In the trial we randomised 419,000 men from approximately 600 practices, general practices across the UK. 300 general practices were randomised to the intervention arm and 300 general practices were invited to the control arm which was no screening. In the intervention arm the practices were asked to invite all men aged 50-69 to come to the practice and to receive counselling about having a PSA test and in the control arm that was just standard NHS practice which is effectively no screening. So those are the two groups we compared, approximately 200,000 men in the intervention arm, 200,000 men in the control arm. We followed up those men for ten years to look at their outcomes and the main outcome we were interested in was prostate cancer mortality.
How did the screening intervention affect prostate cancer mortality?
The intervention did what we wanted it to do: it increased the detection of prostate cancer. So for every thousand men in the intervention arm we identified 43 prostate cancers compared to 36 prostate cancers identified in the control arm. So the intervention increased the detection of prostate cancer by seven in every thousand men. However, despite the increase in the detection of prostate cancer after ten years of follow-up the death rates from prostate cancer were identical – three per thousand in the intervention arm and three per thousand in the control arm. There was a statistically significant increase in the detection of prostate arm in the intervention compared to the control arm.
What grade of cancers were found in the intervention arm?
A large number of prostate cancers identified in the intervention arm were of the low grade type – Gleeson six or less. What we were hoping to do was identify more of the aggressive type prostate cancers, Gleeson seven or above, in the intervention arm however that didn’t actually transpire.
What are the treatment recommendations for someone with this low risk cancer?
What we did do in this screening trial is we had an embedded treatment trial because in reality no-one knows the best treatment for low risk prostate cancer. Should you treat it at all? In other words, is it better just to actively monitor men? So one of the treatment arms in our nested treatment trial, which was called ProtecT, was to actively monitor men with repeated PSA testing over time. The other intervention was radical prostatectomy and the other intervention was radical radiotherapy. So nested within our screening trial was also a treatment trial in which we tried to identify what would be the best treatment for screen detected prostate cancer.
Are these results available?
Those results have been published, so that nested trial was called the ProtecT trial, those results were published in September 2016 in The New England Journal of Medicine. What they showed after ten years was that in men with localised prostate cancer the risk of prostate cancer mortality was extremely low. In all three arms only 1% of men died of prostate cancer and there was no difference in prostate cancer mortality whether you had active monitoring, whether you were randomised to radical prostatectomy or whether you were randomised to radical radiotherapy. There was a suggestion of a halving of the risk, however, of prostate cancer metastases if you had radical treatment compared to active monitoring and what we’re now doing is following up the men longer term to see if longer term there will be a difference in mortality when you compare those treatment options.
What about the missed cases?
What we found with PSA testing is not only did we identify low risk prostate cancer that was never, ever going to harm men in their lifetime but we also found actually that the intervention missed lethal prostate cancers. So we have both false positives – those men who will never be affected in their lifetime, but the other problem with PSA testing is the false negative rate. So even in men with a low PSA we found that those men with a PSA less than the threshold that you would normally investigate further we still found approximately 68 deaths in men who had a low PSA in the intervention arm out of a total of around 550 deaths from prostate cancer.
What would you advise people to do based upon these results?
Our message is that we are very much in favour of the current NHS guidelines. These are published and they are online as part of something called The Prostate Cancer Risk Management Programme. Under The Prostate Cancer Risk Management Programme within the NHS all men over the age of 50 are entitled to ask about a PSA test, go to their GP and talk about it. What we’re saying is that men should really be aware of both the potential benefits and the potential harms of PSA testing.
So our strategy to reduce what we call over-detection, that is detecting too many prostate cancers which would never harm a man in his lifetime, was to do a single PSA screening. There has been a trial in Europe that suggests if you do multiple PSA screening, for example every three or every four years, there may be a small reduction in mortality. However, that is offset against the potential harms from radical treatments such as impotence or incontinence. So our advice is that men can go and ask their GP for a test if they wish but they should really discuss the potential benefits and harms. They should be aware that a single screen is unlikely to work, that if they have repeated PSA testing that may produce a small reduction in mortality and they will have a smaller risk of dying from prostate cancer, but offset against that is a risk of impotence and incontinence.
Based on this research, what should we be doing to improve?
What this does show is we do need new technologies, new biomarkers, for prostate cancer, particularly biomarkers and technologies that can identify those cancers that we call more aggressive, those cancers that will potentially progress and harm a man in his lifetime. The advances in those technologies include better imaging. So, for example, there have been recent studies looking at the value of something called multiparametric MRI scanning. Multiparametric MRI scanning prior to a biopsy may be able to identify more aggressive prostate cancers and actually may make biopsy itself a safer procedure by allowing the urologist to target what tissue he takes and target the more aggressive looking prostate cancers. So that is definitely an advance and may lead to better screening in the future and better outcomes for men. There are other advances that are being looked at, they are still under investigation but they include genetics and protein biomarkers. Some of those genetics and protein biomarkers are promising.
Other potential advances to make PSA-based screening more effective for reducing mortality but with less side effects are to do something called risk stratified screening. So, for example, one option being discussed is to have a man have a baseline PSA test at a young age, say 45 or 50, and then use that to decide whether the man needs screening at all in the future or to decide the frequency of PSA testing. Now, all those technologies and all those biomarkers and risk stratified screening, they’re all being looked at currently. It may be difficult, actually, to do another trial of such modalities but one thing we are doing is we are contributing our data to large international efforts at modelling the potential effects of using those new modalities, so using statistical modelling but using our real life data to try and understand what would happen if we used those modalities in a revised screening programme.
What would your take home message be?
I think it’s really important that men in the UK understand that prostate cancer is the second leading cause of cancer death. There are now nearly 12,000 men dying a year of prostate cancer and there is absolutely no doubt that we do need better methods of detecting those cancers that are aggressive and that will progress to cause a man harm. PSA testing, unfortunately, on its own is a blunt tool. It, of course, will identify some men with more aggressive prostate cancer but we know that from other studies that even if you do repeated PSA testing and you do reduce mortality for every one man whose life is saved approximately 26 men will be treated for prostate cancer who would never have needed that treatment and they will have to suffer the consequences of that treatment which includes side effects such as impotence and incontinence. So the bottom line is we do need some form of early detection of aggressive prostate cancer because it’s an important cause of death. PSA testing is a blunt tool and we really now need to focus our efforts on identifying better methods of detecting those aggressive prostate cancers that do need treatment.
