New serum biomarker appears to detect colon cancer

ORLANDO-A serum marker, colon-cancer-specific antigen-2 (CCSA-2), detects nearly 97% of colorectal cancers and can differentiate between advanced adenomatous polyps and less ominous ones, according to investigators from Johns Hopkins and the University of Pittsburgh who have collaborated on research in this area.

Using proteomic analysis of the nuclear structure of proteins, the investigators previously determined that two similar antigens-CCSA-3 and CCSA-4-could distinguish persons with colorectal cancer from those without. They now believe that a third antigen, CCSA-2, may be an even better indicator, as it can categorize degrees of abnormality according to the level of the antigen in the blood.

“CCSA-2 has characteristics that are clearly distinct from the other CCSA markers,” said Robert H. Getzenberg, PhD, of the Johns Hopkins University School of Medicine. He presented the findings at the 2008 Gastrointestinal Symposium (abstract 276).

Applied to serum samples from 260 individuals who underwent colonoscopies, CCSA-2 detected 96.4% of colorectal cancer cases and 91.3% of advanced adenomas. The respective specificities were 81.7% and 79.2%. The specificity is far superior to that of carcinoembryonic antigen (CEA), the protein now used as a marker, Dr. Getzenberg noted.

Compared with the other CCSA antigens, CCSA-2 is unique because it correlates serum levels of the antigen with the size of the adenoma: The higher the level of biomarker, the larger the growth, with the highest elevations indicating the presence of cancer. This allowed investigators to distinguish between advanced adenomas and non-advanced adenomas, and between low-risk polyps and normal colon, he said.

The investigators plan to validate the study findings in 500 subjects in a multicenter study.

How they did it

In describing the laboratory approach to this discovery, Dr. Getzenberg explained that proteomics focused only on nuclear structural proteins allows investigators to analyze a reasonably sized subset of proteins, to eliminate abundant and interfering components, to study proteins of low abundance or specialized function, and to concentrate on fractions of proteins with biologic importance.

“We are not just randomly looking at proteins, but focusing on a small subset that represents less than 1% of the cell’s total proteins. These proteins are relatively insoluble and have different parameters than other proteins, and therefore are not often studied by general proteomic approaches,” he said.

By comparing tissue from colon cancer specimens, normal adjacent colon, colon polyps, and normal donor tissue, they identified a series of seven nuclear matrix proteins that were cancer specific, which they labeled CCSA. CCSA-2 is expressed at the junction of advanced adenoma and invasive tumor. CCSA-3 and -4 are expressed earlier in the evolution of adenomatous polyps. Assays were developed to detect these proteins in the blood.

Central to the studies is the examination of chromatin remodeling by the modification of non-histone proteins. These changes may have importance in the cancer process. “We believe we may be looking at a nuclear structure that is an important biomarker and may also play a role in the cancer process,” he said.

Bundling markers

Jordon D. Berlin, MD, associate professor of medicine and clinical director of GI oncology at Vanderbilt University, commented: “CCSA-2 shows promise, but further testing is needed.”

In the long run, he said, “markers being developed for screening will probably work best as part of a panel, or bundle, for instance, CCSA-2 plus CCSA-3, CCSA-4, CEA, and possibly others.”

Dr. Berlin noted one problem in this research area: “Many of the biomarkers we have heard about are under proprietary rules that are interfering with the development of the markers we need. Fortunately, this problem is now being addressed.”