Expert Perspectives in the Evolving ALK+ NSCLC Paradigm

In a wide-ranging discussion, Alexander Drilon, MD; Julia Rotow, MD; Thomas E. Stinchcombe, MD; and Ashish Saxena, MD, PhD, provide expert insight on the latest groundbreaking developments in the evolving treatment landscape for patients with ALK-positive non—small cell lung cancer (NSCLC).

OncLive^®: In the NSCLC landscape, what are the biomarkers that are routinely tested and what technology is currently being implemented for testing?

Alexander Drilon, MD

Drilon: Biomarker testing is critical in non—small cell lung cancers and in patients who have advanced disease. In the past, we only knew about sensitizing EGFR mutations and recurrent gene fusions, including ALK. Now we have several other mutations, fusions, and copy number changes that are actionable, as we say, meaning that patients can be matched and [receive] active targeted therapy; either 1 that’s approved by the FDA or other regulatory agencies or maybe 1 that’s in the NCCN [National Comprehensive Cancer Network] guidelines or already in clinical trials. Examples of these are BRAF V600E mutations, for which the dabrafenib/trametinib combination is approved, and RET fusions. We saw interesting data on selective RET inhibitors, MET exon 14 alterations, [and] NTRK fusions, for which we have a tumor agnostic approval for larotrectinib and entrectinib.

The baseline message here is that we can’t get away with simple gene testing anymore, and it becomes much more practical and cheaper to do one big comprehensive test and investigate all these other drivers up front whenever possible.

Stinchcombe:When you look across the lung cancer landscape, we’re having more and more targeted therapy. We now have ALK, EGFR, ROS1, which are the big 3, [and] we added BRAF and we also have NTRK. I anticipate we’re heading for a drug for BRAF rearrangements coming up. If you look across the landscape,…the focus should be on broad-based testing and really making sure that all patients who are released with adenocarcinoma or nonsquamous cytology get tested with a broad platform. The number of genes that we have to [test] now makes the single sequential testing, or several single genes, incomplete.

The thing that…facilitated [broad platform testing] is the use of ctDNA [circulating tumor DNA]. Recently, ctDNA [testing] was compared with standard routine tumor testing in the NILE trial [NCT03615443].…I think it really showed…that the ctDNA [testing] was similarly effective, but the turnaround time was much shorter, around 7 days.…That’s very important when we’re dealing with patients who are very symptomatic and who want to start treatment yesterday.

Rotow: For patients with a new diagnosis of lung cancer, biomarker testing up front is critical to select the best treatment option in the metastatic setting. This includes the PD-L1 marker, but also NGS [next-generation sequencing] testing, if possible, looking for targetable oncogenic driver mutations that can be done as a smaller panel, looking for the big 4 targetable mutations. If at all possible, [it] is best…done as part of a broader NGS panel, so you can pick up on some of these less common or emerging mutations, including RET and TRK, HER2, and MET, which may have options…approved off label or in clinical trials. I think finding these mutations for patients is critical to connect them with all their treatment options.

In the ALK space, biomarker testing, apart from the identification of the ALK mutation, currently doesn’t play a role in the indications for any of these drugs. [However,] I think clinically… [biomarker testing] is going to become increasingly important, especially as we have more and more patients who receive multiple different ALK inhibitors and…start to develop compound resistance mutations. Understanding if they have ALK second-line resistance mutations may help us in the future [to] select among treatment options.

There [are] also some interesting data suggesting that in patients with resistance who do not have an ALK mutation there may be an important biomarker [indicating] a lower probability of response to these later-generation ALK inhibitors. They still can be treated with these drugs. Following that and understanding how it influences probability of response would be important.

Saxena: The only way you’ll know if you have an ALK-mutated [NSCLC] is to test for it, and that’s why tests are so crucial for patients that have stage IV NSCLC. They can be done in a number of ways. There can be immunohistochemistry testing sometimes…to see if they detect the protein in the tumor cells. FISH [fluorescence in situ hybridization] testing has been the standard for a while, but now also [NGS is] used generally in the form of larger panels, not just testing for ALK, but testing in a broader array of mutations at once in panels. So those are very important because unless you test for it you won’t actually know that you have an ALK-mutated lung cancer. And now there are so many good options for it, so it is important to test.

Are there any other emerging biomarkers in the space that are promising?

Stinchcombe: I think the real promising biomarkers [are those that specify] what to do at the time of resistance on the next generation ALK TKIs [tyrosine kinase inhibitors].… There [have] been some retrospective studies, some small case series, showing an association with a certain mutation and efficacy with a certain drug. What we’re really looking to do is…a prospective study.…We have to face the fact that some of these patients may have multiple ALK mutations, and as you treat with more ALK inhibitors, you might develop more ALK tyrosine kinase mutations. [Understanding] how to best target the patients with multiple mutations and ALK-independent mechanisms of resistance is a big step forward.

Saxena: There are some data about the particular variants of ALK fusions [indicating] that some of them may predict what types of mutations happen later on…. There are some investigations [to determine] whether knowing [the] specific type of ALK fusion may impact what type of mutation you’ll see going on, which then impacts what drug you might use.

Julia Rotow, MD

Rotow: I think for ALK patients only [the] presence or absence of ALK mutations by tissue genotype testing may come to be important as a biomarker. The presence of an ALK mutation at resistance often will suggest an ALK-dependent mechanism resistance, whereby treatment with an ALK TKI may offer [a] benefit for patients without a secondary, new ALK resistance mutation. There is the thought that they may pretend to bypass pathway activation leading to resistance, which is less well targeted with the new ALK TKIs, although response rates still occur. [However, it] may also suggest the need for future combination therapy strategies targeting to bypass these mechanisms.

Drilon: For ALK fusion—positive lung cancer, a question… is whether or not immune therapy is very active for patients with these cancers. Now, we have known [that], historically, if you look at immunotherapy given alone for ALK fusion—positive lung cancers, the odds of response weren’t great, meaning the chances of having a decent response rate are lower compared with unselected patients. The question becomes, “If we do something like a combination of chemotherapy and immunotherapy, will that improve the likelihood of benefit over immune therapy alone?”

We have some preliminary data from [the] IMpower150 [NCT02366143] study that if you do chemotherapy with an immune checkpoint inhibitor, like atezolizumab, there is a quarter of patients with EGFR and ALK who benefit. Although if you look at the data a little more closely, it seems [they were] probably clustered within the EGFR-mutant lung cancer. It remains to be seen if a big prospective trial will show a definite benefit to doing chemoimmunotherapy for [patients with] ALK fusion—positive lung cancer, but it certainly is a question that we need an answer to soon.

Insight Into Selecting Among the Multiple Treatment Options for Patients With ALK-Rearranged NSCLC

What are the current strategies in the treatment of patients with ALK+ NSCLC? If you can, please discuss up-front and later-line settings.

Drilon: Technically, [patients with] ALK-rearranged lung cancer have many options and turns of systemic therapy. And if they have advanced disease, obviously chemotherapy is an option, [which is] the prior standard of care. And now we actually have a wealth of different ALK-directed TKIs that are available both in the first-line setting or after a patient progresses.

Rotow: It’s a really exciting time in the treatment of ALK-rearranged lung cancer. We have increasing numbers of treatment options for these patients [that] I think are better tolerated, and we should improve activity for controlled brain metastasis, which is very important for patients. Up front right now, we have alectinib, which is approved for the first-line space and I think most widely used. We also have crizotinib and ceritinib, previously approved as well, which have primarily been supplanted by alectinib in current practice. In the later-line setting, we [also] have options, including [the recently approved] lorlatinib, which is a next-generation ALK TKI that can be used for patients who have progressed on prior ALK TKI therapy.

Saxena: In patients [with a diagnosis of] ALK-positive [NSCLC], the standard of care is to treat them with a [TKI] that inhibits that particular oncogene. [There are] now about 4 different frontline options for ALK-targeted NSCLC: crizotinib, ceritinib, brigatinib, and alectinib.

Crizotinib was the first ALK inhibitor available on the market, but subsequent ALK inhibitors have been shown to be better in terms of progression-free and overall survivals. So now, crizotinib has sort of fallen to the side. I would say at this point, probably the ones most used up front are alectinib and, possibly, brigatinib. And then, in progression, lorlatinib has been approved for patients that have progressed on either crizotinib or another ALK inhibitor or on alectinib, brigatinib, and ceritinib.

What factors do you consider when selecting treatment?

Drilon: When I select treatment for a [patient with] ALK-fusion lung cancer, I basically look at a number of factors, including the clinical outcomes of prospective testing. Obviously, things that are highly prized are a good response rate and durable disease control, and that’s manifested as a prolonged median progression-free survival [PFS]. Then finally, given that many of these patients will have a proclivity for developing brain metastasis, control in the [CNS] is also key.…Beyond the [drug’s] activity, other factors like the adverse event profile also play very strongly in my decision. And beyond that, the activity for certain disease states; [for example,] if somebody’s refractory to a prior [TKI], how good is a next-generation ALK TKI at establishing disease control.

Thomas E. Stinchcombe, MD

Stinchcombe: We look at the first step, which is to confirm [that patients] actually have ALK-positive [NSCLC]. I think we’re concerned about false positives and false negatives, so if there is a patient who has high clinical suspicion, we may even do another biopsy.

We started using ctDNA, and I think if that’s positive then it’s enough to prescribe an ALK TKI. If it’s negative, then I always worry about a false negative or the sensitivity with the assay. The first part is due diligence to [ensure] everybody gets tested to make sure they do have an ALK rearrangement.

Generally, most patients get a TKI up front because the response rates are so high, and they come on relatively quickly. At that point, I generally wait until the tumor testing or ctDNA testing is back. I don’t initiate chemotherapy until we have all the molecular studies back.

Rotow: I think a key factor is knowing what prior treatments [patients have] received. A patient who has had prior crizotinib has a very different resistance profile [from] a patient who has never been treated with an ALK inhibitor or who has had multiple generation, later-line ALK-targeted drugs. It is also very important to understand CNS status. You want to choose a drug that has a good track record of CNS activity, and [in terms of] patient comorbidities and preference regarding [adverse] effect profiles, particularly as we get more and more treatment options, understanding those profiles can become more important.

Saxena: We look at the studies. There are some drugs that have been compared head-to-head, [such as] crizotinib versus alectinib. Alectinib had better overall survival and [PFS]. Similarly, brigatinib has been compared with crizotinib and has been shown to have better overall survival. So generally, those have been the ones we typically use rather than crizotinib. Other things, such as [adverse] effect profile, can come into the picture. Ceritinib can have some gastrointestinal [adverse] effects, which may make it difficult for some patients to tolerate—in the studies, a number of patients had to reduce [their ceritinib dosage]. Alectinib and brigatinib happen to be a little more well tolerated, so those are generally what we start off with.

Can you please provide insight into how initial therapy selection influences later-line treatments?

Drilon: Initial therapy selection definitely influences your options for later-line therapies. Crizotinib was the prior standard of care for ALK fusion-positive lung cancers but based on the results of the ALEX trial [NCT02075840] and the J ALEX trial [JapicCTI-132316], alectinib is considered the standard of care. And this drug is a later-generation drug compared with crizotinib, meaning that it has good CNS activity [and also] can inhibit some resistant mutations that merge with crizotinib. So now, if you have a certain powerful drug with a wider scope, the kind of resistance that may emerge with alectinib, it can depend [on] the type of resistance that we see with crizotinib. So that factors into whether or not later-generation TKIs like brigatinib or…lorlatinib can work in this setting.

So far, we have seen some data for brigatinib after alectinib in 2 series. The data from 1 was retrospective [and] the response rate was 17%, but in another series with a small N, [the] response [was] 40%, so I think we still need larger numbers to see exactly how active the drug is in that setting. Certainly, we know that the other pill, lorlatinib, has also shown about a 30% to 40% response rate after alectinib. Only time will tell what the best TKI is after someone progresses on alectinib, but chemotherapy is always an option for these patients.

Ashish Saxena, MD, PhD

Saxena: With all of these ALK inhibitors, patients eventually develop resistance to the therapy, and that can be due to secondary mutations that develop or that are [caused by] insufficient suppression of ALK through some bypass pathways. With crizotinib and some patients treated with alectinib, there are particular mutations you can see after using them where the later drugs, such as brigatinib or lorlatinib in the second line, are sensitive to those mutations. When you use a particular drug in the first line, it can sometimes affect [the] mutations you might see when they do progress.

Stinchcombe: A challenge we face is that we moved our next-generation ALK TKIs to the first-line setting, and although patients have done much better, when they do progress I don’t think there is a clear path forward. You can argue for a chemotherapy, which has a response rate of [approximately] 40% to 45% [and] PFS [of] 68 months. You could also use lorlatinib, which is approved in the second-line setting [approximately] a response rate of about 40% and a [PFS of] around 8 months and, importantly, a high rate of CNS activity there.

I think the real debate in the field is the role of repeat tumor biopsy or ctDNA biopsy and [the] selection of patients based on the presence or absence of ALK mutation or the specific ALK mutation. That’s really where the field wants to go with more personalized care in the resistance setting, because it’s not really clear what the best strategy is at this point.

We talked a lot about the ALK mutations, but up to 50% of patients will not have an ALK mutation. Another big question that necessitates further investigation is whether that’s due to the insensitivity of our testing methods or [whether] they have [an ALK mutation] and it was below our ability to detect, or [whether] they truly have ALK-independent mechanisms of resistances.

Rotow: So, first-line treatment selection certainly does influence later therapies that are used with these patients. We do know that the resistance profile is different for earlierversus later-generation ALK inhibitors. For example, [the] crizotinib spectrum of secondary ALK mutations [are] very different from those that you might see with a later-generation drug, where you see mutations [such as] the G1202R solvent front mutation, which is more difficult to tolerate [and] to target and is resistant to most available ALK TKIs.

[Also exciting] in this space is the availability of lorlatinib, which is approved for use after later-generation ALK inhibitors, including alectinib, and has activity against this particular mutation. It is very promising because it gives our patients with that mutation a new treatment option.

Insights Into Recent Trial Data of First-Line TKIs and Emerging Questions on the Standards of Care

Please discuss the results of the ALTA-1L study that were recently presented. Can you provide insight regarding the use of brigatinib, potential concerns about its safety, and anticipated changes to its use in practice upon approval?

Drilon: Brigatinib is a next-generation [TKI] that was designed to have a very wide spectrum of activity, not just against the wild-type ALK fusion but also against many different resistant mutations that might emerge with prior therapy. It’s not surprising that the drug was previously approved for [patients with] ALK-rearranged lung cancers who had progressed on crizotinib, the prior standard of care, or who for some reason were intolerant to crizotinib.

Now, with the ALTA-1L trial [NCT02737501], this begs the question, “Is brigatinib better than the previous standard of care?” Patients with ALK-fusion lung cancer were randomized to brigatinib or crizotinib. And no surprise—because this was a drug that was designed with a much wider spectrum of activity as mentioned and also with meaningful CNS activity—we saw that the trial results showed there was a substantial difference or improvement in the PFS compared with crizotinib. We don’t generally look for an improvement of objective response rates because they’re in the same ballpark for these TKIs. But what we value when choosing a newer drug over a prior standard of care is seeing whether or not it prolongs disease control, and we saw that. Finally, we also saw that the drug did have meaningful activity in the CNS, and, if you look at the curves of intracranial PFS, we saw those diverge for crizotinib and brigatinib. These very exciting data establish that in comparison to the old standard of care, crizotinib, we’re seeing substantial differences with next-generation drugs.

In terms of safety, brigatinib does have a unique [adverse] effect profile. Compared with the other drugs we use in this space, there is a lead-in dose [for brigatinib] where you got half of the dose at 90 mg daily, then after a week you got up to twice that dose or the recommended base dose of 100 mg daily moving forward. That was done because…some patients had pulmonary or early pulmonary events with brigatinib. It’s unclear what the exact mechanism of action of this pulmonary toxicity is, but the general frequency if you look at the ALTA-1L trial is low—less than 5% for early pulmonary toxicity. The number is 3%, and it seems to be attenuated if you start with the 90-mg dose up to 180 mg. This, in general, can be manageable. If someone does experience early pulmonary events, then the dose can be reduced from 90 mg.

Other things we look out for with brigatinib are things we might see with other TKIs, [such as] gastrointestinal issues [and] change in liver function tests, even though some other drugs have a higher frequency in transaminitis. One signal that we did see emerge was [an] increase in the CPK [creatine phosphokinase], indicating that the drug may cause some inflammation in [the] muscles or myositis. Again, [this is] something that can be managed with drug and dose modifications.

Stinchcombe: [The results of the ALTA-1L trial] show a consistent benefit in terms of PFS, [with a hazard] ratio of 0.49, median about 25 months versus 11 months on the crizotinib. I think it’s important that the crizotinib performs very similarly. It’s also very interesting [that] the patients with brain metastases had a much more robust benefit from brigatinib compared [with that from] crizotinib. The toxicity profile was manageable, with most of the toxicities being asymptomatic or laboratory-based toxicities. We’re fully anticipating that brigatinib will get first-line approval within the next year.

With brigatinib, there’s always been a little bit of concern because it has these rare, early-onset pulmonary symptoms, generally within the first week. It’s part of the reason why they developed the 90-mg daily [dosing] for the first week and then [increasing] to 180 mg. When I use it, I generally see the patients [during] that 1 week afterward because if those events are going to occur, they’ll most likely occur within the first week. I [also] want to assess their symptoms. I think that’s the unique toxicity of brigatinib that’s different from the other ALK TKIs.

[We have seen that brigatinib shows] efficacy in a second- line setting as well as robust efficacy in the first-line setting. One of the questions with brigatinib...was the activity after a next-generation ALK TKI. We had some preliminary data from ASCO [showing] a response rate of about 40%, in an admittedly small group of patients (about 20), in PFS of 6 to 7 months. I think we’d like to see more data about brigatinib after a next-generation TKI rather than having people just empirically use it.

Rotow: Brigatinib is a next-generation ALK inhibitor… approved now for a little while [for use] in the second-line space. However, ALTA-1L is very exciting because it compared the use of brigatinib in the ALK TKI—naïve setting to the use of crizotinib in the same area. [It’s] very similar to the ALEX trial, which compared the use of alectinib to crizotinib in the same space. In ALTA-1L, which is a randomized phase III trial, they enrolled just under 300 patients and treated them with 1 of these 2 agents. The results were very exciting in the initial analyses, which demonstrated an improved hazard ratio for PFS for these patients. And I’m looking forward to seeing those data as they continue to mature, particularly in regard to median PFS and overall survival. As the data become available,...it [could mean] good news for patients because it [will potentially] give them an additional option in the first-line setting, based on that data. Of course, we’re waiting to see whether there will be regulatory approval for a new indication and awaiting the longer-term follow-up for that drug.

Saxena: The ALTA-1L study compared brigatinib with crizotinib and showed that brigatinib had a better response rate and PFS…. It then became approved for the frontline use of patients with ALK-mutated NSCLC. The way it affects practice is that we now have another option we can use in the frontline setting. Alectinib and brigatinib haven’t been compared head-to-head so it’s not really clear which one would be more effective. It just gives people another option for a drug.

In earlier-phase studies, there was some concern with the particular dosing of brigatinib in the 180-mg dose where there was more pneumonitis than expected. The delivery of the drug is now changed so you start off with a lower dose for about a week. If there’s no sign of any pulmonary toxicity, you move onto the 180-mg dose. That regimen was what was used in the ALTA-1L study, and there weren’t any significant safety signals that were very concerning, to me at least. There was more hypertension in the brigatinib group. More CPK elevation and some lab abnormalities but overall, it was pretty well tolerated.

Can you discuss other data that we’ve seen in this space, such as the updated ALEX trial results, and how the results may have influenced treatment selection?

Rotow: The ALEX study was the other big phase III randomized study…that compared alectinib to the use of crizotinib [in] the ALK inhibitor-naïve setting. We’ve seen prior results from this study that showed increases in PFS for these patients, including good and improved CNS activity with alectinib versus crizotinib. The updated results published recently demonstrated an additional follow-up for the ALEX study with another 10 months; it continues to show reassuring increased PFS for patients who received alectinib and show ongoing improvement in CNS activity compared with crizotinib as well. The overall survival data is still immature. I’m looking forward to seeing that, but there is a trend toward survival benefit.

Stinchcombe: I think there have been 3 phase III trials of alectinib compared with crizotinib, and these have established alectinib as a standard of care at this time. It hasn’t been until recently that we’ve seen the CNS penetration of alectinib and responses for people with small asymptomatic brain metastasis. For patients with asymptomatic brain metastasis, many times [when] they are treated with a drug we do some surveillance, then use radiotherapy if we need to later. If you look at these trials, we are now seeing PFS of up to 36 months; this is really a dramatic change in the therapies and the prognosis, and these patients live with a very good quality of life.

The ALEX trial met its primary end point of improvement in PFS by independent radiological review. But importantly, they kept following these patients and by investigator review there was a definable benefit with really durable PFS…about 36 months, which is very robust for a median PFS.

Saxena: The updated ALEX results were really exciting, and they showed PFS of over 3 years with alectinib. And I think knowing that data and being comfortable telling patients that the median PFS is 3 years makes alectinib more often used because we have more data with it. It’s very good data, and it’s very hopeful for the patient to see these things, so having this information definitely impacts how we talk to our patients about the lines of therapy.

Drilon: Knowing that the ALEX trial shows a positive signal in terms of alectinib being better than crizotinib, the next major question is, “Is there anything else that’s active in the first-line space for patients?” And we know that lorlatinib has explored treating ALK TKI—naïve patients in this first-line setting, and we are seeing if the drug is active. What we don’t know just yet is if the durability of disease control will exceed what we’re seeing with alectinib or brigatinib.…If you look at lorlatinib, it was designed to hit more resistance mutations then these other drugs, if you look at particular series.…Time will tell whether or not lorlatinib is a drug that may eventually become the standard of care for these patients, but we just need to establish more data.

What questions remain in the space following the outcomes from these trials? What answers will be needed to influence practice patterns?

Rotow: Raised first off in the ALTA-1L data is going to be the first sign [of the] use of brigatinib versus alectinib, especially right now [as] we only have approval for alectinib in that space. But if brigatinib does move forward, it’ll be interesting to see how the use of one or other of these drugs plays out clinically based on [adverse] effect profiles as we get longerterm follow-up data.

I hinted earlier that there are questions for patients without ALK mutations that are resistant to bypass pathways; how can we define and design a better combination targeted therapy option for these patients and what should therapy be? That’s a real unanswered question that’s the real subject [of] clinical [and] preclinical research.

Saxena: We don’t really know between alectinib and brigatinib if one of them is more effective. They haven’t been compared head-to-head. Brigatinib can sometimes cover certain resistance mutations that alectinib is not sensitive to, so it’s unclear if starting with brigatinib you would have an even longer PFS. And then there’s always the question of when you progress on these second-line therapies like lorlatinib, what is the next step? Most likely it would be chemotherapy, but developing other drugs is always something everyone is looking at, and if there [are] combinations of other ALK drugs like chemotherapy that would work better.

Stinchcombe: Remaining questions will come from the trials for additional next-generation agents in the first-line setting. I think we’re going to be faced with multiple nextgeneration drugs that are approved in the first-line setting and we’re going to have to make a clinical determination based on patient preference, physician preference, and [adverse] effect profile as to which one would be best suited for the patients.

The question about the optimal sequencing of these agents remains prevalent. A consideration, very common I think, was [that] as we look forward, we’re going to have a very crowded frontline space with a lot of questions in the second-line space.

Alexander Drilon, MD, is research director, Early Drug Development, Early Drug Development Service, Memorial Sloan Kettering Cancer Center; Julia Rotow, MD, is a medical oncologist, Lowe Center for Thoracic Oncology, Dana-Farber Cancer Institute; Thomas E. Stinchcombe, MD, is a professor of medicine, Department of Medicine, Duke University School of Medicine, Duke Cancer Institute; and Ashish Saxena, MD, PhD, is an assistant professor of medicine, Department of Medicine, Division of Hematology & Medical Oncology, Thoracic Oncology Service, Weill Cornell Medical College.