The dangers of prostate testing

Research published in the USA this week casts doubt on the usefulness of the PSA test (prostate specific antigen) in preventing deaths from prostate cancer.

In a carefully designed study (1), researchers from the Department of Veterans Affairs looked at a group of over 500 patients who had died of prostate cancer, and compared their records with a similar group who had been diagnosed with prostate cancer but had not died from it. They found that those in the first group were just as likely to have been screened for prostate cancer using the PSA test as those in the second group. In other words, PSA testing conferred no survival benefit for these patients.

The debate about the pros and cons of PSA has raged since its introduction in the late 1980s, and this latest research does not provide all the answers. Another recently published study from Canada (2), with a similar design, showed that PSA screening might confer a moderate protection against the development of metastatic prostate cancer (that is, advanced cancer which has spread to other parts of the body) although not necessarily affect overall mortality.

In the long run, big, long-term, randomised controlled trials are required to tell us more. Two of these, involving thousands of men in Europe and the USA, will report their findings in the next few years. For other types of screening tests, for example mammography for breast cancer, such trials were carried out before the widespread introduction of the tests (although the outcomes of those trials are controversial; see What mammography misses, by Michael Baum). For prostate cancer screening, as one author pointed out, ‘the new paradigm seems to be, “screen until trials prove screening does not work, then stop” (maybe)’ (3).

The failure, despite lots of research, to show conclusive big mortality reductions from PSA screening does tell us that whatever the benefits may be, they are not enormous. And the risks of prostate cancer screening are such that it should not be entered into lightly. As usual in debates about screening, there is no such thing as a harmless test. PSA testing is harmful in two ways.

Firstly, it has a false negative rate of at least seven per cent. This means that for every hundred men who have the test, 10 will have a raised PSA. All these 10 men will need a prostate biopsy, but only two or three of them will have cancer. So seven or eight cancer-free men will end up having uncomfortable biopsies of their prostate, with potentially serious side-effects such as septicaemia.

Secondly, even more problematically, PSA testing does not differentiate between aggressive and potentially fatal cancers, and those which might have had a benign course, never troubling their host. Post-mortem studies have shown that around 40 per cent of men who die aged over 70 have prostate cancer (4). In the past, most of them would have been blissfully unaware of this fact. In the era of PSA testing, much more of this hidden and harmless prostate cancer is being picked up, so the incidence of prostate cancer has shot up (almost doubling in the USA).

One critic has accurately described this as the ‘the eradication of a disease: how we cured symptomless prostate cancer’. Formerly symptomless prostate cancer now has a new symptom: ‘a disabling state of anxiety resulting from (men’s) knowledge of their PSA level.’ (5)

This would all be fine if we had a safe, effective cure for prostate cancer. Things are advancing slowly in this area, but current treatment methods, mainly surgery or radiotherapy, are radical and potentially disabling. More over, their advantages in terms of survival in the long term are uncertain.

All this ambiguity does little to stop the relentless drive of men’s health groups like the Prostate Cancer Charity to make men ‘aware’ of prostate cancer. This includes of course the infamous ‘squish’ recorded by Ricky Gervais in September last year, as he imitated the noise of a rectal examination, in a radio advert encouraging men to visit their GP. ‘The [new American] research suggests that we should be looking in other directions for an effective test – one that detects the prostate cancers that might kill if left undetected’, said Chris Hiley, head of policy and research at the charity (6).

This may be true but is not exactly helpful to GPs and hospital specialists, who are left with the difficult job of trying to explain to patients who’ve had their ‘awareness raised’ (for awareness read anxiety levels) that testing for prostate cancer may not be a great idea.

Liz Frayn is a doctor in the south of England.

(1) Concato, J et al. The Effectiveness of Screening for Prostate Cancer: A Nested Case-Control Study. Archives of Internal Medicine 2006. Vol 166:38-43 2006

(2) Kopec, JA et al. Screening with prostate specific antigen and metastatic prostate cancer risk. Journal of Urology 2005. Vol 174:495-499

(3) Barry, Michael J. The PSA Conundrum. Archives of Internal Medicine 2006. Vol 166: 7-8

(4) Coley, CM et al. Early detection of prostate cancer: Part 1: Prior probability and effectiveness of tests. Annals of Internal Medicine 1997. Vol 126 (5):394-406

(5) Tannock, I. Eradication of a disease: how we cured symptomless prostate cancer. The Lancet 2002; 359:1341-42

(6) Quoted in the Guardian, 10 January 2006