Research may help predict risk for prostate cancer based on gene expression patterns

Jul 15 2004

According to a study published in the July 15 issue of the Journal of Clinical Oncology, genes expressed in benign tissue adjacent to prostate cancer tissue are much more similar to those expressed in prostate cancer tissue than previously thought.

This finding, the first of its kind, may help predict populations both at risk for prostate cancer and for disease progression based on gene expression patterns, say researchers at the University of Pittsburgh.

"It is not clear what molecular events are responsible for the progression of prostate cancer to a lethal form of the disease," said Jian-Hua Luo, M.D., Ph.D., senior author of the study and assistant professor, department of pathology, University of Pittsburgh School of Medicine. "But by exploring the biology of prostate cancer through the identification of genes and patterns of gene expression, we can more precisely understand what genetic changes cause the disease to progress and develop therapeutic targets to prevent its progression at an earlier stage."

In the study, Dr. Luo, also director of the gene array laboratory at the University of Pittsburgh, and colleagues used high throughput quantitative analysis to genetically profile prostate cancer tissue and noncancerous prostate tissue samples. They analyzed 152 human tissue samples including 66 samples of prostate cancer tissue, 60 samples of benign prostate tissue adjacent to the tumor, 23 samples of donor prostate tissue free of genitourinary disease and three prostate cancer cell lines. Through the analysis, the researchers identified a set of 671 genes whose expression levels were significantly altered in prostate cancer tissue compared to disease-free tissue and found that patterns of gene expression in benign adjacent prostate tissue were much more similar to prostate cancer tissue than to disease-free tissue.

According to Dr. Luo, the gene expression patterns of benign adjacent tissue were significantly overlapped with those of prostate cancer and distinctly different than the disease-free tissue. Furthermore, the adjacent tissue was so genetically altered that it resembled a cancer field effect, undergoing genetic changes similar to prostate cancer, even though it was morphologically benign tissue.

"It appears that genetic alterations in the prostate occur in parts of the gland that otherwise look benign," said Joel Nelson, M.D., professor and chairman, department of urology, University of Pittsburgh and co-author of the study. "We have long suspected a so-called field change in the prostate gland containing cancer, meaning some alteration has occurred throughout the prostate tissue. This study lends support for such a hypothesis."

The researchers also created a gene model using GeneSpringTM software to predict the aggressiveness of the disease and found that the expression profile model was more than 80 percent accurate in predicting the aggressiveness of the disease.

"Since only a fraction of prostate cancers are metastatic, identifying variables that predict the behavior of a prostate cancer tumor based on gene expression patterns should prove important in clinical management of the disease," said Dr. Luo. "The results of this study are a first step in that direction."