Treatment of locally recurrent rectal cancer, results and prognostic factors.

PURPOSE: Assessment of the results and prognostic factors in patients with locally recurrent rectal cancer treated with curative intent. PATIENTS AND METHODS: Forty patients with an isolated pelvic recurrence of rectal cancer were studied retrospectively. The treatment consisted of radiotherapy alone or combined with chemotherapy and/or surgery performed between January 1992 and July 2001. Radiotherapy was given with a 3-4 fields technique (6-15 MV), five times a week. The median radiation dose was 50 Gy (range 25-66.6 Gy). Twenty-five patients underwent salvage surgery. Five patients were treated with concomitant chemotherapy (5-fluoro-uracil/leucovorin) (5FU/LV) during the 1st and 5th week of radiotherapy. RESULTS: Twenty-two of the 40 patients were male. The local recurrence free survival after 3 and 5 years, respectively, was 49 and 39%. Male gender was the only independent factor associated with failure of local control. The 3 and 5-year overall survival of the total group was 36 and 19%, respectively, with a median survival of 26 months. CONCLUSION: In a selection of patients in the treatment of locally recurrent rectal cancer valuable local palliation if not cure, can be reached. A multimodality approach seems to offer the best chances in this threatening situation.

A favourable pathological stage after neoadjuvant radiochemotherapy in patients with initially irresectable rectal cancer correlates with a favourable prognosis.

Initial treatments of locally advanced rectal cancers focus on local control, as local relapse of a rectal cancer is correlated with a high morbidity and mortality. We studied the effect of neoadjuvant radiochemotherapy on advanced rectal cancer patients in relation to downstaging, local relapse and survival. Post-treatment pathological staging, local relapse and survival were analysed in 66 patients from a single institution. 43 patients had irresectable cancer as determined by laparatomy (n=42) or rectal examination (n=1). These 43 patients received 45-56 Gy preoperatively with 5-fluorouracil (5-FU) and leucovorin (350/20 mg/m(2)x5 day (d)) in weeks 1 and 5 during the radiation therapy. 23 patients had primary resectable tumours with a T1-2 stage. Of the initially irresectable tumours 79% became macroscopically resectable, in 74% a R0 resection was performed. In 6 of 34 (18%) surgical specimens, no tumour was found (pT0), 7 patients had small tumour remnants (pT1-2). In these pT0-2 tumours, no local relapses occurred (observation period of median 4.5 years, range 18-87 months). In the 21 patients with pT3-4 tumours 3 local relapses were seen. In the 23 patients with primary resectable T1-2 tumours the relapse rate was 4%. Downstaging of an initially irresectable rectal tumour to pT2 or less results in a local relapse rate and overall survival that correspond with the rates in primary resectable cancer with the same T classification.

Diagnostic, surgical and medical aspect of the midgut carcinoids.

This review covers the incidence, prognosis, diagnosis and treatment of midgut carcinoids with emphasis on the surgical and peri-operative aspects. Midgut carcinoids are rare neuro-endocrine tumours which become manifest once they have metastasized to the liver. Treatment of metastatic disease may include radical resection but is usually palliative. The tumour grows relatively slow. Besides the biochemical effects resulting in the carcinoid syndrome, patients may suffer from mechanical mass effects of the tumour. Medical treatment can alleviate the biochemical effects of the tumour, but has a limited effect on tumour growth. The introduction of octreotide was a milestone in palliation of these symptoms and has led to more aggressive treatment protocols. Treatment aimed at cytoreduction of hepatic metastasis and diminished secretion of bioactive amines may achieve good palliation. Cytoreduction may be performed by means of surgery, hepatic arterial ligation, (chemo)embolization, cryosurgery, radio-frequency ablation, internal radiation or even liver transplantation. The role of these options will be discussed in this review.

Adrenocortical adenocarcinoma in an MSH2 carrier: coincidence or causal relation?

A woman is described who developed an ovarian adenocarcinoma, 3 metachronous colorectal adenocarcinomas, and a primary adrenocortical adenocarcinoma. Genetic investigation of the mismatch repair genes MLH1 and MSH2 showed a germline mutation in MSH2. Colorectal and ovarian carcinoma belong to the tumor spectrum of hereditary nonpolyposis colorectal cancer (HNPCC). Adrenocortical adenocarcinoma, however, has never been described as 1 of the HNPCC-associated tumors. To investigate whether the adrenocortical adenocarcinoma in this patient was caused by the MSH2 germline mutation, determination of microsatellite instability (MSI) and immunohistochemical analysis were performed on 1 of the colorectal tumors and the adrenocortical adenocarcinoma. MSI and general loss of MSH2 protein expression could be seen in the colorectal tumor but not in the adrenocortical adenocarcinoma. Therefore, it is highly unlikely that the adrenocortical adenocarcinoma found in this patient was due to her genetic predisposition for HNPCC. HUM PATHOL 31:1522-1527.

Changes in mortality of acute myocardial infarction as a function of a changing treatment during the last two decades.

Forty years ago, after the establishment of coronary care units, a significant decrease in mortality of acute myocardial infarction was noted. Twenty years ago, the break-through of thrombolysis realized once again a significant decrease in mortality. In this study we compare, in a rather small community hospital, the mortality and safety of thrombolytic therapy in acute myocardial infarction with a more conventional, conservative medical therapy. We examined all cases of acute myocardial infarction between 1978 up to 1998 inclusive, concerning treatment and mortality rate after a six month period. To be included in the study, acute myocardial infarction had to fulfill particular inclusion criteria. A total of 1863 cases of acute myocardial infarction were included. The mortality rate of patients with acute myocardial infarction treated with thrombolytic agents was strikingly lower and statistically very significantly different (p < 0.001) in comparison with the mortality rate of patients treated with heparin or coumarine derivatives. The mortality rate dropped from 10.57% in the coumarine group and from 14.95% in the heparin group to 5.41% in the alteplase group, to 4.95% in the anistreplase group and 4.00% in the streptokinase subgroup. The complications directly connected to the treatment did not seem to be different between the five groups, and they were also not more frequent by using thrombolytic agents. In the last 20 years, better preventive measures (life habits, diet, medication) and trials to better control the risk factors have not influenced greatly the average amount of cholesterol in patients with an acute myocardial infarction. Also the percentage of patients with high blood pressure has hardly decreased over the last 20 years. The mortality associated with acute myocardial infarction has decreased significantly with the use of thrombolytics. In most cases, thrombolytics are administered routinely and safely. In this way, they are the first choice therapy for myocardial infarction in smaller hospitals. To obtain excellent coronary patency, thrombolytic agents with a long half-life and with PAI-1 resistance are required in the future. The current measures and medical therapies seem to be insufficient to control the risk factors for coronary atherosclerosis.

Selective use of preoperative radiotherapy in the treatment of cancer in the lower two thirds of the rectum.

PURPOSE: The authors review the result of the selection of patients with a low rectal cancer for pre-operative radiotherapy. METHODS: The selection was based on the findings of digital examination eventually combined with surgical staging consisting of bimanual palpation during a staging laparotomy or "trial" operation. This selection was used to divide the patients into three groups: one where local radicality could be expected from primary surgery (group 1), one with deeply infiltrating, but mobile tumours requiring 10 x 3 Gy pre-operative radiotherapy (group 2) and one with fixed or borderline resectable tumours requiring protracted pre-operative radiotherapy with 55-59 Gy (group 3). One hundred and one patients were eligible for this study. A resection aiming for pelvic radicality was carried out in 94 patients: primary resection in 38 (group 1), surgery subsequent to 10 x 3 Gy pre-operative radiotherapy in 20 (group 2) and 55-59 Gy in 36 (group 3). RESULTS: The calculated risk of local recurrence at 5 years was 15% (95% C.I. 4-27) for group 1, 8% (95% C.I. 0-20) for group 2 and 30% (95% C.I. 16-44) for group 3. The calculated 5 years survival for the 3 groups was respectively 60%, 49% and 39%. CONCLUSION: The overlap in local recurrence rate between the three groups suggests a substantial downgrading by this approach of selective use of pre-operative radiotherapy in the patients with the most advanced tumour. Notwithstanding recent improvements of imaging techniques there still is a place for the staging laparotomy in the selection of the treatment strategy for advanced rectal cancers.

Pelvic exenteration for advanced primary rectal cancer in male patients.

Between November 1987 and January 1996 pelvic exenteration for primary rectal cancer was carried out in 11 male patients. Two underwent a primary resection with subtotal cystectomy. In the remaining nine patients, the treatment commenced with a staging laparotomy and the fashioning of an end colostomy of the descending colon, followed by preoperative radiotherapy (50-56 Gy in 5 weeks). Total exenteration with uretero-ileo-cutaneostomy was carried out 4-6 weeks later. In two patients, the exenteration was performed despite the detection of hepatic metastases during the second laparotomy. The pathological staging was T4 in six and T3 in five patients. The immediate postoperative course was uneventful in eight patients and their hospital stay averaged 20 days. Serious postoperative complications prolonged the hospital stay of three patients. Three died during the first 6 months. Recurrent disease caused the death of three of the eight remaining patients: one locoregional (7.5 years after surgery), one from pre-existing hepatic metastases (18 months after surgery) and one from the combination of locoregional and distant recurrence (15 months after surgery). Four patients are on follow-up without evidence of disease, 96, 43, 23 and 22 months after surgery. One patient is alive 20 months after exenteration, with two pulmonary metastases having been recently excised. We conclude that, notwithstanding the morbidity rate, total pelvic exenteration is an acceptable option for advanced primary rectal cancer in male patients. Preoperative radiotherapy should be administered for cancers seated in the lower two-thirds of the rectum and perioperative chemotherapy should be considered seriously.

Effects of hemicolectomy on bile acid metabolism in relation to colon carcinogenesis in man.

Bile acids are probably important in colon carcinogenesis. Regional differences in bile acid metabolism within the colon were studied to illuminate the preferential distal occurrence of colon cancer in Western countries. Faeces (24 h) were collected for bile acid measurement from 25 patients with hemicolectomy (nine left and 16 right) and 17 adenoma patients with an intact colon (control subjects). Duodenal bile and cytolytic and alkaline phosphatase activity of faecal water were also studied. The median percentage of deoxycholic acid (DCA) was lower in the hemicolectomy groups [left 48% (range 38-57%), right 45% (2-62%) vs. control subjects 59% (38-70%), P < 0.05]. In duodenal bile, the proportion of DCA in left [4% (1-25%)] was lower than in the patients with right hemicolectomy [19% (0-69%)] and control subjects [24% (7-50%)] P < 0.05. Faecal concentration of protonated DCA was higher in those with right hemicolectomy (0.101 mumol g-1) than in those with left hemicolectomy (0.048 mol g-1), which coincided with a higher cytolytic [right 49% (3-93%), left 2% (1-37%)] and alkaline phosphatase activity [right 6.7 U mL-1 (1.2-40.1 U mL-1), left (2.0 U mL-1 (1-25.7 U mL-1), both P < 0.02]. These findings suggest differences in bile acid metabolism between the proximal and distal colon that may contribute to the disparity in cancer risk.

The anatomical substrate for a difference in surgical approach to rectal cancer in male and female patients.

This review emphasizes' gender related anatomical differences warranting a difference in surgical approach to the problem of rectal cancer in men and women. Differences in the anatomy of the bony pelvis, the pelvic viscera and the lymphatics of the rectum, inspired the authors to extend the margins of the rectal resection in the anterior plane in female patients. Between 1978 and 1992 a rectal resection was carried out for cancers confined to the pelvis in 158 patients. Of these patients 152 were available for review, 95 male and 57 female. In 24 out of 57 female patients extension of the rectal resection towards the genital tract by en bloc excision of posterior vaginal wall and/or uterus was considered necessary to be confident about obtaining tumour free margins. After a median follow-up of 8 years the risk of local recurrence and cancer related death were significantly lower in female patients.

Neo-adjuvant chemotherapy with carboplatin, 4-epiadriamycin and teniposide (CET) in locally advanced cancer of the cardia and the lower oesophagus: a phase II study.

Disappointing results after surgery alone for locally advanced adenocarcinoma of the cardia and the distal oesophagus (stage IIIB/IV) prompted us to combine surgery with neo-adjuvant chemotherapy. With a remission rate of about 70% the combination of etoposide, adriamycin and cisplatin (EAP) has been considered to be the superior treatment, but it has inherent severe toxicity. The authors conducted a phase II study of combined treatment with Carboplatin, 4-Epiadriamycin and Teniposide (CET) to ameliorate this toxicity and to evaluate the effectivity of this regimen in patients with these unresectable tumours. A regimen of 4 cycles of Carboplatin 300mg/m2, 4-Epiadriamycin 80mg/m2 and Teniposide 100mg/m2 was administered intravenously. Treatment cycles were repeated every 3 weeks in patients with initially unresectable adenocarcinoma of the gastro-oesophageal junction proven at laparotomy and/or CT scanning. Nineteen patients were studied and 17 underwent a second laparotomy in an attempt to resect the tumour radically. Eleven patients (65%) of the re-explored group achieved tumour reduction, enabling resection with curative intent. Recurrences, however occurred in 9 patients after a median of 9.5 (4-42) months. One patient died postoperatively as a result of pulmonary embolism. Only one patient remained free of disease after 42+ months. Leucopenia and thrombocytopenia of WHO grade 23 occurred in 58% and 37% of the patients, respectively. This regimen appears to be effective in patients with locally advanced adenocarcinoma of the cardia and the distal oesophagus. Although it can be used in an outpatient setting, the overall toxicity is relatively high and the results are comparable with other less toxic regimens.

5-Fluorouracil/leucovorin/interferon alpha-2a in patients with advanced colorectal cancer. Effects of maintenance therapy on remission duration.

BACKGROUND: A Phase II study of combination treatment with 5-fluorouracil (5-FU), leucovorin (LV), and interferon alpha-2a (IFN) in patients with previously untreated metastatic colorectal cancer was previously reported by the authors. Therapy was on an outpatient basis and consisted of LV 60 mg orally every 8 hours days 1-3, IFN 18 x 10(6) IU subcutaneously days 1-3, and bolus 5-FU 750 mg/m2 intravenously days 2-3. Treatment was repeated every 14 days, until a maximum of 8 courses was administered. A response rate of 54% [95% confidence interval (CI), 34-72%] was observed. However, remission duration after cessation of therapy was short, with a median duration of 5 months (range, 2 to 13+ months). METHODS: Subsequent to the above study, patients on this induction treatment who achieved remission received maintenance therapy: the above described 3-day regimen every 6 weeks until progression, for a maximum of 2 years. RESULTS: Fifty-three patients were enrolled in the induction regimen and 18 out of 29 patients who achieved remission received maintenance treatment. In 50 assessable patients 3 complete recoveries and 26 partial recoveries were observed for a response rate of 58% (CI, 43-72%). Median remission duration of patients receiving maintenance therapy was 9.4 months (CI, 8.4-10.3 months) compared with 4.7 months (CI, 3.2-6.2 months) for patients without maintenance therapy. Median overall survival of all patients was 16.6 months. Toxicity of maintenance therapy was confined to WHO grade 2. CONCLUSIONS: The high response rate of this 5-FU/LV/IFN regimen holds true in a larger group of patients. The group that received maintenance treatment had a remission duration of just over 9 months, the maintenance regimen added little toxicity.

Long-term disease-free survival after isolated peritoneal metastasis from colonic cancer.

We present the case of a 29-year-old female patient with an isolated peritoneal metastatic mass in the Douglas pouch, following ileocecal resection for a Dukes C2 colon cancer of the caecum. As initial treatment, four courses of continuous infusion with epiadriamycin were administered. The effect on the tumour size was marginal. Palliative radiotherapy (33 Gy) resulted in a reduction of the tumour size and subsequently a wide posterior exenteration could be performed. Five years after the initial diagnosis the patient is still in good health with no evidence of tumour recurrence. We sincerely believe that a maximum effort aiming for cure is warranted in selected patients with localized residual or metastatic peritoneal colon cancer, even if the initial prospects seem less favourable.

Thymidylate synthase levels in tumor biopsies from patients with colorectal cancer.

Catalytic activity and FdUMP binding characteristics of thymidylate synthase (TS) were determined in 22 tumor biopsies of patients to be treated (15) or just treated (7) for colorectal cancer with 5-fluorouracil and leucovorin. In 19 samples both parameters could be determined and were found to represent a wide range (15-20 fold). High values of activity correlated with no response, but low values of binding or activity were found in responders as well as in non-responders. It is concluded that the determination of parameters related to TS function is feasible in tumor biopsies and its relevance for clinical management deserves further study.

Effect of extent of anterior resection and sex on disease-free survival and local recurrence in patients with rectal cancer.

Results are presented following 119 curative resections for rectal cancer performed on 47 women and 72 men. Throughout the study it was policy to remove part of the female genital tract when the rectal tumour impinged on the uterus and/or the posterior vaginal wall. After a median follow-up of 7.5 years, local recurrence occurred in three of 46 women and 15 of 71 men (P = 0.03). The survival rates at 5 years were 71 per cent for women (95 per cent confidence interval 56-83 per cent) and 60 per cent for men (95 per cent confidence interval 50-71 per cent) (P < 0.05). The risk for distant metastasis was comparable, suggesting an influence of local recurrence on survival. Reduction of the local recurrence rate coincides with the higher proportion of anterior extensions of surgery in women (19 of 47) than in men (two of 72).

The clinical staging of rectal cancer in patients treated by preoperative radiotherapy.

We describe the results of clinical and (or) surgical staging used by the same surgeon to select a group of 41 patients with advanced rectal cancer for preoperative radiotherapy. Fifteen patients with resectable but advanced rectal cancer were subjected to a short course of radiotherapy (30 Gy in 10 days), immediately followed by resection. High dose preoperative radiotherapy (50-56 Gy in 5 weeks) was administered to 26 patients with borderline resectable or fixed cancer. Adequate resection of the tumour was possible in 21 of these 26 patients 4 weeks after the end of the radiotherapy. A total of 36 patients thus underwent resection after preoperative radiotherapy. No radiotherapy related acute or late morbidity was seen. On 31 December 1992 the results were investigated retrospectively. The median time since entering into the study was 87 months (range 27-141). During the follow-up, pelvic recurrence was detected in six patients; one patient had concomitant distant metastases. The local recurrence free survival at 5 years calculated by the Kaplan-Meier method was 72% (95% CI 58-85). Distant metastases without local recurrence developed in 11 patients. The calculated survival at 5 years was 45% (95% CI 30.5-59).

Anaerobes and their fermentation products in faeces of patients with familial adenomatous polyposis before and after subtotal colectomy and ileorectal anastomosis.

In view of the importance of short-chain fatty acids (SCFA) for the colonic epithelial function and their possible relation with the reported spontaneous regression of rectal polyps after subtotal colectomy with ileorectal anastomosis, we compared the SCFA concentrations in faeces of five familial adenomatous polyposis (FAP) patients before, and in 10 FAP patients after operation to each other and to those of 10 healthy controls. Anaerobe cultural counts and concentrations of organic acids were also investigated in the same faecal samples from FAP patients and controls. The preoperative cultural counts were not significantly different from those of the controls. After colectomy, the Bacteroides (P < 0.002; P < 0.02), and bifidobacteria counts (P < 0.02; P < 0.005) were decreased compared to controls and untreated FAP patients. The ratio of acetic acid to total SCFA significantly increased, which coincided with a proportional decrease of the other SCFA. The frequently observed regression of rectal polyps after colectomy cannot be explained from these results. One would rather expect a possible stimulation of polyp growth and dedifferentiation of colonic epithelial cells.

Phase I-II study of the addition of alpha-2a interferon to 5-fluorouracil/leucovorin. Pharmacokinetic interaction of alpha-2a interferon and leucovorin.

5-Fluorouracil (5-FU) activity has been improved by the use of leucovorin (LV) or alpha-2a interferon (alpha-IF). We investigated the feasibility and activity of addition of alpha-IF to a 5-FU/LV regimen. A phase I study with 26 patients (14 previously untreated, 12 previously treated) with disseminated cancer was conducted. 15 patients were treated with 5-FU/LV and 11 with 5-FU/LV/alpha-IF. The 5-FU/LV regimen consisted of escalating doses of 5-FU bolus intravenously on days 2 and 3, combined with repeated oral LV on days 1, 2 and 3. Treatment was every 2 weeks. In the 5-FU/LV/alpha-IF schedule, 18 x 10(6) U alpha-IF subcutaneously daily was added on days 1, 2 and 3. The phase I study was followed by a phase II study of 5-FU/LV/alpha-IF at the established 5-FU dose in 29 previously untreated patients with disseminated colorectal cancer. The optimal 5-FU dose in both parts of the phase I study was 750 mg/m2/day. Mucositis, diarrhea and leucopenia were dose limiting. Although alpha-IF added its own toxicity (fever, flu-like symptoms, fatigue), it did not decrease the optimal dose of 5-FU. In the phase II study 28 patients were evaluable for response: three complete responses and 12 partial responses were observed (response rate 54%, 95% confidence interval, 34 to 72%). Pharmacokinetics of oral LV was performed in patients treated with and without alpha-IF: significantly higher serum levels of LV and 5-methyltetrahydrofolate were found after alpha-IF addition. Influence of alpha-IF on gastrointestinal absorption or renal clearance could be excluded. In conclusion, this 5-FU/LV/alpha-IF combination seems active in metastatic colorectal cancer. The pharmacokinetic interaction between alpha-IF and LV may play a role in the activity of this regimen. Controlled studies are necessary to establish the value of addition of alpha-IF to 5-FU/LV regimens.

Recent developments in diagnosis and treatment of metastatic carcinoid tumours.

In carcinoid patients a tumour of enterochromaffin cell origin is present, which dependent on the site of origin can result in increased serotonin production. Metastasized carcinoids are often diagnosed by measuring 5-hydroxyindoleacetic acid excretion in the urine. This excretion, however, can be influenced by food intake. On the other hand, serotonin measured in blood platelets is unaffected by food intake and, in addition, is found to be more sensitive. Therapy of metastasized carcinoids is directed at tumour reduction or only reduction of symptoms. Tumour reduction can be achieved surgically or by embolization. Combination chemotherapy has a maximum response percentage of about 33%. Over the last few years, both octreotide and interferon alpha have been used in these patients. They rarely result in a reduction of the tumour size (10-20%). Symptom reduction is achievable in most patients with these agents, however. Recently, increasing knowledge obtained concerning the various serotonin receptors and their antagonists is now being used in the treatment of patients with a metastasized carcinoid. In the future it is expected that the different modalities will be combined increasingly.

Recombinant interferon alpha-2b in patients with metastatic apudomas: effect on tumours and tumour markers.

Malignant carcinoid tumours, islet cell tumours and medullary carcinomas of the thyroid are tumours with similar clinical features. In patients with unresectable or metastatic tumours leukocyte interferon (IFN) and recombinant human (rh) IFN have demonstrated efficacy. Twenty-four evaluable patients with progressive tumours were treated with 2.5 megaunits rh IFN alpha-2b, administered once daily subcutaneously, for a median duration of 7 months (range 0.5-37+). Two carcinoid patients demonstrated a response in tumour size, 80% showed stable disease (SD). Sixty percent of the carcinoid patients with elevated urinary 5-hydroxyindoleacetic (5-HIAA) levels reached a biochemical partial response of the urinary 5-HIAA levels (median duration 13.5 months). In the patients with an islet cell or medullary tumour and an elevated tumour marker, the marker did not further increase. Of the 12 carcinoid patients evaluable for a symptomatic response, ten (83%) experienced a relieve of symptoms. IFN alpha-2b dose reduction or discontinuation due to toxicity was necessary in three and ten patients, respectively. No neutralising IFN alpha-2b antibodies developed despite prolonged treatment. In conclusion, IFN alpha-2b had a beneficial effect in patients with progressive tumours, while long-term IFN alpha-2b treatment did not augment neutralising antibodies. In view of the IFN alpha-2b-related toxicity, administration of IFN alpha-2b on alternating days may be preferable.

Changes in rectal epithelial cell proliferation and intestinal bile acids after subtotal colectomy in familial adenomatous polyposis.

Subtotal colectomy and ileorectal anastomosis in familial adenomatous polyposis patients can induce temporary regression of adenomas in the rectum. The mechanism for this phenomenon is unclear. We evaluated the effect of colectomy on rectal mucosal proliferation, in relation to changes in bile acid metabolism. Four familial adenomatous polyposis patients were studied before and 3-6 months after surgery, and eight others 7-22 years postoperatively. Within 6 months after surgery, the size of the proliferative zone of the colonic crypts was found to be reduced (P less than 0.05). The proliferative activity of total colonic crypts was not affected within this period. More than 7 years postoperatively, increased cell proliferation of total crypts (P less than 0.02), as well as mid (P less than 0.05) and basal (P less than 0.05) crypt compartments, were observed compared to shortly after colectomy. In duodenal bile, deoxycholic acid was absent shortly after operation, whereas several years after operation only a small fraction (2%) was present. Fecal secondary bile acid excretion diminished after colectomy and did not change several years postoperatively. In postoperative stools only, small proportions of ursocholic and ursodeoxycholic acids (about 5% each) were consistently found. As subtotal colectomy causes a temporary decrease in the length of the proliferative zone of rectal crypts toward a normal pattern, this may explain regression of rectal polyps. This temporary effect may be mediated, at least in part, by decreased amounts of cytotoxic secondary bile acids in the rectal lumen.