"A Cause for Action: The Time for Change"
Tuesday 9th November 2004, The Congress Centre, London
Dr Chris Parker, Senior Lecturer & Honorary Consultant in Clinical Oncology, Institute of Cancer Research & Royal Marsden Hospital
Professor Hamdy, Professor of Urology, University of Sheffield
Vivienne Parry: Now we come to another research double act, as it were. This time, the theme is "Early Intervention" and I am going to welcome this Morecombe and Wise of early intervention - Dr Chris Parker who is a clinical oncologist for cancer and researcher at the Institute of Cancer Research and Professor Freddie Hamdy, who is Professor of Urology at Sheffield University and the principle investigator on the ProtecT study is going to speak second. So Chris.
Dr Chris Parker: Thank you Vivienne. I hope I am Wise which makes Freddie, Morecombe. "What Benefits of Early Intervention?" obviously attracts enormous controversy and very often strong feelings. I am going to try to separate out the factual from the controversial by briefly talking to you about three truths about prostate cancer that are generally accepted with very little debate and then go on to the more controversial aspects [Slide 2]. So the first truth is the harder you look, the more you find. So if you look back to the 1970s and 1980s, the lifetime risk of being diagnosed with prostate cancer for a man in the UK was around about 6%, give or take [Slide 3]. And nowadays, in those countries where PSA testing is common and where a high PSA leads to a biopsy, such as in the United States, the life time risk is now close to 20% [Slide 4]. An almost three fold increase. This is really no surprise, because we have known for a long time that in elderly men at post mortem - men who have died from heart disease, strokes, or whatever - the large majority of them had evidence of prostate cancer sitting there in the prostate which hadn't caused any problems. Now in the past we had found just the 6% of men in whom the cancer did cause a problem. Now with PSA testing, we can find 20%.
And that brings me on to the second truth which directly follows from that, which is that prostate cancer is not what it used to be. And what do I mean by that? Well, in the pre-PSA era, we found prostate cancer when it caused problems. When it caused symptoms. That might be pain in the bones from secondaries and very often the cancer ultimately proved fatal. Now with PSA testing, with early diagnosis, those same cancers can be diagnosed maybe 10, maybe even 15 years earlier. But perhaps even more important, we are also discovering many cancers which would never been discovered in the past, because they never were never going to cause any problems. And that's what I mean when I say that prostate cancer is not what it used to be. So screened, detected, early prostate cancer today is a completely different animal from prostate cancer that was diagnosed because of symptoms 20 years ago. This is illustrated by these figures [Slide 8]. On the left you see data from Albertsen on watchful waiting of cancers that were diagnosed in the 1970s and 80s and for men in their late 60s, with a Gleason score of about six, around about 1/3 of them died of their disease. This is without surgery, without radiotherapy, without any curative treatment at all. About 1/3 of those prostate cancers proved fatal.
Nowadays, we know that the outcome is going to be much better than that. We don't know how much better, because we haven't had the time to find out. But it has been predicted by Nicholson and Harland that if you take men of the same age - same Gleason score - and follow them without any curative treatment at all, it is predicted that fewer than 10% of those men would die of prostate cancer - the majority dying of other causes. Prostate cancer is not what it used to be.
And so it follows on from that, that the treatment can be worse than the disease. Now, Roger showed you this slide earlier on which demonstrates that there is no doubt whatsoever that some men benefit from curative treatment of prostate cancer [Slide 10]. That is not in dispute. But if you have been found to have an irrelevant prostate cancer, you can't possibly benefit from treatment. And there is of course a price to be paid for having treatment. And so the Scandinavian data shown here just illustrates the risks of erectile dysfunction and urinary incontinence associated with radical prostatectomy compared with watchful waiting [Slide 11]. Now that may be a price worth paying if you are treating a potentially life threatening cancer. But, on the other hand, if you have got an irrelevant cancer, it is most definitely not a price worth paying. The treatment can be worse than the disease.
Okay, so that is relatively the easy factual bit. Nobody really disputes that to any great extent. Now here we come on to the more tricky, controversial aspects. Who needs treatment and who doesn't? So a man diagnosed today with PSA screen-detected early prostate cancer faces a dreadfully difficult decision. Does he put his care in the hands of a surgeon, have a major operation with the risks we have talked about, because he may well have a life threatening cancer? Or on the other hand, does he choose to have no treatment? This is where active surveillance comes in. It is a response to that dilemma. What it aims to do is to individualise management by targeting surgery or radiotherapy to those who men who need it and avoiding unnecessary treatment in those men who don't. It is done by a combination of frequent PSA testing and repeat prostate biopsies and using those results to target treatment.
So we have put a lot of importance on the rate of rise of PSA. And here are PSA results from two of my patients who have been on surveillance for the last couple of years [Slide 15]. The one started with a PSA of 5 which has risen to about 7 over the last year. The other with a PSA of 10 which has edged upwards only slightly in the past two years. What we do is we plot out the log PSA against time and then draw a straight line through the points [Slide 16]. It is the slope of that line which we take as a guide to prostate cancer behaviour. We come up with a figure known as the PSA doubling time: the time it would take for the PSA to double if these cancers were left alone without any treatment. You can see the man with a PSA starting off with 5 has a doubling time of four years. Whereas the chap with the PSA of 10 has a doubling time of more than 50 years.
Now in my book, it is the chap with the lower PSA, the shorter doubling time who has the tiger - who has the significant prostate cancer and he stands to benefit from surgery or radiotherapy. Whereas the man with the higher PSA and the longer doubling time appears to have a pussycat and he will be better off left alone without any treatment at all. That's the theory.
The most experience to date with active surveillance comes from Toronto. The details aren't important [Slide 18]. I just want to illustrate that they have now published their early results on over 200 men with carefully selected favourable screened-detected prostate cancers and they have targeted treatment according to PSA trends and benefit findings. For what it is worth, this seems to be a feasible approach and, four years in, 2/3 of men have avoided the need for any treatment in this way.
What perhaps is the most interesting of all is if you look at the distribution of PSA doubling times in those 200 men who have been on surveillance without treatment and you can see enormous variation. Now all these men have favourable cancers. Low Gleason scores, low PSAs, low stage and yet they behave in extraordinarily diverse ways. You've got some men with short doubling times, just 2, 3, 4 years. Significant cancers that need treating. But you've got many men with doubling times in the decades or even longer. Now I suggest that these cancers are better off left alone.
Another point to make about active surveillance: we do active surveillance because we think that it is in the patient's best interests in the hope that they will avoid unnecessary treatment. But it is also a wonderful setting for research to try and find new, better ways of distinguishing between tigers and pussycats. So at the Institute of Cancer Research, we have now got over 200 men who are on active surveillance. We are collecting PSA doubling time and biopsy information to see how their cancers behave and then we are testing their prostate tissue, their blood, their urine, for all sorts candidate markers that we hope - some of which we hope - will be able to tell us how to distinguish tigers from pussycats. I am very excited about the future in that way.
How not to treat prostate cancer? I have been described as an expert in how not to treat prostate cancer which I have always taken as the compliment that I am sure it was intended. But the point I want to make here is that there is more than one way of not treating prostate cancer [Slide 21]. In particular, what I have been describing to you is active surveillance, where we take young, fit men who are suitable for surgery, suitable for brachytherapy, suitable for radiotherapy, and we monitor them very carefully to see whether they need that treatment or not. And that treatment is curative. And it is completely different from the way that prostate cancer was not treated for decades. So older men with short life expectancies have been managed by something called watchful waiting which is essentially doing nothing unless symptoms of progressive, advanced prostate cancer develop and then they are treated gently with hormone therapy. Now I really don't mind what these two polices are called, but I think it is tremendously important to distinguish them from each other, because they are chalk and cheese. One is curative, the other is palliative. They are different. I think that is a point that is not emphasised as often as it ought to be.
In the future, there may be new ways of not treating prostate cancer. We had a question about that a few moments ago and I am very excited about active surveillance as a setting for testing nutritional interventions. So the idea is that you take men with early prostate cancer who have chosen to be followed by active surveillance and then you evaluate the effect of nutritional interventions, diet or whatever in terms of the effect on PSA doubling time, in terms of the effect on repeat biopsies. This is a trial design which we are trying to get developed at the moment which would involve 500 men and would give answers in the space of three or four years. So it is something that is rapid and efficient when one compares it with the massive undertakings which have been taken in the past to try to evaluate nutritional interventions.
And lastly, to screen or not to screen? It is obviously highly controversial and I am not going to get into the question. Like John, I am a coward, but what I do want to say is that if active surveillance were accepted as the norm for men who were diagnosed with screen detected early prostate cancer, then it would change the context for the screening debate because the main argument against PSA screening is over treatment. It is the impotence and the incontinence that is caused by treatment which is worse than the disease. It is finding cancers which are irrelevant as well as ones which are significant and then treating them. So if surveillance were the norm and the pussy cats were left alone and the tigers treated, then PSA screening becomes a more attractive proposition than it is at the moment. Thank you.
Vivienne Parry: I can feel a lot of questions out there, but just hang on to them at the moment and we will come to them at the end of Professor Hamdy's speech.
Professor Freddie Hamdy: Thank you very much ladies and gentlemen for giving me the opportunity to speak to you. Chris has set the scene extremely well. We are going to talk about screening and other modalities of treatment of prostate cancer and who gets any benefits from it.
Well let's start with a simple question. The first one - and I am grateful to you Chris for preparing this: "Does screening reduce mortality from prostate cancer?" I am just going to show you these two graphs [Slide 3]. And in these two graphs what you can see on the left hand side is the incidence of prostate cancer in North America - at the top with the see-through circles versus England and Wales at the bottom. And you can see that in the late 80s / early 90s there was a peak of prostate cancer incidence which was brought about by the introduction of PSA testing which was not adopted in the UK at the same speed. And therefore the incidence rate in the UK stayed relatively stable whilst it increased in the US, coming down later on because all the back log of prostate cancer had been cleared. Now if you look at the right hand side graph you will see that the mortality in the US started to drop in the early 90s, mid 90s - and the North Americans started to think that this had happened because of the PSA screening and because of very aggressive treatment of early disease. Then some smart epidemiologist from Bristol called Stephen Oliver published this paper in the year 2000 looking at the mortality from prostate cancer in England and Wales and he found that it was also dropping in England and Wales and I can tell you that it was not because of radical prostatectomies. In 1994 there were 37 radical prostatectomies performed in the UK. Now we are well over 1,000, possibly reaching 2,000.
So what does this tell us? It tells us that yes, there is a reduction in mortality, there is an increase in incidence, but the reduction in mortality is not necessarily caused by treatment. It may be caused by other factors including environmental, hygienic and dietary.
Look at this very troublesome graph and curves showing you the mortality from prostate cancer in various parts of the world [Slide 4]. Where you can see the green circles these are western countries where the prostate cancer mortality is dropping and the other countries it is either stable or going up. There is absolutely no epidemiological rationale to explain why in certain countries of Europe the mortality is coming down and in others it is going up. The practices are very similar everywhere and there is no clear explanation for this phenomenon.
So where are the dilemmas [Slide 5]? Again, Chris has stated them, so I will reiterate. The first one is that there is a very significant risk of over-diagnosis if one tests with PSA. The recent prostate cancer prevention trial figures show that at very low levels of PSAs - 0 to 4 - and Roger emphasized that this morning, you have an incidence of prostate cancer of about 17%. But what cancers are we picking up? Are they worth picking up? Are they going to cause any harm? We simply don't know.
And then there is the significant risk of over-treatment. Let us just work out through a simple calculation. If you take a million asymptomatic men who agreed to receive a PSA test, approximately 100,000 will receive biopsies, 20,000 will have cancer, 10,000 may receive a radical prostatectomy, 10 will die, 300 will have some degree of incontinence, and 4,000 will have erectile dysfunction. If that is the game we want to play in the absence of evidence that this makes any difference, well, that becomes a difficult choice.
And the final dilemma is that there is a lack of Level 1 evidence from randomised controlled trials that aggressive treatment of clinically localised disease improves survival and / or quality of life in these patients.
There are ongoing surveillance, screening and treatment studies in the US, Canada and Europe, and we really ought to await the results of these trials before making recommendations about screening as a public health policy.
The next important question in the next five minutes - "Does radical treatment improve survival in clinically localised prostate cancer? What is the ideal treatment?" It cures everybody, it is given as an outpatient, it has no side effects and it costs nothing. But like the ideal wife or the ideal husband, the ideal treatment does not exist.
Now, we are always obsessed with effectiveness and we are very much prepared to sacrifice cost and side effects because effectiveness is so obvious, so overwhelming, therefore we must have it. But we must also be careful that effectiveness does not perish in front of cost and side effects.
We can study men with prostate cancer with low, intermediate and high risk. But we know that the lowest patients will do well whatever you do to them [Slide 9]. The high risk patients will not do well whatever you do to them [Slide 10]. And I have always wondered, with such an abun dan ce of options for patients with any prostate cancer which are on the shelf, how does one get a patient on the operating table to do a radical prostatectomy today? I found the answer. And the answer is it is the surgical imperative. And what does the surgical imperative say? It says: "If a surgeon tells you that without an operation you will die, as that as a result of the operation you may die, you have no choice but to have the operation". What is very interesting is who actually made this quote (Adolf Hitler).
Now we want to prolong life. But let's go back to some philosophy and let's define what life is. Life - and this is probably the best definition - is "A sexually transmitted condition with 100% mortality". So we are looking for quality of life and, yes, the quality of life may be removing the prostate but the question we have to ask - Who is the quality of life for? - is it for the patient or is it for the surgeon? And unfortunately, ladies and gentlemen, this boat is not mine and the ladies on board are not my nurses [Slide 13]. And really the dilemma can be summarised by the fact that there is absolutely no Level 1 evidence from randomised trials that either screening or treatment of screen detected cases prolongs survival or improves quality of life.
So how are we addressing this in the UK ? You have heard the word ProtecT mentioned several times. It is in the information booklet, the Progress Report from the NHS that you have in your hands. It is a randomised controlled trial of testing and treatment in early prostate cancer. It is taking place in nine regions of the UK spreading from Scotland down to the Southwest and the Southeast [Slide 15]. The study design is simple [Slide 16]. There was a call out from the Health Technology Assessment Panel of the NHS R&D programme back in 1997 / 1998 requesting ideas and proposals for trials on prostate cancer. We submitted a Phase I proposal for the feasibility study of a randomised controlled trial. The reason we asked for a feasibility was because there was a misconception that such a trial would not be possible. In other words, it would not be possible to diagnose young men with prostate cancer and not to treat them and to put them on an active monitoring arm.
So the feasibility was about test finding - What is the best way of recruiting the men and piloting the conduct of the main trial? It was successful. We then engaged into Phase II trials which will last until 2006 to conduct a major three arm randomised trial and it is about comparing active monitoring with radical prostatectomy with radical radiotherapy. The end point is very simple. It is survival at ten years from the date of the diagnosis.
So what have we learned from the pilot study [Slide 17]? This really has set a track record in terms of conducting very large multi-centre, important randomised controlled trials. Pilot studies are extremely helpful and useful. We've learned first of all that it is possible to undertake the main trial in the UK with good randomisation rates. We've learned that primary care is the ideal setting for these men to be seen by trained nurses. That equipoise - which means a frame of mind where the three options which are offered to the patient, weigh the same - exactly the same - for the recruiter so that he can sell the ideas to the patient. Unless there is this equipoise, even body language will put off a patient from being recruited and randomised.
Trained nurses provide excellent counselling and they do the bulk of the work. I am very pleased to see two or three of them sitting in the back here, because without our nurses, this trial would not exist.
Three-arm randomisation is the preferred option. Training of the recruiters is essential and monitoring of the recruiter is equally important.
What is it? It is about finding out what is best. You'll have to guess where Roger is, where Malcolm is and Chris is. I would hate to guess, but it is all about finding what is best for whom. These are our case finding targets by 2006 [Slide 19]. We want to invite 233,000 men between the ages of 50 and 69. 50% to 60% of them will attend and we will collect approximately 3,000 cancers. And this is what we have achieved by September 2004 [Slide 20]. We have sent out 85,000 invitations and 42,000 attended. We have got so far 1,100 cancers: of those ¾ are localised. It is likely that will not be able to complete recruitment by 2006 simply because many of the new centres have started late rather than early, so we are asking HDA to extend the study by 24 months.
The randomisation to June 2004 [Slide 21]. O ut of 712 cases we got an average of 60% to 70% randomisation and of those, 95% have elected the three arm study which means that the two arm has now gone out of the protocol altogether and the rest of the patients have selected their own treatment. Music to the ears of Chris is that the majority have elected active monitoring.
Now just to finish, the Department of Health have asked us to consider alongside the issue of treatment, to test the issue of screening. And because we have an ideal comparison on, and that is men who are not approached by the ProtecT study to take part, we are now going to be able to address this issue at the same time. So effectively this converts the ProtecT study into the intervention arm of a randomised controlled trial of screen which will answer both questions [Slide 22]. Out of 800 practices in all the nine centres, half will be randomised to ProtecT and the other half to no intervention at all. These will be left for men to be managed using standard NHS practice and the others you have heard will go through the ProtecT process.
The primary outcomes will be mortality at ten years and hopefully we will have some answers which are much awaited by the Department of Health and other countries about whether screening and treatment makes a difference in prostate cancer.
None of this work can be done by one or two, or three persons. I am ever so grateful to the growing group of people who are heavily involved who spend so much time dedicated to this study and to the Department of Health and HDA for spending in excess £14 million in making this happen. Thank you very much indeed.
Vivienne Parry: Thank you very much indeed for that.
Delegate: It's all very sad. I'm afraid they presented too late. How do we get round the issue without screening?
Professor Freddie Hamdy: I think that is a very good question and I can tell you what I think testing screening will reveal. It is probably going to reveal that - this is speculation, this is a hunch - that in low risk disease, small volume disease patients, it is unlikely for us that we will show much of a difference if we were to treat these patients aggressively. There is a very interesting example in Austria where there is a Tyrol study called the Deholsten. It is not a randomised control one. It is an observational study of screening. So if you are on holiday in the Tyrol and you go and buy a newspaper and you get offered the PSA test and if you say thank you, you get offered the radical prostatectomy test. They have been doing this since 1993 - so for the last ten years. What they have observed is that in 1993 when they started screening, 1/3 of their patients had clinically localised disease. Four years later, by 1997, they found that 2/3 of the men had clinically localised disease and they also found a difference in mortality. And they have misinterpreted the reduction in mortality to the fact that they were treating clinically localised disease, whilst in fact what they were doing was throwing the net in the sea picking up the fish - some small fish but a lot of big fish. The big fish are the locally advanced disease that were not presented yet and these patients were treated immediately and therefore there may be indirect circumstantial evidence that if locally advanced disease is detected earlier than presenting with symptoms, we may be affecting the mortality. I think that stage migration - this is what we call it - stage migration is going to go in favour of screening, but not for the small cancers, not the localised cancers. Probably for the more advanced, but it is speculation. I don't know if this has answered your question.
Vivienne Parry: Chris do you want to add to that?
Dr Chris Parker: I like to liken screening to the lottery. Because everybody pays for their ticket and very few people win the jackpot. Somebody who is in fact harbouring a nasty prostate cancer would win the lottery if he were screened. I believe it could make the difference between life and death. But I think the price of entering the lottery is the risk of running with incontinence and impotence from a cancer which in fact would have been better off if you had not been diagnosed. So that is the balance.
Vivienne Parry: Gentleman over here.
Delegate: Thank you. Really this is addressed to Dr Chris Parker who has given us a breath of fresh air in the patient-led prostate cancer support associations. We have always traditionally, for the sake for men who have no symptoms, pressed for PSA testing. We have in our naivety always assumed that active surveillance would then be applied and that you experts would get on with the business of sorting things out and advising us, but we know to our cost and I hope that Professor Mike Richards is listening to this, that active surveillance really does not necessarily go on. So we really have to come down to earth, don't we, somewhere? Now breast cancer screening is not much more predictive or it may not be as predictive as PSA testing and digital rectal examinations. But it goes on. On the face of it, it saves lives. It does provide information to those who are likely to suffer from it. So let's have bags of active surveillance after PSA and digital rectal examinations.
Dr Chris Parker: I would just say that the ProtecT trial is going to be enormously important in comparing active surveillance against immediate treatment and until then men are faced with making a decision without complete evidence. In my experience, active surveillance is more popular amongst patients than it is amongst clinicians, but that is not on complete evidence. So I think that the ProtecT trial is tremendously important.
Professor Freddie Hamdy: If I may, I will make a comment and I only heard this two weeks ago at a WHO Consensus on prostate cancer. I heard much to my horror that only 3% of men with prostate cancer in the whole of the European Union are entered into clinical trials. I think this is an unacceptable figure, particularly in view in the dilemmas and uncertainties which we face with this disease. We need to have many, many more men entered in good quality, well managed trials.
Vivienne Parry: I am going to take one more question, for I know that I am going to be shortening tea.
Delegate: If more men agree to randomised testing, what do we say to those men who are given placebos and find that their cancers have grown to a mortal rate, while there is a protection against this, I am aware, and I think you should make it clear.
Professor Freddie Hamdy: Let me try to understand. Is your question about the men who know that they have cancer and do not receive immediate treatment or is it about men who are not offered PSA?
Delegate: Well, it could be either. It could be someone who has not been tested.
Professor Freddie Hamdy: Well it is very different. We are not preventing the men who are not tested from being tested, but we are not offering the testing as part of the trial, which is a different matter. We know that the rate of testing in the UK for PSA is 6%. If you look at Germany or Austria it is 65%. So we have a very low rate of testing anyway which makes the UK a very fertile ground to be able to do these studies. We have very little contamination which is what we call it. So that is for those who have not been tested. So those who are the ones who are tested and have cancer, we don't call it surveillance, we call it monitoring. The reason why we have moved away from watchful waiting is the very sterling work that Jenny Donovan has been doing in Bristol with qualitative research listening to every conversation that we have with the patient. Listening to the reactions of the patients and watchful waiting for many men meant you watch, while I die. We had to change that because it was the wrong terminology. What we are doing is not watchful waiting. What we are doing is active monitoring. So they have a very rigid programme of active monitoring and we want to try and maintain them within the curability window. We don't want to lose them out of the curability window, but we expect that at least 80% of these men will stay within the curability window well beyond ten years.