Abstract
There are currently no reliable diagnostic and prognostic
markers or effective treatments for malignant pheochromocytoma. This
study used oligonucleotide microarrays to examine gene expression profiles
in pheochromocytomas from 90 patients, including 20 with malignant
tumors, the latter including metastases and primary tumors
from which metastases developed. Other subgroups of tumors included
those defined by tissue norepinephrine compared to epinephrine contents
(i.e., noradrenergic versus adrenergic phenotypes), adrenal versus
extra-adrenal locations, and presence of germlinemutations of genes predisposing
to the tumor. Correcting for the confounding influence of noradrenergic
versus adrenergic catecholamine phenotype by the analysis of
variance revealed a larger and more accurate number of genes that discriminated
benign from malignant pheochromocytomas than when the
confounding influence of catecholamine phenotype was not considered.
Seventy percent of these genes were underexpressed in malignant compared
to benign tumors. Similarly, 89% of genes were underexpressed in
malignant primary tumors compared to benign tumors, suggesting that
malignant potential is largely characterized by a less-differentiated pattern
of gene expression. The present database of differentially expressed
genes provides a unique resource for mapping the pathways leading to
malignancy and for establishing new targets for treatment and diagnostic
and prognostic markers of malignant disease. The database may also
be useful for examining mechanisms of tumorigenesis and genotype
phenotype relationships. Further progress on the basis of this database
can be made from follow-up confirmatory studies, application of bioinformatics
approaches for data mining and pathway analyses, testing in
pheochromocytoma cell culture and animal model systems, and retrospective
and prospective studies of diagnostic markers.
Close Window