Immuno-Oncology Approaches Continue to Advance Across Gynecologic Cancers

Michael J. Birrer, MD, PhD

Single-agent and/or combination immunotherapy approaches continue to reign supreme in the treatment of patients with endometrial cancer. Several recent trials have reinforced their role in the first and second lines, and some research suggests that single-agent immunotherapy may reduce the need for chemotherapy in specific patient subgroups, according to Michael J. Birrer, MD, PhD.

In an interview with OncLive^® following an Institutional Perspectives in Cancer webinar on gynecologic cancers, Birrer, chair of the event, highlighted the effect of improved molecular characterization, the clinical significance of the KEYNOTE-775 trial (NCT03517449), and recent changes in the frontline treatment landscape in advanced endometrial cancer.

He also expanded on advancements in ovarian cancer as discussed by his colleagues from the Winthrop P. Rockefeller Cancer Institute, including the shifting role of PARP inhibitors and antibody-drug conjugates (ADCs), and the potential use of cancer vaccines to maximize the efficacy of immune checkpoint inhibitor (ICI) regimens in this space.

Birrer is vice chancellor, director of the Winthrop P. Rockefeller Cancer Institute, and director of the Cancer Service Line at the University of Arkansas for Medical Sciences (UAMS) in Little Rock, Arkansas.

OncLive: How has increased molecular profiling and the evolution of biomarkers helped guide treatment decisions in endometrial cancer?

Birrer: [Molecular profiling has] revolutionized our way of looking at patients with endometrial cancer. Historically, the most [information we gathered] was estrogen receptor and/or progesterone receptor status, and pathology grade. Then we had [tumor] stage, and that was one-size fits all. The Cancer Genome Atlas [TCGA], along with several other [efforts], began to stratify the molecular features of endometrial cancer and correlate [them] with prognosis. I presented data from the TCGA [analysis] and showed 4 subsets based on mutation status that correlated very well with overall survival [OS]. Then we validated those subsets in the PORTEC-2 study [NCT00376844], which is primarily a European database.

There's some difference in terminology between us and the Europeans, but [ultimately both analyses identified] up to 4 subsets. [The first is] a small group of [patients with] POLE mutations. [These patients have] a very high level of mutation and do extremely well. In fact, the field is moving [towards considering whether] they even need therapy. The second group, which is a much larger group, is mismatch repair. We've known that [this is a key subgroup] for a while because of Lynch syndrome, which has been characterized for some time. [These patients] also have a lot of mutations, but not as [many] as [those with] POLE mutations. The third group [is patients who have] copy number[–high] differences. This [subtype] is like the serous tumors. [This endometrial carcinoma subtype is] a lot like ovarian cancer and has a fair amount of mutation. There's a fourth group, which is a grab bag, meaning [it has a] non-specific molecular phenotype. We call [this group] copy-number low [in the US]. [Molecular classification is] a way to stratify patients, and it's already worked its way into National Comprehensive Cancer Network [NCCN] guidelines [for endometrial cancers]. A lot of us are certainly [using], if not all of it, most of it, and it's having an impact on how we treat patients and who we put on clinical trials.

Could you discuss how data from KEYNOTE-775 study improved the identification of patient populations that benefit from second-line immunotherapy combinations within endometrial cancer?

By anybody's measure, [KEYNOTE-775] was a home run. [It] showed statistically significant and dramatic differences in progression-free survival [PFS] and OS. [It had been a long] time [since] we had a prospective, randomized phase 3 trial in metastatic endometrial cancer where we saw an OS advantage [vs physician's choice of chemotherapy]. This was a terrific study. [It was] well conducted and [provided] great benefit for our patients. We already know that patients with mismatch repair deficiency [dMMR] are going to benefit from an immuno-oncology [IO] drug. [However], the patient population who really benefits from the KEYNOTE-775 regimen is the mismatch repair–proficient [population (pMMR) who are] microsatellite stable [MSS], and we don't have a lot [of effective therapeutics] for them. Having a combination that provides not just a PFS improvement, but an OS improvement is [therefore] very important.

At the 2023 SGO Annual Meeting, we saw data presented on both the phase 3 RUBY (NCT03981796) and NRG-GY018 (NCT03914612) trials of upfront ICIs plus chemotherapy in recurrent or advanced endometrial cancer. What unmet need did these trials address, and how might these regimens change the standard of care in the first-line setting?

Both [RUBY and NRG-018] are practice-changing, strongly positive studies. I had mixed feelings when I [first] saw the data. On one level, [these data are] wonderful to hear, but on the other [hand] I wasn't surprised. These studies basically moved immune-oncology [IO] therapy from the second line into the first line in combination with chemotherapy. Moving these agents earlier is a natural evolution. The studies included dMMR patients, and we know they're very sensitive to IO agents. In that sense, [the positive result] was not surprising to me at all.

There are caveats to the studies. NRG-018 used pembrolizumab [Keytruda] and RUBY used dostarlimab-gxly [Jemperli]. [These] are 2 different agents, [but] they work roughly the same. It's [also] important to note that for NRG-018, MSS patients were included. The suggestion from the outcome of the study is that even though those patients generally don't respond very well to single-agent [immunotherapy], combining it with chemotherapy [was effective]. Perhaps it releases more neoantigens. We don't really understand it, but even those patients [with MSS endometrial cancer] benefited from the combination. RUBY did the same thing. It included patients with carcinosarcoma and clear cell [histology], which I congratulate the [authors] on. These are histologic subtypes that are frequently left out [of clinical trials], and then we don't know [how to best treat] these patients. It may be that because [these histologies were included], results [from RUBY] are not quite as striking as NRG-018. However, the [results were] statistically significant [although] the confidence interval gets very close to 1. [RUBY also] included MSS patients, and they benefited too. [These trials have] already worked their way into the NCCN [guidelines], and I have colleagues who are already trying to use these [agents] in this setting. It [depends on the] insurance company. It'll be interesting to see how the FDA rules. I think it's a slam dunk for NRG-018 because it's a narrower group, and the numbers look a little stronger. I don't know if RUBY is going to get an all-comers approval. The numbers aren't quite as strong, and they included a lot of these other histologies. I hope they do [grant approval], because the numbers look good, and it would be helpful for us. We should know sooner than later if the FDA is going to approve these agents.

So, what's next? During my presentation, I mentioned that there's a real question [of whether] patients [with] dMMR, microsatellite unstable tumors need chemotherapy. Could they just get the IO agent [instead]? We're going to be answering that [soon]. The other question connected to KEYNOTE-775 is that lenvatinib [Lenvima] and pembrolizumab is going to be brought up earlier into this group [in the phase 3] LEAP-001 trial [NCT03884101]. That should read out in another 6 months to a year.

What is the significance of the withdrawal of treatment indications for niraparib (Zejula) and olaparib (Lynparza) in heavily pretreated patients as discussed by your colleague Heather R. Williams, MD? What might be done to address the need for approved treatments in this population?

There's good news and bad news. The bad news is that these indications have been withdrawn. The good news is these are incredibly small markets, and they were disappearing anyway. What do I mean by that? What was withdrawn was an indication [from the phase 3] SOLO3 [NCT02282020], ARIEL4 [NCT02855944], and QUADRA [NCT02354586] trials. These were heavily pretreated patient populations in the third or fourth line where we could use a PARP inhibitor for the treatment of patients with ovarian cancer. Because of [the phase 3] NOVA [NCT01847274] and ARIEL3 [NCT01968213] trials, [as well as the] ATHENA [NCT03522246], SOLO1 [NCT01844986] and PRIMA [NCT02655016] studies, PARP inhibitors have moved to much earlier [lines] in ovarian cancer. At 5 years and after having received 3 or 4 lines, how many patients are left that never saw a PARP inhibitor? Almost none. These were shrinking indications anyway, and I think that is one of the reasons the companies were inclined to voluntarily pull back [the indications]. I don't think it changes our management very much.

Why did they do it? They did it because the FDA asked them to in discussion, because the OS appeared to be negatively affected by the PARP inhibitor [unlike] PFS. Most of these [trials supporting the indications] looked at either response or PFS, [which] looked good, but the OS looked like it was detrimental to have the PARP inhibitor. You've got to remember that these are incredibly small trials, and they're not powered for OS.

The two [withdrawals] that are more concerning are in the platinum-sensitive recurrent setting, which is [where] NOVA [evaluated] niraparib maintenance and ARIEL3, which tested rucaparib [Rubraca]. Our concern is in the homologous recombination–proficient [HRP] population. I rarely use PARP inhibitors [in that population] anyway, because even though it's statistically significant, [the survival benefit is] about 3 months. That's a lot of money and [other] costs for 3 months. For the homologous recombination–deficient [HRD] population, [however], there's real benefit. There's concern that the FDA might keep marching along and we'll end up with just patients [with BRCA mutations] in the indication. If that's the case, then we've got a serious problem. In my view, you're going to be leaving out a lot of patients in that HRD population, some of whom do benefit from the use of any one of those PARP inhibitors. We're keeping our eye on this.

Shifting to Dr Williams’s second presentation on ADCs, how can ADCs like mirvetuximab soravtansine-gynx (Elahere) address the need for effective yet less-toxic regimens in platinum-resistant ovarian cancer?

ADCs have been around for a long time, probably 30 years, but the technology has completely changed. We now use the appropriate drugs. We used to use standard chemotherapy, [but] for these [agents] to work we need to use powerful drugs, so you get that bystander effect. We're using better antibodies, and our linkers are now proprietary and more efficient. They hold on to the drug when it's in the blood, minimizing toxicity, and then they let it go when it hits its target in the tumor. All this took quite a while to work out, but our colleagues in lymphoma and leukemia have been doing this for years, and [ADCs have also] been in breast cancer for quite a while. Now [they're moving into] solid tumors [like] gynecologic cancers, including cervical cancer. I'm very excited about it. For platinum-resistant ovarian cancer, it's an unmet need. We need another 4 or 5 drugs [in the armamentarium]. Right now, we have basically 3 [options] and they [are not very effective]. [The exception is] weekly taxol, which is not too toxic and has a response rate between 20% and 25%. Gemcitabine and pegylated liposomal doxorubicin are not very effective. We need more drugs, and ADCs fit that [bill].

Mirvetuximab soravtansine is an example of an ADC that has a good response rate [of] over 40% [based on findings from the phase 3 MIRASOL trial [NCT04209855] and a PFS up to [approximately] 6 months. It fits perfectly with what we planned in [the phase 3] FORWARD-1[/GOG 3011 trial (NCT02631876)], which was ultimately a negative trial. MIRASOL was conducted in a slightly different fashion [from a] biomarker standpoint. There's a whole wave of ADCs coming [down the pike] with better technology and drug-antibody ratios [DAR] up to around 10 to 15. These agents [could not only] be effective for [patients with] platinum-resistant [ovarian cancer], but [may] begin to replace taxanes, which [are the cause of many fears regarding] treatment in ovarian cancers. Carboplatin is pretty gentle, but taxol is toxic [and can cause patients to] lose their hair [or] develop neuropathy. Anything that could replace that will be a huge step forward.

Why do ICI approaches remain so largely ineffective in ovarian cancer, according to the presentation given by Martin Cannon, PhD? How could the use of cancer vaccines such as Th17 dendritic cell vaccines potentially improve low response rates?

The short answer is that we don't know why IO has been such a failure in ovarian cancer.The initial argument has been that this is a tumor that has a DNA repair abnormality, and it should [therefore] be the poster child for IO. [However], this was a very naive view of ovarian cancer because the DNA repair abnormality in ovarian cancer is homologous recombination deficiency [HRD]. That means that large pieces of DNA fly around in a tumor cell. These are not single-base substitutions or mutations like you see in lung cancer [that are caused by] smoking. That creates neoantigens, which is very different. The DNA that's flying around in ovarian cancers is [instead] creating large chromosomal abnormalities. If you look at the mutation rate of ovarian cancer, it's in the lower third of all tumors. That's why someone like Martin Cannon is, in my view, [correct about how to approach IO use]. He's not just reflexively saying, "We'll use an IO drug, and [everyone] will be happy." He's thinking about it carefully, and he's combining IO with a vaccine that's carefully and scientifically [designed]. [That strategy] has a much better chance of being successful, and that's just what we need.

Would you like to highlight any trials in the field of gynecologic cancers that are being conducted at the Winthrop P. Rockefeller Cancer Institute?

I run the Phase 1 Cancer Clinical Trials Unit here [at UAMS Winthrop P. Rockefeller Cancer Institute] and am also active in the phase 2 gynecologic oncology [space]. I would like to point out 2 trials which dovetail nicely with what we just talked about and are sort of the next wave [in gynecologic oncology]. [The first] is a phase 1 trial of a first-in-class cyclin E/CDK2 [inhibitor (NCT05252416)]. This is a pathway that's not been targeted before [in this space] because cyclin E is hard to target. Since the complex is a CDK2, and this is a CDK2 inhibitor, it's perfect. The pharmacokinetics and on-target specificity are remarkably good, and it's an oral drug. For all those reasons, we are very excited about it. Why is it relevant? Well, we know cyclin E is amplified and overexpressed in about 20% to 25% of ovarian cancers, and it ascribes a very bad prognosis. These patients have tumors that are resistant to chemotherapy. [Overall], it's the right target and it looks like a good agent. [The trial is] up and running, and we are almost finished with the dose-escalation [portion].

The second trial is the monoclonal antibody upifitamab rilsodotin [XMT-1536; UpRi]. [This] antibody binds to NaPi2b, which is another target that's overexpressed on ovarian cancer cells much like folate receptor alpha. The technology [utilized] here is a polymer on the antibody, which allows [for] up to roughly 12 drug molecules per antibody, so the DAR is very high. This is going to be an effective drug, and it's already in the platinum-resistant queue. We're opening the [phase 3] UP-NEXT study [NCT05329545] in [patients with] platinum-sensitive recurrence. There's a bit of a hold on [the trial] right now because a couple of patients have had some bleeding issues, and we're sorting through that. But I remain very optimistic about this [agent].