6 PI-RADS Lexicon and Atlas

Standardizing Terminology in the PI-RADS Document

We decided to standardize the terminology used in the PI-RADS document. Our next speaker, Kasha Macor, has really led this effort of producing a lexicon so that we can all start speaking the same language.

Kasha is the associate professor of radiology, urology and oncology in the Department of Radiology at Johns Hopkins, and also the James Buchanan Brady Urologic Institute and the Sydney Kimmel Comprehensive Cancer Center at Hopkins.

Overview of Lexicon and Atlas in PI-RADS

Good morning. It's a pleasure to participate in this course, and I'm going to follow Dr. Weinreb's presentation with sort of an overview of what lexicon and atlas in PI-RADS will look like. And also take you through several examples of cases and how the scoring system that Jeff has just described would be utilized.

Let's move on. I'm going to talk about the lexicon terms. I'm going to highlight only several of the terms that are being used because I'd like to focus on imaging components that it's so important to all of us. And I want to again highlight the sectors that are used on the sector map. And you'll see a lot of repetitive themes here.

And our goal throughout this course is to have you familiarized with the concept and also by seeing over and over and over the same concept from more generic sites such as in the atlas and definition part to the technology aspects that will be discussed in the second portion of today's course. And then through some other concepts that will be based on the scoring systems that I'm going to introduce to you.

And I'd like to emphasize that we will be looking at both normal and pathological imaging findings that are being called observations on Multiparametric MRI, definition lexicon level. And we'll review this specific PI-RADS assessment schemes with some case examples.

Prostate Sector Map and Anatomy

Let's plunge back to the prostate sector where I'm showing only 36 out of the 39 components of the sector map with correlation to anatomy from Dr. Hader. You already had an introduction to the anatomy, and you will see that the sector map represents both the sagittal coronal and from the base seminal vesicles base of the prostate, mid gland apex, and the membranous urethra components. All this together form 39 sectors and 36 are part of the prostate anatomy.

You already heard about the important contribution of the central zone, which is well the market on this coronal image, not to be confused with cancer, it's very symmetric, sort of triangular in shape. It typically extends only to the mid portion of the prostate, the ve and on the axial situated image, which is represented on your right side, you will see the central zone right there, very symmetric surrounding the ejaculatory duct. And that is represented on this diagrammatic part of the base, which is labeled central zone cz.

So you will see that we are highlighting important anatomical structures, anterior fibromuscular stroma, which we can see very well on both the T2 and the diagram, the transition zone with those hyperplastic nodules, both the anterior component and the posterior component, which you can also appreciate. And then as we move to the peripheral zone at the base, you can see the anterior portion and the posterior portion of the peripheral zone. Again, you can very see it very well illustrated on the corresponding example, and you will see that on the labeling scheme.

We use, of course, the laterality right and left, as we normally would use on our MR imaging. Same for the scoring table. And also to be consistent with pathology, which some institutions have it reversed. We want to make sure that when we report the location of the abnormalities, it is consistent with the MR imaging and our scoring sector mapping.

We also going to be using all the zones and already alluded to anterior fibromuscular stroma. We are going to look at the transition zone, central zone and peripheral zone at all levels. And you can see the contribution of different anatomical components changes from the base to the apex. And this is certainly represented in the diagram.

So how are we going to take it? The lettering abbreviations that I already mentioned, and I'll just you can see that we are using certain abbreviations for lettering right and left, sort of obvious that you have highlighted there. And they are both represented on the diagram and also represented on this table here. The goal is to be consistent as Jeff already mentioned, and the way the prostate is divided is divided into half posterior and half anterior.

We no longer using 17 millimeters, which was suggested in PI-RADS version one as the length of the core biopsy sample because we recognize that prostates come in many sizes and shapes and we want to be consistent for our sector mapping. Therefore, we dissect the prostate through the mid horizontal line and then of course through the midline to the right and left. And then we have evenly a spaced cortex.

And if you look at this example and that very well defined lesion in the peripheral zone on the right side, which you can really see very well, you may decide how and where are we going to locate it on the sector map. Now, if you recognize that that lesion is being seen at the level of the base, at the level of the peripheral zone, then and also posterior, so we have the base, we have the right side for laterality, we have the posterior aspect of the peripheral zone. You will be able to clearly localize it in a very specific sector, which then will be reported.

Lexicon Terms for Abnormalities

Some terms in the PI-RADS have certain definitions, and I will highlight a few that are of the most importance when it comes to abnormality of findings on MRI. The key concept is going to be is it focal or is it nonfocal abnormality?

Okay. So among focal abnormalities will be we allowing using term of focus, which is some localized abnormality, distinct from neighboring tissues and not quite a three-dimensional structure. We allow using term lesion, which is a pathologic condition that is a definitive abnormality standing out from the background tissue. We allow term of mass, which is going to be a three dimensional space occupying structure, again, different from the background tissue. And we also allow term nodule, which will be classically a small lump.

For a nonfocal abnormality. We are going to be looking at the not localized to a single focus abnormality, and this could be diffused widely spread. This could be multifocal when there are multiple foci that are distinct from the neighboring tissue, and they may be regional, meaning they will conform to some anatomical boundaries such as the prostatic sector, perhaps sex stand in the old terminology, perhaps an anatomical zone or even lobe.

When it comes to abnormality shape, we have several descriptors such as random oval, which will have a certain implications. The important terms here is the lenticular, because this cress presenting lenticular abnormality will be an important finding to recognize on imaging. And you will hear me say that in a few minutes.

Another interesting concept is to have the tear shape water drop shaped pathology, which actually may have certain implication in tumor staging specifically of the cancer at the apex as it extends along the sphincter, the wedge shape abnormality. Having the shape of sort of v pie shape is an important concept that we will try to look for because it will have a certain meaning. Then linear obviously and irregular terms will have some implications.

When it comes to margins. We also are going to be focused on two major concepts on circumscribed being well-defined and non circumscribed, perhaps ill-defined, blurred obscured or irregular. And even speculated in some pathologists that we see in the prostate. This can be actually outlined these three encapsulated organized chaos and the eraserhead sign. You already heard about the the eraserhead sign are important concepts that I want to highlight.

The encapsulated is going to mean bounded by distinct uniform, smooth, low signal line, and will typically imply a BPH nodule. Organized chaos has been introduced in the PI-RADS version one to the peak, the heterogeneous mass within the transition zone that among all the chaos has an organized boundary to it. Again, almost synonymous to BPH nodule. And then in the eraserhead or sign, which will pertain to transition zone, again, contrary to organized chaos, we are going to look at the hyperintense irregular abnormality with blurred margins as if smashed and on the eraserhead drawing.

Diffusion Weighted Imaging and ADC Mapping

So here comes the origin of this definition on diffusion weighted imaging and apparent diffusion coefficient mapping ADC mapping. We are going to be looking for regions of restricted or impeded, which are synonymous in the PI-RADS areas of diffusion, which are known to be indicative of cancer presence. And I will not go into the techniques, you'll hear some more on diffusion as techniques specifications, but we are going to be focusing on both assessment of ADC mapping and high B value diffusion imaging.

We are going to use the terms of hyperintense, iso and hypointense in a traditional fashion. We have been using in both T2 and also diffusion weighted imaging as well as ADC mapping categories. And you already heard about simplification of the DCE and certainly we are going to focus on positive and negative DCE findings.

What is important is to recognize that truly we are only focusing on the early phase wash of imaging because this is going to be this very origin initial component of the contrast the arrival within the cancer tissue that we will be looking at. And when we will be looking at positive DCE, we are going to be looking again at the term focal, very important concept, focal nonfocal as I mentioned already, so for positive DCE focal early enhancement corresponding to focal peripheral zone or transition zone abnormality on T2 and or diffusion weighted imaging, the negative DCE is going to be defined as lack of early enhancement.

So again, remember the early phase of enhancement and focality, or lack of thereof is going to be a very important discriminator. So in negative DCE we will have a lack of early enhancement. There's going to be perhaps a diffuse enhancement that will not correspond to a T2 or diffusion weighted imaging findings. When we have however, focal enhancement that will correspond to what looks clearly by definition BPH, again, it's going to be negative DCE in terms of suspicion of cancer being present.

PI-RADS Criteria

Let's dive into the PI-RADS criteria. So Jeff mentioned to you that for each imaging modality or parameter sequence, we are going to use a scale of five, score of five, and this is going to be defined for both the peripheral zone and transition zone separately. Okay? And then there's gonna be an overall score that will be the PI-RADS assessment score.

So let's go through some imaging examples for peripheral zone situated imaging findings. What will be score one uniformly hyperintense normal tissue. Classic, very easy, what will be two linear wedge shape or diffuse mild, very important concept of mild hypo intensity versus moderate or marked hypo intensity that we are introducing to distinguish abnormalities from benign categories to perhaps indeterminate or suspicious categories.

So for two linear wedge shape abnormalities such like this or perhaps diffuse mild hypo intensity with indistinct margins, all this will be scored as two, as we know. This could be scores, post biopsy, postin, inflammation, atrophy, and prostatitis. Remember we are focusing on T2 peripheral zone. Okay? So we are taking this out of the context of the overall exam at this point. So I want you to pay attention to that. Let's look at three.

Three is really problematic. The three is sort of non four, non two, but we are trying to be more specific. So three will be non circumscribed random or mild hypo intensity, perhaps such as this. There are some benign linears and then there's more confluent, but it's really not markedly distinctively hypo intense it's there sort of occupies certain territory there, but it's just barely mildly hypo intense.

Now, this has somewhat more diffuse. Who knows what's going on there. Maybe there is a little bit of focal abnormality here. Hard to put the margins on it. It's definitively mild. Okay? And it's also close to the ejaculatory duct at the base. You may debate, well, is it the central zone? Is it a true nodule? How about this wedge shape? Perhaps maybe, maybe more localized definitively, not quite circumscribed, not mild hypo intensity. So I can see how some would say, well, it is focal, right? So we said focus is so important, but again, the mild degree of hypo intensity sort of keeps us in that perhaps three category.

Peripheral Zone T2 Scoring

On four. There's gonna be no question that we have more of a circumscribed. We can really put our boundaries around homogeneous and moderate hypo intense focus or mass or nodule that will be confined to the prostate. And we added the size of 1.5 cm, which I think makes a lot of sense in terms of the clinically significant disease as we debated for a long time.

And there is no question in your mind, I'm sure that this is sort of, who knows what that is. It's kind of hypo mildly intense, but this is a very discreet and hypo intense nodule. Moderately hypo intense happens to measure just 12 millimeters. How about this? Very discreet. There are some other hypo intense regions, but this is a discreet nodule less than 15 millimeters in the peripheral zone definitively score four.

How about this? You have an apex, some maybe benign abnormalities, and here you have this hiding, very discreet markedly hypo intense small nodule, left anterior apex peripheral zone, but less than 1.5. We think it's cancer, but we give it four less than 1.5.

So what's gonna be five, similar like four, however more than 1.5 in cm and beyond the size such as this discreet nodule, no question. This is a large mass long abutment along the capsule on the left. Same with this. There is a mass effect at the right base, very hypo intense large over 1.5 cm nodule that now causes mass effect and like in this case, not only mass effect invasive features invading into recto prostatic angle here.

So these are the definitively highly suspicious for clinically significant prostate cancer peripheral zone. Okay, we can confirm coronal images extending extra prostatic. Okay, so these are the clear fives.

Transition Zone T2 Scoring

How about the transition zone? Well, I had hard time to find a homogeneous intermediate signal intensity in a normal transition zone. I did find one, and this patient actually is in his forties, so he has a little bit of the transition zone, mildly hypo intense and very homogeneous. Okay? And then some intermediate diffuse signal in the peripheral zone, but that is as normal as it gets before we get to BPH, which is classically here.

Circumscribed hypo intense or heterogeneous with this hyperplastic nodules with nice rim, dark rim of encapsulated nodule classic for BPH two, bigger organized chaos all over the place. However, there is a clear boundary around this entire transition zone. We know that there is this pseudo capsule compressed between the transition zone and the peripheral zone that we can visualize as T2 dark. So this is a classic BPH.

Now, the three category in the transition zone is again, quite problematic. We see a lot of patients with BPHs and you know, they have some hyperplastic nodules, some dark signal that looks more infiltrative, but it's still confined within the transition zone. And the question is, are this really infiltrative? Are this part of this stromal hyperplasia? And sometimes we are just not sure margins obscure. If it's heterogeneous signal and it's sort of not bounded by a capsule, perhaps three, some even seem like this one, it's at the base, there's BPH nodules, there is some signal. Does it really go to the peripheral zone? Does it violate the zonal anatomy or is it volume averaging? And again, we correlate with coronal. Sometimes we can tell. And remember, this is just situated imaging, right? So we are not yet combining others, so we trying to assign a score here. So if we don't, not sure, okay, it may have obscured margins because it's not encapsulated would be three.

Now four much easier if we find it such as this non circumscribed. That's number one, homogeneous number two, and moderately hyperintense nodule in the right anterior transition zone, less than one cm. There's no question that's cancer. However, we assign score four because it's less than 1.5 cm and perhaps this one doesn't quite go across. Doesn't quite necessarily demonstrate yet invasive features. This one has this sort of ill-defined non circumscribed homogeneous characteristics here. The question is, you know, are these same pathology or are these encapsulated components of the transition zone here? But clearly there is a nodule there that when we measure it is less than 1.5, we give it four.

Now for five, we have very similar findings with homogeneous non circumscribed, moderately hypo intense ill-defined abnormality with the eraserhead sign. Remember the smudged margin sign of cancer transition zone, like this case large over 1.5 cm, okay? Invasive. This has invasive behavior grows across to the other side, okay? And also goes to the anterior portion of the peripheral zone. In this case, you can't even make the transition zone anymore. There is diffusely infiltrative tumor around and extending to the anterior pre prostatic fat through the anterior fibromuscular stroma. Invasive features more than 1.5. That's your transition zone T2 score five.

Diffusion Weighted Imaging Scores

So now let's move on to the diffusion weighted imaging. And I'll have to hustle a little bit. I think I highlighted to you the concept, how we are going to look at this and just present several of the examples now.

So for peripheral zone diffusion weighted imaging score one, no abnormality, okay, on a DC or high B values. So we really don't see much here. Two, there may be indistinct mildly hyperintense on ADCs right there, right? We don't know. Is it part of this tissue around the ejaculatory duct? Is it really real on the high B value? Really no high signal. It's gonna be two such as this linear and maybe a little bit, but clearly benign linear morphology on three focal mildly hyperintense on ADC. Remember we are not looking yet at this. We are looking at the peripheral zone, okay? And somewhat hyperintense or maybe iso intense on high B value in three, we may have a mildly low signal on ADC map and barely maybe hyperintense on high B value on four, we are going to have moderately and focal again, focal peripheral zone, no question markedly abnormal and corresponds to markedly hyperintense high B value and less than 1.5 cm restriction on ADC high signal on high B value four if less one than 1.5 if more and invasive features, we have five. Okay, large tumor in the peripheral zone, very distinctively hyperintense, high B value similar for the diffusion for the transition zone with little caveats. I'll show you in a moment.

We look at the ADC map, really not much there. Nothing shows up. Always look at T2. Okay? Very important that for transition zone, we always come back to the T2 for correlation. For T2, we have a sort of BPH nodule, ADC low all over the place, nothing really focal, nothing really shows up, maybe, maybe not, doesn't quite correspond on high B value. Again, look at T2, which looks clearly BPH, organized chaos organized there three more problematic. Again, we may have some regions of restricted diffusion, some more geographic within the encapsulated nodule. There may be or may not be. Sometimes we can tell we need to look again on the situated image and see if any of this dark signal corresponds and if we cannot tell, probably it's going to be three.

Now this is a very distinct abnormality. Not only it has focality to it markedly hypo intense on ADC map. Okay, hyperintense on high B value and lenticular shape invading tumor in the anterior fibromuscular stroma here, less than 1.5, score four. Okay, you look at T2, remember there's this tumor, not very hypo intense, but when you put this all together that is different than the rest of the BPH there for five, you have the this restricted diffusion throughout clearly invasive, very high signal on high B value and clearly invasive tumor actually extending beyond the level of the prostate.

DCE Examples

For DCE, remember, positive is focal early enhancement corresponding to a focal abnormality. And that's what we are going to have for peripheral zone. The hyperintensity on early DCE post contrast injection. You can use color maps to show it in the peripheral zone, focality corresponding to diffusion weighted image or, and or T2. Another case for transition zone. This time we have little nodule right there anteriorly. Now it's very hard to say, well it's part of BPH, is it not part of BPH? It's certainly focally positive here you look at T diffusion, no question. That is cancer. Now co negative, both peripheral zone diffuse mild enhancements, some BPH. You look at the color map all over the place, a little bit here, a little bit here. Most importantly, when we look at the T2, nothing peripheral zone and some clearly BPH nodule like pathology in the central gland.

Overall PI-RADS Assessment

So the overall assessment that Jeff mentioned already is going to focus on likelihood of the presence of clinically significant cancer based on the individual parameters that we are going to combine into this assessment one through five as he described. And just two examples quickly how we can put all this together. And again, you will hear more examples throughout the course.

So for the overall PI-RADS assessment, we are going to take all these sequences together. Now, this is the your scoring card. Okay? That's the scoring card that you already saw. That is a scoring card for peripheral zone where the diffusion is your dominant scoring. So you have diffusion, ADC, high B value, T2 and DCE example. We are going to be looking at the focal nodule right there on the left, high B value image focal T2 and positive. Okay, so here it is, the nodule measured 11 millimeters. Remember, size is important. So these are the findings we're going to put together. And if you look at the score, just think of it, where would you put this? 11 millimeter nodule. And if you set four, that's it, it's four, the looks cancer, doesn't matter what other show it's going before, happens to have some matches on the other sequences, which only make our confidence higher.

Example for transition zone, little tougher this time, this is your scorecard for transition zone PI-RADS overall assessment, example T2, high B value ADC, and post contrast kinetic map. And also early post contrast image. You will see on the left that there is a large, somewhat circumscribed, but T2 dark nodule that demonstrates focal restriction, no question about it. Also corresponding positive DCE. Okay, so the question, how are we going to deal with this? Is this encapsulated extruded BPH nodule that has all this worrisome findings. So this is a one of the challenging cases. We measure this and this is 22 millimeters. So now the only thing that doesn't fit here, it's not a sharp eraserhead here, okay? But this is a 22 millimeter nodule. T2 dark focal restriction matches all. So what to do with this?

Well, I decided that this may be one of the problematic cases. So the score actually in this one, in my opinion and my experience, if I see encapsulated but fitting all the criteria, and I know there's disagreement about it, I would really think this is perhaps five. If you go by size or four, if you go by three, I think you know, you really need to put five. So you'll get to four category. So I think this one may be anywhere in this re Now I said what I said based on just example. And that alludes to to, you know, in consensus this happened to be also biopsied by the MRI fusion. That's another vendor that we use at Hopkins, that performs this fusion. And this happened to be prostatic duct adenocarcinoma, four plus four, clearly corresponding to this nodule.

So with that, thank you very much.