Article Text

Download PDFPDF

The Fas–FasL system and colorectal tumours
  1. C A Rubio1,
  2. B Jacobsson1
  1. 1Gastrointestinal and Liver Research Laboratory, Department of Pathology, Karolinska Hospital, 171 76 Stockholm, Sweden; Carlos.Rubio{at}
    1. M W Bennett2,
    2. J O'Connell2,
    3. A Houston2,
    4. J Kelly2,
    5. J K Collins2,
    6. F Shanahan2,
    7. G C O'Sullivan3
    1. 2Department of Medicine, National University of Ireland, Cork, Ireland; j.oconnell{at}
    2. 3Department of Surgery, National University of Ireland

      Statistics from

      Request Permissions

      If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

      We read with interest the article of Bennett et al,1 showing that the coexpression of Fas–Fas ligand (FasL) did not elicit increased apoptosis in colonic tumour cells. They suggested that colonocytes acquire resistance to Fas mediated apoptosis early in the transformation process. The Cork Group has greatly contributed to unravelling the riddle of the mechanism(s) behind the (CD95/APO1) receptor (Fas)–FasL system, which enables apoptosis to occur in neoplastic tissues.

      Using a rat neu transgenic (rNeu-TG) mouse model, Céfai et al demonstrated that the FasL mediated escape from immune rejection of breast tumours correlated with the apoptosis of infiltrating T cells.2

      During studies of similar colorectal lesions as those reported by Bennett et al, we investigated the occurrence of infiltrating lymphocytes.3 In that study, intraepithelial lymphocytes (IELs) were recorded in 70% of 102 hyperplastic polyps and tubular adenomas, in > 80% of 75 villous and serrated adenomas, in 14 of 28 incipient adenocarcinomas, and in only nine of 50 advanced carcinomas. IELs were CD3 positive but major histocompatibility complex class II (MHCII) and TIA1 negative in the normal colorectal mucosa and in hyperplastic polyps. In contrast, MHCII and TIA1 were upregulated in IELs and Fas was downregulated in dysplastic cells from tubular, serrated, and villous adenomas, and in incipient and advanced carcinomas.3 Proliferation and transmission electron microscopy studies provided no indication that the adenomatous cells were undergoing apoptosis. On the other hand, the accumulation of apoptotic granules between the base of the dysplastic cells and the basement membrane strongly suggested that the apoptotic bodies belonged to wandering cells trying to enter the dysplastic epithelium from the subjacent stroma.4 Whereas the neoplastic associated lymphocytes are often intraepithelial in colorectal adenomas, they are often peritumoral in advanced carcinomas.5 Patients with advanced rectal carcinomas who have peritumoral lymphocytes have a good five year tumour free interval, but those lacking peritumoral lymphocytes do not.5

      Against that background, some questions seem pertinent, namely: What is the role played by the Fas–FasL system in advanced colorectal carcinomas lacking IELs? In contrast, what is the role played by peritumoral lymphocytes on the Fas–FasL system? These questions remain unanswered for patients without treatment because they will eventually succumb to the carcinoma, independent of the presence or the absence of intratumoral or peritumoral lymphocytes. However, in treated patients, thc peritumoral lymphocytes seem to master the “metastatic” tumour cells because after tumour removal by surgery those patients have a better tumour free five year interval than operated patients carrying tumours without peritumoral lymphocytes.5 It has been suggested4 that the “struggle” between IELs and adenoma cells with downregulated Fas molecules prevents rapidly proliferating adenomas from becoming clinically apparent cancer; a process that usually takes between 10 and 20 years. These notions are substantiated by the finding that a colonotropic carcinogen inducing slow growing colonic adenomas in rats often evoked IELs and pronounced apoptosis, whereas carcinogens able to induce fast growing adenomas seldom elicited IELs or apoptosis.6

      The importance of Fas–FasL in surgically removed advanced colorectal carcinomas, either having or lacking peritumoral lymphocytes, remains elusive. However, that knowledge could be seminal for the design of immunotherapeutic strategies against advanced colorectal malignancies.


      Authors' reply

      We appreciate the comments of Rubio and Jacobsson regarding our recent paper in the Journal of Clinical Pathology,1 which demonstrated early upregulation of Fas ligand (FasL) in colonic tumorigenesis. It is interesting that these authors previously found a progressive decrease in intraepithelial lymphocytes (IELs) during the progression from hyperplastic polyps through to adenomas, then incipient adenocarcinomas, and finally advanced adenocarcinomas.2 The potential role that FasL counterattack plays in this decrease in IELs has not yet been established, and may prove difficult to do. However, it is interesting that apoptotic bodies have been observed at the margins of adenomas, and these could represent lymphocytes undergoing apoptosis while attempting to infiltrate the dysplastic lesions.3 Although we can only speculate that FasL counterattack plays a role in immune escape in adenomas, there is substantial evidence that FasL causes apoptosis and depletion of tumour infiltrating lymphocytes (TILs) in colonic adenocarcinomas. In a study of 41 colorectal adenocarcinomas, Okada and colleagues recently demonstrated a significant association between FasL expression in the tumours and apoptosis of TILs.4 Furthermore, a high rate of apoptosis of TILs was associated with metastases and significantly poorer prognosis4; this corroborates other evidence that low numbers of TILs are generally associated with a worse prognosis. We have found that even within individual colonic adenocarcinomas, regional variation in the expression of FasL correlates with apoptotic depletion of TILs5; FasL positive tumour nests had fewer TILs and increased apoptosis of TILs relative to matched FasL negative tumour nests within the same tumours. Indeed, we have found that the presence of a vigorous inflammatory response within colorectal cancers is associated with clearance of micrometastases and improved survival.6 A good antitumour immune response appears to be vital for the containment of primary tumour growth, and also to prevent metastasis. As an important mediator of immune downregulation, expressed early in colonic tumorigenesis, we agree with Rubio and Jacobsson that FasL could represent a target for future immunotherapeutic approaches.