Hello, and welcome to today's ISLC STARS webinar. We're going to be talking about cancer drug development in the era of precision medicine. Next slide. My name is Janet Freeman-Daley, and I will be your moderator for this session. I'm a lung cancer patient, a cancer research advocate, co-founder and director of the Ross Wonders, and a STARS staffer. Our speaker today is William Powell, MD, PhD, head of Roche Pharmaceutical Research and Early Development. Previously, Dr. Powell was professor of medicine, cancer biology and pathology at Vanderbilt University, director of Division of Hematology Oncology and director of personalized cancer medicine at the Vanderbilt Ingram Cancer Center, and a practicing medical oncologist. He pioneered precision medicine for lung cancer by identifying sensitivity of lung cancers with certain gene mutations to the drugs we now call targeted therapies. Dr. Powell, how would you like me to address you? Janet, you can just call me William. That's fine. Great. Thank you, William. Next slide. Before Dr. Powell begins, or William begins, we have some housekeeping notes. If you would like to download today's slides, you can access them by clicking on the link in the chat. That's a link to the webinar page on the IASLC website. The recording of this session will also be available there within the week. Your camera and microphone will remain off for this webinar. Please enter any questions you have for the Q&A session using the Q&A button at the bottom of the webinar page. We will not be using the raise hand function or the chat function for Q&A. However, you may use the chat function for other discussions. Also, please note that William cannot give medical advice during this webinar. We will not be taking any questions about individual patient cases. Next slide. This webinar is part of the IASLC STARS program. STARS stands for Supportive Training for Advocates on Research and Science. This program aims to increase the number of lung cancer patient research advocates that are equipped to provide accurate translation of lung cancer science and research for other patients and their caregivers and to bring the patient perspective to lung cancer research and policy. Funding for this webinar is provided by STARS sponsors Amgen, Bristol-Myers Squibb, Genentech, and Lilly Oncology. A link to the STARS webpage is in the chat. Next slide. Here are our disclosures. Now for the reason we're here. William, please tell us about cancer drug development in the era of precision medicine. Thanks a lot, Janet. And it's really great to be here and speaking to the STARS of the IASLC, which has been a great organization as an advocate for lung cancer and lung cancer research. So greetings from Basel. I hope you guys are all okay, healthy and safe. And for the next 25 minutes or so, I'll just give a overview, including historical context for targeted therapies and now precision medicine in the 21st century. So I'm dating myself a little bit, but when I was a fellow in training in medical oncology, this was the view of lung cancer. So this was in the early 2000s. And as you all know, basically there were two main types of lung cancer, non-small cell lung cancer and small cell lung cancer. This was based on probably more than 100 years of just looking under the microscope. And it wasn't any more differentiated than that. And if you think about it, it's very odd. It's like calling an apple a non-orange, like an orange and a non-orange. But anyway, so you have small cell lung cancer and then non-small cell lung cancer. And then we knew non-small cell lung cancer had three distinct histologic subtypes, adenocarcinoma, squamous cell carcinoma, and large cell carcinoma. So this was the scientific view of lung cancer at that time. And during our training, this was a key paper that came out in 2002 in the New England Journal, one of the top journals that helps guide treatment in medicine. And there was basically a study of thousands of patients with quote unquote modern chemotherapy doublets. So modern chemotherapy doublets meaning cisplatin and paclitaxel or taxol, cisplatin and gemcitabine or cisplatin and gemzar, cisplatin and docetaxel, taxotere and carboplatin and paclitaxel. And this is a Kaplan-Meier curve, where on the x-axis is the months of survival. And on the y-axis, you have the percent surviving. And so at time zero, when everyone starts on the trial, they're all alive 100%. And then over time, unfortunately, patients do not survive. And you can see that basically there's modern chemotherapy doublets, four of them, but there's no difference in survival. So when you were trained as a fellow, basically we were told that you choose the treatment based on the patient's other potential side effects and the kind of side effects that they might tolerate. And also, you have no idea if it's going to work in a patient until you give it to them. So you give it for two cycles, get a scan, and then you'll be able to say whether it's working or not. And then a third of the time, tumors will shrink. A third of the time, tumors will stay the same. And then a third of the time, tumors will grow. So, as you can imagine, it's completely unsatisfactory as a doctor to have this kind of approach to treating lung cancer. And so many of us at the time, including myself, were very, very interested in new therapies that were coming out at the time. And in the early 2000s, the EGF receptor, or epidermal growth factor receptor, tyrosine kinase inhibitors, EGFR2KIs, were just being developed. And one of the first ones was Gefitinib, which we now know as Eressa. And while I was in the laboratory, I was also then in the clinic as a fellow and seeing very dramatic responses in some patients. This is actually a chest X-ray, or serial chest X-rays, from a patient I saw who had come from upstate New York. She had had multiple lines of chemotherapy. She had come in a wheelchair on oxygen. And basically, she told us that she was a never smoker, meaning she had smoked less than 100 cigarettes in a lifetime, and that she had no other options. And so we put her on Gefitinib at the time. It was an expanded use protocol. And literally within five days, we got this X-ray, and you can see all the fluffy stuff over here, which was tumor, had very, very quickly resolved. It was quite incredible, but the most incredible thing is she called three days later to say that she felt like a new woman. She had actually come off of oxygen. She had been out of a wheelchair and really had a very, very dramatic response. Unfortunately, the number of people who had such a dramatic response was really only a minority of patients, really only about 10% in Western populations. And then when we studied what were the clinical characteristics of patients who might have such a great response, it turned out to be women, more likely to be Asian, and then have specifically adenocarcinoma histology, and then also have no history of smoking. So we and others set out to really try to figure out why is this subset of patients who may benefit, and is there a better way to predict who's going to benefit based on some kind of genetic marker. And to make a very long story short, we and others then found in 2004 that there indeed was a genetic basis for why some patients, or only one out of 10, responded dramatically to EGF receptor tyrosine kinase inhibitors like jabutinib and erlotinib. This is a complicated slide, but it basically depicts the EGF receptor protein. And then here in blue, there's the signaling portion, the part that we call the tyrosine kinase portion. And then to make a very long story short, we found that there were four types of mutations, and predominantly there's what we call today exon-19 deletions, and then point mutations, which change a leucine to an arginine at position 858. I won't go too much into details about the science, but it was very clear then that patients who had EGF receptor mutations were the ones most likely to benefit from these drugs. Now, just to tell you how long it sort of then takes history to accept these findings, I will go into that. But before I want to do that, I just want to tell you that there are four main types of genomic changes that can cause cancer. This is at a very high level. So on the far left, you have base substitutions. This means that a mutation actually changes an amino acid, like a leucine to arginine at position 858 in EGFR. Then you can have what are called insertions and deletions. This means in the DNA, you can have things inserted into it or deleted from it, which leads to aberrant proteins. And the example here is an exon-19 deletion. You can also have what's called copy number alterations. Basically, this means you have too many copies of the gene. In normal cells, you have one copy from your mom and one copy from your dad, but in cancer cells, you can get abnormally high copies, and an example of that is metamplification. And then the fourth type is rearrangements, and that's where one part of a chromosome where the DNA gets abnormally linked to another part of the chromosome, and it puts two proteins together that aren't supposed to be together, and that leads to a rearrangement or a fusion. This will become relevant for when I talk about precision medicine in general. Anyway, so we had found these AHF receptor mutations, but what we really needed to do was understand how common it was, and how often are we seeing it in patients, and also can we then figure out why some patients really benefit, like I showed you in the chest X-ray, but then over time, basically after a year, their tumors start to grow again. And we call that disease progression acquired resistance, and unfortunately, it occurs in many, many cancer patients, and back in the 2000s, I just want to bring you back to that time, there was very little understood about acquired resistance, especially in lung cancer. And at the time, there was another disease called chronic myelogenous leukemia, or CML, that's a type of hematologic malignancy, and there was another drug called imatinib or Gleevec that had just been approved in 2000. And learning from our colleagues in hematology, we saw that when patients initially responded and then progressed on imatinib, that they would have specific changes in the DNA in the target of imatinib. And so we thought, well, maybe the same thing is happening in EGF-Arminian lung cancer, where you get an EGF receptor tyrosine kinase inhibitor, it really benefits you for a long time, and then over time, it starts to grow again because of changes in the target. Now, how could we prove that? Well, the reason I'm talking to you about this is because we actually had to establish a new kind of protocol. I was at Memorial Sloan Kettering at the time. And we actually had to try to get tissue, cancer tissue, from patients whose tumors had responded and then progressed. Now, at the time, everyone said when we started this, it was unethical to biopsy patients after their cancers were growing, that patients wouldn't volunteer for that, you know, we would have a hard time getting people to enroll. But we tried anyway, because scientifically, it was really important for us to understand that. And so we wrote this protocol, and basically said, you know, you had to have responded to erlotinib or jafitinib, you had to have had a response, then you had to have your tumor grow, and then you had to have a tumor of big enough size so that we could actually biopsy it. And then we, after patients signed consent, we would actually then put a needle or some kind of biopsy to get the tissue, and then we would analyze it at the DNA level. And this is an example, for example, of a patient who I also saw. This is a computed tomography scan, you know, where patients are like a hot dog sort of put through a tube and then sliced, and the head is at the front and the feet are at the back. And then you can see serially what's happening in the lungs, for example. And here, you can see this patient, this is the chest area, here's the heart, of all of these white fluffy things are tumor. Initially, when this patient started erlotinib, then her tumor shrank, almost disappeared. But then after about two years, the tumor started to grow again, specifically here. She also had a growing lesion in the bone. So she was one of the first patients to consent for the rebiopsy, and we analyzed bone tissue as well as lung tissue. And basically, in the first six patients, we were able to find one of the key resistance factors that led to why patients who responded to erlotinib or gefitinib, why they eventually had their tumors start growing again. Basically, from analyzing the tissue from the patients whose tumors grew, we were able to show they actually had a second mutation in EGFR. And that second mutation was called T790M, that basically changed the threonine to methionine position 790. And this is a picture showing you what that mutation might do. So here is an example in the background here of the EGFR receptor, and then this is the threonine or the T in a normal tumor. And this is the drug erlotinib in the pocket. But when you change the T to an M, methionine, you can see that it starts to encroach here upon the drug and basically doesn't allow for good binding of the drug. And that's a potential explanation for why patients eventually get resistance. Now, you might ask, why is that important? Well, before I get into that, let me just highlight. So this was 2005. We had found EGFR receptor mutations in 2004, then we found a resistance mutation in 2005. But it didn't take till 2009 where the rest of the world community accepted that EGFR receptor mutations were actually a predictor of response. So you might think that's a short period of time. You might think that's a long period of time. You know, it took about four or five years for the whole community to be convinced that EGFR receptor mutation testing was valuable. And so before that, actually, some places did it and some places didn't. And your care was highly dependent upon who believed in EGFR receptor mutation and who didn't. Anyway, the first study that showed the benefit was this ARESA pan-Asian study called I-PASS. And again, these are Kaplan-Meier curves. So it gets a little complicated. There's four Kaplan-Meier curves. So patients with advanced lung cancer were randomized to either ARESA or chemotherapy. And in the overall population, you can see that there was some benefit in the geofitinib arm versus the chemo arm. But the progression basically was very quick in both. If you turned out to have an EGFR receptor mutation, then you did much better on ARESA than if you did chemotherapy. And the opposite was seen if you didn't have a mutation. You actually did better on chemotherapy than you did on ARESA. And so this was really the first evidence from a randomized study that EGFR receptor mutation testing was beneficial. Now, I talked to you about the T790M mutation. The reason that's so important is that we were then able to design a drug to overcome T790M-mediated resistance, or I should say AstraZeneca. And basically, that drug became osemertinib. And so just understanding the resistance mutation allowed us to then go back to the laboratory and to the chemists and say, can you overcome this mutation? And essentially, that did become possible. And osemertinib was approved in 2015. And so from 2005 to basically 2015, in a short time, 10 years, we were able to identify the resistance mutation and then have a drug approved to overcome that. Now, I'm happy to say that since the early 2000s, almost 20 years later, there's actually been great progress in molecularly targeted therapy for oncology. This is not an exhaustive list. And it's not just lung cancer. But it's incredible to see that all of these targeted therapies here are really just targeting patients that are defined by some kind of molecular aberration. And so patients now have multiple options, really depending upon the tumor mutation status that we can identify. And then even in the course of treatment, there are multiple options, which were not previously available. And this is all by subsetting patients and no longer just looking at the histology or what the tumors look like under the microscope. Now, you might say, has there really been an impact, though? What's really been incredible to see is there was a paper in the New England Journal this year in 2020, put out by the National Cancer Institute, looking at the impact of targeted therapies in lung cancer. And specifically, in the top here, they looked at trends in incidence and incidence-based mortality. So the good thing is the incidence has been going down in lung cancer, which is not attributed to the new therapies, but possibly attributed to decreased smoking, etc. But they specifically looked at when EGFR first-line therapy was first approved. I told you testing was first adopted in 2009, but first-line EGFR therapy wasn't approved until 2013. And then when you look at the survival rates, actually, you can see that survival trends among men and women basically started to increase just around the time that EGFR therapies were approved in the first line. So this is great progress. You know, we need to continue to do more. I would say this paper didn't even account for Alcfusion or Rosfusion or even immunotherapy, which has been the new advances that have come since the story about EGFR that I told you about. Now, I also mentioned that, you know, patients now have multiple options. And basically, at different courses in the treatment paradigm, patients can now benefit from these different options depending upon the mutation status of their tumor. This was a really interesting paper from Alice Shaw that was published in the New England Journal in 2016, where she showed the clinical course of a patient of hers with Alcfusion positive lung cancer. And here, the patient had a biopsy, was shown to have an Alcfusion, got chrysotinib, then had another biopsy, then ended up switching to ceritinib, then was on another agent for a clinical trial. Unfortunately, at this time, there were no other options but got chemotherapy. But then the patient had another, because of the time in between the first chrysotinib and the second time, and because the tumor grew again on chemotherapy, she was tried again on chrysotinib. Then there was a new Alcfusion inhibitor called lorlotinib. And basically, the patient then benefited from that. The patient actually had a resistance mutation, showing that she would likely benefit from lorlotinib. But then after initially having her tumor shrink and grow again, then she underwent another biopsy. And it turned out the biopsy showed that she might benefit from chrysotinib, which was the first agent set that she got. And so basically, then a patient was able to go back on chrysotinib. So this really just highlights to you how mutation status and biomarker testing can really help guide therapy during the clinical course of a patient who's receiving targeted therapies for specific mutation-driven diseases. Now, I've touched briefly upon immunotherapy. So as molecular targeted therapies were being developed, so were immunotherapies. And these are therapies that harness the immune system to attack cancer. I don't have time to go into all the details of how those discoveries were made. But first, really, the pioneering work was done in melanoma, a different kind of cancer, a skin cancer. And even in melanoma, first in the early 2010s, studies showed that targeted therapy were better than chemotherapy. So this is a Kaplan-Meier survival curve, where once again, is on the x-axis and progression of survival on the y-axis. The standard of care for all the 2000s was chemotherapy, something called the carbazine. But then in the mid-2000s, melanoma, half of them were shown to have a specific mutation called a BRAF mutation, a B600E mutation. And then basically a drug was developed to overcome that, Vemurafenib. And this was the first study to show that Vemurafenib targeted therapy was better than chemotherapy, just like I showed you with ARESA in the I-PASS study. Well, only a short six years later, immunotherapy was shown to be even more potentially effective in melanoma. And here, what's really incredible, these are Kaplan-Meier curves from patients who were treated with Novolumab, which is a PD-1 checkpoint inhibitor, or a combination of immunotherapies. But the details on that are not important. The really key important factor, if you look at these Kaplan-Meier survival curves, is that over time, instead of this curve going down to zero, this curve is starting to flatten out. So you can see that now almost 20 percent of patients are living 36 plus years. And here, similarly in a separate study, you see now survival rates that we've never ever seen before. And just look at this study over here, where you can see the curve goes down to close to 10 percent. So now with immunotherapy, we're really starting to see long-term survival is possible as well in patients with metastatic disease. Now what does that mean for lung cancer? This slide summarizes a lot of work done in the entire field about the benefit of immunotherapy in lung cancer. The short takeaway is that overall survival in non-small cell lung cancer has definitely improved with the introduction of checkpoint inhibitor therapy. And you can see that in the evolution and history of what the therapies were here on the x-axis, and then the overall survival here on the y-axis. So you can see, for example, when I was a fellow, here's platinum doublet chemotherapy. Even before that, people were always only using single-agent chemotherapy. Then we have addition of Avastin or Bevacizumab, but you're always sort of around the year mark. Once you get to chemotherapy, which is cancer immunotherapy plus chemotherapy or cancer immunotherapy plus other therapy, you're starting to get close to the 20-month survival rate. So it's pretty incredible. And then even with immunotherapy, we have some markers where we can figure out who is most likely to benefit. And the best marker we have right now is something called PD-L1 status. This is an indication of how highly expressed the target, it's one of the targets for PD-1, the checkpoint inhibitors is. And you can see that if you have higher expression of PD-L1 in your non-small cell lung cancer tumor, then you have a higher likelihood of surviving longer. We also see this cancer immunotherapy improving overall survival in extensive stage small cell lung cancer. Most of the advances I talked to you about with molecular targeted therapy have been in non-small cell lung cancer, but cancer immunotherapy is also benefiting patients with small cell, where we've gone from under a year with chemotherapy to now starting to be more than a year with chemotherapy plus cancer immunotherapy. I won't go into too much detail on the rest of the slide here. Now, I think this is the slide that really summarizes where we are today. Compared to 2000, when I showed you that lung cancer was classified according to histology and what it looked like under the microscope, we have a much better way of classifying lung cancer. And the most important thing is it's clinically relevant. And so, what we've seen over the past 15 years or 16 years or so is increasing parts of the lung cancer pie sort of addressed by specific mutations. I talked in 2004 about discovering EGF receptor mutations. Also known at the time was something, a mutation in a gene called KRAS, which is another signaling protein. By 2014, we had many other subsets. And then today, you can look at a gene called And then today, you can see we have many, many different subsets, some of which are only 1% to 2% of lung cancer. And then we can also overlay PD-L1 status for the benefit of immunotherapy. Importantly, this shows you that an all-comer approach to treating lung cancer doesn't make any sense anymore. If only one out of 100 patients has a ROS fusion, for example, you know, first, not everybody should get a ROS inhibitor. But also, it's important to identify the one out of 100 who does have that ROS fusion so that you can get the right therapy to the patient. Incidentally, the current president of ISLAC, Tetsuya Mitsudomi, was among the first to identify the MET splice mutations. And there was just recently a MET inhibitor approved for that as well, that particular subset. Now, where are we going in the future? I think this is great progress. And as you saw from the survival curves, you know, hopefully we're also making a huge impact on survival, as well as the quality of life for patients. In the future, I think we'll see continued evolution of precision medicine. And one of the ways that we'll continue to do that is through liquid biopsies. So liquid biopsies are basically drawing blood from patients who have cancer, and then using DNA in the blood to identify what mutations are potentially in the tumor. The reason this is so critical is about 30% of lung cancer patients actually don't have enough tissue to biopsy. I showed you some of the pictures of where you have to put a needle in a tumor. Unfortunately, up to a third of patients don't even have a good sample. So if you could just get a blood draw and then analyze your blood to see specifically what type of mutation your lung cancer has, it would simplify things a lot. And, you know, there's already an example of a study that's doing that. It's called the BFAST study, where patients with lung cancer basically get a blood-based assay, NGS stands for next-generation sequencing. And then based on their specific genetic mutation that's found in the blood, they're then assigned therapy. And in one study, for example, we showed that patients who got electinib and were found to have an outfusion did just as well as if we had biopsied the tumor. So I think you will expect to see a lot more activity in this space, and it could really simplify the way patients are treated. And it can also help us monitor therapy. So if you're on treatment, and then you're getting blood draws, you can potentially see resistance earlier than we've ever seen before. Another advance that we're going to see is something that we call real-world data, and the use of synthetic controls. As you know, most of our clinical trials, when we have to get a drug approved, we actually randomize patients to the new therapy versus the standard of care. And this is a requirement usually from regulatory agencies to show that our new therapy is better than the standard of care. Well, if you only have 1% of patients who have that specific alteration, it might become hard to do a randomized study because you just might not find enough patients or it might take you too long to find patients. But if we can use real-world data, we may be able to then use in parallel data that's being collected on patients who get standard of care versus single-arm studies where patients are actually getting a treatment. So for example, if we had a single-arm treatment where patients got electinib, but then we followed patients in the real world who were getting chemotherapy, who had an outfusion at the same time, then we could actually use that data to show that electinib actually would be better than standard of care. And here's just an example where we have used that kind of real-world data. For example, when Roche was trying to get electinib approved in countries throughout the world, we didn't always have all the randomized studies that the regulatory agencies wanted. But through use of real-world data, we were able to show that electinib was superior to seritinib, and the regulatory agencies actually used that for approval. So we expect that this could lead to even faster approval and more options for patients in the future. So that was a whirlwind tour of the history of molecularly targeted therapy in lung cancer. These are the main messages that I wanted to leave you with. First of all, it's really scientific breakthroughs and new paradigms that are leading to new and improved outcomes in lung cancer. So back in the 2000s, we looked at lung cancer based on histology and what it looked like under the microscope, and everyone was getting chemotherapy. And basically, as I showed you, there was no benefit from any modern chemotherapy doublet, and there was what we call a therapeutic plateau. So we knew we had to do better, and better meant really trying to understand the genetic basis of these diseases, finding these clinically relevant molecular subsets, and then finding therapies to overcome those targets. And that has led to an all-comer from an all-comer approach to a molecular subset approach, and now there's lots of options for patients, even with just one percent or even half percent frequency mutation. And I did want to just make a plug that patient biopsies before and after treatment are really, really important. I know patients have to undergo a lot of extra steps to do that, but it's really important for us to gain new insights and develop new medicines. Then I talked about cancer immunotherapy, which has also led to increasing advances and more survival rates, and even there, some patients are actually living much, much longer than we'd ever expected. But unfortunately, immunotherapy only really benefits about 20 to 30 percent of patients right now, so we need to continue to do more research there and understand who's benefiting and who's not, and then figure out how to overcome that. And then I talked briefly about a couple new approaches in the era of precision medicine, one of which is liquid biopsies and the other one which is real-world data, and I think you'll see more of that in the near future, and those will lead to even more advances in our ability to treat lung cancer. So with that, I just want to thank you. I want to thank ISLAC for inviting me to speak to you today. I want to thank Janet and Kristen and Aubrey also for helping to organize the session, and with that, I will take any questions. Thank you. Thank you, William, for sharing your perspective. We do have time for questions, and they are getting busy, so the first question is from Terry. She's talking about the drugs being able to treat the brain. As you know, the blood-brain barrier keeps a lot of the drugs out of the brain, and some of the newer generation TKIs can now treat the brain effectively, but we really haven't had a lot of years of experience with drugs that treat the brain. What have we learned about short- and long-term effects of these drugs and how effective they are? Yeah, so you know, that's a great question. So the blood-brain barrier is something that nature has put there to prevent, I guess, toxins from getting in the brain, but at the same time, it makes it difficult to get drugs into the brain. As you mentioned, there have been advances with that. For example, like osimertinib and alektinib and other drugs do get into the brain. We know that patients with brain mets have their tumors shrink, but still the concentration of drug that gets in the brain is lower than in the rest of the body. So I think on one hand, we need to do more just to figure out how to get more drugs in the brain, and that will take additional research. But I think your question is more related to other long-term side effects from the medicines in the brain. And there, you know, I think it's always a risk, a therapeutic, it's a risk-benefit ratio, right? If you have a tumor in the brain, it's more important to treat them versus having some side effects from the medicine in the brain. I think many of the molecular target therapies we have are pretty, how would I say, mostly don't have neurological side effects. There are some that have some visual disturbances and maybe some other neurological challenges, but the ones that are serious, you know, they would never get approved. So I think over the long term, I don't see any negative consequences of most of the drugs that we have developed that can go into the brain. But I think we need to continue to do better on the research that we can get more targeted and higher concentrations in the brain. Thank you. Another question has come up is about the sequence of the development of gefitinib versus getting it approved for EGFR. For most of the targeted therapies, we find the target, we go through clinical trials, and the drug gets approved. Gefitinib got approved before we actually were sure that it targeted EGFR. Can you talk a little bit about that? Yeah. So, I mean, gefitinib was developed at a time when we thought that just EGFR receptor was an important target in lung cancer, but we didn't even know that mutations existed. So, you know, that's just the way sometimes science works, unfortunately. If we know about the target, then it's much easier to try to develop a therapy for it. And so, for example, after ALK fusions were found in lung cancer in the late 2000s, then there was efforts to develop the ALK inhibitors. But unfortunately, I would say in drug development, it's only after the fact sometimes that we find out that there's a particular subset that benefits the most. So, I think what I'm saying is you have to do both. You have to do target-driven discovery. So, if we have great targets, like, say, KRAS-T12C, and then we really try to figure out how to drug that, that's one way forward. But you could argue that the immunotherapies, for example, were developed in the absence of a real target, and then only afterwards PD-L1 was found to be a better predictor. So, drug development, unfortunately, just works in both ways. Okay. Another question that comes in, you mentioned the drug imatinib, Gleevec, for chronic myelogenous leukemia. And right now, statistics are showing that for CML, some of those patients are now achieving almost normal lifespans. Is that possible for lung cancer, do you think? Yeah. So, again, that's a great question. It's very hard to compare cancer to cancer. You know, there's more than 200 types of cancer, and then even at the molecular level, there's way more different tumor types. One of the things about CML, especially if you catch it early, is there's very little what's called molecular heterogeneity. So, for example, it looks like if you catch it very, very early, most of the cells have the same mutation, for example, and then they don't have a lot of resistance mutations. The more cells you have in the body that's cancerous, then the more mutations you can have. And so, the more advanced, the more resistance can develop. So, I would say I think it's possible. Right now, lung cancer seems to be a lot more heterogeneous and have a lot more mutations. But that said, I still know that there's patients with EGF receptor mutations that have lived 10 plus years. And so, it is possible and also other alterations. So, it's possible, but I don't think we fully understand exactly who those are right now. But immunotherapy also is leading to that tail, what we call the tail on the Kaplan-Meier curve, where we're also seeing 10% to 20% of patients live for years that we haven't seen before. And it's really encouraging. So mostly on targeted therapies, we've been focusing on a single driver, but now we're finding that there are some other drivers, passenger mutations possibly, that might have some impact on survival. And one of the questions that's come up is about small cell transformation. Is this occurring as a resistance mechanism? And what kind of research is going on to treat that? Yeah, so the person is asking about small cell lung cancer transformation. So what's happened is patients who originally have non-small cell lung cancer, usually adenocarcinoma, over a long period of time on treatment, start to have progression. And then when you re-biopsy it, it turns out they have small cell instead of non-small cell. What this really means is under the pressure of the drug, in an evolutionary system, the cells have become more de-differentiated. This is like a response to an external pressure. And it is a consequence. Even back in 2005, I think, in 2005, we and others showed that it happened with gifitinib and erlotinib. And then now people have shown it with other targeted therapies. There is still a lot of research going on, on how can an adenocarcinoma transform into a small cell. If you really want to get into the science, some of it has to do with loss of some tumor suppressor genes, such as p53 and RB. But I would say there's not a great insight as to how to treat those yet. Right now, I think people are using chemotherapy for small cell. But there's still a lot of research that needs to be done to particularly target small cell transformation after a long-term on targeted therapy. Incidentally, it doesn't just happen in lung cancer, it also happens in prostate cancer and some other cancers. Oh, okay. I didn't know that. Yeah, so after androgen therapy in prostate cancer, you can also get small cell transformation. Okay. Another question. Are there any fourth generation TKIs in the works that you know of that are being researched or in clinical trials? Fourth for EGFR, I'm assuming, or just in general? The question was general. Yeah, I mean, there's a lot of companies and even small biotechs that are going after next generation TKIs in multiple subsets. For example, even in EGFR, there are companies going after a very specific mutation called C797S, which happens after resistance to osomertinib. And then, for example, in ALK fusion, positive lung cancer, NTRK, multiple examples. I think what's incredible now is using modern scientific techniques, you can also try to predict what is the spectrum of mutations that will occur at the very same time that you're developing the drug, so that you could then be prepared, hopefully, with the next generation in time for a patient. But if it's not in time, then it would be for subsequent patients. Okay. A question from Jill. There's a distinct difference in approach to treatment depending on biomarkers. Shouldn't there be reflex biomarker testing upon diagnosis, like there is for non-small cell lung cancer, but across all non-small cell lung cancer types? If I understand the question, should it be applied to squamous cell and large cell carcinoma? Is that the... Yes. And even maybe neuroendocrine? Well, I'll probably say something that wouldn't follow guidelines. So I don't want to violate any guidelines. But essentially, I mean, I think molecular... To me, molecular diagnosis would trump a histological diagnosis. So every once in a while, we will find an age of receptor mutation in squamous cell carcinoma or neuroendocrine. So I would be supportive of that, but I know that there's guidelines that specifically say like adenocarcinoma has reflex testing. Okay. Another question has come in regarding tumor markers, as opposed to doing biomarker testing, looking for genomic variations. There's some tumor marker used in other cancers like CEA or other things to indicate whether a patient is progressing. Are those useful in lung cancer? Can we use those to figure out what the next treatment should be? Yeah, that's a really great question. So right now, there's no great biomarker like that in lung cancer. You have CEA-19-9 in ovarian cancer, you have CEA in some colon cancers, not all colon cancers. And then some other markers. I think the most promising data look like it'll be the circulating tumor DNA. So there are lots of studies now looking at CT DNA to see if those can be better predictors of whether the cancer is under control or not. I think a lot more research still needs to be done, but it might not be a protein marker like CEA or the CEA-19-9, but more like the CT DNA. Okay. So another question from Anne Marie, we've kind of had a wrench thrown in cancer development this year by COVID-19. Can you talk about how that is affecting research trials and drug development? Yeah. I mean, COVID-19 obviously has been affecting us all pretty much around the world. What we have seen is that, unfortunately, because of the hospitals being inundated by patients with COVID and also the requirements for the intensive care units and other like phase one units and things to be taking care of COVID patients that we have seen an impact. For example, clinical trial enrollment has slowed down. I think it really depends on country. You know, the US has been particularly impacted. Some of the European ones over the summer were doing better. Some of the Asian ones as well. But I think in general, we're seeing a slowing down of clinical trial enrollment and accrual. The other thing that we're seeing is unfortunately that patients are not going to the doctor as much. So they're also skipping screening tests and other things, which then means that maybe some patients will be diagnosed with their cancers at a later stage rather than an early stage. So, you know, I think we really have to, you know, cancer is not waiting for COVID. So we need to really, you know, get the epidemic or pandemic under control and then also make sure the hospitals and the doctor's offices are safe enough for patients to go in and get the treatment that they need. Okay. Thank you. There have been some questions that come in about liquid biopsy. The question is, since tissue is the gold standard, but we know liquid biopsy comes back faster and requires less tissue. Do you think that we could at some point rely on liquid biopsy or do liquid biopsy first and then follow up with tissue to confirm? How would you handle that? Yeah. I mean, I think that's where we'll make some progress. So for example, with the BFAS study that I talked to you about, that's where in particular we did liquid biopsies to find outfusions. And then if we found an outfusion in the liquid biopsy, then we treated patients with alectinib. And then we saw the same response rates that you would see if you had a tumor biopsy. So we have filed that with the FDA and, you know, if approved, then it would be the one of the first times I think that we, the community is saying, regulatory agents are saying, yeah, you only need a liquid biopsy and not a tumor biopsy. So I think we will make progress. That was the point I was trying to make there. That'd be great. Yeah. I'm staying on the topic of biomarkers. We have a couple of biomarkers for immunotherapy, PD-L1 and tumor mutation burden that looks at the amount of mutations in the cell, but in general, they don't seem to be quite as reliable at predicting how well a patient will respond. We have patients who have high PD-L1 who don't respond and low PD-L1 who do respond. Are we looking for better biomarkers for immunotherapy? Yeah. So, you know, I think we're getting a little bit spoiled, I think, from having 70, 80% response rates with these genetic driver mutations. So I think that's the ideal. You know, if you have an EGFR7 mutation, you have a 70, 80% chance of benefiting. And then if you don't, then you're unlikely to benefit. But you're right. I mean, right now we don't have great tests for immunotherapy, and there's a lot of activity going on to try to identify better ones. I think one conclusion right now is it's just going to be more complicated than we anticipated. But I think the ultimate goal is to try to get up to the 70, 80% predictability rates. Okay. So, you mentioned a little bit about drug combinations. Drug combinations seem to be one of our best hopes for overcoming this acquired resistance. But if you try to put all the different drugs together, there are thousands of combinations. How do we determine whether or not a particular combination is our best bet before going to a clinical trial? You're talking for an individual patient? Yeah. I mean, I think that's where... Or even in general, for the population. Yeah. I think some of it is rational, and then some of it is empiric. So rational meaning the more we understand about the biology of the tumor and how the tumor can so-called escape, then we can do more rational combinations. An example of that is in EGFR mutant lung cancer, about 20% of the time you get amplification of another gene called MET, and then you no longer need... You sort of bypass signaling from EGFR. So if you know that ahead of time, then you can say, hey, I need to have an EGFR and a MET combination. And indeed, people have shown that that combination can overcome resistance. But I would say there's other ones, especially with cancer immunotherapy, where a combination with a different kind of TKI has been shown to be super beneficial, not in lung cancer, but in renal cell carcinoma, where I think it would have been hard to predict which combination would have been the most effective. So I think we need more rational, biologically driven combinations, and then we also need to unfortunately still do some empiric testing to figure out what maybe some unexpected combinations that we didn't think about. I think what you're asking, though, is how do you sift through a thousand combinations versus having the most rational ones? I think there, you know, it really depends on the science and how much rational understanding there is of the tumor and what you're trying to treat. Okay. We mentioned a little bit about commutations, and there's also been some recent research that finds that for several different genomic drivers, we seem to be having resistance popping up in the same pathways. What kind of research is going on with relationship to these commutations or acquired new mutations? Yeah, I mean, I think it's, again, there's a lot of active research going on. I think it all requires tissue biopsies or tumor biopsies and then really understanding what can occur. I think the most important thing is, you know, how do you also have a database where you can keep track of and collect all of these kinds of data? Because if you start to do one-offs and one patient has this, one patient has that, and then no one's sort of putting that all together, even if you found that commutation, you might not know how to treat it. I think we also could do, and I think the abrogate community could also do a great job on how do we figure out, how do we pool all this information so that if a new patient has such and such mutation, you could just say, hey, in the database, has anyone had this commutation before and what did they get and did it benefit them or not? I think that would also improve our ability to do precision medicine. Well, you led me right into my next question here. Some of the patient groups that are focused on specific biomarkers, like the EGFR resistors, ALK positive, the Ross Wonders, are considering ways to collect their real-world data so that they can help accelerate research. In your view, what sort of questions in real-world data would be most useful for us to gather? Yeah. Yeah, that's a great question. So, you know, there is this privacy, data privacy issue. So I'm assuming that if you collect that data, somehow the data privacy issue is addressed. We would be doing it using an IRB-approved question. Yeah. Okay. Yeah. Yeah. So I think, you know, what people are really interested in, you know, researchers, pharmaceutical companies, et cetera, is obviously how does a patient do? So then the data has to just be very, very clean, so to speak. So, you know, we need to know how long patients were on therapy. So the starting date, you know, when did they stop that therapy, what actually happened on that therapy? Did the tumor grow? Did it shrink? Did it stay the same? All these kinds of variables, in addition to, you know, gender, age, smoking history, et cetera. I would say it's not 100% easy because, you know, there's a company called Flatiron Health, which is actually owned by Roche, but they actually pay an army of nurse practitioners to clean the data. Cleaning the data means putting it in a structured format so that it's all the same and searchable. So I would encourage maybe the advocates to come up, and I'm happy to put you in touch with the Flatiron folks, too, you know, where you can somehow say these are the key pieces of data, and then when we put it in, it will always be in the same structured format so that when you actually go to do a search, it's easy to find. You know, medicine is complicated because there's multiple terms for the same thing, but if people are using five different terms for the same thing in five different databases, then it makes it very hard for them to be comparable or useful. Okay, we have just a little time left, so I'm going to spring a question on you that you may not have thought about. What are, besides providing real-world data, what are some other ways in which patients can help researchers accelerate research? How can we partner with you? Well, that is a great question. I mean, patient centricity is key to drug development. So first is really understanding what the needs of patients are, and really, you know, teaching the scientists not only what's going on in the tumor, for example, but also what's clinically important and how can a medicine actually be beneficial. So that's really important to convey to the scientists and everyone working on drug development. And it even comes down to side effects and other things, right? What's tolerable, what's not, what is ultimately going to be a benefit? That's one. I think the second one, as I highlighted before in the talk, was just somehow allowing for access to tissue. Hopefully with liquid biopsies, it'll be easier, but you know, sometimes we still will probably require tumor biopsies. And it's just invaluable in terms of giving insights as to what's happening on those treatments. Still, you know, now when new therapies get approved and then patients get them, still sometimes I hear that, you know, patients don't want to go under biopsy after their tumors progress because there's not necessarily another option that's offered to them. But then if you don't get that tumor biopsy, then we can't understand what's happening and then try to develop the next medicine. But I do understand it's a big burden to have to go through this biopsy, so not everybody can do that. But they are really helpful when patients can do that. So those are a couple. I was going to say, are there other opportunities for patient research advocates who are becoming knowledgeable about research and the science to work with industry to be able to help make protocols more patient-friendly or give you ideas of the questions that are important to patients? And if someone were interested in that, how might they go about it? Yeah, so first of all, I mean, you know, in the NIH and with granting agencies, I think patient advocates are also playing a huge role in terms of trying to help guide scientists into what makes sense from a patient perspective in terms of the trials that are being proposed. But I would also agree with you that even in pharmaceutical companies, it's important for us to consider, you know, what you said, patient-friendly trials, right? I mean, there's some times when you need a trial where you need more tissue or more blood sampling and things, but there are other times when patient centricity and really, you know, making it as easy for the patient as possible is really critical. So I would encourage all of you to, you know, I think most pharmaceutical companies also take into account patient advocacy groups and appreciate the feedback that's given on various trials. You know, I think one of the things I would tell you is that a lot of times the people that are writing those trials, they haven't actually had to sit in the clinic and walk through all the things that we asked them to do. So even explaining that, I think is really important. And another point I would say is, you know, when I was a fellow, consent protocols were like for the patients were like three pages. And my understanding is they're now like 20 pages or something like that. So, you know, I think it goes both ways. There are probably regulatory agencies and IRBs and things that are also requiring more, but you know, but if we can simplify the whole process, I think it would make it easier for everybody. Well, that's all the time we have. Thank you, William, for your time and for your clear answers. There are many questions that we didn't get to, and the STARS program will make an effort to collect those questions and post them on the website. Thank you for joining me today. And to our audience, please keep an eye out for an email that will be sent with program evaluation later today. And thank you. And that's the end of this webinar. Thanks so much.

precision medicine

genetic mutations

targeted therapies

EGFR mutations

gefitinib

osmirnib

immunotherapy

biomarkers

liquid biopsies