- Email: [email protected]
An Intragastric Fecal Microbiota Transplantation Program for Treatment of Recurrent Clostridium difficile in Children is Efficacious, Safe, and Inexpensive
ARTICLE IN PRESS THE JOURNAL OF PEDIATRICS • www.jpeds.com
ORIGINAL ARTICLES
An Intragastric Fecal Microbiota Transplantation Program for Treatment of Recurrent Clostridium difficile in Children is Efficacious, Safe, and Inexpensive David E. Brumbaugh, MD1,2, Edwin F. De Zoeten, MD, PhD1,2, Amy Pyo-Twist, RN3, Sara Fidanza, RN3, Shannon Hughes, PA4, Susan A. Dolan, RN, MS5, Jason Child, PharmD6, and Samuel R. Dominguez, MD, PhD4,5 Objective To assess the safety, efficacy, and relative expense of a nurse-led fecal microbiota transplantation (FMT) program for the treatment of recurrent Clostridium difficile infection (CDI).
Study design Retrospective cohort study design in children aged 1-18 years with recurrent CDI. The intervention was an intragastric FMT with stool derived from a donor stool bank. Primary outcome was resolution of diarrhea at 3 months post-transplantation. A secondary analysis compared charge data associated with FMT by intragastric delivery vs administration by colonoscopy or nasoduodenal tube. Results A total of 47 intragastric FMT procedures were performed in 42 children (median age 9 years) with recurrent CDI. Response to treatment varied by disease status, with 94% success in previously healthy children, 75% in medically complex children, and 54% in children with inflammatory bowel disease (P = .04). FMT via intragastric delivery showed lower facility and professional charges by 85% and 78% compared with delivery via colonoscopy and radiology-placed nasoduodenal tube, respectively. The use of stool derived from a donor stool bank decreased charges by 49% compared with charges associated with the use of a donor who was a relative. Conclusion A nurse-led intragastric FMT procedure using stool derived from a donor stool bank is a relatively inexpensive and efficacious treatment for recurrent CDI in children. Intragastric FMT success in children was attenuated by the presence of underlying disease, particularly inflammatory bowel disease. (J Pediatr 2017;■■:■■-■■). he prevalence of Clostridium difficile infection (CDI) is increasing in both adult and pediatric patients.1,2 Antibiotic therapy remains the standard first-line therapy for CDI, but the significant rise in highly virulent strains of C difficile and recurrence rates between 15% and 30% have challenged clinicians to discover alternative therapies for the treatment of CDI.3-5 Fecal microbiota transplantation (FMT) has emerged as the standard of care in the treatment of patients with recurrent CDI for whom antibiotics have not been successful.6,7 Although series of FMT for the treatment of recurrent CDI in pediatric patients have demonstrated efficacy rates equivalent to adults, overall reported experience in children remains small.8,9 There are variables of FMT that differentiate treatment protocols and potentially influence treatment efficacy and/or expense. These include the source of donor stool, route of donor stool delivery, setting of the procedure, and the type of healthcare professional who performs the procedure. The most common route of FMT delivery is via the lower gastrointestinal (GI) tract, either by colonoscopy or enema, with reported treatment success of 88%-95%.10 However, colonoscopy in particular is an expensive procedure that usually requires general anesthesia in children. In small series, the delivery of FMT into the upper GI tract has demonstrated efficacy approaching that of colonoscopy.9,11,12 The donor source for FMT traditionally has been a firstdegree relative, requiring the donor to undergo expensive testing to assess for potentially transmissible infectious agents before providing a donation to a single recipient. With the development of public stool banks, however, there is an opportunity for spreading the costs of donor testing across multiple FMT recipients.13 With heterogeneity in FMT protocols and gaps in our knowledge of treatment efficacy in children, there is opportunity to evaluate the value of this procedure in pediatrics. Value can be defined as healthcare outcome relative to cost.14 Therefore, the goal in maximizing value for the FMT procedure is to deliver donor stool at the lowest possible cost without sacrifice of clinical success, which has been approximately 90%.7 We describe the initial outcome and charge-savings data from a nurse-driven pediatric FMT program for the treatFrom the 1Section of Gastroenterology, Department of ment of recurrent CDI disease in children. Specific program design elements Pediatrics, University of Colorado School of Medicine;
T
CDI FMT GI IBD NGT RN
Clostridium difficile infection Fecal microbiota transplantation Gastrointestinal Inflammatory bowel disease Nasogastric feeding tube Registered nurse
2Children’s Hospital Colorado, Aurora, CO; 3Department of Nursing, Children’s Hospital Colorado; 4Section of Infectious Diseases, Department of Pediatrics, University of Colorado School of Medicine and Children’s Hospital Colorado; 5Department of Epidemiology, Children’s Hospital Colorado, Aurora, CO; and 6Department of Pharmacy, Children’s Hospital Colorado, Aurora, CO
The authors declare no conflicts of interest. 0022-3476/$ - see front matter. © 2017 Elsevier Inc. All rights reserved. https://doi.org10.1016/j.jpeds.2017.10.016
1 FLA 5.5.0 DTD ■ YMPD9530_proof ■ November 30, 2017
THE JOURNAL OF PEDIATRICS • www.jpeds.com intended to decrease costs of FMT included delivery in an ambulatory setting, intragastric administration of stool, use of a donor stool bank, and performance by a nurse rather than physician.
Methods The study was a retrospective cohort design with data extracted from the electronic health record. The study was approved by the Colorado Multiple Institutional Review Board. Inclusion criteria included all patients who received intragastric FMT at Children’s Hospital Colorado from March 2015 through September 2016. All patients received FMT as treatment for recurrent CDI, as defined by a positive fecal C difficile polymerase chain reaction test (Xpert C. Difficile; Cepheid, Sunnyvale, California) in combination with characteristic symptoms of diarrhea, hematochezia, or crampy abdominal pain. To be eligible for FMT, patients had to have greater than 2 episodes of CDI and have failed at least 1 course of therapy with oral vancomycin. Children were defined as medically complex if a diagnosis preceding CDI required chronic care by a pediatric subspecialist. All parents or guardians provided informed consent for FMT, and patients older than 7 years provided assent. Pretreatment for the procedure included a minimum of 5 days of oral vancomycin that was discontinued 2 days before FMT. All patients were treated with ranitidine for 24 hours immediately preceding FMT. On the day of FMT, patients had a nasogastric feeding tube (NGT) placed by a registered nurse (RN) in the ambulatory clinic. Some patients with a pre-existing gastrostomy tube received their intragastric FMT via their current tube. Correct intragastric placement of the NGT was confirmed by abdominal radiography. Once placement was confirmed, the FMT procedure was started. Donor stool was received from a stool donor bank (OpenBiome, Cambridge, Massachusetts) in 30-mL frozen aliquots or oral capsules and stored in a −20°C freezer. OpenBiome maintains a rigorous safety program that requires regular laboratory screening of its donors.15 A 30-mL aliquot was thawed immediately before the procedure. Once thawed, the 30-mL stool aliquot was infused over 2 minutes into the stomach through the gastric tube and flushed with 60 mL of tap water. Patients were observed by the RN for 10 minutes in the clinic before NGT removal, then an additional 10 minutes after NGT removal before discharge from the clinic. All families were contacted by phone or secure electronic correspondence at 2 weeks and 3 months post-FMT to assess treatment response and side effects of treatment. Treatment success was defined as the absence of diarrhea, hematochezia, and abdominal pain at 3 months post-FMT or, when symptoms persisted, a negative C difficile test. Patients who experienced a recurrence of symptoms before 3 months postFMT were offered the option to receive a second FMT in a procedure identical to the first. If symptoms were resolved with the second FMT, those patients were considered a treatment success.
Volume ■■
•
■■ 2017
Hypothetical analysis of procedural charges was conducted by obtaining average charge data from Children’s Hospital Colorado for the following standard procedures: diagnostic colonoscopy under general anesthesia and fluoroscopicassisted nasoduodenal tube placement. Charges included both technical and professional fees for these procedures. Charges associated with the actual FMT procedure included 2 hours of RN staffing, an ambulatory clinic charge to account for the 2 hours use of a clinic room, and both technical and interpretation charges associated with the abdominal radiography for NGT placement confirmation. Equipment costs (feeding tubes, tape, syringes, etc) were rolled into the procedural and ambulatory clinic charges. For comparison of charges by stool donor source, we obtained charge data from a standard commercial laboratory (LabCorp, Burlington, North Carolina) for the following tests: stool polymerase chain reaction for C difficile toxins A/B, stool bacterial culture, stool cryptosporidium and giardia antigen assay, stool ova and parasites, serum for HIV serotypes 1 and 2, hepatitis A (Anti-HAV IgM), hepatitis B (HBsAg and AntiHBs), hepatitis C (Anti-HCV IgG and IgM), and rapid plasma reagin. Charges for stool donor bank aliquots were based on procurement charges from OpenBiome, 20% of standard shipping charges (assuming 5 aliquots per shipment), and hospital pharmacy storage and administration charges. The Fisher exact test was used to assess whether group disease status (previously healthy, inflammatory bowel disease [IBD], and medically complex) and route of delivery (NGT, gastrostomy tube, and capsule) impacted treatment outcome.
Results During the study period, 42 unique patients received FMT at Children’s Hospital Colorado. The median age of the study population was 9 years (range, 1-18 years) and 45% were male. Thirty-one percent of the study population had IBD (Table I; available at www.jpeds.com) and 29% were medically complex with oncologic, metabolic, cardiopulmonary, or neurologic diagnoses (Table II; available at www.jpeds.com). Thirty patients were treated successfully with an initial FMT. After not responding to their first treatment, 5 patients elected to have a second FMT procedure within 3 months, and 2 of those patients achieved treatment success. The outcome of treatment success, therefore, was achieved in 32 of 42 patients (76%). There was a significant difference (P = .04) in treatment outcome by the presence of underlying disease (Figure 1). Ninety-four percent of otherwise-healthy children were treated successfully with FMT, whereas only 75% of medically complex children responded successfully. Treatment success dropped further to 54% in children with IBD. Success by method of upper GI tract FMT delivery was 71% by NGT and 67% by gastrostomy tube (P = .99). The single treatment by ingestion of 30 oral capsules was successful. The only reported postprocedural complication was vomiting within 24 hours of the FMT, which occurred in 6 of 47 (13%) procedures. In all cases, the vomiting was a single, selflimited episode that did not require medical treatment.
2
Brumbaugh et al FLA 5.5.0 DTD ■ YMPD9530_proof ■ November 30, 2017
■■ 2017
ORIGINAL ARTICLES
Figure 2. Overall hospital and professional charges associated with FMT by delivery method.
Figure 1. Pediatric FMT success rate by disease status. *P < .04 for difference in FMT success rate by disease status (Fisher exact test).
Total facility and professional charges were modeled for FMT as if administered via diagnostic colonoscopy under general anesthesia or via radiologist-placed nasoduodenal tube (no sedation). These hypothetical procedural charges were compared with charges from our FMT procedure, accomplished via NGT placed by a RN in a hospital outpatient setting (Figure 2). In this analysis, a stool bank was modeled as the stool donor source, and no differences in stool aliquot size or cost were assumed. Compared with FMT via colonoscopy ($7767) and nasoduodenal tube ($4998), FMT administration via nurse-placed NG tube ($1139) decreased total healthcare charges by 85% and 78%, respectively. Standard outpatient commercial laboratory charges for screening of a single first-degree relative as stool donor amounted to $1154 per stool donor. By comparison, the charges associated with procurement, storage, and administration of banked stool donor material from OpenBiome totaled $628, a 46% reduction in charges for donor stool.
Discussion In this study, administration of FMT via upper GI tract delivery to children with recurrent CDI achieved a high rate of treatment success (94%) in otherwise-healthy patients. These data add to previous reports suggesting that upper GI tract
delivery is as effective as lower tract GI delivery in treating recurrent CDI; and frozen, unrelated donor stool is as effective as fresh donor stool obtained from a close relative. Furthermore, we demonstrate that FMT delivery by this method is safe, well tolerated, and inexpensive compared with FMT delivered by colonoscopy or nasoduodenal tube. However, using this FMT technique, we observed attenuated treatment success in children with significant comorbidities, particularly IBD. We do not know whether our attenuated treatment success in children with IBD and medically complexity represents a limitation of our treatment strategy or represents actual real-world efficacy in these specific pediatric patient populations. There are limited data reporting outcomes from FMT in pediatric patients with both IBD and recurrent CDI. Using FMT delivery via colonoscopy, Russell et al described success in 2 of 3 patients whereas Hourigan et al observed 100% success in 5 patients.8,16 There is a significant challenge for patients with IBD in gauging the relative contribution of IBD activity vs CDI in driving patient symptoms. The presence of infection may trigger an inflammatory cascade and IBD flare that persists even after resolution of the infection. In the setting of IBD, FMT also has been associated with worsening IBD symptoms in up to 14% of recipients.17 Asymptomatic C difficile carriage occurs in patients with IBD at a rate 6 times that of healthy controls.18 Asymptomatic carriage may confound interpretation of C difficile testing in this population, lending plausibility to the theory that our diminished FMT success in patients with IBD was due in part to IBD activity and C difficile colonization rather than active CDI. Alternatively, larger numbers of donor bacteria may be required for treatment success of recurrent CDI in patients with IBD. Delivery of greater bacterial counts may be
An Intragastric Fecal Microbiota Transplantation Program for Treatment of Recurrent Clostridium difficile in Children is Efficacious, Safe, and Inexpensive FLA 5.5.0 DTD ■ YMPD9530_proof ■ November 30, 2017
3
THE JOURNAL OF PEDIATRICS • www.jpeds.com achievable via colonoscopy compared with upper GI tract delivery, as aliquots of stool donor specimens typically are larger in FMT via colonoscopy and the passage of bacteria through the stomach and small bowel may reduce the viable bacterial load reaching the colon. However, the ultimate success of FMT may not be attributable directly to the size of the bacterial transfer. Emerging data suggest that specific strains of bacteria, specifically those critical to production of secondary bile acids, are deficient in patients with recurrent CDI and restored in patients who achieve treatment success.19 The presence of non-IBD medical complexity, including preexisting oncologic and neurologic disease, also decreased the likelihood of treatment success. This group of children may be at greater risk of CDI recurrence post-FMT as the result of increased contact with healthcare facilities, greater levels of antibiotic exposure, and the presence of indwelling GI feeding devices.20,21 There are important considerations about treatment safety in the medically complex pediatric population. Patients with complex craniofacial anatomy may not be appropriate for NG insertion. Patients at high risk for gastroesophageal reflux and/or aspiration, including patients under general anesthesia, are not good candidates for intragastric FMT delivery.22 If there is concern for undiagnosed IBD or other GI pathology, FMT via colonoscopy may be preferable, as this procedure has the potential to be both diagnostic and therapeutic. Our approach to FMT in children, summarized as a nursedriven procedure using upper GI tract delivery of stool acquired from a donor stool bank, has many benefits. First, delivery by upper tract incurs a fraction of the healthcare charges compared with the traditional route of FMT by colonoscopy. Second, we have shown treatment success comparable with published data in otherwise-healthy children receiving FMT via colonoscopic administration. Lastly, our data show that use of a donor stool bank is as effective as use of related donor stool, again decreasing expense. If CDI continues to increase in incidence in the pediatric population, we anticipate that FMT will be increasingly sought by patient families and providers. Lowering charges associated with FMT will ensure that this highly effective therapy is accessible to patients. Strengths of this study include the relatively large number of subjects, the largest reported series of FMT in children. The assessment of treatment response was consistent for all subjects at 3 months post-FMT. Our study has several limitations: the research was retrospective, so although we used a standardized data collection form, clinical evaluations and treatment may have differed depending on the individual providers who were caring for these children. As there is currently no assay available to differentiate C difficile colonization from infection, interpretation of C difficile testing remains challenging in children with high rates of colonization, specifically children with IBD or frequent contact with hospital environments.23 Several children with IBD classified in our study as FMT treatment failures may have had active IBD with C difficile colonization.
Volume ■■ In conclusion, our study provides important treatment efficacy data from a novel, nurse-led FMT program in a heterogeneous patient population at a tertiary pediatric hospital. We demonstrate very high levels of treatment success in previously healthy children using a relatively inexpensive FMT procedure. Treatment success was attenuated significantly in patients with IBD. ■ Submitted for publication Jun 15, 2017; last revision received Sep 26, 2017; accepted Oct 12, 2017 Reprint requests: David E. Brumbaugh, MD, Children’s Hospital Colorado, 13123 E. 16th Ave, B290, Aurora, CO 80045. E-mail: david.brumbaugh@ childrenscolorado.org
References 1. Ricciardi R, Rothenberger DA, Madoff RD, Baxter NN. Increasing prevalence and severity of Clostridium difficile colitis in hospitalized patients in the United States. Arch Surg 2007;142:624-31, discussion 31. 2. Deshpande A, Pant C, Anderson MP, Donskey CJ, Sferra TJ. Clostridium difficile infection in the hospitalized pediatric population: increasing trend in disease incidence. Pediatr Infect Dis J 2013;32:113840. 3. Redelings MD, Sorvillo F, Mascola L. Increase in Clostridium difficilerelated mortality rates, United States, 1999-2004. Emerg Infect Dis 2007;13:1417-9. 4. Lessa FC, Gould CV, McDonald LC. Current status of Clostridium difficile infection epidemiology. Clin Infect Dis 2012;55(suppl 2):S6570. 5. Shields K, Araujo-Castillo RV, Theethira TG, Alonso CD, Kelly CP. Recurrent Clostridium difficile infection: from colonization to cure. Anaerobe 2015;34:59-73. 6. Surawicz CM. Clostridium difficile infection: risk factors, diagnosis and management. Curr Treat Options Gastroenterol 2015;13:121-9. 7. Kassam Z, Lee CH, Yuan Y, Hunt RH. Fecal microbiota transplantation for Clostridium difficile infection: systematic review and meta-analysis. Am J Gastroenterol 2013;108:500-8. 8. Russell GH, Kaplan JL, Youngster I, Baril-Dore M, Schindelar L, Hohmann E, et al. Fecal transplant for recurrent Clostridium difficile infection in children with and without inflammatory bowel disease. J Pediatr Gastroenterol Nutr 2014;58:588-92. 9. Kronman MP, Nielson HJ, Adler AL, Giefer MJ, Wahbeh G, Singh N, et al. Fecal microbiota transplantation via nasogastric tube for recurrent Clostridium difficile infection in pediatric patients. J Pediatr Gastroenterol Nutr 2015;60:23-6. 10. Gough E, Shaikh H, Manges AR. Systematic review of intestinal microbiota transplantation (fecal bacteriotherapy) for recurrent Clostridium difficile infection. Clin Infect Dis 2011;53:994-1002. 11. van Nood E, Dijkgraaf MG, Keller JJ. Duodenal infusion of feces for recurrent Clostridium difficile. N Engl J Med 2013;368:2145. 12. Aas J, Gessert CE, Bakken JS. Recurrent Clostridium difficile colitis: case series involving 18 patients treated with donor stool administered via a nasogastric tube. Clin Infect Dis 2003;36:580-5. 13. Kazerouni A, Burgess J, Burns LJ, Wein LM. Optimal screening and donor management in a public stool bank. Microbiome 2015;3:75. 14. Porter ME. What is value in health care? N Engl J Med 2010;363:247781. 15. Openbiome. Quality and Safety Program Somerville, MA. https:// static1.squarespace.com/static/50e0c29ae4b0a05702af7e6a/t/59304 bdf3e00be24622791fb/1496337376745/The+OpenBiome+Quality+ %26+Safety+Program.pdf. Accessed October, 2017. 16. Hourigan SK, Chen LA, Grigoryan Z, Laroche G, Weidner M, Sears CL, et al. Microbiome changes associated with sustained eradication of Clostridium difficile after single faecal microbiota transplantation in children with and without inflammatory bowel disease. Aliment Pharmacol Ther 2015;42:741-52.
4
Brumbaugh et al FLA 5.5.0 DTD ■ YMPD9530_proof ■ November 30, 2017
■■ 2017
ORIGINAL ARTICLES
17. Kelly CR, Ihunnah C, Fischer M, Khoruts A, Surawicz C, Afzali A, et al. Fecal microbiota transplant for treatment of Clostridium difficile infection in immunocompromised patients. Am J Gastroenterol 2014;109:106571. 18. Hourigan SK, Chirumamilla SR, Ross T, Golub JE, Rabizadeh S, Saeed SA, et al. Clostridium difficile carriage and serum antitoxin responses in children with inflammatory bowel disease. Inflamm Bowel Dis 2013;19:2744-52. 19. Staley C, Kelly CR, Brandt LJ, Khoruts A, Sadowsky MJ. Complete Microbiota engraftment is not essential for recovery from recurrent Clostridium difficile infection following fecal microbiota transplantation. MBio 2016;7.
20. Eyre DW, Walker AS, Wyllie D, Dingle KE, Griffiths D, Finney J, et al. Predictors of first recurrence of Clostridium difficile infection: implications for initial management. Clin Infect Dis 2012;55(suppl 2):S77-87. 21. Sandora TJ, Fung M, Flaherty K, Helsing L, Scanlon P, Potter-Bynoe G, et al. Epidemiology and risk factors for Clostridium difficile infection in children. Pediatr Infect Dis J 2011;30:580-4. 22. Baxter M, Ahmad T, Colville A, Sheridan R. Fatal aspiration pneumonia as a complication of fecal microbiota transplant. Clin Infect Dis 2015;61:136-7. 23. Dominguez SR, Dolan SA, West K, Dantes RB, Epson E, Friedman D, et al. High colonization rate and prolonged shedding of Clostridium difficile in pediatric oncology patients. Clin Infect Dis 2014;59:401-3.
An Intragastric Fecal Microbiota Transplantation Program for Treatment of Recurrent Clostridium difficile in Children is Efficacious, Safe, and Inexpensive FLA 5.5.0 DTD ■ YMPD9530_proof ■ November 30, 2017
5
THE JOURNAL OF PEDIATRICS • www.jpeds.com
Volume ■■
Table I. Subject IBD disease type and treatment response Subject no.
Sex
IBD type
FMT success?
No. FMT attempts
F F M M M F M M F M F M M
Ulcerative colitis Crohn's disease Ulcerative colitis Ulcerative colitis Ulcerative colitis Ulcerative colitis Ulcerative colitis Ulcerative colitis Crohn's disease Crohn's disease Crohn's disease Ulcerative colitis Ulcerative colitis
Yes Yes Yes No Yes No No No Yes Yes No Yes No
1 1 1 2 1 1 1 1 1 2 1 1 1
1 2 3 4 5 6 7 8 9 10 11 12 13 F, female; M, male.
Table II. Diagnoses of subjects with medical complexity and treatment response Subject no. 14 15 16 17 18 19 20 21 22 23 24 24
Sex
Diagnoses
FMT success?
No. FMT attempts
M F F F F M M M F M M M
Rhabdomyosarcoma Epilepsy, malnutrition Disorder of mitochondrial metabolism Intestinal dysmotility, neutropenia, seizures, neurofibromatosis Celiac disease, tuberous sclerosis, seizures, autism Genetic syndrome, seizures, dysphagia Pancreatitis, seizures Rhabdomyosarcoma Anoxic brain damage, cerebral palsy, tracheostomy Chronic lung disease, feeding difficulties, venous thrombus Complex congenital heart disease, chronic pulmonary disease Seizures, intellectual disability, cardiofaciocutaneous syndrome
Yes Yes No Yes Yes Yes No Yes Yes Yes No Yes
1 1 1 2 1 1 1 1 1 1 1 1
5.e1
Brumbaugh et al FLA 5.5.0 DTD ■ YMPD9530_proof ■ November 30, 2017
ORIGINAL ARTICLES
An Intragastric Fecal Microbiota Transplantation Program for Treatment of Recurrent Clostridium difficile in Children is Efficacious, Safe, and Inexpensive David E. Brumbaugh, MD1,2, Edwin F. De Zoeten, MD, PhD1,2, Amy Pyo-Twist, RN3, Sara Fidanza, RN3, Shannon Hughes, PA4, Susan A. Dolan, RN, MS5, Jason Child, PharmD6, and Samuel R. Dominguez, MD, PhD4,5 Objective To assess the safety, efficacy, and relative expense of a nurse-led fecal microbiota transplantation (FMT) program for the treatment of recurrent Clostridium difficile infection (CDI).
Study design Retrospective cohort study design in children aged 1-18 years with recurrent CDI. The intervention was an intragastric FMT with stool derived from a donor stool bank. Primary outcome was resolution of diarrhea at 3 months post-transplantation. A secondary analysis compared charge data associated with FMT by intragastric delivery vs administration by colonoscopy or nasoduodenal tube. Results A total of 47 intragastric FMT procedures were performed in 42 children (median age 9 years) with recurrent CDI. Response to treatment varied by disease status, with 94% success in previously healthy children, 75% in medically complex children, and 54% in children with inflammatory bowel disease (P = .04). FMT via intragastric delivery showed lower facility and professional charges by 85% and 78% compared with delivery via colonoscopy and radiology-placed nasoduodenal tube, respectively. The use of stool derived from a donor stool bank decreased charges by 49% compared with charges associated with the use of a donor who was a relative. Conclusion A nurse-led intragastric FMT procedure using stool derived from a donor stool bank is a relatively inexpensive and efficacious treatment for recurrent CDI in children. Intragastric FMT success in children was attenuated by the presence of underlying disease, particularly inflammatory bowel disease. (J Pediatr 2017;■■:■■-■■). he prevalence of Clostridium difficile infection (CDI) is increasing in both adult and pediatric patients.1,2 Antibiotic therapy remains the standard first-line therapy for CDI, but the significant rise in highly virulent strains of C difficile and recurrence rates between 15% and 30% have challenged clinicians to discover alternative therapies for the treatment of CDI.3-5 Fecal microbiota transplantation (FMT) has emerged as the standard of care in the treatment of patients with recurrent CDI for whom antibiotics have not been successful.6,7 Although series of FMT for the treatment of recurrent CDI in pediatric patients have demonstrated efficacy rates equivalent to adults, overall reported experience in children remains small.8,9 There are variables of FMT that differentiate treatment protocols and potentially influence treatment efficacy and/or expense. These include the source of donor stool, route of donor stool delivery, setting of the procedure, and the type of healthcare professional who performs the procedure. The most common route of FMT delivery is via the lower gastrointestinal (GI) tract, either by colonoscopy or enema, with reported treatment success of 88%-95%.10 However, colonoscopy in particular is an expensive procedure that usually requires general anesthesia in children. In small series, the delivery of FMT into the upper GI tract has demonstrated efficacy approaching that of colonoscopy.9,11,12 The donor source for FMT traditionally has been a firstdegree relative, requiring the donor to undergo expensive testing to assess for potentially transmissible infectious agents before providing a donation to a single recipient. With the development of public stool banks, however, there is an opportunity for spreading the costs of donor testing across multiple FMT recipients.13 With heterogeneity in FMT protocols and gaps in our knowledge of treatment efficacy in children, there is opportunity to evaluate the value of this procedure in pediatrics. Value can be defined as healthcare outcome relative to cost.14 Therefore, the goal in maximizing value for the FMT procedure is to deliver donor stool at the lowest possible cost without sacrifice of clinical success, which has been approximately 90%.7 We describe the initial outcome and charge-savings data from a nurse-driven pediatric FMT program for the treatFrom the 1Section of Gastroenterology, Department of ment of recurrent CDI disease in children. Specific program design elements Pediatrics, University of Colorado School of Medicine;
T
CDI FMT GI IBD NGT RN
Clostridium difficile infection Fecal microbiota transplantation Gastrointestinal Inflammatory bowel disease Nasogastric feeding tube Registered nurse
2Children’s Hospital Colorado, Aurora, CO; 3Department of Nursing, Children’s Hospital Colorado; 4Section of Infectious Diseases, Department of Pediatrics, University of Colorado School of Medicine and Children’s Hospital Colorado; 5Department of Epidemiology, Children’s Hospital Colorado, Aurora, CO; and 6Department of Pharmacy, Children’s Hospital Colorado, Aurora, CO
The authors declare no conflicts of interest. 0022-3476/$ - see front matter. © 2017 Elsevier Inc. All rights reserved. https://doi.org10.1016/j.jpeds.2017.10.016
1 FLA 5.5.0 DTD ■ YMPD9530_proof ■ November 30, 2017
THE JOURNAL OF PEDIATRICS • www.jpeds.com intended to decrease costs of FMT included delivery in an ambulatory setting, intragastric administration of stool, use of a donor stool bank, and performance by a nurse rather than physician.
Methods The study was a retrospective cohort design with data extracted from the electronic health record. The study was approved by the Colorado Multiple Institutional Review Board. Inclusion criteria included all patients who received intragastric FMT at Children’s Hospital Colorado from March 2015 through September 2016. All patients received FMT as treatment for recurrent CDI, as defined by a positive fecal C difficile polymerase chain reaction test (Xpert C. Difficile; Cepheid, Sunnyvale, California) in combination with characteristic symptoms of diarrhea, hematochezia, or crampy abdominal pain. To be eligible for FMT, patients had to have greater than 2 episodes of CDI and have failed at least 1 course of therapy with oral vancomycin. Children were defined as medically complex if a diagnosis preceding CDI required chronic care by a pediatric subspecialist. All parents or guardians provided informed consent for FMT, and patients older than 7 years provided assent. Pretreatment for the procedure included a minimum of 5 days of oral vancomycin that was discontinued 2 days before FMT. All patients were treated with ranitidine for 24 hours immediately preceding FMT. On the day of FMT, patients had a nasogastric feeding tube (NGT) placed by a registered nurse (RN) in the ambulatory clinic. Some patients with a pre-existing gastrostomy tube received their intragastric FMT via their current tube. Correct intragastric placement of the NGT was confirmed by abdominal radiography. Once placement was confirmed, the FMT procedure was started. Donor stool was received from a stool donor bank (OpenBiome, Cambridge, Massachusetts) in 30-mL frozen aliquots or oral capsules and stored in a −20°C freezer. OpenBiome maintains a rigorous safety program that requires regular laboratory screening of its donors.15 A 30-mL aliquot was thawed immediately before the procedure. Once thawed, the 30-mL stool aliquot was infused over 2 minutes into the stomach through the gastric tube and flushed with 60 mL of tap water. Patients were observed by the RN for 10 minutes in the clinic before NGT removal, then an additional 10 minutes after NGT removal before discharge from the clinic. All families were contacted by phone or secure electronic correspondence at 2 weeks and 3 months post-FMT to assess treatment response and side effects of treatment. Treatment success was defined as the absence of diarrhea, hematochezia, and abdominal pain at 3 months post-FMT or, when symptoms persisted, a negative C difficile test. Patients who experienced a recurrence of symptoms before 3 months postFMT were offered the option to receive a second FMT in a procedure identical to the first. If symptoms were resolved with the second FMT, those patients were considered a treatment success.
Volume ■■
•
■■ 2017
Hypothetical analysis of procedural charges was conducted by obtaining average charge data from Children’s Hospital Colorado for the following standard procedures: diagnostic colonoscopy under general anesthesia and fluoroscopicassisted nasoduodenal tube placement. Charges included both technical and professional fees for these procedures. Charges associated with the actual FMT procedure included 2 hours of RN staffing, an ambulatory clinic charge to account for the 2 hours use of a clinic room, and both technical and interpretation charges associated with the abdominal radiography for NGT placement confirmation. Equipment costs (feeding tubes, tape, syringes, etc) were rolled into the procedural and ambulatory clinic charges. For comparison of charges by stool donor source, we obtained charge data from a standard commercial laboratory (LabCorp, Burlington, North Carolina) for the following tests: stool polymerase chain reaction for C difficile toxins A/B, stool bacterial culture, stool cryptosporidium and giardia antigen assay, stool ova and parasites, serum for HIV serotypes 1 and 2, hepatitis A (Anti-HAV IgM), hepatitis B (HBsAg and AntiHBs), hepatitis C (Anti-HCV IgG and IgM), and rapid plasma reagin. Charges for stool donor bank aliquots were based on procurement charges from OpenBiome, 20% of standard shipping charges (assuming 5 aliquots per shipment), and hospital pharmacy storage and administration charges. The Fisher exact test was used to assess whether group disease status (previously healthy, inflammatory bowel disease [IBD], and medically complex) and route of delivery (NGT, gastrostomy tube, and capsule) impacted treatment outcome.
Results During the study period, 42 unique patients received FMT at Children’s Hospital Colorado. The median age of the study population was 9 years (range, 1-18 years) and 45% were male. Thirty-one percent of the study population had IBD (Table I; available at www.jpeds.com) and 29% were medically complex with oncologic, metabolic, cardiopulmonary, or neurologic diagnoses (Table II; available at www.jpeds.com). Thirty patients were treated successfully with an initial FMT. After not responding to their first treatment, 5 patients elected to have a second FMT procedure within 3 months, and 2 of those patients achieved treatment success. The outcome of treatment success, therefore, was achieved in 32 of 42 patients (76%). There was a significant difference (P = .04) in treatment outcome by the presence of underlying disease (Figure 1). Ninety-four percent of otherwise-healthy children were treated successfully with FMT, whereas only 75% of medically complex children responded successfully. Treatment success dropped further to 54% in children with IBD. Success by method of upper GI tract FMT delivery was 71% by NGT and 67% by gastrostomy tube (P = .99). The single treatment by ingestion of 30 oral capsules was successful. The only reported postprocedural complication was vomiting within 24 hours of the FMT, which occurred in 6 of 47 (13%) procedures. In all cases, the vomiting was a single, selflimited episode that did not require medical treatment.
2
Brumbaugh et al FLA 5.5.0 DTD ■ YMPD9530_proof ■ November 30, 2017
■■ 2017
ORIGINAL ARTICLES
Figure 2. Overall hospital and professional charges associated with FMT by delivery method.
Figure 1. Pediatric FMT success rate by disease status. *P < .04 for difference in FMT success rate by disease status (Fisher exact test).
Total facility and professional charges were modeled for FMT as if administered via diagnostic colonoscopy under general anesthesia or via radiologist-placed nasoduodenal tube (no sedation). These hypothetical procedural charges were compared with charges from our FMT procedure, accomplished via NGT placed by a RN in a hospital outpatient setting (Figure 2). In this analysis, a stool bank was modeled as the stool donor source, and no differences in stool aliquot size or cost were assumed. Compared with FMT via colonoscopy ($7767) and nasoduodenal tube ($4998), FMT administration via nurse-placed NG tube ($1139) decreased total healthcare charges by 85% and 78%, respectively. Standard outpatient commercial laboratory charges for screening of a single first-degree relative as stool donor amounted to $1154 per stool donor. By comparison, the charges associated with procurement, storage, and administration of banked stool donor material from OpenBiome totaled $628, a 46% reduction in charges for donor stool.
Discussion In this study, administration of FMT via upper GI tract delivery to children with recurrent CDI achieved a high rate of treatment success (94%) in otherwise-healthy patients. These data add to previous reports suggesting that upper GI tract
delivery is as effective as lower tract GI delivery in treating recurrent CDI; and frozen, unrelated donor stool is as effective as fresh donor stool obtained from a close relative. Furthermore, we demonstrate that FMT delivery by this method is safe, well tolerated, and inexpensive compared with FMT delivered by colonoscopy or nasoduodenal tube. However, using this FMT technique, we observed attenuated treatment success in children with significant comorbidities, particularly IBD. We do not know whether our attenuated treatment success in children with IBD and medically complexity represents a limitation of our treatment strategy or represents actual real-world efficacy in these specific pediatric patient populations. There are limited data reporting outcomes from FMT in pediatric patients with both IBD and recurrent CDI. Using FMT delivery via colonoscopy, Russell et al described success in 2 of 3 patients whereas Hourigan et al observed 100% success in 5 patients.8,16 There is a significant challenge for patients with IBD in gauging the relative contribution of IBD activity vs CDI in driving patient symptoms. The presence of infection may trigger an inflammatory cascade and IBD flare that persists even after resolution of the infection. In the setting of IBD, FMT also has been associated with worsening IBD symptoms in up to 14% of recipients.17 Asymptomatic C difficile carriage occurs in patients with IBD at a rate 6 times that of healthy controls.18 Asymptomatic carriage may confound interpretation of C difficile testing in this population, lending plausibility to the theory that our diminished FMT success in patients with IBD was due in part to IBD activity and C difficile colonization rather than active CDI. Alternatively, larger numbers of donor bacteria may be required for treatment success of recurrent CDI in patients with IBD. Delivery of greater bacterial counts may be
An Intragastric Fecal Microbiota Transplantation Program for Treatment of Recurrent Clostridium difficile in Children is Efficacious, Safe, and Inexpensive FLA 5.5.0 DTD ■ YMPD9530_proof ■ November 30, 2017
3
THE JOURNAL OF PEDIATRICS • www.jpeds.com achievable via colonoscopy compared with upper GI tract delivery, as aliquots of stool donor specimens typically are larger in FMT via colonoscopy and the passage of bacteria through the stomach and small bowel may reduce the viable bacterial load reaching the colon. However, the ultimate success of FMT may not be attributable directly to the size of the bacterial transfer. Emerging data suggest that specific strains of bacteria, specifically those critical to production of secondary bile acids, are deficient in patients with recurrent CDI and restored in patients who achieve treatment success.19 The presence of non-IBD medical complexity, including preexisting oncologic and neurologic disease, also decreased the likelihood of treatment success. This group of children may be at greater risk of CDI recurrence post-FMT as the result of increased contact with healthcare facilities, greater levels of antibiotic exposure, and the presence of indwelling GI feeding devices.20,21 There are important considerations about treatment safety in the medically complex pediatric population. Patients with complex craniofacial anatomy may not be appropriate for NG insertion. Patients at high risk for gastroesophageal reflux and/or aspiration, including patients under general anesthesia, are not good candidates for intragastric FMT delivery.22 If there is concern for undiagnosed IBD or other GI pathology, FMT via colonoscopy may be preferable, as this procedure has the potential to be both diagnostic and therapeutic. Our approach to FMT in children, summarized as a nursedriven procedure using upper GI tract delivery of stool acquired from a donor stool bank, has many benefits. First, delivery by upper tract incurs a fraction of the healthcare charges compared with the traditional route of FMT by colonoscopy. Second, we have shown treatment success comparable with published data in otherwise-healthy children receiving FMT via colonoscopic administration. Lastly, our data show that use of a donor stool bank is as effective as use of related donor stool, again decreasing expense. If CDI continues to increase in incidence in the pediatric population, we anticipate that FMT will be increasingly sought by patient families and providers. Lowering charges associated with FMT will ensure that this highly effective therapy is accessible to patients. Strengths of this study include the relatively large number of subjects, the largest reported series of FMT in children. The assessment of treatment response was consistent for all subjects at 3 months post-FMT. Our study has several limitations: the research was retrospective, so although we used a standardized data collection form, clinical evaluations and treatment may have differed depending on the individual providers who were caring for these children. As there is currently no assay available to differentiate C difficile colonization from infection, interpretation of C difficile testing remains challenging in children with high rates of colonization, specifically children with IBD or frequent contact with hospital environments.23 Several children with IBD classified in our study as FMT treatment failures may have had active IBD with C difficile colonization.
Volume ■■ In conclusion, our study provides important treatment efficacy data from a novel, nurse-led FMT program in a heterogeneous patient population at a tertiary pediatric hospital. We demonstrate very high levels of treatment success in previously healthy children using a relatively inexpensive FMT procedure. Treatment success was attenuated significantly in patients with IBD. ■ Submitted for publication Jun 15, 2017; last revision received Sep 26, 2017; accepted Oct 12, 2017 Reprint requests: David E. Brumbaugh, MD, Children’s Hospital Colorado, 13123 E. 16th Ave, B290, Aurora, CO 80045. E-mail: david.brumbaugh@ childrenscolorado.org
References 1. Ricciardi R, Rothenberger DA, Madoff RD, Baxter NN. Increasing prevalence and severity of Clostridium difficile colitis in hospitalized patients in the United States. Arch Surg 2007;142:624-31, discussion 31. 2. Deshpande A, Pant C, Anderson MP, Donskey CJ, Sferra TJ. Clostridium difficile infection in the hospitalized pediatric population: increasing trend in disease incidence. Pediatr Infect Dis J 2013;32:113840. 3. Redelings MD, Sorvillo F, Mascola L. Increase in Clostridium difficilerelated mortality rates, United States, 1999-2004. Emerg Infect Dis 2007;13:1417-9. 4. Lessa FC, Gould CV, McDonald LC. Current status of Clostridium difficile infection epidemiology. Clin Infect Dis 2012;55(suppl 2):S6570. 5. Shields K, Araujo-Castillo RV, Theethira TG, Alonso CD, Kelly CP. Recurrent Clostridium difficile infection: from colonization to cure. Anaerobe 2015;34:59-73. 6. Surawicz CM. Clostridium difficile infection: risk factors, diagnosis and management. Curr Treat Options Gastroenterol 2015;13:121-9. 7. Kassam Z, Lee CH, Yuan Y, Hunt RH. Fecal microbiota transplantation for Clostridium difficile infection: systematic review and meta-analysis. Am J Gastroenterol 2013;108:500-8. 8. Russell GH, Kaplan JL, Youngster I, Baril-Dore M, Schindelar L, Hohmann E, et al. Fecal transplant for recurrent Clostridium difficile infection in children with and without inflammatory bowel disease. J Pediatr Gastroenterol Nutr 2014;58:588-92. 9. Kronman MP, Nielson HJ, Adler AL, Giefer MJ, Wahbeh G, Singh N, et al. Fecal microbiota transplantation via nasogastric tube for recurrent Clostridium difficile infection in pediatric patients. J Pediatr Gastroenterol Nutr 2015;60:23-6. 10. Gough E, Shaikh H, Manges AR. Systematic review of intestinal microbiota transplantation (fecal bacteriotherapy) for recurrent Clostridium difficile infection. Clin Infect Dis 2011;53:994-1002. 11. van Nood E, Dijkgraaf MG, Keller JJ. Duodenal infusion of feces for recurrent Clostridium difficile. N Engl J Med 2013;368:2145. 12. Aas J, Gessert CE, Bakken JS. Recurrent Clostridium difficile colitis: case series involving 18 patients treated with donor stool administered via a nasogastric tube. Clin Infect Dis 2003;36:580-5. 13. Kazerouni A, Burgess J, Burns LJ, Wein LM. Optimal screening and donor management in a public stool bank. Microbiome 2015;3:75. 14. Porter ME. What is value in health care? N Engl J Med 2010;363:247781. 15. Openbiome. Quality and Safety Program Somerville, MA. https:// static1.squarespace.com/static/50e0c29ae4b0a05702af7e6a/t/59304 bdf3e00be24622791fb/1496337376745/The+OpenBiome+Quality+ %26+Safety+Program.pdf. Accessed October, 2017. 16. Hourigan SK, Chen LA, Grigoryan Z, Laroche G, Weidner M, Sears CL, et al. Microbiome changes associated with sustained eradication of Clostridium difficile after single faecal microbiota transplantation in children with and without inflammatory bowel disease. Aliment Pharmacol Ther 2015;42:741-52.
4
Brumbaugh et al FLA 5.5.0 DTD ■ YMPD9530_proof ■ November 30, 2017
■■ 2017
ORIGINAL ARTICLES
17. Kelly CR, Ihunnah C, Fischer M, Khoruts A, Surawicz C, Afzali A, et al. Fecal microbiota transplant for treatment of Clostridium difficile infection in immunocompromised patients. Am J Gastroenterol 2014;109:106571. 18. Hourigan SK, Chirumamilla SR, Ross T, Golub JE, Rabizadeh S, Saeed SA, et al. Clostridium difficile carriage and serum antitoxin responses in children with inflammatory bowel disease. Inflamm Bowel Dis 2013;19:2744-52. 19. Staley C, Kelly CR, Brandt LJ, Khoruts A, Sadowsky MJ. Complete Microbiota engraftment is not essential for recovery from recurrent Clostridium difficile infection following fecal microbiota transplantation. MBio 2016;7.
20. Eyre DW, Walker AS, Wyllie D, Dingle KE, Griffiths D, Finney J, et al. Predictors of first recurrence of Clostridium difficile infection: implications for initial management. Clin Infect Dis 2012;55(suppl 2):S77-87. 21. Sandora TJ, Fung M, Flaherty K, Helsing L, Scanlon P, Potter-Bynoe G, et al. Epidemiology and risk factors for Clostridium difficile infection in children. Pediatr Infect Dis J 2011;30:580-4. 22. Baxter M, Ahmad T, Colville A, Sheridan R. Fatal aspiration pneumonia as a complication of fecal microbiota transplant. Clin Infect Dis 2015;61:136-7. 23. Dominguez SR, Dolan SA, West K, Dantes RB, Epson E, Friedman D, et al. High colonization rate and prolonged shedding of Clostridium difficile in pediatric oncology patients. Clin Infect Dis 2014;59:401-3.
An Intragastric Fecal Microbiota Transplantation Program for Treatment of Recurrent Clostridium difficile in Children is Efficacious, Safe, and Inexpensive FLA 5.5.0 DTD ■ YMPD9530_proof ■ November 30, 2017
5
THE JOURNAL OF PEDIATRICS • www.jpeds.com
Volume ■■
Table I. Subject IBD disease type and treatment response Subject no.
Sex
IBD type
FMT success?
No. FMT attempts
F F M M M F M M F M F M M
Ulcerative colitis Crohn's disease Ulcerative colitis Ulcerative colitis Ulcerative colitis Ulcerative colitis Ulcerative colitis Ulcerative colitis Crohn's disease Crohn's disease Crohn's disease Ulcerative colitis Ulcerative colitis
Yes Yes Yes No Yes No No No Yes Yes No Yes No
1 1 1 2 1 1 1 1 1 2 1 1 1
1 2 3 4 5 6 7 8 9 10 11 12 13 F, female; M, male.
Table II. Diagnoses of subjects with medical complexity and treatment response Subject no. 14 15 16 17 18 19 20 21 22 23 24 24
Sex
Diagnoses
FMT success?
No. FMT attempts
M F F F F M M M F M M M
Rhabdomyosarcoma Epilepsy, malnutrition Disorder of mitochondrial metabolism Intestinal dysmotility, neutropenia, seizures, neurofibromatosis Celiac disease, tuberous sclerosis, seizures, autism Genetic syndrome, seizures, dysphagia Pancreatitis, seizures Rhabdomyosarcoma Anoxic brain damage, cerebral palsy, tracheostomy Chronic lung disease, feeding difficulties, venous thrombus Complex congenital heart disease, chronic pulmonary disease Seizures, intellectual disability, cardiofaciocutaneous syndrome
Yes Yes No Yes Yes Yes No Yes Yes Yes No Yes
1 1 1 2 1 1 1 1 1 1 1 1
5.e1
Brumbaugh et al FLA 5.5.0 DTD ■ YMPD9530_proof ■ November 30, 2017