Gene expression profiling of peritoneal metastases from appendiceal and colon cancer demonstrates unique biologic signatures and predicts patient outcomes.

Published

Journal Article

BACKGROUND: Treatment of peritoneal metastases from appendiceal and colon cancer with cytoreductive surgery and hyperthermic intraperitoneal chemotherapy (HIPEC) shows great promise. Although long-term disease-free survival is achieved in some cases with this procedure, many patients have recurrence. Oncologists have treated such recurrences of appendiceal cancer similarly to colorectal carcinoma, which has been largely ineffective. This study uses gene expression analysis of peritoneal metastases to better understand these neoplasms. STUDY DESIGN: From a prospectively maintained database and tissue bank, 41 snap frozen samples of peritoneal metastases (26 appendiceal, 15 colorectal) from patients undergoing HIPEC with complete cytoreduction and more than 3 years of follow-up underwent global gene expression analysis. Distinct phenotypes were identified using unsupervised hierarchical clustering based on differential gene expression. Survival curves restratified by genotype were generated. RESULTS: Three distinct phenotypes were found, 2 consisting of predominantly low grade appendiceal samples (10 of 13 in Cluster 1 and 15 of 20 in Cluster 2) and 1 consisting of predominantly colorectal samples (7 of 8 in Cluster 3). Cluster 1 consisted of patients with good prognosis and Clusters 2 and 3 consisted of patients with poor prognosis (p = 0.006). Signatures predicted survival of low- (Cluster 1) vs high-risk (Cluster 2) appendiceal (p = 0.04) and low-risk appendiceal (Cluster 1) vs colon primary (Cluster 3) (p = 0.0002). CONCLUSIONS: This study represents the first use of gene expression profiling for appendiceal cancer, and demonstrates genomic signatures quite distinct from colorectal cancer, confirming their unique biology. Consequently, therapy for appendiceal lesions extrapolated from colonic cancer regimens may be unfounded. These phenotypes may predict outcomes guiding patient management.

Full Text

Duke Authors

Cited Authors

  • Levine, EA; Blazer, DG; Kim, MK; Shen, P; Stewart, JH; Guy, C; Hsu, DS

Published Date

  • April 2012

Published In

Volume / Issue

  • 214 / 4

Start / End Page

  • 599 - 606

PubMed ID

  • 22342786

Pubmed Central ID

  • 22342786

Electronic International Standard Serial Number (EISSN)

  • 1879-1190

Digital Object Identifier (DOI)

  • 10.1016/j.jamcollsurg.2011.12.028

Language

  • eng

Conference Location

  • United States