THE PSEUDOMEMBRANOUS COLITIS PUZZLE - PUTTING THE PIECES TOGETHER

The acceptance of diarrhoea as a natural accompaniment of antibiotic treatment can be dangerous, as diarrhoea may signal the onset of a more serious disturbance such as pseudomembranous colitis [I]. This condition was thought to be rare until an 'epidemic' of case reports in the early 1970s pointed to an association with lincomycin and clindamycin. In fact, anyanti­biotic may cause colitis. The exact incidence is uncertain, and probably depends on a variety of factors. A 10-year world­wide prospective survey of 26,000 inpatients recorded no cases of colitis caused by drugs given in hospitals, but there were seven resulting from outpatient treatment - 6 from the same New Zealand hospital. All cases occurred within II months, and 5 were due to lincomycin. Pseudomembranous colitis is difficult to diagnose. Clinically, it varies from mild diarrhoea to fulminant colitis. Symptoms usually start after 4-10 days of treatment but may appear I or 2 weeks after stopping treatment. Colitis is most common in the elderly, who are least able to survive an attack, but it may occur at any age and should not be overlooked in children. 'The possibility of pseudomembranous colitis should be considered in all patients given antibiotics or whose condition deteriorates after surgery.' The key to diagnosis is sigmoidoscopy. The pseudomembrane is charac­teristically a raised film or plaque (grey, white or yellow) and the surrounding mucosa may show oedema, hyperaemia, or contact bleeding; or be normal. A rectal biopsy specimen should also be taken, but a barium enema should not be per­formed. The radiological appearances are not diagnostic and the procedure may be dangerous. Laboratory tests may show many abnormalities - especially hypoalbuminaemia and leucocytosis and a raised white cell <:ount. The mortality rate is high - 27 -44 %, or possibly less if more milder cases were diagnosed and included. Opinions vary on treatment. All anti­biotics should be stopped and more severely ill patients should be given steroids and IV fluid replacement. Until last year the cause was unknown, but the discovery of a toxin in the faeces suggested involvement of a Clostridium species - now thought to be Cl. difficile. The organism must be highly sensitive to vancomycin, which protects hamsters against colitis, and non-absorbable sulphonamides.

'The new tools, then. in our management of pseudomembranous colitis are the screening test for toxin in the stools or in biopsy specimens; careful stool culture for pathogens; and the specific treatment with vancomycin. But we need to know whether other organisms and other toxins are concerned. and which antibiotics may be used to combat them.'

Quick Diagnosis and Prompt Supportive Treatment are Essential to Reduce Mortality A review of 39 suspected cases showed 28 with histologically proved pseudomembranous colitis in I hospital since July 1975 [2]. These patients had all received antibiotics - 6 for infections without surgery and 22 after major surgery. All had diarrhoea, 6 had fever with clinical signs of sepsis and 3 had abdominal pain mimicking anastomotic dehiscence after col­onic resection. ine of 22 patients in the postoperative group died, and in all but 2, pseudomembranous colitis was not diag­nosed until necropsy. Of the 13 postoperative patients who survived, 4 had minimal symptoms, but 9 had a protracted ill­ness and were in hospital with diarrhoea for over a month. 17 patients had white blood counts over 15,000/mmJ and 18 had albumin concentrations below 30g/L. All except I patient had abnormal bacterial flora, with signs of overgrowth of Staphylococclls aureliS in 2 and Clostridium spp. in 5. High titres of faecal toxins were found in all 7 patients investigated for these. Only 4 of 12 patients with histologically proven disease responded to codeine and diphenoxylate hydrochloride within 5 days of starting treatment. Of 4 patients treated intensively with steroids and IV fluids, only I improved in 72

/ hours, and only I of 3 others responded to cholestyramine. One patient treated with vancomycin responded immediately.

,Clostridium difficile Identified as a Cause" o Eight patients with pseudomembranous colitis and 20 with postoperative diarrhoea were studied for presence of Clostridium spp [3]. The faeces of all patients with pseudomembranous colitis had high titres of io.ull, m!utralised at 18 hours by Cl. sordellii antitoxin but not by the other antisera. Cl. difjicile was isolated from all patients with colitis and in 6 with postoperative di~rhoea. Ct. difjici/e toxin was neutralised by Cl. sordellii antitoxin. Seven strains of Cl. difficile were sensitive to metronidazole, sulphonamide, and vancomycin, and resistant to penicillins, cephalosporins, aminoglycosides, lincomycins, tetracycline and erythromycin. 'The demonstration that the cytopathic effect of Cl. difflcile is in­distinguishable from that produced by faeces and is neutralised by Cl. sordelliiantitoxin suggests that Cl. difficile is the causative agent of pseudomembranous colitis.' It may also playa part in postoperative diarrhoea. Little is known of itS distribution, but it was found in the faeces of 40 % of infants in one study. Possibly small numbers of the organism occur in the intestinal tract of healthy adults, and may multiply in certain conditions to cause postoperative diarrhoea or pseudomembranous colitis according to their potential for toxin production. o In another study a cytotoxic substance was detected in stool specimens from all 4 patients with pseudomembranous col­itis and in I of 54 with antibiotic-associated diarrhoea [4]. The clostridia strains from these patients had similar phenotype

i characteristics and were classified as Ct. difficile.

[I] Editorial : British Medical Journal I: 669 ( 18 Mar 1978) [2] Kappas. A . et al.: Ibid I: 675 (18 Mar 1978)

[3] George, R.H. et al. : Ibid I: 695 (18 Mar 1978) .• [4] Bartlett. J .G . et al.: New England Journal of Medicine 298: 531 (9 Mar 1978)

INPHARMA 1st April. 1978 p11