Dr. Laurence Klotz presented “Active Surveillance: Who is the Right Patient and What is the Right Protocol?” at the 27th annual International Prostate Cancer Update meeting on Wednesday, January 25, 2017.
Keywords: prostate cancer, MRI, gleason, biomarkers, focal therapy, biopsy
How to cite: Klotz, Laurence. “Active Surveillance: Who is the Right Patient and What is the Right Protocol?. January 25, 2017. Accessed Aug 2019. https://grandroundsinurology.com/active-surveillance-right-patient-right-protocol
Active Surveillance: Who is the Right Patient and What is the Right Protocol?
Ladies and gentlemen, it’s really a great pleasure to be here. It’s one of my favorite meetings. I call this shrinking the gray zone, and you’ll see why. That is Toronto looking its best. So we published now 15 years ago this paper which, in retrospect, was kind of pivotal. It’s a group of patients that we started managing with surveillance prospectively beginning in around 1997 with informed consent. It was considered experimental and highly risky. We’ve learned a lot since then. We know the problem of over-treatment has been alluded to. No one disputes this.
We know much more about the bad actors and the nature of who has of the occult high grade disease, the misclassification of risk. We know about the predictive value baseline parameters, we have talked about that a lot, actual clinical parameters, just the extent of disease, PSA density, race, and age are quite useful. We know about the flaws about PSA kinetics. Most groups, including ours, a rapid PSA rise is a trigger for intervention. That was not a good strategy and resulted in a lot of over-treatment. We use it now as a trigger for further evaluation, but not treatment.
Multiparametric MRI is a game changer. New modeling studies that show basically it’s a very robust approach even if the mortality rates start to creep up with longer follow-up. And more than that, more than 2,000 publications ─ I don’t know if you can see that ─ from a Medline search of active surveillance in prostate cancer. It’s really an area of very intense investigation all over the world.
Now I hope─ I’m less assured now than I was 20 minutes ago hearing some of Dave Crawford’s comments, but I think what we know and almost everyone agrees with, is that Gleason 3 and Gleason 4 are like night and day, that the molecular genetics of most Gleason 3 is normal. The metastatic potential is approximately zero. I have spent the last ten years trying to collect cases of pathologically characterized Gleason 6 that have metastasized. I’m not convinced there is one really well characterized patient where we had a radical P, only Gleason 6 by contemporary standards who went on to metastatic disease.
Even if there is a handful, it doesn’t matter, whereas Gleason 4 has all of the molecular hallmarks of cancer. And the major Achilles’ heel, the limitation of surveillance is that we underestimate higher grades of cancer which is present in around 25% to 30% of these patients. It doesn’t mean it’s all going to result in the death of the patient if it is left untreated, but it certainly puts them at a higher risk. In contrast, true biologic grade progression where the patients Gleason 3 progresses to a 4 or a 5, it is much less common. And our estimate based on serial biopsies, and Hopkins has come up with about the same estimate, is that this occurs in about 1% of the cases per year usually in a large field of Gleason 3.
So that is one of the significances of the volume of 3. You do get pre-histologic adverse genetic alterations. Dave comments on the 20% that are bad. That doesn’t mean the patient is going to die of prostate cancer if he is left untreated, so we still have to learn what the significance of those pre-histologic molecular changes are. And finally we are in the MRI biomarker era, I am not going to go into that in much detail. I used to talk about how there were these two diverse pathways of genetic development of three which was a proliferative pathway versus 4 and 5 which is a dysplastic pathway kind of like bladder cancer.
But there is increasing evidence now that the differences are more quantitative than qualitative. This is a recent study by Mark Ruben that looks at genomic alterations, and you can see Gleason 3’s on the left, and you move through the higher grade cancer, the alterations are there, you can see it here. But they are much fewer and the amplitude is much less. And so clearly this is partly a matter of degree. One example of this is PTEN loss, which we heard about. PTEN is considered one of the key events in progression of prostate cancer, the loss of PTEN. And this study from Hopkins showed that PTEN loss, while it did occur in Gleason 3, it almost all occurred ─ these were a group of patients with Gleason 7 cancer ─ it almost all occurred in the Gleason 3 that was associated with a higher grade cancer.
In other words, PTEN loss was very rare in 3+3 alone or 3+ only a small amount of 4, so you can see that here, the PTEN, this is Gleason 6 in the white. That is one observation that I am going to provide an explanation to. And this is the well-known paper case report by Haffner, the guy who had a radical P and died 17 years later, and they did whole genome sequencing from multiple bone mets in every sub-area in his prostate. And the surprising thing was the molecular hemology between the little microfocus of Gleason 6 and the metastasis, not the extensive Gleason 4.
So does this mean that Gleason 3 can really be a metastatic cancer? Well, here is the explanation I like, and we heard about exosomes earlier. And the thing that wasn’t mentioned about the exosomes which have been learned in the last three or four years is that these exosomes contain RNA, and they are shed thousands per day by each cell. Each cancer cell is shedding these exomes and they are taken up by neighboring cells. They are also in the circulation and they are taken up by distant cells, and they adversely influence the biological behavior of those cells.
So this has been shown in animal models. A high grade prostate cancer converts a low grade cancer in the other flank to a more aggressive phenotype. It’s like a kid that has been raised well with good values, he goes into a bad neighborhood and falls under the influence of a gang and becomes a drug dealer. And this has huge implications for all kinds of things that we do, I think, but it also explains the Haffner case and it explains how PTEN loss is more common in 3 when it’s associated with 4. It is probably at least in part related to this extra-cellular vesicle transfer or exosome uptake that contains adverse RNA that adversely affects the cell.
We just had this discussion. I believe the Gleason 6 doesn’t have metastatic potential. It is still a cancer, and it can invade just like gliomas, just like basal cell cancers, doesn’t metastasize, and has the characteristics of cancer. I am not going to change the name, but it is a non-metastasizing cancer with a very important concept. If it doesn’t metastasize, why do we care about the volume? For those of you who were there at Dave Crawford’s session from six to seven this morning, he talked about this 48-year old who had extensive Gleason 6, and he operated on him.
And he considered this as a victory of the diagnostic algorithm. Frankly, I don’t know why he operated on him because that guy, there is no way that patient had higher grade cancer. He had been so thoroughly evaluated. You don’t have to treat a patient with high volume Gleason 6, but you have to look hard to make sure that they don’t have anything more. It’s a marker for the presence of higher grade cancer. This is one meta-analysis that shows papers that have looked at this predicting for disease reclassification which basically means upgrading, the PSA density which is a surrogate for cancer volume.
Race, in most papers, is predictive for an increased risk of higher grade cancer, the extent of core involvement, the number of cores and so on. So that is the significance of higher volume 6. You have to re-scrutinize that patient. MRI, perhaps template biopsies, whatever your preferred approach is. There is a range, there is about nine or ten prospective ─ large prospective ─ series now in the literature. There is a range of selection criteria and there are really two extremes. We were at one extreme because we included 3+4 or a PSA between 10 and 20.
The Hopkins group was the other extreme, only Epstein criteria of one or two positive cores, no more than 50% involvement with low PSA density, and the others sort of fall in between these two extremes. So we now have a lot of data from these series that allow us to define what the risk is with these different criteria, so we had about 22% who were intermediate risk of either Gleason 7 or a PSA of above ten. These were mostly older patients, but more than a third were under age 70. And we now have with a long follow-up 30 patients out of 1,000, 3% have developed metastases, 15 prostate cancer deaths at 1.5%, and so that is a meaningful experience and we learned something from that.
Just to emphasize, this was a publication that just came out from Toronto showing that the eligibility criteria or surveillance don’t really influence the likelihood of being treated. So this shows all of these different groups that each have their own intervention criteria, and basically about 25% to 30% are treated by five years. So this is our data, and we had out here an actuarial 15 year prostate cancer mortality of 5%, and this was published about a year and a half ago. And you know I used to think, gee, we’ve done so well, 95% of these prostate cancer patients have had long term survival free of progression.
And then we started to look at how we could predict for the patients who had a metastasis, and essentially the presence of Gleason 4 at baseline was an incredibly powerful predictor for the progression of metastatic disease. So this shows cost-specific survival by low versus intermediate risk at 3% versus 11%, a hazard ratio for metastatic progression of almost 4. It is very meaningful, and this shows that it was all the Gleason. The PSA above ten really had no significant, slight, but not really very significant impact on the likelihood of metastasis whereas the presence of Gleason 4 had a huge impact.
And this is the most compelling data. This is recursive, portioning analysis breaking the patients down the Gleason─ the patients down into each little subgroup. So Gleason 6 on the left, they all did fine; Gleason 7, any Gleason pattern 4 resulted in at least a 20% 15-year prostate cancer metastasis rate despite close monitoring, treatment if they looked like they were getting worse and so on. you see down here is the best group, a PSA of less than 10, percent of cancer less than a third, 3+4, 30% metastasis rate at 15 years. So we have become more restrictive as a result of that, less enthusiastic about surveillance for Gleason 3+4, notwithstanding that biomarkers and MRI are playing a role in some of these patients.
Now in contrast to us, you have the Hopkins group restrictive approach, Epstein criteria only, 0.5% prostate cancer mortality at 15 years. So in a sense that is a victory. They did better. We had actuarial 5% but actually only 1.5% mortality overall. So you’ve got these two extremes. We can offer surveillance with our historic approach to 50%. They offer to about 20% and clearly there is a middle road. The middle road is to be more restrictive about the Gleason 7, but more inclusive about Gleason 6. And my understanding ─ and Alan may correct this ─ is that the Hopkins group is now more inclusive about Gleason 6 beyond Epstein so there has been kind of a convergence of approaches.
Now we are in the MRI era, so this is an example I like. This is a guy who had a systematic biopsy showing one positive core, Gleason 3+3, he had this MRI, and there is the lesion. And there is no way you would pick this up with a systematic biopsy, 4+3 on targeted biopsy, and here is the pathology. There is his Gleason 4+3 cancer, you know so I mean this speaks volumes. Without MRI, this guy would have remained on surveillance and maybe he would have been one of the ones who did badly, so MRI.
The main point, this is a summary of the performance of MRI. This field is on fire. Like there is a new MRI paper every single month with more data. The NPV for clinically significant cancer which is the key metro cure, it’s quite variable. Anywhere from 65% or so to 100%; I just want to focus on this Pannebianco paper published about a year ago. It had 1,100 patients in a randomized study, NPV of 100. There was not a single patient with a negative MRI who had clinically significant cancer on subsequent systematic biopsies, and 99% positive predictive value.
We haven’t done so well. I’m actually presenting this data for the first time. This is a study we just finished called the ASSIST study, a randomized study of men on surveillance randomized between systematic biopsy and MRI with targeted and systematic biopsy. We wanted to see if the MRI was negative, could we forego the biopsy. No difference in the rate of upgrading between the two arms. The upgrading was achieved with only two cores in the MRI group versus 12 cores in the systematic biopsy group, but look at our performance of the MRI, and this is a multicenter study. It’s more of a real world study, I would say. The positive value for PI-RADS 5 was 33%. It was 99% in the Pannebianco study.
The positive predictive value for four was in the same ballpark; the negative predictive value was 85% versus 100%. Around 8% of our patients with a negative MRI had a Gleason 7 or higher. And this just shows the data. This is the targeted biopsy group, no cancer Gleason 6; 8% had systematic biopsy with Gleason 7, and 6% the other way around. So, I don’t know about MRI. It works, it adds a lot. It’s not, I don’t think, as good as some of the literature suggests. It’s also got a role on men on surveillance. In follow-up it is a typical example, a stable lesion Gleason 6 down here. Here it is a year later and it hasn’t changed. That has to mean something versus progression on MRI. So I think increasingly it’s being incorporated into the longitudinal follow-up on these patients, although we are still waiting for an outcome on this.
So biomarkers, now I am not going to add to the extensive discussion we’ve had, but I want to show one piece of data that didn’t get discussed this morning. I thought it was a fantastic session. This was published a few months ago by Wei. They did whole exome sequencing in a small group of patients who had multifocality. So these are four different patients, and these are different cancers within the same prostate. And because of this whole genome sequencing, they could look at the genes that are incorporated into Oncotype GX, Prolaris, and Decipher. And what do you see? Surprise, surprise; heterogeneity.
Look at this. I mean this patient, very negative versus very positive. Most of the signals are in the same direction, but a big difference in amplitude. So you know in the debate which we have been alluding to between MRI and biomarkers, which is best and when should they be used, to me the heterogeneity issue continued to bedevil assays based on the assessment of individual biopsy cores. And this is some evidence that that is the case. But however these pan out, we are into the era of image-based molecular paradigm. There are many, many advantages to this and it reassures the patient.
I’m sure we are going to reduce that 15 year mortality rate from 3% or 4% to less than 1% by early identification of the bad actors. My belief is that most of those patients who metastasize had large anterior cancers that we completely missed. A lot of them had already metastasized at the time of diagnosis or at least perhaps they were amenable to cure, but weren’t treated early enough. So that is going to change dramatically as we incorporate these new tests.
Guidelines; this I think is kind of relevant in the context of the biomarker discussion, in the biomarker discussion we had this morning. So here is our guideline, the ASCO guideline, which basically adopted the Cancer Care Ontario Guideline word for word with one exception, and the NICE guidelines. So they are pretty consistent, you know a serial biopsy, PSA, other tests. Our approach was MRI when clinical and pathologic findings are discordant; ASCO, other tests remain investigational for both MRI and biomarkers; NICE says everyone should have an MRI. They have gone out on a limb with that.
Five ARI’s was the one area we said may have a role; ASCO said no clear role; NICE doesn’t comment on it. The biomarkers have not been incorporated into these guidelines as of yet. Maybe that will change in the next couple of years. Just to show uptake, this is our six or seven different studies, and basically all over the world, New Hampshire, Michigan, the MUSIC collaboration, Australia. This is CAPTURE, Matt Cooperberg in the yellow. This is Sweden, and more than 90% of low risk patients managed with surveillance. And this was just published from Canada. I am very proud of this; this is uptake and surveillance in low risk patients by individual center.
You can see there is at least 65%, and in some cases there were more than 95%. So you know the pattern is there. Actually this is the same slide. Finally I just want to talk briefly about some things that you can do, the opportunity to improve the health of the patient on surveillance. So you know a lot of these guys, you have a healthy 60-year old found to have microfocal Gleason 6. You may be the first doctor he has seen in years and there is an opportunity to intervene. The obvious things are smoking cessation, weight loss. This was a study that Ian Thompson just published that showed you might think─ you tell the patient, okay, you’ve got prostate cancer, you need to improve your health, no change in weight after the diagnosis of prostate cancer in their cohort.
This was a study called Active Holistic Surveillance that was published recently. Aaron Katz, probably some of you know, some of the advice that they give to the patients, broccoli, Omega-3, this is an anti-inflammatory, Genikinoko, whatever that is, etc. I mean you can get really carried away with this stuff, but certainly the patients like to hear about this. Exercise, long story, we are going to hear about that tonight probably, but many ways in which exercise may delay the progression of prostate cancer. This is quite an interesting review article published recently, Enhancing Active Surveillance in Prostate Cancer, the Potential of Exercise Medicine for anyone who is interested.
And this is just a list of the lifestyle studies that are currently ongoing in men on surveillance. And not to go into detail, but there is a huge interest in this area in intervening with some kind of lifestyle, diet, and micronutrient intervention during this period to reduce the risk of progression. We are doing a metformin study in surveillance in Toronto. So, finally, here is the gray zone. So I would say until about 2012 maybe and the U.S. Preventive Service Task Force recommendation which influenced practice, this was the gray zone. It was huge.
Almost everything we did was in the gray zone. T1a, you know, micro focal 6 after a TERP [phonetic] and no one was treating that. But with the rest you could do whatever you wanted. Now the gray zone has shrunk, and I think it’s fair to say for most people, they think that low volume Gleason 6 should not be treated. A higher volume, higher grade disease rather should be treated, and now the gray zone is the patients with extensive disease, Gleason 6, very young men under 50, a small amount of Gleason 7, there is a debate about what─ a small amount of Gleason 4 in Gleason 7 patients.
The patient who has got a restrictive diffusion on his MRI but low grade on biopsy, I don’t know what to do with those guys. A high PSA density is also kind of a red flag. Clearly we have a way to go. We incorporate biomarkers and personalized medicine ideally, no more gray zone. We know exactly what to do and what the risk in each patient is and how aggressively to treat them, so that is the goal. And I will conclude there. Thank you very much.