| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
EDITORIAL |
From the University of Wisconsin Comprehensive Cancer Center, Madison, Wisconsin.
Correspondence: Address correspondence to: John E. Niederhuber, MD, University of Wisconsin Comprehensive Cancer Center, K4/614 Clinical Cancer Center, 600 Highland Avenue, Madison, WI 53792; Fax: 608-263-8613; E-mail: niederhu{at}biostat.wisc.edu
In the May issue of Annals of Surgical Oncology, two articles addressed the importance of obtaining more prognostic information concerning resected colorectal carcinomas. Dr. Yasuda and colleagues1 have studied the significance of lymph node micrometastasis and upstaging the patient’s tumor using immunohistochemistry with a mouse monoclonal anti-human cytokeratin antibody (CAM 5.2), which specifically recognizes cytokeratins 8 and 18. Of the conventional histopathologic staging parameters, the presence of micrometastases was a significant predictor of failure in patients with Dukes’ B colorectal cancer, when four or more nodes were involved or when micrometastases were found in N2 or higher level nodes.
As the authors note, immunohistochemistry to search for micrometastases has several significant variables. These include the careful search for all nodes within the specimen, assignment of nodes to the correct level by the surgeon and pathologist, and even more importantly, the thickness of sections and number of sections analyzed. As noted, it has been reported that a total thickness of 27µm is maximal for detecting micrometastasis. An important question is whether immunohistochemistry for upstaging of colorectal cancer has advantages over, or is equal to, reverse transcription-polymerase chain reaction (RT-PCR) analysis. The cost benefit of each procedure will also be important when selecting the method to be used. Although not addressed in this article, if sentinel lymph node techniques would prove to be applicable to colon cancer, this would permit an extensive analysis of one or two lymph nodes. This more limited analysis would be cost effective.
The second manuscript, from Professor Guillou’s group2 in Leeds, England, reports on an example of a normal tissue protein that is produced during embryonic and fetal development but is undetectable in normal adult tissues. This gene and its protein, termed Survivin, is selectively expressed in tumor cells. This specific protein is an ambiguous inhibitor of apoptosis, and the analysis by the Leeds group and other recent reports make a strong case that Survivin should be included as a molecular marker of prognosis in colorectal cancer.
Increasing the sensitivity of conventional histopathology is important to accurate tumor staging and to determine the risk of developing recurrent disease. The future, however, clearly lies with the molecular characterization of a patient’s tumor in addition to the more accurate and sensitive staging of the nodes at risk. Some of the genes involved in tumor development and progression, and their proteins, are detectable only within the tumor cell. Others are produced and secreted by the tumor and, therefore, are detectable in serum. Some are produced by the host’s normal tissue in direct response to tumor growth. As a result, the complete picture of the process in each patient may be quite specific and is potentially complex. Nevertheless, it seems that such an analysis of a patient’s tumor will ultimately provide a unique "fingerprint" for the patient’s oncology team and provide a guide to determine therapy. The authors are careful to note that not all of the genetic abnormalities detectable appear to influence risk.
The levels of expression of Survivin and a growing number of identified molecular markers vary from patient to patient for a specific cancer, such as colorectal, and also vary with tumor type, such as gastric, lung, breast, etc. This makes the analysis and characterization of a patient’s tumor tremendously complex, but strongly emphasizes the need, noted earlier, to be able to do a complete genetic/protein characterization of each patient’s tumor.
Determination of the complete molecular pattern of a patient’s tumor will provide a number of important potential applications. These include: (a) as mentioned, the ultimate classification/staging of the tumor; (b) assessing risk for failure and, therefore, determining appropriateness of adjuvant therapies; (c) providing monitors of therapeutic efficacy; and (d) ultimately providing specific molecular targets for therapy.
Both of the investigations reported in the May issue relied on the development of highly specific monoclonal antibodies as the tool to identify and quantitate the presence of these tumor markers. In the one case, the use of anti-cytokeratin antibody to detect an antigen on malignant cells in lymph nodes increases the ability to detect micrometastases in nodes that would otherwise be classed as negative by standard hematoxylin and eosin staining. Although this is important, it may not have the ultimate power that is illustrated by the use of monoclonal antibodies to detect specific proteins within or associated with the tumor cell. Such studies attempt to increase our understanding of the potential role of over-expressed or aberrantly expressed proteins in the processes of cancer cell growth and cancer metastasis.
During the past decade, investigators have focused on the analysis and characterization of single genes, such as Survivin, in the specific cancer. Current and future technologies of gene arrays and single nucleotide polymorphism analysis will permit the detection of 50,000 genes in a single assay. The ability to profile the specific mutation/polymorphism patterns with risk of failure and with the response to different therapies is increasingly imminent. These two studies emphasize the direction of future research and point to a day when such profiling will be used to design the required molecularly targeted therapy.
Received for publication March 30, 2001. Accepted for publication April 9, 2001.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
