Identification of a DNA Methylation Marker That Differentiates Malignant Mesothelioma from Reactive Mesothelial Cells on Effusion Samples.
Phyu P Aung, Anand Pathak, Liqiang Xi, Armando C Filie, Mark Raffeld. National Cancer Institute, Bethesda, MD
Background: Differentiating reactive mesothelial (RM) proliferations from malignant mesothelioma (MM) in effusions is a challenging diagnostic dilemma for pathologists. Robust, highly sensitive and specific assays are needed to solve this diagnostic problem. Epigenetic changes to DNA, including gene specific promoter methylation and global hypomethylation, are known to occur during tumorigenesis, and there have been several recent studies of MM describing global gene methylation patterns. In this study, we analyzed promoter methylation of several candidate genes for their utility in differentiating RM cells from MM in effusions.
Design: Cell block sections of 31 effusions from patients with MM involvement (17) or with RM proliferations (14) were retrieved from the files of the Cytology Section, Laboratory of Pathology, NCI. Guided by published methylation array data, we selected several candidate genes for detailed promoter methylation analysis. These included several members of the TRAIL/Death Receptor-Decoy Receptor pathway (TRAIL, DcR1, DcR2, Caspase 8), and the retinoic acid responder protein 1 gene, RARRES1. We also assessed Line 1 promoter methylation, as a surrogate for global DNA hypomethylation. Promoter methylation was analyzed using sensitive and quantitative bisulfite pyrosequencing assays.
Results: Of the TRAIL receptor pathway genes, DcR1 displayed the most dramatic methylation differences with MM cases having an average methylation of 27% and the RM cases, 5.8% (p= 0.0009). Using an optimal cutoff for this biomarker as determined by receiver operator characteristic (ROC) curve analysis, this marker had a sensitivity of 88% and a specificity of 100%. While promoter hypermethylation of both DcR2 and TRAIL was also highly sensitive for MM, the specificity of both assays was relatively low, rendering both markers unsuitable for diagnostic use. Neither caspase 8 nor RARRES1 showed significant methylation differences between MM and RM. Only three cases of MM showed significant global hypomethylation as assessed by Line 1 analysis.
Conclusions: Our studies indicate that promoter methylation of Decoy Receptor 1 (DcR1) is a promising marker that differentiates MM cells from RM cells in cytologic effusions. Furthermore, global hypomethylation, using Line 1 as a surrogate marker, was of limited use in distinguishing MM from RM. The identification of DcR1 hypermethylation as a highly sensitive and specific molecular marker for MM has significant implications for the diagnosis of MM.
Wednesday, March 2, 2011 1:00 PM
Poster Session VI # 75, Wednesday Afternoon