3 Pathology and MRI Correlation

Introduction

Massou is the professor of radiology at the University of Toronto, clinician scientist at Ontario Institute of Cancer Research, senior Scientist at Sunnybrook Research Institute. And he's now the chief in the department of Medical Imaging at the Sunnybrook Health Sciences Center in Toronto. And Massou is going to talk some more to us about pathology and MRI correlation.

It's a great pleasure to be here. I'm gonna cover in the next 40 minutes or so, we need to know in every disease area, which is rad path correlates. It is something that is not intuitive. It's not like a renal cell cancer or metastatic disease elsewhere. When you talk about prostate cancer, you really do need to understand these correlates.

I have no conflicts of interest or other disclosures.

Prostate Anatomy

I think when you read about prostate anatomy, one of the challenges is that the literature going back the last three, four decades can be a little confusing. And there's been a general consensus that when we talk about the prostate and the various areas and zones within the prostate that we're gonna stick with McNeal's anatomy. So I'm gonna go through that first. And there are four distinct zones. The peripheral zone, the central zone, and the central zone is the one that I think is one of the most confusing concepts. It gets mixed up with some other terminology. The transition zone, this is where you see a lot of the BPH nodules and anterior fibromuscular stroma. So these are the four distinct zones, and we'll go through a few images now to go through that anatomy.

So here's a relatively young man's prostate, axial coronal images, T two weighted. And what you see here is, first, I just wanna show you the ejaculatory ducts. You can see these pretty much with or without an endorectal coil on every MR if the patient hasn't moved very much and they're coming down from the seminal vesicles into the verumontanum.

Then we have the central zone. The central zone is an area of tissue that is around the ejaculatory ducts, often a little lateral to the ejaculatory ducts. And it is a slightly dark T two signal? It's quite variable, but it's often mildly dark to quite dark on T two weighted imaging. And this makes cancers in this area more challenging to pick up. It's also an area that exhibits more signal abnormality in prostatitis. And so this is one of the more challenging areas we have in a source of potential false positives when you're reading prostate MR. So if you're new to prostate MR you might be calling a lot of false positive reads.

The third area is then where most of the prostate glandular tissue is, which is in the peripheral zone. And this is that C-shaped tissue along the posterior aspect. There's more of it at the apex relative to the transition zone as you get more BPH. And then we have the transition zone itself, where we see a lot of the BPH nodules occurring. This patient also has a anatomic variant, which is not that infrequent, which is a prostatic utricle cyst. This is a mullerian remnant, and something that you should just be aware of as a potential normal finding often is just a T two bright cystic looking thing. It can occasionally have hemorrhage or other things in it.

But now let's go through the same anatomy axially. Again, just multiplanar T two weighted imaging here. And as we come down, you can see the peripheral zone dominating in a young patient at the base, but in older patients, again, you'll see more BPH. And as we come down, you'll see this continuation of the detrusor muscle becoming more fibrous. And this is the anterior fibromuscular stroma, this area here. And then you have retropubic veins. And so a lot of veins, and I'll show you some more pictures of that around the prostate.

So this is the anterior fibromuscular stroma, this extension of fibrous tissue from the urinary bladder along the anterior margin of the prostate. Notice how there is still a cuff of vessels in front of it that you need to be aware of. And as we come into the most inferior aspect of the prostate, a lot of peripheral zone here.

So let's look at this coronal image and just show you some of the structures around the prostate. So as we look at T two weighted imaging, what you'll see if we look on this side is all these worm-like structures. And one of the problems, of course, with T two weighted imaging is that slow flowing venous blood will look very much like the unsuppressed fat. And so you need to be aware of these cuff of veins all the way around the prostate. And so again, this periprostatic or venous plexus can be a source of false positives in your reading.

Second thing is, as you come towards the base of the prostate and you're looking at the neurovascular bundles, you'll see this sort of flaring or feathering of these veins going out, and important not to confuse these as you're looking at them. Axial volume averaging through venous walls with extraprostatic extension. And it's really minimizing these false positive reads that's gonna bring value to MR in the current clinical context.

Levator ani muscles cuff either side of the prostate. And you'll hear the term prostatic capsule used. There isn't a true capsule of the prostate. This is just some compressed tissue at the margin of the gland. So there's not actually a true prostatic capsule. So that dark line you see on T two is not a true capsule.

So in talking about these veins and diffusion weighted imaging, one of the things you need to be aware of is, particularly at the apex on axial imaging, you'll see these large structures in many men on either side of the gland. And the prostatic capsule at the apex, or this I'll use the word capsule in quotes, is not particularly clear. And so these veins can exhibit fluid fluid levels in them. And this is thought to be related to a hematocrit level that develops with the very slow flow. You'll actually see this in many men in the corpus cavernosum on the pelvic MRs. You'll see this fluid fluid level there. And the problem with that is that actually the ADC of that hematocrit is actually quite low. So on diffusion weighted imaging, you can sometimes call a false positive if you're looking at a small vein along the margin at the apex in particular. So just be careful, be wary of that. Look at your multiplanar T two weighted imaging when you're looking at those, particularly at the apex.

The other thing that I think may become of more relevance in terms of T two anatomy is the anatomy of the sphincter. The voluntary sphincter of the urethra. And it looks like you'll see this on the coronal images. You'll see it as a donut on T two weighted images. And this is it here, these two dark bands. There's a varying degree of intraprostatic extent of this sphincter, and there's at least one publication describing how the varied anatomy in this can affect continence after radical prostatectomy. So I think you should at least be aware of what the sphincter looks like. It may grow in importance as prostate MR continues to get widespread adoption. So this is the urethral sphincter.

So that's some basic anatomy. And focusing on the central zone as a T two dark area that is really posterior to the transition zone around the ejaculatory ducts as one key point to emphasize, and really these periprostatic veins as a potential pitfall.

Gleason Scoring

Now, what about Gleason scoring? And Dr. Choyke touched on this already. I think it's very, very important that we understand what's going on with Gleason scoring today and how it's changed over time. What's happened is that there has been a change starting in 2005 with the way pathologists do Gleason scoring. And what's happened is that there's been a more restrictive definition of what Gleason three pattern is. And so things that in the past would've been called Gleason six cancer, pre 2005, are now being called Gleason seven cancer, particularly three plus four types. So this cribriform pattern, you'll hear that term used by pathologists. Many of the cribriform types have moved from Gleason three into Gleason four post 2005, 2006.

That's in a sense, good news for us in MR in the sense that the threshold of Gleason seven disease, Gleason score seven disease, is very important, and potentially more valid as a threshold. Because some of the bad Gleason six actors have been culled out with this new scheme. We need to learn more about this, but just keep this in mind. And retrospective studies now done have essentially shown that patients who get radical prostatectomy for current classification of Gleason six disease at the time of diagnosis don't die of prostate cancer. It's really when they have Gleason seven under the new scheme that you're really at threat for that disease specific mortality. So very important to keep this in mind. Have a discussion with your pathologists, and radiologists. Being on the same page with the Gleason scoring is incredibly important to building the confidence with your urologists that you're doing the right thing. If you're on a different page with your pathologist, there'll be a lot of heterogeneity in the rad path correlates that you're getting back. And this could lead to problems.

Now, as we heard from Peter, the single most important prognostic factor, the major prognostic factor remains Gleason score, which is a summation of Gleason grades. And the biggest sort of innovation we have coming forward in the last few years with multiparametric MR is this negative correlation between ADC and Gleason score. That means the higher your Gleason score, the more restricted the tumor is, particularly in the peripheral zone. And so this has allowed us to tell urologists and our oncology colleagues that yes, this cancer is something that is now hitting the Gleason seven threshold. You need to do something about it. It's very important. So that's a key point. You've heard it once already. I'm just emphasizing it again.

Prostate Cancer Distribution

Prostate cancer distribution. This is very important again, for us to recognize. 'Cause there will be a difference in the look of your MRs and where the cancers are, depending on whether you're seeing a patient pre-first biopsy, post first biopsy with high PSA or on active surveillance or pre-op. And the principle difference is that if they've never had a biopsy before, the vast majority of cancers, more than 70% sit in the peripheral zone posteriorly, usually in the mid to apical gland. And this is where standard systematic biopsy is usually done. So if a patient hasn't had a biopsy yet, PSA is not a horrible test. And so their PSA went up, they saw their doctor, they saw their urologist, they come for an MR. You're gonna see a lot of posterior cancers in areas that would normally have been picked up by a biopsy. So key point here is that in biopsy naive patients, posterior cancers and peripheral zone cancers are the commonest location that you're gonna see.

Transition Zone Cancers

What about transition zone cancers? Transition zone cancers are something that we need to know how to interpret very, very well, because the one of the commonest indications we see is patients with discordance high PSA repeat negative biopsy. And a lot of the cases you're gonna see over the next two days are in fact that their patients with, you're gonna see these big anterior cancers, and they certainly do occur. Up to 30% of cases are non peripheral zone cancers. And we're gonna see a lot more of those in our referral base than a urologist might see in their general referral base going to systematic biopsy. So vast majority peripheral zone in biopsy naive, and in the general population, but in our cohorts, especially with high PSA negative biopsy, more anterior distribution. And we can see why I won't spend much time on this slide, but if the biopsy pattern is this, this is where the misses are gonna be.

Transition Zone Tumors vs. BPH

So what about the nature of transition zone tumors? We need to be able to distinguish these from BPH. BPH, benign prostatic hyperplasia, or nodular hyperplasia as a pathologist would refer to it, is something that can be a challenge sometime to distinguish from prostate cancer. And T two is king here. So diffusion weighted imaging is king in the peripheral zone. T two is king in the transition zone. None of these is absolute, but you need to look at everything. But really what we want is we want an ill-defined margin for prostate cancer in the transition zone on T two, we want a lenticular shape, amorphous looking tumor. Of course, if you see extracapsular extension, that makes it easy. And this idea of local invasion, it looks like it's chewed through the other nodules. And compressed fibromuscular stroma.

So this is a typical couple of cases, PSA high prior negative biopsies. And we can see the tumors. There are lenticular, amorphous, locally invasive, extraprostatic extension just on T two weighted imaging alone. And this has been described as the charcoal smudge sign. Other, there's other terms that people have used. It looks like you sort of wet your thumb on a charcoal drawing and smudge through that architecture. This patient moved a lot during their exam. And what you can see here is how even though T two is king in the transition zone, looking at all the parameters are gonna help you come to more definitive diagnosis. So just to emphasize that.

False Positives and Pitfalls

Now, what about BPH? Let's talk about not showing case of cancer after cancer after cancer. But let's talk about the false positives. This is one of our biggest challenges in the learning curve for interpreting prostate MR is getting around false positives and getting that confidence to say, this is not an equivocal call for prostate cancer. So we face some challenges with BPH, when we look at this case here, typical BPH patient, big transition zone compression of the peripheral zone here. What we see is a number of round nodules. These the ones that are easiest to interpret, have this dark border around them. So very well-defined. We're gonna call these benign nodules. Nothing to worry about, regardless of their other appearances on diffusion and T two, there are other nodules that sometimes look like they're sitting within larger nodules. They sometimes don't have a boundary. And when you start to look at the diffusion, you say, well, there's all these areas of restriction in here that are looking pretty bad on these ADC maps. This is the cancer, this is the index tumor, this peripheral zone cancer. But what about all this stuff in here?

And this is one paper looking at the ADC values of cancer on the far left stromal hyperplasia, which is a type of hyperplasia that occurs in these BPH nodules and glandular hyperplasia on the far right here. And there is a scale here, a sliding thing. So the idea that stromal hyperplasia and inflammation will cause restriction is very important. And here's what's going on pathologically. When you look at a BPH nodule, you can have anywhere from a lot of these cystic spaces and glandular proliferation on one extreme to a lot of proliferation of the stroma without much glandular hyperplasia. And you can have anything in between, or you can have inflammatory cells mixed in. This one will look more T two dark, the one on the left, and will exhibit a more restricted pattern on diffusion. The one on the right will have more T two bright areas exhibit less restriction. And again, anything in between. So you can have mixes, and this is the challenge we face in interpreting BPH, and you'll hear a lot more about interpretation of the transition zone in the next day and a half.

Inflammation, Fibrosis, and Granuloma

What about another false positives? Well, fibrosis and granuloma with inflammation, or inflammation in general is a very common pathology finding on biopsy. Here's an extreme case of what looks like a really ugly tumor, maybe a hint at its inflammation and it's very angular shape, but I think you'd be hard pressed to call its inflammation out of the gate. This is granulomatous inflammation. The prostate margin is ill-defined. This looks pretty ugly. This is all granulomatous inflammation and fibrosis, this degree of inflammation you will see more frequently in patients who have been on BCG therapy for bladder cancer. So something to be aware of. Do not interpret prostate MRs without getting the biopsy results. It sometimes is like pulling hen's teeth to get the biopsy results. And the PSA from your referrers, if you don't do it, you are going to have problems.

There should be positive systematic biopsies for cancer on the left side extensively. If this was all real, so this was all negative, what about inflammation? Can ADC help? And so there is a trend here in terms of distinguishing prostatitis from high grade prostate cancer on this graph on the right. These are all inflammatory cells infiltrating into the prostate on this pathology image. But there's a lot of overlap. And so you're gonna see examples where we're relying very much on morphology and nodularity and perhaps a little bit of ADC thresholding to try and tease out prostatitis as we move forward. But it is a challenge. Again, the history of a known confirmed acute prostatitis or chronic prostatitis is important in not over calling.

You will hear other terms. And I remember when I first started, I didn't know what these terms were, and what to make of them, until I started to realize, and that pathologists are like us. They need hedging terms. And that's what ASAP is. Atypical small acinar cell proliferation. This is where the pathologist thinks it's a cancer, it might be a cancer. And that's the way I look at it, that they're really not sure. And so they call ASAP. So if your patho, some pathologists don't use that term at all, but if your pathologist is using that term, it doesn't mean there's no cancer there. It means you need to look really hard because he or she's pretty worried about it.

What about high grade PIN? High grade prostatic intraepithelial neoplasia is associated with, or thought to be associated with a higher risk of harboring prostate cancer somewhere in the prostate. It's become more controversial recently as to what the strength of that association really is. The nature of the imaging changes associated with high grade PIN are poorly understood, I would say at this point. But so you will see these terms. They are things that should push up your risk factor profile in interpreting the MR and looking at the likelihood of significant disease.

Multifocal Disease and Index Tumor

Well, we've already heard that prostate cancer is multifocal. And this has always been the principle argument against focal therapy against everything we're doing with MR is, let me get this straight. Radiologists. You want to take a cancer that is multifocal in 80% of patients, and you want to use imaging to tell us that we should only be doing targeted biopsies and we should only be doing focal therapy. I would say that the power of imaging has been such that in the last number of years as this has come forward, that the whole paradigm has shifted as urologists have understood, and radiation oncologists and others have understood what they're missing by not looking at images. And it's this is still a work in progress, but I think there's been a big change.

But you need to know this for interpretation, is that the average number of cancers, this includes Gleason six cancers now is more than one. It's just over two. So seeing 'cause you see one cancer doesn't mean you shouldn't be looking carefully for a second cancer. But there usually is a dominant focus. And this is our primary role. It's okay to miss a tiny cancer somewhere in the prostate. What you don't wanna miss is the dominant driving cancer for that patient's outcome. And this is our job. Find the index tumor. I'll talk a little bit tomorrow. I'm gonna talk about active surveillance, about our second, I think our second major job is, but our first one is to not miss the index tumor. Extraprostatic extension is usually from the dominant focus.

So I already talked about the index lesion. So here's a case, and just to show you what happens in a typical read. So endorectal MR. Can see a prostate cancer. We're all gonna call this, this is the ADC, this is the permeability map. Notice how it's not necessarily the same size on the permeability map as the tumor, and but we see just on the permeability map alone, another little here. And boy, it's really hard to know if that's a real finding. And then there's a whole bunch of other stuff that's nonspecific, but I'm gonna call this one for sure. PIRADS five under the old classification. Four under the new classification. But here's the pathology, but that's okay. And as long as the urologist and you and the pathologist understand what our mission is and what we're doing, the fact that we picked up the dominant lesion and probably gave a PIRADS three or four score to the second lesion is good enough.

This was sort of a another thing that we came across through the evolution of interpreting prostate cases and understanding this difference between low and higher grade tumors. This is an old case now, T two weighted imaging. Nothing ADC map, nothing early wash in symmetric enhancement in the peripheral zone, but really nothing focal here, no index tumor. This would be a negative case. Nothing to call. But then we were doing a study and this comes back, what's going on here? How can you have such a big cancer? And that's where all this subtlety started to come in about the nature of the spread and growth of prostate cancer. And this is now well recognized. Many people are talking about this. And in fact, prostate cancer, we are not just talking about the malignant glandular elements. We're talking about an interaction between the prostate cancer itself, the stroma, the surrounding fibromuscular tissue. And when there's loose stroma and benign glands and loose stroma around tumors, and they're sparsely, sort of like a like we saw a salt and pepper shaker earlier. So imagine you just had salt and pepper glands, malignant glands spread out throughout the stroma. These are potentially these are more invisible to T two weighted imaging and diffusion. And this is very important concept.

When we look at malignant glands that are densely together, that's below the red line. These are we all gonna see fibrosclerotic areas around malignant glands and dense stroma around malignant glands. We see all of this stuff on MR very easily, but when we've got loose stroma and sparse glands, it's more problematic. And so think of prostate cancer. Next time you're in your kitchen and looking at your nice marble or granite. Think about it like that in terms of the way it can grow and the way it can spread. This is important for focal therapy. And here's just a selection of white areas on pathology that have this sparse pattern that don't really show up on MR very well, particularly on T two and diffusion weighted imaging. So most cancers do have regions that will show up, but their boundaries certainly are not well defined.

Tumor Volume and Margins for Focal Therapy

So how are we gonna measure index tumor volume? And the way that people are thinking about this now is not to do direct measurements of volume, but to talk about thresholds. And so another key point I wanna make for this talk is to say, if you're gonna take the half cc threshold for a tumor volume at pathology, MR does quite well sensitivity. Multiparametric MR sensitivity 88% specificity 95%. So make sure urologists understand that this is a size dependent diagnosis you're making as it is in many areas of the body. And that half cc or larger tumors is certainly one threshold you can pick and a good one to pick, because it's well recognized by the urology community. The other one you can pick is just a maximum diameter of the tumor. And you'll hear different things from different people. Four to five millimeters would be one.

And then I think to finish off, I just wanna talk, 'cause this course is gonna talk about biopsy and focal therapy. And this is work from London, Ontario in Canada. Others are doing this work as well, is if we want to tell a radiation oncologist or we want to tell a urologist who's gonna deliver focal therapy, what margin they should put to eradicate the whole tumor, how are we gonna do that? Because I just told you that we don't necessarily see the true boundary of the tumor. And anyone looking at the MR will wanna draw a line right around that tumor. If you do that, you will miss components of the cancer within that focus. We don't know what the significance of missing those is, but certainly the desire is to eradicate. And so I'm just gonna go back one if I can just to show that graph.

So what I wanted to show you here was that this is where the DCE may rise back in importance. We don't know yet, but certainly the T two and ADC combined give a margin. And we may find that the DCE adds some rationale. It's still up in the air, but these margins are big anywhere from five to 10 millimeters to have a greater than 90% probability of encompassing the entire tumor. And this has implications for regional treatment. So we need to start thinking in terms of GTV CTV and PTV, like our radiation oncology colleagues. One approach we're taking with one technology, which is transurethral delivered energy, is to actually recognize that the growth patterns are often asymmetric in the peripheral zone and to give about a seven to 10 millimeter margin on either side of the tumor as we treat peripheral zone lesions.

Summary

So to summarize, the prostate has three separate zones and anterior fibromuscular stroma with distinct MR appearances. The central zone is the one that you need to make sure you're up to date on the peripheral zone in transition. Zone have different background appearances and require different methods of interpretation. Prostate cancer is a multifocal disease, but there's often a dominant index tumor. And the lower the ADC, the higher the Gleason score. And finally, Gleason three cancer at that moment in time for the patient. If it's Gleason six that is, its only Gleason three, cancer is likely clinically insignificant. And then watch out for the temptation to overcall the margins.

Well, thank you very much.