Why I’m skeptical of a test for newly diagnosed prostate cancer

american cancer societyComing to an office near you: a new test that can “confidently” predict whether or not you need to have aggressive therapy for your newly diagnosed prostate cancer.


That’s what the press reports would lead you to believe. And it’s really going to catch your attention if you’re one of the tens of thousands of men who will have to decide what to do if you are diagnosed with prostate cancer that has what we call “favorable characteristics.”

And with the test coming to market, you would assume that your doctor would have a good understanding of whether or not it works based on the available studies and information. But guess what? The likelihood of that is pretty low, because your doctor has probably been reading the same press reports as the rest of us, since the scientific studies that doctors should rely on to make decisions about this test are simply not available. But the website promoting the test is there for all to see.

Do I sound a bit skeptical? Well, maybe I am. Because if the PSA experience has taught us anything about testing for prostate cancer, it is that we should learn the evidence before we leap. And in this current circumstance, we don’t have much-if anything-to learn from outside of company press releases and promotional materials and media reports coming from the scientific meeting where an abstract (#2131) of the research supporting the test was presented recently.

Treating for prostate cancer really is an important question. This is a cancer that the American Cancer Society estimates prostate cancer will diagnosed in close to 239,000 men in the United States in 2013. About 30,000 men will die from prostate cancer this year. Many of the men who are diagnosed will be treated fairly aggressively for their cancer, either with surgery or radiation. Many of the men who are treated will have side effects from the treatment. However, only a limited number of those men will really benefit from that treatment.

The dilemma we face, both as patients and as physicians, is that we don’t know which men really need treatment. We have some crude guidance, but in fact we are holding out hope that research studying gene patterns from the cancer tissue will tell us which men need treatment and which do not. There is considerable research underway to help us figure that out, which becomes increasingly important now that virtually every reputable organization agrees that testing to find prostate cancer “early”  should be done, if at all, only when the patient fully understands all the benefits and risks of doing so.

So, what do I know about this new test? It’s based on a limited genetic analysis of prostate cancer tissue taken at biopsy. Based on a news report from a reputable source, the company that invented it is now making the test available to the public at a cost of approximately $3800.

I have reviewed the abstract and I know that two development studies and one validation study were done. I know that the validation study was done on 395 patients and concluded that the test predicted prostate cancers that had “aggressive characteristics,” such as a larger size or higher grade. I know that the research was performed at reputable institutions by highly regarded physician researchers.  The company press release claims that the test “strongly predicted disease aggressiveness offering information beyond currently available clinical factors, such as PSA and biopsy Gleason Score to help physicians and their prostate cancer patients confidently [their word] choose the most appropriate treatment based on an individualized risk assessment.”

The press release adds that the test it will help more men avoid “unnecessary treatment” and will lead to more men choosing active surveillance as an alternative to surgery or radiation.

What I don’t know is what the abstract doesn’t say: how many men with favorable characteristics in their tumor tissue had “bad” prognostic scores, and did those “bad” scores actually correlate with bad outcomes over 8 or 10 years of follow-up? (There were some statements in the media reports to that effect, but not in the abstract, and that is the only information that I can even begin to rely on). I

don’t know anything about what was said or presented in the actual presentation of the abstract yesterday and whether-as may happen in a scientific meeting-there was any discussion of the results, or anyone who didn’t agree with the recommendations or conclusions. I don’t have a publication in a peer reviewed journal that gives me access to all of the information I would need to make an informed decision whether or not to recommend this test to a patient, or even consider it myself if I was diagnosed with prostate cancer.

And if I don’t know enough based on the abstract, website, news reports and press releases to make a medical decision, neither does your physician. This is not a criticism of the researchers who reported the research. They are outstanding, well-accomplished and very well published. But if I were a busy doctor and wanted to know more, I would have difficulty getting answers to the questions that I had.

I tried to find answers to my questions by going to the medical literature for information. I will grant that it may be there, but I had difficulty finding it. I did find the abstract on line. There were also some abstracts presented at prior scientific meetings, which were cited in the company’s literature list (which itself was difficult to negotiate if looking specifically for information on this test), but no specific published journal articles came to light.

This brings us to one important point: who should get this test. The company says if you meet certain criteria-which are helpfully listed in a “quiz” on their website-you should ask your doctor about the test and together you can make a decision on whether or not you should get it. But your doctor probably knows only as much about this test as you do.

We should expect more before we take this leap, like the data in a generally accepted format. Data by press release and media reports is not data, my friends.

One other point that you may not be aware of: unlike prescription drugs in the United States, laboratory tests developed and done in a single lab — which is the case with this particular prostate cancer predictive test — are not subject to rigorous review by the Food and Drug Administration. This particular laboratory has a track record of reliability in the medical community, but that does not mean the claims made for the test are reviewed by the FDA as are applications for new drugs. I have seen situations where laboratories have made great claims for their particular test which do not stand up to scrutiny when carefully reviewed by others. I don’t know that is the case here, but then again I don’t have the type of information I would be looking for to make any assessment about the test at this time.

I must put in a word of serious caution about my comments that is very important: this might be a fantastic test. It might be the answer we are looking for when it comes to telling us what we need to know about patients who appear to have low risk prostate cancer and would be candidates for active surveillance or watchful waiting. But let’s have the data so we can make the interpretation before we blast the media and the web with all sorts of testimonials that boot strap limited studies to grand promises. Diagnostic tests and treatments that are not subject to critical review have a tendency to become self-fulfilling, such as the enthusiasm generated for robotic surgery. And based on some comments in some of the press reports, including the one referenced above, some experts agree with a cautious approach. However, those are not the comments that appear to make the headlines.

As my colleague  Dr. Otis Brawley, who is the chief medical officer of the American Cancer Society, said to me recently, faith based medicine is not evidence based medicine. After what we have been through with prostate cancer, we should have learned our lesson to demand the evidence before we demand the test.

J. Leonard Lichtenfeld is deputy chief medical officer, American Cancer Society. He blogs at Dr. Len’s Cancer Blog.

Leave a Comment

Most Popular

Get KevinMD's 5 most popular stories.