NEW YORK (GenomeWeb) – A genomic and molecular analysis of esophageal carcinomas has indicated that they are not a single type of cancer.
Researchers from The Cancer Genome Atlas Research Network led by Adam Bass at the Dana-Farber Cancer Institute examined more than 160 esophageal carcinomas for copy number, SNP, methylation, and other differences. As they reported today in Nature, they found that esophageal squamous cell carcinomas and esophageal adenocarcinomas differ not only at the histopathological level but also at the molecular level. In addition, the researchers uncovered that esophageal squamous cell carcinomas more closely resembled head and neck squamous cell carcinomas, while esophageal adenocarcinomas were more similar to chromosomally unstable gastric adenocarcinomas.
"These analyses call into question the premise of envisioning esophageal carcinoma as a single entity," Bass and his colleagues wrote in their paper.
The TCGA team collected fresh-frozen tumor samples as well as germline DNA from 90 patients with esophageal squamous cell carcinoma and 72 patients with esophageal adenocarcinoma . The researchers then compared those samples in an analysis drawing on whole exome, mRNA, and microRNA sequencing as well as SNP array and DNA methylation profiling data. Subsets of the esophageal carcinoma samples also underwent low-pass whole-genome sequencing and reverse-phase protein array proteomic analysis.
Using the algorithm MutSig, Bass and his colleagues noticed differences in genomic alterations among the esophageal cancer types.
For instance, esophageal squamous cell carcinomas commonly harbored TP53, NOTCH1, and TGFBR2 mutations, while esophageal adenocarcinomas had TP53, CDKN2A, and ERBB2 mutations. Additionally, esophageal squamous cell carcinomas had focal amplifications affecting SOX2, TERT, and FGFR1 and deletions at RB1, while esophageal adenocarcinomas had recurrent amplifications affecting VEGFA, ERBB2, and GATAG and recurrent SMAD4 deletions.
At the same time, the researchers' mutational and copy number data indicated that cell cycle regulators were frequently altered in both cancer types, as were receptor tyrosine kinases, members of the TGF-β pathway, and chromatin-modifying enzymes.
This suggested to them that although similar pathways might be affected in both types of esophageal carcinoma, different genes seem to be altered. This, they added, likely reflects differences in pathophysiology and could indicate that patients might benefit from different therapeutic approaches.
"These molecular data show that histological subtypes of EAC and ESCC are distinct in their molecular characteristics across all platforms tested," the researchers wrote. "Our analyses therefore argue against approaches that combine EAC and ESCC for clinical trials of neoadjuvant, adjuvant, or systemic therapies."
Clustering of esophageal squamous cell carcinoma data further split them into three subtypes, the researchers reported, one of which, dubbed ESCC1, was marked by changes in the NRF2 pathway. In addition, they found that the gene expression pattern of this subtype looked like that of lung squamous cell carcinoma and head and neck squamous cell carcinoma subtypes.
Subtype ESCC2 tumors, meanwhile, had higher mutation rates in NOTCH1 or ZNF750 as well as greater leukocyte infiltration, and ESCC3 subtype tumors harbored mutations predicted to affect the PI3K pathway.
As there has been debate about how to classify cancers that develop near the gastroesophageal junction, Bass and his colleagues compared the molecular profiles of these esophageal cancers to those of tumors that developed in nearby tissues. Esophageal squamous cell carcinomas, they reported, more strongly resembled head and neck squamous cell carcinoma, but esophageal adenocarcinomas more closely resembled gastric cancer.
When they jointly analyzed esophageal adenocarcinomas and gastric cancers, they further found that esophageal adenocarcinomas and chromosomally unstable gastric tumors formed a distinct group. This, the researchers said, could indicate that gastric and esophageal chromosomally unstable tumors might not be distinct tumor types and that gastroesophageal adenocarcinomas could exist on a continuum of subtypes.