• 2022-09
  • 2022-08
  • 2022-07
  • 2022-06
  • 2022-05
  • 2022-04
  • 2021-03
  • 2020-08
  • 2020-07
  • 2020-03
  • 2019-11
  • 2019-10
  • 2019-09
  • 2019-08
  • 2019-07
  • br Declaration of interest none


    ★ Declaration of interest: none.
    Corresponding author at: University of Bern, Institute of Pathology, Murtenstrasse 31, Bern, CH-3008, Switzerland.
    E-mail address: [email protected] (I. Zlobec). 1 Authors contributed equally.
    1. Introduction
    Tumor DETA NONOate
    in colorectal cancer is recognized as an important prognostic factor [1]. The presence of tumor buds, defined as single tumor cells or small tumor cell clusters (up to 4 cells), leads to worse overall and disease-free survival and is associated with advanced tumor stage, lymphatic and venous invasion as well as lymph node and distant metastasis.
    Co-expression of CK/VIM in tumor buds and stroma 19
    Tumor budding has often been referred to as a hallmark of epithelial-mesenchymal transition (EMT) but the morphology of tumor buds, as well as immunohistochemistry and RNA profiles do not suggest a full transition to a mesenchymal state [2,3]. A closer look at tumor budding in the literature suggests that only a subset of tumor buds shows nuclear expression of β-catenin and absence of membranous E-cadherin, which are considered “classic” signs of EMT, but this differs between tumor types (eg, more frequently found in colorectal cancers but almost never reported in pancreatic ductal adenocarcinoma) [3-7].
    On the other hand, laser capture microdissection studies in both colorectal cancers and oral squamous cell carcinomas show an up-regulation of EMT-related genes in regions captur-ing tumor buds [8,9]. For example, De Smedt and colleagues report a shift from an epithelial profile in the tumor center to a more mesenchymal one in regions that have captured tumor buds [8]. Our group has also recently shown an association be-tween high-grade tumor budding and the mesenchymal/poor prognostic subgroup of the Consensus Molecular Subtypes (CMS4) characterized by up-regulation of matrix remodeling genes, as well as EMT- and cancer stem cell–related genes [10,27]. Interestingly, the presence of cytokeratin (CK) is often used as a marker for the visualization of tumor budding, which as a marker for epithelium inherently contradicts the EMT hy-pothesis [11]. Some authors refer to “partial” EMT, which sug-gests a hybrid state in which epithelial and mesenchymal conditions can simultaneously be observed [2].
    If partial EMT exists, then tumor cells with both epithelial and mesenchymal characteristics, for example, expressing both CK and vimentin (VIM) should be detectable. We hy-pothesize therefore that a subset of tumor buds is in fact in a state of partial EMT, which can be observed by co-expression of both CK and VIM.
    However, tumor cells are not the only double-positive cell type in colorectal cancers. Fibroblasts proximal to the mesothelium are also immunoreactive to CK and VIM by im-munohistochemistry. Until now, this type of stromal reaction has simply been dismissed as “reactive” but has not been well-characterized. Chen and Borges show that local perito-neal injury associated with tumor invasion is characterized by activation and proliferation of serosal stromal cells that ex-press CK and mesothelial cell-like features in gastrointestinal tumors [12]. The degree of this stromal proliferation is associ-ated with the degree of injury, extent of tumor invasion and type of tumor. Other studies have postulated that during peritoneal metastasis, the mesothelium is replaced by a strong stromal reaction, characterized by activated myofibroblasts, in a process named mesothelial-mesenchymal transition (MMT) [13,14]. These fibroblasts have been considered by several authors as CAF (Cancer-Associated Fibroblasts) which pro-mote tumor progression through communication with cancer cells [15-17]. However, there are multiple theories about the origin of CAFs which are still unclear.
    The aims of this study are (1) to determine whether CK/VIM co-expressing tumor cells exist in colorectal cancer and whether these cells can be attributed to a subgroup of