In situ validation of an intestinal stem cell signature in colorectal cancer.

OBJECTIVE: Wnt/Tcf, Lgr5, Ascl2 and/or Bmi1 signalling is believed to define the mouse intestinal stem cell niche(s) from which adenomas arise. The aim of this study was to determine the relevance of these putative intestinal stem cell markers to human colorectal cancer. DESIGN: 19 putative intestinal stem cell markers, including Ascl2 and Lgr5, were identified from published data and an evaluation of a human colorectal gene expression database. Associations between these genes were assessed by isotopic in situ hybridisation (ISH) in 57 colorectal adenocarcinomas. Multiplex fluorescent ISH and chromogenic non-isotopic ISH were performed to confirm expression patterns. The prognostic significance of Lgr5 was assessed in 891 colorectal adenocarcinomas. RESULTS: Ascl2 and Lgr5 were expressed in 85% and 74% of cancers respectively, and expression was positively correlated (p=0.003). Expression of Bmi1 was observed in 47% of cancers but was very weak in 98% of cases with expression. Both Ascl2 and/or Lgr5 were positively correlated with the majority of genes in the signature but neither was correlated with Cdk6, Gpx2, Olfm4 or Tnfrsf19. Lgr5 did not have prognostic significance. CONCLUSION: These data suggest that 74-85% of colorectal cancers express a Lgr5/Ascl2 associated signature and support the hypothesis that they derive from Lgr5(+)/Ascl2(+) crypt stem cells, not Bmi1(+) stem cells. However, Olfm4 was not found to be a useful marker of Lgr5(+) cells in normal colon or tumours. In this large series, Lgr5 expression is not associated with increased tumour aggressiveness, as might be expected from a cancer stem cell marker.

Liver transplantation following donation after cardiac death: an analysis using matched pairs.

Grafts from donation after cardiac death (DCD) donors are used to increase the number of organs available for liver transplantation. There is concern that warm ischemia may impair graft function. We compared our DCD recipients with a case-matched group of donation after brain death (DBD) recipients. Between January 2002 and April 2008, 39 DCD grafts were transplanted. These were matched with 39 DBD recipients on the basis of identified variables that had a significant impact on mortality. These were used to individually match DCD and DBD patients with similar predictive mortality. We compared patient/graft survival, primary non-function (PNF), and rates of complications. Of all liver transplants, 6.1% were DCD grafts. PNF occurred twice in the DCD group. The incidence of nonanastomotic biliary strictures (NABS; 20.5% versus 0%, P = 0.005) and hepatic artery stenosis (HAS; 12.8% versus 0%, P = 0.027) in the DCD group was higher. One-year (79.5% versus 97.4%, P = 0.029) and 3-year (63.6% versus 97.4%, P = 0.001) graft survival was lower in the DCD group. Three-year patient survival was also lower (68.2% versus 100%, P < 0.0001). Our study is the first to use case-matched patients and compare groups with similar predictive mortality. There was a higher incidence of NABS and HAS in the DCD group. NABS were likely a result of warm ischemia. HAS may have been due to ischemia or arterial injury during retrieval. The DCD group had significantly poorer outcomes, but DCD grafts remain a valuable resource. With careful donor/recipient selection, minimization of ischemia, and good postoperative care, acceptable results can be achieved.