|Home | About | Journals | Submit | Contact Us | Français|
Clostridium difficile (CD) infection is almost always confined to the colon causing a spectrum of illness ranging from diarrhoea to fulminant colitis. CD infection of the small intestine has been described but the identification of CD toxin in the stoma effluent of a patient with an end ileostomy is rare. We describe a 91-year-old woman, with a history of proctocolectomy for ulcerative colitis, presenting with profuse ileostomy diarrhoea after a course of antibiotics. Ileostomy effluent was positive for CD toxin but the patient died despite appropriate treatment. This suggests that the small intestine is susceptible to CD infection in antibiotic-treated patients many years after a colectomy. CD enteritis should be considered in all patients with increased ileostomy diarrhoea despite the absence of a colon.
Clostridium difficile (CD) infection is a major cause of hospital acquired infection. It causes a toxin-mediated mucosal inflammatory process and is associated with increased morbidity and mortality, especially in elderly hospitalised patients. CD infection is typically a colonic infection and has been only rarely described affecting the small intestine.1
We describe a patient who had diarrhoeal symptoms due to hospital-acquired CD infection in her ileostomy effluent many years after colectomy.
A 91-year-old woman was admitted to hospital from a residential home in November 2007 with a 3-day history of profuse ileostomy diarrhoea associated with anorexia, nausea, lethargy and reduced mobility. The patient reported that she had been changing her ileostomy bag with an increased frequency and had been experiencing leakage of green and watery effluent. Apart from two transient episodes of subacute bowel obstruction, she had no prior history of similar bowel disturbances since having an ileostomy. Nobody else in the home or hospital had suffered with a diarrhoeal illness.
Two weeks before admission she had been discharged following a 3-week inpatient stay with a pubic ramus fracture complicated by Staphylococcus aureus bacteraemia requiring a 10-day course of intravenous meropenem. This antibiotic was selected as the patient had a documented penicillin allergy and empirical antibiotics were required pending blood cultures. No other antibiotics had been administered.
Her past medical history included a subtotal colectomy and end-ileostomy for ulcerative colitis in 1971, transitional cell carcinoma of the bladder, acute pancreatitis in 1991 and pulmonary embolism in 1995.
On admission, she was febrile (37.5C) but normotensive. Systemic examination was otherwise unremarkable. Abdominal examination revealed an old vertical midline scar and ileostomy in the right iliac fossa.
Laboratory investigations revealed a neutrophil leucocytosis (10.3×109/l), renal impairment (urea 26.3 mmol/l, creatinine 128 µmol/l) and an elevated C reactive protein (280 mg/l). Initial stool cultures for salmonella, shigella, Escherichia coli 0157, campylobacter and CD toxins from the ileal effluent were negative. Stool viral assays were not determined.
The sudden onset of profuse diarrhoea and systemic inflammatory response supported an infectious aetiology, although the differential includes a recurrence of inflammatory bowel disease. A viral aetiology was thought to be less likely as was other small bowel pathology such as coeliac disease, malabsorption, short bowel syndrome and drug side effects.
She was managed empirically with intravenous meropenem in case of a recurrence of her staphylococcal sepsis, which had previously responded to meropenem. During the admission she developed vasculitic lesions on the finger tips and a trans-thoracic echocardiogram raised the possibility of vegetations on the mitral valve leaflets. Therefore, antibiotics were rationalised to teicoplanin and gentamicin to cover possible Staphylococcal endocarditis.
The diarrhoea persisted and a per-stoma ileoscopy revealed a normal ileal mucosa. Repeat culture from the ileostomy effluent was subsequently reported to be positive for CD toxins (day 15) and she was started on oral metronidazole. However, in spite of a reduction in volume of effluent and improved consistency, her general condition deteriorated and she died in week 6 of a hospital-acquired pneumonia after several weeks of debility and frailty.
CD colonisation of the colon is common in patients following treatment with broad-spectrum antibiotics. Pseudomembranous colitis is generally confined to the colon with a demarcation at the ileocaecal valve.2 Although CD has been described in jejunal aspirates,3 small bowel involvement occurring in patients with an ileostomy remote from the time of the colectomy for inflammatory bowel disease is rare.4–7
A literature search revealed 15 case reports of CD enteritis with prior abdominal surgery in 11 of 15 cases.1 7 An additional series described 12 cases of CD enteritis; 6 with an ileostomy, of whom only 1 had underlying inflammatory bowel disease.8 This patient developed CD infection within 2 weeks of surgery. There are only six reported cases of ileostomy CD diarrhoea following inflammatory bowel disease, including one patient who developed a postoperative CD infection.4–9 Risk factors for CD enteritis include old age, recent admission to hospital and recent antibiotics. Our case is only the fifth patient described with CD enteritis occurring at least 1 year after the creation of the stoma in inflammatory bowel disease.
Susceptibility to CD infection in the small bowel may be due to morphological changes such as colonic-type metaplasia and partial villous atrophy,10 or an altered bacterial flora that resembles faecal flora following an ileostomy.11 The long time interval between the original operation and CD infection supports this. The trigger in this case was likely to have been the broad-spectrum antibiotics used prior to the onset of diarrhoea leading to hospital-acquired CD infection. The rate of nosocomial CD infection in this institution is 0.13% of admissions. Symptoms include fever, diarrhoea, nausea, vomiting and abdominal pain. Diagnosis is confirmed by toxin assays. The absence of CD toxin at presentation may be due to a false-negative result but highlights the need of repeat testing if the index of suspicion is high. Clinical management is similar to that of CD infection with an intact colon. CD infection is associated with increased mortality, particularly in patients with advanced age and co-morbidities. We suspect that the CD enteritis was a manifestation of her poor overall health and was, therefore, a contributory factor rather than a cause of her death.
The clinical entity of CD infection in an ileostomy is rare and not well-known even among gastroenterologists. This may be a rarity but the high profile of CD within the community warrants all clinicians to consider this even in patients who do not have a colon.
Competing interests None.
Patient consent Obtained.