In a conversation with CancerNetwork® following the 19th Annual New York GU Cancers Congress®, Rahul Aggarwal, MD, touched upon on the shifting paradigm of neuroendocrine prostate cancer (NEPC) based on a presentation he gave at the conference.1 While de novo NEPC is rare, Aggarwal emphasized that treatment-emergent NEPC (t-NEPC) has become a formidable clinical challenge, appearing in an estimated 10% to 15% of patients as a resistance mechanism to androgen receptor (AR) pathway inhibition. Given that these patients often face the poorest survival outcomes in hormone-resistant settings, identifying and targeting this phenotype is a top priority for the field.
Throughout the discussion, Aggarwal detailed the “clinical gestalt” that necessitates metastatic biopsies, such as the presence of liver metastases or a low PSA despite a high disease burden, and the significance of genomic markers like RB1, TP53, and PTEN loss. He explored the transition toward platinum-based regimens and the “exciting” frontier of novel surface targets, including DLL3, SEZ6, and CD46. From the utility of liquid biopsies and epigenetic profiling to the emergence of T-cell engagers and antibody-drug conjugates (ADCs), Aggarwal delivered a comprehensive look at how precision medicine is finally making inroads into prostate cancer’s most aggressive variants.
Aggarwal is program leader of Genitourinary Oncology and associate director for Clinical Research at the University of California San Francisco Comprehensive Cancer Center.
CancerNetwork: Would you be able to give a brief background of your presentation that you gave at NY GU?
Aggarwal: It was a real pleasure to speak at the NY GU event, and there was a lot of good discussion [between] Maha Hussain, MD, FACP, FASCO and Daniel P. Pretrylak, MD. The organizers had asked me to speak about targeting DLL3 and neuroendocrine prostate cancer. More broadly, neuroendocrine prostate cancer is an area of focus of mine, and the reason is it’s the highest unmet need in prostate cancer; these are the patients that have the worst survival outcomes in the hormone-resistant setting, and therefore, [we’re] trying to develop new treatment strategies.
This is a disease entity that’s quite uncommon at the time of diagnosis, less than 1% of all new prostate cancer diagnoses. Clearly in the treatment-emergent setting, we see the [development] of neuroendocrine prostate cancer commonly. The exact incidence is hard to pin down. We think somewhere in that 10% to 15% range based on the data that’s been published, including from our center. It’s focused on that: “What are the survival outcomes for these patients? What are some of the emerging therapies,” and critically, “How do we identify these patients?” Knowing that [for] a lot of patients, it’s difficult to get a metastatic biopsy, which is currently our gold standard in terms of making the diagnosis.
Given the prevalence of t-NEPC as a resistance mechanism to AR pathway inhibition, what specific clinical characteristics would trigger a re-biopsy in the current treatment landscape?
It’s this clinical gestalt of who we think is [developing t-ENPC]. That being said, NCCN guidelines recommend a metastatic biopsy in the hormone-resistant setting for all patients with an accessible lesion, because we have to remember that there are other clinically actionable things we learn from that biopsy, including repeat genomic testing if it’s been done before or genomic testing for the first time, to look for a BRCA2 or other associated alterations. Microsatellite-high [disease]––we see that in several percent of patients. Then the histopathology, which we use to look at neuroendocrine which is one resistant feature. You can also have patients that may lose AR expression but also don’t have neuroendocrine markers, what we call double-negative phenotype, and we know that prognostically, that’s also an indicator of more aggressive disease.
I would argue that there’s a lot of reasons to think about a metastatic biopsy for patients, which we do commonly in other disease types. Yes, there are some clinical triggers where we’re going to try to push hard to do a biopsy, any patient that has a liver metastasis, I’m going to think about a biopsy, regardless of what the PSA is or other clinical features. If a patient has a known RB1 mutation based on prior genomic testing, that’s going to be a scenario where I’ll make note that, yes, I want to get a biopsy on this patient at some point in their treatment course, ideally of a metastatic lesion, to look for neuroendocrine [disease]. Low PSA in relation to disease burden on imaging, PSMA-PET–negative, [and] soft tissue disease is sometimes a trigger. Not all those patients [have neuroendocrine type disease], but certainly you’re enriching for neuroendocrine detection when you biopsy those PET-negative lesions. These are some of the clues that we might use in terms of guiding biopsy decisions.
Considering the roles of both RB1 and TP53 loss in lineage plasticity, are clinicians close to using these alterations as definitive biomarkers to facilitate a patient switch to a platinum-based regimen?
We have randomized phase 1/2 trial (NCT01505868) data from Ana Aparicio, MD, in the group at MD Anderson, where they did a trial of cabazitaxel (Jevtana) with or without carboplatin.2 This was not only in patients with RB1 or TP53 mutations but was enriched for [these] patients. If you look at the subgroup analyzes from that study, it was those patients that had dual loss of 2 or 3 tumor suppressors, so RB1, TP53, and PTEN. If you had 2 or more genes with alterations or deletions, those are the patients that seem to benefit from the addition of platinum to cabazitaxel in that randomized trial.
Now this was a phase 2 trial, and it was a subset analysis, but, clearly, we know this as a group of patients that has worse outcomes. If clinically it makes sense to add in platinum-based chemotherapy coupled with the genomic features, I use that combination quite frequently in patients that fit those criteria. Whether it’s practice-changing for everyone that treats prostate cancer, it’s not randomized, phase 3 data, but it is compelling.
For aggressive variant disease with AR expression, how is the introduction of platinum chemotherapy vs continued AR targeted therapy considered, and is there a rationale for a combination of both?
The terminology is important, because when we say aggressive variant or aggressive phenotype, really, we are talking about MD Anderson-generated criteria that can either be clinical, which was goes back to the original paper with carboplatin and docetaxel, where there was a set of clinical criteria, including visceral metastases, presence of lytic bone metastases, et cetera.3 That’s the clinical definition of aggressive variant.
More recently, there’s a molecular definition of aggressive variant based on RB1, TP53, and PTEN alterations. again, 2 or more loss in those patients that have that treatment, emergent, aggressive variant phenotype. Yes, we often do think about platinum based chemotherapy, but first we know that some of these patients, of course, still benefit from AR targeted therapy, and it may still be an AR driven disease, at least in the subset of the prostate cancer within that patient, knowing that there’s a lot of intra and inter tumoral heterogeneity. Even as we’re thinking about starting a patient on platinum based chemotherapy, typically in the treatment emergent setting, we’re also still continuing that patient on androgen deprivation therapy. We don’t want to let the AR responsive portions of that patient’s disease start to accelerate, and so we often will still maintain antigen blockade in that context.
What emerging surface targets do you find to be the most promising for ADCs or T-cell engagers in this neuroendocrine subset?
This is where it’s been exciting, to think about the increased interest amongst drug developers, pharmaceutical companies, and academic institutions in this disease entity in a way that just wasn’t present 5 or 6 years ago. There’s a lot of exciting both targets and different therapeutic modalities you covered. ADCs and T-cell engagers––both are exciting classes of therapies. I would add, potentially radioligand therapy. On top of that, in terms of this neuroendocrine prostate space, DLL3 is a target where we already have prospective clinical data that there are some patients that can derive meaningful benefit from DLL3-targeted treatment.
In neuroendocrine prostate cancer, it comes down to the patient selection with DLL3. It can be heterogeneous in terms of the DLL3expression. This is not like small cell lung cancer, where that biomarker is a lot more homogenous and a high rate of expression in most, if not all, patients. If we can use either biopsies looking at DLL3 expression, or potentially molecular imaging, these are ways to really enrich patients for DLL3-targeted treatments.
A couple of other targets to keep an eye on [include SEZ6]. SEZ6 is a target that is looking promising. We’ll [likely] hear more about it in the context of ADCs and potentially other treatment modalities. That’s one to keep an eye on. Then there are others that are traditionally thought of as prostate adenocarcinoma targets but might have less lineage dependence and might still be expressed in the neuroendocrine setting. We’ve been working on a target against a protein, CD46 here at my institution that seems to be expressed in both adenocarcinoma and in neuroendocrine [disease]. We hopefully will hear more about that. B7H3 the expression is a bit lower in neuroendocrine [disease], but still expressed. There may be opportunities to think about B7H3-targeting in neuroendocrine prostate [cancer], the way that’s being considered in small cell lung cancer as well. These are a couple of the targets that are emerging and hopefully [there’s] more to come.
How will the use of tissue biopsy evolve with circulating tumor cell and cell-free DNA liquid biopsies emerging to help identify the transition to neuroendocrine phenotype?
That’s where the field is going. We’re still probably not quite yet there terms of, do we have a clinical grade-ready assay that can detect neuroendocrine [disease] and use it for the purposes of enrollment, like an integral biomarker in a clinical trial. I don’t think we’re quite there yet with either the CTCs or cell-free DNA. We’re moving in that direction, in particular the methylation profiling from circulating tumor DNA seems to do an excellent job of segregating the adenocarcinoma from the [patients with neuroendocrine disease], because we know that epigenetic dysregulation is really a biologic hallmark of this disease.
A lot of work [is] going in that setting. Molecular imaging is overlaid on top as well, as we get lineage-specific tracers like PSMA-PET and others that are coming may help us to identify patients as well, or at least enrich for those that we should need to be thinking about the possibility of neuroendocrine differentiation.
Is there anything else you would like to highlight that we might not have touched upon?
I want to reiterate the excitement in this area, knowing that it’s a huge unmet need, and we’re really trying to encourage more interest and focus in this area, and we can make a meaningful clinical impact for our patients as we get some of these therapies into the clinic.
ReferencesAggarwal R. Neuroendocrine prostate cancer: spectrum, recognition, and management implications. Presented at: 19th Annual New York GU Cancers Congress®; March 13-14, 2026; New York, NY.Corn PG, Heath EI, Zurita A, et al. Cabazitaxel plus carboplatin for the treatment of men with metastatic castration-resistant prostate cancers: a randomised, open-label, phase 1–2 trial. Lancet Oncol. 2019;20(10):1432-1443. doi:10.1016/S1470-2045(19)30408-5Aparicio, AM, Harzstark AL, Corn PG, et al. Platinum-based chemotherapy for variant castrate-resistant prostate cancer. Clin Cancer Res. 2013;19(13):3621-3630. doi:10.1158/1078-0432.CCR-12-3791