University Of Michigan Scientists Reveal Prostate Cancer's Molecular Fingerprint. Study Links Proteins To Patient Prognosis

Until now, physicians have been unable to decode these fingerprints, which hold the key to understanding the relationship between gene expression and future prognosis for men with prostate cancer. But a new study from the University of Michigan Medical School, published in the Aug. 23 issue of Nature, offers scientists their first look at the genetic and molecular profile of prostate cancer.

"Our study has important applications in the diagnosis, prognosis and treatment of prostate cancer," says Mark Rubin, M.D., a co-author of the Nature paper and an associate professor of pathology and urology in the U-M Medical School. "The ultimate goal is to help physicians determine which patients need immediate, aggressive treatment and which can be watched and treated conservatively."

Researchers at the U-M's Comprehensive Cancer Center analyzed prostate tissue samples from 50 men and found nearly 200 genes or gene fragments in which expression profiles varied consistently, depending on whether the tissue was normal or malignant.

U-M researchers used more than 80 complementary DNA microarrays to assess gene expression profiles in four types of tissue. These included normal prostate tissue from men with and without prostate cancer, tissue with benign changes, localized prostate cancer and aggressive, metastatic cancer. Tissue samples were obtained from the U-M Prostate Specialized Program of Research Excellence (SPORE) tumor bank, funded by the National Cancer Institute and directed by study co-author Kenneth Pienta, M.D., a professor of internal medicine and surgery in the U-M Medical School.

"Previous prostate cancer studies focused on one gene at a time," says Arul Chinnaiyan, M.D., Ph.D., an assistant professor of pathology in the U-M Medical School, who directed the study. "Using microarray technology, we were able to look at thousands of genes in prostate cells simultaneously. This is important, because it is most likely that many genes are involved in the development and progression of prostate cancer - each controlling a different step in the process."

While some of the genes identified in the U-M study are well known to cancer researchers, many others have never before been associated with prostate cancer. Two of these new genes are hepsin and pim-1, which could turn out to be important new clinical biomarkers for prostate cancer, according to Rubin.

When U-M scientists tested more than 700 prostate specimens for the presence of hepsin protein, the highest levels were found in pre-cancerous tissue - the type seen just before prostate cancer develops, according to Chinnaiyan. Lowest levels were found in benign prostate tissue. While hepsin's exact function is still unknown, U-M researchers suspect it plays a key role in establishment of tumors. Pim-1, a known cancer-causing gene, also was highly expressed in prostate cancer. Importantly, levels of hepsin and pim-1 protein both were correlated with patient prognosis.

To ensure accuracy, the genetic profile of normal human prostate tissue was used as an experimental control at each stage of the study. "Our reference point was always normal tissue, not generalized cell lines," says Chinnaiyan. "When we say a gene is up-regulated or more frequently expressed in prostate cancer, it means as compared to normal prostate tissue -- often normal tissue from the same patient, which was adjacent to the tumor."

"Without the U-M's prostate tissue bank this work would not have been possible," says Rubin. "We were also fortunate that U-M is one of only three medical schools in the United States with a rapid autopsy program where tissue samples are obtained within hours of death," adds Chinnaiyan. "Since RNA is prone to rapid degradation, it's impossible to track gene expression in cells from autopsied tissue, unless RNA is isolated soon after the patient dies."

According to Chinnaiyan, the next step is determining the functional role for each gene identified in the study as having an association with prostate cancer. "This paper will generate a great deal of work for researchers in many laboratories," he predicts.

"It has been 15 years since the Prostate Specific Antigen (PSA) test became available in 1987," says Rubin. "This approach could give us many new diagnostic tests within three to five years. Eventually, it could lead to a diagnostic kit physicians could use to determine the best treatment and prognosis for their patients with prostate cancer."

The research was supported by the National Cancer Institute's Specialized Program of Research Excellence in Prostate Cancer (#P50 CA 69568). The U-M has applied for a patent on prostate cancer gene expression profiles for future diagnostic and therapeutic use.

Others U-M scientists involved in the study include: Saravana M. Dhanasekaran, Ph.D., research fellow; Terrence R. Barrette, research associate; Debashis Ghosh, Ph.D., assistant professor of biostatistics in the U-M School of Public Health; Rajal Shah, M.D., assistant professor of pathology; Sooryanarayana Varambally, Ph.D., research fellow; and Kotoku Kurachi, Ph.D., professor of human genetics.