Michael Birrer, MD, PhD

Case: A 70-Year-Old Woman With Endometrial Cancer

A 70-year-old postmenopausal woman presents after experiencing abnormal uterine bleeding for about 3 months. She has a grown daughter, underwent menopause at 52 years of age, and her father had prostate cancer in his late 60’s.

PE: Notable for large uterus and tenderness upon palpation

Endometrial biopsy: endometrioid adenocarcinoma, FIGO grade 1

Surgery: ELAP TAH BSO with bilateral pelvic node dissection

Pathology: grade 2 endometrioid adenocarcinoma, 15 negative pelvic nodes, invasive 1.9 cm of 2.4 cm myometrium

Molecular testing shows MSI-high, MMR proficient, and HER2-

Postoperative radiotherapy: vaginal cuff brachytherapy to a dose of 21 Gy in 3 fractions

12 months after completing radiotherapy, she presented with new RLE edema and right hydroureter

She then was treated with carboplatin/paclitaxel chemotherapy which was well tolerated

15 months later the patient has disease relapse with metastases to the paraaortic lymph nodes

There’s a tremendous amount of excitement in the field, and I share it. The first direction is moving all these agents up front and getting them up earlier in the treatment of endometrial cancer. We can take pembrolizumab—we now have dostarlimab approved also—and move it up earlier, closer to the initial diagnosis, combining it with chemotherapy. A patient gets a biopsy. It’s positive. They get surgery, but they’re at higher risk. They would get adjuvant chemotherapy if they’re microsatellite unstable, for sure. It makes a lot of sense to add an I/O [immuno-oncology] agent at that point, which would likely increase the fraction of patients who never recur. That’s the first step: moving them all up front. I don’t think we’re going to see much from a toxicity standpoint. For tumors that are microsatellite stable, it’s the same thing, moving lenvatinib-pembrolizumab up front. It may even be better than carboplatin-paclitaxel because of the data we have so far. The patient gets diagnosed, and they’re at high risk for recurrence. Instead of getting adjuvant chemotherapy, they might get adjuvant lenvatinib-pembrolizumab.

The second direction is to move them up front. A lot more patients are going to benefit, but some aren’t. The questions become, “Why are they not responding to immunotherapy? What are the mechanisms of resistance? Can we combine I/O combinations with other I/O targets, or can we use I/O with other agents other than lenvatinib to improve the response rate?” That’s going to be a little trickier than what I just said. What I just said about moving it up front—all that is happening. We’ve got randomized phase 3 trials ongoing. But what I just described is going to require some hard-core biology. We need to biopsy those tumors and look at them. If they’re microsatellite unstable but not responding to single-agent I/O, what’s the mechanism of that? Why is that happening? Is it that T cells aren’t trafficking into the tumor? Is it that they have a different target or mechanism other than PD-1? The whole field of immuno-oncology has matured to the point that we’re looking a little further downstream in terms of mechanisms of resistance.

Patients with endometrial cancer have been a forgotten group of patients. It was mostly a postmenopausal patient population, many of whom had hypertension, diabetes, and heart problems. It’s not a huge market, so companies stayed away. We’re over that—companies are in there—but I’m not sure the oncology field recognizes that. I still get patients referred to me whose oncologist said that with endometrial cancer recurring after chemotherapy, there’s nothing to offer. We need to bang the drum and make it clear that terrific results from KEYNOTE-158 and KEYNOTE-775 are home runs.

Videos

Closing out his discussion on advanced endometrial carcinoma, Michael Birrer, MD, PhD, shares his hope for future treatment advances.

Advanced Endometrial Cancer: Future Directions in Care

When the whole microsatellite-unstable phenomenon broke through, it came from TCGA [The Cancer Genome Atlas] of endometrial cancer, and then pembrolizumab splashed onto the scene. We knew we had a new option for about 35% of endometrial cancers, but the other 65% to 75% are microsatellite-stable tumors. What do we have to offer for them? Along comes this combination of lenvatinib plus pembrolizumab. Lenvatinib is a bit of a dirty kinase. It’s not as narrow and specific as others, but that’s probably why it works—meaning it’s inhibiting a number of tyrosine kinases, 1, 2, or 3 of which then synergize with the immune modulation from pembrolizumab.

Initial studies were phase 2 trials, most coming out of [Johns] Hopkins, showing this tremendous response rate, pushing toward 50%. But more important, [they had] a disease control rate around 75%, meaning there were patients who didn’t get a response but had shrinkage and control of their tumor. This combination looked interesting, but KEYNOTE-775 was the home run, directly examining this combination in comparison with standard of care, which is chemotherapy. This showed not only a progression-free survival advantage but an overall survival advantage. I would ask anybody in the audience, “How many times have you seen a randomized phase 3 trial in endometrial cancer showing an overall survival advantage?” It’s very rare.

For years, we used doxorubicin based on no objective survival advantage data. Here we have a combination that’s terrific and very effective, and it’s targeted to the microsatellite-stable population, which is the majority. The only thing to remember is that while pembrolizumab has a very defined toxicity profile, lenvatinib is a little harder as a tyrosine kinase inhibitor. It has hypertension, which can be impressive and quick; diarrhea; and fatigue. People just need to realize that the FDA-approved dose is 20 mg, which I use, but a lot of my colleagues start at 10 or 14 mg because they’re worried about the toxicity.

Expert perspectives on Study 309/KEYNOTE-775, which combined pembrolizumab with lenvatinib in patients with advanced endometrial carcinoma.

Combination Pembrolizumab-Lenvatinib in Advanced Endometrial Cancer

Before, it was easier: carboplatin, paclitaxel, and that’s it. We’d debate whether we wanted to use doxorubicin. It’s gotten more complicated, but that’s good for our patients because there are choices. In this case, with a tumor in the outer 50% of the uterus, a grade 2, I probably would have given the patient adjuvant chemotherapy rather than radiotherapy. But I tend to be very aggressive because I’ve never had a patient get into serious trouble with carboplatin-paclitaxel. That might have been slightly different.

Let’s say the patient was treated with radiotherapy and then, when the tumor recurred, we use carboplatin-paclitaxel. It’s perfectly reasonable. But it recurs, and we move on to pembrolizumab. I completely concur with that. Given what I just said about KEYNOTE-158, it’s hard to imagine that re-treatment with chemotherapy, let alone with any of the agents that I have outlined, would be anywhere as near as effective as pembrolizumab as a single agent. I agree on that.

What you find for endometrial cancer is that there are patients who will get adjuvant chemotherapy or treatment for metastatic disease and, for whatever reason, have a long disease-free interval. Let’s say you have a patient—with grade 3 endometrial cancer, for example—who gets adjuvant chemotherapy and then 2½ to 3 years later recurs. There are individuals who will re-treat with carboplatin-paclitaxel. The reason they do is because the platinum-free interval is so long. But I have to point out that the field is contaminated a little because most patients who treat endometrial cancer treat ovarian cancer. In ovarian cancer the platinum-free interval is clearly defined. When patients recur with a long platinum-free interval, re-treatment predicts for a reasonable response. There are no comparable data in endometrial cancer to correspond on that. The molecular reason is because ovarian cancer has a homologous recombination deficiency; most endometrial cancers do not. So I’m not sure that analysis makes a lot of sense. In this case, given that you had an MSI [microsatellite instability]–high tumor, going on to pembrolizumab makes a lot more sense than re-treating with carboplatin-paclitaxel.

A lot of people won’t count adjuvant chemotherapy as a line because it’s not a line for a metastatic disease. You’ve removed everything, and you’re just treating the patient. I think that’s nonsensical. The tumor is being exposed. Any tumor that’s left is being exposed to chemotherapy, so that’s a line to me. If they recur, then you’re into second line, and a lot of the indications for pembrolizumab are second-line indications. What are the scenarios? A patient gets surgery like this patient, radiation occurs, nothing else, and then it recurs and the patient gets chemotherapy. That’s the first line and then, pembrolizumab is coming in as the second line. If this patient had gotten it instead of radiation and adjuvant chemotherapy then recurred, the adjuvant would have been counted as first line, and the recurrence would have been second line. I would have used pembrolizumab at that point. I hope that’s clear. Because we have choices, it’s gotten a little murky. Just remember that when you give chemotherapy, it doesn’t matter if you can see the tumor. You’re treating tumor, so that’s a line.

There are patients on KEYNOTE-158 who are almost flattening the curve for overall survival. A lot of us feel for those patients who have microsatellite-unstable endometrial cancers, because for a subset of those, the tumor does not come back within 2 to 3 years and they’re likely cured. It’s not everybody. As I said, the response rate is only 50%, but there’s a percentage of patients who don’t respond and so the tumor progresses. I’ll keep my fingers crossed that she’d be in the group that did really well, and then the tumor would disappear—which I’ve seen—and wouldn’t come back. If she’s not in that group, then she falls into an interesting group of patients for whom there’s a lot of research ongoing. When patients who have microsatellite-unstable tumors either respond or don’t respond to pembrolizumab, what’s the mechanism to resistance? We don’t know that, but within about 5 years we’ll have a much better idea. And if this patient fell into that group, one could think about the other agents I’ve talked about. The mTOR inhibitors would be there. Even though this isn’t a stable tumor, there’s no reason why one couldn’t add lenvatinib to pembrolizumab, just as you do with microsatellite-stable patients. You also have bevacizumab as an antiangiogenic agent, so there are options for that patient.

Insight on the optimal sequencing of therapy in patients with advanced endometrial cancer based on MMR and MSI status.

Advanced Endometrial Cancer: Optimal Sequencing of Therapy 

 There’s a rationale for the use of immuno-oncology drugs like pembrolizumab in tumors that have a DNA-repair abnormality, particularly if the DNA-repair abnormality is mismatch repair. This results in the slipping of DNA as it gets replicated so that you add an extra base here and there, which then produces an abnormal protein that the body hasn’t seen. We call that a neoantigen. These mismatch repair–abnormal tumors are loaded with new proteins that the body recognizes as foreign. The immune system reacts. And the tumors are clever, so they begin to put the T cells asleep, if you will, through a number of mechanisms, including PD-1. A drug like pembrolizumab, which activates T cells by inhibiting that pathway, anti–PD-1, will activate all these T cells, which will then fight the tumor. That’s why we see such dramatic responses in these types of tumors to those agents.

That was the rationale. It evolved while the trials were ongoing, but it made a lot of sense. The classic trial for this tumor is KEYNOTE-158, which was a basket trial. It was essentially designed across many tumors, not just endometrial cancer. There were 2 cohorts, D and K, that had endometrial cancers in there, and many had mismatch repair abnormalities. It was a basket trial, so these aren’t randomized prospective phase 3 trials. These are single-arm trials. The fact that the FDA eventually approved the drug based on that tells you how dramatic the results were. If you look at a summary of the endometrial portion of KEYNOTE-158—and Dave O’Malley’s paper, a great summary in the 

—demonstrated that in MSI [microsatellite instability]–high endometrial cancers, response rate was close to 50%.

More important, we’re talking about recurrent metastatic endometrial cancer, which has already been treated with chemotherapy. These are patients who can’t be cured or haven’t been able to be cured. These are patients whose progression-free survival [PFS] is very short. They don’t do well. In this study, there was close to 50% response rate. The duration response, as reported in that review paper, wasn’t reached, meaning that they have incredible durable responses because once you activate those T cells, they keep fighting the tumor. The overall response had not been reached. The median PFS was 13.1, but that’s a little misleading because those are patients who’ve actually recurred. The main takeaway is the high response rate in a previously untreated tumor and 2 of the durations of response were home runs, which is quite spectacular.

There were no surprises in KEYNOTE-158 in terms of toxicity of this drug or tolerability. The immuno-oncology drugs do have some autoimmune phenomenon, hyperthyroidism, and thyroid abnormalities between 10% and 15%. That was present here. Others on the list were absolutely in line with what we’re used to for immuno-oncology drugs. There was no additional signal. In general, my experience in what was reported in the KEYNOTE-158 was a very tolerable regimen. These patients were frequently on this for up to 35 cycles. That’s a lot of treatment, and that reflects that it was very well tolerated.

Michael Birrer, MD, PhD, considers the role of pembrolizumab in advanced endometrial cancer given results from KEYNOTE-158.

KEYNOTE-158: Pembrolizumab Use in Advanced Endometrial Cancer

It’s all changed now because of our better understanding of the molecular biology of this tumor and new effective agents. Chemotherapy is still carboplatin-paclitaxel, and it’s standard of care. You’ll occasionally see a single-agent carboplatin in patients who are older and more fragile. Normally, it’s the combination and still very effective. With response rates, they’re very acceptable. It’s the combination that’s used in both adjuvant chemotherapy for high-risk tumors and recurrent tumors in the metastatic setting.

A subset of endometrial cancers are very well differentiated and hormonally driven like type 1 tumors that are ER [estrogen receptor] and PR [progesterone receptor] positive. They behave so well and indolent that hormonal agents are perfectly acceptable. Standard of care is Megace or tamoxifen. Some people even use a combination of those. You can use gonadotropin inhibitors, and you can also use aromatase-type agents. All of them give reasonable response rates in these well-behaved tumors. The new kids on the block are the immuno-oncology drugs. These drugs, in particular, are active and effective against tumors that have microsatellite instability. The mechanism behind that is the fact that when you have mismatch repair, you get shifts in the DNA and form proteins that have neoantigens. These neoantigens are proteins that the body has never seen. The immune system recognizes them, and the immuno-oncology drugs activate the T cells, which then fight the tumor. It’s a brilliant new approach for these tumors, and it’s quite effective.

There are other agents, like tyrosinase inhibitors. A lot of these endometrial cancers have PI3 kinase mutations or 10p deletions that activate the PI3 kinase pathway. If you go to mTOR, a protein way down on the end of that pathway, a number of mTOR inhibitors are on the market are moderately effective. Finally, we’ve known for a long time that antiangiogenic agents such as bevacizumab are active. Memorial Sloan Kettering [Cancer Center] had a phase 2 trial showing 15% to 20% response rates for bevacizumab, which was very well tolerated. Antiangiogenic agents are important. Lenvatinib, which is a somewhat dirty kinase inhibitor, certainly affects blood vessel formation for the tumors, but it probably does other things and synergizes in a very effective way with pembrolizumab. That’s my summary of the major agents for endometrial cancer.

A broad look at the systemic and targeted therapies available for patients diagnosed with advanced endometrial carcinoma.

Treatment Armamentarium for Advanced Endometrial Cancer

Endometrial cancer is usually characterized into 2 types: type 1 and type 2. Type 1 is best thought of as hormone driven, or estrogen driven. These ladies are usually postmenopausal. They have a high BMI [body mass index] and because of that, they have circling levels of estrogen. That causes endometrium to proliferate and eventually transform. When these tumors are examined, they are ER [estrogen receptor] and PR [progesterone receptor] positive and responsive to hormones. They tend to be better behaved than the other type of endometrial cancer. They invade the uterine wall less aggressively. They’re well differentiated and responsive to hormones.

The other type of endometrial cancer is much more aggressive. Type 2 is not hormone dependent. It doesn’t respond to hormones. Hormones aren’t even part of its risk factors or genesis. It’s characterized by papillary serous tumors of the endometrium. Carcinosarcomas and clear cells are in that group. Endometrial tumors that are poorly differentiated grade 3 are put into this type 2. Those are very aggressive, recur early, and metastasize very early. For hormone-driven tumors, the risk factors are obesity, unopposed estrogens, age, and so forth. Although those tumors behave a little better, we don’t know what causes the type 2 tumors. They have almost ubiquitous 

 mutations. They have a lot of genomic rearrangements in it. For instance, pap serous almost look like ovarian cancer. This patient had an endometrioid adenocarcinoma. I’m assuming this was a hormone-driven tumor. What’s interesting is that she started out as grade 1, but when the entire tumor was taken out, it was grade 2. This is that intermediate group, sometimes very hard to classify pathologically. A lot of grade 2s behave like grade 3s, and some behave like grade 1s.

From a pathology standpoint, it’s usually not difficult to tell endometrial cancer. There’s some confusion occasionally when you get a papillary serous tumor. Maybe it has already spread and is sitting on the ovary. Did it come from the ovary into the endometrium, or did it come from the endometrium into the ovary? That could be a pathologic issue, but that’s pretty uncommon and rare. The molecular analysis has changed the entire field over the last 5 to 10 years has been the recognition of microsatellite instability in some of these tumors. It’s standard of care now that these tumors have to be evaluated for that. We usually do defective MMR analysis, which is an immunohistochemical [IHC] stain for the 4 proteins involved in mismatch repair: MLH1, MSH2, MSH6, and PMS. That’s the standard of care in pathology because they can now do this IHC analysis. This patient had that and the molecular analysis, where they’re looking at microsatellites. They had a terrific work-up and identified a tumor that was molecularly mismatched repair unstable.

Let me emphasize a couple of points. First, there are 2 tests for this: IHC, which is present in most pathology suites’ standard of care, but also molecular analysis, looking at a select number of microsatellites. We use the IHC test. But sometimes, when our IHC test is normal, we’ll also go to the molecular test. Both are perfectly acceptable. What are we doing it on? Unfortunately, when you do an endometrial biopsy, a lot of times it’s not sufficient material to be able to do these tests. Usually, we’re testing the material that came out at the hysterectomy procedure where unfortunately you have a lot more tumor, but that allows you to do the test.

There’s some debate about if the tumor comes back later. Should you do a rebiopsy? That wouldn’t be necessary in this case, because we’ve already proven that the patient’s tumor has microsatellite instability. But if it were normal, microsatellite stable, and then the tumor came back a year or 2 later, would that be worth rebiopsying and looking at? Yes, it’s possible that the tumor can acquire a somatic, meaning in the tumor itself, even later in the natural history, and that there could even be a methylation event for some of these repair enzymes. Rebiopsying a patient with a stable tumor is certainly reasonable, and I’ve done that.

For those of you who might not be in an area where the pathologist or the pathology suite is capable of doing this, these of course can be done with central commercial companies in the United States. It’s very easy. They turn them around within 2 weeks.

Focused discussion on risk factors that may drive endometrial cancer followed by advice on optimal diagnostic work-up. 

Michael Birrer, MD, PhD | Authors

Articles

Advanced Endometrial Cancer: Future Directions in Care

Combination Pembrolizumab-Lenvatinib in Advanced Endometrial Cancer

Advanced Endometrial Cancer: Optimal Sequencing of Therapy

KEYNOTE-158: Pembrolizumab Use in Advanced Endometrial Cancer

Treatment Armamentarium for Advanced Endometrial Cancer

Endometrial Cancer: Risk Factors and Diagnostic Work-up