Professor Declan Murphy:
Thank you, Alberto. Thank you, ladies and gentlemen, for the invitation to this fantastic meeting. So we’ve been asked to discuss this in two frames of mind. One is conventional imaging and one is, again, the disruptive novel imaging that we’ve just been hearing about. I have no relevant disclosures. So let’s think first about clinically node positive by conventional imaging and the EAU guidelines regarding staging of high risk and localized cancer advises to perform conventional imaging at the moment, cross-sectional imaging with CT or MR.
And you can see from the definition of a positive node, these are quite large lesions by the time they are detectable on a CT and therefore quite a lot of cancer. And disappointingly, as you all know, CT therefore performs very poorly for staging pelvic lymph nodes with sensitivities of less than 15% typically in this high risk localized or locally advanced population. And although MR performs a little better, these are still disappointing as sensitivity figures, thereby explaining why a lot of patients will relapse following local treatment in the pelvis.
And then regarding the role of surgery for high risk locally advanced prostate cancer, no positive. The EAU Guidelines have evolved a little bit over the years. And at present, they say that in highly selected patients with node-positive prostate cancer, surgery can be offered but only as part of a multimodal approach. So we, therefore, need to counsel our patients very carefully and also of course counsel them about other options as you’ll hear shortly about radiotherapy or systemic therapy. And the guidelines also say that even with this recommendation, only limited evidence exists for the role of surgery in node-positive disease.
So let’s look at the two papers that inform this recommendation. The first is a combined series from MAYO clinic and from Alberto at San Rafael. And these are 302 patients with pathological node-positive disease. So these are PN1 as it turns out. And they went back and looked at the clinical staging of these patients and demonstrated that only about 17% of them were node-positive on the original CT scan, despite all of these patients being pathologically node-positive. And with a long follow-up, they wanted to see if there was a difference in survival in the patients who were clinically node-negative and clinically node-positive. The hypothesis being there would be no difference. And you can see from the patient profile, this is consistent with a group of patients who all have node-positive disease at pathology.
And what you see here is the fact that there is no difference in cancer-specific mortality, whether patients are clinically node-positive or node-negative and who are undergoing surgery, albeit, of course, this is a retrospective series. And they wanted to look at what the predictors of mortality were and as you might expect, high Gleason scores and the number of positive nodes were predictors for a cancer-specific mortality on multivariable analysis, but clinical node status was not a predictor of survival and thereby they conclude that clinically node-positive patients on the preop CT scan should not necessarily be a contra-indication to surgery because these patients seem to do quite well.
The second paper I’ll show you is from the National Cancer Database in the US of almost 3000 patients with clinically node-positive. So the previous was pathologically node-positive. So these are clinically node-positive and Firas Abdollah and colleagues look to see whether patients who were managed with ADT alone versus those who were managed with local therapy plus ADT had a difference in survival. And what you see is about two-thirds of patients did have local therapy in these node-positive, clinically node-positive scenario, including patients having surgery and radiotherapy. And note here in the patients having surgery, only about one third of them subsequently went on to have ADT. And that’s because, of course, many of these patients were pathologically node-negative, revealing the poor performance characteristics of CT. So using complex statistics to try and allow for the limitations in the retrospective study design, they showed overall that the patients who received local therapy, either surgery or radiation did much better than those who received ADT alone. So I think that’s an important message that local therapy of one type or another has value here.
And second, they went on to look at whether the type of local therapy made a difference, whether it was surgery or radiotherapy. Again, allowing for the study design here and what they showed is no difference, basically, whether the patient had a radical prostatectomy or radiotherapy plus or minus ADT didn’t seem to influence the likelihood that they were going to have durable longterm survival.
So my take-home messages to you on conventional imaging is that it performs very poorly for staging clinical node status when compared with pathological findings at node dissection. There is some low quality evidence to support surgery for clinical node-positive disease, but this should be discussed within a multimodal conversation with the patients. Of course, we’ve just heard a fantastic intro into the role of novel imaging in this whole space, so I’m going to wind everybody up a little bit more by talking about this.
This is our updated systematic review, which was shown briefly already of about 5,000 patients undergoing a PET scan. Most of these are biochemical recurrence, but we have enough data now to do a systematic review on the primary lymph node staging, and this is from that European urology paper. And you can see from this impressive curve that the specificity, of course, is 96-97% and the sensitivity is very good for detecting lymph nodes in the primary staging setting at about 77% albeit the usual limitations of the papers that I’ve shown you here. But clearly I think those of us who work a lot with this technology understand that we do see lymph nodes with much better performance and Stefano also showed something about the EAU guidelines and these are very reasonable comments I think in the current guidelines, which do not recommend novel imaging for primary staging, but they make some nice comments.
They comment that, first of all, as I’ve just shown you novel imaging seems to be more sensitive. Yes. However, these are really important questions. The benefits of detecting these lesions earlier is not clear and especially if there’s disease you can see only on a PET scan, metastatic disease especially, it’s not clear if these sorts of lead patients should be treated systemically or whether they should have aggressive local and metastasis directed therapy approaches. And I think that’s very important. This is the management impact that we’ve discussed already and they also highlight that there are trials underway, evaluating management impact and I’m going to show you a little bit about one of those studies.
This is the pro PSMA study which we’ve been running in Australia, supported by the Movember Foundation and PCF Australia and coordinated by ANSA, and this is a trial of 300 men with high-risk localized prostate cancer who were randomized to either get a CT and bone scan first or to get a PET scan first and then we record the management impact and then they cross over and then we can rerecord the management impact.
So it’s set up to measure accuracy. It’s a prospective trial with very tight quality control on all aspects of the imaging and secondly to evaluate the management impact. And this study as with all our PSMA related studies in Australia, recruited very quickly, well ahead of time, and we locked the database for followup just recently. And we are preparing to preserve to present the primary endpoint of that. Michael Hofman will do that at EAU next year, if it’s a positive primary endpoint.
So that I think will answer some of the questions about management impact. And whether it’s right or wrong, we will show the accuracy, but we will show what the management impact is in this study. And I’ll show you two cases from proPSMA, some may just to highlight two different aspects. Here’s a typical high-risk prostate cancer patient with a big M or a lesion and you can see it very nicely on the PET CT.
He was randomized to a CT first and this is N zero by CT criteria and then he crossed over to the PET scan and this is not unusual. Of course, you see these small internal iliac lymph nodes. And you may argue here that PET scanning hasn’t actually helped much because there’s not much management impact. This patient would have been recommended to have an extended pelvic lymph node dissection which would have discovered these nodes if he had an adequate dissection. However, here’s a different scenario. A 46-year old in the trial presents with lower urinary tract symptoms, diagnosed with high-risk prostate cancer, randomized to CT and bone scan first, which were normal. So in any other paradigm, we have a young, very fit patient with high localized, locally advanced prostate cancer would have been offered a conventional treatment. And you can see on the PET scan of the pelvis, he’s got obvious pelvic lymph nodes which were not apparent on CT.
People sometimes wonder how can you not see these on CT? But of course on the PET scan there’s a lot of avidity, there’s a flare effect. But when you look on the CT, these are less than 10 millimeters. But unfortunately, he’s also got more established disease. He’s now an M1A patient because he has extensive retroperitoneal lymphadenopathy. Very obvious, high SUVs. We don’t biopsy these anymore. We used to five years ago when we weren’t certain, so what do we do now? There’s very profound management impact for this type of patient. Should we still proceed with primary therapy, which is what he would be offered anywhere else in the world at a young fit patient like this, or should we believe that he’s got metastatic disease and offer him best systemic therapy plus perhaps some local therapy to the prostate, à la STAMPEDE, et cetera.
So this is a classic conundrum we see with high quality imaging. But remember those of you who have doubts about this type of imaging, this is real. This is prostate cancer. Okay. So I think it, you can’t unsee this when you’ve shown it to a patient. So we can’t just ignore it and persist with the 20 or 30 year old imaging paradigms. We have to somehow embrace it into the decision making and try not to harm patients.
So I think these are the really important considerations. The accuracy is one thing, but it’s the management impact, I think that’s the most profound thing here as Ian Davis talked about. And I think my bottom line here is that I do think those of you running very large phase three randomized trials of metastatic or M0 high-risk prostate cancer should embrace novel imaging in your trial design. You don’t have to do the whole 2000 patient series with novel imaging, but build it into the study design so we can record accuracy and management impact.
Finally, a little extra advantage we have in PSMA-detected node diseases. The ability to identify these nodes a little better at the time of surgery because it’s not easy sometimes to find those tiny internal iliac nodes that are seen on the preop PET scan. But this group from Munich, led by Matthias Eiber and Tobias Maurer before he moved to Hamburg, developed a nice technique for salvage lymph node dissection, which can now move into the primary setting where they inject indium labeled PSMA prior to surgery and then intraoperatively using a combination of an acoustic probe and a visual probe, they can identify them. So what you see on the left is the probe about to be deployed into the abdomen to detect these lymph nodes with a signature you can hear, if they turn the sound up please. And on the bottom you can see the dense PSMA staining on this. And on the right you’ll see that as they take out the lymph node packets, they can detect and verify that they’ve resected these nodes.
So I think the PSMA may also be just an opportunity to do higher quality targeted surgery using this type of technology. And they’ve reported nice data in terms of accuracy here. So my take-home message is on novel imaging. It’s self-evident to those of us who use the technology that the sensitivity and specificity is better, but we do need the objective evidence and the appropriate PSMA trial should provide that for us. It’s very disruptive and I think it’s the management impact that is the most considered challenge we face in this era when we have widespread access to PET scanning in some countries.
So in conclusion, I think there is some good evidence for local therapy in cN1 M0 prostate cancer by conventional imaging. And we’ll hear about radiotherapy as that option for which there’s a lot of evidence, but surgery can be an option as part of a multimodal approach. I think it’s fair to say that novel imaging provides a much more accurate picture of the disease extent, but we need to very carefully evaluate the management impact. And as a result of that, I think we will select better patients for surgery. Thank you very much indeed.