Robert A. Figlin, MD, FACP: Hi Roy, Bob Figlin. Steven Spielberg Family Chair at Cedars Sinai Cancer in Los Angeles, the ensign professor and deputy director at Yale. Nice to see you again.
Roy S. Herbst, MD, PhD: Great to be here.
Dr. Figlin: We finally get the overall results from ADAURA. It's an absolutely critical trial for people to understand how to use osimertinib in EGFR-positive resected disease. You just gave the plenary presentation for that, so summarize the data for us.
Dr. Herbst: Oh yeah, I'm still riding on a high from that. It was so well-received. We presented ADAURA at the virtual plenary three years ago, the disease-free survival, which was in fact the primary endpoint. At that time we did not have survival data. It was just way too early. But the disease-free survival was a hazard ratio of 0.17, so 83% improvement in disease-free survival. But the trial needed to make sure.
Some updated disease-free survival were presented earlier this year. The hazard ratio came up just a little bit, but we're still in the 70% to 80% improvement in disease-free survival, and everyone was waiting for the overall survival. It was just presented. The bottom line for stage II and IIIA patients completely resected, chemotherapy or not, who then received osimertinib had a hazard ratio 0.49. So a 51% improvement in risk of death. That was quite good, I think. And then, patients who were stage I, II, and III, also had a hazard ratio of 0.49, so a 51% reduction.
That was better than expected. I would've expected a good survival result based on the disease-free survival. But now to be in the hazard ratio less than 0.5, I think is really very nice. It sort of gives us more confidence in the whole paradigm that taking a targeted drug and using it in the right patient early can really have an influence on their overall outcome.
Dr. Figlin: Knowing that you think about clinical trial design and endpoints, are you surprised with the difference between disease-free survival hazard ratio and overall survival hazard ratio?
Dr. Herbst: No, because we only treated patients for three years, so there might be something there. Some patients might need more. By the way, some patients might need less. But disease-free survival is a much more pure endpoint. I think if you look at the disease-free survival and you look at the forest plot and all the subgroups, they're all quite significant, and, to the last, left of one. Of course, in overall survival, we have far fewer events at this point, and all the point estimates are to the left of one, but there still are a couple of groups that might cross the hazard. It's a secondary exploratory analysis. But no, disease-free survival is always a little bit more clean. I have done many trials in my time. I've never seen survival like this, certainly in lung cancer, where you have this sort of magnitude of difference. It just speaks to the progress we've made.
When we started all this, we had chemotherapy to offer for these patients, and we were having curves that you could barely put the pointer in between. We called that a positive; five percent to seven percent improvement. We had to do thousand patient trials in order to see the difference, and meta-analysis for chemotherapy. That's still valid. We should still use some of those drugs. Now, in the right patients, targeted therapy can have an amazing effect, as well.
Dr. Figlin: Let's shift gears. I know this is a complicated area, but the trial designs are nice because you get the patient that's already EGFR positive. You know their next-generation sequencing, you know their EGFR status. In the real world, our colleagues in the community often see I to IIIA non-small lung cancer patients long before they know what their next-generation sequencing is going to be. How should we be approaching those patients now, with your data showing such a dramatic benefit from EGFR-targeted therapy?
Dr. Herbst: I mentioned this in the plenary. We have to have EGFR sequencing at the time we see these patients. Certainly, for those who are non-squamous, and I actually like to do the squamous patients as well, but certainly for a non-squamous patient, you want to know the EGFR status. You also want to know PD-L1 status because immunotherapy is in play in the neoadjuvant setting as well. It really speaks to the need for wherever you are, wherever you're practicing, to have access to a multimodality tumor board that includes your pathologist there to help you with the EGFR and the biomarker testing. Then you also, of course, need to have a surgeon so that you can look at the tumor and decide, is this a surgically-resectable tumor or not? Because if you're going to do surgery, then you have to then say, "What's the molecular characteristics?"
Do you go the ADAURA route? Or some might say, "Do you go one of these perioperative routes?" And we have Checkmate 816, which is a standard of care, neoadjuvant nivolumab plus chemotherapy. Now we've just seen data here at ASCO from Merck, the KEYNOTE 671, there's an AEGEAN trial at the American Association for Cancer Research annual meeting. There's a lot out there. It really does speak to the need to have a discussion before you set out to treat the patient. We need these tests early. It's got to almost be a reflex now to get EGFR testing in patients.
Dr. Figlin: Just to be clear, in patients in the real world, you think that doctors should be waiting for systemic therapy until such time as they know the PD-L1 status and the EGFR status before deciding upon whether to go neoadjuvant versus adjuvant therapy or the EGFR-targeted therapy?
Dr. Herbst: Unless there's some medical reason that one cannot wait, which might occur, but with either centers doing it themselves or most places, I think we ship it out, and you should be able to get it back within a week to 10 days, assuming you have the mechanism to ship it out and get it back. I think yes. I would want to have that information. When I see a patient, and we have a multimodality tumor board, I probably have one tomorrow morning. I'll have to call in. But what we will do is I'll ask the surgical group: “Is this a resectable patient?” Then I'll ask, “What is their genotype benefit?” And if they turn out to be EGFR-mutated, we're not going to want to use neoadjuvant immunotherapy there. If they are surgical, then you might say, “Can this patient tolerate a new immunotherapy, and do they have any contraindications?” Then we have to discuss it with a patient.
I think now, I think that's a very reasonable way to go. There are still some patients who will get chemoradiation too. We mustn't forget that some patients probably shouldn't go to surgery. One thing that really worries me sometimes is that, in a non-EGFR-mutated patient, people will take a big tumor and think they're going to shrink it with an neoadjuvant. Those are probably the patients that never get to surgery.
Dr. Figlin: Exactly. Lastly, just again for clarity, for our community oncologists, it's tissue-based next-generation sequencing and PDL, not circulating tumor DNA.
Dr. Herbst: For the first diagnosis, yes, I want it to be tissue-based. I've had a couple of cases where at least you have tissue that shows you've got a lung carcinoma then and there isn't enough for molecular analysis. Sometimes you can accept a blood-based test in that setting, but certainly if it's negative, if the blood's negative, I want to do it on the tissue because it's a sensitivity issue.
Dr. Figlin: There's not complete concordance between blood and tissue.
Dr. Herbst: I think the blood is pretty specific but not that sensitive. It's all evolving. We are personalizing the treatment of lung cancer in early disease, and we need to get the word out. It's very important.
Dr. Figlin: Lastly, before I let you go, are your surgeons on board? I mean surgeons, when they see something that can be removed, it is what they learned to do, telling them to wait a bit, has that been a challenge for you at Yale?
Dr. Herbst: For the ADAURA, I think they're going to be more on board now that we have survival. I do often hear from the surgeons, "Where's the survival?" Although again, in ADAURA, you can get the patient after surgery too. For the neoadjuvant, it's coming around. We're actually doing trials now. So that's a wonderful area to look at new drugs and new combinations.
So yes, I think that, even in the academic setting, it's taking a while to come around. I can imagine in the community it might even take a little bit longer of. But again, you have to know which patients you can do this on. The surgeons need to gain experience with it. Once a surgeon sees that they remove a tumor and it's melted away, and they look at it in the microscope and it's a pathological complete response, you only need to see a few of those, and you're getting 17% to 20%. That's pretty good. Now the other 80% we're going to have to figure out what to do with them. That's going to be the challenge for us for the next few years.
Dr. Figlin: Great to see you.
Dr. Herbst: You too. Thanks a lot.
This transcript has been lightly edited for clarity.
