Declan Murphy: Thank you, Neha. Thank you very much for the invitation. We have a very short break from PSMA before business, as usual, is restored I'm sure. emailed me and asked me to speak about prostate cancer pathology. My main disclosure here is that I am clearly not a pathologist, so I have no idea why she asked me to speak about this. I did reflect on why she asked me to talk about it. Why she would ask a urologist to speak about prostate cancer pathology? Why ask a... I'm even simpler than [inaudible], to speak about this. I do have history, it's catching up on me. Because after Sweeney read out the charted results, I joked that we urologists should learn how to give docetaxel, it's not that difficult. Is it really? I'm wondering whether also, I can update my simple approach to life by talking about prostate cancer pathology and make it simple enough for me to understand.

It used to be simple. Just 10 years ago, there were really only three things about the prostate pathology report. It was Gleason 6, it was Gleason 7, or it was Gleason 8/9/10. That, really, was how we risk stratified our patients. Maybe the number of cores was relevant. But today it's a lot more granular to everybody's great benefit. There are a lot more things we consider, and the ISUP panel have done a great job advancing at this field.

The three things that have been identified for discussion today, I think are the three most important things for we as clinicians to consider. So I'm going to talk about those. But already, I can tell you my take-home message is that knowledge of these variants is important. Knowledge of what the ISUP panel recommend is important for our clinical practice. And working closely with our specialist, prostate cancer pathologist is important. So if you remember nothing else, please embrace that in your practice going forward, because these factors impact greatly on how we stratify our patients, and therefore, the management recommendations we make for them.

Let's talk about these three variants. First of all, cribriform. The definition is seen there. This is one of the four classic appearances that make up Gleason pattern 4. There was a seminal moment, excuse the pun, for cribriform pathology in 2014, when the ISUP panel made a very big change. Because prior to that, Gleason 6 cancer was allowed to have some cribriform appearances. But in 2014, based on data showing that cribriform drove poor outcomes, the panel said that cribriform, whether it is small gland or large gland, should all be considered Gleason pattern 4. That has been one of the biggest changes in prostate cancer pathology over the past 30 years.

It was a good decision, because this paper that was published a year later, highly cited, went and looked at how important cribriform was for predicting how Gleason 7 patients will do grade group 2, grade group 3. They compared the cribriform appearance versus the non-cribriform appearance. What you see is that, Gleason 7 patients with cribriform appearance have much worse biochemical recurrence, much worse progression to metastasis, much worse disease-specific and even, overall survival.

So that's a really important message for us when we're managing our grade group 2, grade group 3 patients. It is that the presence of cribriform is very important for counseling our patients and managing how we look after these tumors. The same, by the way, applies to Gleason 8 cancers, which another Dutch group also published. I won't show you today.

I think this is the big impact of cribriform on our practice, is that some favorable intermediate-risk prostate cancers are candidates for active surveillance. Of course, all low-risk patients are, but what about intermediate risk, the really big population? Yes, some of them are, but we have to be careful. I think this is my take-home message for you. ISUP 2 grade group 2 cancers without cribriform, may be candidates for active surveillance. Whereas grade group 2 cancers with cribriform growth are not candidates for active surveillance, which is what the guidelines currently say. So when I see a newly diagnosed grade group 2 prostate cancer patient, which is a huge amount of our practice, that is the first thing I look at.

Cribriform clearly outperforms percentage pattern four, for example, in predicting the likelihood of a poor outcome. Whilst it's not well understood exactly why that is, it is essential for pathologists to report the presence of cribriform, which they do, because the consensus guidelines recommended. But for us clinicians to appreciate how important it is to look for the presence of cribriform on these patients.

We will move and talk about the other two areas I've been asked to speak about, which are IDCP intraductal carcinoma and ductal carcinoma. We wrote a nice review about this in science, translational medicine, a couple of years ago with our translational team. That's a very detailed analysis of the pathological and molecular features of IDCP and ductal, which I will just refer to briefly today. But if you're interested in this topic, that's a really nice piece of work led by our colleagues in the science field.

Silke Gillessen: [inaudible] by the way, the reason why you were invited.

Declan Murphy: Really? It's not just you tormenting me. Thank you, [inaudible]. This is a definition of IDCP. There is confusion because people think, "Well, they've both got the word ductal in their title." However, IDCP is of course, much more similar to cribriform growth. So the topic we've just discussed, the difference being that in IDCP, the basal cell layer is still preserved. That is the pathological feature that determines IDCP versus cribriform. But sometimes clinicians get confused about, "Oh, it's ductal. Ductal is really bad," but it's not. This is much more like cribriform. That's the way I think about it.

What I want to do is explain to you how we accidentally came across IDCP as a research interest in our group in Melbourne. Gail Risbridger and her team have an excellent PDX program for localized prostate cancer. We were PDX-ing a bunch of my BRCA2 patients undergoing radical prostatectomy, when completely accidentally, we noted that IDCP was growing very floridly in the BRCA2 PDXs. Whereas, when we went back and looked at the PDXs for our regular high risks, we found just very modest levels of IDCP, a classic accidental discovery. We then wanted to know whether the presence of the IDCP in the BRCA patients was driving a worse outcome, and it was. So these are all our BRCA2 patients from that particular series, as stratified into whether they had IDCP or not. Clearly, the ones with IDCP had a much worse overall survival at 10 years.

So, an accidental discovery that got the community thinking, how important is this? We then went and did a systematic review to look at the importance of IDCP and the relationship to higher-risk cancers. We clearly showed that although IDCP is still under-reported and underappreciated, you can see, as we get intermediate higher-risk metastatic patients, you're much more likely to see IDCP in your pathology.

A question that arises then is, if we see IDCP, knowing now that it is associated with higher grade cancers and perhaps is driving some poor outcomes in the BRCA2 patients, should we be looking for DNA repair defects if you've got a patient with a newly diagnosed cancer with IDCP? [inaudible], who's here, and colleagues, then went and looked at this in about 150 unselected patients with recurrent or metastatic prostate cancer. Did germline sequencing on them and noted, rather typically, about 14% of patients had germline mutations. But their question was, are they seeing more IDCP or ductal in those patients compared to a non-DNA? And yes, is the answer. Up to half of the patients with the germline mutation had IDCP versus only a very small number with germline. They, therefore, recognize that the presence of this type of histology appears to be associated with DNA repair defects. And therefore, it may be important to identify those defects if your patient has IDCP or ductal. So there you go.

There are a couple of papers saying we should all be doing germline sequencing. That's exactly what the NCCN did on the foot of these papers and a couple of others. They recommended that in patients with a history of prostate cancer with introduction or cribriform, in deep pathology, we should consider germline testing. Case dismissed. Not. Because then, Elena Castro, a very smart clinician-scientist who's an expert in this area, tweeted out last year, a link to her paper, just published. Which Emmanuel was involved with as well, saying that there was no association between germline BRCA2 mutations and the presence of IDCP in a really nice study they did. So, the opposite of what we might have observed in these other studies that were really just hinting towards it.

I conclude about that... Ignore that latter topic for a second. Understand that IDCP is associated with higher-risk prostate cancer. It may or may not be associated with BRCA2 prostate cancer. It's likely still under reported. The way I think about IDCP is really exactly like cribriform. So the first few moments when I spoke to you about the value of it. When we see IDCP, I consider, this is like cribriform. It still has the basal cell layer. So pay attention, clinicians, to the presence of it, and consider it to be a marker of aggressivity while we consider whether or not the BRCA2 thing is real.

Finally, ductal. Ductal is a very specific morphology. It is the commonness of the variance outside acinar adenocarcinoma, but it's still very rare. It is really pure in only about 0.2 to 0.4%. But up to 10 or 12% of patients have mixed acinar adenocarcinoma and ductal. We did a systematic review on this last year and showed that the presence of ductal definitely drove a higher risk of metastatic disease at presentation. They often present with lower urinary tract symptoms, T3 disease, et cetera. Often the PSA is not very elevated, which, of course, is explained by the morphology. This is a recent patient of mine, a doctoral cancer patient. This is a cystoscopy. We're looking in the bladder, and now we're pulling back into the prostate. This was prior to his prostatectomy, and this is the classic appearance for urologists of a ductal cancer. Looks almost like transitional cell carcinoma arising from the [inaudible]. They present with this type of bleeding and lower urinary tract symptoms, often T3 disease.

This is the last paper I'll show you from MD Anderson. Brian Chapin produced this lovely series from MD Anderson showing that when patients have ductal, compared to just high risk, they have poor outcomes from metastasis-free survival and overall survival, whether they have surgery or radiotherapy. So it's a bad cancer. A question is, if you add in ADT in the adjuvant or salvage setting, does that improve outcomes? No, it doesn't make a difference. These patients do badly. They did a nice molecular analysis and just concluded that there is resistance. There are features that make them a little resistant to ADT, and clearly, we need to understand ductal cancers a little bit better.

My take-home messages, as you see, it is rare. It has worse presentation at staging, except for PSA. Poor prognosis. Poor response to ADT, and we do really need to focus on ductal independently in trials going forward. My final message is, these are important. These pathological variants standardized reporting as driven by ISUP are essential. We must work closely with our pathologists, and we must address some of those lingering questions I highlighted in clinical trials going forward.

Thank you very much for the invitation.