Adaptation and carcinogenesis in defunctioned rat colon: divergent effects of faeces and bile acids.

Because the composition of faeces modulates colorectal carcinogenesis, promotional effects of the secondary bile salt sodium deoxycholate (SDC) were compared with those of dilute homogenised faeces (12.5% w/v) or saline alone in rat colon isolated from the faecal stream as a Thiry-Vella fistula (TVF). Each fluid was used to irrigate a group of TVFs 3 times per week for 12 weeks. Other rats had TVF without irrigation or colonic transection and reanastomosis (sham TVF). Operations followed a 6-week course of azoxymethane injections. At sacrifice 15 weeks postoperatively crypt depth and tumour yield were reduced to the same extent in both the non-irrigated TVFs and the SDC-irrigated TVFs, when compared to shams. Irrigation with faeces and saline completely restored crypt depth and partly restored tumour yields to the levels in shams. Tumours were smaller in the SDC group than in the other 4 groups. While tumours developed mainly in the left colon of shams, there was significantly more even distribution in the TVFs. Exclusion of the colon from the faecal stream leads to mucosal hypoplasia and impaired carcinogenesis. Irrigation with faeces or saline partly reverses these changes. Deoxycholate has no such effect and clearly is not co-carcinogenic in this model.