All right. Very good. So, I was asked to give a talk about adenocarcinoma in situ of the cervix. And I think it's important, tying in with the last case that we, the case that was just presented, that was a case that showed squamous carcinoma, carcinoma in situ or CIN3, had invaded into the endocervical glands. This, what I'm going to talk about is different, because what I'm going to talk about is an adenocarcinoma that arises primarily from the endocervical glands. So, it's not a squamous carcinoma. It's a glandular carcinoma. And when you get back a report from the pathologist that shows, as the previous case did, that there is squamous HSIL extending into the endocervical glands, that's not invasive carcinoma, and it's not adenocarcinoma in situ. As Dr. Felix said, though, it does have a slightly worse prognosis than if it did not invade into the endocervical glands. We treat it pretty much the same as CIN3 that does not invade into the gland. The word invade is wrong. That does not extend into the glands, because there's no invasion below the basement membrane. Okay. I have no financial conflicts. I'm going to start out with a hypothetical case. Here's a 27-year-old parrot 2. Her screening test is positive for HPV type 18. She had a pap test along with it, cytology, that reported back atypical glandular cells favor neoplasia. And so, the question that's asked is, what's the next step in our workup? Let me show you what's going on in this slide. I hope you can see my pointer here. But these large flat cells with small nuclei in them, those are normal squamous cells. This is a large cluster of cells that looks like it comes right out of an endocervical gland. It's a ball of cells. And they have a basal nucleus, and they have a little streaming cytoplasm coming from looking just like an endocervical cell. But they also have a number of cells that have very large, very dark, very irregular nuclei. And here's another cluster that has separated from this. And all of these have very different sized nuclei. And so, the basal cell, basal nucleus with the streaming cytoplasm leads the pathologist to say that this is most likely an endocervical cell. And the fact that the nuclei are very large, very dark, and irregular in shape, leads them to say that this is something where the glandular cells are atypical. And I, the pathologist, think maybe this is a neoplasia. Okay. So, she comes in for colposcopy, because the next step, according to the algorithms, and I'll go through those at the end, is colposcopy and endocervical curatage. The first thing that strikes me about this cervix is that it is very, very shiny. It's making a lot of mucus. And sometimes, endocervical adenocarcinomas will do that. My eye is drawn immediately to this atypical blood vessel. All right. Notice that it starts here, and it makes a sudden right-hand turn, and then it makes another right angle, right-hand turn here. We want our blood vessels to arborize, to have branches that come off of them. This one looks a bit normal. Okay. So, you've got a central trunk. You've got a narrower branch coming this way, a narrower branch coming this way. This one has a trunk that gets thinner here, gets wider again, has a branch here that's wider than the trunk is, cuts off abruptly. That's very worrisome. The biopsies and the ECC were consistent with CIN3, squamous disease. And you can see here, again, we've got a nice basement membrane, and we've got all these dark nuclei coming up to the surface. So, that's CIN3, but also adenocarcinoma in situ. And you'll see that this would be a normal endocervical gland. It's got very tiny nuclei at the basement, and one layer, and it's got this streaming cytoplasm. Here you've got what looks like multiple layers of nuclei, and they're very large nuclei. But again, the basement membrane is intact, so it's not invasive. Just to confirm, the pathologist put on a P16 immunohistochemical stain, and both the squamous and the adenocarcinoma in situ were positive. Okay. So, what we're going to do in the next 20 minutes or so, and I'm going to talk fast and maybe skip some slides because time is short. We're going to talk about adenocarcinoma in situ, how difficult it is to diagnose it, and I'll give you some clues and some hints, and we'll go on with the management recommended for atypical glandular cells and for AIS. AIS, adenocarcinoma in situ, is a pre-invasive lesion of the endocervical glands. The incidence, at least in the United States, is increasing over time. The average age is about the same as the average age that we would see for CIN3 or early invasive squamous cancer, 35 to 37. There is an average interval between the mean diagnosis of AIS and the mean diagnosis of invasive adenocarcinoma of five years. This suggests that the AIS is, in fact, a pre-invasive lesion, and most of the time when we diagnose AIS, there is also a squamous high-grade lesion sitting right next to it. Unfortunately, up to 15% of AIS diagnoses will also, at the same time, have an invasive adenocarcinoma if you look harder. This is often not diagnosed on cervical cytology. I'll show you a slide showing that HPV screening is a better way to lead you to a diagnosis of AIS than cytology is. The diagnosis often happens by accident. I think most of the AISs I've diagnosed in my career, I have done a biopsy or a leap for HSIL, for squamous disease, and the pathology report came back, oh, and by the way, there's adenocarcinoma in situ. 10% of these lesions are multifocal. Most of them are adjacent to the squamo-columnar junction, but many are not. Many of them will skip across normal columnar epithelium and show up deep in the glands higher up, and it does not obey the usual colposcopy rules that we teach. I'm not going to spend much time on this slide except to point out that there are multiple different histologic types of adenocarcinoma of the cervix, and about 15% of them are not HPV-associated. The non-HPV-associated types occur in women who are a bit older, and they include types such as gastric AIS, clear cell, mesonephric, endometrioid. It's just something to keep in the back of your mind that while HPV testing is still the best way to screen for it, perhaps if you have the ability to do cytology, co-testing might pick up a few more. There's a lot of discussion on the primary principal HPV type associated with AIS. This is a classic graph that many of you have seen. This is from Javier Bosch's study and I think Sylvia de San Jose's studies back in the early 2000s that show the different HPV types associated with normal HSIL, squamous cell carcinoma, and adenocarcinoma. I want you to pay attention to the pink one and the orange one. The pink one down here is HPV-16. The orange one is HPV-18. As you can see in the squamous HSIL, it's mostly type 16 and a small maybe 10% additional type 18. Once we get to invasive cancer, now we've got 50% type 16 and maybe 20 to 25% type 18. If we're looking at adenocarcinoma, this is invasive adenocarcinoma, it's about 50-50. As the bullet point shows, HPV-18 is associated with 8% of HSIL, but 35-50% of AIS. For that reason, many of us are sure if we have a patient with HPV-18, we will include an endocervical sampling, an ECC, along with our initial workup. This is just a slide showing that this is out of 332,000 women. There were 97 cases of glandular disease, and 60% of them were negative on Pap test, and 85% of the adenocarcinomas were negative on Pap test, but they were 80% positive on HPV. The Pap is not a very good way of picking up glandular lesions. This goes through the Bethesda Clature for Pap test for cytology, and they recommend dividing it into atypical glandular cells. Your pathologist may report the Pap as atypical endocervical cells, atypical endometrial cells, or atypical glandular cells not otherwise specified. Endometrial cells can come down through the cervical os and be picked up on the Pap test. Sometimes, the pathologist can't tell the difference between endocervical cells that look abnormal and endometrial cells that look abnormal. There is also that category that we started out with in the original case, in which they say these are atypical glandular cells, favor neoplasia. I'm not sure what it is, but I'm worried about it. Then there is also a frank diagnosis of endocervical adenocarcinoma in situ. The old terminology that some pathologists will still use is AGUS, A-G-C-U-S. This terminology went away in 2001. When atypical glandular cells show up on cytology, 11% of them actually represent high-grade squamous disease. Another 3%, so far fewer, will represent adenocarcinoma. Of interest, when you get AGUS on a Pap, 1.4% showed endometrial hyperplasia. Unlike AGUS, where malignancy is less than 1%, if you've got atypical glandular cells, there's a 5% chance that this is hiding a cancer. If they say favor neoplasia, that jumps up to 21%. We take this diagnosis very seriously. Of interest in this study, among the cancers that they found were endometrium and endocervix, which you would expect, squamous cancer of the cervix, which you would expect, but also cancer of the ovary, fallopian tube, metastatic cancer from the colon and the breast. How does that show up on a Pap test? Well, the theory is that the metastatic to the peritoneum are picked up by the fallopian tube and wander down onto the cervix where they're picked up by your Pap test. Okay, colposcopy. Most of the lesions, as I mentioned, lie within the transformation zone or close to the squamo columnar junction. This particular diagram obviously shows a cervix here, and this part here is what you can see on colposcopy or VIA. This part here is what you're going to remove on a leap. In this particular illustration, the AIS is way down here because all of these are endocervical crypts lined by endocervical glands that can then undergo malignant transformation. Your cone biopsy or leap will pick up any of them in this area, but they're going to miss this. This is why an ECC is important, and this is why our management takes all that into consideration. So colposcopy or VIA prediction is difficult to impossible. The features of AIS overlap with this lesions, so if you see something that looks well, you may pick up the AIS with your biopsy or your leap there. Unfortunately, they also look fairly close to immature squamous metaplasia, which might be normal. So you have to have a suspicion and you have to biopsy anything that looks funny. The features that I'm going to show you of AIS are the same features that we might see for adenocarcinoma, except adenocarcinoma is more likely to have friability, necrosis, surface ulceration, all the features that you know to look at when you're looking for cancer. So the surface pattern may look like normal ectopy, like a great big ectropion. You may see endocervical papillae that are coalescing. They may be a variable size, irregular shape, and they may be confused with immature metaplasia, and I'll show you pictures. If atypical glandular cells are reported on cytology, please biopsy anything that looks abnormal. We may see acetyl white lesions. Not uncommonly, there'll be milky white, maybe surrounded by glandular epithelium, because remember, that's where this arises. You may see patchy red and white tissue on the cervix. You will frequently see atypical vessels. The descriptions include terms like root-like. This means that they are bulbous, wider in the middle, some looking like blood blisters or looking like a yam or sweet potatoes. They may be hairpin glomeruloid vessels, little thin vessels that turn back on themselves. Willow branch-like. The term that I like is character writing, because I'm not totally sure what it means, except somebody is writing to you in a language that you're not familiar with, instead of the Latin alphabet. It may look like some of those letters. Really strange variations on the branching pattern that you normally see. There is a type of adenocarcinoma called mucous secreting that may have large gland openings, and one of the characteristics of these is that they may not have that nice thin rim of aceto-whiteness that you expect with gland openings. You may also see comfortable glands, and again, as I showed earlier, copious mucus. So here is a 29-year-old. She had atypical glandular cells on cytology, and this is what her cervix looks like. And looking at this initially, you can see this nice white rim going all the way around, and you're tempted to say, well, that's the squamo-columnar junction. Then you look inside here, and you see that this looks solid, and you think, well, maybe that's the squamo-columnar junction. So the first thing you notice is that there are patches of red and white tissue. Here is what I would consider to be an atypical coalescence of the papillae. It just doesn't look right. Here are some of those root-like vessels. Look at this great, big, bulbous, looking like little blood blisters there. In addition, here's a little vessel that turns back on itself. Here's another little vessel that starts and stops abruptly. Here is a great big one that is branching, but it doesn't look like it's branching normal. There are others in this picture as well. This is a lady that I saw several years ago, and her pap test was ascus, and she was HPV positive. But what I noticed was, again, here is your squamo-columnar junction, and here are these endocervical papillae, and all of a sudden here is another white patch that's going up into the canal. And when we biopsied that, she had both squamous high grade and adenocarcinoma in situ on the biopsy. Here's another one. You would look at this and say this is an abnormal-looking os, but I think that it's abnormal because she maybe had a previous leap. It has that appearance to it. But look up inside. You've got these patches of aceto-whiteness, and this could be new metaplasia, but it looks more like coalescence of the endocervical papillae. Look at this blood vessel here. It starts there. It stops there. It starts here again. It's fat. It may have a branch, but mostly it's just this long white tree trunk with nothing growing out of it. That would have me worried. This is a 43-year-old with atypical endocervical cells favor neoplasia on her pap test, and her biopsy did in fact show AIS. So the first thing that strikes me looking at this is she has a fairly wide ectropion for a 43-year-old, and it has that shiny mucousy effect to it. She has these large gland openings that are not surrounded by that thin rim of aceto-white. Here's a normal gland opening. You can see this little white rim going around it there. That's normal. And she's got these bulbous root-like blood vessels there, and under the green filter you can see some of this character writing. It looks more like a squashed bug to me than character writing. There is an old thin monograph by a Japanese pathologist named Yuki, U-E-K-I, who described the features of adenocarcinoma of the cervix. He talked about a muscular networking, which looked very much like this, which I'm calling a fibrous lattice. But you can see that there is fibrous tissue here that's going every which way. And on biopsy, again, you can see all of these cells with pseudostratification, multiple layers of nuclei building up on it, and that was adenocarcinoma in situ. And in fact, there was a locally invasive component. This is the only time I've seen this outside of Yuki's textbook. This is from a patient we examined in China several years ago, and I was very excited to be able to get a picture from the pathologist there. This is another case that we examined. I believe we saw her at Fudan University in Shanghai. This is before acetic acid. She's 29 years old. Very normal looking, healthy cervix. You put on the acetic acid, and what jumps out at you is this little acetyl white area that's totally surrounded by columnar epithelium. This is not connected to the squamo-columnar junction. It should not be there. Now here's also an acetyl white patch here that is obviously squamous CIN. But on biopsy, and I don't know which was which, biopsy they reported both CIN2 up here, and I think this is the focus of adenocarcinoma in situ. This is another type of adenocarcinoma that I've only seen a couple of times, but once you've seen it, you'll remember it. This again is before acetic acid. She's 27 years old. Atypical glandular cells favor neoplasia. So if you are doing cytology, and you get this diagnosis, that means you're going to biopsy whatever looks funny. So the first thing you see up here, your eyes are drawn immediately to these blood vessels and these large white cystic structures. These are just normal Nabothian cysts. Protruding out of the cervix is this filmy, velvety-like substance. And you wonder, what in the world is that? I looked at her next with just the high-powered magnification and the green filter after cleaning her off with a little saline. And all of a sudden, what jumps out are these very dramatic spaghetti vessels. Here's a hairpin vessel, and you might want to call this a glomeruloid if you use that term. But these are obvious abnormal blood vessels. Now watch what happens when we put the acetic acid on. All of a sudden, it agglutinates. It all jumps together. So you're going to look at this, and you're going to say, I have no idea what this is, but I'm going to biopsy it. This is a good place to do some biopsies. Again, if you look up close, you can see these two railroad tracks here. So these are atypical vessels. I don't know if these are atypical or if they're just little patches of blood in clefts. And on histology, there is this very unusual papillary pattern, which is P16 positive. And the diagnosis was an endocervical villoglandular adenocarcinoma. You remember that early slide that I showed you that listed all the different types of adenocarcinomas. Villoglandular is responsible for, I think, maybe 8%, a very small number of adenocarcinomas of the cervix. It's P16 positive, so that tells you that it is HPV-mediated as well. All right, I'm going to spend a very few minutes going over the management algorithms. These are from the ASCCP guidelines, and they're based on cytology plus HPV. So all categories of atypical glandular cells on cytology, your first step is going to be colposcopy and endocervical sampling. And you'll notice that it says all subcategories and any HPV result. So if you get back atypical glandular cells, HPV-negative, you're still going to do colposcopy and endocervical sampling. If the patient is at some increased risk for endometrial hyperplasia or cancer, you're also going to do an endometrial biopsy. So ASCCP has set an age cutoff at 35. If they are 35 or older and have atypical glandular cells, their workup includes an endometrial biopsy. If they're under 35 and have the stigmata of chronic anovulation, abnormal uterine bleeding, obesity, et cetera, even though they're under 35, you're also going to do endometrial sampling. If the biopsy comes back showing atypical endometrial cells, then your first line is going to be endometrial sampling plus an ECC. You can do colposcopy at the same time. And the reason for doing that might be that if the endometrial biopsy doesn't show any disease, then your next step is going to be colposcopy. Those of us who have patients who may get lost to follow-up, I'm going to do all three at the same time. I'm going to start with a colposcopy, then do an ECC, then do an endometrial biopsy. So what happens if you've got atypical glandular cells or atypical endocervical cells, not otherwise specified, and your original biopsy is negative? There's nothing there. Well, the recommendation is to bring them back in a year for co-testing. And if that's normal, bring them back in another year. So this is closer follow-up than we do with squamous disease. And if both of these are negative, then we put them on a three-year regimen for a while. If, on the other hand, it shows HSIL, but no glandular disease, then we manage that as appropriate for HSIL. If your initial Pap test was atypical glandular cells favor neoplasia or adenocarcinoma in situ, and your initial workup may show some disease, but it doesn't show AIS or cancer, then what they're telling you is you're not done. Because your initial cytology was favor neoplasia or AIS. The next step is going to be a cone biopsy or possibly a leap. Many of us will start out with the cone if AGC favor neoplasia is our diagnosis. Now, the little asterisk here is important because it says your diagnostic excisional procedure should provide an intact specimen with interpretable margins. And for most of us, an intact specimen means a cold knife cone as opposed to a leap, because leap very often comes in multiple fragments, and that's not good enough. Also, an endocervical curatage is recommended. So I'm going to finish off with what happens when your biopsy shows adenocarcinoma in situ. First of all, as I mentioned, an intact specimen, an excision is your next step. You're going to do a cold knife cone. A leap with a top hat excision, that little excision above the the initial leap, that's not acceptable. The excision should be at least 10 millimeters long, and if childbearing's completed, it should be twice that size. An ECC is preferable. If the margins are positive on your initial cone, repeat the excision until the margins become negative. Margin status is a predictor of residual disease, recurrent disease, and progression. The recurrence risk is 2% if the margins are negative, but rises to 19% if margins are positive. So you want to keep doing repeat cones until either you have no cervix left, or until the margins are negative. Once you've done your cone, the next step is going to be a hysterectomy. If the margins are negative, a simple hysterectomy is recommended. If the margins are still positive, then some people will do a simple hysterectomy, but many will do a modified radical. Surgical assessment of the lymph nodes is acceptable at the time of hysterectomy. As I mentioned, the age at which AIS appears overlaps the age at which many people want to get pregnant. And so if fertility is desired, you do your initial cone, and if margins are negative, then you can continue to follow her. What does that mean? That means co-testing plus an ECC every six months for three years. And if those are all negative, then continue to do it annually for another two years, and then every three years indefinitely until she has her hysterectomy. And the recommendation is she should get that hysterectomy after her childbearing is completed. So if the margins are not negative on that initial cone, if they're positive, the recommendation is to do a repeat cone and then another cone. And if negative margins cannot be achieved on multiple conizations, then a hysterectomy is recommended. Fertility sparing should be done only in select cases after discussion with the patient of the risks of recurrent AIS or cancer. Obviously, if she has a hysterectomy, her fertility is over, but she needs to weigh which is a bigger concern of hers, not having children or having invasive cancer. When somebody opts for fertility sparing and they've had their children, okay, then they should have a hysterectomy. And I've asked multiple oncologists over the years, well, what happens if four or five years out she is still totally negative? Does she still need a hysterectomy? And the answer has been consistently, yes, she should get the hysterectomy. If she opts not for the hysterectomy, then continued surveillance is acceptable as long as HPV tests are consistently negative. If they become positive, she should get a hysterectomy. And after the hysterectomy, she should continue to get HPV-based testing up till 25 years or more. The HPV-based testing should follow the same kind of surveillance that we've been talking about, which would be cytology and PAP at six-month intervals times three, then annually times two, then every three years until at least 25 years. So in conclusion, atypical glandular cells on cytology should be referred for colposcopy and ECC with EMB endometrial biopsy as appropriate. Management of AIS on biopsy is hysterectomy unless childbearing is considered. If concerns for childbearing exceed the concerns for cancer, cone margins must be negative, close follow-up for at least 25 years, and hysterectomy after childbearing is recommended. And I thank you. And I think we're out of time, but I will hang on if people want. And if there are questions, I'll be happy to answer.

adenocarcinoma in situ

cervical cancer

HPV screening

colposcopy

endocervical curettage

histological examination

hysterectomy