FMT. The patient reported no adverse events until 17 days after the final FMT dose (May 2019), when a fever (38.9°C) and cough developed. He was found to have an infiltrate on chest radiography and was given levofloxacin for pneumonia. The patient returned 2 days later because of a lack of clinical improvement, and he was noted to have growth of gram-negative rods on blood cultures obtained at the initial presentation. Treatment with piperacillin–tazobactam was initiated, and the patient was admitted to the hospital. A diagnostic paracentesis was obtained 4 days after initiation of antibiotic therapy; the findings from ascites fluid studies were not consistent with spontaneous bacterial peritonitis, with a total white-cell count of 384 cells per cubic millimeter (31% neutrophils) and bacterial culture with no growth. The organism that grew in the blood culture was identified as ESBL-producing E. coli. The patient’s antibiotic therapy was switched to a carbapenem, and he completed a 14-day course of meropenem (inpatient therapy) followed by ertapenem (outpatient therapy). His condition has remained clinically stable since completing antimicrobial therapy. A follow-up stool sample was screened for ESBL-producing organisms on selective medium and was negative.
A 73-year-old man with therapy-related myelodysplastic syndrome was admitted for allogeneic hematopoietic-cell transplantation after reduced intensity conditioning (melphalan and fludarabine) with the use of a graft from an HLA-mismatched unrelated donor. The patient also received prophylaxis against post-transplantation graft-versus-host disease with high-dose cyclophosphamide on days 3 and 4 after stem-cell infusion, followed by sirolimus and mycophenolate mofetil, both starting on day 5. The patient was enrolled in a phase 2 trial to preemptively administer FMT oral capsules before and after allogeneic hematopoietic-cell transplantation (NCT03720392). He received 15 FMT capsules on day 4 and on day 3 before hematopoietic-cell transplantation. Cefpodoxime prophylaxis was initiated on day 1 before transplantation according to institutional standards to minimize the risk of gram-negative bacteremias, given that the patient was allergic to fluoroquinolones. On day 5 after stem-cell infusion (8 days after the last FMT dose), the patient had development of fever (39.7°C), chills, and altered mental status. Blood cultures were drawn, and cefepime therapy was promptly started for febrile neutropenia (absolute neutrophil count, 0 cells per cubic millimeter). Later that evening, hypoxia and labored breathing developed, and the patient was transferred to the intensive care unit to undergo intubation and mechanical ventilation. Preliminary blood culture results indicated the presence of gram-negative rods. His antibiotic regimen was expanded to include meropenem. Despite maximum supportive measures, the patient’s condition worsened, and he died from severe sepsis 2 days later. The final results of blood cultures showed ESBL-producing E. coli.
Strain Identification and Follow-up
The occurrence of these two bacteremias with similar resistance patterns (Figure 1A) prompted investigation. FMT capsules from all lots are routinely frozen for possible future analyses. We confirmed that Patients 1 and 2 both received FMT capsules from the same lot from the same donor. Each of three lots of capsules from that donor was found to contain ESBL-producing E. coli with a resistance pattern similar, but not identical, to the blood isolates from the patients, as determined by means of culture (Figure 1A). Frozen fecal samples that were obtained from Patients 1 and 2 before FMT were cultured and found to be negative for ESBL-producing organisms.
Genomes of ESBL-producing E. coli isolated in blood samples from the patients and in a stool sample from the donor and a control strain were sequenced by personnel at Day Zero Diagnostics. Genomic relatedness between samples was calculated by means of whole-genome sequencing and single-nucleotide polymorphism (SNP)-based analysis. Genomic DNA from the bacterial samples was sequenced on the iSeq 100 System (Illumina), reads were mapped to a closely related reference genome, and high-quality SNPs were identifed in each sample. SNP distances between pairs of isolates were computed by counting the number of variable sites (see Section 2 in the Supplementary Appendix). In silico multilocus sequence typing and serotyping showed that the three isolates belong to the same multilocus sequence type 131 and serotype O25:H4. SNP analysis revealed high genetic similarity among the three isolates, which either had no SNP differences or differed by 1 SNP across 4.5 million base pairs examined (Figure 1B). The isolates exhibited much lower genetic relatedness (distance of 121 to 124 SNPs) to the control E. coli isolate of multilocus sequence type 131 and serotype O25:H4 derived from a blood culture that had been obtained at our hospital in 2017 (Figure 1B). Similarly, a genomic analysis of nine previously sequenced multilocus sequence type 131 isolates (six isolates of O25:H4 serotype and three of O16:H5 serotype) derived from clinical urine samples with no known epidemiologic links (also obtained at our hospital) revealed that they all had a distance of more than 100 SNPs from the FMT-linked isolates. Genomic studies of E. coli outbreaks have shown epidemiologically linked isolates that had a distance from each other of less than 10 SNPs to be a part of a transmission cluster.18-20 Because there was a distance of at most 1 SNP, the isolates from the donor and the patients were deemed with high confidence to be clonal organisms.
A total of 22 patients received FMT capsules generated from this donor (6 recipients in the two trials described above and 16 recipients who received FMT for the treatment of recurrent or refractory C. difficile infection). All recipients were subsequently contacted and informed, and stool screening with an ESBL selective medium was offered. The results are shown in Table 2. For the patients in the two trials discussed herein, stool samples that had been obtained before FMT were available; the samples were cultured, and all showed no growth on screening plates. No pre-FMT samples were available for the 16 recipients who underwent FMT for the treatment of C. difficile infection. Of the 12 tested patients who received FMT capsules generated from this donor, 5 had post-FMT samples that grew organisms on an ESBL selective medium. Of the 7 tested recipients who underwent FMT for recurrent or refractory C. difficile infection, 4 had post-FMT samples that showed growth on an ESBL selective medium that was morphologically consistent with E. coli. The number of antibiotic-resistant genes contained within the fecal microbial metagenome has been shown to be much higher in patients with recurrent or refractory C. difficile infection than in healthy persons, and screened donors have even lower numbers of resistant genes.21 When FMT is successful, the recipient’s metagenomic burden of antimicrobial resistance genes mimics that of the donor.21 Although we cannot conclusively attribute positive screening results for ESBL-producing organisms in other asymptomatic recipients to FMT, the rates of positive tests are, in our opinion, unexpectedly high and probably represent transmission through FMT. Stored capsules from 32 FMT preparations from 10 other donors (2014 to 2018) were subsequently screened, and none showed any growth on ESBL screening plates.
Although rare cases of bacteremia and death after FMT have been described,13-16 we report two cases of bacteremia caused by an organism that was transmitted through FMT, as shown by means of genome sequencing. The stool donor had no risk factors of multidrug-resistant organism carriage and had donated fecal material before we included routine tests for ESBL-producing organisms in our donor-screening protocol. Given that the frozen capsules from all preparations derived from this donor were positive on ESBL screening, it is likely that this donor material would have been disqualified if screened before donation. Subsequent screening for ESBL phenotypes in the other FMT recipients suggests further FMT-related transmission; however, confirmatory identification of ESBL-producing E. coli and an analysis of the relative relatedness of these isolates to one another were not performed. To date, we have not identified related infectious complications involving ESBL-producing organisms in other recipients.
Our two patients had risk factors for bacteremia. Patients with advanced cirrhosis are at heightened risk for bacterial translocation because of increased intestinal permeability, impaired reticuloendothelial system function, and portosystemic shunting.22 Early after allogeneic hematopoietic-cell transplantation, patients are at risk for infections from enteric microorganisms because of compromise of the intestinal barrier related to conditioning therapy and concurrent neutropenia. We have revised our protocols on the basis of these events of transmission through FMT. Both patients also received oral antibiotics (rifaximin and cefpodoxime in accordance with usual care) near the time of FMT, and we hypothesize the timing of this antibiotic treatment might have been important with respect to the selection and maintenance of drug-resistant organisms.
These serious adverse events were promptly reported to the FDA and supervising institutional review boards. Our trials were halted, and the FDA placed a national alert on its website and mandated additional screening for those conducting FMT research under FDA supervision.23 Antimicrobial resistance patterns were similar but not identical, which shows that these data alone cannot fully inform clonality assessments. There is growing evidence showing that highly related isolates can display divergent profiles on antibiogram or pulsed-field gel electrophoresis.24 It is therefore possible that previous studies have underestimated the risk of infection related to FMT, given the substantial overlap in bacterial species between donor and recipient. In our cases, we were able to show transmission through FMT because the organisms were easily identified and genome sequencing was performed. ESBL-producing E. coli are not inherently more virulent, although the treatment of this type of infection is clearly more challenging.25-27 The carriage of antimicrobial resistance genes can either confer a survival advantage or, at other times, extract a fitness cost.
Despite the infectious complications reported here, the benefits of FMT should be balanced with the associated risks when considering treatment options for patients with recurrent or refractory C. difficile infection. The development of defined, cultured, therapeutic microbial mixtures is an obvious and important future goal. Ongoing assessment of the risks and benefits of FMT research is needed, as are continuing efforts to improve donor screening to limit transmission of microorganisms that could lead to adverse infectious events.
Funding and Disclosures
Supported by a grant from American College of Gastroenterology (to Dr. Bloom).
Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.
Drs. DeFilipp and Bloom contributed equally to this article.
This article was published on October 30, 2019, at NEJM.org.
We thank the patients and their families for providing permission to report adverse events and for their participation and contribution to research; Febriana K. Pangestu and Ian C. Herriott at Day Zero Diagnostics for their sequencing efforts; and the personnel at the Massachusetts General Hospital Clinical Microbiology Laboratory for their assistance in the identification, susceptibility testing, and confirmatory extended-spectrum beta-lactamase testing of the clinical and donor isolates.
1. Cammarota G, Ianiro G, Tilg H, et al. European consensus conference on faecal microbiota transplantation in clinical practice. Gut 2017;66:569-580.
2. Mullish BH, Quraishi MN, Segal JP, et al. The use of faecal microbiota transplant as treatment for recurrent or refractory Clostridium difficile infection and other potential indications: joint British Society of Gastroenterology (BSG) and Healthcare Infection Society (HIS) guidelines. Gut 2018;67:1920-1941.
3. Gupta A, Cifu AS, Khanna S. Diagnosis and treatment of Clostridium difficile infection. JAMA 2018;320:1031-1032.
4. Kassam Z, Lee CH, Yuan Y, Hunt RH. Fecal microbiota transplantation for Clostridium difficile infection: systematic review and meta-analysis. Am J Gastroenterol 2013;108:500-508.
5. Van Nood E, Vrieze A, Nieuwdorp M, et al. Duodenal infusion of donor feces for recurrent Clostridium difficile. N Engl J Med 2013;368:407-415.
6. Youngster I, Russell GH, Pindar C, Ziv-Baran T, Sauk J, Hohmann EL. Oral, capsulized, frozen fecal microbiota transplantation for relapsing Clostridium difficile infection. JAMA 2014;312:1772-1778.
7. McDonald LC, Gerding DN, Johnson S, et al. Clinical practice guidelines for Clostridium difficile infection in adults and children: 2017 update by the Infectious Diseases Society of America (IDSA) and Society for Healthcare Epidemiology of America (SHEA). Clin Infect Dis 2018;66:987-994.
8. DeFilipp Z, Hohmann E, Jenq RR, Chen YB. Fecal microbiota transplantation: restoring the injured microbiome after allogeneic hematopoietic cell transplantation. Biol Blood Marrow Transplant 2019;25(1):e17-e22.
9. DeFilipp Z, Peled JU, Li S, et al. Third-party fecal microbiota transplantation following allo-HCT reconstitutes microbiome diversity. Blood Adv 2018;2:745-753.
10. Krajicek E, Fischer M, Allegretti JR, Kelly CR. Nuts and bolts of fecal microbiota transplantation. Clin Gastroenterol Hepatol 2019;17:345-352.
11. De Leon LM, Watson JB, Kelly CR. Transient flare of ulcerative colitis after fecal microbiota transplantation for recurrent Clostridium difficile infection. Clin Gastroenterol Hepatol 2013;11:1036-1038.
12. Shogbesan O, Poudel DR, Victor S, et al. A systematic review of the efficacy and safety of fecal microbiota transplant for Clostridium difficile infection in immunocompromised patients. Can J Gastroenterol Hepatol 2018;2018:1394379-1394379.
13. Quera R, Espinoza R, Estay C, Rivera D. Bacteremia as an adverse event of fecal microbiota transplantation in a patient with Crohn’s disease and recurrent Clostridium difficile infection. J Crohns Colitis 2014;8:252-253.
14. Trubiano JA, Gardiner B, Kwong JC, Ward P, Testro AG, Charles PG. Faecal microbiota transplantation for severe Clostridium difficile infection in the intensive care unit. Eur J Gastroenterol Hepatol 2013;25:255-257.
15. Solari PR, Fairchild PG, Noa LJ, Wallace MR. Tempered enthusiasm for fecal transplant. Clin Infect Dis 2014;59:319-319.
16. Baxter M, Ahmad T, Colville A, Sheridan R. Fatal aspiration pneumonia as a complication of fecal microbiota transplant. Clin Infect Dis 2015;61:136-137.
17. Clinical and Laboratory Standards Institute. Performance standards for antimicrobial susceptibility testing: CLSI supplement M100. 29th ed. Wayne, PA: Clinical and Laboratory Standards Institute, 2019.
18. Grad YH, Lipsitch M, Feldgarden M, et al. Genomic epidemiology of the Escherichia coli O104:H4 outbreaks in Europe, 2011. Proc Natl Acad Sci U S A 2012;109:3065-3070.
19. Peacock SJ, Parkhill J, Brown NM. Changing the paradigm for hospital outbreak detection by leading with genomic surveillance of nosocomial pathogens. Microbiology 2018;164:1213-1219.
20. Schürch AC, Arredondo-Alonso S, Willems RJL, Goering RV. Whole genome sequencing options for bacterial strain typing and epidemiologic analysis based on single nucleotide polymorphism versus gene-by-gene-based approaches. Clin Microbiol Infect 2018;24:350-354.
21. Millan B, Park H, Hotte N, et al. Fecal microbial transplants reduce antibiotic-resistant genes in patients with recurrent Clostridium difficile infection. Clin Infect Dis 2016;62:1479-1486.
22. Wiest R, Lawson M, Geuking M. Pathological bacterial translocation in liver cirrhosis. J Hepatol 2014;60:197-209.
23. Important safety alert regarding use of fecal microbiota for transplantation and risk of serious adverse reactions due to transmission of multi-drug resistant organisms. Silver Spring, MD: Food and Drug Administration, June 13, 2019 (https://www.fda.gov/vaccines-blood-biologics/safety-availability-biologics/important-safety-alert-regarding-use-fecal-microbiota-transplantation-and-risk-serious-adverse).
24. Gordon NC, Pichon B, Golubchik T, et al. Whole-genome sequencing reveals the contribution of long-term carriers in Staphylococcus aureus outbreak investigation. J Clin Microbiol 2017;55:2188-2197.
25. Lavigne JP, Blanc-Potard AB, Bourg G, et al. Virulence genotype and nematode-killing properties of extra-intestinal Escherichia coli producing CTX-M beta-lactamases. Clin Microbiol Infect 2006;12:1199-1206.
26. Deschamps C, Clermont O, Hipeaux MC, Arlet G, Denamur E, Branger C. Multiple acquisitions of CTX-M plasmids in the rare D2 genotype of Escherichia coli provide evidence for convergent evolution. Microbiology 2009;155:1656-1668.
27. Beceiro A, Tomás M, Bou G. Antimicrobial resistance and virulence: a successful or deleterious association in the bacterial world? Clin Microbiol Rev 2013;26:185-230.
Citing Articles (416)
This protocol was reviewed by the Food and Drug Administration in 2018 and did not mandate screening for ESBL-producing organisms. CRE denotes carbapenem-resistant Enterobacteriaceae, ELISA enzyme-linked immunosorbent assay, HIV-1 human immunodeficiency virus type 1, HIV-2 human immunodeficiency virus type 2, MRSA methicillin-resistant Staphylococcus aureus, PCR polymerase chain reaction, and VRE vancomycin-resistant enterococci.
The body-mass index is the weight in kilograms divided by the square of the height in meters.
Extended-spectrum beta-lactamase (ESBL) screening was performed with the use of HardyCHROM ESBL medium (Hardy Diagnostics). FMT denotes fecal microbiota transplantation.
Original ArticleMay 15 Original ArticleMay 15 Original ArticleMay 12