How Pain Can Serve as a Useful Biomarker for Prostate Cancer

An interview with Andrew J. Armstrong, MD ScM FACP, Professor of Medicine, Surgery, Pharmacology and Cancer Biology, Director of Research, the Duke Cancer Institute Center for Prostate and Urologic Cancers Divisions of Medical Oncology and Urology, Duke University.

Interviewed by Daniel Tennenbaum, MD, chief resident at the Maimonides Medical Center urology residency program in Brooklyn, New York.

This is the second part of a three-part conversation with Dr. Armstrong. Watch part three here.

Dr. Tennenbaum: In regards to the patients that you were discussing who light up brighter or more dim with PSMA PET scans, is there any effort to further subtype these patients where they’re carrying multiple phenotypic biomarkers? What’s the next step for patients like that?

Dr. Armstrong: Yeah, so at ASCO this past year, we looked at the VISION Study and presented data on the standardized uptake value the SUV, and we looked at different patient characteristics, the SUV max, which would be the brightest lesion that you have, or the SUV mean across the whole body where it’s more of a representation of the whole body intensity. So for example, a patient could have a very high SUV max, but might have lots of lesions that are dim, and that would be very different than a whole body SUV mean where the whole patient is bright.

And we showed that the SUV mean at a whole body level is much more important in determining progression-free and overall survival with this PSMA lutetium targeting. Also, pattern of spread’s important. So encompassing another phenotype, which is pattern of spread into that SUV can give you kind of a two-way reading. So patients with liver metastases tend to do more poorly than patients with lymph node only metastases even across the same range of brightness scales. So there you start to incorporate a little bit of biology, seed and soil and metastasis biology along with the brightness of the tracer.

Dr. Tennenbaum: Sure. You had mentioned a few of the phenotypic biomarkers in your introduction, and one of them that you’ve mentioned a few times is that of pain, cancer-related pain. I find that to be somewhat interesting because it’s perhaps the only subjective biomarker that I could think of. Why do you think it is that pain serves as a useful biomarker? How does it help predict the patient’s disease course?

Dr. Armstrong: Well, it’s a great question. I mean, you and I have probably both seen many patients where they have widespread disease but zero symptoms and then the opposite patient where they have very minimal disease, but a lot of pain. And so there are factors secreted by the tumor in the tumor microenvironment space that regulate symptoms. And the consequences of those symptoms can be very profound on patient quality of life. Some of the trials that have been done and completed successfully required either absence of those symptoms or presence of those symptoms in the context of a phase three study, for example, PROVENGE was absence of symptoms and Radium required symptoms. Moving those treatments into new contexts would require new studies. And so when the FDA approves the drug, it’s going to be approved based on the context in which it was studied and not beyond that context.

So for example, we don’t often give PROVENGE to patients with significant symptoms like pain requiring opiates because it hasn’t been studied in that context. And likewise, Radium is not typically given in very, very early disease before symptoms develop. And honestly, the studies that did that showed a higher fracture rate and a black box warning was added to Radium because of that. And so I think we have to be followers of evidence-based medicine, and symptoms can be quantifiable, they can be objective. You can have patient reported outcomes using validated metrics and surveys that either capture the pain scale or the amount of opiates a patient’s receiving like an analgesic scale.

Dr. Tennenbaum: Understood. Thank you. You had mentioned a few minutes ago about using radioligand therapy founded on Lutetium and PSMA. Is it fair to say that the field of theranostics falls under the umbrella of phenotypic precision medicine?

Dr. Armstrong: Absolutely. And it’s moving more so. Certainly the VISION Study was that proof of principle that you can improve overall survival in a very late stage setting, post docetaxel, post AR therapy, and that’s where the FDA approval is right now. But ongoing and completed studies are now testing this in pre-chemotherapy or chemo naïve castrate resistant settings, hormone sensitive metastatic settings, but all using that PSMA PET scan. It’s possible that as we move into even earlier settings, microscopic settings that you might not need a PET scan, but that hasn’t been proven yet. And it’s also possible that as we move into combination approaches, the PET scans might not be needed, particularly if there’s other biomarkers that could determine the presence of PSMA.

I think PSMA is pretty ubiquitous in prostate cancer. About 85% of patients have PSMA detection. But even within that group, there’s a wide range of expression, like I mentioned on the PET scan. And we see the same thing with circulating tumor cells. We see some men that have all of their tumor cells that are circulating in their blood be bright on PSMA and some that there’s heterogeneity, meaning that if you targeted those tumor cells, likely you would select for the cells that lack it and that would clonally evolve. And right now we’re working on newer radioligand therapies that target other molecules besides PSMA and PET scans are being developed as companion diagnostics for those therapies like DLL3 or CD46 or STIP1, PSCA. So there’s a lot of excitement in this field of precision medicine based on the VISION Study and what that means.