ROS1 is interesting in that it shares significant homology with 

. They're close to 70 to 80% homologous proteins, if you will. 

 was discovered in the mid-1980s. And to my knowledge, I don't know that we understand the normal functioning of 

 nor do we understand what the ligands for this receptor may be. Again, about 10 years ago, it was discovered in non-small cell lung cancer, [NSCLC] that about 1% of patients had a 

 fusion. Remarkably similar impact like we've talked about with 

 in that this with the fusion becomes an oncogenic driver. It's ligand-independent, even though we don't know what the ligands are. It seems to be independent of a ligand and turns on the growth pathway and survival pathways, giving the cancer cell an advantage over the normal cells. Of course, this also has been a target for the development of drugs, and we have a couple of drugs approved in this setting for specific use in 

-positive patients. And again, should be part of your comprehensive genomic testing platform in advanced non-small cell lung cancer.

 fusions are present in about 1-2% of non-small cell lung cancer. These are gene fusions or chromosomal rearrangements, most common partners CD74. That accounts for about 40% of them. There are a lot of others that occur a little less frequently, ECR, SLC-34A1, STC4, there are at least 10 others. The significance of the fusion partner clinically isn't known. It'll take a little while to uncover those differences given the rarity of these specific fusions. As of today, it does not influence our treatment decisions.

: Every patient that walks in my door should – in my clinic is tested for 

 rearrangements. There are challenges to this. We know that

-rearranged lung cancer is roughly 1 to 2% of all lung adenocarcinoma. That said, when we identify these rearrangements, we can wed them to targeted therapies. There are 2 now that are approved, and potentially a 3rd coming down the pike. The best way to interrogate for 

 rearrangements is probably through a RNA-based next-generation sequencing platform. We order tissue analysis for both DNA and RNA, specifically RNA to identify these fusions. We also order a liquid biopsy. Again, liquid biopsies are probably not the optimal way as a stand-alone test to detect fusions. But supplement to tissue may be helpful. There are gaps in testing for sure. We know that based on ASCO [American Society of Clinical Oncology meeting] data in 2021 that we are not doing the best job of next-generation sequencing for our patients with non-small cell lung cancer. And we need to understand what the best platforms are to interrogate fusions. And in that regard, both DNA and RNA, specifically RNA-based sequencing platforms are important. And importantly doing both tissue and plasma on our patients. A liquid biopsy and the tissue biopsy, both can enhance the capture of alterations that are potentially targetable. We need to address these gaps. We need to do better. But I’m confident that over the next 3 to 5 years – there’s just too many good therapies now – that all of us will be doing a good job of interrogating lung cancer tumors to identify all actionable mutations including 

Videos

Ben Levy, Mark Socinski, and Stephen Liu explain the role of a ROS1 alteration in NSCLC and its value as a targetable biomarker.

ROS1 Alterations in Advanced NSCLC

There are multiple standard-of-care therapies for patients with advanced ALK-rearranged lung cancer. Crizotinib, the first, was developed as a MET inhibitor initially, but it had activity in ALK-rearranged lung cancer. When you compared crizotinib to chemotherapy, there was an improvement in progression-free survival [PFS] and intracranial-response rates for those who received crizotinib as compared with patients given chemotherapy. Thus, crizotinib has been the control arm of every subsequent trial that has solidified a next-generation ALK-directed therapy. Multiple trials have compared next-generation ALK-directed therapies with crizotinib. Of course, the ALEX trial compared alectinib with crizotinib and showed an improvement in PFS. Trends toward overall survival [OS] are also there. Because the ALEX trial came first, alectinib is considered a standard of care, and it is my standard of care for most patients with ALK-rearranged lung cancer. The ALTA-1L trial then compared brigatinib with crizotinib and showed similar outcomes. The data are not as mature as are those for the ALEX trial, but they show improvements in PFS and intracranial response. We’re following the OS and, again, trends toward OS. Because of that, brigatinib is also considered a front-line standard of care. The most recent data is from the CROWN study, which compared lorlatinib, another next-generation ALK-directed therapy, with crizotinib. Not surprisingly, the results showed an improvement in PFS. All of these trials showed improvements in response rates and in intracranial response rates, so lorlatinib is another potential option. These drugs have different toxicities. Because the ALEX trial offers the most mature data, and because I’m a creature of habit, I have generally used alectinib front line for most of my patients with ALK-rearranged lung cancer, but there are other options on the table. Right now, the front line data that we have is with alectinib as seen in the ALEX trial, with brigatinib in the ALTA-1L trial, and with lorlatinib in the CROWN trial. All are potential therapeutic options. Again, alectinib seems to be my go-to first drug for patients with ALK-rearranged lung cancer.

Over the past several years, data sets addressing optimal therapy in the ALK-positive population have emerged. We have several studies with alectinib in different populations—the global ALEX trial, J-ALEX trial in Japan, and the ALESIA trial in China. All of these trials compared alectinib with crizotinib therapy and showed superiority of the primary outcome, which was largely PFS. We had a recent update on OS from the J-ALEX trial, which probably had the longest follow-up. The problem with interpreting the OS, which was not different between the crizotinib and alectinib arms, was that there was a remarkably high crossover rate to alectinib from the crizotinib arm. More than 70% of patients crossed over to alectinib, which may blur a survival advantage. When you look at the data from the ALEX trial, which did not allow crossover from the crizotinib arm, there is superior OS. In fact, the median survival on the alectinib arm has not yet been reached in follow up, and many patients are beyond 5 years. This is a population of patients that may live for years when diagnosed and treated appropriately. Following those alectinib trials, we have the ALTA-1L trial and the J-ALTA trial in Japan. They were looking at brigatinib, again comparing it with crizotinib as the control arm. And once again, crizotinib use produced inferior results. The primary end point was PFS. Very similar results were reported in ALTA-1L trial, and a small cohort in the J-ALTA trial showed the superiority of brigatinib over crizotinib. The most recent data set that was added to this experience was the CROWN trial, which compared lorlatinib with crizotinib. Again, the primary end point of PFS was superior with lorlatinib versus crizotinib. All 3 of these agents have beaten our former standard of care, crizotinib, when compared head-to-head. We need to give crizotinib some relief, because it was never intended to be an ALK inhibitor. It came into clinical early phase 1 trials as a MET inhibitor. True ALK inhibitors that have more specificity and potency for ALK are superior to the old standard of crizotinib. Interestingly, the 3 agents we've spoken about were initially proven in patients who had failed crizotinib. But now all 3 of them have shown superiority over crizotinib. Which is the best one to use in this setting now that we have 3 positive trials with these agents? It's very difficult to do cross-trial comparisons. The variant of ALK makes a difference. Co-mutations, such as P53, can make a difference in the survival outcomes. We may not know all of those specifics from all trials, which makes cross-trial comparisons somewhat treacherous in this setting. And toxicity profiles differ. They all target ALK and have similar efficacy end points, but they have different toxicity profiles, which demands consideration. For years, our standard, based on the results of the J-ALEX and the ALEX studies, has been alectinib. Personally, data from either the ALTA-1L trial or the CROWN trial would not change my first choice, which has been alectinib, in this population. All of these agents are highly active, and one can't be faulted for using any one of them in this setting. Tailoring the decision on the adverse effect profile instead of efficacy may be legitimate. 

A particular issue in the ALK-positive population of patients with NSCLC is the propensity for CNS [central nervous system] involvement at the time of diagnosis. We reviewed the results of the BFAST trial, in which about 40% of patients had brain metastases at the time of initial diagnosis; however, there is certainly a risk of developing brain metastases over time. Alectinib, brigatinib, and lorlatinib have quite good CNS penetration and CNS activity. The response rates are quite good. Response rates to alectinib in patients with brain metastases in the BFAST trial was in the 90% range, and that's also true for the other agents. CNS progression, particularly when it's symptomatic, can be devastating in certain patients. Importantly for radiation oncologists if brain metastases are diagnosed, the knee-jerk reflex often is to radiate them. These brain metastases are often asymptomatic. Initially, we can spare the patient the toxicity of brain radiation, particularly whole-brain radiation. The ALK-positive population tends to be a bit younger, and the median age is typically in the mid- to early-50s. These patients are typically never-smokers and otherwise healthy, so you don't want to expose them to the risk of CNS radiation toxicity unless necessary before they really need radiation. And many of these brain metastases can be controlled. When second-generation ALK inhibitors are compared specifically with crizotinib, which has much less CNS activity and penetration, the CNS progression rates are dramatically lower than found with crizotinib or chemotherapy in this population.

For ALK fusion-positive lung cancer, it really is an embarrassment of riches. We have a lot of very active, very effective, and well tolerated drugs available. Alectinib, brigatinib, and lorlatinib all have shown superior results when compared with crizotinib in randomized trials, but they do have some differences in terms of efficacy, safety, and convenience. They all are very effective with high response rates. They do have a lot in common. When compared with crizotinib, they all have been associated with better PFS. All are highly CNS-penetrating, which is particularly important in ALK fusion-positive lung cancer because of the high tropism for brain metastases with this subset. Use of any of these agents generally produces PFS and the durability of response of at least 3 years, on average, before initial resistance occurs. Certainly, there are differences between them. When comparing their PFS with that of crizotinib, the HR by investigator assessment for alectinib was 0.43 in the ALEX trial and for brigatinib was 0.43 in the ALTA-A1L trial; for lorlatinib, it was quite a bit lower at 0.28. There's no question that lorlatinib is a highly effective drug, possibly because it may have more CNS penetration. But this may be one of the rare cases in oncology where we choose not to use our most potent drug first, and that reflects the differences in toxicities. Lorlatinib can raise cholesterol [levels], which is usually straightforward but still can be a challenge for an oncologist. The toxicity of concern is the impact on cognition and neurologic function. It may cause mild and temporary dizziness or moodiness or something as serious as seizures, hallucinations, depression, or personality changes. I've seen a wide range of cognitive impact from lorlatinib therapy. Patients treated with lorlatinib must be watched closely, because the effects can be mitigated with dose reduction if they're recognized early. But sometimes they're subtle. In a busy clinic, we may not get to spend a lot of time with the patient, and we may miss that they're no longer able to balance their checkbook, that they get lost when they're trying to go home, or that they're wandering on the streets. If patients don't have a strong support system or people around them who can identify or report these changes, it can be concerning. When we look at alectinib, it's a very well-tolerated drug. Some patients notice myalgias, fatigue, and, sometimes, constipation. Liver function-test abnormalities are a little common. Brigatinib is also extremely well tolerated. It famously has a low incidence of early-onset pulmonary events in which the DLCO [diffusing capacity for carbon monoxide] transiently drops. Interestingly, these events self-resolve with continued dosing, and a lot of patients may not really notice them. With the current approved dosing of 90 mg given once daily for a week followed by 180 mg given daily, they're actually pretty rare. Overall, the drug is very well tolerated. There also are a few practical, nonclinical factors to consider. Alectinib therapy involves 8 pills each day, and brigatinib therapy involves 1 pill each day. Overall, however, they're both very effective, highly well-tolerated agents. And we're lucky to have all of them. I really don't think there's a wrong answer among all of those.

With so many options available for advanced ALK fusion-positive NSCLC, how do we choose among alectinib, brigatinib, or lorlatinib? Patients need to be watched a little more closely with different surveillance strategies, because I have some concerns about cognitive toxicity with lorlatinib therapy. I've seen some severe cases in the past. I'm probably a little less likely to use lorlatinib in the first-line setting, although I might consider it if there's a particularly heavy CNS burden. Generally, I'm looking to alectinib and brigatinib. When we look at patients with ALK-positive lung cancer, the PFS HR ratio for lorlatinib was the lowest among the 3 agents. The survival is much, much longer than any one drug's PFS on those studies, suggesting that patients really should receive multiple lines of therapy. Therefore, I'm pretty comfortable reserving a drug like lorlatinib for later use. Between alectinib and brigatinib, both drugs are excellent and very effective, offer fast responses, and are extremely well tolerated. Really, it's a coin flip. Some nonclinical factors that could be considered include pill burden and cost, and that's something that's unpredictable. If there are differences in out-of-pocket cost for a patient based on their insurance coverage, copays, or formularies, then that probably will sway me, because both are similar and effective. Looking forward, we must see if there are differences in terms of resistance to these drugs and how that might influence sequencing. But as of today, all of these are appropriate options

It’s easy to decide what to do with a patient with advanced stage ALK-rearranged lung cancer. We have multiple options, and we generally start with alectinib in my practice. But what do you do when there’s disease progression on alectinib? And the progression is heterogeneous. There can be oligoprogression where there’s just one site that’s growing, there can be progression only intracranially, or there can be systemic progression where you really need to think about a therapeutic switch. And that’s where we really talk about resistance and what’s going on. There are multiple ways that resistance occurs in patients with ALK-rearranged lung cancer. First, you can have amplification of the ALK fusion that can confer resistance to the drug you've given. Second, bypass signaling pathways, other pathways like EGFR or c-Kit, can be engaged and cause resistance. Perhaps one of the areas that’s being most studied as a mechanism of resistance is mutations in the ALK-kinase domain. And these occur following alectinib treatment. We’re starting to gather data on how common they are. One mutation that we look at is G1202R. And this is important, because this mutation may be targetable by a next-generation ALK-directed therapy, lorlatinib. At the time of progression, I generally test for mechanisms of resistance. We look for these secondary mutations in the ALK-kinase domain. And if there is a G1202R or other mutation, we generally will leverage lorlatinib. What to do for patients that don’t have mutations? That is where things get a little bit trickier. Chemotherapy or lorlatinib could be leveraged. We’re just beginning to scratch the surface on how to sequence therapies for ALK-rearranged lung cancer and what to do at the time of resistance.

Ben Levy, Mark Socinski, and Stephen Liu describe the current treatment options for patients with ALK+ NSCLC. They also summarize recent clinical data for ALK inhibitors.

Clinical Management of ALK+ Advanced NSCLC

ALK was initially discovered around 1994 in a patient with large-cell lymphoma. It subsequently was discovered in patients with lung cancer by Japanese investigators in 2007. In terms of normal development, what does ALK normally do? It is believed to play a pivotal role in cellular communication and in the normal development and function of the nervous system, and that's derived from information on animal models in this setting. It's been an interesting journey from that initial description in large cell lymphoma to being part of our standard thinking and our standard evaluation in non–small cell lung cancer (NSCLC).

 If we think about the role that ALK plays in lung cancer development, we know that it is an oncogene. In lung cancer, we have chromosomal rearrangements or gene fusions that can transform a normal cell into a cancer. These fusions typically retain the kinase domain of ALK, and they link to the amino-terminal portion of various partners, most commonly EML4. We're not sure [about] the significance of these partners. It may be important, but it doesn't impact treatment today. If we look at something like splice variants just within EML4-ALK, there are at least 15 different variants that are named in the order that they were described. The most common is variant 1, and the next most common is variant 3. An examination of those two variants, which [represent] 40% to 45% of the EML4 fusions, shows their differences. Variant 3 is associated with a much worse prognosis and very different resistance patterns. About one third of patients with variant 3 have the G1202R point mutation that confers resistance to a lot of TKIs [tyrosine kinase inhibitors]. You don't see such resistance in patients with variant 1, so there are differences in these ALK fusions. Today, they don't guide our therapy; in the future, I'm willing to bet that they will.

 ALK fusion should be a standard part of the initial evaluation of any patient with advanced-stage NSCLC. The presence of an ALK fusion allows you to proceed without directed therapy, which is highly effective. I'll talk about that in a moment. We test for ALK, both with IHC [immunohistochemistry] and next-generation sequencing. This is one of the examples in which IHC is an accepted testing method for this biomarker. Like many of these targeted fusions and oncogenic drivers, if you find one, you tend not to find any other driver mutations, except for rare situations where you may have 2. I've had a couple of cases where you see both EGFR mutations and ALK fusions. For the most part, they're mutually exclusive. And if you find one, you can act on that and go down that pathway of treatment. We know we have a number of different ALK agents, the first one being crizotinib. And then the second-generation drugs include agents like alectinib, ceritinib, brigatinib, and lorlatinib in this setting. In my practice, it's standard to test plasma for these genetic alterations, and that leads me to discuss the BFAST trial. This ongoing, multicenter clinical trial is examining the use of next-generation sequencing in the blood, the identification of molecular alterations, and the outcomes after the patient is relegated to a specific therapy. There are multiple cohorts in the BFAST trial, and we’ve recently seen some data for the ALK-positive cohort. In the BFAST trial at the time of this reporting, there are about 2200 patients who were screened; 119 were found to be ALK-positive. That's a rate of about 5.4%, which is what we would expect in this setting. These patients were treated with alectinib at the standard dose of 600 mg given twice daily. EML4 was the most common fusion partner at 84% of the time. About 44% of the ALK population had a co-mutation with P53. Interestingly, the tumor mutation burden was, at a median, only 2. This is a low TMB [tumor mutational burden] population. Now, again, these were identified by plasma-based testing. With use of alectinib, the overall confirmed response rate was 87% by investigators and 92% by independent radiologic review. At 1 year, 75% of those patients were still in response. About 40% of patients at baseline had asymptomatic brain metastases, which is an issue with the ALK-positive population. The response rate in the measurable brain metastasis population at baseline was 91%. This underscores the clinical usefulness of plasma-based testing. Again, in my practice, I do both tissue and plasma testing. But this data from BFAST shows that you can detect ALK at a similar frequency in blood as you do in tissue, and the outcomes with alectinib therapy in this experience were quite good. This reinforces the clinical usefulness of plasma-based next-generation sequencing.

 The prognosis for patients with ALK-rearranged lung cancer is favorable. It is quite different than for patients with other mutations that we see in lung cancer. That underscores how important it is to test for these fusions/rearrangements. We have a lot of data now. The updated overall survival analysis from the ALEX trial, which compared use of front-line alectinib versus crizotinib in patients with advanced ALK-rearranged lung cancer, shows that more than 60% of patients are still alive at 5 years. This is remarkable in the context of what we generally saw maybe 10 years ago, which was a median survival of only 12 months for patients with stage IV lung cancer. We have now moved to 5-year survival rates of more than 50%. Again, this underscores the importance of testing for the rearrangement. You can’t give the drug unless you test. Patients with ALK-rearranged lung cancer generally present with advanced-stage disease, like many of our patients who are driver mutation-positive. A malignancy with this mutation or fusion has a proclivity to go to the CNS [central nervous system] or brain. Many patients present with brain metastases. Importantly, the drugs that we leverage in this setting can cross the blood-brain barrier and elicit intracranial responses in the brain. For all these reasons—good drugs, better outcomes, unprecedented survival times—we need to identify these fusions up front and give the right targeted therapy. And in this case, we have many targeted therapies for our patients with ALK-rearranged lung cancer.

Ben Levy, Mark Socinski, and Stephen Liu explain the role of an ALK alteration in NSCLC and its value as a targetable biomarker.

ALK Alterations in Advanced NSCLC

 In non–small cell lung cancer (NSCLC), we have several DNA alterations. Some are mutations, and some are gene rearrangements or fusions with a chromosomal rearrangement that creates a driver fusion. We also have amplifications in certain situations that are potentially important, so the testing platform needs to have the ability to detect all sorts of alterations, because it's not simply one. In some instances, as in ALK disease, there is an accepted immunohistochemical, or IHC, test. Again, that is looking at protein expression, not DNA or RNA expression. In understanding the heterogeneity of what you're looking for in different situations, mutations are different than fusions, and protein expression is different than either DNA or RNA alteration. Techniques such as PCR [polymerase chain reaction] or FISH [fluorescence in situ hybridization] are different than next-generation sequencing for these sorts of things. In my practice, my standard of care is a next-generation sequencing platform that does both DNA and RNA analysis to maximize the detection of these alterations, hopefully at the time of initial diagnosis.

 When you think of what tests to use to detect these fusions, it's important to be familiar with the process. There are a few different ways to identify gene fusions or chromosomal rearrangements, which refer to the same thing. In the past, we have relied on the process called FISH, which is a bit labor-intensive and involves visually identifying an abnormal spatial relationship between 2 areas in the gene. There are a few of these assays that are FDA approved, but they have imperfect sensitivity. We can also look at protein expression. For example, ALK IHC is approved by the FDA as a rapid, fairly straightforward test to detect an actionable ALK alteration. Immunohistochemistry is a bit less validated for others, like ROS1 or NTRK, where positive IHC results probably need to be confirmed with next-generation sequencing. In general, I strongly recommend multiplex testing and the use of next-generation sequencing for all patients. We're well past the tipping point where this is cost-effective. It provides more efficient use of the tissue at hand and a better use of time. If we're using all of these tests up front, we're much less likely to run out of sample before we perform all of the necessary tests. There are several studies that now show it's more cost-efficient to order multiplex testing upfront. It saves time, and, overall, it is just a better test. Many commercial platforms of next-generation sequencing are DNA-based. Some are RNA-based, and there is a potential advantage to using RNA sequencing for some targets, particularly those with large introns or complex events (eg, NTRK, NRG1, ROS1). RNA sequencing has a better sensitivity. It's important that we understand the tests we're ordering, because the stakes are quite high. If we miss one of these drivers, not only do we deprive a patient of a highly effective, tolerable treatment, but we send them down the route of immunotherapy, which will be ineffective and will make subsequent targeted therapy much more dangerous. When a biomarker is present, it opens a whole catalog of different treatment options. It's important we don't deprive patients of that benefit. The most important part of treating advanced NSCLC is proper upfront biomarker testing. I can't stress how important it is to do full comprehensive genomic testing with next-generation sequencing, ideally with RNA-based sequencing, because it is so important for planning therapy. If a driver ulceration is undetected, not only are we going to deprive someone of an effective targeted drug, but we're going to go down the route of immunotherapy, which is ineffective and which makes subsequent targeted therapy much more dangerous. An important tenet of treating lung cancer is identifying what those alterations are up front, right at the beginning. And we're not nearly as good as we think. When we look at results from the Molecularly Informed Lung Cancer Treatment in a Community Cancer Network: A Pragmatic Consortium (MYLUNG) data that was presented at [the 2021 American Society of Clinical Oncology meeting], we find a low bar of EFGR, ALK, ROS1, BRAF, and PD-L1. These 5 markers are established and are straightforward. We have known about them for years, and we have approved agents available to us on the shelf to target them. With that low bar of 5 targets, how often are patients with advanced lung cancer getting tested? In a strong robust clinical network, like the US Oncology Network, it was less than half. That is a low bar using a test like EGFR, which is pretty straightforward. EGFR is a marker that we've known about for coming on 2 decades, yet 1 in 5 patients are not getting tested for it. As good as we are at testing, as aware as we think we are, we are missing patients. And if we're missing patients, that means we are not treating them appropriately. There’s a lot of room for improvement. We can certainly do a lot better.

Gene rearrangements are an important genomic driver in NSCLC, specifically adenocarcinoma. There are now approved therapies for ALK rearrangements, RET rearrangements, ROS rearrangements, and NTRK rearrangements. It’s important that we test for these fusions in our patient population. There are probably best platforms or best next-generation sequencing platforms to do this. By using both DNA and RNA, by leveraging DNA and RNA next-generation sequencing platforms, we’re better able to identify these fusions. And with that said, every patient who walks through the door in my office with an advanced adenocarcinoma of the lung gets tested for these gene rearrangement fusions. This is part of a larger panel that we do. ALK, ROS, RET, and NTRK are rarer than other types of mutations, like KRAS G12C or EGFR, but they are actionable. They are seen in lung cancer. At the very minimum, patients are tested up front, before we decide on a particular therapy. The question then becomes, “When patients develop disease progression, what do we do?” It is important to test for mechanisms of resistance for patients with these fusions who receive targeted therapy, because we’re learning more and more about novel therapies and testing novel therapies that may target these mechanisms of resistance. We may look at something like an ALK fusion. You can target these fusions with therapies, but there are mutations that occur as a consequence of resistance, and some of these mutations may be targeted with other therapies. And this is the way we’re heading. At the very minimum, I encourage testing upfront for these fusions, specifically with a DNA- and RNA-based platform approach, and then again at the time of resistance to identify these mutations. We may then try to leverage either standard-of-care therapies or therapies that are being tested in clinical trials against these mechanisms of resistance.

 Genomic testing extends beyond the frontline setting. When we think of testing, we're trying to inform our treatment. Cancers continue to evolve, but they don't always evolve in the same way. When we identify a genomic driver up front and then start a targeted therapy, we're putting a selective pressure on that tumor. And these tumors will continue to evolve when the cancer progresses. If we can better understand that progression, then perhaps we can better plan our next lines of therapy and add other targeted agents if we see specific actionable alterations. But these biopsies will also reveal histologic transformation. We would estimate about 15% will have transformation, either small cell or squamous cell, and that would certainly influence what we do next. Understanding how drugs stop working and how cancers evolve is just as important. If a genomic driver is present at the beginning, we start a targeted agent. When the cancer progresses, repeating a biopsy and biomarker testing is critical to the optimal management for that patient.

Ben Levy, Mark Socinski, and Stephen Liu explain the various types of genetic alterations and protein expressions that are targetable in NSCLC and the crucial role of comprehensive genetic testing at the time of diagnosis and at the time of disease progression.

Testing for Gene Alterations and Protein Expression in Advanced NSCLC

Our increased understanding of non–small cell lung cancer (NSCLC) biology has led to the important development of novel therapies and improved outcomes for our patients. This has been the whole personalized therapy story in solid tumor oncology, which has been led by the lung cancer field. We have a better understanding of the genomic underpinnings of the tumor. We can now wed particular targeted therapies to these molecular niches. This has been exciting for physicians, but, more importantly, it’s been a big win for patients. We have data [showing] that if we know the patient’s [genetic] target or mutation and give them a targeted therapy, they do better than if they have a target and don’t receive targeted therapy. But you can’t give genotype-directed therapy unless you know the genotype. We started out with EGFR, but the list of relevant alterations that we can now test for in patients with lung cancer is growing. By defining or identifying these mutations, we can wed patients to therapies. This has led to improved outcomes, quality of life, and the potential ability to circumvent the need for chemotherapy. It’s been incredibly exciting to be in this field. I’m just talking about targeted therapies and haven’t even mentioned immunotherapies. Again, understanding the biology of a tumor also may help leverage specific immunotherapies. Clearly, the story has been about identifying mutations and wedding those mutations to targeted therapies to improve outcomes for our patients with advanced stage lung cancer.

We now have FDA-approved treatments for specific biomarker-defined subsets of lung cancer that include ALK, ROS, RET, and NTRAK. These are important to target the EGFR set, the paradigm. We now have extremely effective and well tolerated agents for ALK [-positive disease]. Crizotinib and ceritinib were superior to chemotherapy, and alectinib, brigatinib, and lorlatinib were all superior to crizotinib. All 5 of these drugs are approved [by the FDA]. The newer agents (alectinib, brigatinib, and lorlatinib) are typically effective for years and are given as pills that you can prescribe via a pharmacy today. These drugs have been fully approved and work, on average, for about 3 years and, for some, much longer. It is quite remarkable, and we have important efforts ongoing to further improve outcomes. [We are] never satisfied with what we have. For ROS1 [-positive tumors], we have 2 drugs that are FDA-approved. Crizotinib is a very versatile agent, and, most recently, entrectinib was approved by the FDA in 2019. Another agent, repotrectinib, or TPX-0005, has been given a breakthrough therapy designation [by the FDA] in early results, [but it is] not yet approved. RET fusion [-positive tumors], which currently represent about 1% to 2% of lung cancer cases, can also be effectively targeted. For many years, we were repurposing drugs off the shelf that were approved for other cancers, like kidney cancer and thyroid cancer. Drugs like cabozantinib and vandetanib have shown modest activity and notable toxicity from a lot of their off-target activity. We have selective inhibitors that are much less toxic, extremely well-tolerated, and highly effective. When used as initial therapy, pralsetinib, an FDA-approved RET inhibitor, produced a response rate of 88% in the first-line setting, which was really remarkable. With a 96% disease control rate, [you can] go in and know with confidence that this drug will work. Selpercatinib, another FDA-approved RET inhibitor, showed a frontline response rate of 85%. These responses are rapid, they're deep, they're very durable, and, in many cases, they last for years. Likewise, for tumors with end-track fusions, we have approval of larotrectinib and entrectinib, another versatile drug. These are FDA approved, have high response rates, and are potent gene inhibitors. What's unique about these drugs is they're both active and are both FDA-approved across different tumor types. 

In NSCLC, there are several unmet needs. Largely, there is the percentage of patients that gets comprehensive genomic testing. It still is far below the plateau that it should be, which is 90-plus percent of patients. The list of either mutations or alterations, amplifications, fusions, or all the alterations we see in NSCLC is growing and, hopefully, will continue to grow. The list of targeted therapies approved by the FDA for specific biomarker-defined populations is also growing. We have just had the first drug for the G12C subset of the KRAS mutations, which is a phenomenal advance in the overall therapeutic options for these patients. All of these really good drugs have a nice therapeutic index, a clinical activity that we typically don't see with chemotherapy or I-O [immuno-oncology] therapy in these populations, [but] none of them can be used unless you make the molecular diagnosis. And we still see that comprehensive genomic testing covering all the actionable alterations is still not at the level where we need it to be. There are lots of different reasons for that in a lung cancer population. For example, it's often not easy to get adequate tissue from this population. There's been underuse of plasma-based next-[generation] sequencing. And there has been some misunderstanding of the pros and cons of different testing platforms in terms of what you can expect. We need to continue educational efforts to stress the importance of comprehensive testing. The standard of care for both accuracy and efficiency is next-generation sequencing. All of my patients with stage IV disease get tissue-based testing and plasma-based testing on an NGS [next-generation sequencing] platform that does both DNA and RNA. The one thing you'd never want to miss in a lung cancer population is a targetable driver alteration, again, because of the activity of these agents.

An increased understanding of driver gene alterations in advanced non–small cell lung cancer (NSCLC) has led to the development of targeted therapies that offer improved patient outcomes. Ben Levy, MD; Mark A. Socinski, MD; and Stephen Liu, MD, share their thoughts on the emerging treatment landscape and remaining unmet needs.

Overview of Targeted Therapies, Improved Outcomes, and Unmet Needs in Advanced NSCLC

Stephen Liu, MD, an associate professor of medicine, director of thoracic oncology, and director of developmental therapeutics at Georgetown Lombardi Comprehensive Cancer Center, discusses the rationale for the ongoing phase 1/2 ARROW trial (NCT03037385) in 

 fusion–positive non–small cell lung cancer (NSCLC). 

 fusion–positive NSCLC tend to have adenocarcinoma, be non-smokers, and have less robust responses to immunotherapy than patients without oncogenic drivers, says Liu. As such, this subset of patients represents a group in need of novel therapies in NSCLC, Liu explains. 

Pralsetinib (Gavreto) was approved by the FDA in September 2020 for the treatment of adult patients with metastatic 

 fusion–positive NSCLC based on findings from the ARROW trial, in which pralsetinib induced a 57% overall response rate and a 5.7% complete response rate in this patient population. Updated findings from the ARROW trial, which were presented during the 2021 ASCO Annual Meeting, reported that pralsetinib continued to demonstrate robust and durable antitumor activity in patients with 

 fusion–positive NSCLC, as well as other advanced solid tumor types. 

Notably, the addition of pralsetinib to the NSCLC armamentarium offers patients with 

 fusions a highly selective and effective agent that is well tolerated, concludes Liu. 

Stephen Liu, MD, discusses the rationale for the ongoing phase 1/2 ARROW trial in RET fusion–positive non–small cell lung cancer.



Dr. Liu on the Rationale for the ARROW Trial in RET Fusion+ NSCLC

Stephen Liu, MD, an associate professor of medicine, director of thoracic oncology, and director of developmental therapeutics at Georgetown Lombardi Comprehensive Cancer Center, discusses common 

 fusion partners in patients with non–small cell lung cancer (NSCLC).

The most common fusion partner in NSCLC is CD74, as indicated by multiple studies, Liu says. Additionally, several other fusion partners have been identified such as 

. Previous findings have reported almost 20 

Looking forward, experts in the field are seeking to understand the impact of fusion partners on the underlying disease biology and if they could guide treatment decisions. Currently, this impact is not understood, Liu concludes.

Stephen Liu, MD, discusses common NRG1 fusion partners in patients with non–small cell lung cancer.

Dr. Liu on Common NRG1 Fusion Partners in NSCLC

Stephen Liu, MD, an associate professor of medicine, director of thoracic oncology, and director of developmental therapeutics at Georgetown Lombardi Comprehensive Cancer Center, discusses the rationale for examining afatinib (Gilotrif) in patients with non–small cell lung cancer (NSCLC) harboring 

 fusions are different from other fusion events that occur in patients with NSCLC, as they don’t lead to a constitutively active kinase, Liu says. NRG1 is the ligand for the transmembrane receptor. The fusion brings the EGF-like domain of NRG1 into close proximity with its receptor, which is either ErbB3 or HER3. HER3 then tends to heterodimerize with HER2, although it could also do so with HER4, HER1 or EGFR, Liu explains.

Current strategies are focused on targeting receptors, as effective agents are already available, Liu notes. Afatinib is an orally bioavailable pan ErbB kinase that was developed to inhibit the abnormal signaling downstream of 

Stephen Liu, MD, discusses the rationale for examining afatinib in patients with non–small cell lung cancer harboring NRG1 fusions.

Dr. Liu on the Rationale for Examining Afatinib in NSCLC Harboring NRG1 Fusions

Stephen Liu, MD, an associate professor of medicine, director of thoracic oncology, and director of developmental therapeutics at Georgetown Lombardi Comprehensive Cancer Center, discusses emerging research on antibody–drug conjugates (ADCs) in non–small cell lung cancer (NSCLC) that were presented during the 

2020 World Conference on Lung Cancer (WCLC)

Several important areas of research were covered during the virtual meeting, according to Liu, with findings on investigational ADCs being of particular interest, Liu says. Last year, data read out with fam-trastuzumab deruxtecan-nxki (Enhertu), which was shown to elicit a response rate of 61.9% in patients with HER2-positive NSCLC, according to the results of the DESTINY-Lung01 trial. At this year's meeting, additional results evaluating the agent in patients with 

 overexpression were shared, and the agent was found to demonstrate antitumor activity irrespective of expression levels.

Additionally, data regarding the Trop-2–targeted ADC datopotamab deruxtecan (DS-1062) were also revealed. Preliminary results showed that the agent elicited an overall response rate of 23% when given at a dose of 4 mg/kg, 21% at 6 mg/kg, and 25% at 8 mg/kg per blinded independent central review. As Trop-2 is frequently overexpressed in patients with NSCLC, this is a potentially efficacious option that may impact other indications in the space, Liu concludes.

Stephen Liu, MD, discusses emerging research on antibody–drug conjugate in non–small cell lung cancer that were presented during the 2020 World Conference on Lung Cancer.



Stephen Liu, MD

Articles