PSA test

Why men shouldn't rely on the prostate cancer test - by the doctor behind its discovery

By Dr Richard Ablin
Last updated at 8:05 AM on 5th May 2010

While I was working in cancer research more than 40 years ago, I was looking for a clue which might lead to a new treatment for cancer. That's when I discovered a unique protein. This was what is called an antigen  -  and this antigen is found only in the prostate gland.

But frankly, I was a bit disappointed. I'd hoped to find an antigen that occurred only in cases of prostate cancer. This one appeared even when there was no cancer.

Now, ironically, that protein  -  the prostate-specific antigen (PSA)  -  is world-famous and has become the basis of the principal screening tool used for prostate cancer, which every year kills around 10,000 British men.

But if you think that I revel in my discovery, you're wrong. In fact, I believe testing for PSA in healthy men has been a huge waste of money and led to many men  -  particularly in the U.S., where there is an unofficial national screening programme  -  being left unnecessarily frightened, frustrated, impotent and incontinent.

Had it not been for the test, yes some lives might have been lost. But, woefully, due to the PSA test, many, many more cancers have been detected that have not merited any treatment.

The majority of these patients  -  an estimated million men  -  could have lived painlessly with their tumours until the natural end of their lives. Instead, they've undergone pointless and painful medical therapies from surgery and radiation to even possibly chemotherapy.

Of course, I am not against treatment for prostate cancer. In 1970, when I identified PSA, I was working with two urologists in Buffalo, New York, on cryosurgery  -  using freezing techniques to kill off malignant cells in the prostate.

In the process of analysing patients' responses to cancer and cryosurgery, I noted that with both treatments the amount of PSA present in the blood fell  -  suggesting the therapy was working.

During monitoring, the patients' levels went back up, which suggested the cancer had not been completely eradicated. So in 1986, the Food and Drug Administration (FDA) approved the subsequently developed PSA test as a 'harbinger' for recurrence, deciding it could be used for monitoring or prognosis.

In other words, the PSA was a tool for checking the effectiveness of treatment  -  if levels rose, it meant the patient needed more therapy. But then several urologists saw that this non-invasive blood test for PSA could offer patients an alternative to the more traditional rectal examination.

This is, naturally, not popular and its degree of accuracy in detecting prostate cancer in the early stages is low.

In 1987, a landmark paper appeared in the New England Journal of Medicine which claimed the PSA test could be used to detect cancer. This caused a storm in the U.S.  -  everybody began using it. As one colleague put it, PSA testing was implemented with a fervour which would not disgrace a medieval inquisition.

Of course, I was worried by the paper and this new passion for PSA testing, but there was little I could do to stop the flood.

The numbers of prostate cancers detected and treated with radical prostatectomies nearly trebled in the U.S. from 39 per 100,000 to 90 per 100,000. Despite this being a major operation which can carry a 30 per cent chance of impotency and a 60 per cent chance of incontinence.

In 1994, the PSA test was officially approved specifically for diagnosing cancer.

What a mistake. The FDA never evaluated the benefit versus the risk of PSA testing.

And, as I'd discovered 24 years before, the antigen was not cancer specific. It's not just present in men with cancer, but also in those suffering from prostatitis (often caused by an infection of the gland), or benign prostatic hyperplasia (BPH), a fairly common condition affecting men of 40 and older.

Yes, if there is an extremely high level in the blood something is going on and needs looking at. But lower levels could easily indicate the other two diseases of the prostate.

Part of the problem is that when PSA screening was initiated, the figure set by which you pass or fail (rather like the figure set for 'healthy' levels of cholesterol) was arbitrary, as no one knew what the correct amount was  -  or even if there even was one.

In the end, officials set the pass mark at four nanograms per ml. If a patient had a figure of four or lower, the likelihood (they said) was that he didn't have cancer. If it was above 4, there was a strong possibility he did and follow-up leading to a biopsy was in order.

The pass mark struck me as foolish and confusing. In fact, data swiftly showed that of men who had a score between four and 10 ng/ml, 80 per cent had benign prostatic hyperplasia, not cancer.

Of those who scored less than four, 40 per cent actually did have cancer. So healthy men were being tested and becoming terrified. There was mass over-diagnosis, yet many men with cancer were being reassured they did not have it  -  and were being sent home.

All this because they were using a test  -  developed to check the progress of the disease  -  as a diagnostic tool. Another issue was the enthusiasm of the pharmaceutical industry, which had vested interests in increasing the number of men found to have prostate cancer who could then be treated.

It is not unreasonable to say the disease was hyped by those who stood to make money from drugs.

A problem I also felt was never adequately addressed was the issue of severity  -  we know that some prostate cancers are more aggressive than others.

For example, most men who live to their 80s will have some degree of prostate cancer, but it may not matter  -  their cancers will never become symptomatic, let alone life-threatening.

Sometimes these cancers are called 'cats' and 'tigers' to distinguish them  -  but I think of them as 'turtles' and 'rabbits' placed in a box. The turtles crawl around the box going nowhere. But the rabbits can jump out of the box at any time. You can't ignore the rabbits, but it doesn't make sense to treat the turtles.

Yet, if I say you have detected cancer due to having a PSA test at an age when you are at risk (over 50) that pointed you towards a biopsy but there may not be a need to treat it, people look at me as if I'm crazy.

And while I am well aware the word 'cancer' is perhaps the most feared word in any language, cancer of the prostate is unique. Prostate cancer is an age-related disease.

For example, if you take samples from men who don't have any symptoms aged 60 to 70, 65 per cent will have cancer. But is the cancer a 'turtle' or a 'rabbit'?

Instead, I advocate watchful waiting. In practical terms, this means getting a PSA test, then getting another periodically, for example every six months. Then you can follow the trend and if you're not seeing an increase in PSA in the absence of symptoms, then there is no point in a biopsy.

I've tried to tell people that they're using this test the wrong way. I find it very sad that my discovery is used this way. Billions have been spent promoting it for something which it can't do, when that money could have been spent on areas which might have led to a better test.

We know PSA screening, at least in the manner it has been used, is not the answer  -  studies in the U.S. have shown that there is no difference in the death rate between those screened and those not screened. Studies from Europe last year showed that we're mistreating 48 people just to save one life.

But I do sense a sea change. The American Cancer Society has suggested that doctors should make their own minds up about screening. In the UK, you are already much more judicious.

Those who want to monitor PSA levels in their bloodstream could ask their doctor to do so every six months or on an annual basis, particularly if they have a family history of prostate cancer.

So I'm not saying PSA testing does not have a use. But asking it to detect cancer is not what it is fit for. It's time to stop trying to make a silk purse out of a sow's ear.