Genomic Test Predicts Metastasis Risk in Prostate Cancer

Zosia Chustecka

February 22, 2013

A new test used after prostatectomy can indicate whether a man is at high risk or low risk for metastases, and therefore whether or not he is a candidate for additional treatment. The test, known as Decipher (developed by GenomeDx Biosciences), could help prevent overtreatment, according to Ketan Badani, MD, urologic oncologist at Columbia University in New York City.

He presented study results at the 2013 Genitourinary Cancers Symposium (GUCS) in Orlando, Florida, which showed how physicians who were provided results from this genomic test altered their management decisions.

The test is currently available to select physicians and their patients as part of ongoing clinical studies. The company hopes that the test will become more widely available in 2013.

The test measures 22 markers associated with aggressive prostate cancer, and uses samples of paraffin-embedded tumor tissue that have been removed during prostatectomy.

The resulting percentage score indicates whether a man is at high or low risk for metastases, and therefore whether he is a candidate for further treatment, such as radiation and/or androgen-suppression therapy, Dr. Badani explained in an interview.

The Decipher test is similar to the OncotypeDx genomic test used in breast cancer, he noted. Both provide a risk stratification based on biologic material in the tumor tissue. "We haven't had this in prostate cancer before," Dr. Badani added.

Currently, physicians evaluate clinical and pathologic information, and rely mainly on their clinical judgment to decide which patients need further treatment and which can be followed with observation, Dr. Badani explained. However, this decision is "highly variable from doctor to doctor," he said.

"I think the information provided by this genomic test can help to standardize that decision," he said.

"The truth is, we probably overtreat a lot of these patients.... A lot of these men with high risk factors do not develop metastases, but we don't know who that is going to be, so our gut feeling is to treat everybody," he explained.

"This new test can help us decide who is most likely to develop metastases, which means we can spare a lot of patients from having additional treatment, with all the attendant side effects," he said.

Another Study

Several nomograms are available to help with the treatment decision-making process. One is the CAPRA-S nomogram (based on standard clinicopathologic parameters), which was used in another study also reported at GUCS.

Matthew Cooperberg, MD, from the University of California, San Francisco, and colleagues conducted this study in a contemporary cohort of patients who had undergone radical prostatectomy. From a group of 1010 patients treated at the Mayo Clinic from 2000 to 2006, a case–control design was used to analyze a subset of 219 men with 1 or more high-risk features and available paraffin-embedded tissue.

After a median follow-up of 6 years, 27 of the 219 had died from prostate cancer.

CAPRA-S had identified 103 men as being at high risk. Dr. Cooperberg and colleagues also used the Decipher test in their study. It identified only 49 men as being at high risk, 19 of whom died from prostate cancer. Decipher identified the other 54 men as being at low risk, only 1 of whom died of prostate cancer.

Notably, the Decipher test was able to "down-risk" more than half of the men who had been identified as high risk by the CAPRA-S nomogram, Dr. Cooperberg and colleagues note.

A multivariate analysis found that both tests were independently prognostic of prostate-cancer-specific mortality, with hazard ratios of 1.62 for Decipher and 1.22 for CAPRA-S. However, combining the 2 tests was more accurate than either test alone, the researchers conclude.

Changing Management Decisions

Dr. Badani and colleagues showed that physicians who were presented with information from the genomic test altered their management decisions. This happened to a greater extent than expected, he told Medscape Medical News.

Their study involved 20 urologic oncologists from 18 institutions who were asked to review the cases of 12 men who had undergone radial prostatectomy.

First, they used standard clinicopathologic information, including data on prostate-specific antigen levels before surgery (and after surgery, where available) and on the pathologic stage and grade of the tumor. "These are the factors that we use right now when deciding whether a patient needs further treatment after prostatectomy," Dr. Badani explained.

Next, they were asked to use information from the Decipher genomic test to make the treatment decision.

After the genomic test results were unblinded, treatment recommendations had changed in 43% of cases. Specifically, in the cases for which further treatment had been recommended on the basis of clinicopathologic information, the recommendation changed from treatment to observation in 31%.

The urologic oncologists reported that the genomic test results influenced their recommendation in 63% of cases, and said that the information was clinically relevant in 84% of cases.

"Our study demonstrates that practicing urologists will utilize genomic-based patient-specific information and will trust these data to make treatment decisions for their patients," Dr. Badani said in a statement.

Comments on the new test appear in press materials released by the American Society of Clinical Oncology, which is one of the sponsors of GUCS. Leonard Gomella, MD, FACS, who was a member of the GUCS news planning team, said that "the development of this genomic prostate score provides a unique tool with more information than is currently routinely available to allow physicians and their patients to choose the prostate cancer active surveillance approach with confidence."

Both studies were funded by GenomeDx. Several coauthors from both studies are employees of GenomeDx.

2013 Genitourinary Cancers Symposium (GUCS): Abstract 196 and Abstract 60.