NSAID-induced polyp regression in familial adenomatous polyposis patients

NSAIDs inhibit prostaglandin synthesis. In 1983, Waddell et al first reported that sulindac, a NSAID (Clinoril), caused regression of rectal adenomatous polyps in several patients with familial adenomatous polyposis, an inherited form of colorectal cancer. Subsequently, NSAIDs have been used as chemopreventive agents in animal carcinogenesis models and adenoma regression had been confirmed in human trials with sulindac. This article summarizes these developments and describes possible mechanisms of colorectal neoplasia chemoprevention.     

Tissue prostaglandin levels in familial adenomatous polyposis patients treated with sulindac

BACKGROUND: Recent work has demonstrated a correlation between frequency of aspirin ingestion and colorectal cancer prevention. Sulindac, another nonsteroidal anti-inflammatory drug (NSAID), has been shown to cause polyp regression and a fall in cell proliferation in patients with familial adenomatous polyposis, who are destined to develop colorectal cancer unless the colon is removed. However, the mode of action of NSAIDs in colorectal carcinogenesis prevention remains to be determined, although a prostaglandin-mediated mechanism seems likely. METHODS: Rectal or duodenal biopsies from 20 patients with familial adenomatous polyposis, who had been randomized to sulindac or placebo, were analyzed for prostaglandin (PG) E2 and F2 alpha levels before and after treatment. RESULTS: A significant fall in prostaglandin E2 and F2 alpha levels was seen in patients who were on sulindac; this correlated with a visual improvement in number and size of polyps in the same patients (P = 0.0096; PGE2, P = 0.036; PGF2 alpha, Spearman's rank correlation). CONCLUSIONS: Nonsteroidal anti-inflammatory drugs may prevent colorectal cancer by their inhibition of prostaglandin synthesis. Prostaglandins may be implicated in carcinogenesis through an increase in cell proliferation, through immunosuppression, by increasing neovascularization, or via a mutagenic effect.     

Evaluation of various screening and surveillance methods in colorectal carcinoma

Colorectal cancer is a common disease which is almost wholly preventable by early removal of adenomatous polyps. Screening should be offered to all persons without risk factors from the age of 50. Selection of the appropriate screening programme should take into account personal preference, local expertise and insurance coverage. Endoscopic screening and surveillance investigations should be strongly encouraged in all persons wit risk factors such as (1) previous treatment of colorectal adenomatous polyps or cancers, (2) ulcerative colitis, (3) patients with hereditary colorectal cancer syndromes and (4) first degree relatives of patients with colorectal cancer. The following four strategies are available for candidates > 50 years without risk factors: (1) faecal occult blood testing (annually), (2) flexible sigmoidoscopy (every 5 years), (3) a combination of both (1 + 2) strategies and (4) coloscopy (every 10 years). Coloscopy should be performed after a positive test result in strategy programs 1-3. Results from prospective randomized trials are available only for faecal occult blood testing, showing an approximately 15% reduction of mortality in the screening group. The potential for reduction of colorectal cancer mortality has been estimated at 30-70% and 60-90% for flexible sigmoidoscopy and coloscopy respectively. However, no results from prospective randomized trials are presently available. Cost-effectiveness analysis has not shown relevant differences between the four different screening strategies.     

Colorectal cancer detection and screening

Colon cancer is a leading cause of death in the United States and is estimated to cause 56,500 deaths during 1998. Most cancers evolve from adenomatous polyps. Screening asymptomatic average-risk individuals is recommended to reduce colorectal cancer mortality by detection and removal of adenomatous polyps.     

Detection of decay-accelerating factor in stool specimens of patients with colorectal cancer

BACKGROUND & AIMS: Colorectal cancers have an increased expression of decay-accelerating factor (DAF). The aim of this study was to determine whether stool specimens of patients with colorectal cancer contain increased amounts of DAF. METHODS: DAF was measured using an immunoassay in the stool specimens of 40 persons with colorectal cancer, 18 with colorectal adenomatous polyps, 13 with upper gastrointestinal cancer, and 41 without gastrointestinal disease. RESULTS: Stool DAF concentrations in patients with colorectal cancer (0-9.8 ng/g stool; median, 1.6 ng/g) were significantly higher than those in patients with adenoma (0-6.4 ng/g; median, 0 ng/g) (P < 0.05), patients with upper gastrointestinal cancer (0-3.1 ng/g; median, 0 ng/g) (P < 0.05), and subjects without gastrointestinal disease (0-3.4 ng/g; median, 0 ng/g) (P < 0.01). Resection of colorectal cancers caused a marked decrease in stool DAF concentrations. The stool DAF test was positive in a substantial portion of patients with colorectal cancer whose tumors were small ( < 2 cm), at an early TNM stage, or unassociated with fecal occult blood positivity. The sensitivity of the test for colorectal cancer was 55%, and the specificity was 85%. CONCLUSIONS: The measurement of stood DAF deserves evaluation as a test for detection of colorectal cancer.     

Aspirin suppresses the mutator phenotype associated with hereditary nonpolyposis colorectal cancer by genetic selection

Nonsteroidal anti-inflammatory drugs (NSAIDs) are well-known cancer preventives, which have been largely attributed to their antiproliferative and apoptosis-inducing activities. In this study, we show that microsatellite instability (MSI) in colorectal cancer cells deficient for a subset of the human mismatch repair (MMR) genes (hMLH1, hMSH2, and hMSH6), is markedly reduced during exposure to aspirin or sulindac [or Clinoril, which is chemically related to indomethacin (Indocin)]. This effect was reversible, time and concentration dependent, and appeared independent of proliferation rate and cyclooxygenase function. In contrast, the MSI phenotype of a hPMS2-deficient endometrial cancer cell line was unaffected by aspirin/sulindac. We show that the MSI reduction in the susceptible MMR-deficient cells was confined to nonapoptotic cells, whereas apoptotic cells remained unstable and were eliminated from the growing population. These results suggest that aspirin/sulindac induces a genetic selection for microsatellite stability in a subset of MMR-deficient cells and may provide an effective prophylactic therapy for hereditary nonpolyposis colorectal cancer kindreds where alteration of the hMSH2 and hMLH1 genes are associated with the majority of cancer susceptibility cases.     

Molecular biology of colorectal cancer

Colorectal cancer is a significant cause of morbidity and mortality in Western populations. This cancer develops as a result of the pathologic transformation of normal colonic epithelium to an adenomatous polyp and ultimately an invasive cancer. The multistep progression requires years and possibly decades and is accompanied by a number of recently characterized genetic alterations. Mutations in two classes of genes, tumor-suppressor genes and proto-oncogenes, are thought to impart a proliferative advantage to cells and contribute to development of the malignant phenotype. Inactivating mutations of both copies (alleles) of the adenomatous polyposis coli (APC) gene--a tumor-suppressor gene on chromosome 5q--mark one of the earliest events in colorectal carcinogenesis. Germline mutation of the APC gene and subsequent somatic mutation of the second APC allele cause the inherited familial adenomatous polyposis syndrome. This syndrome is characterized by the presence of hundreds to thousands of colonic adenomatous polyps. If these polyps are left untreated, colorectal cancer develops. Mutation leading to dysregulation of the K-ras protooncogene is also thought to be an early event in colon cancer formation. Conversely, loss of heterozygosity on the long arm of chromosome 18 (18q) occurs later in the sequence of development from adenoma to carcinoma, and this mutation may predict poor prognosis. Loss of the 18q region is thought to contribute to inactivation of the DCC tumor-suppressor gene. More recent evidence suggests that other tumor-suppressor genes--DPC4 and MADR2 of the transforming growth factor beta (TGF-beta) pathway--also may be inactivated by allelic loss on chromosome 18q. In addition, mutation of the tumor-suppressor gene p53 on chromosome 17p appears to be a late phenomenon in colorectal carcinogenesis. This mutation may allow the growing tumor with multiple genetic alterations to evade cell cycle arrest and apoptosis. Neoplastic progression is probably accompanied by additional, undiscovered genetic events, which are indicated by allelic loss on chromosomes 1q, 4p, 6p, 8p, 9q, and 22q in 25% to 50% of colorectal cancers. Recently, a third class of genes, DNA repair genes, has been implicated in tumorigenesis of colorectal cancer. Study findings suggest that DNA mismatch repair deficiency, due to germline mutation of the hMSH2, hMLH1, hPMS1, or hPMS2 genes, contributes to development of hereditary nonpolyposis colorectal cancer. The majority of tumors in patients with this disease and 10% to 15% of sporadic colon cancers display microsatellite instability, also know as the replication error positive (RER+) phenotype. This molecular marker of DNA mismatch repair deficiency may predict improved patient survival. Mismatch repair deficiency is thought to lead to mutation and inactivation of the genes for type II TGF-beta receptor and insulin-like growth-factor II receptor. Individuals from families at high risk for colorectal cancer (hereditary nonpolyposis colorectal cancer or familial adenomatous polyposis) should be offered genetic counseling, predictive molecular testing, and when indicated, endoscopic surveillance at appropriate intervals. Recent studies have examined colorectal carcinogenesis in the light of other genetic processes. Telomerase activity is present in almost all cancers, including colorectal cancer, but rarely in benign lesions such as adenomatous polyps or normal tissues. Furthermore, genetic alterations that allow transformed colorectal epithelial cells to escape cell cycle arrest or apoptosis also have been recognized. In addition, hypomethylation or hypermethylation of DNA sequences may alter gene expression without nucleic acid mutation.     

Inherited colorectal polyposis and cancer risk of the APC I1307K polymorphism

Germ-line and somatic truncating mutations of the APC gene are thought to initiate colorectal tumor formation in familial adenomatous polyposis syndrome and sporadic colorectal carcinogenesis, respectively. Recently, an isoleucine-->lysine polymorphism at codon 1307 (I1307K) of the APC gene has been identified in 6%-7% of the Ashkenazi Jewish population. To assess the risk of this common APC allelic variant in colorectal carcinogenesis, we have analyzed a large cohort of unselected Ashkenazi Jewish subjects with adenomatous polyps and.or colorectal cancer, for the APC I1307K polymorphism. The APC I1307K allele was identified in 48 (10.1%) of 476 patients. Compared with the frequency in two separate population control groups, the APC I1307K allele is associated with an estimated relative risk of 1.5-1.7 for colorectal neoplasia (both P=.01). Furthermore, compared with noncarriers, APC I1307K carriers had increased numbers of adenomas and colorectal cancers per patient (P=.03), as well as a younger age at diagnosis. We conclude that the APC I1307K variant leads to increased adenoma formation and directly contributes to 3%-4% of all Ashkenazi Jewish colorectal cancer. The estimated relative risk for carriers may justify specific clinical screening for the 360,000 Americans expected to harbor this allele, and genetic testing in the setting of long-term-outcome studies may impact significantly on colorectal cancer prevention in this population.     

Cyclooxygenase-2 inhibitors in tumorigenesis (Part II)

The rate-limiting step in arachidonate metabolism is mediated by enzymes known as cyclooxygenases (COXs). These enzymes catalyze the biosynthesis of prostaglandin H2, the precursor of molecules such as prostaglandins, prostacyclin, and thromboxanes. The COX enzyme family consists of the classical COX-1 enzyme, which is constitutively expressed in many tissues, and a second isozyme, i.e., COX-2, which is induced by various stimuli, such as mitogens and cytokines, and is involved in many inflammatory reactions. Because nonsteroidal anti-inflammatory drugs (NSAIDs) inhibit both COX-1 and COX-2, these drugs also cause unwanted side effects, exemplified by gastrointestinal bleeding. Accumulating evidence indicates that NSAIDs can reduce the incidence of colorectal cancers in human and experimental animals and can reduce the number and size of polyps in patients with familial adenomatous polyposis. This Part II (of a two-part review) focuses on the growing clinical and experimental evidence that NSAIDS and COX-2 inhibitors can influence the risk of colon (and possibly of other) cancers.     

Medical residents' colorectal cancer screening may be dependent on ambulatory care education

Colorectal cancer results in significant morbidity and mortality in the United States. Screening is a critical component of cancer prevention. However, research has suggested that physicians may inconsistently adhere to surveillance guidelines. Since residency training can significantly impact upon future practice patterns, assessment of postgraduate colorectal cancer education is important. This retrospective chart review of patients > or = 50 years of age compared screening performed by resident physicians' in different internal medicine residency programs at The George Washington University Medical Center. Resident physicians who received multiple lectures in colorectal cancer surveillance or were required to document performance of screening on a medical record preventive care summary form performed significantly more rectal examinations (P < 0.0004), fecal occult blood testing (P < 0.00001), and flexible sigmoidoscopies (P < 0.00001) when compared to other resident physicians. Postgraduate education should employ multiple education techniques and reinforcement procedures to increase physician compliance with cancer screening.     

Sulindac causes rapid regression of preexisting tumors in Min/+ mice independent of prostaglandin biosynthesis

Several lines of evidence strongly link prostaglandins (PGs) and leukotrienes (LTs) to cancer of the intestine. Several studies have reported a 40-50% reduction in mortality from colorectal cancer in individuals who routinely consume nonsteroidal anti-inflammatory drugs, possibly by inhibiting cyclooxygenase activity. However, the role of eicosanoids in this process is still unclear. The heterozygote Min/+ mouse model, like patients with familial adenomatous polyposis, carries a nonsense mutation in the adenomatous polyposis coli (APC) gene that results in the spontaneous development of intestinal adenomas (100% incidence). This study investigated the association between eicosanoid biosynthesis, intestinal tumor load, and the chemotherapeutic effect of the nonsteroidal anti-inflammatory drug sulindac during early and preexisting phases of tumor growth and development as well as residual effects after drug withdrawal. Administration of sulindac (320 ppm) to Min/+ mice reduced the tumor number by 95% but did not alter the levels of PGE2 and LTB4 in intestinal tissues. Increasing PGE2 and LTB4 levels by 44% with dietary arachidonic acid supplementation had no effect on tumor number or size. When sulindac was added to the arachidonic acid-supplemented diet, tumor number was reduced by 82%, whereas eicosanoid levels remained elevated. In Min/+ mice with established tumors, treatment with sulindac for 4 days reduced tumor number by 75%, and continual administration of sulindac was necessary to maintain a reduced tumor load. In summary, alterations in eicosanoid formation were not correlated with tumor number or size in the Min/+ mouse model; thus, the antitumor effect of sulindac seems to be PG independent.     

Measurement of histidinohydroxylysinonorleucine and hydroxyproline in skin collagen by reversed-phase high-performance liquid chromatography after 9-fluorenylmethyl chloroformate labeling

The management of intra-abdominal desmoid tumors in patients with familial adenomatous polyposis (FAP) is very difficult. Non-steroidal anti-inflammatory drugs (NSAIDs), anti-estrogenic agents, and steroids are most commonly used, because surgical removal of these tumors may result in severe morbidity, with local recurrence being common. We report a patient with FAP and intra-abdominal desmoid tumors that regressed markedly after prednisolone therapy. The patient, a 38-year-old woman, had undergone total colectomy and ileorectal anastomosis with a diagnosis of FAP with colon cancer. Approximately 17 months after the surgery, she noticed an elastic firm lump in the abdominal wall. She also experienced lower abdominal distension. Computed tomography (CT) of the lower abdomen showed an invasive heterogenous low-density mass occupying the intra-abdominal space. She was treated with sulindac, NSAID, at 300 mg/day, the diagnosis being intra-abdominal desmoid tumors. She exhibited an intestinal obstruction about 9 months after the initiation of sulindac therapy. We changed the treatment and began prednisolone (initial dose, 40 mg/day). This treatment was continued for two years; subsequently, the lesions regressed markedly. She is currently well, more than 3 years after the withdrawal of prednisolone.     

Colorectal lesions predisposing to cancer

Precursors of colorectal carcinoma are adenomatous polyps, sporadic or arising in familial adenomatous polyposis and Lynch syndrome and chronic inflammatory lesions related to ulcerative colitis and Crohn's disease. The adenoma-carcinoma sequence is well established and early detection and removal of colorectal adenomas is thought to prevent colorectal cancer in high risk asymptomatic persons, i.e. subjects over 45 years, with personal or familial history of adenomas and colorectal cancers. The precancerous potential of adenomatous polyps varies according to tissue type, with increased risk with the extent of the villous component, high grade of dysplasia, large size greater than 1 cm and multiple adenomas. The development of de novo colorectal cancer from normal mucosa with flat adenomas has been recently emphasized. The risk of colonic cancer in patients with ulcerative colitis and Crohn's disease is controversed.     

A meta-analysis of the effects of nonsteroidal anti-inflammatory drugs on blood pressure

BACKGROUND: Nonsteroidal anti-inflammatory drugs (NSAIDs) have well-known gastrointestinal and renal toxic reactions. Effects of NSAIDs on blood pressure are less appreciated. A meta-analysis was performed to determine the hypertensive effects of NSAIDs and rank them by magnitude of change in mean arterial pressure (MAP). METHODS: A literature search of published English-language studies of NSAIDs and their effects on blood pressure was done. Studies were included if they met the following criteria: (1) the studies were intervention studies; (2) NSAIDs at any dose or aspirin at doses of 1.5 g/d or greater were included; (3) documentation of blood pressure was provided; and (4) the studies were 24 hours in duration. Studies were excluded if 20% or more of their participants dropped out or if the dose of antihypertensive drugs was adjusted while the subjects were taking NSAIDs. The major outcome was change in MAP while patients were receiving NSAIDs. Each NSAID arm was extracted from its trial. Information on possible confounders, including subject age, trial quality, amount of dietary salt intake, and whether study subjects were hypertensive or normotensive, was recorded. We calculated the average change in MAP on each NSAID, adjusting for confounders. RESULTS: Fifty-four studies with 123 NSAID treatment arms met inclusion criteria. The mean age of subjects was 46 years. Of the 1324 participants, 1213 subjects (92%) were hypertensive. The effects of NSAIDs on blood pressure were found solely in hypertensive subjects. Among these, the increase in MAP after adjusting for amount of dietary salt intake was 3.59 mm Hg for indomethacin (57 treatment arms), 374 mm Hg for naproxen (four arms), and 0.49 mm Hg for piroxicam (four arms). The MAP decreased by 2.59 mm Hg for placebo (10 arms), 0.83 mm Hg for ibuprofen (six arms), 1.76 mm Hg for aspirin (four arms), and 0.16 mm Hg for sulindac (23 arms). The effects on MAP by using placebo, sulindac, and aspirin were statistically significantly different from indomethacin. CONCLUSIONS: In short-term use, NSAIDs vary considerably in their effect on blood pressure. Of the drugs studied, indomethacin and naproxen were associated with the largest increases in blood pressure. The average effects of piroxicam, aspirin, ibuprofen, and sulindac were negligible.     

A truncated poisson regression model with applications to occurrence of adenomatous polyps

Adenomatous polyps are considered as the dominant precursor lesion of colorectal cancer. A phase III colorectal cancer prevention trial, conducted by the Arizona Cancer Center, concerns the ability of wheat bran fibre supplement to reduce the recurrence of adenomatous polyps. All participants in the study are to have had colorectal polyps detected and removed during a baseline (qualifying) colonoscopy within three months prior to enrolment. In this paper, our interest focuses on occurrence of adenomatous polyps at the baseline colonoscopy. We use a truncated Poisson model to fit these types of data. We develop a regression model to assess the effects of explanatory factors on the positive counting variable. We fit truncated Poisson parameters by a log-linear regression model and estimate regression parameters by the maximum likelihood procedure. Finally, we apply it to the baseline colonoscopy data from the Wheat Bran Fiber study.     

Colorectal cancer--chemoprevention

Colorectal cancer is the second leading cause of death from cancer in The United States. During the last fifteen years, emphasis has been placed on identification of high risk patients and families and outline of appropriate surveillance regimens for normal and high risk patients for colorectal cancer. Parallel to this effort, abundant clinical data has been accumulated that chemoprevention of colorectal cancer with nonsteroidals and aspirin may be possible. Interruption of prostaglandin metabolism appears to be one of the mechanisms of action but not the only therapeutic arm. Currently, sulindac, aspirin, calcium and selenium supplementation are attractive recommendations to at risk patients awaiting results of clinical trials. Other agents in development add excitement to the concept of colorectal cancer chemoprevention.     

A preventive registry for hereditary nonpolyposis colorectal cancer

Hereditary nonpolyposis colorectal cancer (HNPCC) is a genetic disorder characterized by a strong family history of colorectal and extracolonic cancers, usually at a young age. This article presents a new provincial service for families with HNPCC. The Steve Atanas Stavro Familial Gastrointestinal Cancer Registry at Mount Sinai Hospital is accruing patients that meet a set of criteria establishing a putative diagnosis of HNPCC. The objectives of the Registry are to develop and assess patient pedigrees, to coordinate screening procedures for at-risk persons, to maintain a prospective database of patient information, to provide education and support for families and to contribute to research. To date, surgeons and patients are the most common referral sources, while oncologists and geneticists are the least common. The ultimate goal of the HNPCC service is the secondary prevention of cancer and a corresponding decrease in mortality for HNPCC family members.     

Polyamine-enhanced NMDA receptor activity: effect of ethanol

Gastrointestinal symptoms and lesions are often associated with the clinical use of non-steroidal antiinflammatory drugs (NSAIDs). An open-label, single arm multicenter Italian study evaluated if misoprostol, a prostaglandin E1 analogue with gastroduodenal mucosal protective activity, was effective in the prevention and treatment of NSAID-induced gastroduodenal lesions. Patients affected by rheumatoid arthritis (RA) or osteoarthritis (OA), in treatment with NSAIDs and suffering from gastric symptoms or gastroduodenal lesions related to NSAID use, were admitted to the study. Gastrointestinal and arthritic symptoms were assessed before and after 4 weeks co-administration of an NSAID (the most frequent was diclofenac, used in 35% of the RA and in 22% of the OA patients, followed by piroxicam and tenoxicam respectively) + misoprostol (200 mcg two times daily in 58% of the cases, 200 mcg three times daily in 39%, 200 mcg four times daily in 3%). On admission and after 4 weeks of therapy a gastrointestinal endoscopy was performed to evaluate the condition of the gastroduodenal mucosa. Final results showed that: (i) NSAID-related gastric lesions were more frequent than duodenal lesions; (ii) when patients were given misoprostol and NSAIDs, 96% of them did not develop gastric lesions and 97% did not develop duodenal lesions; (iii) even when NSAID therapy was continued, gastric or duodenal lesions induced by NSAIDs healed or in any case did not worsen in 92% and 91% respectively of the cases; (iv) during the period of coadministration of NSAIDs+misoprostol, NSAID-related UGI symptoms disappeared or improved in 77% of the cases.(ABSTRACT TRUNCATED AT 250 WORDS)     

Chemopreventive efficacies of aspirin and sulindac against lung tumorigenesis in A/J mice

Non-steroidal anti-inflammatory drugs (NSAIDs) are among the most widely prescribed drugs. In this study, we demonstrated the efficacy of aspirin to inhibit lung tumorigenesis in A/J mice. Lung tumors (9.9 tumors/mouse) were induced by the tobacco-specific nitrosamine, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), administered in drinking water between week 0 and week +7. Groups of mice were fed sulindac (123 mg/kg diet), acetylsalicylic acid (ASA; 294 mg/kg), non-buffered Aspirin (294 mg/kg) or buffered Aspirin (294 mg/kg) in AIN-76A diet from week -2 to the end of the bioassay (week +23). These doses are comparable to the maximal doses recommended for humans. ASA and non-buffered Aspirin were the most effective inhibitors and reduced lung multiplicities by 60 and 62%, respectively. Sulindac inhibited lung tumor multiplicity by 52%. Inhibition by buffered Aspirin was not statistically significant. We evaluated the efficacies of NSAIDs to inhibit NNK activation by h1A2 v2 cells expressing human P-450 1A2. Salicylates, at doses of 500 microM and 1 mM, had no effect on NNK activation. Sulindac and its sulfide and sulfone metabolites (1 mM) inhibited NNK metabolism by 90, 92 and 65%, respectively. We observed a 76% inhibition with SKF 525A, a P-450 inhibitor. Taken together, these results indicate that salicylates and sulindac could be equally effective as chemopreventive agents, but they could differ in their mode of action.     

Flat cancers do develop in the polyp-free large intestine

PURPOSE AND BACKGROUND: Qualitative and quantitative analysis of many flat early cancers that have been discovered during the last decade led us to recognize that a flat route of cancer development de novo is as important a route as the polypoid one. We aim to prove through a longitudinal study that these flat early cancers indeed develop in flat mucosa and not in an adenomatous polyp. METHODS: From January 1, 1990, to July 31, 1994, 554 patients underwent at least two colonoscopies. These patients consisted of 364 males, and average age was 59 years. We searched for flat early cancers developing in polyp-free colorectal mucosa on or after a second colonoscopy. Polyp-free mucosa here means an intestinal segment proved to possess no adenomatous polyp during the preceding colonoscopies, irrespective of the presence of polyps elsewhere. RESULTS: Four flat early cancers were found developing in polyp-free colonic mucosa in four patients. Average age of the patients was 67 years. Locations of the cancers were the transverse (3) and descending colons (1). The shapes were all depressed, and average size of the lesions was 11 mm. Two lesions were endoscopically resected, and two by surgery. CONCLUSION: These four depressed cancers developing in polyp-free mucosa show that flat early colorectal cancers do arise de novo and not from an adenomatous polyp having collapsed on itself.