Heather Wakelee, MD, discusses the latest developments with EGFR inhibitors and immunotherapy agents in non-small cell lung cancer.
Heather A. Wakelee, MD
The scope of non—small cell lung cancer (NSCLC) has been progressively transforming every few months, with regulatory decisions and emerging regimens making their way through the pipeline, most recently with EGFR inhibitors and immunotherapy agents.
For example, the second-generation EGFR inhibitor dacomitinib was found to reduce the risk of disease progression by more than 40% and resulted in an average 6.5-month improvement in response duration versus gefitinib (Iressa) as a first-line treatment for patients with advanced, EGFR-mutant NSCLC, according to data from a phase III trial presented at the 2017 ASCO Annual Meeting.1
However, dacomitinib’s potential impact could be lessened, as the results of the phase III FLAURA trial, which compared osimertinib (Tagrisso) with gefitinib or erlotinib (Tarceva) in the frontline setting of patients with EGFR-mutant NSCLC who harbor the acquired T790M mutation, demonstrated that osimertinib significantly improved progression-free survival (PFS).
In the immunotherapy realm, Heather Wakelee, MD, says that oncologists are still grappling with the May 2017 FDA approval of pembrolizumab (Keytruda) with pemetrexed/carboplatin as a frontline treatment for patients with NSCLC regardless of PD-L1 expression, which was based on results from cohort G of the small, phase II KEYNOTE-021 trial. In the cohort, the median PFS was 13.0 months with the addition of pembrolizumab versus 8.9 months for chemotherapy alone (HR, 0.53; 95% CI, 0.31-0.91; P = .0205).2
During the 2017 OncLive® State of the Science SummitTM on Advanced Non—Small Cell Lung Cancer, Wakelee, associate professor of medicine (oncology) at the Stanford University Medical Center, discussed the latest developments with EGFR inhibitors and immunotherapy agents in NSCLC.Wakelee: The field of EGFR-mutant lung cancer keeps growing as far as the number of agents we have to help people. There has also been movement in the best testing techniques. Since EGFR mutations were identified, the primary tests have been based on tissue. Now, there is a movement toward some of the liquid biopsies with plasma testing. There was the approval of the first one of those about 1 year ago to look for the primary activating mutations, and now for one of the resistance mutations, as well.
When a patient comes in with newly diagnosed lung cancer with an EGFR mutation, we have 3 drugs to choose from. There is a lot of discussion around which one is going to be best for individual patients, looking at the differences in toxicity and potency. In the United States, we are still using a lot of erlotinib, afatinib (Gilotrif), and gefitinib, which are also very good choices.
At the 2017 ASCO Annual Meeting, we heard about dacomitinib, which wasn’t really a new drug but it is one we hadn’t heard about in many years—so it wasn’t on anyone’s radar anymore. However, they did a head-to-head study, the ARCHER 1050 trial, of dacomitinib versus gefitinib and showed that dacomitinib had a significantly better PFS, well in excess of 1 year at 14 months. For most of the phase III trials with any of the first-line EGFR tyrosine kinase inhibitors (TKIs), it has been under 1 year or close to it, but not in that range. That was very interesting and exciting to see that bar get moved.
The issue, though, is that the drug is more toxic and we have seen that in other agents where there is more potency and more toxicity, so a lot of patients needed dose reductions. Therefore, it is going to be interesting to see how that plays out. Once the drug actually makes it to the market in the United States, I am not sure we know exactly what is going to happen there.
One of the challenges is that with one of the next-generation drugs, osimertinib, there is a head-to-head study right now of osimertinib versus other first-line EGFR TKIs. If that ends up showing superiority, there is a chance that will change. It is better tolerated.
There is a lot that is going to depend on how good the PFS is, because right now with dacomitinib, erlotinib, gefitinib, or afatinib, once one of those drugs stops working, we have the next step, which is to go to osimertinib. If you start with osimertinib, we don’t have that next step. Those are going to be the questions we have to grapple with. Is it going to be better to go from a first-generation drug to a third-generation drug? Or, should we just skip and start with the third-generation drug? We don’t know that yet. If it turns out that it is better to start with the third-generation drug, then, unfortunately, dacomitinib came too late for us to really be using it too much.
I also talked about the resistance that happens. We know that for the first- and second- generation drugs, the T790M resistance mutation is the most common mechanism of resistance and that is where osimertinib plays a good role. However, there are patients who don’t have T790M. For them, we have chemotherapy; we don’t have as much else to offer them. There are a lot of combination trials being done: afatinib with EGFR antibodies, afatinib with other compounds and other different strategies, and MET inhibitors. However, we don’t have a definite [answer that] this is what we should do, so there is a lot of research still happening in that field. Well, osimertinib is a very potent EGFR TKI. Therefore, it [targets] the activating mutations very well, including some of the less common activating mutations. It doesn’t work against exon 20 insertion, and that is an area where there is still a lot of investigation happening with some ongoing trials. However, there is nothing that we know of that works against exon 20 right now.
Osimertinib works well against T790M. However, if a patient developed a bypass pathway, osimertinib is not going to work for that. If the patient developed MET overexpression as the primary resistance mechanism, it is not necessarily going to work in there, either. There are certainly limitations to it but for the patients with T790M—which is the majority—it is very effective.
For some of these other resistance mechanisms, could bevacizumab (Avastin) play a role? Some physicians prefer to use an off-label combination of bevacizumab and erlotinib.
Regarding erlotinib/bevacizumab as a first-line combination versus erlotinib alone, we have phase II data from Japan that did show an improvement in PFS that was kind of in the range that we saw with dacomitinib. It was actually 16 months. However, that was a randomized phase II trial and, because of that, there hasn’t been wide adoption of it. Also, there is the fact that patients are then coming in for infusions every 3 weeks.
There is a larger trial that is being conducted and completed in Japan; however, we haven’t seen data yet. There was also a US study that is still enrolling. There is also a study of erlotinib with or without ramucirumab (Cyramza) which is a VEGF-R antibody. Those [trials] have the potential to change practice but they haven't yet.
There is not as much known if you have already developed resistance to erlotinib, afatinib, or gefitinib. What role does bevacizumab play there? Can you add it later? Some of us will do that if someone is starting to have slow progression—maybe then, you add bevacizumab. But, there is really no data to support that. There is an ongoing trial looking at osimertinib with bevacizumab, but we haven’t seen any data from that either. At this point, I am not using osimertinib as a first-line drug unless someone has a de novo T790M mutation. Otherwise, at this time, I am still using one of the first-generation drugs. We don’t have dacomitinib available in the United States, so I don’t have to answer that one.
As far as deciding between erlotinib, gefitinib, and afatinib, I talk with patients about the fact that afatinib is a little bit more potent, but it also tends to have more toxicity. Erlotinib is the more standard one we use in the United States. It is sort of the basis for a lot of the clinical trials still. Gefitinib is what I will think about giving to my patients who are a little bit more frail, just because it is better tolerated. Also, for a lot of my patients who are Asian, gefitinib has been a more standard EGFR TKI in Asia for a very long time. Therefore, it is something people are used to hearing about, so I will give that often in that setting, too. The world of immunotherapy is rapidly evolving, and it is important to step back and think: where did we start with it? The first indications were all in the second-line setting, so patients who previously had chemotherapy were randomized to then get either docetaxel or 1 of the immune-based drugs. There was almost the same design for nivolumab (Opdivo), pembrolizumab, and atezolizumab (Tecentriq). In all of those trials, they all showed superior OS to docetaxel. The hazard ratios were not that far off from each other—in the .7 range, a little bit better here and there—but very similar.
The differences were a bit around selection so, with nivolumab, atezolizumab, and pembrolizumab, higher PD-L1 expression was tested in different ways and seemed to correlate with some better outcomes. However, in all [of the trials] there were also patients with very low PD-L1 expression who were still benefitting. For the pembrolizumab study, patients had to have some expression, whereas in the other trials patients didn’t have to have any. We are still grappling with [the question]: what does that mean in the second-line setting?
Where things have changed the most in the past year was the first-line indication. The KEYNOTE-024 trial was for patients with high PD-L1 expression [via] the IHC 22C3 pharmDX assay used for pembrolizumab. For those over 50% or at least 50%, patients were randomized to either get chemotherapy or pembrolizumab. There was superiority not just in survival, but in response, PFS, and OS with pembrolizumab in that setting. That rapidly changed practice.
We had some follow-up at the 2017 ASCO Annual Meeting by Dr Julie Brahmer looking at how many people were still doing well. It was pretty high numbers around 18 months. They also were looking at the PFS2, which are patients who responded, stopped responding, and then went to chemotherapy. However, the benefit from chemotherapy seemed to be higher than one would have anticipated, so there are questions about perhaps there is some sensitization toward chemotherapy that can happen. A lot of investigation is happening there. A lot of work is also looking at what else can we do to rev up that response; what drugs can we add to it? Can radiation play a role? There are many studies. How many different angles are we looking at?
At this time, there is not really a clear winner there. One of the IDO inhibitors was talked about at the 2017 ASCO Annual Meeting, but those were pretty small numbers and careful selection, so I don’t know if that is going to be the answer. There is a lot looking at the CTLA-4 inhibitor plus the PD-1 inhibitor combination, and that data is evolving and looking promising. However, we are not clear about where that is going to go—first-line, second-line, third-line—and what toxicity questions we have.
We are also looking at chemotherapy and, of course, there was an FDA approval for chemotherapy plus pembrolizumab. However, that was a pretty surprising approval for most of us because it was based on a randomized phase II study that was small, and while there was a PFS benefit in some of the subgroups, there was no OS benefit. In some of the subgroups, especially if you look at the PD-L1 expression between 1% and 49%, chemotherapy was actually better than the combination. So, we were really surprised by that indication.
Also, there is the fact that the label doesn’t exclude patients with EGFR and ALK mutations, and that is one of things we feel is critical. If you look at who actually went on those trials, first-line patients with EGFR or ALK abnormalities were not allowed on those trials, and first-line patients with EGFR mutations or ALK translocations really should get a focused targeted drug, even if they have high PD-L1 expression. Many of them do, but that doesn’t mean they’re going to respond the same way as someone who didn’t have a driver mutation. That is a really important point, and one that most, but not all, of the community knows. It’s hard because the patients really push to get immunotherapy if they have PD-L1 expression; they have heard that message. However, they don’t always put it in the context of understanding that if they have an EGFR or ALK abnormality, that is a better pathway to start with. With atezolizumab, we are waiting for the first-line data. It has its second-line approval and third-line approval for patients. If they have high PD-L1 expression, they are going to get pembrolizumab in the first-line setting but, in the second-line setting, atezolizumab is being used. It is something I am prescribing for my patients.
Nivolumab still has a role in the second-line setting. The data are just as good as it is with any of the other drugs; it was just the first-line trial where things didn’t go quite the same way. That has raised some questions; whether they are warranted or not is another study. The combination trials that are being done are involving various different combinations of all those drugs.
We haven’t heard much about durvalumab (Imfinzi), but there is going to likely be more this fall with the PACIFIC trial, which was the first one in patients with stage III disease. It was for patients who had chemoradiation, and then went on to get durvalumab or not. We have only seen that little press release that said it was positive, but we don’t know any of the details yet. The expectation is we will be hearing more about that in the fall, and that could get durvalumab to be talked about it quite a bit more.
Avelumab (Bavencio) is available but not in lung cancer. It’s on the market for other indications and, certainly, there are ongoing trials in lung cancer, so we'll find out more about it. There are many others being developed by other companies. As far as how are we going to pick which one to use, I don’t know yet. As other first-line trials read out, we perhaps won’t all be giving pembrolizumab in the first-line setting for the high PD-L1 patients; maybe we still will. As far as toxicity differences, we haven’t seen those very clearly. It is different than with the TKIs.
With these immune drugs, most patients do okay. Some get fatigue and some have absolutely horrible things happen to them. People are becoming more aware of the risk of the horrible [adverse events], how to best manage them, make sure that we are watching those patients closely, and educating patients to tell us when things aren’t going well so we can initially start steroids and help them get through that.
Editors Note: This interview took place prior to AstraZeneca’s announcement of the phase III FLAURA trial’s findings, as well as the FDA’s decision to grant durvalumab a breakthrough therapy designation as a treatment for patients with locally advanced, unresectable NSCLC.