Prostate cancer tests are great for some but bring hidden problems to many
24 May 2011
The question of prostate specific antigen (PSA) testing for prostate cancer is not whether any individual has ever benefited from a PSA test, to which the answer is yes.
The question is whether a PSA test should be recommended for all men in a population screening program, to which the answer is clearly no.
Why is this?
Simply, the PSA is not accurate enough. In particular, if a patient has cancer, the PSA can't distinguish between a prostate cancer that will never cause a problem in a man's lifetime from one that could cause his death.
A population screening test must be shown to decrease deaths. The result is that men who have PSA tests are being over-diagnosed and over-treated for prostate cancers that would never have caused them harm.
Given that treatment can result in impotence and incontinence, it is doubtful that a man would choose this treatment if it weren't life-saving.
So what are the chances?
Two large international studies were run to compare a group who were routinely having PSA tests with those who were not.
The study in the USA showed no benefit in survival from prostate cancer for those who were having PSA tests.
This study has been criticised because there were men in the comparison group of the study who had a PSA test anyway which would make a difference more difficult to find.
A large European study did find a small improvement (20%) in prostate cancer survival rates in patients who received the testing.
But for every man whose life was saved, 48 additional men had to be treated, and risk the side effects of treatment and yet would not expect a survival benefit.
These are men who would not have died of prostate cancer but still have the psychological burden of a cancer diagnosis, even if they decided against treatment.
There have been other studies reported. A subgroup of the European study from Göteborg suggested a better survival difference of over 40% over 14 years, but still with a very high rate of over diagnosis.
On the other hand, a study from Stockholm which monitored men for 20 years showed no decrease in death from prostate cancer for those men having PSA testing.
So how could you tell every man that they must have a PSA test?
Rather than population screening - where every man is advised to have a PSA - men should discuss their individual risk of prostate cancer with their GP.
Men should have an opportunity to find out the potential benefits and harms resulting from a PSA test, and make their own informed choice.
There are also suggestions that one-off testing at an age before most prostate cancers occur might be helpful but that would require large trials before being able to be recommended.
There is one further downside of continuing the PSA testing debate and that is diverting attention from the real need to find a better test.
We need to fund research in this area and follow up promising early results from tests which use genetic fingerprints to distinguish which prostate cancers need treating.
This blog post also appeared on The Conversation.