 What other agents are we looking at? Some of them I already mentioned. We’re looking at PARP inhibitors and tumors, which behave like ovarian cancers, like serous cancers. Can we predict the response in those?

There are the PI3-kinase mutations, so we have these inhibitors. The problem with that has obviously been the hyperglycemia, but we’re getting better at figuring that out. When we look at these agents, some of the other options are taking the hormonal therapy as they have in breast cancer and trying to combine it with some of these newer agents. We have an agent approved in breast cancer, the CDK4/6 inhibitors. We’re now looking at it in uterine cancer, combining that with hormonal therapy and trying to see improved results in patients who seem to have hormonally receptive tumors.

We’ve talked about the HER2 [human epidermal growth factor receptor 2]/neu candidates, those antibody-drug conjugates also being developed in breast cancer. Why wouldn’t they work in uterine cancer? We’re studying those. Dr Joyce Liu presented a WEE1 inhibitor, looking at it in serous cancers of the uterus, with very interesting results. As we look at these, it comes back to what I said: Check the next-generation sequencing, look for a targetable mutation, and find a clinical trial. Even outside clinical trials, there are some options for these patents.

As we look at endometrial cancer, we need to understand what’s going on regionally. I still believe there is a role to check lymph nodes. Sentinel lymph nodes is now a minimally invasive surgery. Why not get that information? I really want us to see if radiation is adding much of anything. We really need to answer this question. I say that, but if I have a patient with local-regional disease and know that radiation has been shown to be beneficial, I have a hard time omitting it.

I’m old-school. I admit it. My friend Dr Brian Slomovitz is going to beat me up for saying that. But that’s a really important question. Most important is this: I used to look at these patients after they completed carboplatin-paclitaxel and say, “Sorry, there’s not much else for you. We can try some hormones.” Now these patients are living years. Just in my practice yesterday, I had 2 patients who have a recurrence for 4 years. It’s almost unheard of prior to this new era that we live in. We are in the most exciting time of drug development ever. Let’s continue to get our patients these options. The best way to do that is through clinical trials.

Videos

Promising Novel Targets for Endometrial Cancer Management

 One thing with uterine cancer is that we’re finally starting to get some trials. We actually have agents out there. We have opportunities for new agents, and there are a lot of exciting new agents. I have never been more excited for uterine cancer developmental therapeutics than I am today. There are 21 large-phase trials ongoing. That’s more ongoing right now than we’ve had in probably the previous 50 years. Immune agents are at the center of this.

We’re trying to find ways to move the immunotherapy from a recurrent setting into the up-front setting. We look at the RUBY and AtTEnd trials. We look at GOG-18. What are those trials? Backbone of carboplatin-paclitaxel plus or minus I/O [immuno-oncology] agent. That’s advanced or recurrent disease. Some people would argue noncurative. What are we doing now? We’re taking patients with curative intent. My partner Dr Floortje Backes is PI-GYN-20, which is looking at patients who have intermediate-risk uterine cancer to whom you’re going to give either whole pelvic radiation or vaginal cuff brachytherapy. They’re MMR [mismatch repair] deficient or MSI [microsatellite instability] high, and they’re randomizing them to I/O [immuno-oncology], in this case pembrolizumab, or not, placebo controlled.

We’re trying to increase the curative intent with those patients who we know will have a decent chance of recurrence. Looking at KEYNOTE-B21 again—patients with curative intent. In these patients, it’s a carboplatin-paclitaxel backbone and once again looking at pembrolizumab, but more a higher-risk population of patients: advanced disease, higher risk, serous cancers. In addition, we’re now allowing radiation in that trial to see how those patients do, and that’s physicians’ choice. We’re allowing them to also utilize an agent that some of us believe may help.

What else? We’re looking at lenvatinib-pembrolizumab. I can’t believe they did this: They went head-to-head to chemotherapy backbone, carboplatin-paclitaxel in the first-line setting. We already know from a recent press release that in the recurrent or advanced up-front setting with the prior therapies—as I said, the progressive setting—lenvatinib-pembrolizumab beat physicians’ choice. We already have a press release saying that this is positive.

Now we’re waiting to see if lenvatinib-pembrolizumab beat carboplatin-paclitaxel. That is going to turn the world on its head if that’s positive. It’s a pretty big bar, but at the very least we’re going to see how they compare. The AtTEnd trial is also looking at patients with curative intent. We have the MITO END-03 trial, in which Italian colleagues are looking at patients with curative intent. We have 21 trials. What else do we have? We have better drugs—antibody-drug conjugates—targeting HER2 [human epidermal growth factor receptor 2]/neu. A couple of trials are looking at patients who would use trastuzumab and whether to either add to that, or in the recurrent setting.

There’s also combining some antivascular therapy, pan TKIs [tyrosine kinase inhibitors], and I/O therapies. So there are different ways. Looking at potentially PARP inhibitors in patients who act more like ovarian cancer, HR deficiency—homologous recombination deficiency—not mismatched repair deficiency. Seeing if bringing PARP inhibitors plus other agents is going to improve the outcomes of those patients.

Remember what I said at the beginning as we started talking: There was nothing for second line just a few years ago. Once we get beyond these agents that we’re talking about, pembrolizumab plus or minus lenvatinib, depending on the patient population, the options are really limited. We are going to have the options for those patients or, most important, a potentially higher chance of curative intent at the first line of therapy.

I almost forgot 1 of the biggest trials, the RUBY trial, which is probably going to be done with enrollment shortly. Once again, this trial is looking at carboplatin-paclitaxel plus or minus dostarlimab in these patients in the first-line setting. All these options are going to improve the outcomes of our patients, but let’s try to get them on clinical trial.

 It’s amazing that we have 2 regimens: monotherapy immunotherapy as well as the lenvatinib-pembrolizumab combination for patients in the recurrent setting. Those are game changers for our patients. We do have some complete responses. Time will tell how long those complete responses hold. That’s pretty exciting to think about as well. But patients who have partial responses for long periods of time will, unfortunately, eventually progress and will need another line of therapy. The nice thing is coming off these studies with responses, they actually feel better enough that they can get another line of therapy, which is a change from our prior options.

It’s exciting to know that there are a number of very promising agents in the pipeline that are entering larger clinical trials. I’m going to highlight 2 of them. The first has been led by my good friend Dr Joyce Liu from Dana-Farber Cancer Institute. This uses the WEE1 kinase inhibitor adavosertib in women with recurrent uterine serous carcinoma. This is a group of patients we talked about earlier in this session who have loss of 

 and very poor prognosis based on that. There may be some benefit from chemotherapy but very high risk for recurrence. Fortunately, as I just mentioned, they do tend to respond well to pembrolizumab and lenvatinib, so that would definitely be my first choice. But we have emerging data with adavosertib that this is also a compelling agent in this setting.

Dr Liu ran a single-site phase 2 trial that has 34 patients with recurrent uterine serous carcinoma. Several of them have had many lines of therapy. This isn’t a group of patients who came up 1 line and then onto adavosertib. They are very heavily pretreated. The median number of prior lines was 3. This is a different world from when I was training. You had a response rate of 30%, which is quite high when you think about it in this heavily pretreated population. Six-month progression-free survival was almost 60%. Even if they didn’t respond, you had 60% of women not progressing at the 6-month mark, which is a nice signal for clinical efficacy.

If you add the adverse events that we know well, for WEE1 kinase inhibition with neutropenia, in about 30% of patients;these are grade 3 toxicities. Patients with anemia are at about 20%, and then about 25% have grade 3 or higher fatigue. We still need to work on the dosing of this plus the population. This has moved into a larger multisite single arm, so everyone will get adavosertib. But it is moving into what will be hopefully an accelerated approval strategy to get this agent approved and available for patients with recurrent uterine serous carcinoma.

Encouragingly, there is a lot of biomarker work being done in Dr Liu’s study, which should be coming out in the press very shortly. Then there’s the upcoming study looking at 

 amplification as potentially a biomarker for patients who will respond even better to this than the whole intent-to-treat group. It doesn’t identify the patients who only respond, so it’s not an eligibility criteria, but keep an eye out for 

 as an emerging biomarker for several diseases—especially surrounding use of this class of agents. This is an exciting advance. The study is just launching, so hopefully several of you will be opening it. I think it will open and close fairly quickly given the excitement surrounding this agent.

Switching gears a little away from immunotherapies to tyrosine kinase inhibitors and kinase inhibitors, focusing on hormonal interventions and other strategies. We’ve been talking about endometrial cancer and loss of 

 and different subtypes, but a lot of our patients with endometrial cancer have endometrioid types. They have estrogen receptor–positive tumors, and we’ve been using hormonal modulation with pretty good efficacy for a long time. Can we improve on that?

There’s a lot of interest coming from our colleagues in breast cancer about using CDK4/6 inhibitors. As you know, CDK4/6 inhibitors, such as abemaciclib, palbociclib, and ribociclib, are used with aromatase inhibitors in patients with hormone receptor–positive 

-negative breast cancer in the front line. There are a lot of data in this hormone receptor–positive breast cancer population with not only progression-free but overall survival data. Would that translate to hormone receptor–positive endometrial cancer as well?

Dr Mansoor Mirza and the ENGOT [European Network for Gynaecological Oncological Trial] group presented this at ESGO [European Society of Gynaecological Oncology International Meeting]. It’s also been published. This is the ENGOT-en3 study. This is a randomized double-blind phase 2 study in patients with recurrent endometrial cancer, and they were randomized to the CDK4/6 inhibitor palbociclib plus letrozole, compared with letrozole. Letrozole is probably not the first choice of any of us for hormonal management in an endometrial patient. We probably would be using medroxyprogesterone with or without tamoxifen. That’s 1 weakness of the study, but it is cleaner to look at it vs letrozole. I do think the control arm is something we have to think about it.

But be that as it may, it was a randomized phase 2, really well-designed blinded study, and they looked at progression-free survival as an end point. They had 77 patients enrolled, so it was a nice-size study. What you saw was that the use of palbociclib and letrozole significantly improved the progression-free survival compared with letrozole and placebo. The median progression-free survivals were 8.3 vs 3 months, respectively, and the hazard ratio was 0.56. Even though I don’t love the letrozole as a control, I don’t know that medroxyprogesterone would have been that much better. It may have been a little better, maybe 4 or 5 months, but the combination would have still won.

These are really compelling early data. You did see the traditional adverse effects. We know these well from breast cancer, with palbociclib. Neutropenia is the big thing. Grade 3/4 neutropenia is 42% with palbociclib vs none in the letrozole arm. That’s well known from the breast cancer literature. It’s 1 of the known adverse events with CDK4/6 inhibition. These are compelling early data. There are planned studies that are just launching in the United States, a bit of a me-too design, but just trying to confirm this. I’m guessing this regimen will get some sort of an NCCN [National Comprehensive Cancer Network] listing at some point. Then we’ll have to see where it fits in the lineup of options for patients with recurrent disease.

We have data from Dr Brian Slomovitz’s work with everolimus and letrozole in the recurrent, but also in the untreated population of patients with hormone receptor–positive endometrial cancer. That looked very compelling instead of chemotherapy. I know I’m probably not alone: There are some patients with recurrent advanced disease that I start on everolimus-letrozole, but I used to start them on medroxyprogesterone-tamoxifen, and you would get years on that. I’m interested in how this story plays out and how I would sequence this particular regimen in individual patients based on their presentation, performance status, and expectations of efficacy. These are very compelling data. Time will tell where it ends up in the sequencing of things, but it’s definitely something to keep an eye on.

Novel Strategies in Development for EC Management

 Let’s talk about the sequencing in patients with recurrent or advanced uterine cancer. As I’ve said, carboplatin-paclitaxel is my backbone. I’m always looking for a clinical trial, particularly in advanced and recurrent patients. We have to do better for these patients. What happens if they’re not a candidate for a clinical trial—if they were just diagnosed with breast cancer and it excludes them? My backbone is carboplatin-paclitaxel. I’ll utilize that in the recurrent disease to tolerance or toxicity. The reason being that I don’t have very good luck getting complete responses in people, and like in ovarian cancer, stop them and maintain them.

However, I recently have utilized the carboplatin-paclitaxel, and if they have a tumor that is low grade or ER+ [estrogen receptor positive] or PR+ [progesterone receptor positive], I switch them to a hormonal-based therapy like aromatase inhibitors. I’ve used tamoxifen-based combination—probably 1 of the most underutilized combinations that we have. It’s well studied in the GOG [Gynecologic Oncology Group], usually very well tolerated, and it gives them a break from the cytotoxic drugs. If after 6 to 8 cycles of the cytotoxic they’re starting to get toxicities, look to transition them into a better-tolerated regimen like hormonal therapy.

If they have a serous cancer and they’re HER2 [human epidermal growth factor receptor 2]/neu positive, then I obviously continue the trastuzumab. I stop the backbone of the carboplatin-paclitaxel after 6 cycles, then I continue trastuzumab until toxicity or progression. If a patient progresses through carboplatin-paclitaxel and they’re MSI [microsatellite instability] high or MMR [mismatch repair] deficient, then it’s really easy: pembrolizumab. If they’re proficient, it’s really easy: lenvatinib-pembrolizumab.

If a patient has a great response or it’s been years since they’ve seen carboplatin-paclitaxel because they’ve received it for adjuvant therapy, I may go back to carboplatin-paclitaxel. But if they’re proficient or deficient, I look for immunotherapy. The response rates there are amazing. Unlike chemotherapy in the recurrent setting, I do see complete responses. It’s probably only about 10% of patients, but I do see complete responses.

Therapy Sequencing Through Multiple Lines of Therapy in EC

 As I look at the experience of lenvatinib-pembrolizumab, we’ve treated a fair amount of patients. Once we had the indication, patients didn’t have any other option. I’ve used a lot of lenvatinib-pembrolizumab. I have an amazing team in place. I have a full-time nurse and a full-time nurse practitioner on my team, and they are 100% essential. In addition, we have a specialty pharmacist who sits down with the patient when prescribed to give them information and prepare them. I talk to them, the nurse talks to them, the specialty pharmacist talks to them, and then we have weekly video conferences with the patient to manage symptoms.

It is absolutely essential to educate the team about the problems these patients will have so that they get recognized and we don’t turn a small problem into a big 1. My experience is that the data in the clinical trials showing 70% to 80% dose reduction are absolutely accurate. Most patients will require dose reduction. Sometimes, I’m surprised when a patient who I think is going to require dose reduction actually tolerates the medication better than I predicted. What are some of the other things I worry about? You have to make sure you’re keeping an eye on the creatinine clearance. Patients who come in with preexisting renal disease with a creatinine clearance less than 40 mL/min should be dosed with only 10 mg of lenvatinib.

I’ve talked about how I practically manage the dose reduction, which is the 18-mg prescription. Some people have managed it by starting at a lower dose, but that always makes me a little worried. There’s some hint in the data that the loading dose of the higher dose may be helpful. I’m not sure. My concern is starting at a lower dose. If they tolerate it, fine. Then you give them the higher 14 mg, they have a little bit of symptoms, and they want to return to the 10 mg.

These patients are going to have symptoms. We have to continue to get better at managing their control, which every single day my nurse, nurse practitioner, and I are learning. We continue to improve. There are now some really nice guidelines and patient education that shows when these toxicities occur. Most of the dose reductions with lenvatinib are going to be in the first or second cycle. Then the I/O [immuno-oncology] toxicities will be beyond that. Of those, I always worry about the diarrhea. Is it the immune-associated colitis? The diarrhea will occur in the first cycle or maybe the second cycle, so it’s very unlikely to be I/O associated.

Some of the dermatologic conditions are usually able to be managed. I have become very good friends with my multidisciplinary collaborators. Our dermatologists have also instituted a remote consult where we can take a picture and send it to them. I’ve had a lot of run-ins recently with endocrinology. I have a couple of patients on single-agent I/O or the combination who have had significant development of type 1 diabetes. I’ve had a recent run of other complications for which I’ve had to involve rheumatology—some of them on single-agent I/O, some on a combination, but not all of them on lenvatinib-pembrolizumab.

This is across the spectrum of the I/O toxicities. The multidisciplinary approach is very important. It’s very important to educate your nursing team, nurse practitioner, physician extenders, and most important, your patients.

 This is the accelerated, now likely confirmed, approval of lenvatinib and pembrolizumab, the already-established approval of pembrolizumab for mismatch repair–deficient patients and the pending approval of dostarlimab. We used to have nothing in this space that helped patients, and now we’re going to have 3 new regimens for patients with recurrent endometrial cancer. You would select it based on biomarker selection. Either mismatch repair deficient can get monotherapy or mismatch repair intact can get the combination. We’ll have to see the data from the phase 3 study, where there were patients with mismatch repair deficiency enrolled on that, and see if we need to use the combination in that patient population or just stick with monotherapy. That’s 1 of the questions I still have about the data. Sometimes the enemy of good is better. If I can do really well with a monotherapy and have less toxicity, should I stick with that or should I go for broke and use both in a mismatch repair–deficient patient? That’s 1 of the clinical questions that remains outstanding.

The phase 3 study won’t answer that because, if you remember, the control arm is chemotherapy, not single-agent immunotherapy. We can get a sense from the efficacy in the mismatch repair–deficient population compared with monotherapy of what we think the add-on is from the lenvatinib and that group. Unquestionably, in a mismatch repair–intact population, the combination is the new standard of care.

While that response rate and durability of response rate importantly are very exciting, this combination is challenging to use because of the toxicity. If you recall, the starting dose of lenvatinib is 20 mg, and if you look at Dr Vicky Makker’s original paper, over 80% of patients required a dose reduction because of mainly—not entirely—lenvatinib-related toxicity. Fatigue and diarrhea were some of the main ones. You can also get some PPE [palmar-plantar erythrodysesthesia] about 8 weeks in from starting lenvatinib or from increasing the dose. So there are some toxicities that are a challenge, especially if you’re starting it in a patient who is already somewhat ill.

Those of us who have been using this for a while have developed our own strategies for how to start it. This is off-label, so I’m just going to say that. The label is 20 mg. There are certainly providers that start at 20 mg and dose reduce in the event of toxicity. I tend to start at 14 or 18 mg. Honestly, I prescribe 18 mg because it comes as 10 mg and two 4s. I start them at 14 mg for the first month or 2 and make sure we’re not going to have hypertension—another big issue, fatigue, or diarrhea. I just see how they’re doing. If they tolerate that very well and it usually works, and they at least have stable disease or some disease reduction, then they have that other 4 mg. And I just add it on, bump them up to 18 mg, and then keep them on that for a while. That’s really where I’ve stayed.

My patient population is potentially a little older and a little more medically frail than other places, but I haven’t really taken anyone up to 20 mg. I’ve definitely seen efficacy. I was a little dubious when I first heard about this. I wasn’t convinced because of the toxicity. But now that I’ve used it and I’ve gotten some familiarity with the drug and comfort with how I’m going to dose escalate it, I’m using it much more frequently. It has worked in patients who are pretty heavily pretreated, and we’ve used things instead of this combination. It has worked. It’s January. I have a patient right now who I did not think was going to make it to Christmas. She not only made it to Christmas but got to drive to Florida and see family, and she’s continuing on with stable disease right now. I’m pretty amazed. We’re 6 months in.

This is a transformative regimen, but it takes education around the adverse effects and an office that understands the patients on this drug and can respond very promptly to a patient call with problems, so that you can adjust, hold, and maintain patients on this very effective therapy for as long as possible. If they come off it or are constantly being held because of uncertainty, it won’t be very effective. Drugs don’t work if you don’t take them. This is 1 that’s fantastic, but it requires quite a bit of intentional education to use. I think we’re all up for that challenge.

Impact of Pembrolizumab/Lenvatinib Approval in EC

 Let’s move on to KEYNOTE-146, which I refer to as LEN-PEM because it gets a little difficult to say lenvatinib and pembrolizumab. Dr Vicky Makker reported the final analysis that was published in the 

], looking at and diving down into the different aspects of these patients. When we look at the final report, Vicky does a wonderful job describing that there is a difference between patients who have received prior chemotherapy for adjuvant vs patients who have just received carboplatin-paclitaxel for recurrent disease. Meaning: If they have recurrent or advanced disease, they get carboplatin-paclitaxel and then move right into lenvatinib-pembrolizumab compared with a patient who received carboplatin-paclitaxel as adjuvant and then recurs and gets the agents of lenvatinib-pembrolizumab.

In the total population of patients, we see a response rate, depending on the report, of about 37% to 38%. We see this 37% to 38% in the entire cohort of patients, which is just under 100.

Dr Makker does a wonderful job in the most recent report looking at the differences between patients who have been treated with carboplatin paclitaxel vs those who received adjuvant therapy.

This is very interesting. When we look at the population of patients, the patients who received carboplatin-paclitaxel and then went right into LEN/PEM [lenvatinib-pembrolizumab] had a response rate of 41%, while the patients who had previously received carboplatin-paclitaxel for adjuvant and local-regional disease as the adjuvant setting had a 57% response rate. When we add up the numbers, they’re slightly different because there’s another group of patients who had more than 1 prior therapy, and that group of patients is going to be a little lower. So the entire population is about 37% to 38%.

Those who went right from carboplatin-paclitaxel are at about 41%, while those who received previous adjuvant therapy are at 57%. What does that mean to me? That means I always quote to my patients 37% or 38% response rate. But if I take a patient who had received previous therapy of carboplatin-paclitaxel for adjuvant, their response rate is up to about 57%. This is a small group of patients. We need to see the confirmatory trial, which will hopefully be reported shortly to drill down into this answer.

The toxicity with LEN/PEM [lenvatinib-pembrolizumab] is real. You have all the associated I/O [immuno-oncology] toxicities, then you bring in lenvatinib. So let’s talk about the I/O–associated toxicities. We need to learn from our lung cancer and melanoma experts because they have so much more experience with these agents. When I discuss with a patient the toxicity associated with pembrolizumab or other I/O agents, I talk about a 30% to 40% chance of an I/O-associated toxicity, which would require me interrupting treatment and potentially giving steroids. Of that 30% to 40%, 5% to 10% will have toxicities that will potentially require hospitalization.

We’ve gotten better at managing these toxicities, so I suspect that number is closer to 5% now. But within that population, we have about 1% to 2% of patients who will have life-threatening or life-altering toxicities. Somebody becoming a type 1 diabetic or losing their function of the pituitary is life altering. That’s rare. But if it occurs, it is obviously life altering. Combining with that lenvatinib, what are the other challenges with lenvatinib? Hypertension, diarrhea, some skin changes can be an issue, hair changes, some whitening of the hair. The ones I worry about are the diarrhea and the hypertension.

These are patients who I give the antidiarrhea directions to and make sure they have the medications at home. We also give them the prescription. I have colleagues who make them fill a prescription so they have something at home while we give them the prescription, because we always know these high blood pressures occur on the weekend without our usual team in place. We give the patient a prescription if they need them, and we make sure we have them monitored closely. What else have we done? We’ve instituted weekly videoconferencing with the patients. One thing that has come out of the COVID-19 [coronavirus disease 2019] era is the ability to perform videoconferencing.

We now have once-a-week videoconferencing for the first cycle and potentially second cycle to make sure we’re on top of these toxicities. You have to remember that about 80% of patients on lenvatinib-pembrolizumab will require dose reduction. Some people would argue 88% if you include the dose interruptions. The usual starting dose is 20 mg. That’s the recommended dose. Personally, 20 mg can be a challenge because they get only two 10-mg tabs.

If you dose reduce, you’re going right from 20 to 10 mg, and I’m telling you that 70% to 80% are going to be dose reduced. What we do is order the 18 mg. That comes with a 10-mg tab and two 4 mg tabs. I feel much more comfortable with a 10% adjustment from 18 to 20 mg. It’s unlikely that we’re impacting efficacy. Then when that patient needs to be dose reduced, I can dose reduce according to the package insert down to 14 mg, and if I need to go down to 10 mg, they still have that option.

Safety/Efficacy Results From EC Cohort of KEYNOTE-146

 I just finished talking about monotherapy immunotherapy and how landscape changing and world changing it has been for women with mismatch repair deficient [dMMR] recurrent endometrial cancer, which is about 30% of women who present. You see that across the board in studies. That’s really important, but what about the other 70% who don’t have mismatch repair deficiency? Are they just left with our standbys of doxorubicin and weekly paclitaxel with expected response rates of 15%? That hardly seems fair. Fortunately, they are not left with that. They actually have an equally good option, also accelerated approval, and hopefully will be confirmatory approval very soon with the combination of pembrolizumab and lenvatinib.

This is a really interesting combination. Dr Vicky Makker from Memorial Sloan Kettering Cancer Center has led developments of this combination in endometrial cancer. This originally was an investigator-initiated trial. Those of you who run IITs [investigator initiated trials] know that they can sometimes result in tremendous opportunities for our patients; this is an example. Congratulations to Dr Makker, who recognized that both drugs individually had some efficacy. Lenvatinib, just to remind you, is an oral tyrosine kinase inhibitor that targets a number of things: VEGF1, VEGF2, VEGF3, FGFR, PDGFR, BIK, and MEK. I may be even forgetting 1. It’s kind of a dirty tyrosine kinase inhibitor, mainly acting in an antiangiogenic space. We know from prior studies that antiangiogenic agents have efficacy in endometrial cancer. A very large, over 100-patient, single-arm phase 2 was done with single-agent lenvatinib with a response rate of 14%. It has some activity, certainly, but it’s not overwhelming.

Similarly, if you look at pembrolizumab in microsatellite-stable patients, or those patients who don’t have mismatch repair deficiency, you see a similar response rate. It’s about 10% to maybe 15%. There is some efficacy with each drug, but it’s not great in a microsatellite-stable population.

There’s a lot of rationale for combining them. When you think about that VEGF, it is a drug that promotes angiogenesis, and that’s how we all think about it, but it’s also incredibly immunosuppressive. If you have high levels of VEGF around your tumor, you’re creating a microenvironment that is incredibly immunosuppressive with high levels of immunosuppressive macrophages or Tregs but not a lot of cytotoxic CD8 cells. When you block VEGF with tyrosine kinase inhibitors—cediranib is very good at this, lenvatinib actually may be a little better—you’ll not only block the antiangiogenic effect but also convert this immunosuppressive environment into a less immunosuppressive environment, allow more CD8 T cells to come in, and by doing so enhance the efficacy of immune agents such as pembrolizumab.

This was demonstrated in elegant animal models and then moved into IIT that I talked about, which was a single-arm study that launched in a few centers in the United States and used the combination of lenvatinib and pembrolizumab in patients who had disease that had recurred after frontline paclitaxel and carboplatin. So many patients were enrolled who were mismatch repair deficient, but most of the patients were mismatch repair intact. They found a remarkable response rate in this group of patients who heretofore had nothing active in their toolbox. They had an overall response rate of almost 40%, several standard deviations above what would be expected with standard options.

In addition to this, about 11% of patients had complete response rates, which is something we had never talked about in this disease space. Suddenly, we’re putting patients into some kind of a remission, which is pretty remarkable. At the time this was presented, the duration of response had not been reached, but it’s right about a year, with most responders going well beyond 6 months.

When you look at the breakdown among the different histologies—and these are small numbers, so we can’t hang too much on it— it is quite exciting. There were 5 patients who had clear cell tumors and each of them had a reduction in their tumor size. Not all of them were a RECIST response of a 30% reduction, but all of them had a reduction in their tumor size. If you looked at the patients who were classed by the serous, their response rate was even higher than the overall group. So this combination appeared to be synergistic in a group of patients who would not be expected to respond to immune checkpoint inhibitors alone.

This got very quickly accelerated approval, not only in the United States, but simultaneously in Australia and Canada as well through a very interesting program. As with all accelerated approvals, there was an ongoing confirmatory trial that compared the combination vs standard medicines, but we have not seen the data. We hope to see it in spring, but the press release has just come out and stated that it met the primary end points and was superior. The anticipation is that this will receive confirmed approval and then hopefully will be available outside the United States as a global approval as well, which is an incredibly important advance for women with recurrent disease. We have 2 options now based on molecular subtypes, both of which with a high likelihood of helping patients for a durable period of time. That’s a world of difference from where we were a few years ago.

Rationale for Combination Therapy in Advanced EC

 One of the most exciting developments in the field of therapeutics for endometrial cancer has been immunotherapies. We’ll talk quite a bit about monotherapy immunotherapy in this module, then later we’ll talk about combinations. For women whose tumors harbor mismatch repair deficiency [dMMR], microsatellite instability [MSI], and maybe even tumor mutational burden [TMB]—that’s not a validated biomarker yet, but we’re certainly looking at it—use of monotherapy immunotherapy has been a game changer for women with recurrent disease.

Prior to the pembrolizumab approval, there was nothing FDA approved. Our options for second-line treatment of endometrial cancer was doxorubicin or weekly paclitaxel. We have phase 3 data for both of those agents in that setting with a response rate of 15%. That’s the best our patients could expect. There wasn’t really an option for third-line treatment of endometrial cancer because our patients, unfortunately, weren’t living to third and fourth line. When you think about it, that’s not too long ago. Now we’re in this immunotherapy era. It’s such a different environment for our patients, thankfully, with monotherapy immunotherapy.

Pembrolizumab was the first to get approval based on a very large basket trial, and some patients with MSI-high endometrial cancer were included, which was fantastic. Now we have dostarlimab. The GARNET study was presented in 2019 and published in 

 in 2020. This was led by Dr Ana Oaknin from Spain. The GARNET study was a huge phase 1 study looking at the efficacy of dostarlimab, which is an anti–PD-1 monoclonal antibody, in patients with recurrent endometrial cancer. They had a number of molecular cohorts, but the 1 we’re talking about today is those patients who have deficient mismatch repair proteins or microsatellite instability. They also had to have had at least 1 prior line of chemotherapy. This is second line and beyond recurrent disease.

This is a very large study. They accrued over 100 patients into this study, so it’s a very large safety database as well as an efficacy database. The efficacy was what we would expect in this population: fantastic. If you look at the 71 patients who were confirmed deficient mismatch repair, the complete response rate—which I don’t think we’ve ever talked about in this disease setting—was almost 13%, with an additional 30% being a partial response. The overall response rate was about 43% or 44%. An additional 15% to 16% were stable disease, which often can also be a win in this disease setting if it’s long-standing. The duration of response was so long that it wasn’t even reached, so we can’t report that at this time.

The other thing that’s important to note is that because this was such a large study, we have nice data for safety. The safety was what we would expect from this class of agents, with very low rates of immune-related toxicities. Those toxicities are real. As we’ve been using these agents more commonly, we’re all getting more familiar with recognizing them and intervening early to prevent them from becoming severe. The toxicities are uncommon but can be important.

We saw low rates of anemia in this study at about 3%. That’s probably disease related. We saw a very low rate of 2% colitis, which is 1 of the adverse effects we always focus on. And then diarrhea had a rate of only 2%. Those were the grade 3 or higher treatment-related adverse events. That’s a very manageable adverse-effect profile and compelling clinical efficacy. This was 1 of the largest studies done just in women with endometrial cancer, so it gives us a really nice data set, proving what we all suspected: This is an active drug.

We’re hopeful that dostarlimab will gain accelerated approval in this setting and be available for us to use in addition to pembrolizumab. It’s always great to have more than 1 option for our patients. From a cost standpoint, for whatever reason, drugs are slightly different from one another. If 1 is working but not tolerable, another may be tolerable. We don’t have data for that, but we’ve certainly seen that with other classes of drugs—PARP inhibitors, for example. Some people tolerate 1 and not the other. We’ve certainly seen that with tyrosine kinase inhibitors. Having more than 1 similar asset available in a disease space is important for a number of reasons—most important access, so that everyone, if for whatever reason that they may not be able to access 1 drug, can get the other and benefit from durable, tolerable efficacy. This is really exciting. Congratulations to this study team.

GARNET Clinical Trial Regimen in Advanced Disease

 As we look at the data in MSI [microsatellite instability] high from KEYNOTE-158, I was honored to present this at last year’s ESMO [European Society for Medical Oncology Congress] in a featured poster. This data has now been published in the 

], looking at response rates in MSI-high endometrial cancers of 57%. That was markedly better than the original reports, which were about 35% to 37%. As we look at these patients, the clinical benefit is beyond the objective response rate of 57%, with most patients achieving some clinical benefit.

It’s very important when evaluating the biomarkers—at the very least in recurrence, but you could argue at the time of diagnosis—to know what the mismatched repair protein expression is. It’s very important if that is with advanced disease. I would argue to also check PCR [polymerase chain reaction] for MSI.

The impact of pembrolizumab availability has been marked. We previously didn’t have a decent second-line drug for endometrial cancer. It has completely changed the way I look at a patient with recurrent uterine cancer. I’ve always said there’s no possibility of curative intent in recurrent cancer. But now I have to think twice before I say that. When looking at this 57% or 58% response rate, about 10% plus or minus will get complete responses.

All of us who use these agents have several patients with recurrent uterine cancer in our practice who have been cured. We are in a completely new world, and all of us who treat these patients see this. To deny a patient who has MSI-high or an MMR [mismatch repair]–deficient uterine cancer the access to pembrolizumab, I would argue, is not adhering to standards of therapy.

Let me clarify 1 point. When I was talking about access to pembrolizumab for patients who are mismatch repair deficient, I was talking about patients who have recurrent tumors or who have already been treated with carboplatin-paclitaxel. Combining pembrolizumab with carboplatin-paclitaxel has not been adequately studied. Ongoing trials are evaluating whether it adds to carboplatin-paclitaxel. When I’m referring to pembrolizumab for mismatch repair–deficient patients, I’m referring to the recurrent setting. I would not recommend utilizing it instead of carboplatin-paclitaxel at the current time.

Many people ask me what other tumor markers are checked, such as PD-L1 expression and TMB [tumor mutational burden]. I’m pretty adamant about not checking PD-L1 expression. It has not been shown to be predictive in uterine cancer, and we don’t have an indication. TMB is a different story. Do I send it separately? Not usually, but if it’s part of a commercial package. When I check NGS [next-generation sequencing], I definitely will include it. Obviously with the indication in pembrolizumab, I haven’t found that if somebody is an MMR-proficient tumor that they are then TMB positive, but it’s probably worth checking if you’re sending it off. I don’t normally check it separately.

Second-Line Checkpoint Inhibitor for MSI-H Disease

 I mentioned that our understanding of stage in a recurrence and now our emerging understanding of the molecular subtypes is really impacting how we assign adjuvant therapy with the diagnosis of endometrial cancer. You can look at 2 recently published manuscripts that come from different eras to get a sense of this. We’re excited that we saw Gynecologic Oncology Group [GOG] protocol 0209 finally published in 2020 by Dr David Miller. Congratulations.

However, this study was done over a decade ago. They enrolled women with advanced or recurrent disease. These could be patients whom we would consider adjuvant treatment—maybe they had a surgery with a positive node, all the way to stage IV disease. They randomized them to receive paclitaxel and carboplatin, which was the experimental arm at that time, vs the standard of care at the time, which was the triplet of paclitaxel, doxorubicin, and cisplatin given over 2 days. It was a noninferiority design and paclitaxel and carboplatin was found to be noninferior and, therefore, an established standard of care that many of us had moved to already many years ago. But now we have the data to back that up. When you look at the stratification in that study, there was no mention of molecular subtyping. We just weren’t doing that in that era.

Fast-forward to another publication from this year, which was the PORTEC-3 trial. This was a European study that was questioning the benefit of adding chemotherapy to radiation therapy in patients with stage I through III disease as adjuvant therapy. They did extensive molecular profiling as an exploratory analysis in about 500 women. They looked at the markers that I previously mentioned in this module. They looked at loss of 

, and mismatch repair deficiency. They also had a group of patients in whom no significant mutations were found. They looked at the prognostic significance of those 4 classifications then they looked at the significance of those classifications on whether a patient received chemotherapy with their radiation therapy.

As I mentioned previously, the groups definitely were prognostic. When you look at the patients who had loss of 

, their 5-year recurrence-free survival was only 48%, compared with 98% for women whose tumors had a 

 mutation, 72% for women with mismatch repair deficiency, and 74% for those with no significant mutation noted. There definitely is a prognostic significance associated with 

Layered on to that are suggestions of predictive biomarker potential with these markers as well. If you look at the patients with loss of 

 who did or didn’t receive chemotherapy with their radiation—everyone received radiation—the 5-year recurrence-free survival for chemotherapy was 59% vs only 39% if they only received radiation. This group of patients may receive more benefit from chemotherapy than those who don’t have a loss of a 

, they all did so well that it was hard to show a difference. It was 100% vs 97% for chemotherapy vs no chemotherapy. If you looked at the patients with mismatch repair deficiency, it was 68% 5-year recurrence-free survival with chemotherapy vs 76% without. Those are pretty close. The confidence intervals overlap, so it’s probably not significant. The question we need to answer is whether we need to use chemotherapy in that group. Similarly, for those patients with no significant mutation profile, there wasn’t a statistically significant difference in the response.

This is an exploratory analysis so we can’t hang our hat on this, but if you look at the group of patients who have loss of 

, the difference between 59% and 36% is a striking difference in recurrence-free survival and may give us more insight into who needs adjuvant chemotherapy with or without radiation with endometrial cancer.

If you think back to GOG-0249, which was based on high-intermediate risk criteria and added chemotherapy to radiation vs radiation alone, it did not show a difference in this high-intermediate risk group. Maybe that’s because we were still including a group of patients who were just too low risk for recurrence. It’s really focusing on potentially this group of patients with loss of 

. That may be the group who needs chemotherapy. That will have to be validated in a study designed around that end point, but this PORTEC-3 exploratory analysis is certainly a step in the right direction. From a practice standpoint, being off–clinical trial gives you some information about when you add chemotherapy and when you potentially don’t need to in these patients. It’s very important information. I’m glad this was published this year.

Adjuvant Chemotherapy +/- Radiotherapy: Recent Data

 The challenges of advanced disease of uterine cancer are most profound in that we still continue to hang with the history of radiation therapy. Historically, all these tumors were treated with radiation therapy. But every time we do a trial comparing radiation with chemotherapy or radiation with observation, we’re called into question the importance of radiation. With the recent PORTEC-3 trial, we now see that the combination of radiation and chemotherapy may provide benefit. But one could argue that radiation did not add much; it was the chemotherapy.

Some of the challenges we need to look at are: What’s the most important modality, and how many modalities do we need? The most important modality is chemotherapy, but radiation may play a role—especially in locoregional disease, meaning lymph node metastasis. What else is challenging? If we have locoregional disease only with curative intent, how much do we have to do it? Is radiation alone enough? Do we even need radiation? Is chemotherapy alone enough? Are there patients we could just observe after surgery? How about the patients who had a hysterectomy but did not have lymph nodes removed? Do we need to remove those lymph nodes? Every trial we’ve had on surgical staging of patients comes into question, but I digress.

When we’re looking at the advanced uterine cancer and we start thinking about patients who have metastasized beyond the locoregional, their outcomes are very poor. What do we know? The tried-and-true carboplatin-paclitaxel is still our go-to and the basis of each of our backbones of these clinical trials. That’s being questioned with the current LEAP-001 trial looking at lenvatinib and pembrolizumab vs the standard carbo-Taxol [carboplatin, paclitaxel]. We’ll talk about that in a little while.

There are many challenges in the treatment of recurrent uterine cancer. One is that if they’ve already received carboplatin-paclitaxel in adjuvant therapy, should we go back to carboplatin-paclitaxel? Or should we look at some of our new agents, like pembrolizumab in MSI [microsatellite instability] high? Or lenvatinib and pembrolizumab in MSS [microsatellite stable] patients?

We talk about platinum resistance and sensitivity in uterine cancer, but we do that with almost 0 data to influence it. We’ve extrapolated from uterine cancer, but when we really look at it, there’s no way we should be extrapolating. Is there a group of patients I would absolutely put back on carboplatin-paclitaxel? Yes, if it has been years since their prior adjuvant therapy. But as we look at the patients who have received carboplatin-paclitaxel, their response is decreased when they receive it a second time. We need more data on this population. Do I go back to it? Absolutely. Do I have data to support that? No. We need better data to support when we should go back to carboplatin-paclitaxel. I use a lot of the other agents available to us, as far as standard of care or clinical trial—rather than going back to carboplatin-paclitaxel—unless the interval is very long. Is it 2 years, 3 years? Probably 2 years.

Treatment Challenges: Advanced Endometrial Cancer

 We’ve learned a tremendous amount in the past 3 to 5 years about the importance of molecular profiling in patients with endometrial cancer, both in early stage and, probably more important, for treatment selection in advanced-stage and recurrent-stage disease. We’ll talk a little more about that throughout this module.

What we’ve really learned is that all patients, regardless of stage or histology at the time of diagnosis, must have their tumors tested for mismatch repair [MMR] proteins, plus or minus microsatellite instability, and every institution now has its own algorithms developed. I’m pleased to hear that this has rolled out very quickly both in the community and academic centers, so at least patients with a new diagnosis are having mismatch repair proteins tested. This is important because it identifies patients who are at risk for having Lynch syndrome, which is a genetic predisposition to developing endometrial cancer.

Regardless of risk for recurrence, if we diagnose women who have this genetic predisposition, we can prevent them from developing a colon cancer later in their life through appropriate risk-based screening once we’ve cured them from their endometrial cancer. We can do cascade testing for relatives and try to create “previvors” and do risk-reducing surgeries and screenings that prevent them from ever-developing Lynch-associated cancers, which also includes some ovarian cancers. This is incredibly important from a family testing standpoint and prevention standpoint but also in terms of identification of women who will potentially benefit from immunotherapies. We’ll talk about that a little later in this segment.

The same goes for women who are already advanced, have recurrent disease, or are at high risk for recurrent disease. We are testing everyone for mismatch repair proteins. This is less a must, but many of us are doing next-generation sequencing [NGS] for these patients. There, you get tumor mutational burden, microsatellite instability, and 

. You also get some less frequent, but very important from a predictive biomarker standpoint, findings of 

 mutations are less common but still present, and there are others. 

 may be targetable in the future. There are a number of these molecular findings that do impact selection of therapies in the recurrent setting. In the advanced or recurrent setting, next-generation sequencing is certainly warranted.

In terms of thinking about patients who are newly diagnosed and looking at their molecular subtype and histology, we are increasingly understanding the impact on prognosis. For many of you listening to this, it has been a journey. We’ve talked about patients who were intermediate risk if you think back to Gynecologic Oncology Group [GOG] protocol 33, and then high-intermediate risk was determined from GOG-99, based on grade, depth of invasion, and age. We ran clinical trials trying to figure out who was at higher risk for recurrence and how we could impact that recurrence based on risk stratification. Now, when you layer in the molecular profiling on top of that, we have an enriched manner by which we can assess someone’s prognosis and maybe even their appropriateness for adding chemotherapy to radiation therapy in an adjuvant setting based on these molecular subtypes.

When we’re talking about our patients who have loss of 

, that’s mainly going to be your high-grade serous patients but also some high-grade endometrioid, patients who have 

, patients who have mismatch repair deficiency, and a group of patients who have no significant mutations identified. The prognosis in each group varies if you use that group of patients with no mutations as the baseline. Our worst prognosis is among patients with loss of 

. Patients with mismatch repair are somewhat in the middle. We are starting to tailor therapies based on those molecular subtypes and moving a little from our prior way of looking at things by intermediate and high or immediate risk. It’s very much an emerging paradigm for selection of therapy for the right patient.

Genomic Assessments for Metastatic Disease

 It appears that endometrial cancer is actually increasing in its incidence. Why is that? Well, we know as a society we’re getting heavier, and that’s not helping. That only explains the higher number of lower-grade endometrioid cancers. We are also getting older, and as we age, we are seeing higher incidence of these type 2 or high-grade-type uterine cancers. As we age and get heavier as a population, that directly impacts the incidence and prevalence of endometrial cancer.

As the risk stratification of uterine cancer continues to change and metamorphose over time, we still use the classic definitions of histology, endometrioid, serous, and clear cell. Those are still very important. What else are very important? The uterine factors, looking how deep the tumor goes into the uterus, and looking to see if the cervix has stromal involvement. The tubes and ovaries still apply. Those are very important, along with metastatic disease to the regional lymph nodes, or intra-abdominally. That’s extremely important.

What else have we developed? Now we’re looking at the molecular changes in these tumors, and they may predict better than the histologies. 

 expression, mismatch repair [MMR] protein deficiencies, and 

, are all very impactful on the behaviors of the tumors and may be more important than histology.

But I’m not willing to give up histology. As we move forward, we still need to know these important aspects. As we look forward to the different classifications, the TCGA [the Cancer Genome Atlas] started to do this: low mutational burden, high mutational burden, 

. Each of these categories continues to be refined, and some additional data continue to come out that we’re able to apply to the clinical patient population at the direct patient care. Meaning, it’s part of not just research trials but everyday care: 

, MMR deficient, MSI [microsatellite instability] high.

Kathleen Moore, MD, Stephenson Cancer Center

Articles