Experts on advanced ovarian cancer give detailed insights on the role of biomarker testing and identify important biomarkers.
Robert L. Coleman, MD, FACOG, FACS: This is a good point where we can bring up potential risk factors that you alluded to a bit. What about genetic risk? What do we think about that?
Bradley J. Monk, MD, FACOG, FACS: In the past, we used to say, let‘s take a family history so we can predict who’s at familial risk for epithelial ovarian cancer, and obviously we can’t. Certainly younger patients are more likely to have a germline mutation, but because we can‘t predict, all organizations recommend that we test every patient for germline mutations. Now, these are inherited mutations, and again, it used to be just BRCA, now it‘s a panel of 47, 48 genes, and we can identify maybe as high as 20% of patients who [have mutations that] run in families. Age of onset and family history is not always informative, so we test everyone. I think the world understands germline testing for all, but we don‘t tumor test for all. Please tell us when, why, and what we should do for the tumor test, and how it‘s complementary of germline.
Robert L. Coleman, MD, FACOG, FACS: That’s really important now as we think about the types, what we do with the information. One of the things you mentioned is that we were interested in learning about BRCA status in families because we would be able to provide counseling. A lot of that was that there were some potential benefits of that information to the patient because it was a prognostic factor. In other words, patients who had that mutation tended to have higher response rates to chemotherapy and do a little better. Now we’re using it in a predictive way, so that we are aligning the aberration or the finding in the germline, in which we would assume would be in the tumor, and then in the tumor alone if it doesn‘t occur in the germline, to provide information with respect to the drugs that we want to use. In this case, we’re focused on PARP inhibitors because they’re aligned with efficacy in those conditions. The tumor testing, depending on what’s available in certain situations, we‘re fortunate in the United States that we can usually get these done at the same time, send blood and tumor tissue, so we can sort out what is a germline event and what’s going on in the tumor itself, so we can then use that information to help design therapy.
There are other biomarkers that we attain in our patients. Many people know about CA-125 [cancer antigen 125], this is a circulating antigen that originally was a target for immunotherapy many decades ago. We learned that it was not a great target for that because the extracellular component of that molecule was fragile and broke off, and was floating. So we were able to capture that as a tumor-associated antigen. CA-125 is another biomarker that’s frequently monitored during therapy; as of yet we haven‘t been able to use it as a predictive biomarker for treatment. Certainly, there’s that option, that as we learn more about ovarian cancer tumor microenvironment, there may be other biomarkers we may be able to leverage in the future. Those are the typical ones. It’s important that we get that information from the tumor tissue. I don’t know what your thoughts are about this. We see a fair number of patients who get neoadjuvant chemotherapy, so a diagnosis might be made on the basis of cytology with initiation of chemotherapy, and then patients will go to surgery after neoadjuvant chemotherapy and get a biopsy sample done at that time. That can sometimes be problematic because of the quality and quantity of the tissue. What are your thoughts about relying on cytology alone, for instance, to make a diagnosis?
Bradley J. Monk, MD, FACOG, FACS: I’m passionate on [core biopsies], and I think you are too. The idea is to put a cannula, like a 14-gauge needle, in and go in and out 20 times. The first time I asked the interventional radiologist to do that, he asked, “You want me to stick her 20 times?” I said, “No, I want you to stick her once, but go in 20 times.” So now we get sufficient tissue. Here‘s the way I would partition it. We have companion diagnostics, immunotherapy predictive biomarkers, and then what I call biomarkers of interest The companion diagnostic today is a genomic instability score [GIS], or a loss of heterozygosity, LOH, which together mean HRD [homologous recombination deficiency]. Basically, there are several excellent laboratory [tests], in the clinical trials they’ve been run by Foundation [Medicine], greater than or equal to 16%. Then there’s MyChoice, which is greater than or equal to 42%. Another companion diagnostic based on your hard work is folate receptor alpha…greater than 75%, and that‘s very exciting, it’s CDx. That will always be done in most of the labs that you do in folate receptor alpha, it’s IHC [immunohistochemistry], and GIS and LOH is genomics. The second would be IO [immunotherapy] markers. There are some ovarian tumors, generally not high-grade serous, that have a high tumor mutational burden, and MSI [microsatellite instability] high.
The third category are what I call biomarkers of interest. I like to look for CCNE, which is opposite of BRCA, and predicts an opposite prognosis. Now we can target that through clinical trials, as you know. The HER2 world is exploding. The other would be ARID1A, particularly in clear cell endometrial. Did I miss anything? What other biomarkers would you look at when you would look at a comprehensive profile?
Robert L. Coleman, MD, FACOG, FACS: I think those are what we have aligned with us right now. As we‘ll talk about what’s available in the recurrent space with NaPi2b expression, that may be another one we may be adding as the ADC [antibody-drug conjugate] world expands and becomes informative, those types of things might become available to us. The other one that you hear about a lot, but I don‘t know how to act on it, is estrogen and progesterone receptor, which we see in some of our endometrioids. We also see that in high-grade serous. The issue there is, is it informative in the way we might use it in breast cancer? A lot of people listening in might be thinking, well, shoot, I get breast cancer markers and these are part of that profile, should it work in ovarian cancer? We’ve come to understand that in high-grade serous at least, expression of that receptor doesn't necessarily translate into enzymatic activity or biological activity as a target. But those would be other ones we could potentially try to get.
Transcript edited for clarity.