Oliver Sartor, MD: There’s a big difference between genotypic and phenotypic biomarkers. If we want to look at genotypic biomarkers, we either have to look at the tumor and get a biopsy of the tumor—put a needle in the tumor—or we have to look at circulating tumor DNA. Although, insights can be provided by germline DNA. It turns out that not everybody has a contemporaneous biopsy available if you’re looking at advanced prostate cancer. I see patients who were diagnosed 8 or 10 years ago with a prostate biopsy in the community eventually end up finding their way into my practice and they have bone-only metastatic disease. Bone-only metastatic disease is hard to define in terms of good tissue that we can use for genetic typing. We could often find excellent markers for circulating tumor DNA. And with BRCA1, BRCA2, and ATM, there are circulating tumor DNA biomarkers approved by the FDA, in particular the Foundation assay that the FDA has approved.

On the other hand, getting tissue isn’t so easy. If we look at large trials, like the PROfound trial [NCT02987543], about 30% of patients had difficulty getting a signal. They’re using next-generation sequencing to define these molecular biomarkers. But 30% of the time, you don’t get good assays back. The molecular biomarkers are really good. I use them, but there are still limitations. On the PSMA [prostate-specific membrane antigen] PET [positron emission tomography] scan, which is now a phenotypic marker, we’re able to get a PET scan in virtually all patients for whom reimbursement can be solved. We still have problems with PSMA PET scan reimbursement. But if we get the PET SCAN, which we can do in a large percentage of patients today, then we can define that tumor in terms of PSMA uptake very readily.

In the VISION trial, which used lutetium-177–PSMA-617, comparison was made between the metastasis and the uptake you see on PSMA relative to liver. If it was higher than liver, then you were eligible for the trial. We also had some exclusionary criteria, which we may cover in a bit. Nevertheless, phenotypic biomarkers with the use of PET scans can be used for the vast majority of patients for whom reimbursement is available. The genotypic biomarkers depend in part on the availability of tissue, which may be limited.

Scott Tagawa, MD, MS, FACP: There are different phenotypes, which is a little different from genotype. There can be a lot of different alterations in the genes, and things can appear the same. In fact, we have many different genes that are different between species, but species tend to look the same when we compare humans with other humans vs humans compared with other animals.

When we look at certain genes, such as BRCA2, if we have a tumor that is altered in BRCA2, that looks more or less the same most of the time. A scan of a patient with a BRCA2 alteration vs a patient without doesn’t look so different. However, there are different phenotypes that go beyond the genes that are there or expressed, and that may have to do with different types of tumors. I happened to mention before when we talked about different types of images, such as PSMA, that with a different type of cancer from the prostate, some people would refer to it as neuroendocrine and some people refer to it as small cell. There’s small cell neuroendocrine. There are aggressive variants. But we haven’t fully identified differences because there are biopsies that would look the same.

Different genes may look a bit different, but there are different phenotypes—such as a low androgen receptor [AR] pathway and low PSA predilection for liver or parenchymal brain metastasis—having to do with the small cell neuroendocrine phenotype. This can be assessed via imaging as well as with biopsy. We can look at the biopsy under the microscope, and there may be differences in terms of histology and imaging characteristics, such as those more likely to have an AR-type of a PET, such as DHT [dihydrotestosterone] or PSMA, or those who may be more avid on FDG [fluorodeoxyglucose], or if we were to look at some metastatin as extrapolating from some of the neuroendocrine and GI [cancers].

Transcript edited for clarity.