Abstract
Background:
Fecal microbiota, live-jslm (RBL; REBYOTA®), is the first Food and Drug Administration (FDA)-approved, single-dose, rectally administered, microbiota-based live biotherapeutic product for preventing Clostridioides difficile infection (CDI) recurrence. Alternative routes of administration are of clinical interest.
Objectives:
Evaluate the safety and efficacy of RBL administration via colonoscopy.
Design:
Retrospective analysis of electronic medical records of participants administered RBL via colonoscopy under FDA enforcement discretion.
Methods:
The number of participants with treatment and/or procedure-emergent adverse events (TEAEs) was evaluated. Treatment success and sustained clinical response, defined as the absence of CDI recurrence within 8 weeks and 6 months, respectively, were evaluated.
Results:
TEAEs were experienced by 75% (6/8) of participants; most were mild to moderate in severity, and none due to RBL or its administration. Most participants had treatment success (80%; 8/10); 75% (6/8) had sustained clinical response.
Conclusion:
Real-world safety and efficacy of RBL administered via colonoscopy were consistent with clinical trials of rectally administered RBL.
Keywords
Background
Clostridioides difficile infection (CDI) causes approximately half a million infections per year in the United States, leading to as many as 30,000 deaths annually.1,2 Approximately 35% of patients treated for an initial episode and as high as 65% of patients with one or more prior recurrences may experience recurrent CDI (rCDI).3–6 Gut dysbiosis and exposure to C. difficile increase the risk of CDI and rCDI, with symptoms ranging from mild to severe diarrhea and additional complications such as toxic megacolon, sepsis, and death. 7 Although standard-of-care (SOC) antibiotic therapy with vancomycin or fidaxomicin is recommended for primary and recurrent episodes of CDI, antibiotic therapy does not directly address underlying gut dysbiosis. 8 Because the likelihood of subsequent CDI increases with each successive episode,3,9,10 there is a need for safe and efficacious approaches to microbiome restoration. 11
Microbiota-based, live biotherapeutic products (LBPs) were evaluated for prevention of rCDI in randomized controlled trials, demonstrating efficacy and safety in patients with rCDI.12,13 Fecal microbiota, live-jslm (REBYOTA®, abbreviated as RBL, previously RBX2660), is the first single-dose, rectally administered, microbiota-based, LBP approved by the US Food and Drug Administration (FDA) to prevent rCDI in individuals aged ⩾18 years following SOC antibiotic treatment for rCDI. In prospective phase II clinical trials and a placebo-controlled pivotal phase III trial, RBL reduced the recurrence of CDI and provided sustained clinical response at least 6 months after administration.13–15 RBL responders reported durable responses up to 24 months after administration in a phase II open-label study. 15 To evaluate the efficacy and safety of RBL, these clinical studies required stringent inclusion and exclusion criteria, and participants with inflammatory bowel disease, irritable bowel syndrome, and immunocompromising conditions were excluded.
As part of the Assured Active Treatment (AAT) compassionate use program, participants ineligible for enrollment in prior clinical trials who required additional therapy for the prevention of rCDI were provided RBL at clinical trial sites in accordance with FDA enforcement discretion (ED) policy. 16 Based on this regulatory guidance, participant eligibility to receive rectally administered RBL under ED in this retrospective analysis was assessed by site physicians to ensure a broad study population that more closely mimicked real-world practice. Most treatment-emergent adverse events (TEAEs) were mild to moderate in severity and comparable between comorbidity subgroups and the overall population. 17 As with the results of prospective phase II and phase III trials of RBL, there were no serious adverse events (AEs) assessed as related to RBL. At 8 weeks, 82.8% of RBL recipients had treatment success, of which 88.7% had a sustained response through 6 months.
Clinical guidelines have historically recommended microbiome restoration with fecal microbiota transplantation (FMT) administration via colonoscopy or oral capsules; rectal administration is suggested if other approaches are unavailable. 18 RBL is unique in that it can potentially be administered via colonoscopy without alteration of the active material. Colonoscopic administration may be preferred in certain clinical cases, leading to interest in the use of RBL via this route. Since a subset of participants in the AAT program received RBL via colonoscopy based on investigator discretion, this descriptive subgroup analysis evaluated the safety and efficacy of RBL administered via colonoscopy for the prevention of rCDI.
Methods
This subgroup analysis includes participants from a prior retrospective, multicenter, single-arm, real-world analysis of electronic medical records (EMRs) of participants with CDI aged ⩾18 years who received RBL under ED between 1 November 2015 and 30 September 2019, across five study sites in the United States, as described previously. 17 Briefly, site investigators determined participant eligibility, including obtaining and evaluating diagnostic testing, RBL dosing regimens, and methods of RBL administration. Each RBL dose consisted of a 150-mL suspension containing between 1 × 108 and 5 × 1010 colony-forming units (CFU) per milliliter (mL) of fecal microbes, including >1 × 105 CFU/mL of Bacteroides. An on-study episode of CDI was defined as a record of treatment of CDI (either FMT or SOC antibiotics) or CDI-related diarrhea during the 6-month study period. Laboratory diagnostic testing was not required to confirm an on-study episode of CDI. Efficacy and safety results were analyzed for two different populations: (1) the full analysis set (FAS; N = 94), which included all participants who met eligibility criteria and received RBL during the study period and (2) the primary safety set (PSS; N = 64), which included participants naïve to RBL administration who met the study requirement for continuously comprehensive medical records for 6 full months of safety follow-up. Participants without comprehensive medical records were conservatively considered failures even if they did not experience a CDI event and were excluded from the primary safety analysis. The secondary safety set (SSS; n = 30) included participants who were enrolled but excluded from the PSS because they received RBL during a prior clinical study, did not have continuous medical records for 6 months after administration of RBL, or exited the study early. All occurrences of participants receiving RBL via colonoscopy were at two sites. As RBL administration via colonoscopy was not described in the original study protocol, no standard methods of colonoscopic administration or participant selection were defined. 17 Methods of RBL administration via colonoscopy were at the investigator’s discretion.
The primary objective was to evaluate the number of participants with RBL- and/or procedure-related TEAEs, defined as AEs, occurring on or after the day of RBL administration via colonoscopy. Supportive safety analyses included the number of TEAEs of special interest [gastrointestinal (GI) AEs, fever, chills, fatigue], the relatedness of TEAEs, the severity of TEAEs, and the onset of new chronic conditions relative to RBL administration or poor prognosis of preexisting conditions. Safety analyses were conducted only in participants with comprehensive medical records. Treatment efficacy parameters evaluated included treatment success (absence of CDI recurrence through 8 weeks after the final dose of RBL administered for the qualifying CDI event) and sustained clinical response (absence of CDI recurrence at 8 weeks and through 6 months after the final dose of RBL administered for the qualifying CDI event).
The protocol received IRB/research ethics board approval before commencement at participating sites according to local requirements (Arkansas Gastroenterology, SSU00119272; and Kaiser Permanente Springfield Medical Center, FWA00002344). Statistical analyses were conducted using descriptive statistics. As a retrospective subgroup analysis, there were no methodological measures to address potential sources of bias.
The reporting of this study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement. 19
Results
This analysis included 10 participants who received RBL under ED via colonoscopy for the qualifying event. Of these participants, four were in the PSS and six were in the SSS. Two of the six participants in the SSS did not have continuous medical records, and five of the six participants received RBL rectally in a prospective study prior to RBL administration under ED. Participants (n = 10) were mostly female, aged 22–80 years, and White (Table 1). Medical histories included GI disorders (32 occurrences), psychiatric disorders (11 occurrences), and vascular disorders (10 occurrences). The most common medical comorbidities were gastroesophageal reflux disease (eight occurrences), anxiety (five occurrences), hypertension (five occurrences), gastritis (four occurrences), depression (three occurrences), diverticular disease (three occurrences), insomnia (three occurrences), and vitamin D deficiency (three occurrences). Among eight participants who received RBL via colonoscopy and had comprehensive medical records, six (75%) reported TEAEs; most were mild to moderate in severity; GI disorders and infections were the most common AEs (Table 1). The specific moderate or severe AEs were as follows: participant 2 (CDI, urinary tract infection), participant 9 (upper respiratory tract infection, herpes zoster), and participant 10 (vulvovaginal candidiasis, cellulitis); participant 4 had two separate episodes of ileus. No TEAEs, including ileus and infections, were assessed as related to RBL or its administration.
Demographics and AEs for up to 6 months in participants administered RBL via colonoscopy.
A full CDI history was not required for participation in the study, and the episodes presented are complete as captured in the provided medical history records. AEs by system organ class were gastrointestinal disorders (n = 8), infections (n = 8), investigations (n = 1), metabolism and nutrition disorders (n = 1), musculoskeletal and connective tissue disorders (n = 1), nervous system disorders (n = 2), and reproductive system and breast disorders (n = 1). The preferred term for the AE in the system organ class investigations was urine analysis abnormal.
Medication used for treating a qualifying CDI event.
Participants 1 and 6 lacked comprehensive medical records up to 6 months after RBL administration. Unavailable medical record data are indicated by ‘–’.
Includes all events related and unrelated to CDI.
Participant 2 experienced one CDI-related event.
Participant 4 experienced two CDI-related events.
AE, adverse event; CDI, Clostridioides difficile infection; ED, enforcement discretion; F, female; FDX, fidaxomicin; M, male; N, no; PSS, primary safety set; RBL, fecal microbiota, live-jslm; VAN, vancomycin; Y, yes.
Treatment success rates of 80% and 100% were reported in FAS and PSS, respectively, for participants who received RBL via colonoscopy (Table 2). As these participants were a small subset of a larger study cohort, treatment success rates for participants who received RBL rectally are included for reference. Two participants who received RBL via colonoscopy had incomplete EMRs and were conservatively classified as having treatment failure. Of eight participants who experienced treatment success at 8 weeks, 75% (6/8) had sustained clinical response at 6 months. Participants 4 and 2 had CDI recurrence at days 59 and 103, respectively; participant 4 received a second dose of RBL via colonoscopy at 6 months after initial administration.
Treatment success at 8 weeks and sustained clinical response through 6 months after RBL administration via colonoscopy.
RBL doses included participants who received (as one course) one, two, or three doses of RBL via rectal administration. Data for RBL doses administered rectally have been described previously. 17
Two participants were classified as having treatment failure due to the lack of comprehensive EMRs up to 6 months after RBL administration via colonoscopy.
EMR, electronic medical record; FAS, full analysis set; PSS, primary safety set; RBL, fecal microbiota, live-jslm.
Discussion
In this subgroup analysis of a prior retrospective study, participants who had RBL administered via colonoscopy had an 80% (8/10) treatment success rate after 8 weeks of the last RBL dose; 75% of these participants had a sustained clinical response at 6 months. Most TEAEs were mild to moderate in severity, and none were assessed as related to RBL or its administration via colonoscopy. The 80% treatment success rate may be a conservative estimate given that two treatment failures were due to loss to follow-up. This analysis represents the first real-world evaluation of the safety and efficacy of a microbiota-based LBP administered via colonoscopy to prevent rCDI. Although a colonoscopic administration strategy is considered resource-intensive, frequently involves sedation, and is invasive for patients, some physicians prefer this delivery method because it permits direct visualization of the colonic mucosa while delivering microbiota-based LBP and allows for assessment for other pathologic findings including inflammatory bowel disease, polyps, and colon cancer. 20
The safety profile of RBL administered via colonoscopy was comparable to phase II and phase III randomized controlled trials and open-label studies using a rectal administration strategy. An integrated analysis of three phase II trials (PUNCH CD, PUNCH CD2, PUNCH Open-Label) and two phase III trials (PUNCH CD3, PUNCH CD3-OLS) showed proportions of participants with TEAEs were 60.2% and 66.4% for participants administered placebo and RBL, respectively. Most TEAEs were mild to moderate in severity. 12 The majority of TEAEs were related to preexisting conditions such as CDI recurrence and GI-related comorbidities. 12 Similarly in this analysis, TEAEs were reported in six of eight participants (75%) who received RBL via colonoscopy; the most common TEAEs were GI disorders and infections. Most of these participants had one or more GI comorbidities (e.g. gastritis, microscopic colitis, irritable bowel syndrome, diverticular disease, a history of megacolon, intestinal metaplasia of the gastric mucosa, and cholelithiasis) with or without medical histories denoting potentially complex anatomical challenges (e.g. colectomy, liver transplantation, colonic polypectomy). As in the integrated safety analysis from the broader clinical study program, no infections were assessed as related to RBL, likely a result of FDA-recommended pathogen screening. 12
The treatment success rate of 80% and sustained clinical response rate of 75% for participants administered RBL via colonoscopy were comparable to results from the phase III PUNCH CD3 randomized controlled trial (70.6% Bayesian model-estimated treatment success and 92.1% sustained clinical response), 13 and the AAT program (82.8% overall response rate and 88.7% sustained clinical response). 17
The main limitation of the current analysis is the small number of participants administered RBL via colonoscopy (n = 10). Treatment effects via colonoscopy could not be fully evaluated due to the lack of a placebo control arm. Since this analysis provides only a general framework for RBL administration via colonoscopy, specific administration methods or patient selection data were not prospectively collected. Because all participants were included in the original retrospective study under ED, irrespective of response to RBL, and a qualifying CDI event was defined by the first exposure to RBL, potential treatment outcome selection bias was likely addressed.
Conclusion
RBL administered via colonoscopy demonstrated no new safety signals from previous trials. Future studies including more patients and more robust study design, such as the CDI SCOPE trial (NCT05831189), will help validate the real-world effectiveness of RBL when administered via colonoscopy.
Footnotes
Acknowledgements
The authors thank all the participants and their families and caregivers as well as the investigators and site staff. Medical writing support, under the guidance of the authors, was provided by Paul Sobol, PhD, Robin Isaac, PharmD, and Ian Mbano, PhD (ApotheCom, Yardley, PA, USA) and was funded by Ferring Pharmaceuticals, Parsippany, NJ, USA.
