Faecal microbiota transplantation for refractory C lostridium difficile infection

Abstract

Faecal microbiota transplantation (FMT) has become a part of the treatment algorithm for Clostridium difficile infection (CDI), particularly for recurrent infections when antibiotics have diminishing efficacy. Notably, despite a significant proportion of patients suffering from refractory disease, there is a general lack of evidence describing the use of FMT in this patient cohort. We present here a case of successful treatment of refractory CDI in a patient under critical care.

Background

Clostridium difficile infection (CDI) is a very common, nosocomial infection with a high healthcare burden due to its protracted course and frequent relapses despite appropriate medical therapy. The emergence of faecal microbiota transplantation (FMT) as an effective, durable therapy for CDI has rapidly altered the treatment paradigm for patients with this condition. Society guidelines worldwide now reflect this and recommend the use of FMT particularly in recurrent CDI. FMT is also recommended for refractory CDI; however, the evidence for this practice is limited.

Case presentation

We report a case of successful FMT for refractory CDI in a patient under critical care. The patient, a 71-year-old woman, was admitted electively for surgical reversal of loop ileostomy. The ileostomy was formed 1 year prior, following repair of a strangulated hernia and associated small bowel resection. Medical history included large congenital atrial septal defect with left-to-right shunt (and resultant pulmonary hypertension), mitral and tricuspid regurgitation, atrial fibrillation, hypertension and hypercholesterolaemia. Medications included warfarin, bisoprolol, ramipril and frusemide.

Following surgery, postoperative recovery was stable for 72 hours. On postoperative day 4, clinical deterioration secondary to vomiting was noted, together with significant oxygen desaturation, thought to represent pulmonary aspiration. Two doses each of cefuroxime and metronidazole were administered. High dependency unit admission was arranged for administration of high flow oxygen therapy. Lobar consolidation with high inflammatory markers and fever was treated empirically with 5 days of co-amoxiclav and a single administration of gentamicin. On postoperative day 9, there was a large gastrointestinal (GI) bleeding event, which was found to be originating from the anastomotic site. This required interventional angiography and embolisation of a branch of the superior mesenteric artery which was successful in achieving haemostasis. Further infective lung pathology with Escherichia coli sp was initially treated with 6 days of piperacillin-tazobactam and gentamicin. Despite this, septic shock intervened with Candida albicans in sputum and urine, Enterococcus sp in surgical wound and E. coli sp in urine. The piperacillin-tazobactam and gentamicin course was extended to 13 days on review with the addition of fluconazole for 7 days. Cardiac failure and acute kidney injury necessitated intensive care unit (ICU) admission for invasive ventilation, vasopressors and renal replacement therapy. Persistent growth of E. c oli in sputum was treated with 5 days of ciprofloxacin, following which antimicrobial therapy was ceased.

Following clinical improvements, a percutaneous tracheostomy was placed 11 days after ICU admission to aid ventilatory weaning. At this time, the patient developed abdominal pain and profuse, watery stools (Bristol stool chart type 6 and 7, up to seven times per day) together with fever up to 38°C. The patient was isolated and a faecal management tube was inserted rectally, and large daily outputs (>300 mL/day) were noted. Clostridium difficile testing (including detection of GDH, PCR for toxin gene and toxin A/B EIA) was positive. Fidaxomicin (200 mg two times per day) was started, according to our local policy which recommends this agent first line for all adult patients, as previously described.¹ Despite 10 days of fidaxomicin therapy, ongoing profuse, watery diarrhoea was observed. Repeat testing for C. difficile remained positive and the patient had a repeat course of fidaxomicin, which also made no significant symptomatic difference. The organism was not cultured, so we were unable to perform antimicrobial susceptibility testing; however, resistance to fidaxomicin is rare with a recent study reporting no resistant strains among 1889 tested.² The clinical course was further complicated by a hospital-acquired pneumonia. A resistant E. coli and Pseudomonas aeruginosa were isolated on sputum sample and treated with 5 days of meropenem, following which there was good clinical resolution.

Although there were signs of clinically severe CDI (white blood cell count 20×10⁹/L, albumin 30 g/L, low-grade fevers and impaired renal function), this was confounded by the numerous concomitant, contributing medical complications outlined above. Regardless, as the CDI was refractory to standard medical therapy, FMT was arranged. Antibiotics were discontinued 48 prior to FMT, as per our unit’s protocol. Our unit performs FMT using 200 mL of frozen-thawed material derived from >50 g of screened donor faeces. Healthy volunteers donate to our stool bank and are unknown to the recipient, they receive no financial incentive, other than travel expenses. Our methods and protocol for FMT are described elsewhere.³ The FMT suspension was delivered via colonoscopy, although due to severe diverticular disease it was not possible to reach the caecum. As a result, the suspension was delivered into the proximal sigmoid colon.

Outcome and follow-up

Despite suboptimal delivery of the FMT suspension, the patient had a significant and rapid clinical response. Within 24 hours, the volume of diarrhoea from her rectal tube had reduced drastically and within 48 hours it was possible to remove the rectal tube altogether. The consistency of her stools improved as well and she was able to be stepped down from the ICU 3 days after her FMT. She was discharged home from hospital 3 weeks after her FMT with no adverse clinical sequelae. The patient remains symptom free at the time of writing, 8 months after the procedure.

Discussion

Antibiotic therapy to treat primary CDI has historically had a high recurrence rate, with studies showing treatment failure rates of 15%–30%.^{4 5} Recurrent C. difficile is associated with severe disruption to the colonic microbiome, most commonly through antibiotic exposure, resulting in loss of colonisation resistance. The rationale of FMT is to restore the normal colonic microbiome.

The literature to date regarding FMT in C. difficile is quite heterogeneous and mainly focusses on patients with recurrent (rather than refractory) disease. The terms ‘refractory’ and ‘recurrent’ are often used interchangeably and therefore limit the ability to perform informative subgroup analysis. The randomised trials investigating the utility of FMT in CDI included only 19 patients with refractory CDI and were not powered to assess a difference between these groups.^{6 7} Numerous case series that coincidentally included patients with refractory disease have been published, but the specific outcomes for these patients are not separately reported. In addition, there are no studies that compare FMT in refractory CDI to standard therapy, mainly due to the small numbers of such patients. As a result, while major guidelines do recommend the use of FMT in refractory CDI, this is often based on limited published data and extrapolation from the recurrent CDI population.⁸ Often inconsistent terminology is used when referring to recurrent CDI. We recommend avoiding the term resistant infection as this implies non-susceptibility to antimicrobial agents. Refractory C. difficile is usually used to indicate poor or partial response to treatment. Recurrent CDI suggests initial improvement or complete resolution of symptoms followed (after a variable period of time) by their reappearance. Recurrence can be further classified as a relapse (caused by the same infecting strain) or reinfection (with a new strain).

There is a paucity of evidence guiding the use of FMT in patients admitted to the ICU, primarily because the bulk of published trials excluded patients in the ICU and those requiring circulatory support or vasopressors.^9–12 As a result, the use of FMT in this select population is supported only by case reports and case series.^13–15 The studies show that admission to ICU or critical illness does not preclude the use of FMT, as the only alternative may be a surgical approach, such as colectomy or loop ileostomy, which has a mortality rate of over 50%.¹⁶ Kelly et al have even shown FMT to be effective in immunosuppressed patients with few adverse events. It is therefore reasonable to consider the use of FMT in a broader range of patients than that suggested by the initial trials, especially when alternative therapeutic options are limited or have significant associated risk.

Recently, Hvas et al showed FMT, preceded by 4–10 days of vancomycin 125 mg four times a day, was superior to fidaxomicin or vancomycin alone in the treatment of recurrent CDI.¹² However, once again, no clear distinction was made between refractory and recurrent CDI and severe disease or critical illness was an exclusion criteria, slightly limiting the study’s generalisability. No difference could be detected between fidaxomicin and vancomycin, although the study was underpowered to detect this difference. Fidaxomicin has previously been shown to be non-inferior to vancomycin with regard to mild to moderate CDI, with the benefit of lower recurrence rates.⁵ Two courses of fidaxomicin failed to make any noticeable difference in our patient’s clinical symptoms.

As a result of a failure to improve with medical therapy, the patient proceeded to have an FMT. We determined that the patient would be able to tolerate a colonoscopy, and hence employed this modality rather than nasojejunal delivery. Colonoscopic delivery has been shown to have a higher response rate compared with an upper GI mode of delivery based on pooled response rates (95% vs 90%, p=0.02).¹⁷ Hence, our practice is to perform FMT via the colonoscopic route in the majority of cases (provided there is no contraindication to colonoscopy).

Any algorithm for the treatment of patients with refractory CDI in the ICU is based on a sparse body of evidence. Patients like the one reported, with refractory (rather than recurrent) disease, and those admitted to ICU are rarely reported on in the published literature. The case presented advocates the use of FMT in refractory CDI, including carefully selected patients admitted to the ICU or with moderate-severe disease. Colonoscopic delivery is recommended in those who can safely tolerate sedation, although an upper GI route is acceptable in the remaining patients. We also suggest broadening the inclusion criteria in future studies investigating FMT in refractory CDI to better understand its utility in treating this condition.

Learning points.

• Refractory Clostridium difficile infection (CDI) is a common and debilitating condition, and can be challenging to treat.

• Faecal microbiota transplantation (FMT) should be considered as part of the therapeutic options for refractory CDI.

• FMT can be used in patients in the ICU and in immunosuppressed patients.