Crypt Cell Production Research Articles

Infection Risks During Longterm Rituximab Therapy Change Over Time

A new method for the analysis of small-intestinal crypt-cell kinetics using routine peroral diagnostic biopsies is described. Untreated patients with childhood and adult coeliac disease and adults with the gluten-sensitive enteropathy of dermatitis herpetiformis were studied, together with groups of adult and childhood controls. In the classical flat avillous mucosae the increase in crypt size was found to be three-dimensional. The number of proliferating cells per crypt was shown to be markedly increased, and an even greater rise in the crypt-cell production rate was demonstrated. A significant increase in the mitotic index was also confirmed in the avillous mucosae. On the basis of these findings it is suggested that the characteristic crypt morphology in glutensensitive enteropathy can be explained as an adjustment to accommodate the expanded mass of proliferating and maturing cells necessary to support the augmented cell production rate. We may speculate that this in turn is a response to a pathologically rapid loss of cells from the mucosal surface.

The Calcium-Sensing Receptor - A Driver of Colon Cell Differentiation

Dietary Ca(2+) reduces colon cell proliferation and carcinogenesis, but it becomes ineffective or even tumor-promoting during carcinogenesis. It appears that Ca(2+) and the colon cell CaSR together brake the massive cell production in normal colon crypts. The rapid proliferation of the transit-amplifying (TA) progeny of the colon stem cells at the bases of the crypts is driven by the "Wnt" signaling mechanism that stimulates proliferogenic genes and prevents apoptogenesis. It appears that TA cell cycling stops and terminal differentiation starts when the cells reach a higher level in the crypt where there is enough external Ca(2+) to stimulate the expression of CaSRs, the signals from which stimulate the expression of E-cadherin. At this point the APC (adenomatous polyposis coli) protein appears and some of it enters the nucleus. There it removes the apoptogenesis shield and stops the beta-cateninTcf-4 complex from driving further TA cell proliferation by releasing beta-catenin from the nucleus, and delivering it to cytoplasmic APCaxinGSK-3beta complexes for ultimate proteasomal destruction. Cytoplasmic beta-catenin is prevented from returning to the nucleus by destruction in APCaxinGSK-3beta complexes or locked by the emerging E-cadherin into adherens junctions which link the cell to proliferatively shut-down functioning cells with APC-dependent cytoskeletons moving up and out of the crypt. A common first step in colon carcinogenesis is the loss of functional APC which results in the retention of proliferogenic nuclear beta-cateninTcf-4. This drives the eventual appearance of mutation accumulating, apoptosis-resistant clones the proliferation of which cannot be inhibited by external Ca(2+) because of CaSR-disabling gene mutations.

Calcium, calcium-sensing receptor and colon cancer

There is much evidence that dietary Ca 2+ loading reduces colon cell proliferation and carcinogenesis in humans and rodents, but during carcinogenesis it becomes ineffective or even tumor-promoting. We are beginning to see how Ca 2+ balances the continuous massive cell production in colon crypts by driving the terminal differentiation and eventually the apoptosis of the cells mainly on the mucosal surface, and how this Ca 2+ control is lost during colon carcinogenesis. The rapid proliferation of the transit-amplifying (TA) progeny of the colon stem cells is driven by the so-called “Wnt” signaling mechanism, which involves the stimulation of proliferogenic genes such as those for c-Myc and cyclin D1 and the silencing of the gene for the cell cycle-stopping p21 Cip1/WAF1 protein by nuclear β-catenin • Tcf-4 complexes. TA cells avoid mitotic damage and premature apoptosis by expressing the protein survivin. It appears that TA cell cycling stops and terminal differentiation starts when the cells reach a higher level in the crypt where there is enough lumenal Ca 2+ to stimulate the expression and activation of CaSRs (Ca 2+-sensing receptors), the signals from which stimulate the expression of E-cadherin. Along with this, the APC (adenomatous polyposis coli) protein appears and some of it enters the nucleus. There it makes the TA cells susceptible to the eventual apoptotic balancing by stopping survivin expression and the β-catenin • Tcf-4 complex from driving further cell cycling by releasing β-catenin from the nucleus, and delivering it to cytoplasmic APC • axin • GSK-3β complexes for ultimate proteasomal destruction. Cytoplasmic β-catenin is then prevented from returning to the nucleus by either being intercepted and destroyed by APC • axin • GSK-3β complexes or locked by the emerging E-cadherin into membrane adherens junctions which tie the cell into the sheet of proliferatively shut-down cells with APC-dependent cytoskeletons moving to the mouth of the crypt and onto the flat mucosal surface. A common first step in sporadic colon carcinogenesis is the loss of functional APC which disorients upwardly directed migration and causes the retention of nuclear β-catenin and proliferogenic β-catenin • Tcf-4 complexes as well as genomic instability. Eventually the balance between cell proliferation and terminal differentiation and death is radically tipped in favour of proliferation by the appearance of apoptosis-resistant, survivin-expressing clones of Ca 2+-insensitve cells which are locked into the proliferative, mutation-prone mode because of CaSR-disabling gene mutations which prevent the stimulation of E-cadherin expression and terminal differentiation.

Intestinal metabolism of weaned piglets fed a typical United States or European diet with or without supplementation of tributyrin and lactitol

The aim of the study was to investigate the effects of supplementation of a microencapsulated blend of tributyrin and lactitol (TL) to a standard European (EU) diet without antibiotic growth promoters on intestinal metabolism and mucosa development of weaned piglets and to compare it with a standard US diet containing animal proteins, zinc oxide, copper sulfate, and carbadox. Ninety piglets weaned at 21 d were divided into 3 dietary groups consisting of 5 replicates each: 1) US diet supplemented with 55 mg/kg of carbadox, and 2.5% each of plasma proteins and spray-dried blood cells in the first phase, 3,055 mg/kg of Zn in the first and second phases, and 180 mg/kg of Cu in the third phase; 2) EU diet based on vegetable proteins and no antibiotics; and 3) the same EU diet supplemented with 3,000 mg/kg of microencapsulated TL. The study was divided into 3 phases: 0 to 7, 8 to 21, and 22 to 35 d. On d 7, 21, and 35, animals were weighed, and feed consumption and efficiency were determined. On d 14 and 35, one pig per pen was killed, and the intestinal contents and mucosa from the proximal, middle, distal jejunum and the ileum were sampled. Intestinal wall sections were fixed for histological analysis, and intestinal content was used for VFA, ammonia, and polyamine analysis. Throughout the study (d 0 to 35), the US diet had greater ADG and ADFI than the EU diet (P < 0.05). The EU diet supplemented with TL tended to have 11% greater ADG (P = 0.17). Feeding the EU diet caused a reduction in proximal and middle jejunum villi length by 10% (P < 0.05) and an increase in crypt size in proximal jejunum (P < 0.05) compared with the US diet, probably due to an increased rate of cell loss and crypt cell production. The TL supplementation resulted in longer villi along the jejunum and less deep crypts in the proximal jejunum (+15.9 and -8.9%, respectively; P < 0.05) than the unsupplemented EU diet. The TL diet increased the concentrations of cadaverine and putrescine in the small intestine (P < 0.05) and seemed to increase cadaverine, histamine, putrescine, and spermine in the large intestine by 1.5- to 10-fold compared with the US or EU diet. In conclusion, although the US diet had a greater effect on growth performance and mucosal trophic status than the EU diets, the supplementation with slowly released TL seemed to be an effective tool to partially overcome the adverse effects of vegetable protein diets.

Nimesulide Alters Cell Recruitment into Mitosis in Murine Intestinal Crypts Without Influencing the Cell Production Rate

The use of nonsteroidal anti-inflammatory drugs (NSAIDs) that exhibit COX-2 selectivity is associated with fewer gastrointestinal side effects than seen with more traditional NSAIDs. To determine whether the early effects on cell kinetics in the intestinal mucosal epithelium described after COX-2 selective inhibition are sustained following continuous therapy with these inhibitors, assessments of morphometry and cryptal cell proliferation in the murine small intestinal mucosa were made at 24 hr after treatment with indomethacin, a dual COX inhibitor (10 mg/kg body weight intraperitoneally), nimesulide, a selective COX-2 inhibitor (15 mg/kg body weight intraperitoneally), or vehicle. Nimesulide-treated intestine was elongated beyond control values, in contrast to the shorter indomethacin-treated intestine, but anomalous villous forms were present in both treated groups. Both treatments induced expansion and contraction of proliferating compartments in the crypts in different regions of the intestine but nimesulide did not alter crypt cell production rates, in contrast to the down-regulation induced by indomethacin. These findings may provide some of the fundamental basis for the gut-sparing properties seen in patients treated with COX-2 selective inhibitors.

Nimesulide Inhibits Crypt Epithelial Cell Proliferation at 6 Hours in the Small Intestine in CD-1 Mice

To determine whether the gut-sparing selectivity of cyclooxygenase-2 inhibitors is related to early crypt kinetic mechanisms, this study compared the primary effects on small intestinal mucosal epithelial cell proliferation and morphometry of a nonselective dual cyclooxygenase inhibitor, indomethacin, with a cyclooxygenase-2 selective inhibitor, nimesulide. Indomethacin downregulated the crypt cell production rate in the proximal small intestine, and nimesulide reduced cell proliferation in the proximal and distal small intestine. Compared to controls, there were smaller proliferating compartments in the crypts in midintestinal segments in both indomethacin- and nimesulide-treated groups, but more dividing cells in the distal intestine in indomethacin-treated group. Crypt cellularity, numbers, and width were unchanged from control values in both treated groups, suggesting a reduction in crypt cell emigration. Despite its selectivity for inhibiting cyclooxygenase-2, nimesulide induces similar but widespread initial effects on intestinal cell kinetics when compared to indomethacin.

A study of the effects of pinealectomy on intestinal cell proliferation in infant newborn rats

Study the proliferation rate of jejunum and large intestine crypt epithelial cells, in rats pinealectomized immediately after birth. Twenty-four male Wistar rats were distributed into two groups: Acute group (n=12) and Chronic group (n=12). Six animals of each group were operated for removal of the pineal gland (pinealectomy-PnX), and other six were controls (sham pinealectomy-C). Animals from acute and chronic group were sacrificed 15 and 90 days after the surgery, respectively. In acute group, pinealectomy of new-born rats has not caused significant alteration in cell proliferation (PnX=58.77+/-1.77 and C=60.88+/-1.10 in the descending colon/PnX=31.56+/-0.45 and C=31.73+/-0.47 in the proximal jejunum) and in crypt cell population (PnX=24.92+/-4.82 and C=23.60+/-2.48 in the descending colon/PnX=39.92+/-3.49 and C=44.32+/-5.56 in the proximal jejunum). However, in chronic group there was an uprising crypt cell production per crypt in the proximal jejunum (PnX=57.54+/-2.19 and C=47.19+/-7.3)and in the descending colon (PnX=37.78+/-2.22 and C=17.92+/-2.28). As the increase of intestinal crypts epithelial cells in chronic group is a carcinogenesis predetermining factor, the understanding of the interaction between pineal gland and this event has great importance.

Effect of defunction on cytokinetics and cancer at colonic suture lines

An intestinal suture line potentiates experimental carcinogenesis in its vicinity, probably due to adaptive hyperplasia. By contrast, a defunctioning colostomy causes distal hypoplasia, and fewer tumours develop. Male Sprague-Dawley rats (n = 160) were used to study adaptation (assessed by a stathmokinetic technique) and carcinogenesis (induced by azoxymethane) at an end-to-end anastomosis that was raised in either functioning or defunctioned left colon. Controls had no procedure and other rats had proximal colostomy alone. Defunction had a profound antitropic effect on the colon, reducing bowel length and weight and crypt cell production rate (CCPR) by 22-56 per cent. Anastomotic CCPR was increased by a factor of 2.6 over controls: mean(s.e.m.) values of 12.71(2.85) versus 4.87(0.41) cells/crypt/h (P less than 0.01), but defunction reduced this by 76 per cent to a mean(s.e.m.) value of 3.00(0.52) cells/crypt/h below that in the intact colon. In the left colon, 39-50 per cent of tumours were sited at the anastomosis. Compared with controls there were 77 per cent fewer tumours in defunctioned colon, but they still favoured the site of anastomosis. Neoplasms at the colostomy site accounted for 74-77 per cent of all right-sided tumours. Anastomosis and defunction have powerful but contrasting effects on colonic adaptation and carcinogenesis; when combined they tend to cancel each other out.

Effect of duodenogastric reflux on gastric mucosal proliferation after gastric surgery.

The effect of duodenogastric reflux on cell proliferation and mucosal mass in the stomach was studied. Male Wistar rats (n = 118) were submitted to Polya partial gastrectomy, partial gastrectomy with Roux-en-Y diversion of bile, total duodenogastric reflux or handling of the stomach alone (sham operation). Following oral administration of the carcinogen N-methyl-N'-nitro-N-nitrosoguanidine for 6 months, animals were killed 6, 9 or 12 months postoperatively and their stomachs were examined for crypt cell production rate and mucosal DNA content. Compared with shams, crypt cell production rate was more than twice as high in the gastric remnant 12 months after Polya partial gastrectomy (P less than 0.001) and median DNA content was 31 per cent greater (P = 0.05). After total duodenogastric reflux, DNA content was 62 per cent greater than in shams (P = 0.02), while Roux-en-Y diversion reduced crypt cell production rate by 65 per cent (P less than 0.001). Only Polya gastrectomy increased the number of rats developing gastric carcinomas (9 versus 2 shams; P less than 0.05). Increased mucosal cell proliferation in rats with duodenogastric reflux may help to explain the development of gastric stump cancer.

Kinetic changes in experimental colonic urinary diversion.

An experimental model of colonic urinary diversion was performed on male Wistar rats to see if faeces, urine or a faeces and urine mixture produces tumours. Cell kinetic techniques, including the crypt cell production rate, and cellular DNA measurement were used to examine the effect of urine and of faeces on the colonic mucosa. Urinary diversion into the colon significantly increased colonic crypt activity. Tumours were found at the anastomosis when urine and faeces were mixed at 4 months, and at 6 months with urine alone.

Adaptation to jejunoileal bypass promotes experimental colorectal carcinogenesis.

The co-carcinogenic potential of 85 per cent jejunoileal bypass (JIB) was tested in male Sprague-Dawley rats (n = 81) given 6 preliminary injections of the selective intestinal carcinogen azoxymethane (total dose 90 mg/kg). Controls had sham JIB. Colorectal adaptation was studied 30 weeks postoperatively in rats given injections of vehicle alone. JIB caused 17-33 per cent increments in colonic length, weight and crypt depth; crypt cell production rate was more than doubled (P less than 0.01). Despite lowering body weight by 27 per cent, JIB increased the median number of colorectal tumours per rat from 3 to 8 (P less than 0.01). The findings suggest that evidence of hyperplasia and dysplasia should be sought in patients receiving subtotal jejunoileal bypass for obesity.

Experimental colonic carcinogenesis after gastric surgery

Peptic ulcer surgery may predispose to the subsequent development of colorectal cancer. This experimental study has investigated the effects of gastric operations on colonic cell proliferation, bile acid excretion and carcinogenesis. Male Sprague-Dawley rats (n = 105) underwent sham operation, Pólya partial gastrectomy or vagotomy and pyloroplasty. The carcinogen azoxymethane was administered weekly for 6 weeks thereafter (total dose 60 mg kg-1). When the animals were killed 24 weeks after operation, colons were examined for mucosal mass, crypt cell production rate (CCPR) and tumour yield; faeces were assayed for contents of neutral steroids and bile acids (both total and individual). Morphometric indices and mucosal DNA content were similar in all three groups. Pólya gastrectomy reduced: (1) CCPR throughout the colon (by 42-65 per cent, P less than 0.002); (2) the number of rats with colorectal tumours (26 per cent versus 63 per cent, P less than 0.05); (3) faecal levels of neutral steroids and bile acids, notably hyodeoxycholic acid (P less than 0.01). Although vagotomy and pyloroplasty increased caecal CCPR, there were no consistent differences in faecal steroids and no alteration in tumour yield after the operation. These results fail to support clinical studies suggesting that gastric surgery predisposes to colonic carcinogenesis. Indeed, Pólya partial gastrectomy exerts a protective effect, probably by inhibiting colonic cell proliferation.

The effect of altered luminal nutrition on cellular proliferation and plasma concentrations of enteroglucagon and gastrin after small bowel resection in the rat.

Luminal nutrition is known to have a trophic effect on small bowel mucosa after intestinal resection. Humoral agents, however, may also contribute to this process. Two of the proposed humoral agents, enteroglucagon and gastrin, were therefore investigated after intestinal resection and transection in the rat, and changes in their concentration in the plasma were related to cellular proliferation. Forty-eight male Wistar rats had either 75 per cent proximal small bowel resection or jejunal transection. The animals were further divided into three groups, each with a different nutritional intake. The first group were allowed food ad libitum. The second group were kept under hypothermic conditions which resulted in hyperphagia, while the last group were nourished intravenously. A further 8 animals had a laparotomy only (sham operation). All animals were killed 12 days after operation, plasma enteroglucagon and gastrin were measured, while determination of the crypt cell production rate (CCPR) was used to denote cellular proliferation. In each group resected rats had significantly higher crypt cell production rates and greater enteroglucagon levels compared with transected animals. However, only in the normally fed group was plasma gastrin increased in resected animals, there being no significant difference in the plasma concentration of this peptide in transected compared with resected rats, in both the intravenously fed and hyperphagic groups. In the models studied enteroglucagon appears to be a more likely candidate for a humoral trophic agent than gastrin in intestinal adaptation.

The role of pancreatico-biliary secretions in intestinal adaptation after resection, and its relationship to plasma enteroglucagon.

Two groups, each containing 16 male Wistar rats, had either 75 per cent small bowel resection or jejunal transection; 8 animals from each group; had previously been subjected to pancreatico-biliary diversion. All animals were killed 12 days after the operation, plasma enteroglucagon levels were measured and crypt cell production rate (CCPR) at different sites of the remaining small intestine was measured using a metaphase arrest technique with vincristine. In each of the resected groups there was a significant increase in the CCPR and enteroglucagon levels compared with the transected groups. Furthermore it was found that the CCPR and enteroglucagon levels were higher in the resected group without the pancreatico-biliary diversion compared with the resected group with the diversion. This study, although it confirms the importance of pancreatico-biliary secretions in intestinal adaptation, could also indicate that a humoral factor may be important in the control of intestinal cell proliferation. Our findings do not exclude the possibility that enteroglucagon could be a candidate for such a role.

Relative effects of ileal resection and bypass on intestinal adaptation and carcinogenesis.

Since ileal resection and ileal bypass are commonly performed in man and might stimulate colonic hyperplasia, their co-carcinogenic potential was explored in male Sprague-Dawley rats (n = 135). One week after 33 per cent distal small-bowel resection, 33 per cent distal small-bowel bypass or distal ileal transection (control), animals started a 6-week course of azoxymethane injections (total dose 90 mg/kg ip). Findings in rats killed at 20 and 25 weeks were similar: bypass produced a higher yield of colorectal tumours (4.0 +/- 0.6 per rat: mean +/- s.e.) than controls (2.4 +/- 0.4; P less than 0.05), but resection caused maximal enhancement (5.2 +/- 0.5: P less than 0.01). In rats killed at 30 weeks, however, tumour yields were almost identical. Overall, resection increased colonic tumour yield by 55 per cent (P less than 0.02) and bypass by 32 per cent. Stathmokinetic measurements of crypt cell production rate (CCPR) at 20 weeks showed similar increases after resection and bypass both in residual functioning small bowel (109-200 per cent: P less than 0.01) and in colorectum (63-100 per cent: P less than 0.05). At 30 weeks these adaptive effects persisted, despite an overall increase in CCPR with age. Loss of functioning ileum enhances experimental colorectal carcinogenesis principally by reducing the latent period for tumour development. Resection has a greater effect than bypass probably by producing earlier hyperplasia, though later adaptive effects are similar.

Hypoplasia of defunctioned rectum.

The adaptive response of the large bowel to surgical defunction in man is essentially unknown, although in the rat there is progressive hypoplasia and a reduced propensity to experimental carcinogenesis. Mucosal biopsies were taken from the upper rectum completely defunctioned by a proximal stoma from 2 months to 5 years earlier in 11 patients and from 14 controls without abdominal operations or disease. Samples were established in organ culture and, after 16 h, crypt cell production rate (CCPR) was determined by a stathmokinetic technique. Crypt morphometry was also undertaken. CCPR in defunctioned large bowel was less than half that of controls: (mean (s.d.)-1.96 (0.68) versus 4.65 (0.54) cells crypt-1 h-1, P less than 0.0001). Likewise, crypt length was 24 per cent lower (0.34 (0.05) versus 0.44 (0.04) mm, P less than 0.0001) and crypt width was 38 per cent lower (0.04 (0.01) versus 0.07 (0.01) mm, P less than 0.0001). Rectal defunction causes profound and persistent hypoplasia in man.

Mucosal morphology, cell proliferation and faecal bacteriology in acute pouchitis.

A study was performed to investigate whether acute reservoir ileitis (pouchitis) is associated with specific changes in mucosal morphology, crypt cell kinetics and faecal bacteriology in the ileal pouch. Forty-six patients were studied (ileal reservoir, 36; end ileostomy, ten) using clinical grading, sigmoidoscopy and biopsy; 24 patients with a reservoir were restudied after therapy for 1 month with metronidazole 400 mg three times daily. An index of villus atrophy and crypt cell production rate (CCPR) were determined in all biopsy material. Faecal bacteriology was assessed in 12 patients with a pouch before and after metronidazole therapy. The mucosa of patients with pouchitis was associated with a lower villus atrophy index (P = 0.052), a higher CCPR (P = 0.03) and a higher grade of acute inflammation than that in those without pouchitis. There was no difference in faecal bacterial counts between patients with and without pouchitis. A low atrophy index correlated with a high CCPR (P < 0.001), worse functional score (P < 0.001) and more severe pouch mucosal acute inflammation (P < 0.001), but not with faecal bacteriology. Following metronidazole therapy there was resolution of acute pouch inflammation, increased villus atrophy index (P = 0.049), decreased CCPR (P = 0.049) but no differences in faecal bacterial counts apart from Bacteroides species. These data show that metronidazole therapy does not specifically alter the growth of common faecal bacteria in patients with pouchitis, apart from Bacteroides species. However, metronidazole causes resolution of the typical changes in pouch mucosal morphology and crypt cell kinetics associated with pouchitis.

Individuals with a strong family history of colorectal cancer demonstrate abnormal rectal mucosal proliferation.

Thirty-seven individuals with an increased lifetime risk of colorectal cancer because of family history and a control group (n = 21) with no such family history underwent colonoscopy. No neoplastic lesion was found in any individual. Rectal biopsies of macroscopically normal mucosa were taken 8 cm from the anal margin and mucosal proliferation assessed by the crypt cell production rate (CCPR). There was no difference in CCPR between the control group (median 10 (95 per cent confidence interval 7-11) cells per crypt per h) and those at intermediate risk (lifetime risk between 1:17 and 1:10, n = 14). However, there was a significant difference in CCPR between the control group and those at higher risk (lifetime risk > 1:10, n = 23) (median CCPR 13 (95 per cent confidence interval 10-17) cells per crypt per h, P = 0.004). A high risk of colorectal cancer derived from family history correlates with an increased colorectal mucosal proliferation rate.

Suture technique affects perianastomotic colonic crypt cell production and tumour formation

Suture line recurrence is an important cause of failure after potentially curative resection for colonic carcinoma. Our aim was to determine whether suture technique affected the incidence of perianastomotic tumours in experimentally induced colonic cancer. Sprague-Dawley rats were randomized into three groups. A 1 cm longitudinal colotomy was repaired with four interrupted 6/0 polypropylene monofilament sutures, using either a transmural technique (n = 18) or a seromuscular technique (n = 18). Control animals (n = 18) had a sham laparotomy. All animals received nine, weekly, subcutaneous injections of azoxymethane (total dose 90 mg/kg) starting 6 weeks after laparotomy. Surviving animals were killed 32 weeks after laparotomy. Five animals from each group were given intraperitoneal bromodeoxyuridine (100 mg/kg) 1 h before being killed. At death, perianastomotic tumours occurred more frequently in animals with transmural sutures than in either controls or those with seromuscular sutures. This difference was associated with a greater mucosal bromodeoxyuridine crypt cell labelling index in the transmural suture group. We conclude that a transmural anastomotic suture technique promotes the development of experimental perianastomotic colonic tumours.

Inhibition of intestinal carcinogenesis by dietary supplementation with calcium.

Dietary supplementation with calcium reduces colonic crypt cell production rates in both normal and hyperplastic mucosa. Calcium can bind intraluminally with bile salts and fatty acids thus reducing their mitogenic effect. The protective role of oral calcium on intestinal carcinogenesis (induced by azoxymethane) was tested in 60 male Sprague-Dawley rats submitted to either 80 per cent mid jejuno-ileal resection (n = 30) or jejunal transection (n = 30). Half the rats in each group received calcium lactate 24 g/l added to their drinking water. Rats were killed 25-27 weeks postoperatively. Enterectomy increased colonic tumour yield by 60-106 per cent (P = 0.002-0.005) and duodenal tumour yield by 70-86 per cent. Calcium abolished this effect at both sites, halving intestinal tumour yields in rats with both transection and resection (P less than 0.05). Doubling the dietary intake of calcium inhibits experimental carcinogenesis.