Biomarker Advances Continue to Drive Progress in NSCLC

Siddhartha Devarakonda, MD

The arrival of next-generation sequencing, specifically for patients with non—small cell lung cancer (NSCLC), has allowed physicians to identify molecular abnormalities and potential therapies to match those targets. Though the turnaround time is longer than the standard panel to test for EGFR/ALK/ROS1 mutations, it has the potential to unravel additional mutations that will provide even more information about a patient’s prognosis and likelihood of therapeutic response.

“The biggest point is that lung cancer is not the same as it was 10 years before,” said Siddhartha Devarakonda, MD. “Testing is very important, and if you don’t test, you will not find alterations. Some of these drugs have very good clinical efficacy in inducing responses and prolonging patient longevity. It is really important to test for these alterations.”

OncLive: Can you provide an overview of your presentation?

Devarakonda, MD, a senior fellow at Washington University School of Medicine in St. Louis, lectured on biomarker testing in NSCLC during the 2017 OncLive^® State of the Science Summit in Advanced Non—Small Cell Lung Cancer. In an interview, he shared insight on the importance of testing for these molecular markers, some of the rarer mutations, and what work lies ahead for the field in this area.Devarakonda: Just like you wouldn’t treat breast cancer without knowing if a patient is estrogen receptor—positive, progesterone receptor–positive, or HER2-positive in 2017, it is very important to look for biomarkers in lung cancer.

Some of these have a bearing on patient prognosis, and help to identify the right therapy. One thing to remember is that lung cancer is a molecularly heterogeneous disease; you have small cell lung cancer, squamous cell carcinoma, and adenocarcinoma. These are the most common subtypes. Each of them have a distinct biology, and that is 1 thing we have learned with a lot of sequencing studies.

Adenocarcinomas are typically characterized by alterations in the RAS/RAF pathway. In SCLC, on the other hand, you typically have your TP53 retinoblastoma mutation, but squamous cell carcinoma differentiation and a few structure alterations is what characterizes them.

Of these, the 1 thing that’s clinically important as of today is the receptor TKI pathway in adenocarcinoma, because that is the 1 we have medications for. This can be through EGFR mutations, KRAS mutations, BRAF mutations, translocations in ALK/ROS1/RET, and so on. The bottom line of this is you have a lot of alterations in lung cancer; in adenocarcinoma, you have the receptor TKI signaling pathway. It is important to pick up these alterations because a lot of these can be targeted today with TKIs, so that is one of the big take-home messages here.

Alterations in the RAS pathway tend to be mutually exclusive. One good analogy for that is if you have lights turned on in a room, you don’t have to turn it on a second time. The reason this is important is that sometimes even KRAS is not a targetable alteration. Sometimes, we don’t end up testing for those alterations because if you find it, it is often very unlikely that you will find something else. You don’t have to delay treatment unless the next-generation sequencing panel comes back and your EGFR/ALK/ROS1—the ones that you often test first-line for—are negative.

How do you test for these alterations? The right tests to pick depends on what you’re trying to look for. For mutations, PCR has been the standard test but next-generation sequencing is becoming more popular. The one advantage is if you are trying to do a PCR test, you need to have prior knowledge of what you're looking for. That is easier for KRAS, because KRAS is typically activated by mutations in 3 spots. EGFR, however, is a pretty large gene and then you find a variety of alterations in it about 90% of the time. It is usually the exon 21 point mutation, or exon 19 deletion, but every now and then you do find these rare mutations such as exon 20 and exon 18, some of which you can target with drugs. For instance, afatinib (Gilotrif) has good clinical data from the LUX Lung studies.

It is important to test and that is 1 way of testing. Next-generation sequencing, on the other hand, has the advantage of sequencing the whole gene. You have complete knowledge of what is going on in that gene when you do your test. Secondly, you get all kinds of tissue, as well, because that allows you to do panel testing. You can test for multiple genes at once. The big limitation is that it still takes a good 2 to 3 weeks to get your results. What we end up doing clinically is send out a test for EGFR/ALK/ROS1, which is standard and quicker. If all that is negative, then we send out for the extended panel while we are treating our patients with standard-of-care treatments.

The other biomarker that deserves mention here is PD-L1 testing; obviously, that has moved to the first-line setting. That is through immunohistochemistry (IHC) because it is expression that you are looking for. For the other targets, you can sometimes use IHC. Fluorescence in situ hybridization is a good test, too, and you can look for the translocations in ALK and ROS1. Again, this is where next-generation sequencing comes in. It is attractive because you’re not just looking for mutations, but translocations, amplifications, and these alterations in multiple genes in 1 step.

Finally, where is all of this headed toward? We know that PD-L1 is a biomarker today that predicts response to immunotherapy. As sequencing becomes cheaper, other markers might become more important. We know that mutational burden in a tumor is very important and a very good predictor of patient response to immunotherapy. That may gain popularity going forward, but it needs to be clinically studied and validated.

Are there specific biomarker-driven trials ongoing to get more information on some of these newer targets?

Also, sequencing a tumor gives you an idea of not just stimulating the mutational burden, but also to trying to design specific personalized vaccines against mutant peptides.When it comes to newer targets, the ones to watch out for is the MET exon 14 skipping mutation. There are a couple of good papers that have been recently published in the last year or so, showing that this is a target with a lot of alterations. You get good responses when you use crizotinib (Xalkori) in these patients. Although not approved [in lung cancer], cabozantinib (Cabometyx) is a good drug. As a matter of fact, crizotinib started out as a MET inhibitor, but then it gained traction in the ALK world; but, it’s coming back to MET.

Secondly, another thing to mention is the BRAF V600E alteration. We have some good clinical trial data showing that you can target the BRAF V600E using drugs that you use in melanoma. Both dabrafenib (Tafinlar) and trametinib (Mekinist) basket trials and vemurafenib (Zelboraf) have also been linked with responses in lung cancer that had the V600E mutation.

At what point do you get genetic testing on patients’ tumors?

The 1 other thing would be the NTRK1 mutation and the fact that it gets rearranged in lung cancer and can be a driver alteration; that knowledge is new. Drugs such as entrectinib are something to watch out for, as well.Often, when we biopsy patients at the time of diagnosis we like to get as much tissue as possible. They know that lung cancer is not something where testing is optional; you must do it to get patients the best care.

Even though next-generation sequencing is attractive, we still like to get some more information using targeted sequencing because the turnaround time is faster. We end up testing for EGFR/ALK/ROS1 first, then PD-L1 is 1 more thing we like to test. This is almost reflexive at our institution. Depending on what is available—if they still have tissue left—we like to send out for extended panel. KRAS is something we look for, as well, because it will give us an idea of how extensively we need to test.

What are the pros and cons of next-generation sequencing?

Do you predict that PD-L1 testing will still be just as important going forward?

Also, we have kind of been using the circulating tumor (CT) DNA testing as well in patients who are not fit for a biopsy and don’t have lesions amendable for a safe biopsy. You get your immediate testing done for that, it helps you pick a therapy and, as you keep moving forward, you send out a CT DNA and wait for the results to come back.From a scientific standpoint, there is no disadvantage but you can now do this on CT DNA. Most of the CT DNA tests use next-generation sequencing that basically captures the DNA. Next-gene sequencing is much faster compared with traditional sequencing. As of today, the biggest challenge is its cost. I don’t think these panel tests are approved by all insurance companies; that is also something that delays things for us other than the turnaround time.That is a very tough question. With the immune system, the interaction is not quite as straightforward. It is not just PD-1/PD-L1; it is probably close to 100 molecules in that interface. PD-L1 seems to be a good biomarker based on clinical data we have as of today, but if there is a better one that is something we will need to wait and watch for.

We do have these PD-L1—negative patients who respond, as well. It is a good biomarker, especially given that clinical trials have shown the advantage of testing, especially with pembrolizumab (Keytruda). If another better biomarker would replace that, time will answer that question. Personally, it likely will [replace it], but I don’t predict the future.