Active matrix metalloproteinase-2 activity discriminates colonic mucosa, adenomas with and without high-grade dysplasia, and cancers

Pathologic assessment of colorectal adenomas, a complex task with significant interobserver variability, typically defines the scheduling of surveillance colonoscopies after removal of adenomas. We have characterized the activity levels of pro matrix metalloproteinase-2, active matrix metalloproteinase-2, and matrix metalloproteinase-9 in colorectal adenomas and carcinomas as potential markers of pathologic progression during colorectal tumorigenesis. Endogenous fully activated matrix metalloproteinase-2, in particular, has been studied less frequently in adenomas due to difficulties in detection. For this report, tissues (n = 119) from 51 individuals were extracted and assayed on gelatin zymograms with digital standardization to nanogram quantities of purified active controls. Resulting data were assessed by graphical and multinomial logit regression analyses to test whether matrix metalloproteinase-2 or matrix metalloproteinase-9 activities could discriminate among 4 different types of colorectal tissue (normal mucosa, adenomas with or without high-grade dysplasia, and invasive carcinomas). Active matrix metalloproteinase-2 successfully discriminated among these tissue categories. Median activity for active matrix metalloproteinase-2 increased in a stepwise fashion with pathologic progression from normal mucosa to adenoma without high-grade dysplasia to adenoma with high-grade dysplasia to cancer. Although pro matrix metalloproteinase-2 and pro matrix metalloproteinase-9 activities could discriminate to some extent among tissue categories, those effects did not contribute additional information. Active matrix metalloproteinase-2 activity correlated significantly with histopathologic assessment of colorectal tissues. The ability of active matrix metalloproteinase-2 to distinguish adenomas with high-grade dysplasia from adenomas without high-grade dysplasia may be particularly useful in predicting future colorectal cancer risk for an individual, thus optimizing scheduling of surveillance colonoscopies. (C) 2011 Elsevier Inc. All rights reserved.