The Neuroendocrine Prostate Cancer Spectrum Diagnosis and Treatment Presentation - Eleni Efstathiou
September 17, 2019
Eleni Efstathiou, MD, Ph.D., Associate Professor, Department of Genitourinary Medical Oncology, Division of Cancer Medicine, The University of Texas MD Anderson Cancer Center, Houston, Texas.
Written Coverage: APCCC 2019: Neuroendocrine Prostate Cancer Spectrum - Diagnosis and Treatment
Download Presentation Slides
Eleni Efstathiou: Thank you very much for allowing me to present data and discuss something I'm not an expert on. Thank you. I come as a clinician and a physician, a person who can try to understand looking from the outside as all of us in this room how this field can get better with the help of all of us. And as I am listening very clearly, I am presenting in lieu of Ana. Ana Aparicio has done a lot of this work in our institution and other people in this room have been really fundamental in trying to do it.
These are my disclosures. I think as we sit as clinicians, we are all bothered and concerned with the lack of a terminology that will allow us to better define and understand just by using a word what is this aggressive variant or neuroendocrine prostate cancer spectrum. We spent a lot of time during the last APCCC, and I will end with these slides trying to understand if we can reach a consensus and how to identify it and treat it. I do not think that we actually reached a consensus at any time and nor would we if we included these questions tomorrow, which we will not. Probably will be best to leave it for two years later.
This is a very heterogeneous group. These are tumors that during the course of androgen deprivation appear to become less dependent on androgen signaling and have invariably upward prognosis. This is quite loose, I think we would all agree. And a White Paper was produced a few years back by a group of experts in the field who tried to address these inaccuracies in terminology and come up with a better term, which they could not.
The word neuroendocrine I believe was used mainly because it was reflective of a poor clinical course that was reminiscent of the small cell variant, a very confusing term since neuroendocrine morphology features and markers are not warranted for a tumor to belong in these categories. In fact, it can be even more confusing if we take into consideration that there are some new endocrine pathology features that are not at all associated with aggressiveness, such as all of us who spend time with pathologists have learned paneth cell-like differentiation.
How about aggressive variants of prostate cancer? It is less confusing, but potentially more contaminated with other forms of an aggressive form of prostate cancer. Therapy-related neuroendocrine or small cell prostate cancer. It seems that it's starting to be used more and more, but the concern back then was that if you actually use the term therapy-related, you're introducing a bias and hence clinicians may not opt to go for what has been found to be very positive treatment options.
Androgen-indifferent prostate cancers. Well, it has been shown that some, a few, of these tumors may still respond to novel androgen signaling inhibition, and again, bias should not always be introduced. AR-negative prostate cancer, too limited, as there is mixed expression. Anaplastic prostate cancer. We at MD Anderson, of course, love this term. It went with Ana's name, I believe, but the term was used a lot before in pathology to denote pleomorphic cytology, so it didn't really make sense, right? So we're stuck.
But aggressive variant prostate cancer or NEPC, if you want to call it neuroendocrine, appears to have an increased incidence, and it's been reported now in the more recent papers that we will very briefly discuss, up to 20% after treatments with novel agents potentially. Is it greater awareness? Is it that patients are living longer, or is it indeed a resistance to novel therapies that are evolving? And it is likely to be a combination of the three.
And there has been a more recent NCI workshop on lineage plasticity and androgen receptors that brought together a lot of the groups that were working on it and identified clearly those unmet needs. Understanding how this lineage plasticity occurs, determining, and that is crucial I believe, the temporal contribution and cooperation of the emerging drivers, identifying those preclinical models that we capitulate to the biology with the known limitations and recognized phenotypes of the disease. And finally, all of this will lead to identifying therapy targets and novel trial designs that will be dedicated to the entity as it is defined.
Let me take a little bit of a step through the history of what we have been doing and then try to tie it in with so much work that is out there that there's simply no time. First, there was morphology. Back in days that a lot of us in this room were not involved and others were, there was a lot of interest in that small subset of small cell prostate cancer and its aggressive course and clinical features, and there are some old studies addressing whether there would be responsiveness with the use of platinum, and there was, of course with the disease to come back roaring.
But then comes clinical presentation, and it was identified that there were these aggressive variants that looked different potentially morphologically but had this very similar clinical course to small cell prostate cancer. And of course, you'd just like to check at that point, as a lot did, whether platinum would work for these aggressive clinically cases.
How are these identified? Well, beyond small cell prostate cancer, what we published and others have looked into is that the clinical characteristics might include visceral meds only, liver metastases, lytic bone mets dominantly, bulky nodes or a very bulky prostate mass, low PSA in under 10 and under 4 relative to the volume disease, neuroendocrine markers, serum CEA that is elevated or very high LDH, and primary castration resistance, and there's a little bit of variability across the field. Others say less than 6 months, others say less than 12 months. Again, it's a little bit loose there.
The hypothesis is do shared clinical features, all small cell prostate carcinoma that we can find in what we would identify clinically as an aggressive variant, predict for shared platinum-based chemotherapy combination sensitivity? That was the next trial that was presented that actually looked and confirmed that indeed there is some sensitivity of these variants to platinum and there are of course other trials that looked into it, as well.
And this brought us to the most recent trial that is now impressed at the Lancet oncology and I present some of the data of an operation Paul Corn looking at treating metastatic castrate-resistant prostate cancer with cabazitaxel plus-minus carboplatin with very nice pre-specified stratification based on clinical criteria. The trial was overall positive for our PFS, but if you look at the fourth plot on the bottom, it seems that a lot of it is being driven by those tumors that appear to have clinically the features of the aggressive variant, which is very interesting.
And of course, the next step is to use the wealth of data that has been produced by many investigators to look for potential molecular markers that can be predictive even more accurately, let's say than clinical characteristics. Ana and Paul decided to look at P53 mutation, RB loss, and P10 loss, and they assessed in just a subset of these patients, the presence, and then looked for two out of three of these features and the association with response. And in this limited cohort, and of course not a prospective validation or retrospective, there seems to be a signal for the signature.
What about the rest of the cohorts? There are plenty out there with many, many patients. This is from a year ago from I believe Kim Chi's group and again given to me by Ana, where one can see that you can identify in patients who were treated with abiraterone enzalutamide looking this time at circulating tumor DNA, and I also have some data that I could not present as it's not yet public, on the same type of liquid assay that would suggest that you can identify such patients by liquid biopsies. And in fact, if you look carefully, it looks like those patients who appear to be more resistant to abiraterone enzalutamide may be enriched for such a signature.
If we looked at the Stand Up 2 Cancer efforts, we can see from Eric Smalls' group that there are also a lot of efforts to try to not only clinically and morphologically, but also genomically characterize these small cell-like features. And we can see from a very, very recent paper from Wassim Abida that these types of cancer appear to be enriched for RB loss. You don't see as much the combination with the P53 showing up, but essentially I would say that there is some level of concordance. We are expecting in the future that all of these groups should work more closely together to give us some answers.
For the current state from a practical perspective, clinical criteria appear to help identify the aggressive variant. The caveat, there's a contamination by other molecular subtypes. When it comes to morphology, things like that, one of the Stand Up 2 Cancer teams have focused on, I believe that now there are several current guidelines that recommend sampling metastases, and this is a step forward. We have been all in rooms 10 years ago saying it was not feasible. Now it's in the guidelines. Molecular sub-timing, it's too early. It requires validation. It's not ready for primetime.
Treatment. While treatment of a clinical aggressive variant, consideration of platinum-based combinatorial chemotherapy data appears to be valid, but the evidence remains weak. We do not have even overall survival data. Point for non-purists, though, if there is a contamination based on the clinical criteria, let's say with DDR-driven tumors, there is no practical concern. You're still going to use platinum.
This was our consensus table back in 2017. We did not really reach a consensus. Maybe we were close to one for clinical considerations, and we definitely did not reach one for the use of standard of care versus platinum-based, and I think that even if we use the cards ... We're too close to lunch to do that ... We would probably see the same thing.
Our expectations from the next APCCC data set. We need more precise molecular characterization to help identify subtypes to move away from lumping everyone together. This will enable therapy development, and if the workshop pans out, we might be able to predict this transformation early on and spare some patients from having it happen. Hint, I think a lot of us mentioned, and William was one of the speakers who mentioned that there are a lot of non-PSA progressions within not only the MCRPC studies, but we're seeing them also in the SPA studies. I love this new term. Finally, I think we will have more data on how these patients should be followed. Remember, PSMA imaging is more of a neurosis, and especially when it comes to these patients, it might be a concern.
I would like to end with something that I think we should be amiss to not mention. A week ago, we lost a man who had prostate cancer for 27 years. His contribution to a lot of what we are discussing in this room cannot be forgotten, and regardless of political beliefs or what have you, this was a great philanthropist with a wonderful legacy left for us. Thank you very much.