Gli2 protein expression level is a feasible marker of ligand-dependent hedgehog activation in pancreatic neoplasms.

The hedgehog pathway is known to promote proliferation of pancreatic ductal adenocarcinoma (PDA) and has been shown to restrain tumor progression. To understand how hedgehog causes these effects, we sought to carefully examine protein expression of hedgehog signaling components during different tumor stages. Genetically engineered mice, Pdx1-Cre;LSL-KrasG12D and Pdx1-Cre;LSL-KrasG12D;p53lox/+, were utilized to model distinct phases of tumorigenesis, pancreatic intraepithelial neoplasm (PanIN) and PDA. Human pancreatic specimens of intraductal papillary mucinous neoplasm (IPMN) and PDA were also employed. PanIN and IPMN lesions highly express Sonic Hedgehog, at a level that is slightly higher than that observed in PDA. GLI2 protein is also expressed in both PanIN/IPMN and PDA. Although there was no difference in the nuclear staining, the cytoplasmic GLI2 level in PDA was modest in comparison to that in PanIN/IPMN. Hedgehog interacting protein was strongly expressed in the precursors, whereas the level in PDA was significantly attenuated. There were no differences in expression of Patched1 at early and late stages. Finally, a strong correlation between Sonic Hedgehog and GLI2 staining was found in both human and murine pancreatic tumors. The results indicate that the GLI2 protein level could serve as a feasible marker of ligand-dependent hedgehog activation in pancreatic neoplasms.

The production of coagulation factor VII by adipocytes is enhanced by tumor necrosis factor-α or isoproterenol.

BACKGROUND: A relationship has been reported between blood concentrations of coagulation factor VII (FVII) and obesity. In addition to its role in coagulation, FVII has been shown to inhibit insulin signals in adipocytes. However, the production of FVII by adipocytes remains unclear. OBJECTIVE: We herein investigated the production and secretion of FVII by adipocytes, especially in relation to obesity-related conditions including adipose inflammation and sympathetic nerve activation. METHODS: C57Bl/6J mice were fed a low- or high-fat diet and the expression of FVII messenger RNA (mRNA) was then examined in adipose tissue. 3T3-L1 cells were used as an adipocyte model for in vitro experiments in which these cells were treated with tumor necrosis factor-α (TNF-α) or isoproterenol. The expression and secretion of FVII were assessed by quantitative real-time PCR, Western blotting and enzyme-linked immunosorbent assays. RESULTS: The expression of FVII mRNA in the adipose tissue of mice fed with high-fat diet was significantly higher than that in mice fed with low-fat diet. Expression of the FVII gene and protein was induced during adipogenesis and maintained in mature adipocytes. The expression and secretion of FVII mRNA were increased in the culture medium of 3T3-L1 adipocytes treated with TNF-α, and these effects were blocked when these cells were exposed to inhibitors of mitogen-activated kinases or NF-κB activation. The ß-adrenoceptor agonist isoproterenol stimulated the secretion of FVII from mature adipocytes via the cyclic AMP/protein kinase A pathway. Blockade of secreted FVII with the anti-FVII antibody did not affect the phosphorylation of Akt in the isoproterenol-stimulated adipocytes. CONCLUSION: Obese adipose tissue produced FVII. The production and secretion of FVII by adipocytes was enhanced by TNF-α or isoproterenol via different mechanisms. These results indicate that FVII is an adipokine that plays an important role in the pathogenesis of obesity.

The diagnostic accuracy of high-resolution endoscopy, autofluorescence imaging and narrow-band imaging for differentially diagnosing colon adenoma.

BACKGROUND AND STUDY AIMS: Conventional colonoscopy can result in unnecessary biopsy or endoscopic resection due to its inability to distinguish adenomas from hyperplastic polyps. This study therefore evaluated the efficacy of high-resolution endoscopy (HRE), autofluorescence imaging (AFI), and narrow-band imaging (NBI) in discriminating colon adenoma from hyperplastic polyps. PATIENTS AND METHODS: This was a prospective multicenter study in patients undergoing AFI and NBI examinations. HRE, AFI, and NBI images were classified into two groups based on morphological characteristics, the predominant color intensities, and the visibility of meshed capillary vessels, respectively. Each of the endoscopic photographs were independently evaluated by a single endoscopist. The images were then assessed by three specialists and three residents, the latter having performed < 500 colonoscopies and < 30 NBI and AFI examinations. Diagnostic test statistics were calculated to compare the accuracy in differentiating colon adenoma from hyperplastic polyps for each method. RESULTS: A total of 183 patients were enrolled in the study and 339 adenomas and 85 hyperplastic polyps were identified. AFI and NBI could distinguish adenoma from hyperplastic polyps with an accuracy of 84.9 % and 88.4 %, respectively, whereas HRE exhibited an accuracy of 75.9 %. In the 358 lesions in which the AFI diagnosis was consistent with that of NBI, the accuracy, sensitivity, and specificity were high, at 91.9 %, 92.7 %, and 92.9 %, respectively. During the study comparing specialists and residents, AFI and NBI dramatically improved the diagnostic accuracy of residents from 69.1 % to 86.1 % and 84.7 %, respectively. CONCLUSIONS: Both AFI and NBI are considered to be feasible tools that can discriminate colon adenoma from hyperplastic polyps, and their use may be particularly beneficial for less-experienced endoscopists.

Natural history of branch duct intraductal papillary-mucinous neoplasms of the pancreas without mural nodules: long-term follow-up results.

BACKGROUND AND AIMS: Although branch duct intraductal papillary-mucinous neoplasms (IPMNs) of the pancreas without mural nodules are frequently observed in asymptomatic subjects, the natural history of these lesions has never been studied. The aim of this study was to elucidate the natural history of branch duct IPMNs without mural nodules. METHODS: Eighty-two patients who had no apparent mural nodules on initial examination were selected for follow-up. All subjects underwent examinations by imaging modalities including endoscopic retrograde pancreatography, and were followed-up by regular examinations once or twice a year. Serial changes of the maximum cystic diameter and the appearance of mural nodules were studied during the observation periods ranging from 14 to 148 months (median, 61 months). RESULTS: Nine (11.0%) of 82 patients exhibited obvious progression of cystic dilatation (median, 59 months). Of these nine patients with cystic enlargement, six continued with regular follow-up examinations. Three cases underwent surgical resection, and were pathologically diagnosed as adenoma in two and borderline in one. Four patients (4.9%) showed newly developed mural nodules in dilated branch ducts (median, 105 months). Histological analysis revealed three cases classified as adenoma and one as carcinoma in situ. None of the remaining 69 patients (84.1%) showed any changes in dilated branch ducts (median, 57 months). CONCLUSIONS: Most branch duct IPMNs without mural nodules remained unchanged during long-term follow-up. Although follow-up with careful examination is required to detect newly developed mural nodules in dilated branch ducts, branch duct IPMNs without mural nodules can be followed-up without surgery.

Hepatic iron deprivation prevents spontaneous development of fulminant hepatitis and liver cancer in Long-Evans Cinnamon rats.

Several clinical studies have suggested that excess hepatic iron accumulation is a progressive factor in some liver diseases including chronic viral hepatitis and hemochromatosis. However, it is not known whether iron-induced hepatotoxicity may be directly involved in hepatitis, cirrhosis, and liver cancer. The Long-Evans Cinnamon (LEC) rat, which accumulates excess copper in the liver as in patients with Wilson's disease, is of a mutant strain displaying spontaneous hemolysis, hepatitis, and liver cancer. We found previously that LEC rats harbored an additional abnormality: accumulation of as much iron as copper in the liver. In the present study, we compared the occurrence of hepatitis and liver cancer in LEC rats fed an iron-deficient diet (ID) with those in rats fed a regular diet (RD). The RD group showed rapid increments of hepatic iron concentrations as the result of hemolysis, characteristics of fulminant hepatitis showing apoptosis, and a 53% mortality rate. However, no rats in the ID group died of fulminant hepatitis. Hepatic iron, especially "free" iron concentration and the extent of hepatic fibrosis in the ID group were far less than those of the RD group. At week 65, all rats in the RD group developed liver cancer, whereas none did in the ID group. These results suggest that the accumulation of iron, possibly by virtue of synergistic radical formation with copper, plays an essential role in the development of fulminant hepatitis, hepatic fibrosis, and subsequent hepatocarcinogenesis in LEC rats.

K-ras mutations and cell kinetics in Helicobacter pylori associated gastric intestinal metaplasia: a comparison before and after eradication in patients with chronic gastritis and gastric cancer.

BACKGROUND: Helicobacter pylori related gastric intestinal metaplasia (IM) is considered to be a precancerous lesion. AIMS: To identify the effects of H pylori eradication on K-ras mutations, cell kinetics in IM and histological changes in patients with and without gastric cancers in a one-year prospective study. METHODS: Patients included group A (n = 39), chronic gastritis, and group B (n = 53), intestinal-type early gastric cancer patients who had all undergone endoscopic mucosal resection (n = 25) or surgical resection (n = 28). K-ras codon 12 mutations in IM were examined, followed by DNA sequencing analysis. Proliferating and apoptotic cells were detected with anti-Ki-67 antibody and using the TUNEL method, respectively. RESULTS: The incidence of K-ras mutations in the cancer was only 3.8%. The mutant K-ras in IM was observed more frequently in group A (46.2%) than in group B patients (1.9%) (p<0.005). After eradication, the K-ras mutations significantly declined to 12.8% in group A (p<0.005). The mutation pattern of K-ras codon 12 before eradication was that GGT was mainly changed to AGT (50%) in group A. AGT transformation was not affected by treatment. Apoptosis in IM showed an increase after H pylori eradication in both groups (p<0.05 in group A) although no histological improvement in IM was observed. The monocyte score was significantly higher in group A than in group B (p<0.05); the score improved significantly after eradication. CONCLUSIONS: K-ras mutations in IM do not always play a role in gastric carcinogenesis but cell kinetics, especially apoptosis, in IM may contribute to it. There are early events in K-ras mutations which are influenced by H pylori infection; some mutations may also be selected by eradication. These unstable K-ras mutations in IM may be related to lymphocyte infiltration caused by H pylori infection.

Activation of peroxisome proliferator-activated receptor gamma by troglitazone inhibits cell growth through the increase of p27KiP1 in human. Pancreatic carcinoma cells.

In the present study, we examine whether human pancreatic carcinoma cells express peroxisome proliferator-activated receptor gamma (PPARgamma) and the effect of PPARgamma activation by its selective ligand on cellular growth in pancreatic cancer cells. Immunohistochemical study of resected human pancreata using a polyclonal PPARgamma antibody revealed that PPARgamma protein expression in the nuclei of carcinoma cells was observed in 9 of 10 pancreatic adenocarcinomas. In contrast, normal pancreatic duct epithelial cells in the samples expressed no PPARgamma. Reverse transcription-PCR and Northern blot analysis demonstrated that all four tested human pancreatic cancer cell lines, PK-1, PK-8, PK-9, and MIA Paca-2, expressed PPARgamma mRNA. Luciferase assay in PK-1 cells showed that troglitazone, a selective ligand for PPARgamma, transactivated the transcription of a peroxisome proliferator response element-driven promoter in a dose-dependent fashion. Troglitazone inhibited the growth of all four pancreatic carcinoma cell lines in a dose-dependent manner. Cell cycle analysis by flow cytometry demonstrated that troglitazone induced G1 arrest in PK-1 cells. To examine the role of cyclin-dependent kinase inhibitors in the G1 arrest by troglitazone, we determined p27KiP1, p21CiP1/Waf1, or p18Ink4c protein expression by Western blot analysis in troglitazone-treated PK-1 cells. Troglitazone increased p27Kip1 but not p21Cip1/Waf1 or p18Inkc protein levels in time- and dose-dependent manners. To clarify the functional importance of p27Kip1 in the cell growth inhibition by troglitazone. All these results suggest that PPARgamma could be considered as a possible target molecule for treatment in human pancreatic carcinomas.

Overexpression of hemochromatosis protein, HFE, alters transferrin recycling process in human hepatoma cells.

HFE is a MHC class 1-like protein that is mutated in hereditary hemochromatosis. In order to elucidate the role of HFE protein on cellular iron metabolism, functional studies were carried out in human hepatoma cells (HLF) overexpressing a fusion gene of HFE and green fluorescent protein (GFP). The expression of HFE-GFP was found to be localized on cell membrane and perinuclear compartment by fluorescent microscopy. By co-immunoprecipitation and Western blotting, HFE-GFP protein formed a complex with endogenous transferrin receptor and beta(2)-microglobulin, suggesting that this fusion protein has the function of HFE reported previously. We then examined the (59)Fe uptake and release, and internalization and recycling of (125)I-labeled transferrin in order to elucidate the functional roles of HFE in the cell system. In the transfectants, HFE protein decreased the rate of transferrin receptor-dependent iron ((59)Fe) uptake by the cells, but did not change the rate of iron release, indicating that HFE protein decreased the rate of iron influx. Scatchard analysis of transferrin binding to HFE-transfected cells showed an elevation of the dissociation constant from 1.9 to 4. 3 nM transferrin, indicating that HFE protein decreased the affinity of transferrin receptor for transferrin, while the number of transferrin receptors decreased from 1.5x10(5)/cell to 1. 2x10(5)/cell. In addition, the rate of transferrin recycling, especially return from endosome to surface, was decreased in the HFE-transfected cells by pulse-chase study with (125)I-labeled transferrin. Our results strongly suggest an additional role of HFE on transferrin receptor recycling in addition to the decrease of receptor affinity, resulting in the reduced cellular iron.

Phase II study of irinotecan, leucovorin, 5-fluorouracil and tegafur/uracil for metastatic colorectal cancer.

Irinotecan combined with continuous-infusion 5-fluorouracil (5FU) has been shown to be an effective and tolerable regimen in the treatment of metastatic colorectal cancer (MCRC). Tegafur/uracil (UFT) during 5FU infusion enhances plasma 5FU concentration, mimics continuous 5FU infusion and delivers the drug to target tumor cells. We conducted a phase II trial of four-agent combined therapy for MCRC, giving patients (pts) intravenous irinotecan (30 mg/m2 on day 1), leucovorin (LV, 200 mg/m2 on day 1 and 2), 5FU (300 mg/m2 on day 1 and 2), and UFT (400 mg/day for 14 days). The main endpoint was the objective tumor response rate. Sixteen pts with a good performance status were enrolled from February 2001 to May 2002. The response rate was 19% (3 partial responses), and 13 pts had stable disease. The median time to progression was 5.2 months, and the median survival time was 20.2 months. Considering the low toxicity and reasonable cost, this regimen deserves further investigation.

Cocaine-amphetamine-regulated transcript (CART) acts in the central nervous system to inhibit gastric acid secretion via brain corticotropin-releasing factor system.

Recent study has indicated that cocaine-amphetamine-regulated transcript (CART) is an anorectic chemical in the brain. In the present study, we examined the hypothesis that CART may act in the central nervous system to alter gastric function. Food consumption, gastric acid secretion, and gastric emptying were measured after injection of CART into the cerebrospinal fluid in 24-h fasted Sprague Dawley rats. Central injection of CART inhibited food intake, gastric acid secretion, and gastric emptying. In contrast, ip injection of CART failed to inhibit gastric acid secretion and gastric emptying, suggesting that CART acts in the brain to suppress gastric acid secretion and gastric emptying. In the vagotomized animals, centrally administered CART did inhibit pentagastrin-stimulated gastric acid secretion. The CART-induced acid inhibition was also observed in rats treated with indomethacin, a cyclooxygenase inhibitor. In contrast, pretreatment with central administration of a CRF receptor antagonist, alpha-helical CRF9-41, completely blocked the central CART-induced inhibition of gastric acid secretion. All these results suggest that CART acts in the brain to inhibit gastric function via brain CRF system. The vagal pathway and the prostaglandin system are not involved in the acid inhibition.

Activation of PPARgamma inhibits cell growth and induces apoptosis in human gastric cancer cells.

We investigated the expression of peroxisome proliferator-activated receptor gamma (PPARgamma) and the role of PPARgamma in cell growth in human gastric cancer cells. Reverse transcription-polymerase chain reaction, Northern blot and Western blot analyses showed that a human gastric cancer cell line, MKN45, expressed PPARgamma mRNA and protein. Luciferase assay in MKN45 cells showed that troglitazone, a selective ligand for PPARgamma, transactivated the transcription of a peroxisome proliferator response element-driven promoter. Troglitazone or pioglitazone, selective ligands for PPARgamma, inhibited the growth of MKN45 cells in a dose-dependent manner. Co-incubation of MKN45 cells with troglitazone induced DNA ladder formation. These results suggest that human gastric cancer cells express PPARgamma and that activation of PPARgamma inhibits cell growth and induces apoptosis in gastric cancer cells.

Characterization of prostanoid receptors mediating contraction of the gastric fundus and ileum: studies using mice deficient in prostanoid receptors.

Receptors mediating prostanoid-induced contractions of longitudinal sections of gastric fundus and ileum were characterized by using tissues obtained from mice deficient in each type and subtype of prostanoid receptors. The fundus and ileum from mice deficient in either EP(3) (EP(3)(-/-) mice), EP(1) (EP(1)(-/-) mice) and FP (FP(-/-) mice) all showed decreased contraction to PGE(2) compared to the tissues from wild-type mice, whereas contraction of the fundus slightly increased in EP(4)(-/-) mice. 17-phenyl-PGE(2) also showed decreased contraction of the fundus from EP(3)(-/-), EP(1)(-/-) and FP(-/-) mice. Sulprostone showed decreased contraction of the fundus from EP(3)(-/-) and FP(-/-) mice, and decreased contraction of the ileum to this compound was seen in tissues from EP(3)(-/-), EP(1)(-/-) and FP(-/-) mice. In DP(-/-) mice, sulprostone showed increased contraction. DI-004 and AE-248 caused the small but concentration-dependent contraction of both tissues, and these contractions were abolished in tissues obtained from EP(1)(-/-) and EP(3)(-/-) mice, respectively, but not affected in other mice. Contractions of both fundus and ileum to PGF(2)alpha was absent at lower concentrations (10(-9) to 10(-7) M), and suppressed at higher concentrations (10(-6) to 10(-5) M) of the agonist in the FP(-/-) mice. Suppression of the contractions at the higher PGF(2)alpha concentrations was also seen in the fundus from EP(3)(-/-), EP(1)(-/-) and TP(-/-) mice and in the ileum from EP(3)(-/-) and TP(-/-) mice. Contraction of the fundus to PGD(2) was significantly enhanced in DP(-/-) mice, and contractions of the fundus and ileum to this PG decreased in FP(-/-) and EP(3)(-/-) mice. Contractions of both tissues to I-BOP was absent at 10(-9) to 10(-7) M and much suppressed at higher concentrations in TP(-/-) mice. Slight suppression to this agonist was also observed in the tissues from EP(3)(-/-) mice. PGI(2) induced small relaxation of both tissues from wild-type mice. These relaxation reactions were much potentiated in EP(3)(-/-) mice. On the other hand, significant contraction to PGI(2) was observed in both tissues obtained from IP(-/-) mice. These results show that contractions of the fundus and ileum induced by each prostanoid agonist are mediated by actions of this agonist on multiple types of prostanoid receptors and in some cases modified by its action on relaxant receptors.

Immunoreactive transferrin receptor in sera of pregnant women.

Immunoreactive transferrin receptors in sera of 90 pregnant women of various gestational periods were investigated. The mean concentrations of the serum transferrin receptor in normal males and females were 251 +/- 94 ng/ml and 256 +/- 99 ng/ml, respectively. Serum transferrin receptor levels in pregnant women did not show a significant increase in the early stage of gestation. However, a rapid elevation of the mean concentration of transferrin receptor was observed after 20 weeks of pregnancy. The apparent increase with gestation period suggests that this immunoreactive receptor in the sera of pregnant women is derived from the placental syncytiotrophoblast and reflects the activity of maternofetal iron transport.

Successful treatment of a VIPoma by continuous subcutaneous infusion of somatostatin analogue (SMS 201-995).

A case with WDHA syndrome due to VIPoma is reported. Injection of somatostatin analogue SMS 201-995 was followed by prompt suppression of vasoactive intestinal polypeptide levels (VIP), decreased stool volume, and restoration of the serum potassium concentration to normal. Long-term treatment with SMS 201-995 for up to 20 weeks produced excellent clinical control and a decrease in tumour size. No adverse effects were noted except for localized pain at the site of injection. This was overcome by using a continuous subcutaneous infusion pump which also enabled the effective daily dosage to be reduced and thereby adverse reactions to be avoided.

Expression of syndecan-1 is common in human lung cancers independent of expression of epidermal growth factor receptor.

In order to determine syndecan-1 expression in lung cancer, we examined 115 lung cancer specimens and 17 lung cancer cell lines. Syndecan-1 was immunohistochemically stained with a polyclonal antibody in 115 paraffin-embedded specimens; 84 cases out of 97 non-small cell lung cancer (NSCLC) and eight cases out of 18 small cell lung cancer (SCLC) were positively stained. Simultaneously, epidermal growth-factor receptor (EGFR) was stained; 47 cases out of 97 NSCLC and one case of 18 SCLC were positively stained. No significant correlation was shown between EGFR and syndecan-1 expression (P=0.68). Syndecan-1 mRNA was detectable in 16 of 17 lung cancer cell lines and EGFR mRNA in nine of 17. Eight cell lines had syndecan-1 mRNA as well as EGFR mRNA. PR-39 (1 microM) and 80 pM transforming growth factor-beta(1) (TGF-beta(1)), did not increase expressions of syndecan-1 mRNA and EGFR in five lung cancer cell lines. We concluded that lung cancer had detectable syndecan-1; however, expression of syndecan-1 protein did not correlate with survival time of lung cancer patients.

Recombinant human granulocyte colony-stimulating factor can mobilize sufficient amounts of peripheral blood stem cells in healthy volunteers for allogeneic transplantation.

Peripheral blood stem cells (PBSC) are a good source for bone marrow reconstitution after intensive chemotherapy. The ability to transplant PBSC between allogeneic subjects would be a key step forward in the application of this procedure. For this purpose, we examined the mobilization and recovery of PBSC in healthy volunteers who were given recombinant human granulocyte colony-stimulating factor (G-CSF). Three informed healthy volunteers were injected with G-CSF subcutaneously at a dose of 2.5 micrograms/kg for 6 successive days and at 5.0 micrograms/kg for the following 4 days. The concentration of PBSC was assessed daily by CFU-GM and BFU-E assays, both of which peaked on the sixth to seventh day of G-CSF administration. Comparison between colony assays and hematological parameters revealed that flow cytometry analysis of CD34+ cells by mononuclear cell gating is a rapid and convenient method to quantify mobilized stem cells. The maximum numbers of CFU-GM were 432, 665, and 1386 colonies/ml blood. It is calculated that sufficient amounts of stem cells for transplantation (at least 1.0 x 10(5) CFU-GM/kg) could be obtained by leukapheresis of 5 to 201 blood when the peak was attained. This trial confirmed the feasibility of allogeneic transplantation using PBSC from healthy volunteers who have received G-CSF.

Morphological heterogeneity of myeloperoxidase-positive granules in normal circulating neutrophils: an ultrastructural study by cryosection.

Ultrastructural localization of myeloperoxidase (MPO) in the granules of human circulating neutrophils was examined by cryosection. On careful comparison with the morphological characteristics of the granules by conventional transmission electron microscopic study, large MPO-positive granules were divided into five types by immunocryoultramicrotomy using monoclonal antibody. Double staining of MPO and lactoferrin (or lysozyme) was also performed. Lactoferrin was generally detected in MPO-negative granules. Lysozyme immunostaining was present in MPO-positive and -negative granules. These data may suggest different functions among large MPO-positive granules of human circulating neutrophils.

Morphogenesis of nonpolypoid colorectal adenomas and early carcinomas assessed by cell proliferation and apoptosis.

Nonpolypoid neoplasms, as well as ordinary polypoid tumours, are occasionally found in the colorectum. To clarify whether cell kinetic status affects the macroscopic morphology of colorectal neoplasms, we investigated proliferative indices (PI), apoptotic indices (AI), and the expression of apoptosis-related gene products. We examined 110 colorectal neoplasms comprised of 36 polypoid, 38 flat elevated and 36 depressed tumours. According to WHO's criteria these tumours consisted of 61 adenomas with low grade dysplasia (LGD), 30 adenomas with high grade dysplasia (HGD) and 19 carcinomas with submucosal invasion. Apoptotic cells were detected by TUNEL staining. Proliferating cells and apoptosis-related gene products were assessed by immunohistochemistry for Ki-67, p53, Bcl-2, and Bax antigens. AI were closely associated with macroscopic morphology in adenomas but not in carcinomas. PI were relatively constant among the three macroscopic types in adenomas and carcinomas. Median AI values of polypoid, flat elevated and depressed tumours were 1.8%, 2.1% and 4.6% for adenomas with LGD, 0.8%, 2.4% and 6.2% for adenomas with HGD and 2.9%, 4.0% and 3.6% for carcinomas, respectively. Overall PI were significantly higher in carcinomas than in adenomas with LGD, whereas AI were not different. Although the incidence of expression was significantly higher in carcinomas for p53 and in adenomas for Bcl-2 than the others, the expression of apoptosis-related gene products (p53, Bcl-2 and Bax) was similar among polypoid, flat elevated and depressed tumours. Macroscopic morphology of colorectal adenomas is determined by the apoptosis not by proliferation, and high apoptosis found in depressed adenomas implies their low net growth.