Not all malignant tumors are bad all of the time.

Prostate cancer, for example, often begins as a benign growth that merely causes discomfort in urinating. The condition can continue for months or years before turning nasty and progressing from benign to invasive to metastatic to lethal.

Lethality strikes down 35,000 to 40,000 men in the U.S. every year. Annually, another 200,000 are diagnosed with prostate cancer (PC). It's the commonest neoplasm in American males and the second leading cause of cancer deaths in men.

During that lengthy, uncertain period of tumor progression, patients and their urologists have a lot of decisions to discuss. One of their stand-by biomarkers is prostate-specific antigen (PSA), by which a monoclonal antibody detects and quantifies cancerous prostate cells, as an indicator of malignancy. Clinical and research pathologist Arul Chinnaiyan observes anent PSA's role in medical practice.

"Prostate-specific antigen is a test that's useful in diagnosing PC, but is relatively nonspecific, in that men who have noncancerous conditions - benign prostatic hyperplasia, prostatits and so forth - can also have elevated PSA. Furthermore, PSA levels don't tell us much about prognosis - whether a patient has the aggressive form of PC, or a more indolent, slow-growing disease. PSA is helpful in screening and diagnosing PC, but not prognosis, predicting the course a patient's PC will likely take."

Chinnaiyan, an assistant professor of pathology and urology at the University of Michigan School of Medicine at Ann Arbor, is senior author of a paper in Nature, released today online and dated Oct.10, 2002. It's titled: "The polycomb group protein EZH2 is involved in progression of prostate cancer."

EZH2 takes its acronym from two genes - Enhancer and Zeste - originally discovered in the fruit fly, Drosophila melanogaster. The polycomb protein complex, also identified initially in Drosophila, controls DNA transcriptional memory.

Cancer Progressors Confront Tumor Aggressors

"There are two basic findings in this Nature article," Chinnaiyan told BioWorld Today. "One is more clinically relevant, the other more biological. The clinical aspect," he continued, "is that we've identified a novel protein, EZH2, as being elevated in metastatic prostate cancer. When we go back and look at patients with clinically localized PC who have higher levels of this EZH2 protein, they tend to be progressors. They have a more aggressive disease, rather than a more indolent, slow-growing stage, suggesting that this EZH2 protein might be useful as a signature lethal biomarker. Usually, when a PC patient has widely metastatic disease, which gets into his lungs, liver, brain or bone, it's incurable. So clinically, our EZH2 test will help with treatment decisions - whether such patients should be kept on a watchful-waiting protocol, or prescribed more aggressive treatment, such as surgery or radiation, which has some morbidity of its own.

"EZH2 - the gene encoding EZH2 - resides on human chromosome 7," Chinnaiyan went on. "We were the first to disclose the association of gene and protein to the prostate about a year ago. It will be the first lethal biomarker of PC.

"The EZH2 gene in general is a transcriptional repressor," Chinnaiyan continued, "keeping other genes turned off. So it's part of this cellular memory machinery that essentially maintains cellular identity to cells. It tells which genes to turn on or off and makes muscle cell what it is, skin cell what it is. Perturbation in this memory machinery can sometimes lead to cancer."

The Michigan co-authors used advanced microarrays - DNA chips - to interrogate the genes that are being expressed in prostate cancer. Their chip was made up of about 20,000 cDNAs to various genes. "We were looking for differences between benign and prostate cancer," Chinnaiyan explained, "the genes that distinguish those differences. They showed that EZH2 expression topped the list of 55 genes found to be more active in metastatic PC than in localized prostate cancer.

"We fabricated chips that displayed these 20,000 genes," he went on, "then isolated RNA from benign prostate and clinically localized cancer, as well as metastatic PC. We labeled those RNAs with fluorescent dyes and incubated them on the chips, picking up differences between benign and cancer by color contrast."

Animal Trials Next, Human Patients Later

"If we could find a drug or an RNA interference technique that has limited toxicity," Chinnaiyan suggested, "it would certainly have potential in human trials down the road. But I think our next phase will be animal studies. A recent report showed that RNA interference can be used in rodent models. So now I think our impending experiments would be to see if we can specifically knock down the EZH2 protein in mice, using RNA silencing. Our particular study suggested that therapeutically, if there's a way to knock down EZH2 or its activity, metastatic cells might stop growing. We're just taking advantage of this recently suggested technique in mammalian cells. It's analogous to what people have been doing with antisense, to knock down transcripts. But in this case, it's a small interfering RNA duplex of about 21 nucleotides that we designed against the RNA sequence."

Chinnaiyan expects that between two and three years from now, his tests for PC diagnosis and prognosis will enter clinical trials. "Right now," he observed, "our next step is to validate these findings in a larger patient cohort, at Michigan itself, as well as at institutions elsewhere in this country, to see if our findings hold up. Then our next phase, which we are now developing concurrently, would be to see if we can apply this PC test to needle biopsies. This would serve as a red flag to physicians and help save the lives of men with the most aggressive form of the disease. The studies that we conducted here used 1,000 tissue samples from prostatectomy surgery.

"The university is applying for a patent to protect our gene-expression profiles," he said. "As potential licensees, it is bringing in various companies to see who's interested. Primarily, the most interest expressed so far comes from some diagnostic companies that want to implement our diagnostic and prognostic biomarkers.

"We now have a multicenter trial in the discussion stage," Chinnaiyan said. "We're planning to do it at the Dana Farber Cancer Center of Harvard-affiliated Brigham and Women's Hospital."