Read the Full Video Transcript

Alicia Morgans: Hi, I am so excited to be here today with Dr. Himisha Beltran, who is joining me after a wonderful talk at ESMO 2024, where she talked to us about DLL3 and use of it as a target for neuroendocrine prostate cancer. Thank you so much for being here with me today, Dr. Beltran.

Himisha Beltran: Thanks so much for having me, Alicia.

Alicia Morgans: Wonderful. Well, let's have you go through your talk and then I'll ask some questions at the end.

Himisha Beltran: So, as we all know, targeting cell surface proteins has arisen as an exciting area of drug development, not just in prostate cancer, but also in other cancer types. And DLL3, or delta-like ligand 3, is one emerging therapeutic target in prostate cancer.

DLL3 is not expressed in normal healthy adult tissues, making it an attractive drug target. And we found in prostate cancer, it's not expressed in benign prostate or localized prostate cancer. But as tumors progress to castration-resistant prostate cancer, a subset, 12% of those with adeno histology, will acquire DLL3 expression, albeit very focal. But when tumors transform from an adeno histology to a neuroendocrine or small cell carcinoma, this target becomes highly expressed, as shown here in 76% of cases in this series.

So, we know that neuroendocrine or small cell carcinoma of the prostate rarely arises de novo, but can arise in later stages of disease progression as a mechanism of resistance. This process of lineage plasticity, where tumors shut down their adeno luminal lineage and turn on this neuroendocrine lineage, is an important biologic process. But it also has therapeutic implications, as certain targets can be down-regulated like the androgen receptor as well as PSMA. But we can also leverage or exploit this process therapeutically when thinking about acquired targets like DLL3.

So currently, the diagnosis of neuroendocrine prostate cancer is based on a tumor biopsy. And that biopsy typically shows features of high-grade, poorly differentiated neuroendocrine carcinoma that can look a lot like other small cell carcinomas like small cell lung cancer and others, and also share a lot of the same molecular features as well as therapeutic targets like DLL3. So DLL3 is not only expressed in prostate cancer, these neuroendocrine prostate cancers, but is also expressed in small cell lung cancer, Merkel cell carcinoma, small cell bladder cancer, other poorly differentiated high-grade neuroendocrine carcinomas in the GI tract, and others.

And excitingly, the DLL3-targeted T-cell engager, Tarlatamab, was recently FDA-approved for patients with previously treated small cell lung cancer. So, this is a drug that binds DLL3 on tumor cells and also binds CD3 on T-cells, redirecting T-cells to the tumor microenvironment to elicit an immune response. And this drug was approved based on data presented at ESMO last year, and subsequently published in the New England Journal, from a phase 2 trial in pre-treated small cell lung cancer. So, a patient population that has very limited effective therapeutic options; it's a very aggressive disease. And the response rates in this study were 40%, with many durable responses and promising overall survival rates that led to the FDA approval.

And so, this leads to the important question of whether we should be using this for our neuroendocrine prostate cancer patients, given that this target is expressed. And we often extrapolate from small cell lung cancer when we think about treatment. If you look at the NCCN Guidelines, they highlight that patients who are diagnosed with neuroendocrine prostate cancer, again based on biopsy showing histologic features, could be considered for platinum-based chemotherapy. But beyond platinum, the guidelines refer to small cell lung cancer guidelines, where we can use agents such as lurbinectedin and other second-line chemotherapies. And now that Tarlatamab is approved and will be making it into those guidelines, how should we be giving this to our patients?

There was some data presented at ASCO earlier this year by Dr. Rahul Aggarwal on Tarlatamab in prostate cancer. However, the inclusion criteria were quite broad. It included not only small cell neuroendocrine prostate cancer, but also those that had neuroendocrine differentiation or greater than two tumor suppressor losses in the absence of neuroendocrine histology.

And the data were less impressive, I would say, than the small cell lung cancer data, with overall objective response rate of 10.5% in the overall population, median duration of treatment of 1.4 months. Although there was this one patient that they highlighted that had a long-term response of 25.8 months.

So, how do we resolve these discrepancies between small cell lung cancer and prostate? I think it had a lot to do with patient selection. If you look at that study, there was, as I mentioned, a broader inclusion criteria based on PSA range from zero to 5,000, suggesting that there may have been some AR-driven tumors within that spectrum. And of the evaluable patients, only 56% expressed DLL3. None of those tumors had adeno genomic features. And DLL3 expression was defined as 1% DLL3 expression. So it may be, looking more closely at this study, potentially the patients that were more likely to respond may not have been well represented in this study.

Merck has another DLL3-targeted T-cell engager called MK-6070, formerly called HPN328, which is also again a T-cell engager. But instead of being a bispecific that binds DLL3 and CD3, it's a trispecific that also binds albumin with a goal to extend half-life and minimize nonspecific T-cell activation.

So, this is an ongoing trial. There are multiple dosing cohorts and now combination cohorts, but we did present interim results of this ongoing trial at ASCO earlier this year. This trial is enrolling small cell lung cancer, neuroendocrine prostate cancer, and other high-grade neuroendocrine carcinomas with DLL3 expression. So, it's a phase one study.

But when you look at the patients in this interim analysis that have been treated, there were responses across the spectrum of neuroendocrine carcinomas in small cell lung cancer. The extracranial responses were approaching 50%. And you can see neuroendocrine prostate cancer—not a lot of patients included in this analysis. I think we do need more data, but most patients did have some degree of tumor shrinkage.

And when we look at the swimmer's plots in this trial, there were some patients with neuroendocrine prostate cancer that I was particularly intrigued by, as they were treated beyond radiographic progression due to clinical benefit. And one can imagine that if you have a very aggressive neuroendocrine clone and that may be responding, that progressing lesions... Well, we know we often see mixed features in prostate cancer, may be due to this heterogeneous disease that we see. And points to the need for potentially better ways to diagnose and treat tumor heterogeneity.

I'm excited by work by Jason Lewis and Michael Morris as part of a PCF Tactical Team to develop DLL3 PET imaging, and potentially as a way to complement PSMA PET and F-18 FDG PET in diagnosing neuroendocrine prostate cancer. And I think it may also have implications when thinking about targeting DLL3 and understanding resistance patterns.

We have also been working preclinically with DLL3-targeted T-cell engagers, the MK-6070 drug, where we see potent antitumor activity in PDX models when we treat mice with this drug and human T-cells. We see T-cell infiltration. And we've been developing relapsed and now acquired resistance models, and it's really important to try to understand how tumors resist these therapies. And is it related to DLL3 or other mechanisms where we can also see in patients where responding and progressing lesions may have different levels of DLL3?

This tumor heterogeneity in prostate cancer also points to a potential for co-targeting. As we know, there are a number of emerging therapeutic drug cell surface targets in prostate cancer. And when you look across the spectrum of castration-resistant prostate cancer, there tends to be incomplete overlap of the expression of these targets. And so potentially, we should be thinking more rationally about combining drugs or even bispecific therapies that target more than one antigen.

So just to conclude, DLL3 has emerged as a therapeutic target for small cell neuroendocrine carcinomas, a very aggressive subset of cancers, which includes neuroendocrine prostate cancer. DLL3-targeted T-cell engagers have demonstrated activity in neuroendocrine carcinomas. And there are other DLL3-targeted drugs also in development.

And for prostate cancer, I think patient selection will be important. DLL3 is not expressed in most adenocarcinomas, and therefore heterogeneous tumor features might lead to mixed responses. Biomarkers that capture heterogeneity may help improve patient selection in the future and inform rational combination strategies. And I think understanding resistance patterns may also inform next-line therapy, whether it be another DLL3-targeted agent that's in development or other approaches or combinations. Thank you. Thanks very much.

Alicia Morgans: Great, Himisha. That was fantastic. And this is really such an exciting area of research because our patients who have neuroendocrine differentiation or small cell differentiation have terrible clinical courses in some cases and really very few options to control the disease. As you think about things that are already available, trials that are in process now, it's possible that some patients may actually even have access to things like Tarlatamab right now for neuroendocrine differentiated prostate cancer or small cell prostate cancer. Is this something that you have seen used clinically? And as I understand, it's delivered in the hospital. How is that process going for our patients with prostate cancer?

Himisha Beltran: Yeah, so I don't have experience of using Tarlatamab off-label at the moment. I know this is just recently approved for small cell lung cancer, so it'll be quite a process, as you mentioned, that it requires some coordination of care even for those patients and to try to think about how we could get this for prostate cancer. Obviously, I'm very excited about the ongoing trials in prostate cancer and other cancers to establish T-cell engagers in that space.

I think that with T-cell engagers, you do need... the side effect profiles are related to immune-related side effects, cytokine release syndrome, neurologic symptoms with ICANS. And so, patients do require inpatient hospitalization for the first priming dose and the target dose before being transitioned to outpatient. And I think while most of the CRS events are mild and reversible and occur within the first 24 hours, it's something that we're not yet comfortable with doing as an outpatient. And so, that'll be something I think that the field has to think about more carefully as more and more of these drugs enter the clinic across cancer types.

Alicia Morgans: Absolutely. And to your point about the disease heterogeneity, it's not uncommon to have adenocarcinoma mixed in with neuroendocrine or small cell as you mentioned. Is this a situation where we would both try to treat the adenocarcinoma maximally and also then give maybe a DLL3-targeted agent? How do you think through trying to control the disease overall, knowing that there are these different components?

Himisha Beltran: Yeah, I think when I... Because there's such a spectrum of the disease, I definitely start... when I meet a patient, think about what are the clinical features? And is it a pure small cell? Is it a mixed histology? Is the PSA going up? What's the PSMA PET scan look like? Try to get an idea, and the tumor biopsy, of course, to try to get an idea where on that spectrum of disease it might be. Because we've all seen those patients that have kind of pure small cell carcinoma that look and act a lot like small cell lung cancer. And then those other patients that have maybe an aggressive clinical course, but not the same kind of explosive disease as we might see in small cell carcinomas.

And so I think teasing that out will be important, especially when thinking about single-agent therapy target that's only expressed in neuroendocrine carcinoma. But I hope that the future will be in co-targeting, as we know that we want to intervene early. We want to be able to target mixed populations in patients that are developing lineage plasticity. And we want to be able to have more robust antitumor activity. And so if you target only one component, we may not be seeing the maximum tumor responses that we really want to see.

Alicia Morgans: That makes sense. Well, as you have a final message then for our listeners about DLL3 and the possibilities that exist for neuroendocrine and small cell prostate cancer, what would your message be?

Himisha Beltran: Yeah, I would say that this is a challenging disease to diagnose and treat. And so to refer for clinical trials and start thinking about making sure that we try to learn as much as we can from the patients that we have as well in the context of research studies, so that we can be better at understanding and diagnosing heterogeneity and treating it in the future.

Alicia Morgans: Great. Well, thank you so much for your time and your expertise. We always appreciate hearing from you.

Himisha Beltran: Thanks for having me, Alicia.