John L. Marshall, MD: Sameek, Katie mentioned fusions. We talk about different gene testing, next-generation sequencing, immunohistochemistry [IHC], RNA sequencing, proteins, phosphoproteins, all sorts of different tests that are out there. Could you walk us through the testing that you think is really important to be doing in the bile duct space?

Sameek Roychowdhury, MD, PhD: Katie nicely outlined the wide range of genetic changes. It’s quite the menu of therapies that we can offer patients. Beyond those genes, there are also different types of mutations. For example, the classic mutation is a single-letter change, BRAF V600E, resulting in 1 amino acid change that makes that gene active; it can be targeted with a RAF and MEK inhibitor, which we see in cholangiocarcinoma. Another type of change is the gene amplification classically seen in ERBB2 or HER2 amplification in breast cancer. We see that in cholangiocarcinoma; it can benefit from ERBB2 antibodies or inhibitors. The last one to talk about is fusions. A classic example of that is the BCR-ABL translocation in CML [chronic myelocytic leukemia]. The reason I highlight that is that it’s a little different in terms of detection. The first tests that we had for ABL fusions in CML were based on FISH [fluorescence in situ hybridization]. You can detect a break in the ABL gene or a rearrangement of BCR and ABL together. Today, we’ve evolved to using next-generation sequencing, but not all next-generation sequencing is the same. I think the key differentiation factor for testing is, is your test using RNA sequencing?

Generally, if we’re looking for fusions, RNA sequencing will be better. Nowadays, many of the commercial entities available to us for use incorporate RNA sequencing. There’s DNA testing and RNA testing. The RNA testing is particularly valuable for fusions. As long as you’re doing DNA and RNA testing, you’re likely to capture all these genes, mutations, amplifications, and gene fusions or rearrangements.

The last thing I would say in terms of test choice is the role of liquid biopsy. That uses fragments of circulating tumor DNA that are shed into the bloodstream, and not all tests are great for detecting gene fusions. They’re probably decent for point mutations or gene amplifications, but they’re not perfect. For those liquid biopsy approaches, it’s still preferred to get a tissue-based test for next-generation sequencing. In lieu of having that availability—you can’t do it, it’s not safe, not feasible, time is of the essence—then a liquid biopsy can be helpful, but with the caveat that it may be missing 10% to 40% of the fusions out there. If it’s positive, it’s positive, and if it’s negative, I wouldn’t discount another gene being present.

John L. Marshall, MD: That’s terrific. Milind, let me ask you to drill down on this a little more. Some centers, I think yours initially, said we’re going to build our own panels internally. Some like mine said we’re going to have partnerships with companies that are doing this space so that we didn’t have to keep up with it, we were going to let them keep up with it. There’s still a role for immunohistochemistry here with HER2 testing and PD-L1, MSI [microsatellite instability], and IHC is the quickest way to determine that. At a center where I think you guys have an internal panel, how are you dealing with this expanded testing that’s required to find the right patients?

Milind Javle, MD: Not very well is my quick answer. Most centers have attempted to make the panel without necessarily recognizing the heterogeneity, particularly with rare diseases such as cholangiocarcinoma. We had our own fantastic panel, but it was a struggle to convince people that I’m not going to send tissue there because they don’t check FGFR fusions. At this point, there are 150 fusions. The current platform that we have does detect fusions, but I think there’s variability.

What are the standards here? What are we comparing? How much are the false negatives? I don’t know the answer to this. I feel that we need to partner with the industry or partner in a logical way, so we can have the most up-to-date platforms. All of these approvals that Katie will talk about in a minute have an approved companion diagnostic for a reason, so you don’t do a 50-gene panel in house and then say that this particular genetic mutation doesn’t exist. I think there should be some uniformity and certain standards, which I don’t believe there are at this point.

John L. Marshall, MD: I even think about PD-L1 IHCs, I think there are 6 different antibodies that are used that are companion diagnostic approved. If you did that in your own shop, your pathologist, in theory, is supposed to keep track of which antibody and how to read it out. The target is moving so quickly that it is hard for any of us to keep up. As the 4 of us watch this space just for GI [gastrointestinal] cancers, I am not always sure about what I’m looking at and the result that I have in hand. Our partners who are doing all of the cancers, how do you keep up with this? When those people come knocking on your door, saying our test is the best, we want to do the best we can to help give you the tools to ask the right questions and figure out who to partner with for you and your practice. This was a great discussion on that.

Transcript Edited for Clarity