No abstract

Carcinoma of the large bowel is common in developed countries with a North-West European culture.' Diets rich in fat2 or meat3 and low in fibre4 may account for the higher incidence in developed nations when compared with underdeveloped societies.Faecal bile acid concentrations are quantitatively related to the amount of fat eaten5 and were reported to be raised in populations with a high incidence of large bowel carcinoma.6 Subsequent studies on healthy subjects in Scandinavia only partially confirmed the association between faecal bile acid concentration and the risk of cancer.7In previous studies, patients with established carcinoma of the large bowel had raised levels of bile acids in the faeces8 9 suggesting either a causal relationship or, alternatively, the effect of an obstructing lesion. In other studies patients with adenomatous polyps" had raised faecal bile acid concentrations, while those with familial polyposis10 and ulcerative colitis"1 did not.We report here the faecal bile acid concentrations of normal subjects and compare them with patients at increased risk of large bowel carcinoma and with cancer patients before and after excision of the tumour. Methods PATIENTSPatients at increased risk of large bowel cancer were *Address for correspondence and requests for reprints: D G Mudd

We have investigated the role of ornithine decarboxylase activity in rectal mucosa as a marker for colorectal neoplasia. Biopsies of normal rectal mucosa were taken from 18 patients with adenomas greater than 1 cm diameter, 11 with carcinomas and 16 controls. The mean ornithine decarboxylase activity in normal rectal mucosa of adenoma patients, 6.52 nmol CO2 released h-1 (mg cell protein)-1, was significantly lower than that in controls, 16.8, P = 0.006. The difference in rectal ornithine decarboxylase activities between cancer patients, 3.58, and controls was also significant, P = 0.001. These preliminary results suggest that ornithine decarboxylase may be a useful marker in screening for colorectal neoplasia.

This review examines the evidence for and against an association between cholecystectomy and colorectal cancer.

We have assessed the value of the BICAP electrocoagulation probe in reducing the incidence of further bleeding in patients with upper gastrointestinal haemorrhage. One hundred and twenty-nine patients were studied in a prospective randomized controlled trial. There were 85 male and 44 female patients, age range 16-92 years. Forty-five patients had stigmata of recent haemorrhage (visible vessel or spot) and were randomized during endoscopy to 24 control and 21 treatment patients. Seven control patients rebled compared with nine treated patients (Fisher's exact probability test P = 0.44). The transfusion requirements of control patients (3.9 +/- 3.2 units) was not different from that of treated patients (5.7 +/- 3.7 units): Mann Whitney U test, P = 0.06. In the treatment group there was no difference in the operation rate. Also, the number of probe applications between patients with further bleeding and those with no further bleeding was similar (11.6 +/- 5.5 and 11.0 +/- 5.75 respectively). Access with the probe was considered inadequate in 50 per cent of lesions, but this did not correlate with the incidence of rebleeding. Stigmata of recent haemorrhage were significant in predicting rebleeding (P = 0.0019 Fisher's exact probability test). Overall mortality rate of 3.2 per cent was low and was not influenced by electrocoagulation or presence of stigmata of recent haemorrhage. We have not shown that BICAP bipolar electrocoagulation reduces the incidence of rebleeding in upper gastrointestinal haemorrhage.