Published online before print December 28, 2007

This Article Full Text (Rapid PDF) All Versions of this Article: jmoldx.2008.070077v1 10/1/93    most recent Purchase Article View Shopping Cart Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Melnikov, A. A. Articles by Levenson, V. V. Search for Related Content PubMed PubMed Citation Articles by Melnikov, A. A. Articles by Levenson, V. V.

Copyright © 2008 American Society for Investigative Pathology
Journal of Molecular Diagnostics, doi:10.2353/jmoldx.2008.070077

Accepted for publication October 2, 2007.

Article

Array-Based Multiplex Analysis of DNA Methylation in Breast Cancer Tissues

Anatoliy A. Melnikov^*, Denise M. Scholtens^*, Elizabeth L. Wiley, Seema A. Khan, and Victor V. Levenson^*@

From the Robert H. Lurie Comprehensive Cancer Center,* and the Departments of Preventive Medicine, Pathology, and Surgery, Feinberg School of Medicine, Northwestern University, Chicago, Illinois

^@ To whom correspondence should be addressed. E-mail: levenson{at}northwestern.edu.

	Abstract

Abnormal DNA methylation is well established for cancer cells, but a methylation-based diagnostic test is yet to be developed. One of the problems is insufficient accuracy of cancer detection in heterogeneous clinical specimens when only a single gene is analyzed. A new technique was developed to produce a multigene methylation signature in each sample, and its potential for selection of informative genes was tested using DNA from formalin-fixed, paraffin-embedded breast cancer tissues. Fifty-six promoters were analyzed in each of 138 clinical specimens by a microarray-based modification of the previously developed technique. Specific methylation signatures were identified for atypical ductal hyperplasia, ductal carcinoma in situ, and invasive ductal carcinoma. Informative promoters selected by Fisher's exact test were used for composite biomarker design using nave Bayes algorithm. All informative promoters were unmethylated in disease compared with normal tissue. Cross-validation showed 72.4% sensitivity and 74.7% specificity for detection of ductal car-cinoma in situ and invasive ductal carcinoma, and 87.5% sensitivity and 95% specificity for detection of atypical ductal hyperplasia. These results indicate that informative cancer-specific methylation signatures can be detected in heterogeneous tissue specimens, suggesting that a diagnostic assay can then be developed.