The polarizing debate over cancer screening tends to get particularly heated when it involves prostate cancer, a leading cause of cancer death in men. On one side are supporters of screening who point to lives saved. On the other are those who think that today’s primary screening tool—a blood test for prostate specific antigen, or PSA—leads to widespread overdiagnosis and overtreatment.

Yet those on both sides of this dispute tend to agree on one thing—the need for better diagnostic tools that can identify whether a particular tumor is likely to grow, spread and become life-threatening. And a recent study by a research group in the United Kingdom could eventually lead to just such a tool. Analyzing genomic “signatures,” the researchers were able to classify prostate cancer into five distinct types with different rates of recurrence. Adam Feldman, a urologist at MGH who specializes in the treatment of prostate cancer, speaks about the implications of this study for better screening and treatment.

Q: What’s the biggest challenge in treating prostate cancer?
A: There are more than 230,000 cases of prostate cancer diagnosed in the United States each year, resulting in 27,000 deaths. So obviously there are a significant number of cases that will not be lethal.

There are actually two big challenges in cancer that has not spread beyond the prostate. The first is to separate those who require treatment from those who do not. The second is to know who is highest risk and may need therapies beyond surgery or radiation.

Q: What does that mean for treatment?
A: Once prostate cancer has been diagnosed, there are three options for management. The patient can have his prostate removed surgically. Another option is to use radiation without surgery to treat the cancer. (Sometimes, depending on the severity of the cancer, hormonal therapy to lower a man’s testosterone is used along with the radiation.) These two treatments have very good success, but they can have a risk of significant side effects as well as the risk of recurrence in the future.

But a third path is active surveillance, in men with a low risk form of prostate cancer who meet certain criteria. This is where we basically say we think this cancer may never affect you.

Q: So why is this new study significant?
A: The authors of this study looked at patients who had surgical treatment, and analyzed the risk of recurrence. They looked at a large set of data that was generated by identifying abnormal changes in both the RNA and DNA of cancer cells, and they winnowed it down to a set of one hundred genes that correlated strongly with the recurrence of cancer after surgery. This showed that patients tended to cluster into five distinct groups based on these genetic changes. And each group had different rates of recurrence.

Next, they took that 100-gene signature and tested it on a second group of patients to see how it might predict outcomes. In fact, they even tested it on a third group of patients to ensure its validity. This repeated testing, or validation of their results, increases the chances that this test will work in other populations, meaning it may actually prove to have broader implications in identifying men with prostate cancer who are at risk of having it recur.

Q: How might all that affect prostate cancer treatment?
A: The authors were able to compare their genetic signature to the parameters we currently use. Their genomic test outperformed all of the standard parameters in predicting recurrence after surgery. It also outperformed several other publicly available genomic tests that are currently on the market.

We know that our current parameters are not perfect, and that if we treated everyone by those parameters, we would overtreat many patients. So an improved test such as this one could be very helpful. And at least in this study, it did appear to help identify those patients at highest risk after surgery.

One important thing to note is that we really don’t know how cost-effective these genomic tests are to use, compared to the standard clinical parameters that we currently use.

Q: Can this test be used for early screening, or is it only for men who have already been treated surgically?
A: One major challenge in prostate cancer is determining which patients would be appropriate candidates for active surveillance. Unfortunately that’s one of the things this research doesn’t address yet, although the authors comment on this need in their discussion. The current study only looked at patients who were already treated with surgery.

To test that, this 100-gene signature would have to be sifted out in biopsy tissue obtained during the initial prostate cancer diagnosis. It would also be interesting to look at patients who have been treated with radiation only, to see what the outcomes are after radiation (again, on biopsy tissue).