Two oncologists offer rationale for both sides of the argument for/against the upfront use of newer tyrosine kinase inhibitors for patients with ALK-positive or EGFR-mutant NSCLC.
Fabrice Barlesi, MD, PhD
With the advent of newer generations of targeted therapies for treating patients with nonsmall cell lung cancer (NSCLC), debate has erupted within the medical community over whether patients should be treated with newer agents upfront, or whether these agents should be held for later use. During the 2017 European Lung Cancer Conference, 2 oncologists took up this debate and offered the rationale for both sides of the argument for/against the upfront use of newer tyrosine kinase inhibitors (TKIs) for patients with ALK-positive or EGFR-mutant NSCLC.
On the pro side, David Carbone, MD, PhD, first raised the issue of sequencing TKIs to achieve a greater overall survival (OS) than would be achieved with 1 agent alone. Fabrice Barlesi, MD, PhD, who spoke against using newer agents upfront, supported sequencing but advocated for using the first-generation agents first in the sequence.
Barlesi is head of multidisciplinary oncology and therapeutic innovations at Assistance Publique Hôpitaux De Marseille, coordinator of the Center of Early Phase Cancer Trials of Marseille, and professor of medicine at Aix-Marseille University in France. He pointed out that according to the results of retrospective analyses, responses and treatment durations are better when agents are used in succession rather than alone.
He acknowledged that the survival advantage with newer agents has exceeded that of the first-generation agents, such as when comparing alectinib (Alecensa) or ceritinib (Zykadia) with the first-generation agent crizotinib (Xalkori) in patients with an ALK rearrangement. In the J-ALEX study of patients without prior ALK inhibitor treatment, for example, the progression-free survival (PFS) was 25.9 months in the alectinib arm (95% CI, 20.3 months to not reached) and was 10.2 months (95% CI, 8.2-12.0) with crizotinib (HR, 0.34; 95% CI, 0.17-0.71; P <.0001).1
“If you look only at the short term, at the PFS, yes, it’s true, you have an advantage to PFS in the ALK-rearranged patients for alectinib and you can see that the PFS is increased,” Barlesi said.
However, sequential treatment of these ALK inhibitors, starting with crizotinib followed by alectinib, led to a median OS of 51.1 months and a median combined PFS of 18.2 months, according to a retrospective study.2According to the IFCT-1302 CLINALK trial, a French retrospective study, OS outcomes would be best when, following crizotinib, patients were treated in the second line with a next-generation ALK inhibitor rather than with an alternative treatment, usually chemotherapy, or best supportive care.3
Carbone, the Barbara J. Bonner Chair in Lung Cancer Research, director of the James Thoracic Center, and professor of medicine at The Ohio State University Comprehensive Cancer Center, alternatively suggested that 2 median PFS values could not be added together to come up with an estimated “time on TKI” calculation. A patient who achieves the median PFS with 1 drug may not achieve the median PFS with a subsequent drug, or vice versa, he noted.
Carbone also noted that this rough calculation does not account for treatment beyond progression. Evidence from clinical trials have shown that there is an added clinical benefit to continuing to treat beyond progression by RECIST criteria. In the ASPIRATION trial, patients with EGFR-mutant NSCLC were treated with first-line erlotinib beyond which are common in NSCLC, the change of treatment can also be delayed to stay on the first-line targeted agent longer, thereby increasing the patient’s OS.
In addition, Carbone said that saving the next-generation TKIs for later makes no sense, and it is better to treat patients with the best treatments first. “History has taught us that in oncology, using the best treatment first results in the best survival,” he stated.
With newer agents that demonstrate prolonged PFS rates, clinicians can pre-empt the development of known resistance mutations, such as T790M in EGFR-mutant disease, and provide a rapid and more reliable relief of symptoms, according to Carbone.
However, Barlesi contended that the use of newer agents instead cuts off additional treatment options down the line. If, for example, a patient were to be treated frontline with the second-generation ALK inhibitor alectinib, the patient would be more likely to develop a G1202R resistance mutation, and none of the ALK inhibitors are sensitive to this resistance mutation by the half maximal inhibitor concentration, mean IC. Even the next-generation agent lorlatinib, which shows the highest sensitivity to this mechanism of resistance of the ALK inhibitors, has a sensitivity very close to the 50 nmol/L IC50 cutoff, at 49.9 nmol/L.5Without any additional ALK inhibitors available to use in this situation in the second line, the next treatment approach would be chemotherapy.
In EGFR-positive NSCLC, the T790M resistance mutation can be treated with the third-generation EGFR TKI osimertinib (Tagrisso), but what can be done when the patient develops a resistance mutation from treatment with osimertinib? Targeted agents are not currently available to treat these new resistance mechanisms, including MET amplification, HER2 amplification, and more.
Barlesi suggested that tumors with actionable targets are heterogeneous, and that each patient and mechanism of resistance is different and should be treated individually.
He also said that the safety profile of the newer agents can be slightly better than that of the first-generation TKIs, but not significantly so. Pointing to the rates of grade 3/4 dyspnea among the different ALK inhibitors, Barlesi cited 4% with crizotinib, 5.7% with ceritinib, 3% with alectinib, and 7% with brigatinib (Alunbrig).6
For both EGFR and ALK TKIs, the central nervous system is a frequent site of disease progression, but the newer drugs have substantial activity in the brain, Carbone said. In the J-ALEX trial, patients with brain metastases at baseline who were treated with alectinib had a significant improvement compared with those treated with crizotinib; the median PFS was not reached in the alectinib arm, while it was 10.2 months in the crizotinib arm.7In patients without brain metastases at baseline, the median PFS was 20.3 versus 10.0 months, respectively.
However, Carbone and Barlesi pointed out that these patients’ brain metastases will usually be pretreated, and in this case, the difference in central nervous system activity is not as significant. The objective response rate of patients with pretreated brain metastases with crizotinib was 33% versus 36% with ceritinib and 28% with brigatinib.6“When we look at the safety, there is maybe a slight advantage, if any, but absolutely not for all the newer targeted agents, and the safety profile should not drive our decision with the data we have to date,” Barlesi stated.
Carbone suggested that the best way to end this debate would be to conduct a clinical trial, rather than adding PFS values together. He pointed to the ongoing randomized, 3-arm APPLE trial that is investigating patients with an EGFR T790M mutation treated with either osimertinib in the first line, gefitinib (Iressa) followed by osimertinib upon RECIST progression, or gefitinib followed by osimertinib upon molecular progression. He said that this will answer questions around whether first-generation or newer agents are better in the frontline, and whether RECIST progression or molecular progression is the right indication for switching treatment.
“I think that a long first PFS really matters, that toxicity matters, that brain penetration matters. And we need to do the right studies with planned crossovers with time on TKI and OS endpoints, as well as paying attention to quality of life,” Carbone concluded.