Effect Measure

Another story about a “new” screening test, this one for prostate cancer, the most common cancer in the US. This one looks for a particular combination of variations in five genes. If a man has all five and a family history of prostate cancer then his risk of is increased by a factor of nine. The researchers who have devised the test have also patented it and plan to sell it for $300 bucks a pop. I, for one, am not buying it (literally or figuratively). First some of the details as given by a press report:

Almost half of prostate cancer patients carry five genetic variations and a family history of the disease, says a breakthrough study in the search for a root cause of one of the most common and deadly tumors.

Having the combination raised the chances of getting the disease 9-fold, according to a study of 4,000 Swedish men published today by the New England Journal of Medicine. While each of the variations increased the cancer risk a small amount, putting them together gave the researchers the power they needed to generate a substantial and consistent finding, they said.

Researchers led by Jianfeng Xu, a professor of cancer biology at Wake Forest University School of Medicine in Winston- Salem, North Carolina, are working with a company, Proactive Genomics LLC, to sell a $300 test based on the findings. (Bloomberg)

I’ve already had a prostate screening test, the prostate specific antigen (PSA) blood test. In fact I’ve had a couple of them. I don’t plan to have any more. Why? Because the PSA is too sensitive and not sufficiently specific. It produces quite a few false positives in the sense that it doesn’t differentiate prostate lesions that are no clinical problem from those that are extremely aggressive and dangerous. And neither does this test. I can’t see what advantage it has over a PSA (which I find so unsatisfactory I have decided not to get any more). It’s even worse. It doesn’t say a thing about whether there is a lesion there or not, just the risk of a lesion. You’d have to have the same test (a PSA or digital rectal exam) as a follow-up.

This work isn’t worthless. If we know what genetic elements are involved they might provide important and useful clues to causes of prostate cancer. And of course it is certainly worth quite a bit to the folks who will be hawking this to the worried well. But spending $300 for no tangible benefit?

No thanks.

Comments

#1 Eric Juve
January 18, 2008

And don’t even mention needle biopsies, not comfortable.
#2 realist
January 18, 2008

revere, I hope you never get prostate cancer, or at least not before you’re really old and soon ready to die anyway.

As I understand it, the PSA test is only helpful if you have it done annually. One test can tell you nothing, only a dramatic increase from whatever your “normal” PSA measurement is indicates possible cancer.

My husband had the PSA test done annually for several years before he had an anomaly that turned out to be prostate cancer. His prostate was removed, and the cancer cells appeared to be contained on one edge of it. Now he has had annual PSA tests done (I think for 5 years post surgery) to make sure no cancer cells escaped into his body.

The real problem is that there is no way to tell if prostate cancer is slow-growing or aggressive, so the only thing you can do is to treat it if your life expectancy is more than 5 years out.
#3 revere
January 18, 2008

realist: I am quite realistic about prostate cancer. I’ve know quite a few vicitims, both professionally and personally. I’ve also had a number of PSAs and a biopsy. I have finally come to the conclusion that for this specific cancer this screening test isn’t worth it (for me). PSA aside, I don’t see the added value of this genetic test. If your personal calculus is such that it is a concern, then an annual PSA is just as good.
#4 jen_m
January 18, 2008

With the mini-GC post, that makes two posts this week about false-positive problems. I wonder what the ROC looks like for annual PSA among the Xu-profile positive?

Here’s hoping some novel non-invasive imaging techniques will come along to replace biopsies. That won’t address the stress that false-positives induce, but at least the poor guys whose PSAs rise won’t have to contend with unnecessary surgeries.
#5 PFT
January 19, 2008

These genetic screening tests are tools for the insurance industry to deny you coverage. Stay away from them unless your insurance is locked in and you never have to apply as an individual.
#6 hurtmyfeelings
January 21, 2008

You mean the PSA test is �too sensitive� in layman�s terms obviously, as high sensitivity, low percent of false positives, is a good quality for a screening test, unless it is necessarily at the expense of specificity, the percentage of correct negatives.

Aside from where global health dollars are best spent, the predictive value of the PSA test could be improved when administered to the high risk group as identified by this new test if indeed it sufficiently identifies those at high risk. However, if I understand both the inventors claims and epidemiology correctly, it looks like the new test has a sensitivity of approximately 50% and a specificity of 90%, and it would still leave many false positives, i.e. those falsely identified as being at high risk, probably only providing modest improvements in predictive value.
#7 tymbuktu
February 25, 2008

WHat about the PSA AND a DRE? Isn’t that much more reliable???
#8 James Conlon
April 9, 2008

I had a rectal exam recently together with a “super” blood test. The urologist tried to explain why his blood test was different than the one I had done a few days earlier by my MD. He said his test evaluates 2 different things whereas my MD test only had 1. I have no idea what these 2 items consist of. I am 71 years old and apparently have an enlarged prostate (who doesn’t it at my age). The urologist called me back a few days later and said that the count was 7 whereas a 4 is considered okay. He also said that he recommended a biopsy. Would that be a biopsy of the prostate? Can you shed any light on that 7 count? And the biopsy scenario? I do not want to undergo this unless clear results can be obtained. I get the impression that there are no clear reults unless one is diagnosed with cancer outright?
#9 revere
April 9, 2008

James: The biopsy is done outpatient with a probe through the rectum. No anesthesia. I’ve had it. Not super comfortable but not terrible. It does give a good diagnosis of malignancy in the prostate, although the clinical significance of some of the lower grade lesions might not be clear. Usually what they pay attention to is how fast the PSA level is going up, not necessarily its absolute value. That would require another test, say in 3 months.
#10 Anonymous
October 4, 2009

So, now digging prostate info. Isn’t it like running one step behind trying to keep up with the pros and cons of the medical world? The testing , biopsy itself raises the psa. I have seen that 3 times with my husband. He had a normal biopsy on a 4psa. Shortly after the number was higher, didn’t know at the time the biopsy itself causes the psa to rise, so another biopsy which was atypical in one area and then the number was higher after that biopsy and then the third biopsy and the same area that was first normal and then atypical is now malignant all leading to another leap with the psa number. Something about observing the experience made me curious if my husband would have been better off sitting and waiting a while with the psa 4 than all of the testing which increased the psa number and possibly caused some trouble. All of this being difficult to sort, especially noticing so much talk about the accuracy of the psa to begin with and the possibility of causing more harm with biopsy, etc. Much of what I read was after the fact, just alarmed into testing initially. My suggestion is to do a lot of reading, pros and cons before you jump on the table.
#11 revere
October 4, 2009

Anonymous: That’s good advice, although you will not find the answer. It depens a little on your age. Over 65 I wouldn’t bother with a PSA (this is the voice of experience talking as well as epidemiologically). I’ve been through it.