Roland S. Croner

Background: Histological phenotype and clinical behaviour of malignant tumours are not only dependent on alterations in the epithelial cell compartment, but are affected by their interaction with inflammatory cells and tumour-associated stroma. Studies in animal models have shown influence of tumour-associated macrophages (TAM) on histological grade of differentiation in colon carcinoma. Disruption of transforming growth factor beta (TGF-beta) signalling in tumour cells is related to more aggressive clinical behaviour. Expression data of components of this pathway in tumour-associated stroma is limited.Methods: Tissue micro arrays of 310 colon carcinomas from curatively resected patients in UICC stage II and III were established. In a first step we quantified amount of CD68 positive TAMs and expression of components of TGF-beta signalling (TGF-beta1, TGF-beta receptors type 1 and 2, Smad 3 and 4) in tumour and associated stroma. Further we analyzed correlation to histological and clinical parameters (histological grade of differentiation (low-grade (i.e. grade 1 and 2) vs. high-grade (i.e. grade 3 and 4)), lymph node metastasis, distant metastasis, 5 year cancer related survival) using Chi-square or Fisher's exact test, when appropriate, to compare frequencies, Kaplan-Meier method to calculate 5-year rates of distant metastases and cancer-related survival and log rank test to compare the rates of distant metastases and survival. To identify independent prognostic factors Cox regression analysis including lymph node status and grading was performed.Results: High-grade tumours and those with lymph node metastases showed higher rates of TAMs and lower expression of TGF-beta1. Loss of nuclear Smad4 expression in tumor was associated with presence of lymph node metastasis, but no influence on prognosis could be demonstrated. Decrease of both TGF-beta receptors in tumour-associated stroma was associated with increased lymph node metastasis and shorter survival. Stromal TGF-beta receptor 2 expression was an independent prognostic factor for cancer related survival.Conclusion: Histological phenotype and clinical behaviour of colon cancer is not only influenced by mutational incidents in tumour cells but also affected by interaction of tumour tissue with inflammatory cells like macrophages and associated stroma and TGF-beta signalling is one important part of this crosstalk. Further studies are needed to elucidate the exact mechanisms.

Introduction: Liver microcirculation disturbances are a cause of hepatic failure in sepsis. Increased leukocyte-endothelial interaction, platelet adherence and impaired microperfusion cause hepatocellular damage. The time course and reciprocal influences of ongoing microcirculatory events during endotoxemia have not been clarified. Methods: Male Wistar rats (232 ± 17 g) underwent cecal ligation and puncture (CLP). Intravital microscopy (IVM) was performed 0, 1, 3, 5, 10 and 20 hours after CLP. Mean erythrocyte velocity, leukocyte and platelet rolling in postsinusoidal venules and sticking of leukocytes and platelets in postsinusoidal venules and hepatic sinusoids were determined. Heart rate (HR), mean arterial pressure (MAP) and portal venous blood flow (PBF) were measured. Blood count and investigation of hepatic enzyme release was performed after each IVM time point. Results: Hepatic platelet-endothelial adherence in liver sinusoids and postsinusoidal venules occurred one hour after the induction of endotoxemia. Leukocyte-endothelial interaction started three to five hours after CLP. A decrease of hepatic microperfusion could be observed at three hours in sinusoids and ten hours in postsinusoidal venules after CLP, although PBF was reduced one hour after CLP. HR remained stable and MAP decreased ten hours after CLP. Hepatic enzymes in blood were significantly elevated ten hours after CLP. Conclusion: Hepatic platelet-endothelial interaction is an early event during endotoxemia. Leukocyte adherence occurs later, which underlines the probable involvement of platelets in leukocyte recruitment. Although PBF is reduced immediately after CLP, the later onset of hepatic microperfusion decrease makes the existence of autoregulatory liver mechanisms likely.

Background: To identify early symptoms of familial adenomatous polyposis with a view to improve early diagnosis and treatment. Diagnosis on the basis of genetic testing is usually limited to where there is a known family history, so FAP is more usually diagnosed on clinical grounds. Except for those identified via FAP registers, the majority of patients are symptomatic at the time of diagnosis. Methods: We undertook a retrospective study of 143 FAP patients treated at the Department of Surgery, University of Erlangen between 1971 and 2000. We identified patterns of symptoms, endoscopic findings and extracolonic manifestations in three age groups. Results: FAP was diagnosed clinically on the basis of symptoms in 84% (120/143) of these patients. Most presented with intestinal symptoms such as colonic bleeding (68%) and diarrhea (42%). All but one of the patients between 20 and 40 years old had rectal polyps (98.7%, 75/76), whereas in those over 40 years old the prevalence was 76% (35/46). Non-specific symptoms such as abdominal pain, fatigue and bloating were less frequent and were mainly reported by patients older than 40. Conclusion: The commonest presenting features of FAP are alteration of bowel habit and rectal bleeding, but both are found in many other conditions. Patients with these findings need immediate endoscopy to allow prompt diagnosis and prophylactic surgery.

Microarray technology has been used by a growing number of investigators and several studies have been published that list hundreds of genes differentially expressed by colorectal carcinoma (CRC) and normal mucosa (MC). On the basis of our own and other investigators’ microarray data, our goal was to identify a common denominator gene cluster distinguishing CRC from MC. Thirty GeneChips (HG-U133A, Affymetrix) were hybridized, 20 with RNA of CRC stages I–IV (UICC) and 10 with MC. Expression signals showing at least a 4-fold difference between CRC and MC (p<0.01) were identified as differentially expressed. In addition, in our integrative data analysis approach only those genes whose expression was altered simultaneously in at least 2 of 5 recently published studies were subjected to an unsupervised hierarchical cluster analysis. We detected 168 up- and 283 down-regulated genes in CRC relative to MC. Twenty-three genes were filtered from the five articles reviewed. An unsupervised hierarchical cluster analysis of these 23 genes confirmed the high specificity of these genes to differentiate between CRC and MC in our microarray data. Colorectal cancer and mucosa could be clearly separated by 23 genes selected for being differentially expressed more than once in a recent literature review. These genes represent a common denominator gene cluster that can be used to distinguish colorectal MC from CRC.