Sartor Discusses the Evolving Role of Biomarkers in Prostate Cancer

Oliver Sartor, MD

There have been significant strides in identifying biomarkers to appropriately treat patients with prostate cancer, according to Oliver Sartor, MD. Although defining mutations and subgrouping patients may take longer than anticipated, he adds, the results could be vastly beneficial.

OncLive: How has the role of biomarkers evolved in prostate cancer?

In an interview with OncLive, Sartor, the medical director of Tulane Cancer Center, discusses current research with biomarkers in prostate cancer, as well as how some treatments may be utilized as both single agents and in combination.Sartor: There has been tremendous evolution in the biomarker space, really over the last year and a half. We can begin to look back and ask questions about what we can do in the way of molecular testing that can allow us to predict how therapies might work or not work.

If we look back at what I thought was important initial information was the discovery of ARV7, which was detected originally by the Johns Hopkins Medicine groups as well as some others. They brought it into the clinic to be able to show that ARV7 would be predictive of resistance to either abiraterone (Zytiga) or enzalutamide (Xtandi) and that helped to move things forward. They were able to do that with circulating tumor cells, and that was an important step because you didn't have to put a needle in the tumor.

Then we look at another important advance by the Stand Up To Cancer group and also funded by the Prostate Cancer Foundation. What they did was determine how to put needles directly into tumors and to define a whole series of somatic alterations and mutations that were clinically relevant. It turns out there's a high number of individuals, probably about 25%, who have defects in their DNA repair pathways. Based on a New England Journal of Medicine article that was published last October, a PARP inhibitor might be able to target that series of individuals who had DNA repair mutations.

More recently, we have begun to show as well as others, that carboplatin may be a particularly good agent within individuals with BRCA2 mutations. There's some unpublished data that we're putting together that would indicate the same. So, we've been moving to try to define DNA repair defects and things that might be able to elicit responses in those patients, and right now preliminarily, the platinums and the potential radiation, which I didn't mention, and PARP inhibitors are areas that need further exploration.

In the real world, how much testing is being done to identify these sorts of mutations?

Are there studies that are going to move that forward?

In addition, we've been able to define increasingly the amplification of AR and the utilization and detection of mutations within an androgen receptor that confer resistance. As we begin to look at the field, we have DNA repair defects, we have androgen receptor amplification, we have mutations, and we have variants of androgen receptors; each of these are looking like they might predictive biomarkers. More work is necessary, but it's a rapidly evolving field and one that is very exciting right now.The current commercial utility is somewhat limited, because ARV7 tests are limited to laboratories. At Johns Hopkins, there's a clinical trial where they're embedding this test into their trial design, but it's a little bit difficult to simply get the tests done. It is the same with the AR amplification and mutations. Now, there are newer companies that are analyzing circulating free DNA and can give you insights into things like DNA repair defect and mutations, BRAF mutations, alterations in the androgen receptor, and we're incorporating it into our practice today, but it's still relatively exploratory.Yes, there are multiple studies that are going to move that forward. There are plans for a study with olaparib (Lynparza) which is a PARP inhibitor. That study is going to be looking prospectively in a multi-institutional manner for those with DNA repair defects. That trial is under design right now. There's also been several companies well known in the prostate cancer field that require PARP inhibitors, and they are going to be developing their own strategies.

Is it challenging when you start to break down prostate cancer into smaller groups to do clinical trials to find patients that fit into these smaller boxes?

With the therapies that are currently available, is there an ideal way to sequence?

In addition, there is data with carboplatin, a miniscule amount of data published, but a group I've been cooperating with is now putting together a much larger series that basically demonstrates that carboplatin is active in those with DNA repair defects. That is particularly true for the BRCA2 gene. There is a multiplicity of steps that will be moving this forward.The answer is yes and no. As the rewards for finding these types of patients goes up, the intensity of the search goes up, but yes, we are segmenting the disease that used to be somewhat monolithic, and now we are creating a series of molecular subsets, some of which are actionable and some of which are not. We really have a lot more to learn, and I think much of the next 5 years will consist of deciphering how we take these molecular changes and translate them into clinical action.When we begin to think to think about sequencing and the combination of agents that are used in advanced prostate cancer—we currently have 6 different agents, and probably 4 different classes of agents—how do we choose these drugs wisely? How do we bring them together in sequence? Is there an optimal sequence? How do we address the issue of combination therapy and is it superior to sequential therapy or not? These are all dominating questions in our field.

There are a series of clinical investigations, some biomarker driven and some not, that begin to address these questions. For instance, one of the questions might be about the utility of novel hormones in combination with radium-223, and there are clinical trials to look at abiraterone with or without radium-223 in patients who might have a bone-dominant disease. The question is, "Can the combination out-perform the individual drug that is used today, either abiraterone or enzalutamide?"

Another example of a combination trial is enzalutamide with or without abiraterone, run by the Alliance group, and they're asking if combination AR-targeted therapy can be superior to using one alone, and in this case it's enzalutamide. As we begin to evolve, there are whole series of tests that need to be performed in the way of clinical trials. Some sequencing questions may be unanswerable because there are so many sequences available, but some combination trials can be focused upon and I think we'll have answers in the next 3 to 5 years.