Monotherapy with a tyrosine kinase inhibitor is the frontline standard of care for patients with advanced EGFR-mutant non–small cell lung cancer, with 3 currently approved agents: erlotinib, gefitinib, and afatinib.
Sarah B. Goldberg, MD
Monotherapy with a tyrosine kinase inhibitor (TKI) is the frontline standard of care for patients with advanced EGFR-mutant non—small cell lung cancer (NSCLC), with 3 currently approved agents: erlotinib (Tarceva), gefitinib (Iressa), and afatinib (Gilotrif). Following acquired resistance, patients who test positive for the EGFR T790M mutation receive osimertinib (Tagrisso), while those with negative tumors undergo chemotherapy.
However, this sequence could undergo a major upheaval as a result of top-line findings from the phase III FLAURA trial, recently announced by AstraZeneca, the developer of the third-generation EGFR inhibitor osimertinib.1 In the 566-patient study, osimertinib improved progression-free survival (PFS) compared with erlotinib or gefitinib in the frontline setting for patients with EGFR-mutant advanced NSCLC. At this point, data have not been released.
“We know that osimertinib has an improvement in PFS compared with erlotinib or gefitinib, which I don’t think is a surprise to anyone,” said Sarah B. Goldberg, MD, MPH, assistant professor of medicine (medical oncology) at the Yale School of Medicine and Yale Cancer Center in New Haven, Connecticut.
“The question is, how much of a progression-free survival benefit? It will really have to be significant enough to overcome the benefit of sequencing the 2 lines of therapy,” she said, during a presentation on sequencing EGFR inhibitors at the 18th Annual International Lung Cancer Congress® (ILCC), hosted by Physicians’ Education Resource® (PER®) July 27 to 29 in Huntington Beach, California. “I think we will have much more data in the next few months and maybe this will change our practice.”The 3 currently approved frontline TKIs have similar efficacy but with varying toxicity profiles, with none showing superior overall survival (OS). The most recently added agent, afatinib, was compared with gefitinib in a head-to-head phase IIb study for patients with EGFR-mutant NSCLC.2
Median OS was 27.9 months with afatinib and 24.5 months with gefitinib, but this was not statistically significant (HR, 0.86; 95% CI, 0.66-1.12; P = .258). The objective response rate (ORR) across the study was 72.5% with afatinib compared with 56% for gefitinib (P = .0018). The median PFS was 11.0 versus 10.9 months, for afatinib and gefitinib, respectively (HR, 0.73; 95% CI, 0.57-0.95; P = .017).
Regardless of the frontline EGFR TKI selected, resistance is inevitable, primarily due to an acquired mutation in T790M. For those with this type of progression, osimertinib quickly became the standard of care, following an accelerated approval in November 2015, Goldberg noted. In March 2017, osimertinib was granted a full approval for pretreated patients with T790M-positive NSCLC, based on findings from the phase III AURA3 trial.3
In AURA3, the median PFS with osimertinib was 10.1 months compared with 4.4 months for pemetrexed plus platinum-based chemotherapy (HR, 0.30; 95% CI, 0.23-0.41; P <.001). The ORR with osimertinib was 71% compared with 31% with chemotherapy. The median duration of response was 8.5 versus 4.2 months for osimertinib and chemotherapy, respectively. OS data were not yet available.
“Because of this trial, osimertinib is now the standard-of-care for patients. It has been approved for T790M-positive disease,” Goldberg said. “After progression, it has become standard to biopsy patients and, if T790M positive, to give osimertinib.”Whether the frontline standard of care will be changed depends heavily on the magnitude of benefit seen with osimertinib. Although data from the phase III FLAURA study have not yet been released, Goldberg referenced phase I data from a 60-patient study of osimertinib in the first-line setting as a potential indicator.4
In this study, the ORR was 67% for patients assigned to 80 mg once daily osimertinib (95% CI, 47%-83%) and 87% with osimertinib at 160 mg (95% CI, 69%-96%). Across doses, the median PFS was 19.3 months with osimertinib and the 12- and 18-month PFS rates were 72% and 55%, respectively.4 “The median PFS with the other agents is about 10 to 12 months, so really this is an impressive PFS,” she said.
If the PFS benefit in this study was also seen in the phase III results, it may be sufficient to warrant switching to frontline osimertinib, Goldberg noted. “I would love to see the overall survival data from FLAURA but that’s not going to happen for a while,” she said. “That’s really what we want to know. I was not surprised at all to see that PFS was positive.”
If osimertinib were to move to the frontline setting, there would be larger sequencing implications for the second-line setting. Following progression on osimertinib, the next treatment would be chemotherapy, which has shown about a 4-month PFS benefit across prior trials, Goldberg noted. In some cases, this could mean inferior outcomes versus the current standard of care (Figure 1).
“With the phase I, PFS [with osimertinib] was 19 months,” she said. “I think for the T790M-positive patients, they might be missing out. They probably have more benefit with their 1 year of erlotinib or afatinib and then 10 months of osimertinib. But you don’t know which patients are going to get T790M and which won’t, so that’s half the patients.”
ILCC conference chair David R. Gandara, MD, did not consider this approach to be an appropriate method for determining a therapeutic sequence, “I would caution against using medians to determine if we’re going to establish a new standard of care,” he said.
“That’s a total misinterpretation of the KaplanMeier curves,” added Gandara, an OncLive® Giant of Cancer Care® in lung cancer and director of the University of California Davis Comprehensive Cancer Center’s Thoracic Oncology Program. “There are many trials where the medians are not different but there are subsequent major differences. Oncologists are fixated on these medians, and we need to get off that.”
Outside of efficacy, other characteristics could come into play when considering the frontline treatment of choice, especially the presence of brain metastases or toxicity. Additionally, patient choice should weigh heavily toward the decision.
Although select patients with brain metastases— namely those with more aggressive lesions—are better treated using local therapy, some are candidates for EGFR inhibitors, Goldberg said. “I think osimertinib has better penetration in the brain compared with other TKIs but that hasn’t been proven yet,” she said. “That [added efficacy] would push me, in a patient with brain metastases, to use osimertinib in the first line.”
In findings from the second-line AURA3 trial,5 those with central nervous system (CNS) metastases experienced benefits similar to those of the full population of the trial. In participants with ≥1 measurable CNS metastasis at baseline, the CNS ORR was 40% (95% CI, 29%-52%) with osimertinib and 17% (95% CI, 7%-32%) with chemotherapy (odds ratio [OR], 3.24; 95% CI, 1.33-8.81; P = .014).
The CNS PFS was 11.7 months with osimertinib compared with 5.6 months for chemotherapy. The differences in adverse events also warrant consideration (Table4, 6-8), although this criterion may not be as convincing as efficacy. “Osimertinib definitely has better toxicity but once the patients get through the first couple weeks of erlotinib or afatinib I don’t think toxicity is big issue,” Goldberg said.In addition to osimertinib, multiple therapies are currently racing toward the frontline setting. Although other contenders are not currently standards of care, they could further disrupt the future paradigm. Chief among these regimens is the combination of erlotinib and bevacizumab (Avastin), which is being explored in the phase III BEVERLY study as a frontline therapy for patients with EGFR-mutant NSCLC (NCT02633189).
“The exploration of erlotinib/bevacizumab is a very important trial that’s being conducted,” Goldberg said. “Hopefully we’ll have data soon to tell us if this is a treatment we should use for our patients.”
In a prior phase II study,9 the median PFS with the combination of bevacizumab and erlotinib was 16.0 months compared with 9.7 months with erlotinib alone (HR, 0.54; 95% CI, 0.36-0.79; P = .0015). The ORR in the combination arm was 69% versus 64% with erlotinib alone (P = .4951). OS data were immature.
Another single-agent TKI, dacomitinib showed signs of efficacy when compared with gefitinib in the frontline setting for patient with EGFR-mutant NSCLC; however, Goldberg expressed uncertainty about the future for this agent, and questioned whether it would become available. “Dacomitinib is not an approved drug yet, and I am not sure it will be, but it has some promising data in the firstline setting,” she said.
In the phase III ARCHER 1050 trial,10 the median PFS with dacomitinib was 14.7 months compared with 9.2 months for gefitinib (HR, 0.59; 95% CI, 0.47-0.74; P <.0001). ORR was not statistically different between the groups. The ORR was 74.9% with dacomitinib versus 71.6% with gefitinib (P = .3883). OS data were not yet mature.If osimertinib were to become the frontline treatment choice, greater emphasis would have to be placed on the second-line setting. Studies have already explored agents for T790M-negative tumors, for which osimertinib is not typically considered an option.
For those with T790M-negative tumors, resistance is primarily driven by activation of bypass signaling, mutations in downstream pathways, or phenotypic changes in the tumor. In early studies of osimertinib, which included T790M-negative tumors, slight efficacy was still seen with the third-generation EGFR inhibitor. In this study, the ORR was 21% with osimertinib.11
“The drug is not approved for this population of patients but it is something interesting to note,” Goldberg said. “The question to ask is, could these patients have been T790M-positive by plasma? The biopsy was done on tissue for these patients but we now know a lot more about finding T790M.”
Current considerations for T790M-negative patients include chemotherapy, afatinib plus cetuximab (Erbitux), or immunotherapy. The combination of afatinib and cetuximab demonstrated promise for patients with acquired resistance to first-line EGFR TKIs in a phase II study.12 In this 126-patient study, the combination had an ORR of 25% in T790Mnegative tumors and a disease control rate of 62%.
The median PFS was 4.7 months. A phase II/III study is exploring afatinib plus cetuximab as a frontline therapy for EGFR-mutant NSCLC. The primary endpoint for the phase II portion is PFS and for the phase III it is OS (NCT02438722).
Goldberg prefers using chemotherapy in the T790M-negative setting. Her go-to regimen is carboplatin plus pemetrexed for patients with adenocarcinoma, although several options are available. In some patients, if eligible, she would also consider bevacizumab.
Given the success seen in other types of NSCLC, immunotherapy could also represent an intriguing option; however, Goldberg does not feel that PD-1/PD-L1 inhibitors should be tried for patients with EGFR-mutant NSCLC, based on the limited data available.
In the KEYNOTE-010 study,13 the PD-1 inhibitor pembrolizumab (Keytruda) was not superior to docetaxel for pretreated patients with EGFR-mutant NSCLC, although this population represented a small subset of the trial. This lack of benefit was observed both for OS (HR, 0.88; 95% CI, 0.45-1.70) and PFS (HR, 1.79; 95% CI, 0.94-3.42).Resistance to osimertinib does eventually develop through a variety of mechanisms that are rapidly being identified (Figure 2).14 “Resistance to osimertinib is a problem, and we don’t yet know how to overcome it,” Goldberg said.
Studies are attempting to uncover regimens that capitalize on changes in the tumor following treatment with osimertinib—a scenario such as T790M following first- or second-generation TKIs. Chief among the potentially targetable resistance alterations is C797S, for which fourth-generation EGFR TKIs are already under development. Among the most notable is EAI045, which has shown efficacy in mouse models.15
Preclinical studies are also assessing combination approaches to overcome resistance to osimertinib, specifically in the third-line setting. In early findings, the combination of brigatinib (Alunbrig) and osimertinib proved effective in vitro and in vivo for NSCLC cells harboring EGFR T790M and C797S mutations.16
At this point, data on the rates of the C797S mutations are from second-line settings, suggesting that this resistance mechanism could occur more frequently in the first-line space. In fact, Goldberg theorized, the alteration potentially could occur at the same frequency as the T790M alteration for first- and second-generation TKIs (50%-60%).
In yet another dilemma in the sequencing story, it is unclear whether current EGFR TKIs would still be effective in patients with EGFR C797S mutations, because the T790M mutation would not be present. As a result, it might still be possible to utilize erlotinib, gefitinib, or afatinib following frontline osimertinib, should this agent become the standard of care. However, data on this are not currently available.
“There’s a lot about this that we don’t know yet,” Goldberg acknowledged. “There may be cases where a patient progressing on a third-generation agent could then respond to a first- or second-generation agent. We see this in ALK-positive patients. But, I don’t think that will be the majority of patients.”