Transcription
Live from Boston, my name is Emilio Madrigal and today is Tuesday, September 21, 2021. I'm remotely joined by my good friend and colleague, Refat.
Today we continue our GYN-OB pathology Pathcast series, and this is the 16th lecture of that series. If you want to see all the other lectures, please feel free to visit our website, that's pathologycast.com, and you'll be able to see all the series and all the lectures for this GYN and OB pathology series on that site.
Today we welcome Dr. Andre Pinto. He's an associate professor of pathology at the University of Miami Miller School of Medicine, and the title of his talk is "Endometrial Stromal Neoplasms." As always, please feel free to post your comments and questions in the Facebook and YouTube chat windows, and Refat and I will be sure to pass those along to Dr. Pinto at the end of the lecture.
With that, I'll turn the microphone over to Dr. Pinto.
Afternoon or good morning, good evening, wherever you are. Thanks to the Pathcast team for the invitation to present here today. My name is Andre Pinto; I practice gynecological pathology at the University of Miami, and I'm happy to be speaking to you all today.
The topic of my lecture today is endometrial stromal neoplasms. I'm going to be honest: when this was suggested to me, I thought, "I don't think I can make an hour's worth of endometriosis tumors." However, I guess I put a lot of pictures here, and I'm going to try to focus a lot on morphology, image chemistry, and some of the differential diagnosis, and I hope that this will be helpful for you.
So let's start with a case presentation, with a case example.
We have a 57-year-old female. She had a history of hysterectomy in an outside hospital for a "quote-unquote" uterine tumor, so we're not sure what the pathology was at that time. She came to our institution presented with multiple pelvic nodules which were suspicious for recurrence of this tumor. What I'm going to show you next is the biopsy of these pelvic masses or pelvic nodules.
This is a medium power view of this, I'll say, spindle to ovoid tumor composed of cells with scant cytoplasm. You don't see a lot of pleomorphism, so the cells look, I would say, quite uniform throughout. There are some blood vessels evenly distributed; I don't see much mitosis, there's no necrosis, so there are not really many features of a high-grade tumor. However, other areas of the tumor did show, as you can see here, brisk mitotic activity, some enlarged nuclei; some of the cells look more epithelioid. The nucleoli in this case show a biphasic or two different morphologies in that tumor.
When we did immunohistochemistry, what's interesting is that depending on the area that you're staining or you apply your IHC, we had different results. In the area that looks more spindle to ovoid and monomorphic throughout, this one I'm listing here is just the most relevant I see. We've done more. So the tumor was keratin negative, TTF-1 negative, and of course, we always throw back stage as gynecologic pathologists. But in contrast, it was ER strongly and diffusely positive, CD10 positive, CK negative, and the reason I'm showing CK is I'll tell you in a bit.
Desmin was mostly focal, with some areas of patchy positivity; H-caldesmon negative and Cyclin D1 negative. The areas that look more epithelioid, in contrast, also show keratin positivity but hormone receptors such as ER were negative, CD10 negative, CK positive, Desmin caldesmon negative, Cyclin D1 strongly positive approximately 80 to 90 percent of the cells.
So with this index case, I think the questions that arise are: what's your diagnosis, and how common is it to have these two distinct IHC profiles in the same tumor? We don't see that very often; usually, we see tumors that we stay in different blocks, and they tend to have the same profile as well. How often does this happen in individual stromal tumors, and is there any other testing that we could do to just be definitive and confirm the diagnosis?
So with this in mind, let's go a little bit of background on endometrial stroma neoplasms according to the WHO criteria. This is the current classification. A lot has changed in the past decade or so with endometrial stromal tumors, but what I'm showing here is the most current classification. We divide those tumors into endometrial stroma nodule, endometrial stroma sarcoma, or I'm going to be referring here as ESS or low-grade ESS. High-grade ESS, which is in the endometrial stroma sarcoma category, is subdivided into low grade and high grade, and pretty much we're relying on the molecular abnormality as the signature of these tumors.
The low grade most commonly shows a translocation involving the JAZF1 gene, while the high grade involves what I'm going to refer to as Y-H. The third category involves B-CORE alterations, and it's alterations because those alterations could be fusions like translocations or it can be mutations, ITDs, or internal tandem duplications.
Finally, there's a diagnosis of exclusion, which tends to be kind of a waste basket. After you rule out all your endometrial stroma neoplasms with all the tests, if nothing stains, if nothing kind of predicts that this is endometrial sarcoma, then we classify them as undifferentiated endometrial sarcoma, which is the preferred term.
With that in mind, as a whole, endometrial stroma neoplasms are the second most common uterine mesenchymal tumor, with 20 to 30 percent of the mesenchymal tumors being ESS or ESTs. As we know, if you practice gynecological pathology, the most common by far is smooth muscle tumors, benign or malignant, and we quite often have to rule those out first before we entertain the diagnosis of an endometrial stromal tumor. It can appear at any decade or any age range, but most commonly in the fourth and fifth decades.
These tumors have unique clinical-pathologic features. For that, I mean that we’re not really grading these tumors into low grade or high grade based on just the cytologic appearance. Actually, these are very distinct tumors in the sense that they have different IHC profiles, they have different molecular abnormalities, and different prognostic outcomes.
So that’s why it’s very important to classify correctly. In terms of prognosis, if you really make the diagnosis of endometriosis stroma nodule, that’s a benign diagnosis, while lower grade endometrial stromal sarcoma tends to have an indolent course, while the high-grade endometrial sarcoma can be and is usually more aggressive. Finally, undifferentiated endometrial sarcoma tends to have a very dismal prognosis.
Let’s talk about each one of them separately.
Endometrial stroma nodules, or ESS, the way these tumors are morphologically presented, the cells that compose these tumors consist of ovoid—very similar to the first picture I showed you—the ovoid to spindle cells that are very similar, if not identical, to your proliferative endometrial-type stroma. As you can see here in the lower picture, if I put some glands scattered throughout, you probably wouldn’t even call that a tumor; you’d think this is just PE-type stroma. That’s how these cells are cytologically, that’s their appearance.
What’s also characteristic of these tumors is to have this prominent small to medium-sized vessels that are consistently dispersed throughout the tumor. One can only make the diagnosis of an endometrial stromal nodule after really making sure that there’s no invasion, which could be lymphovascular invasion or myometrial invasion, which we’re going to discuss in a little bit. These tumors are benign.
What classifies invasion? What’s the original criterion for endometrial stroma that we still apply nowadays? This came from Dr. Tavassoli and Norris back in the day when they allowed up to three projections of up to three millimeters each, meaning if this tumor has five or six or if this projection has a tongue-like infiltration, or they exceed this, technically, you could not call this endometriosis. Another criterion is that no lymphovascular invasion is allowed, so either one excludes the diagnosis of endometriosis. That’s why sampling and making sure you have an excisional specimen is key in order to make this diagnosis.
When we say that, it's because if we do have infiltration into the adjacent myometrium, if we have lymphovascular invasion, then we kind of move away from the diagnosis of endometriosis from a nodule. Now we’re talking about low-grade endometrial stroma sarcoma or grade ESS. Cytologically, it’s almost impossible, if not impossible, to distinguish from individual s t romal nodules.
However, like we discussed about the criteria, they commonly have a very infiltrative type of growth, which tends to be tongue-like, sometimes finger-like, or sometimes even blunt invasion into the adjacent myometrium. Lymphovascular invasion tends to be very easy to find in excisional specimens, which makes the diagnosis of low-grade endometrial stromal sarcoma.
Here I put a little note that size and mitotic activity is irrelevant, meaning you could have an endometrial stromal tumor that would be eight centimeters; it doesn’t matter. It’s still classified or qualified as an ESS if you don’t see invasion into the adjacent myometrium and/or lymphovascular invasion.
Here are good examples of this entity. In the first picture on the upper left side, you see this very characteristic tongue-like infiltration into the adjacent myometrium. On the right side, the same exact morphology that we saw for endometrial stromal nodules, meaning these cells resemble PE or proliferative endometrial-type stroma, and they commonly have, if not, it’s very characteristic, these small arterioles or small vessels consistently present throughout the tumor.
As you can appreciate in this picture, lymphovascular invasion tends to be very easy to identify, as you can see here as well. A few more pictures show lymphovascular invasion—those nests of tumor cells within the capillaries of the myometrium. This is a low-power picture in which you can appreciate both of those features: a very primitive infiltration into the muscle, and also manifold lymphovascular invasion.
Sometimes it’s even difficult to determine if this is LVI or VSI, or if this is actually nests of invasive tumor, but I would say that probably these guys are all lymphovascular invasion; they’re shaping around the vessels. All these here are infiltrative into the myometrium, and here’s a high-power picture to appreciate or to show these nice round capillaries, or small arterioles, present throughout the tumor—very characteristic of low-grade ESS or endometriosis.
One issue that we have, actually, there are two issues I would like to bring up, is how to approach biopsy samples. The first question, the first problem, is how to make sure that you’re looking at in a biopsy sample an endometrial stromal tumor versus, for example, a crushed polyp or just pieces of fragmented benign endometrium.
The key here is to see that once you see a lot of fragments—like a lot of really tissue composed almost always or only of endometrial stroma—that’s a sign that it’s probably not your benign endometrium that got fragmented. Quite often, you see this expansive or sheet-like growth, which you tend not to see that in the normal endometrium.
One key to distinguish that from the polyps is that you’re going to have those uniformly placed glands—benign glands from an endometrial polyp—that point towards that diagnosis rather than an endometrial stromal tumor.
The second issue is: okay, so I’m uncomfortable that actually this is an endometrial stromal nodule—I mean, I’m sorry, an endometrial stromal tumor. How do I know if this is a nodule, like a benign versus a low-grade endometrial stromal sarcoma? I would say that in most cases in endometrial biopsies or curates, that diagnosis cannot be made because, if you recall from the criteria, we need to see adjacent myometrium to qualify for invasion.
If vascular invasion is also key, which if you just have pieces of tumor, you may not see vessels adjacent or far away from the tumor. So in those cases, most authors would just recommend to be descriptive, write a note that if this is an endometrial stroma neoplasm, and in the comment, say that you really cannot distinguish in small biopsy samples between an endometrial stromal nodule versus low-grade endometrial stromal sarcoma, and this will be best made at an excisional mini and hysterectomy specimen.
Immunohistochemistry—CD10 and hormone markers, meaning estrogen and progesterone receptors—are the most classically used markers for endometrial stromal sarcoma or endometrial stromal neoplasms, in general, and it’s really diffuse and positive in the low-grade variants.
You can also use other markers such as WT1; Vimentin is just a non-specific mesenchymal marker. It’s important to know that muscle markers such as Desmin can be positive in endometrial stromal tumors, and that doesn’t make it into the diagnosis of a leiomyoma or leiomyosarcoma.
You have to have in mind that quite often you have to use a panel to distinguish them both, and those muscle markers tend to be stronger in areas of smooth muscle differentiation. The same goes for sex cord markers: if you have areas with a low-grade endometrial stromal sarcoma with sex cord-like differentiation, which I’m going to show you a few pictures, then those areas will stain, for example, for keratin or inhibin.
So that’s something also to keep in mind. Another marker that has been described and validated in the literature is the IFI TM1. I’m not very... we don’t have that in our lab and I don’t have a lot of experience, but there’s some good data out there showing that this is actually a reliable marker for endometrial stromal tumors—low-grade and sometimes even superior to CD10.
So what are the variants that we need to be aware of in order to make the diagnosis? There’s smooth muscle differentiation, which creates this starburst appearance. There's sex cord-like differentiation, and fibers of fibromyxoid differentiation. The fibrous variant ESS with glandular elements, and there are many others—skeletal muscle, pleomorphic; there are many others, but these are the most common ones.
Here’s an example of a starburst or low-grade endometrial stromal sarcoma with smooth muscle differentiation. So this nodule in the center has a highly nice core, and there’s more epithelioid cells. This is a smooth muscle differentiation. If you apply your Desmin or Caldesmon, it tends to be diffusely positive. One has to be aware because sometimes they are very exuberant in a case, so make sure to see that to make the correct diagnosis.
As you can see here in the periphery, we’ll still have the conventional or spindle-type low-grade ESS, which facilitates the correct diagnosis.
Sex cord-like differentiation—so as you can see here throughout this picture, you have areas of more of a trabecular, vague organoid growth amongst these fibrous plaques, which can also be seen in endometrial stromal tumors.
So this is something that if you use your inhibin or your keratin or your SF1—those areas will stain, which proves that that’s some sex cord-like differentiation.
I guess the differential diagnosis here would be, for example, a uterine tumor resembling an ovarian sex cord tumor, which, by default, those tumors cannot have an ESS component.
So once you have a combination of ESS plus these areas that make it into a sex cord-like differentiation of endometrial stromal tumors, the fibromyxoid variant is not more than just a more fibrotic or more highly myxoid background of the areas of the tumor, which gives a more eosinophilic and fibrous appearance throughout this neoplasm.
Finally, sometimes they see glandular elements in the endometrial stromal tumors, which can be challenging. One can wonder whether this could be adenomyosis, could this be endometriosis? This is actually a case that I had in the ovary, which was metastatic from the uterus, that had a lot of glandular elements.
At first glance, it would just look like exuberant adenomyosis or endometriosis, depending on where it is. But I guess the key for the right diagnosis is to make sure that you do have areas of conventional low-grade ESS, and then putting everything together, the overall findings will be consistent with just glandular elements in an endometrial stromal sarcoma.
This is the ideal scenario of an immunohistochemical panel for these tumors: CD10 being the most commonly used, diffusely positive. I didn’t show hormone receptors ER/PR because this is to make an analogy with smooth muscle tumors, which will be positive as well.
Desmin and Caldesmon ideally will be negative, but quite often we see tumors with or without smooth muscle differentiation that do have focal, sometimes even diffuse, staining for Desmin.
So be aware of that. Do not use Desmin alone; do not use CD10 alone because CD10 can also stain smooth muscle. So put together morphology; put together the IHC and, if you have it available, genetic or molecular results to make the distinction, especially if your differential diagnosis is with a highly cellular or a cellular leiomyoma, which is what I’m showing you right here.
So if you just have a high power image, you start wondering, could this be an endometrial stromal neoplasm, or could this be a highly cellular leiomyoma?
The features that I tend to use to make the distinction, other than IHC or immunologic chemistry, is that in leiomyomas, or smooth muscle tumors in general, you tend to have thick-walled blood vessels.
We just reviewed that in the endometrial stromal tumors; those vessels are small, they’re delicate, and they’re evenly distributed in the tumor. Another clue for the diagnosis of a smooth muscle tumor, meaning leiomyoma, is that at the periphery of the tumor, adjacent to the normal myometrium, we see the blending rather than infiltration; cleft-like spaces are also commonly present in smooth muscle tumors, and you don’t really see that very often in endometrial stromal neoplasms.
I’d say Desmin and H-caldesmon can differentiate; H-caldesmon tends to be slightly better to differentiate those two.
Here’s an example of a smooth muscle tumor where the vessels are larger—some of them are staghorn—and you see clefts. At the periphery of the tumor, you see the blending like I was mentioning, like so these tumor cells—this smooth muscle tumor, this leiomyoma—actually merge into the adjacent myometrium, and it’s also a good sign that this is a smooth muscle tumor rather than the infiltration or the tongue-like projections we see with endometrial sarcomas.
One concept that I just wanted to bring out quickly is that there are endometrial stromal tumors with limited infiltration. So what is that? We learned that endometrial stromal nodules aren't allowed to have more than three millimeters in projections, however, what do we do in cases that have marginal irregularity exceeding that allowable for the diagnosis of endometrial stromal nodules but you don’t really see widespread myometrial invasion, which is commonly present in the low-grade stromal sarcomas?
This terminology of endometrial stromal tumors with limited infiltration was created back in 2002. Here’s a good example of it.
I wanted to take your attention to this picture here, that most of the tumor is well-circumscribed. So you say this is an endometrial stroma, but then you have a projection that’ll be maybe four or five millimeters which exceeds the allowable, highlighted by CD10.
What to do? How to classify? For many years, we called them endometrial stroma tumors with limited infiltration because most of the series that came out for many years show that these tumors actually behave pretty much like endometrial stromal nodules. They don’t really metastasize; they don’t recur.
However, more recent data have shown, especially this work here from Dr. McCluggage and Dr. Moore, that some of these tumors can present with distant metastasis even if they’d just qualified as endometrial stromal nodules with limited infiltration.
So just to summarize—if you see a tumor that has very limited infiltration but is slightly above what’s allowed for stroma nodules—probably the best term or classification of those would be endometrial stromal sarcoma with limited infiltration, given that some of these tumors do have malignant potential.
Molecular genetics—these tumors are molecularly classified or qualified by the presence of translocations, commonly involving the JAZF1 gene. This is the most common, and it says 12, but others such as PHF1 and many other gene partners can be fused with JAZF1.
Here’s an example of a break-apart probe where you see the translocation. Most of the time, you don’t really necessarily need a molecular confirmation or molecular diagnosis if it’s classic morphology, classic immunochemistry; sometimes, it can be helpful when the diagnosis is difficult to make.
All right, so that’s everything we had for the low-grade individual stromal circumstances. Endometrial stroma high-grade tumors—this is something that has evolved a lot in the past, let’s say, five years. A lot of the new gene fusion has been described, the classification has changed, and just to remind you that now we use high-grade ESS with the Y-H fusion or translocation, while the B-CORE alterations are the two current WHO groups that we’re going to talk about.
What are the key features to make the diagnosis of a high-grade endometrial stroma sarcoma with a Y-H translocation?
These tumors do show primitive invasion just like we see in low-grade ESS; however, what’s very characteristic of these tumors is this biphasic appearance.
We do have areas that look like low-grade ESS, as you can see here. The tongue-like projections—the cells look, even though it’s a low-power picture, but very uniform. I wouldn’t be able to call this high-grade if I only had those two areas, but in contrast, you have different foci with cells exhibiting high-grade cytology—more epithelioid—and these cells tend to have necrosis, high mitotic activity, which tends not to be present in low-grade foci.
We know that necrosis and metabolic activity is no longer a criteria to call low grade versus high grade; we used to call it that in many many years ago in the old classification. But it’s something that we tend to see much more in the high-grade component.
Here are a few examples of this epithelioid, round component, which would qualify this tumor as a high-grade ESS. But again, some areas do still retain the low-grade phenotype.
So when the first studies or descriptions of this entity came out—about a decade ago—interestingly, Cyclin D1 is one of the genes that are upregulated in this family of tumors. One of those is CCND1, which is the gene that encodes Cyclin D1.
So, can we use Cyclin D1 to make the diagnosis or to help in the diagnosis of high-grade endometrial stromal sarcoma if we don’t have access to molecular tools? The answer is most likely yes.
As you can see in the rest of the presentation, a lot of these high-grade endometrial stromal sarcomas from different families—the different genes overexpress Cyclin D1.
Here’s an example—part B here, the picture B—diffuse expression. We tend to like to see eight more than 70 to 80 percent. So, a patchy 30 to 40 does not classify as overexpressed, so you really want to see diffuse expression of Cyclin D1, and these high-grade components, high-grade foci, are most often negative for CD10, ER/PR, hormone receptors, and CD10.
So that pretty much changes the idea that we had many years ago—that if it’s endometrial stromal sarcoma, it has to express CD10 and hormone receptors—and the answer is absolutely not because we know that the high-grade ESS, the high-grade components, are commonly often negative for those markers.
One thing to be aware of is that these tumors can express KIT or CD117, but if you’re looking at an abdomen or in a recurrence, make sure that you throw out your differential to exclude gastrointestinal stromal tumors.
That was everything I had for the family of the Y-H high-grade ESS.
Most recently, B-CORE became one of the newly described variants of high-grade ESS. There are two types of endometrial stromal sarcomas with B-CORE alterations, which all classify as high grade. The first one is gene fusions and the second one is internal tandem duplication (ITDs).
That’s what I’m referring to here, but morphologically, they are very similar. They tend to show a myxoid stroma in many cases. The cells can be spindle or epithelioid, but how can you distinguish that from a high-grade ESS, the one that we just saw with the Y-H fusion? These ones with B-CORE are not associated with a low-grade component, so there’s no biphasic morphology here—everything looks very similar throughout, and quite often with these features that I just mentioned, which is the myxoid stroma.
As you can appreciate here, they still have the classic morphology, the classic features of primitive infiltration into the myometrium, quite often with vascular invasion, but they tend to be very myxoid quite often, and other things that we tend to see with these tumors is the presence of HPC-like or hematopoietic-like vessels, which can create a little bit of confusion.
If you only have areas that are hypovascular, you start wondering could this be a smooth muscle tumor, could this be a solid fibrous tumor?
But the good thing is that you start to appreciate other features that are common in endometrial stromal sarcomas, such as the hyaline plaques, the vascular invasion, and the myxoid stroma.
Here’s another example showing that myxoid appearance. The cells in this picture look very spindle, relatively—they actually can deceptively blend because they look relatively uniform throughout.
Here’s an example of it: spindle, slightly hypovascular, and they don’t really have a specific pattern of growth; they kind of grow in sheets, sometimes in fascicles, so we need to do immunohistochemistry to make the diagnosis.
One thing to be aware of is that depending on the type of B-CORE alteration, the IHC can vary. One thing that I wanted to point out before anything is that Cyclin D1 still remains as a good marker.
So regardless if you’re discussing for diagnosing or grading endometrial stromal sarcoma over the Y-H translocation versus the one with the B-CORE alteration, Cyclin D1 can be considered as a relatively good marker for high-grade ESS as a whole.
ER/PR tends to be low as well as CD10 tends to be more commonly expressed in tumors that have translocations involving the core, while the ones with the ITD tend to be negative to focal. So just to be aware that ER and CD10 negative do not exclude, as we learned, a diagnosis of ESS, and IDCs are available, and B-CORE immunohistochemistry is not very sensitive or specific but tends to be more reliable if we’re talking about the B-CORE ITD variant.
With the rearrangement, it can be negative; it doesn’t exclude at all. But one thing, before making the diagnosis of endometrial sarcoma with a B-CORE alteration, think about the uterine mixture family of tumors, which can be very very similar and can be a pitfall if you don’t really understand the differential diagnosis.
I would say the most similar tumors that you need to make—a semi-aminostain or sometimes molecular testing to make the differential diagnosis—would be a myxoid leiomyosarcoma. Since we know that smooth muscle tumors are way more common than any other mesenchymal tumors, always have in the back of your head, “Am I dealing with this smooth muscle tumor?” because that will be your most common diagnosis.
The other one will be the high-grade ESS, especially the most B-CORE alteration, and another tumor that a lot of attention has been given in the past few years is the inflammatory myofibroblastic tumor (IMTs). They can look very similar, and here’s an example of that.
These are actually three real cases of these myxoid spindle cell tumors, and it would be very difficult to make the diagnosis without IHC. Yes, you can see the features that are more commonly seen with muscle tumors; that are described the vessels.
Quite often, myxoid leiomyosarcomas actually have a very infiltrative pattern of growth, so it’s actually a characteristic feature of these tumors which is very also similar to inflammatory myofibroblastic tumors, high-grade ESS.
Honestly, when I have a tumor that looks like that, I always do a panel of immunostains to make the diagnosis, and here’s the correct diagnosis.
So this is an inflammatory myofibroblastic tumor. You want to see some of the inflammatory cells associated, but sometimes it’s just very inconspicuous.
Myxoid leiomyosarcoma and B-CORE on the right.
So once I face one of these tumors with myxoid stroma, I tend to do an immunohistochemical panel, which includes muscle markers, which include ESS markers, and include an ALK. So, if you have ALK in your laboratory IHC, that can be very helpful because it’s a good marker for inflammatory myofibroblastic tumors and tends to be consistently negative in those mimics.
Another concept that I wanted to bring up is: is there such a thing as low-grade endometrial stromal sarcoma progressing to high-grade? Since these are tumors that are molecularly defined, does it really happen? Meaning if I had a diagnosis of a low-grade ESS back in the day, and now this tumor is behaving much more aggressively with recurrence and looks morphologically distinct.
The answer is yes; there have been some cases of this progression from low-grade to high-grade. In order to make the diagnosis, you really have to do molecular testing to exclude that what you’re dealing with and you’re trying to say that you call it high-grade is actually a tumor that does carry a JAZF1 gene rearrangement.
This case series, which was published last year, had 12 cases, and the vast majority tested had a translocation or fusion, which is characteristic of the low-grade tumors. However, they started showing some nuclear enlargement, nucleoli, increased mitosis—features that we tend to see with the high-grade family of tumors.
Another interesting thing is that they started losing CD10 and ER, which tend to be very often, if not all the time, expressed in the low-grade variants, and they start losing those. But Cyclin D1, which is a high-grade marker, the good thing is that it was consistently negative.
There’s limited data out there to say, “The fact that it histologically looks higher grade, could that predict aggressive behavior?” Because we know that tumors sometimes look horrendous under the microscope, and that doesn’t correlate with biological behavior.
I’ll say lyomyosarcoma being one of them: we don’t really have a standardized way of grading leiomyosarcomas, and you can have a tumor that is very indolent-looking under the microscope which can behave worse than one that has really pleomorphic findings and histological atypia.
So in these cases, it’s limited, but possibly yes, that could predict aggressive behavior.
These are the pictures of the paper, which you can see here. Be cautious calling this low-grade endometrial stromal sarcoma even in the presence of a JAZF1 or PHF1 translocation because the tumor looks much more epithelioid, brisk mitotic activity—doesn’t really look like your classic low-grade ESS.
Keep aware of this potential progression. Use Cyclin D1, which will be probably negative, and if you have access to molecular testing, that will still confirm that this tumor originally or still carries low-grade ESS signatures.
The last family of these tumors, which I’m not going to spend too much time on, is just that… so we’re aware that this is an evolving group of tumors, and recently this group published B-CORE L1 endometrial sarcomas, also high-grade. Some of them were fusions; some of them were mutations; some of them were deletions.
They looked very similar to the B-CORE family of tumors, the high-grade B-CORE family of tumors, including myxoid stroma, changes collagen plaques, and epithelioid areas. However, the gene that is involved is not B-CORE.
This is just new data; there’s not much more that we know about these tumors because the IHC was very limited. This group is a group from a referral lab, a molecular lab, so they didn’t have much access to IHC or clinical data.
But I wanted to show you that this is an evolving topic, and I’m sure if I’m presented with this lecture in two or three years from now, we’re going to have other family of tumors. As of now, the markers that we have in our hands still can diagnose the vast majority of sarcomas.
Here’s an example of the cases of those B-CORE L1 altered tumors—myxoid stroma. Some areas look more epithelioid, as you can see here in the middle; some areas look more spindle and fascicular.
If you have access to a broad genetic panel, such as next-generation sequencing, I’m not sure if they did NGS or RNA sequencing in that study, but it’s something that may eventually pick up those genes.
To summarize the pictures, this is the Y-H translocated high-grade ESS: epithelioid areas, biphasic tumor, which the areas of each tumor, or each component, will stain differently.
While the B-CORE is quite often spindle, myxoid, HPC-like vessels, but always use your immunohistochemical tools and, if possible, molecular diagnosis.
Finally, the last entity, which is a diagnosis of exclusion, will be our undifferentiated uterine sarcoma. By definition, these are high-grade. So we don’t really call high-grade undifferentiated sarcoma because it’s just like the serous carcinoma of the endometrium: those tumors are not to be graded; they’re always high-grade.
There’s no resemblance to the PE or proliferative phase of the endometrium; there’s no line of differentiation after you do your immunostains, or even by morphology and histologically.
These tumors are characterized by these cohesive, quite often sheets of cells with myxoid morphology, tons of necrosis, destructive myometrial invasion. The markers, like the ones I listed here, are tired; I put some ESS markers here because these tumors are…
We don’t really know if they’re endometrial in origin, like what line of origin they are. They can be very focused here and there for some markers, but in general, they’re negative for most of the markers that we use.
The definition that, you know, in order to call that, is there’s no morphologic or immunophenotypic evidence of any cell differentiation—no skeletal, no smooth muscle, no fibroblast. So this is a diagnosis of an undifferentiated tumor.
And same applies for the lineage markers; for example, if you do like your H-Caldesmon, it tends to be negative. Your myogen, your SMA may have a few areas here and there that are positive, but in general, they’re negative for all the immunos.
They do not have any specific gene fusions or genetic alterations; they can have a lot of chromosomal abnormalities, non-specific mutations, but no really fusions, because if you catch a fusion, if you look at the tumor and you think, “Oh, this is going to be undifferentiated sarcoma,” you send for your molecular frequency, and it comes back, for example, with a B-CORE fusion or any type like the ones that we saw today, then that completely rules out the diagnosis of the undifferentiated.
I can’t stress more, and I’m being repetitive—this is a diagnosis of exclusion after you rule out everything.
Here’s an example of it—marked pleomorphism, cells look super ugly under the microscope. You can have lymphovascular invasion and then you do H-Caldesmon; there’s nothing to be negative.
You do molecular studies; they come back non-specific. Therefore, you classify as undifferentiated sarcoma. As you can see here, pleomorphism, discohesive, very mitotically active tumor.
Let’s go back to our index case, the one that we started with in our patient. If you remember from the beginning of the presentation, there was a tumor that had two different areas—one resembling proliferative type stroma, the second one looking more epithelioid, round, with brisk mitotic activity.
Immunohistochemistry was different. This tumor here was CD10, ER positive; Cyclin D1 negative, while this component here in the same tumor was Cyclin D1 positive, CD10 negative, ER negative.
So what's your diagnosis? Yes, this is the areas here—high-grade, low-grade. This is a high-grade endometrial stromal sarcoma with Y-H fusion, which belongs to the first category that we discussed on the high-grade endometrial stromal sarcomas.
So how common is it to have two distinct IHC profiles in the same tumor? It’s actually very characteristic of this type of tumor, as we see they have a biphasic appearance. The areas that are low-grade tend to stain for low-grade markers, and the areas that are high-grade interestingly stain for high-grade markers.
Is there any other test that you can do to confirm the diagnosis? Again, if you really have access to a molecular diagnosis, I would strongly recommend this to be a tool to be used on those tumors because, differently from the low-grade variants, the high-grade tumors are not as clear-cut because there are so many variants, and there are so many IHC properties in each one of them.
I tend to send those for molecular testing; if you have FISH that would pick up the genes, the B-CORE or Y-H, you can use that as your testing, but now I would say most individuals are using more broad panels such as next-generation sequencing, which then you don’t have to really pick the fusion that you’re looking for.
Those modern next-generation sequencing testing can identify fusions, mutations, and they’ll be more helpful in this scenario.
In terms of prognosis, endometrial stroma nodules are not to be considered malignant; so they are benign. But low-grade to high-grade; the undifferentiated—prognosis gets worse as you move to the right. Most of the low-grade endometrial stromal tumors or sarcomas are treated with anti-estrogenic therapy because these tumors are ER/PR strongly positive, and they feed…
I mean, there’s a good response if you use estrogen therapy; the high-grade ones—because, quite often, they’re hormone receptor-negative—tend to be treated with chemotherapy, and they have an intermediate clinical behavior or outcome.
As you move towards the undifferentiated, these tumors, unfortunately, have very poor prognosis; therefore, chemotherapy also doesn’t really work effectively, unfortunately.
In summary, just to kind of recapitulate all the entities we discussed today:
For endometrial stroma nodules, in order to make a diagnosis you really have to examine the entire surrounding areas of the excisional specimen, which would be preferred because we’re not allowed any myometrial infiltration; we’re not allowed any lymphovascular invasion, and these are, by definition, benign.
Low-grade endometrial stromal sarcomas—these cells are very similar to the ones of endometrial molecule, meaning they’re PE-type stroma. But the difference is that they show tongue-like or very primitive infiltration into the adjacent myometrium and lymphovascular invasion is very easy to find.
Hormone receptors and CD-10 still are classical good markers, and we tend to use them, and they’re characterized by translocation or fusion involving the JAZF1 or PHF1 genes.
Then moving towards the high-grade category, we have the Y-H translocation, which is a tumor with a biphasic appearance—two more cell populations—the areas that look like low-grade, and there’s a little bit of high-grade. Depending on the blocks or the full sites you’re staining, the morphology, the immunochemical panel, the results are going to be different.
So it stains like low grade in the low-grade areas; it stains like high grade in the high-grade areas, meaning ER/CD-10 negative. CD-10 tends to be very helpful.
Number four—the ones with the B-CORE alterations could be fusion; could be internal tandem duplications. We do not have this by appearance; we only have this uniform, vaguely myxoid tumor which can express hormone receptors and can express CD-10 depending on the type of genetic abnormality that is underlying.
Cyclin D1 is a good thing, it’s still reliably positive in most of them, and the genetic mechanism behind this is the translocations or ITDs involving B-CORE gene.
Finally, if you excluded all your tumors and you exclude the leiomyoma as well as the ribosarcoma, then you’re dealing with an undifferentiated uterine sarcoma, for which there are no characteristic genetic abnormalities, and there are no specific IHC markers.
A few take-home points are that endometrial sarcomas are a molecularly defined group of tumors. We’re not just looking at how they appear, how ugly this tumor looks, or grading just based on cytology. Actually, these are diagnostically distinct tumors that carry genetic abnormalities, and each one of them has different IHC properties.
In endometrial biopsy samples, as we know, it's sometimes really hard to make the diagnosis. Be mindful of the limitations—write a note of your differential, and suggest either re-biopsy or an excisional specimen.
CD-10 and ER/PR negativity does not by any means exclude the diagnosis of individual stromal sarcomas because this is something that we’ve learned for the past decade—that these tumors can quite often be negative for those.
So sometimes we have… you know, some of my colleagues who are not gynecologic pathologists said, “Oh, I’m looking at this tumor in the peritoneum or the pelvic area of a woman, and I thought about ESS, but it’s ER-negative and CD-10 negative.”
So that’s when you have to kind of remind them that this, depending on the type of tumor that you’re describing or you’re diagnosing, can quite often be negative.
Final classification is putting together the morphology, the IHC, and, if you have it available, genetic or molecular results, and to be continued because, like I mentioned, we have so much data coming with these molecular sequencings that we’re now doing with these tumors; we’re learning more from it.
So I wouldn’t be surprised if the next WHO in five or six years has a completely different classification from what we have now.
With that, I’ll thank you for your attention, and I’m happy to take any questions if you have.
Thank you so much, Dr. Pinto. Can you hear me?
Yes, I can hear you.
Thanks for this very informative discussion on endometrial stromal neoplasms. I can see some questions online, and I think a couple of questions are regarding biopsy diagnosis.
Here are two questions which I found to be related. The first question is about why endometrial stromal sarcoma diagnosis cannot be made certainly based on purity. I think you have mentioned that. The second question, which is kind of related, is in regard to the matter of establishing the correct diagnosis of endometrial stromal neoplasm in small biopsy. What’s the most appropriate algorithm? Is it microscopic description? What do you suggest?
Right, so to answer the first question: the reason we cannot sometimes make the diagnosis of malignancy in those biopsies is because we may not have—or quite often we don’t have—underlying myometrium to appreciate invasion.
The next question is: why is it difficult to make—is there an algorithm? I’d say first, we’re going to make sure that this is not an endometrial patient that is taking hormone antagonists such as has a progesterone IUD or something that will make those glands very tiny.
Be aware of that because you really want to see the glands in between the stroma in order to call that benign. Sometimes crushed polyps or fragmented endometrial inactive endometrial can be just composed of endometrial stromal cells, but that’s why it’s so challenging.
You’ve got to sometimes write a note that says there are a lot of cells that raise the possibility, but if you don’t see like an expansive, sheet-like growth, you don’t see an absence of glands that are evenly dispersed, I’d be cautious to make that into small myopathy.
Thank you again, Dr. Pinto.
So there are some more questions. Let me read them for you.
So one is, I hope I will read the abbreviation correctly: U-T-R-O-S-C-T is in the differential of ESS. How do we differentiate it?
You’re like the user term of resembling a sex cord over tumor, right? That’s a rare diagnosis.
So first of all, I’d say endometrial stromal sarcomas are much more common than U-T-R-O-S-C-T, that is one thing to consider.
And secondly, endometrial stromal tumors can have sex cord differentiation, but one thing that endometrial stromal tumors will have, even in the presence of sex cord differentiation, is areas that are classic or conventional ESS in the background.
So if you’re looking at a case that has areas of sex cord differentiation that makes you think about U-T-R-O-S-C-T, you’ve got to sample. You’ve got to see other areas that will lead to a diagnosis of ESS, such as the presence of those cells of PE-type stroma, the vascular invasion, the characteristic infiltration into the myometrium.
So basically, I would just call the uterus, once you excluded the presence of a low-grade ESS in that tumor, and if you're still not sure, you can always send for molecular studies. U-T-R-O-S-C-T should not have JAF1 translocations.
Actually, there are some recent data here also about specific genetic abnormalities in U-T-R-O-S-C-T, especially the malignant ones, which are different than the ones we see in ESS.
Thank you! So there is another question. So this is about weak and/or focal expression of markers such as CD10, Desmin, or Cyclin D1.
Does it exclude the diagnosis of undifferentiated sarcoma?
That’s a great question, and I think it’s—I’ve spoken with some of our very prominent soft-tissue pathologists in our institution, Dr. Rhonda Rosenberg, Dr. Montgomery. I had the discussion, especially with Dr. Rosenberg in the past about... I’ve had cases that are very, very focal, for example, a few cells here and there expressing smooth muscle actin, and then you do CD that.
H-Caldesmon is negative. So when do you cross that line? What’s your threshold to coordinate undifferentiated versus— for example—a high-grade ESS or a leiomyosarcoma?
I would say it’s a little bit arbitrary because, you know, the undifferentiated should not express anywhere very limited, but how much is the “little”?
I would say if I see a case that has a few Desmin here and there and then SMA here and there and H-Caldesmon, even though if it’s patchy, I would be more inclined to call that like a cytologically high-grade leiomyosarcoma because there is expression with some smooth muscle and muscle markers.
But this is arbitrary. Ideally, they should be very negative, but if you have a few cells here and there that you don’t think look like a leiomyosarcoma and look like it doesn’t carry any molecular abnormalities that would classify as a leiomyosarcoma or endometrial stromal sarcoma, then you’re dealing with an undifferentiated tumor.
Right, so the next are two related questions I thought regarding molecular changes.
So, one question is: what is the cause for fusion of genetic alterations? This is one question.
The second question from another viewer is: are high-grade ESS, Y-H rearranged, expressing B-CORE?
I’m sorry, dude—I kind of missed a few words on the first and the second questions. Would you mind repeating again?
Yeah, the first is: what’s the cause for fusion of genetic alterations?
What’s the cause for fusion? Well, that could be the determining genetic abnormalities in sarcomas, right?
So when you have translocation of fusions, which is a genetic abnormality, like same way you have mutations or amplification, sometimes we have translocations, and that would be the diagnostic and the driving pathogenic alteration for sarcomas and other tumors that have fusions—two more have translocations, and some have mutations.
I’m not sure if that answers the question.
The cause of it is just a genetic event.
Right. No, I’m also not sure exactly what was the question.
The second question was: are high-grade ESS Y-H rearranged expressing B-CORE?
Oh, B-CORE, yes, I’m sorry.
I’m thinking about Cyclin D1.
I’m not entirely sure because most of the studies were done with the B-CORE gene alteration, which is more commonly in the ones that are ITDs versus the ones that are translocated.
But since I’m only going to say about personal experience, I’m not sure with the literature shows, but I’ve seen tumors that are not B-CORE altered and they do have some staining for the IHC.
So I don’t think it’s a great marker, honestly.
It’s something that we use as a panel, so I wouldn’t be surprised if the Y-H ones can have some positivity, but I don’t know the right data on that.
Right, so the next two questions are again somewhat related I thought and somewhat towards therapy.
So the first question is, what’s the surgical therapy? And also, in regard to certain translocation types, is there differential therapy?
The second one is: are there different types of high-grade uterine sarcomas with different genetic changes that differ a lot in treatment modalities or we can give a final diagnosis of high-grade sarcoma?
Oh, that's a good question. I would say the first question is the surgical treatment, and I would say most of the tumors would be treated with a surgical intervention, and then after that, I’m not sure if all of them would receive for hormonal therapy, will receive anti-estrogenic therapy.
The ones with high-grade would probably receive chemotherapy, but I’m not sure if that’s depending on the stage or if that’s just a general rule.
So since I’m not an oncologist, I wouldn’t be definitive on the answer to that.
And the second question is if the translocation or the endometrial sarcomas, I guess, would define the treatment.
I would say yes and no.
So, if your differential is a low-grade versus high-grade, then the answer is yes, because of the prognosis—the ones that are high-grade, which carry this translocation that we’ve talked about today, would be important to be diagnosed.
So these tumors are treated more aggressively, while the low-grade ones tend to be treated with anti-estrogenic therapy.
Now, the other question is if sarcomas, in general, therapy can be applied to high-grade ESS—that I would; again, since I am not an oncologist, I would not be sure of the answer.
Right, there’s another question. Can we make a diagnosis of endometrial stromal neoplasm outside of the uterus? For example, when it is associated with foci of endometriosis?
Yes, the answer is yes.
One of the terms that is commonly utilized is endometrioid stromal sarcoma. If you really don’t have the uterine counterpart, meaning if this patient had a hysterectomy and was benign—and now she has an ovary—if you diagnose endometrial stroma sarcoma, the correct term for that will be endometrioid stroma sarcoma, and that is correct—can be associated with endometriosis.
Right, so I think Dr. Pinto, we have one last question from our online viewers.
So what's the importance of accurate typing of these tumor categories?
I would say, for—again, it all depends on, you know, which kind of tumors we’re talking about.
I’ll say if you misdiagnose a low-grade for a high-grade or vice versa, I think that’s important because of prognostic implications for treatment purposes; but if you don’t have molecular studies or molecular tools in your lab and you made the correct diagnosis of a high-grade ESS, but you’re not sure if that’s a B-CORE or if there’s a Y-H or if it’s a translocation versus an ITD, I would say that may not be of much significance.
Because these tumors tend to have some of them, like, you know, of course, they defer a little bit here and there in terms of prognosis, but I think they’re all hybrid endometrial stromal sarcomas, meaning the treatment will be the same and the approach on these patients will be the same.
So as long as you correctly classify the endometrial stroma nodule as a nodule, the lower grade is low grade, the higher grade is high grade, and the undifferentiated is undifferentiated. I would say you don’t really necessarily need to subclassify the type of genetic abnormality if you don’t have that tool in your academic institution or if the patient cannot afford the testing.
So as long as you get the umbrella diagnosis correct, I think that’s all we need.
Thank you! I think these are the questions that I could find online from the viewers.
Thank you again, Dr. Pinto, for this excellent talk and all the questions that you answered for our viewers.
You would be very happy to know, Dr. Pinto, that there are a few hundred viewers from so many different parts of the world.
I could keep track of some viewers who joined from as far as Bolivia, Algeria, Brazil, Abu Dhabi, Libya, Mexico, Philippines, India, Greece, Argentina, and of course, there were some from the US as well.
Thanks to our viewers for your continued support to Pathcast. If you like our lectures, please follow our YouTube channel. You can subscribe to the YouTube channel that is Pathcast, and also the Facebook page—you can follow and like it by the same name.
We have a website where all the lectures are organized by subspecialty, and all the previous lectures are available there—all the 16 lectures, including today.
In fact, today’s lecture, you can also find a PDF of the talk on the website which is already uploaded, and the website is pathologycast.com.
We also have a newsletter; you can subscribe to the newsletter so that you can stay updated with all the upcoming lectures.
Our next lecture is on October 8th, and we will have a talk on CNS pathology—so there is the new WHO classification that has come up. Dr. Foster Rodriguez will talk on the 2021 WHO classification of CNS tumors, and he will focus on gliomas—how they are updated in the new classification.
Again, the date is October 8th and it will be at 12 p.m. Eastern time.
Hope to see you then, and thank you again, Dr. Pinto, for this excellent talk and for your time.
Thanks a lot!
Thank you, Refat. Thanks for the invitation, and thanks for all the audience that is watching across the world. Congratulations on Pathcast, on this excellent platform that it’s made available to everyone.
Thank you again!
Thanks a lot!