The ALK Story: Rapid Bench-to-Bedside Journey Helps Set New Pace for Cancer Drugs

Alice T. Shaw, MD, PhD

In August 2007, researchers in Japan described a new molecular aberration in non—small cell lung cancer (NSCLC): a chromosomal inversion that resulted in ALK gene fusions.¹ Less than 4 years later, crizotinib (Xalkori) gained FDA approval for patients with ALK-rearranged NSCLC. And today, there are 3 more ALK-targeting drugs on the market and an application pending for what would be the fifth agent in this class (Timeline).

Such rapid and prolific drug development has been remarkable for NSCLC, a tumor type that just a decade ago had few treatment options. Beyond that, the trajectory of advancements for this subset of patients with the malignancy, estimated at 3% to 7% of those with NSCLC, has helped define a new model of drug development for the genomic era.

The field reached another milestone recently with the publication of the first mature phase III overall survival (OS) data for an ALK inhibitor. Updated findings showed that after a median follow-up of 46 months, the probability of survival at 4 years for patients who received crizotinib during the pivotal PROFILE 1014 trial was 56.6% compared with 49% for those who had received chemotherapy.² Adjusting for the patients randomly assigned to chemotherapy who crossed over to crizotinib upon progression (144 of 171 patients; 84.2%) revealed a nearly 65% improvement in OS with crizotinib therapy (HR, 0.346).

The findings drew praise from David R. Gandara, MD, director of the Thoracic Oncology Program at the University of California, Davis and a Giants of Cancer Care^® award winner.

“Important long-term survival data in ALK+ NSCLC show 4-year survival of about 50%. Equals transformation to ‘chronic disease,’ which is our goal in all advanced cancers,” Gandara said in a comment posted on Twitter. “And newer ALK inhibitors are adding to this advance.”

In recent interviews with OncologyLive^®, 2 architects of that progress, Alice T. Shaw, MD, PhD, and D. Ross Camidge, MD, PhD, discussed the evolution of ALK-targeting therapies, the impact on the drug development paradigm, and the prospects for future advancements in recent interviews with OncologyLive^®.

“What has set the stage for all targeted therapies that have been developed is knowing what the target is and having a diagnostic assay developed to identify patients with this target. That was critical for getting crizotinib developed because we were able to identify the target population who would be most likely to respond to this targeted therapy,” said Shaw, director of the Center for Thoracic Cancers at Massachusetts General Hospital and professor of medicine at Harvard Medical School, both in Boston. She was the lead investigator for the global registration studies into crizotinib and ceritinib (Zykadia), which was approved in 2014.

A Paradigm Shift in Patient Selection

The prospective identification of patients who would most likely benefit, analysis of the dynamic changes in the tumor that lead to resistance to targeted therapies, and development of therapies that have activity in the central nervous system (CNS)—a common site of metastases—set the stage for efficient development of multiple ALK inhibitors and other therapies targeting other gene mutations, regardless of cancer type. However, much work remains to uncover resistance mechanisms with newer-generation ALK inhibitors, which will help optimize the sequencing of ALK inhibitors and guide development of therapeutic combinations that address ALK-independent pathways of resistance.Crizotinib had already shown activity against ALK and MET aberrations in preclinical studies in anaplastic large-cell lymphoma when the ALK chromosomal rearrangement was discovered in a resected tumor of a patient with NSCLC in 2007.¹ Manabu Soda, MD, PhD, and colleagues, in an article published in Nature, identified the fusion of portions of the ELM4 and ALK genes as a distinct aberration in patients with NSCLC (FIGURE, PAGE 20).

“There was a lot of good timing in this [discovery],” said Shaw. “You have a new target discovered and there’s a targeted therapy poised and ready to be tested in a clinical trial. There was some good timing that helped in terms of speeding the development [of crizotinib].”

Shaw and Camidge, a professor of medicine/ oncology and director of thoracic oncology at the University of Colorado in Aurora, said that the enrolling with prospectively identified ALK rearrangements in an expanded cohort of the phase I trial was a key paradigm-shifting strategy for investigating targeted therapies in lung cancer.³ By contrast, previous strategies with EGFR-targeted tyrosine kinase inhibitors (TKIs) involved giving the drugs to all patients with NSCLC and retrospectively analyzing the characteristics of the small proportion of patients who had dramatic responses. According to Camidge, this method was relatively inefficient for identifying which patients would benefit.

Timeline. Milestones in the Development of TKIs for ALK-Positive NSCLC

ALK indicates anaplastic lymphoma kinase; NSCLC, non—small cell lung cancer; ORR, objective response rate; OS, overall survival; PFS, progression-free survival; TKI, tyrosine kinase inhibitor.

“They spent about 5 years trying to figure out who those 10% [of responders] were,” he said. “Although it became clear that some people did remarkably well, I don’t think they homed in on the subpopulation that had a definable acquired genetic abnormality in EGFR mutations.By

contrast, the phase I/II study of crizotinib prospectively focused on the subgroup of patients with an ALK rearrangement after in vitro studies showed dramatic activity of crizotinib in cell lines that had an ALK rearrangement.⁴ “[The researchers] did something that was relatively unique at that time,” Camidge said. “They actually said, ‘We want you to prescreen people for some of these objective molecular abnormalities, and we’ll give you guidance. We think you should look at lung cancer with ALK rearrangements, kidney cancer for MET mutations, and various subcohorts.’ They were testing the molecular sensitivity hypothesis from the very first expanded cohort from the phase I study.”

Shaw agreed that knowing the target and being able to identify the target population most likely to respond to ALK-targeted therapy were critical to the rapid progression from development to approval of crizotinib and later-generation ALK inhibitors. “That was really a pivotal step in the development of the drugs—that we were able to test the drug in the correct population of patients,” she said. “All of our ALK inhibitors—like ceritinib, alectinib [Alecensa], and brigatinib [Alunbrig]—have been tested in that target population and we don’t have to be swimming in all of these patients who really don’t have the target.”

This strategy—testing targeted therapy in a preselected group of patients—was instrumental in studying crizotinib’s efficacy for other uncommon mutations in NSCLC. Crizotinib was approved for ROS1-positive metastatic NSCLC in 2016 after an expanded cohort analysis of 50 patients with ROS1-rearranged NSCLC showed an objective response rate (ORR) of 72%; the drug also demonstrated antitumor activity in patients with MET exon 14-altered NSCLC.^5,6 In May, crizotinib was granted a breakthrough therapy designation for metastatic NSCLC with MET exon 14 alterations that has progressed after platinum-based chemotherapy.

“What’s been exciting is to see that the [drug] activity and approvals have been tissue-agnostic,” Shaw said. “We’re seeing this whole idea of ‘find the right population of patients’ so you can test the targeted therapy in the correct population.”

Figure. EML4-ALK Fusion

Most ALK rearrangements in non—small cell lung cancer involve an inversion within chromosome 2p that results in a fusion of the ELM4 and ALK genes.

Defining CNS Activity

Shaw and Camidge agreed that the paradigm developing targeted therapies in lung and other types of cancer has permanently shifted. “The concept of using a targeted drug in an untargeted way is yesterday’s motif,” said Camidge. “Coming in with a molecular preselection from the earliest trial is very common now.Historically, many clinical trials have excluded patients with CNS metastases, and treating these patients represents a major unmet need. Brain metastases are common with ALK-positive NSCLC and were present in 35% of patients upon enrollment into the PROFILE 1007 trial.⁷ New or progressive brain metastases frequently occur in crizotinib-treated patients and often appear earlier than recurrences or metastases in the rest of the body, Camidge noted. This highlights the lack of CNS activity—a “major liability” of crizotinib, he said.

Traditionally, options for brain metastases include whole-body or stereotactic radiation therapy and surgery; however, these treatments are associated with high toxicity and may not be feasible for patients with poor performance status. Therefore, development of second- and third-generation ALK inhibitors has focused on increasing CNS penetration and activity. In the ALEX trial, the CNS-specific objective response rate (ORR) with frontline alectinib was 36% compared with 28.6% with crizotinib in patients treated with prior radiotherapy. For those without prior radiotherapy, alectinib showed a CNS ORR of 74.4% versus 24.3% with crizotinib.8 In the ASCEND-4 study, ceritinib yielded an overall intracranial response rate of 57% with a median duration of 16.6 months.⁹

The recent breakthrough therapy designation for lorlatinib, a third-generation ALK inhibitor, was largely due to its robust CNS activity, according to Shaw. Data from a phase I/II trial showed that the drug yielded a 90% ORR and 75% intracranial ORR in patients with treatment-naïve ALK-positive NSCLC.10 The FDA is evaluating a new drug application for lorlatinib for patients with ALK-positive metastatic NSCLC who have progressed on 1 or more ALK TKIs; a decision is expected by August 2018.

“One of the major criteria they wanted in [lorlatinib] was that it would penetrate into the brain,” Shaw, a leading investigator into the drug, said. “I think part of the reason lorlatinib is highly effective is because it is able to control the disease in the brain as well as the systemic disease.”

Camidge also noted that the research on CNS activity of ALK-targeted therapy was a key impetus for the European Association of Neuro-Oncology’s recently created evidence-based guidelines on the diagnosis and treatment of brain metastases for solid tumors.¹¹ “The whole eld of ALK inhibitors went from presenting some retrospective messy data saying, ‘We have activity in the brain’ to helping to lead to a set of guidelines that says, ‘If you want to capture data in the brain, this is how to do it,’” Camidge said. “We’re getting much more robust information in terms of a drug’s activity in the brain, and I think ALK has certainly led that field.”

Table. PFS Data for FDA-Approved ALK Inhibitors in Frontline Trials

NE indicates not estimable; PFS, progression-free survival.

Next Steps for ALK Inhibitors

Addressing Treatment Resistance

Overall, the lack of CNS activity with crizotinib helped highlight CNS metastases as a relevant clinical endpoint that needed to be investigated when developing newer-generation ALK inhibitors. “In addition to working on some of the mechanisms of resistance to crizotinib biologically, [the new-generation agents] have to have penetration into the brain,” Camidge said. “I don’t think you could develop a successful ALK inhibitor that doesn’t have CNS penetration these days.Although ALK-targeted therapies initially show efficacy for ALK-rearranged NSCLC, they also create resistence by altering the tumor’s biology. Resistance and disease progression on crizotinib often occurs within 1 to 2 years and were described in case reports soon after the first publication of the drug’s efficacy. Investigators quickly recognized the need to research changes in the tumor microenvironment and resistance mechanisms that occur with a given ALK inhibitor and develop agents that worked on the new biology of the tumor.

“Almost at the same time [that] we were presenting data on the rst publication of efficacy [for crizotinib], there would be case reports of resistance based on somebody who rebiopsied,” Camidge said. “The concept of rebiopsying was very foreign a few years ago.”

Common mechanisms of crizotinib resistance include secondary mutations in the ALK tyrosine kinase domain or amplification of the ALK fusion gene (sometimes in combination with a secondary ALK mutation), and many resistant tumors continue to depend on the ALK gene for growth and proliferation, indicating that sequential inhibition with ALK-targeted therapies is often effective for prolonging survival.

Therefore, many second- and third-generation ALK inhibitors are designed to target common resistance mutations in the ALK gene that occur with crizotinib. For example, most newer-generation ALK inhibitors have been shown to target the L1196M gatekeeper mutation, among others, which regulates accessibility of the kinase adenosine triphosphate binding pocket and is a relatively common mechanism of crizotinib resistance.¹²

Several of these agents have also been tested in the frontline setting. Ceritinib and alectinib recently received approval for frontline use based on data from the phase III ASCEND-4 and ALEX trials, respectively, and Camidge predicted that several other ALK inhibitors soon will be approved for frontline use and eventually replace crizotinib as the first-line agent.^13,14

Optimizing Therapeutic Sequence

“There is a role for sequential ALK-targeted therapies using these increasingly potent ALK inhibitors one after the other,” Shaw said. “Patients can really derive a lot of benefit from that approach, but our hope is to improve on that and use our best molecules up front rather than use them in the order in which they were developed.”Although resistance mechanisms have been crucial for guiding development of ALK-targeted therapies used after crizotinib, patterns of resistance will likely differ when these agents are used in the frontline setting versus second- and third-line settings. Camidge also noted that because the frontline improvements in progression-free survival (PFS) were so remarkable with the new agents (median PFS, 25.7 vs 10.4 months for alectinib vs crizotinib) compared with crizotinib, identifying their patterns of resistance and optimal subsequent lines of therapy could take several years (Table^1,13,14). However, he said that because the patient could be taking the drug for several years, the choice of which ALK inhibitor to use in the frontline setting will likely depend on the adverse effect profile, convenience of dosing, and cost of the drug.

Camidge cautioned that although newer-generation ALK inhibitors target a broad spectrum of ALK mutations induced by the previous ALK inhibitor, a broader spectrum of activity in vitro may not translate into longer duration of disease control. For instance, he said that although lorlatinib has a broader spectrum of activity in vitro than brigatinib, it appears to have a numerically shorter median PFS when doing a lateral comparison of clinical trial data. “It’s not that the mutations are unimportant, but they’re [just] part of the story. Lorlatinib may be a ‘super’ ALK inhibitor, but maybe you need something that has a little more activity against pathways other than ALK,” he said.

Crizotinib’s use in the frontline setting will probably diminish over the next several years, Camidge said , but its roles and that of the other ALK inhibitors will likely target specific mechanisms of resistance. “Every one of these drugs will have some niche role, but we just have to gure out which is the [drug that is] best tolerated and easiest to remember to take to start with,” he said.

Therapeutic Combinations With ALK Inhibitors

The 4 approved ALK inhibitors—all oral medications—were approved with different recommended full dosing regimens: crizotinib, 250 mg twice daily with or without food; ceritinib, 450 mg once daily with food; alectinib, 600 mg twice daily with food; and brigatinib, 180 mg once daily with or without food.Sequential ALK inhibitor therapy leads to ALK-independent mechanisms of resistance in many patients, and combination strategies will be necessary to target these effectively, according to Shaw. She and her colleagues are investigating combinations of lorlatinib and other therapies that target the most common ALK-independent pathways of resistance, although these can be diverse.

“We want to think strategically and target the ones that are most common, but likely it’s not going to be 1 simple combination that solves resistance for everybody,” she said. “We need more research to understand what the most common mechanisms of resistance are and the best way to target them. We need the clinical trials to test these combinations, both in terms of how well they work and in terms of safety. One of the challenges with developing combinations is that we often run up against toxicities that are either much worse than we see with single drugs alone or just completely unanticipated.”

Checkpoint inhibitors, for example, have demonstrated severe toxicities when combined with ALK inhibitor therapy. Group E in CheckMate 370, a single-arm cohort, showed that nivolumab (Opdivo) plus crizotinib led to severe hepatotoxicities in 5 of 13 patients (including 2 deaths) with ALK translocation-positive NSCLC, leading to discontinuation of the combination.¹⁵ “We certainly want to develop immunotherapy strategies for ALK patients, but simply trying to combine an ALK inhibitor with a checkpoint inhibitor is not the right way to go,” Shaw said.

Monotherapy with checkpoint inhibitors is “almost completely ineffective” in patients with lung cancer, Camidge agreed, but he noted that the IMpower150 trial showed that adding atezolizumab (Tecentriq) to bevacizumab (Avastin) and chemotherapy appeared to benefit a group of patients with ALK-positive NSCLC.¹⁶

“The question is, if you completely perturb the environment by throwing everything else at it, can you engender some role for immunotherapy?” asked Camidge, adding that the current role of immunotherapy for ALK-positive NSCLC is experimental.