In a Targeted Oncology case-based peer perspectives live discussion, Heather A. Wakelee, MD, discussed treatment with tyrosine kinase inhibitor treatment options for a real-world Hispanic patient with EGFR-mutant non–small cell lung cancer.<br />
Heather A. Wakelee, MD
In a Targeted OncologyTMcase-based peer perspectives live discussion, Heather A. Wakelee, MD, professor Medicine (Oncology) at Stanford University and discussed treatment with tyrosine kinase inhibitor (TKI) treatment options for a real-world Hispanic patient with EGFR-mutant nonsmall cell lung cancer.
History and physical exam
• A 66-year old Hispanic woman presented to her primary care physician complaining of visual disturbances, nausea, fatigue, and sporadic headaches.
• Medical history: hypertension managed on candesartan; hyperlipidemia managed on simvastatin; no smoking history
-Blood pressure: 148/70 mm Hg; decreased breath sounds in lower left lobe; otherwise negative
• Complete blood count and chemistry work up: within normal limits
• Brain MRI demonstrates a 10 mm right parietal mass at the gray-white junction with vasogenic edema.
• CT of chest, abdomen, and pelvis reveals a 3.4 cm mass in the lower left lobe and several small liver nodules.
• CT-guided transthoracic needle biopsy of the lung lesion shows grade 2 adenocarcinoma, acinar subtype.
• ECOG performance status: 1
• PD-L1: 55% PD-L1 expression on tumor cells (22C3 pharmDx test)
• Molecular panel testing: EGFR exon 19 deletion
Targeted Oncology™:Would you initiate multidisciplinary assessment for treatment of thepatient’s central nervous system (CNS) involvement? If so, would you opt for stereotactic radiosurgery (SRS) before instituting the tyrosine kinase inhibitor (TKI) therapy?
WAKELEE: I’m starting to see more patients who are in this exact scenario and they are started on a three-drug regimen [involving osimertinib (Tagrisso), erlotinib (Tarceva), gefitinib (Iressa)], or pembrolizumab [Keytruda]. For this case, a sample was sent for molecular panel testing and the result was anEGFRexon 19 deletion. There is evidence for using an EGFR TKI versus a checkpoint inhibitor in the first-line setting. She is also symptomatic.
Radiation oncologists might opt for SRS before a TKI, but the trouble might appear 3 years from now when the patient develops edema. Before, we didn’t have to worry about 2 or 3 years after SRS, but now I see it more often with my patients, especially in patients with theALKmutation. I’m looking ahead to the future a little bit more because we’ll need to consider surgery 2 years post SRS.
If the patient were asymptomatic, I wouldn’t consider SRS. Because she’s symptomatic, I might at least talk to her about the potential that in 2 or 3 years we’re going to have to deal with edema. What I find troubling is when the patient comes to me and they have 20 little things [wrong] and they’re not that symptomatic, but they’ve been rushed to have the whole brain treated. In 2 or 3 years, you have a patient starting to show signs of early dementia. I have lost a couple of people to dementia with adenocarcinoma from whole brain radiotherapy. But in this case, it’s reasonable to consider SRS.
Do you prefer to use osimertinib upfront or hold its use in case an EGFR T790M mutation develops on first-generation TKIs?
In our clinic, which is a Kaiser [health plan] system, the patient is given gefitinib in the first line. If they progress, they go onto osimertinib. We have data using erlotinib and bevacizumab [Avastin] or ramucirumab [Cyramza]. We have data supporting gefitinib and chemotherapy that prove PFS [progression-free survival] and OS [overall survival] improvement. We have impressive FLAURA data, but there is also significantly longer PFS data if you give VEGF plus an EGFR inhibitor followed by osimertinib later.1,2
What do the National Comprehensive Cancer Network (NCCN) guidelines on nonsmall cell lung cancer (NSCLC) suggest for first-line treatment?
In the first-line setting, osimertinib is the preferred first-line treatment, according to the NCCN.3There are other recommended agents, including erlotinib, afatinib [Gilotrif], gefitinib, and ramucirumab. In specific situations, the combination of erlotinib and bevacizumab is a possible treatment, too. If a patient starts on one of these therapies and then is found to have anEGFRmutation, I don’t necessarily switch them. In my experience, it depends on how the patient is doing.
What are the key findings from the FLAURA trial?
In FLAURA, patients who received osimertinib had a median PFS of 19.9 months versus patients who received standard of care, which was 10.2 months [HR, 0.46; 95% CI, 0.37-0.57;P<.0001]. Many clinicians have changed their treatment paradigm based on these results, but there are some hold outs. Initially, the HR for OS was 0.63, but later, when the final OS results were presented at ESMO [European Society for Medical Oncolgy], the HR was 0.799.
Some critics of FLAURA suggest that in Asian subsets, the findings aren’t as impressive, but some data looking at a Chinese population demonstrated similar findings. The curves for the Chinese population were almost the same as the curves for the overall population. Half of my patient population in California is of Asian descent and I talk to them about the FLAURA findings. I mention the Asian subset findings, but I also emphasize that the agent is the right drug for them. I have also prescribed osimertinib off-label for atypical patients. I believe there will be some presentations at ASCO [American Society of Clinical Oncology] this year focusing on atypical patients withEGFRmutations. I have a few atypical patients on osimertinib. One in particular started on osimertinib but ended up with cardiac toxicity and QTc prolongation, so now he’s on afatinib.
What do the guidelines say about treating patients who are PD-L1 positive with an EGFR mutation?
According to the NCCN guidelines, if your patient has a driver mutation, don’t give them a checkpoint inhibitor. We started hearing about these patients who weren’t doing well with checkpoint inhibitors in the second line. At the Society for Immunotherapy of Cancer meeting in 2018, there were some data showing that it’s not the best plan to start with checkpoint inhibitors in patients with nonsquamous, advanced NSCLC and 1 or more actionable mutations [EGFR,ALK, orROS1].4
Another study out of University of California, Los Angeles had me worried. When I heard about it, I thought it was unethical. The investigators looked at giving pembrolizumab before a TKI for patients with advanced NSCLC who wereEGFR-positive and PD-L1positive.5In other words, that’s single-agent pembrolizumab in the first-line setting. In the first 11 patients recruited, the investigators only had 1 responder, and it turned out that the patient didn’t have anEGFR[mutation] and wasn’t eligible for the trial, but ended up on it anyway.
Seventy-three patients had high PD-L1 expression, so this is 1 responder out of 11, even although everybody had high
PD-L1 expression. They reported 2 deaths in 6 months, which included 1 death from pneumonitis.
What are some safety concerns associated with the sequencing TKIs plus checkpoint inhibitors?
Not all the data involving TKIs plus the checkpoint inhibitors suggest a high toxicity rate. Crizotinib [Xalkori] has high hepatitis rates and osimertinib has high pneumonitis rates. Some of the others are safer. Erlotinib seems to fare a little better than others. But for the trials that evaluate combinations, the efficacy rates look no different from TKIs alone. That’s why we don’t pair them. We’re not recommending it.
Adam J. Schoenfeld, MD, and colleagues looked at patients who had any kind of PD-L1 or PD-1 blockade and then went
on to receive osimertinib.6The investigators reported severe adverse events [AEs] in 15% of patients. Median onset was
about 20 days after they started the osimertinib and the patients all required steroids.
When the investigators looked at erlotinib, they didn’t see these types of AEs. If patients were given an EGFR TKI first
and then went on to a checkpoint inhibitor, patients were fine because the EGFR TKI is not going to stick around the
system. The checkpoint inhibitor is there and remains for a couple of months.
Please discuss the meta-analysis of second-line therapy evaluating single-agent checkpoint inhibitors versus docetaxel.
When you pool together all the trials that looked at a singleagent checkpoint inhibitor versus docetaxel second line, you’ve got to go back in time to look at the subsets of patients who have anEGFRmutation. That was the only group that did better with the chemotherapy versus single-agent checkpoint inhibitors.7
That’s what gave us the best evidence that the checkpoint inhibitors are not a great idea for most patients who have an
EGFRmutation. This is the first thing that we saw about it. Chee Khoon Lee, MD, and colleagues showed that the patient
[with]EGFRwild-type benefited from the checkpoint inhibitor and the patients [with]EGFR[mutations] did better with
Let me be clear. I don’t want people to think that we should never consider checkpoint inhibitors in patients withEGFR,
because there are exceptions. It’s just important to note that patients withEGFRmutations can have variable responses to
checkpoint inhibitors.7We have to think about PD-L1 differently. Patients with anALKmutation don’t benefit, although many of them have high PD-L1 levels. Patients with a ROS1 mutation don’t show a benefit, although they have high PD-L1 levels. For patients withEGFRmutations, some will show a benefit. I have a couple of patients [who] definitely have anEGFRmutation who are benefitting from using a checkpoint inhibitor.
• The patient received SRS while awaiting molecular studies.
• Osimertinib (80 mg) once daily was initiated.
• She experienced a good partial response.
• She complained of headaches and worsening fatigue 10 months after the initiation of osimertinib.
• CT showed 3 new liver lesions
• Brain MRI showed 1 new lesion
• ECOG performance status: 1
What was the patient’s experience on osimertinib?
The patient presented with a solitary brain metastases, but she also had liver metastases. She received SRS while waiting for results of molecular studies. She had high PD-L1 levels at 55%. In addition, she had an EGFR exon 19 deletion. She was started on osimertinib and had a positive initial response.
Unfortunately, it wasn’t a long response. After 10 months she complained of headache and worsening fatigue. The median
[response] is 18 months, but we have seen patients whose
response is shorter.
Do you send out for a repeat biopsy or do you choose liquid biopsy for mutation testing at this point?
If there is something that looks appropriate to biopsy, then we’ll consider it. But in this case, we would send out for a liquid biopsy. It’s not always covered [by insurance], but that’s the game we have to play. However, I expect to get the results back fairly quickly. By the time I set them up for a regular biopsy, I could have the liquid biopsy test back.
What if the mutation testing reveals that the patient has C797S resistance? How does this influence your decision making?
If the patient has C797S, that is a mutation that’s resistant to osimertinib. You could consider a clinical trial for the patient,
but unfortunately, there is nothing clinically available to them. There are some trials evaluating giving first-generation agents, like erlotinib or gefitinib, which can work against C797S. The challenge is if you give either of those agents, you could lose some of the other clones [and could inadvertently provide expansion of resistant variants].
There aren’t clinical trials available that are specific to C797S. [As I mentioned] some first-generation drugs may work
but if the patient has T790M and C797S, no agents work well. If the patient only has C797S, then sometimes erlotinib and
afatinib can work. Those agents are being evaluated, but duration of response [is] not that long.