Early Detection Research Network
Team Project

Colon Reference Set Application: Robert Getzenberg - Johns Hopkins (2008)

Colon Ref Set App: Getzenberg (2008)
250
Feng, ZidingFred Hutchinson Cancer Research Center
CCSA-3, CCSA-4,
No design specified.
Proteomics
G.I. and Other Associated Cancers Research Group

The goal of this project is to confirm the early validation of CCSA-3 and CCSA-4 as biomarkers for the early detection of adenocarcinoma of the colon and rectum.

Aim 1: Characterize the distribution of CCSA-3 and CCSA-4 values in a blinded set of serum samples obtained from patients with colorectal adenocarcinoma, colorectal adenomas, , hyperplastic polyps, inflammatory bowel disease and normal colons. Aim 2: Estimate the ROC curves for discrimination between: patients with colorectal cancer versus others; colorectal cancer or adenoma versus others; colorectal cancer, adenoma, or hyperplastic polyps versus IBD; colorectal cancer, adenoma, or hyperplastic polyps versus IBD or normal.
1.1   Characterize the distribution of CCSA-3 and CCSA-4 values in a blinded set of serum samples obtained from patients with colorectal adenocarcinoma, colorectal adenomas, hyperplastic polyps, inflammatory bowel disease and normal colons. Both CCSA-3 and CCSA-4 are reported as concentrations on continuous scales. Graphical displays and descriptive statistics will be presented by disease category. If the graphical displays indicate the with-population distributions are significantly, non-Gaussian, the values will be appropriately transformed. Analysis of variance will be used to test the null hypothesis (at a 5% significance level) that the means of the populations are equal. 1.2   Estimate the sensitivity and specificity for discrimination between: 1) patients with colorectal cancer versus adenoma, hyperplastic polyps or IBD; 2) colorectal cancer or adenoma versus hyperplastic polyps or IBD; 3) colorectal cancer, adenoma, or hyperplastic polyps versus IBD; 4) colorectal cancer, adenoma, or hyperplastic polyps versus IBD or normal. For each of the four tests, the sensitivity and specificity will be estimated assuming prevalence of 0.50 and equal false negative and false positive costs. The null hypotheses that sensitivity < 0.75 and specificity < 0.75 will be tested using a one-side exact binomial test at a 5% significance level; rejection of the null hypotheses constitutes an argument for advancing the markers to a full validation study. 1.3   Empirical ROC curves will be plotted for Tests 1-4. Optimal sensitivity and specificity will be estimated depending on assumptions about prevalence and the costs of false positives and negatives. Tests 1, 2 and 3 are of particular interest for clinical management, while Test 4 is more relevant for population screening.
No Data Received at DMCC

No datasets are currently associated with this protocol.


Version 5.0.2