Findings suggest many patients with localized disease can delay surgery or radiation without increasing their mortality risk.
In this interview, Freddie C. Hamdy, FRCS, FMedSci, and Jenny L. Donovan, PhD, FMedSci, share the 15-year clinical and patient-reported outcomes from the ProtecT trial (NCT02044172), which were presented at the 2023 Annual European Association of Urology Congress in Milan, Italy and published in the New England Journal of Medicine.1,2 Hamdy is a professor of urology at the Nuffield Department of Surgical Sciences at the University of Oxford. Donovan is a professor of social medicine at the Bristol Medical School.
Hamdy: In the 1990s, there was a blood test which appeared on the scene, which is PSA [prostate-specific antigen]. That was allowing a lot of men who are asymptomatic to be tested and then to receive prostate biopsies. It has created a huge increase in the diagnosis of early prostate cancer. Most of them were indolent cancers, so not causing much harm, but people were not quite aware of that at the time. The treatment of prostate cancer had evolved, so surgery was reproducible, and radiotherapy was also improving. Many of these men were being given radical treatments, which also had [adverse events] and consequences.
What we realized in the mid-90s is that there was no real evidence out there that the treatments that were given made a difference to these patients in terms of survival from the cancer, had the effects of the treatment been studied properly. In a nutshell, we set up a trial called ProtecT. We did the pilot initially and then we followed full time.
Between 1999 and 2009, at 9 centers in the UK, we tested 82,429 men with a PSA and then biopsies if the PSA was raised. That exercise by itself detected 2965 cancers. Of all the ones who had disease contained within the prostate–or clinically localized disease, as we call it—1643 were then randomized to 1 of 3 options. One was active monitoring, which is a form of surveillance. The other is radical prostatectomy. The third arm was radical radiotherapy. Both surgery and radiotherapy were the conventional treatment options at the time.
Our primary end point was disease-specific mortality. There were other secondary end points, like all-cause mortality. How many patients will develop metastasis in each group? How many would progress? What were the [adverse effects] of the treatment? But more importantly, Jenny led a quality-of-life [analysis] of patient-reported outcomes. Jenny, do you want to describe that?
Donovan: Yes. We asked the men to fill in quite a long study questionnaire, containing all sorts of questions about urinary, sexual, and bowel function and quality-of-life impacts. The men were amazing. They filled in a questionnaire every year, for 12 years of follow-up. [There were] 80% of respondents across all the measures in the arms, so very committed to doing that. We've gotten results of those, which I'll present after Freddie talks through the clinical.
Hamdy: What we wanted to find out is if there was effectiveness in the radical treatments; what is the price that the patient has to pay to get that benefit? [This is] what we call the tradeoff. What is the tradeoff in quality of life that you have to think about when you decide to have radical treatment for prostate cancer?
Hamdy: We started in 1999, and patients [either] received surgery, radiotherapy, or active monitoring. Active monitoring is a form of active surveillance, where the men are followed up with a regular PSA, and then if there's any change in either symptoms or the PSA itself, that triggers a set of investigations. If the investigation suggests that there is an element of disease progression, then a discussion with the patient happened to see if they wanted to switch over to a radical treatment. Very often, it wasn't so much absolute proof of disease progression that happened; it was the patient was getting anxious or the physician was getting anxious, and then they came out of active monitoring and received the treatment.
The patients who were allocated to radical treatments, most of them received the treatments. Over the years, they continued to receive radical treatment. The patients who had active monitoring slowly were trickling over from the active monitoring arm into receiving active radical treatment. By the median 10-year follow-up, about 55% of the men who were allocated to active monitoring had received treatment. When you look at what happened over the 15 years, the men allocated to the radical arms, 90% [for prostatectomy] and 92% [for radiotherapy] had received the radical treatment. In the 5 years between 10 and 15, there was an increase in the number of men who received radical treatment, but the increase was very modest and much less than we thought. It went from 55% to 61%, so just a 6% increase, which we thought was very low. We thought more patients would switch over.
When we published our results in the 10-year median follow-up, which was in 20163, what we found is that there was no impact of the various radical treatments on mortality at 10 years, and that there was no difference between the arms. Whether you received active monitoring, surgery or radiotherapy, according to intention to treat, at 10 years, there was no difference. Everybody was surprised. Everybody thought that these curves would start to separate and that we would have an impact that 15 years. But much to everybody's surprise, including ours, we did not find that difference.
Dying from any cause, again at 10 years, was reported in 2016. There was no difference between the arms as to dying from any cause. The same continued up to 17 years, median 15 years. Again, these little things do not show any significant differences at all. Now, the reason that we thought between 10 and 15 years that we would see a difference in mortality is because at 10-year median follow-up, we saw that if men were receiving radical treatments, there was a reduction in the rate of metastasis by about 50%. It wasn't a high rate of metastasis, but still the radical treatment seems to reduce that by 50%. We thought that because that happened at 15 years, it would translate into a difference in mortality. And in fact, it didn't. The difference in metastasis continued over the 15-year median, up to 17 years. It was still half, but still not high. If you look at the lead percentages, 10% of the men on active monitoring had a risk of developing metastasis and 5% in the treatment arms. Although there was a difference, there was no difference in mortality, and that was quite a major finding that nobody expected.
We looked at 2 new measures. One is what is the difference in the number of men who went on long-term androgen deprivation therapy? The reason we looked at that is because the quality of life of men under androgen deprivation therapy can be affected. Taking hormones in these men reduces libido, and it has impact on cognitive functions, and lots of other things. Exactly the same as metastasis, we saw that there was a difference in the administration of long-term androgen deprivation at 10 years and 15 to 17 years. We also had a difference in local disease progression. This is without the disease spreading in the distance like in the lymph glands or in the bones. This is disease spreading locally on the prostate. We found a difference, which was more than 60%.
The difference with these is that the definition of local disease progression was very subjective. It's probably the weakest subjustification for giving these patients active treatment, and yet some of them did get it but without absolute strong evidence that there was disease progression.
Hamdy: Since 2014, there have been misconceptions about the study. The first one is, when we published our baseline data, in a paper on the Lancet Oncology4 in 2014, we put all the clinical pathological criteria for all the men's information. There was a perception then that about 77% of men had low-risk disease, but we had never actually assessed their risk by stratification by using proper tools and methods, which we did this time. What we found is that using these contemporary risk stratification methods, more than a third of men had intermediate- or moderate-risk, and some had high-risk diseases as well. That changes the outlook toward that cohort, in that it isn't just a low-risk cohort, it has got an important component of intermediate-risk disease.
The other thing that we found is that when you do surgery on a patient, you know exactly what stage and what grade of the disease they had. We looked at all the men who received surgery, and we found that one third of them at surgery, when they received it within 12 months of the diagnosis, had disease, which was already pushing itself outside the prostate, so it was more advanced than we thought. Because it's a randomized design, if there is 30% in 1 arm, we have to accept that there is 30% in each of the other arms as well. You put these together, and it shows that our cohort is actually not a just a low-risk cohort.
The other thing is some people thought we are claiming that men with aggressive high-risk prostate cancer should not receive treatment or should have delayed treatment. We never said that. We are not disputing at all the fact that these men have disease which can be lethal, and they need to have early and aggressive treatment. People also were saying that active monitoring or surveillance as we did, it is now outdated. There are lots and lots of other methods, including imaging, other genetic tests that people may use, and biomarkers. What we're saying is that despite the low intensity that we had with active monitoring, it still has not been able to demonstrate that there was a difference in mortality between the arms.
Finally, again, people are saying that because there's better imaging, there's more biomarkers, then these modern technologies give us a reliable indicator of what is going to happen to the disease at diagnosis. There is no evidence to suggest that any of these biomarkers or imaging, yet, show any difference in the outcomes and in the cancer-specific survival in the long-term.
There's another indication of what happens to men with metastases. At 10 years, these are the numbers of men who had metastasis who were allocated to active monitoring, surgery, or radiotherapy. [In the active monitoring arm] we had 22 men who were alive with metastasis by the end of 2015, and of those, 14% died. For surgery, it was 8, and 25% died. For radiotherapy, it was 10, and 70% died. There is a discrepancy between the percentage. We're not quite sure why. We assume that in the active monitoring arm, the patients were followed up very, very rigorously, and we [waited for] the disease to progress and they were treated early, which is why we have a lower incidence of death in this group.
Hamdy: Firstly, the survival from clinically localized prostate cancer, low and intermediate risk remains very high over the median period of 15 years, and that is irrespective of treatment allocations. It reaches 97% or 3% death, or 97% survival.
Men with metastasis do not necessarily die of prostate cancer, and this is again a new finding, because most people thought if you have metastasis, you're going to die of the disease. We found that that is not the case. It poses a question on whether we should be using metastasis as an end point in clinical trials in the future. Undoubtedly, men who die of prostate cancer have some molecular and genetic lethality features [that] have not been identified yet. These are not easily impacted by the multimodality treatment approaches. The risk stratification methods are not reliable at the moment.
We think the indications for active monitoring or surveillance can be expanded safely to intermediate-risk disease. Although our trial is not a screening trial, it's important to emphasize that. We have very long survival, which we showed up to 20 years after the diagnosis, with or without treatment, with all that metastasis, in itself does not support the introduction of population testing for early detection. Maybe it should be risk stratified and not mass testing. Finally, as I said in the beginning, the treatment decision with the survivals that we have shown and the [adverse events] from the treatments that Jenny's going to explain now, together, the information can be combined for individual patients to measure that tradeoff between the benefit they would get in terms of metastasis and progression, and the [adverse events] in terms of dealing with leakage, sexual function, and bowel function.
Donovan: As Freddie said, the key thing is that the mortality is the same, irrespective of the treatment, but the radical treatments do reduce the risk of metastases developing. The issue is, what's the tradeoff with radical treatments, as opposed to active monitoring? I already mentioned this about the questionnaire that the men completed very well for 12 years. There were some concerns when we published last time that treatments had changed over the years since we did ProtecT. There are new treatments like brachytherapy and modulated radiotherapy. These 3 studies evaluated whether the source of patient-reported outcome measures that we used were the same or different for these techniques. Basically, each of these studies confirms that even with these modern treatments, the patient-reported outcomes were almost identical, except for a bit less toxicity for radiotherapy. These patient-reported outcomes do relate to current techniques.
The overall quality of life, in terms of physical health and anxiety, depression, there's no difference between the groups at any time. That's the overall general health, which is obviously good for the men. But there are some particular symptoms that are different between the different groups.
None of the men wore pads at the beginning of the study, but after treatment, there's a much higher rate of needing to wear pads in the surgery group. Although that recovers somewhat, there still remains 1 in 5, or 20% of the men, needing to wear pads to protect against urinary leakage through that time period. The active monitoring group increased a little bit because, as Freddie explained, some of them changed to radical treatment.
What people have tended to do is only continue studies until about 2 or 3 years or maybe 5, and tend to reach the conclusion that nothing much is happening after that. What we've shown is that these effects affect them in long-term. The higher rate of urinary leakage in the surgery men continues through time and even increases a little towards the end. The active monitoring group, it reduces a little bit and it plateaus, because fewer of them changed to radical treatment, and then not much change in the radiotherapy arm. At the end of the study, the radiotherapy arm was also increasing a little for urinary leakage. These are long-lasting effects that men need to know before they go into these treatments in order to make their choice.
Sexual function is another key area that's affected, or the proportion of men who are able to have erections firm enough for intercourse. Again, at baseline, they're all the same; around 70% are able to have these erections. After treatment, it's much lower among the surgery group. There's also an effect in the radiotherapy group, and the active monitoring group gradually declined over time. Again, these things continue long-term. By the end of the 12 years, all of the men have low levels of sexual functioning, but they get there in quite different ways. For the surgery men, the risk of having poor sexual function is early on, and it stays the same low level. With radiotherapy, there's a bit of a recovery after the treatment, but not anywhere near baseline, and then they decline a bit more slowly. Whereas the active monitoring group declined slowly over time with increasing age and some effect of the radical treatments.
In the paper, we have looked at the active monitoring men who stayed on active monitoring, because of course, if they stay on active monitoring, they avoid these effects. Of course, we see that some of them change, and if they change, they obviously have those effects.
In relation to bowel function after treatment, the radiotherapy group was a little bit worse, but there's no difference long-term between the groups in overall bowel symptoms. We did see a difference in fecal leakage, once a week or more. That was higher in the radiotherapy group at the early stage, then it seemed to recover, and it was the same in the groups. There's then a later increase in the radiotherapy group. It's higher than the other 2 groups by the end. Again, short-term studies will give you the wrong impression. These effects can come later on for radiotherapy.
Donovan: Our key messages are that the [adverse events] of radical treatments can continue to affect someone's life 12 years after treatment. Each treatment approach has a particular profile of effects long-term, and we've produced all these data in all these diagrams, both for the patient-reported outcomes and the clinical findings, so that clinicians could share those with men in their clinical practice. They can help to explain this issue about the tradeoffs, because men newly diagnosed with localized prostate cancer can now have all this information to help them make that decision. It is a difficult one, to trade off between the risk of metastasis with active monitoring vs the harms of treatment with the radical options. We've enabled that to happen. Now they can look at the short-, medium-, and long-term [effects] with our data, and use their own values and priorities. Of course, men will differ whether their sexual life is particularly important at a point in time or not. The whole idea of this is to help men make better informed decisions.
Hamdy: We thought we would probably stop at 15 years, because we would show that there is a difference in mortality, and then that's it, we've answered the question. But the question remains unanswered. We do not know how long it takes for the radical treatments to have an impact on survival. The natural thing is that we should continue to follow these patients up. We've got 21% of the cohort who have died. There's still 79% of the cohort who are still alive. We have 24% of the men who are on active monitoring, who have never received any treatment at all, and are alive and well.
The first thing to say is that we absolutely need to continue to follow them up to 20 years. It is possible that at 20 years, we are then going to see a divergence in the curves that you have seen, that the radical treatment will have impact on mortality. It's also possible that we don't show an effect. If we don't show an effect, what it means is that the radical treatments given to patients with screen-detected disease will never impact mortality. They may impact on other things, like patients not being on long-term hormones, or patients having less symptoms from local progression. But that piece of information is critical for a number of things for a man when they are diagnosed with prostate cancer. The other thing is how can you reconcile the search for screening across all the populations when we know that the treatments don't seem to make a difference for such a long time. It's really important to continue that.
The other thing that is important is what we call translational research. In the men that have died, there is a black box in their biological materials. That black box, which in aviation terms is where all the secrets are of what happened when a plane crashes, each one of these men has a black box in the biological samples that we have from them in the tissue and in the blood and in the DNA. We are planning to look at all these to try and understand where these lethality features at molecular and genetic levels are so that we can find the proper indicators and biomarkers and diagnosis. That's after the diagnosis of prostate cancer, not before. Sometimes before, if we can.
Donovan: On the quality-of-life side, we're keen to look at the impact of the metastases and the long-term hormones to compare those who take the long-term hormones or have metastases and compare with the others to see what differences there are. That would help with the equation. Also to publish some results about why the men changed from active monitoring to the radical treatment, because that's obviously another issue. How do you keep them on it when it's safe, and how do you allow them to change when there's disease progression? At the moment, it's a mix of reasons.
Hamdy: Where we want to get to is when a man decides to have a particular treatment for his cancer, and we are talking about low- and intermediate-risk disease, we're not talking about high-risk disease, they need to be well-informed about what they're going to be like after 5 years, after 10 years, after 15 years, and after 20 years. This is what we're able to provide through this fairly long follow-up.
Donovan: We do know now that they have time to make that decision. They don't have to rush into the decision, which some men are urged to make a decision quickly. These data suggests that men have time to really consider all these issues and weigh them.
Hamdy: We can't claim that we have answered the question of screening with our study, because it was essentially a treatment effectiveness study of screen-detected prostate cancer. The reason we mentioned screening is because we found such a very long survival, with no impact on the detection and the treatment on these patients. We have to question whether population screening is actually a sensible and reasonable thing to do in the way that it's done today, which is by PSA testing, imaging, and doing the biopsies. You suddenly convert a healthy man into a man with cancer. You give them a new identity. You give them a new passport, which is a cancer passport. They have to live with this for the rest of their life.
The most important thing for physicians is that at the moment, what's happening is people do the PSA test, and then they think after, and then they treat the patient with a radical prostatectomy, and they think afterwards. What we are asking physicians and our colleagues to do, is to think before doing PSA, and think before doing the surgery or the radiotherapy. That hopefully, is going to make a difference. At the moment, I don't think anybody who knows all the evidence and all the data out there can conclude that screening should be adopted as a public health policy. I think the evidence is that it is likely to cause more harm than good, if it did happen.
What should happen is targeted screening. Targeted screening means that you have means to identify men who are at risk of developing aggressive disease in the future. If you're able to have these indicators before you test them, whether it's a hereditary factor or genetic factor, then that will be helpful and useful. But to do a mass screening with PSA imaging biopsy is not acceptable.
Hamdy: I think both papers are very complimentary. They're very clear as they're published. The key messages that we gave you are really the most important. Particularly, we want to get rid of the misconceptions about the trial, that it should be not considered because it's all low-risk, and we can't make any conclusions out of it. That is just not correct. We need to make sure that does not make people disregard what's in our conclusions.
Donovan: It applies for the low- and intermediate-risk cancers that are found. We just think men should have the opportunity to understand what the consequences might be of either the PSA test or the diagnosis before they make those treatment decisions.
1. Hamdy FC, Donovan JL, Lane JA, et al. Fifteen-year outcomes after monitoring, surgery, or radiotherapy for prostate cancer. N Engl J Med. 2023. doi:10.1056/NEJMoa2214122.
2. Donovan JL, Hamdy FC, Lane JA, et al. Patient-reported outcomes after monitoring, surgery, or radiotherapy for prostate cancer. N Engl J Med. 2016;375(15):1425-1437. doi:10.1056/NEJMoa1606221.
3. Hamdy FC, Donovan JL, Lane JA, et al. 10-year outcomes after monitoring, surgery, and radiotherapy for localized prostate cancer. N Engl J Med. 2016;375(15):1415-1424. doi:10.1056/NEJMoa1606220.
4. Lane JA, Donovan JL, Davis M, et al. Active monitoring, radical prostatectomy, or radiotherapy for localized prostate cancer: study design and diagnostic and baseline results of the ProtecT radomised phase 3 trial. Lancet Oncol. 2014;15(10):1109-18. doi:10.1016/S1470-2045(14)70361-4.