Pancreatic ductal adenocarcinoma (PDAC) is the most lethal of all the common malignancies and markers for early detection or targets for treatment of this disease are urgently required. The disease is characterised by a strong stromal response, with cancer cells usually representing a relatively small proportion of the cells in the tumor mass. We therefore performed laser capture microdissection (LCM) to enrich for both normal and malignant pancreatic ductal epithelial cells. Proteins extracted from these cells were then separated by two-dimensional gel electrophoresis (2-DE). The limited amounts of protein in the LCM procured samples necessitated the detection of 2-DE resolved proteins by silver staining. Consequently, loading equivalent amounts of protein onto gels was essential. However, we found that conventional means of measuring total protein in the samples were not sufficiently accurate. We therefore adopted a strategy in which the samples were first separated by one-dimensional sodium dodecyl sulphate-polyacrylamide gel electrophoresis, stained with silver stain and subjected to densitometry. Evaluation of the staining intensity was then used to normalise the samples. We found that the protein profiles from undissected normal pancreas and LCM-acquired non-malignant ductal epithelial cells from the same tissue block were different, underpinning the value of LCM in our analysis. The comparisons of protein profiles from nonmalignant and malignant ductal epithelial cells revealed nine protein spots that were consistently differentially regulated. Five of these proteins showed increased expression in tumor cells while four showed diminished expression in these cells. One of the proteins displaying enhanced expression in tumor cells was identified as the calcium-binding protein, S100A6. To determine the incidence of S100A6 overexpression in pancreatic cancer, we carried out immunohistochemical analysis on sections from a pancreas cancer tissue array containing 174 duplicate normal and malignant pancreatic tissue samples, from 46 pancreas cancer patients. Normal pancreatic ductal epithelia were either devoid of detectable S100A6 or showed weak expression only. Moderately or poorly differentiated tumors, by contrast, showed a higher incidence and a higher level of S100A6 expression. These observations indicate that the combination of LCM with 2-DE provides an effective strategy to discover proteins that are differentially expressed in PDAC.

The cross-talk between tumour cells and the surrounding supporting host cells (stroma) is a key regulator of cancer growth and progression. By undertaking 2-DE analysis of laser capture microdissected malignant and stromal components of pancreatic tumours and benign ductal elements, we have identified high levels of S100A8 and S100A9 in tumour-associated stroma but not in benign or malignant epithelia. Immunohistochemical analysis (n = 71 patients) revealed strong expression of both proteins in stromal myeloid cells, subsequently identified as CD14(+)/CD68(- )monocytes/macrophages. Co-immunofluorescence revealed that S100A8 was expressed in a subset of S100A9-positive cells. Correlation of the expression of S100A8 and S100A9 to patient parameters revealed that the microenvironments of tumours which lacked expression of the tumour suppressor protein, Smad4, had significantly reduced numbers of S100A8-immunoreactive (p = 0.023) but not S100A9-immunoreactive (p = 0.21) cells. The ratio of S100A8- to S100A9-positive cells within individual tumours was significantly lower in Smad4-negative tumours than in Smad4-positive tumours (p<0.003). Pancreatitic specimens also contained S100A8- and S100A9-expressing cells, although this was not observed in regions displaying extensive fibrosis. In conclusion, our study provides an extensive analysis of S100A8 and S100A9 in pancreatic disease and highlights a potentially important relationship between pancreatic cancer cells and their surrounding microenvironment.

We identified pre-operative CEA as an independent predictor of OS and DFS on an individual level. CEA offers additional prognostic value to TNM staging and should be requested routinely as part of the pre-operative work-up.

Purpose. Timely administration of adjuvant chemotherapy following colorectal resection is associated with improved outcome. We aim to assess the factors which are associated with delay to adjuvant chemotherapy in patients who underwent colorectal resection as part of an enhanced recovery protocol. Method. A univariate and multivariate analysis of patient data collected as part of a prospectively maintained database of colorectal cancer patients between 2007 and 2012. Results. 166 patients underwent colorectal resection followed by adjuvant chemotherapy. Median postoperative hospital stay was 6 days, and time to commencement of adjuvant chemotherapy was 50 days. Longer inpatient stay correlated with increased time to adjuvant chemotherapy (P = 0.05). Factors found to be independently associated with duration of hospital stay and time to commencement of adjuvant chemotherapy included stoma formation (P = 0.032), anastaomotic leak (P = 0.027), and preoperative albumin (P = 0.027). The use of laparoscopic surgery was associated with shorter time to adjuvant chemotherapy but did not reach significance (P = 0.143). Conclusion. A number of independent variables associated with delay to adjuvant therapy previously not described have been identified. Further work may be required to elucidate the effect that these variables have on long-term outcome.