Analysis of Columnar Lined Esophagus Reveals Less Frequent Mutations in Non-Goblet Cell Metaplasia Than Intestinal Metaplasia
Santhoshi Bandla, Jeffrey H Peters, David Ruff, Shiaw-Min Chen, Chieh-Yuan Li, Kunchang Song, Kimberly Thoms, Virginia R Litle, Thomas Watson, Nikita Chapurin, Arjun Pennathur, James D Luketich, Derick Peterson, Austin Dulak, Lin Lin, Adam Bass, Daivd G Beer, Tony E Godfrey, Zhongren Zhou. URMC, Rochester, NY; University of Pittsburgh, Pittsburgh, PA; Life Technologies, Foster City, CA; Dana Farber Cancer Center, Boston, MA; University of Michigan, Ann Arbor, MI
Background: Gastroesophageal reflux causes the formation of columnar metaplasia of esophageal mucosa, which is classified as either intestinal type (IM) defined by goblet cells or as non-goblet type (NGM). In the United States, the presence of goblet cells is required for the diagnosis of Barrett's esophagus (BE), the strongest known risk-factor for development of esophageal adenocarcinoma (EAC). Worldwide, there is considerable debate regarding the cancer risk of NGM and whether it should be included in the definition of BE. IM is known to harbor frequent genetic alterations that are often observed in adenocarcinoma. The aim of this study was to investigate the frequency of EAC-associated genomic mutations in NGM and IM as a means to infer relative cancer-risk of the two epithelia.
Design: 19 NGM, 16 IM and 5 composite tissues (both NGM and IM) were analyzed along with matched normal squamous specimens using targeted re-sequencing of 20 genes found to be frequently mutated in EAC. Sequence data from metaplasia samples was compared with patient-matched normal DNA in order to filter sequence polymorphisms and identify variants. In addition, mutation data from 66 EAC samples was compared with that from metaplasia samples to see if mutation load could classify disease status. A pairwise Receiver-Operator Characteristic (ROC) analysis was used.
Results: This analysis revealed a total of 16 non-synonymous and 5 synonymous mutations. 2 mutations were observed in independent NGM samples (2/19; 10.5%) in CDKN2A and COL11A. 14 (14/21; 66.7%) mutations were observed in the IM and composite samples including non-synonymous mutations in TP53 (n=4; 24%), SMARCA4 (n=2; 12%), LRP1B (n=3; 18%), ERBB2 (n=1; 6%) and CDKN2A (n=1; 6%) and synonymous mutations in APC, ERBB2 and LRPB1. Mutation load was found to have strong power to classify EAC versus metaplasia. ROC analysis resulted in an area under the curve of 0.954 for EAC vs. any metaplasia and 0.915 for IM alone versus EAC.
Conclusions: Our results show that IM has a much higher frequency of cancer-associated mutations than NGM and is therefore the potential to pose a higher risk for development of EAC. In addition, the mutation load appears to have good potential for detecting disease status in patients with BE. This approach could be applied to cytology specimens for BE surveillance.
Monday, March 4, 2013 9:00 AM
Proffered Papers: Section D, Monday Morning