Gubens Details the Management of Patients With EGFR+ NSCLC in Case Study

Nov 27, 2019

During a Targeted Oncology live case-based peer perspectives discussion, Matthew A. Gubens, MD, MS, reviewed with a group of physicians the localized and systemic treatment options available for the management of patients with non–small cell lung cancer.

Matthew A. Gubens, MD, MS

During aTargeted Oncologylive case-based peer perspectives discussion, Matthew A. Gubens, MD, MS, reviewed with a group of physicians the localized and systemic treatment options available for the management of patients with non—small cell lung cancer (NSCLC). Gubens, an associate professor of thoracic medical oncology, University of California, San Francisco, Thoracic Surgery and Oncology Clinic, Helen Diller Family Comprehensive Cancer Center, explained these options based on a case scenario of a patient withEGFR-mutant NSCLC.


A66-year-old Caucasian woman presented to her primary care physician complaining of visual disturbances, nausea, fatigue, and sporadic headache. Her medical history was notable for hypertension, which was managed on candesartan and hyperlipidemia, which was managed on simvastatin; she had no smoking history.

On physical exam her blood pressure was 148/70 mm Hg and she demonstrated decreased breath sounds in the lower left lobe. Her complete blood count and chemistries blood work werewithin normal limits.

A brain MRI demonstrated a 10-mm right parietal mass at the gray-white junction with vasogenic edema. A CT of the chest, abdomen, and pelvis revealed a 3.4-cm mass in the lower left lobe and several small liver nodules. A CT-guided transthoracic needle biopsy of the lung lesion showed grade 2 adenocarcinoma, acinar subtype. She was diagnosed with stage T2aN0M1c NSCLC and had an ECOG performance status of 1.

Her PD-L1 expression was determined to be 55% on tumor cells by the PD-L1 IHC 22C3 pharmDx test. Molecular panel testing revealed anEGFRexon 19 deletion.

Targeted Oncology™: How would you treat this patient if she tested positive for aROS1mutation?

GUBENS:For aROS1mutation, I would probably use crizotinib [Xalkori] or entrectinib [Rozlytrek], which has aROS1label and probably has better brain penetration, if brain [metastases] are an issue. I tend to use crizotinib first if there [are] no brain [metastases] and entrectinib if there [are], but at least there are more options than there used to be.

What drugs can clinicians use in this setting, and are any options better than the rest?

Osimertinib [Tagrisso], especially outside the United States, is not well paid for. It’s an expensive drug, and it’s not yet approved in some of the other countries that we work with.

There have been compelling studies where investigators have added bevacizumab [Avastin] or ramucirumab [Cyramza], [and results have shown] strong progression-free survival [PFS] results. Compared with erlotinib [Tarceva], you’d expect [a PFS of] 8 to 10 months, but you’re seeing this long PFS [instead].1

At the American Society of Clinical Oncology meeting, [investigators] showed gefitinib [Iressa] plus chemotherapy had improved [not just] PFS compared with gefitinib alone but [also] overall survival [OS].2

It’s tough because I think erlotinib is inferior.

Are there drawbacks to using chemotherapy with a tyrosine kinase inhibitor [TKI]?

A concern I’ve heard is that it makes us nervous to lose one of our [therapy options]. If you give chemotherapy with the TKI and the patient progresses, what do you have [if that doesn’t work]? That’s always in the back of my head, though that OS benefit [is something to consider].

Then the other thing, when we talk to patients, is that there is something so different and unique about an oral therapy that I referred some of our patients. It would take a pretty [big efficacy] difference to convince them that instead of that pill, they still have to come to the clinic every 3 months for their infusion. That’s more of a philosophical thing.

Please describe the FLAURA trial.

The FLAURA trial [looked at] treatment-naïve patients with non—small cell lung cancer [NSCLC]. They had to have one of the 2 canonical mutations, so the exon 19 deletion or the L858R exon 21 mutation. They allowed [patients with] stable central nervous system metastases. [This] speaks to the point that they’re confident about osimertinib’s ability to get through the blood-brain barrier.1

[The brain metastases] have to be asymptomatic. So [a patient] could have 10 as long as they were small and asymptomatic.

The patients were randomized to either the osimertinib arm or the erlotinib or gefitinib arm. Patients were stratified by theEGFRtype of mutation and by Asian or non-Asian populations because there have been some differences [observed between groups] in similar studies. Patients were treated until disease progression; crossover was allowed from erlotinib or gefitinib to osimertinib in the setting of provenEGFRT790M mutation because that was always the credential to get to osimertinib in the second line. They did build that into the crossover.

The primary end point was PFS, and secondarily they also looked at response rate, duration of response, and OS.

What were the efficacy results for this trial?

There was an impressive improvement in median PFS. There was [improvement compared with the] control arm because 10 months is what we figured we got out of our first- and second-generation agents. The median PFS with osimertinib was almost doubled [at 18.9 months]. Keep in mind that when [patients] had T790M, which is about 50% to 60% of the patients on this first-generation agent, only those 50% or 60% of patients received [standard of care], and they had an average PFS of 10 months in that setting.

I think the skin adverse effects [AEs] and gastrointestinal AEs have been impressively better.

OS was reported at the European Society for Medical Oncology 2019 Congress, and this is the first TKI-versus-TKI trial in theEGFR-mutant space to show a median OS benefit [with osimertinib (38.6 vs 31.8 months; HR, 0.799; 95.05% CI, 0.641-0.997;P= .0462)].3

We always think about that OS benefit being watered down by crossover, but there was still an OS benefit to leading with the best agent. Even if a patient crossed over, got chemotherapy, and maybe even got immune therapy, there was still OS benefit in the end.

Is osimertinib a part of the guidelines for the treatment of patients with NSCLC?

This was enough to get the National Comprehensive Cancer Network [NCCN] to put osimertinib on their list of [recommended] agents.4One thing that they talk about is patients with PD-L1—positive,EGFR-mutated [disease] and the use of immunotherapy. NCCN makes a very cautious note that says if there is a presence of an oncogene, that might be a reason not to give a PD-1 inhibitor as a single agent.

In the second line, we know that even if you have exhausted your TKI and first-line chemotherapy, patients withEGFRmutations tend to underperform at every level of PD-L1 [inhibitor therapy]. That’s this moment of caution to say, “Where do we put the checkpoint inhibitor?” A lot of us would put it after you’ve exhausted other options.

Are any studies looking at immunotherapy in a first-line setting for this patient population?

The investigators [of a phase II study] took patients [who were TKI naïve] withEGFR-positive disease and a PD-L1 status of at least 1% and enrolled them directly onto immunotherapy with single-agent pembrolizumab [Keytruda].5

They planned to [recruit] 25 patients [who] had at least 1% expression of PD-L1. The [patients received] standard dosing, and the investigators were looking at overall response rate. It was a 2-stage trial, [but] they stopped at 11 patients because only 1 of the 11 had an objective response. That patient was a lab error; they weren’tEGFRpositive in the end. Interestingly, 73% of the patients had high PD-L1 [≥50%], a level where a lot of us could say treatment with pembrolizumab is going to work well.

Not a singleEGFR-positive tumor with a PD-L1 expression over 50% had response to single-agent pembrolizumab. Then worse, they had 2 deaths within 6 months of enrollment, and one of them was a pneumonitis case.

Do you worry about pneumonitis in other drugs for these patients?

Giving carboplatin and pemetrexed and holding off on [pembrolizumab is] my approach. Part of the reason is that this [phase II study is] a cautionary tale because osimertinib, erlotinib, and gefitinib all have their own risk of pneumonitis. There’s a little higher [risk] than with second-generation agents, as well as immune therapies. Even if you stopped the immunotherapy, [whether it’s] durvalumab [Imfinzi], pembrolizumab, or atezolizumab [Tecentriq], the half-life is long, so you’re giving them together and you’re compounding that risk of pneumonitis.

I think the other cautionary tale is from the company that makes durvalumab. The investigators did a trial called TATTON [NCT02143466], which looked at osimertinib combinations. One of the combinations was osimertinib in the first line with durvalumab. They didn’t see additive pneumonitis. They saw multiplicative pneumonitis, and they had to [suspend enrollment in this arm of] the trial early because they had about 30% significant pneumonitis.6

Not every TKI acts like that, but I think in this patient population, you want to be careful in your eventual use of the immunotherapy. I’m not saying they should never get it, but be mindful of that, and do not combine them.

What second-line studies have had an impact on this patient population?

There’s a meta-analysis of patients who got [a checkpoint inhibitor] in their second line of chemotherapy. These are patients who got EGFR treatment, then received a platinum doublet, and then they received these checkpoint inhibitors. These were the first trials for docetaxel versus a PD-1 or PD-L1 inhibitor that were studies. They did these meta-analyses to look at theEGFRmutation in patients on those trials. These are the studies that got nivolumab [Opdivo], pembrolizumab, and atezolizumab approved.7

The OS HR was 0.69.EGFRwild-type [tumors did] better than those that were mutated. TheEFGR-mutated tumors had an HR of 1.11, so there was a signal that single-agent [therapy] was not going to be especially effective in these patients. These patients may have low tumor mutational burden, and they aren’t inflamed smoker patients. There may be other biological reasons for howEGFRinteracts with the PD-L1 axis. HR 1.11 isn’t 1.8, so I wouldn’t say I would never give a patient the immunotherapy and docetaxel. But it is not the [best treatment] either. [The checkpoint inhibitors were] underwhelming as single agents, even in the end line.

The graphical meta-analysis showed that for the most part, OAK [NCT02008227], CheckMate 057 [NCT01673867], and KEYNOTE-010 [NCT01905657] had HRs favoring docetaxel in this patient population. KEYNOTE-010 [favored the PD-1/PD-L1 inhibitor] a little bit, but the meta-analysis showed that the HR was over 1.11 [for all studies].

CASE (continued)

The patient received stereotactic radiosurgery (SRS) while awaiting molecular studies. Osimertinib at 80 mg once daily was then initiated and she experienced a good partial response.

She complained of headaches and worsening fatigue 10 months after initiation of osimertinib. A CT showed 3 new liver lesions, and a brain MRI showed 1 new lesion. Her ECOG performance status was 1.

Would you normally suggest SRS or surgery for a patient like this?

The neurosurgeons at my institution won’t usually do surgery. They don’t usually do resection unless [the tumor is in a] bad location and there [are] symptoms, but this was 10 mm, so it was accessible.

Our group has been doing more SRS. The textbook would say whole brain [radiation] is reasonable, even in the first place, but I think there is still this idea that SRS has more currency now and it’s situational. If I get 10 new metastases, my gamma knife team will say, “No, you need to do whole brain [radiation] or get her on a better systemic therapy.” It depends on the patients’ symptoms and on the volume, but we have been stereotactically spot welding as well. [The patient should] stay on osimertinib. I think there is consensus that if there is systemic control, stay on the drug that is giving a systemic control with good quality of life.

When is a repeat biopsy appropriate?

I see the arguments for and against doing a stage III [biopsy], but if a patient has stage III cancer and the [tumors] reappear [only] 4 years later, we’ll have technology for testing that is a lot better 4 years later. So do I spend the [the money] then or early on? But 4 years out, I probably would have been a little more formal and approved [the biopsy]. Whereas if it happened 3 months later and it’s the same process. I think it also depends on the accessibility of a tumor, how morbid that would be. That also might be a point where I would think about a liquid biopsy, which would answer the question noninvasively and quickly.

What are the therapeutic options for this patient now?

My colleague would never give erlotinib with bevacizumab up front, but on early indolent progression, she would add in the bevacizumab. One of the things being studied is [standard of care versus bevacizumab].

I probably wouldn’t use it off label. I’d probably add the bevacizumab to chemotherapy. [There are] TKIs or local therapy, depending on the burden of disease. Carboplatin, pemetrexed, and pembrolizumab have no EGFR data, but it’s being done a lot because we like carboplatin and pemetrexed for these patients. The IMpower150 [therapy of atezolizumab, bevacizumab, and chemotherapy] is an option. I have used it a couple [of] times, but it requires a long time in the chair, and it’s expensive.

Angiogenesis inhibitors are not wrong. I see a lot of insurance [denials] because bevacizumab is still expensive. Maybe now with the biosimilars, it’ll be a different insurance situation.

If a patient has limited site progression on immunotherapy, do you use radiation or continue immunotherapy as treatment?

Radiation may reinvigorate the immune system to develop more neoantigens and get some response salvage back. So whether it’s immunotherapy, a TKI, or even occasionally chemotherapy, if [the patient has] a long benefit and then oligo progression, especially if it’s a single site, I like the idea of radiating and continuing. We’re getting more and more successful with insurers paying for it.

What kind of AEs have you seen when using immune checkpoint inhibitors after the TKIs?

On IMpower150, very few. There are not many patients [who] are given immuno-oncology right after completing osimertinib, so my sample size is pretty low. I’ve had a couple of nasty [standard-of-care] pneumonitis cases, and those can get ugly, so I’m nervous about them, but I haven’t seen many yet.


  1. Soria J-C, Ohe Y, Vansteenkiste J, et al; FLAURA investigators. Osimertinib in untreatedEGFR-mutated advanced non—small-cell lung cancer.N Engl J Med. 2018;378(2):113-125. doi: 10.1056/NEJMoa1713137.
  2. Noronha V, Patil VM, Joshi A, et al. Gefitinib versus gefitinib plus pemetrexed and carboplatin chemotherapy inEGFR-mutated lung cancer [published online August 14, 2019].J Clin Oncol. doi: 10.1200/JCO.19.01154.
  3. Ramalingam SS, Gray JE, Ohe Y, et al. Osimertinib vs comparator EGFR-TKI as first-line treatment for EGFRm advanced NSCLC (FLAURA): final overall survival analysis. Ann Oncol. 2019;30(suppl 5; abstr LBA5_PR). doi: 10.1093/annonc/mdz394.076.
  4. NCCN Clinical Practice Guidelines in Oncology. Non-Small Cell Lung Cancer, version 1.2020. National Comprehensive Cancer Network Published November 6, 2019. Accessed November 19, 2019.
  5. Lisberg A, Cummings A, Goldman JW, et al. A phase II study of pembrolizumab in EGFR-mutant, PD-L1+, tyrosine kinase inhibitor naïve patients with advanced NSCLC.J Thor Oncol.2018;13(8):1138-1145. doi: 10.1016/j.jtho.2018.03.035.
  6. ELCC 2016 news: osimertinib combined with durvalumab in EGFR-mutant non-small-cell lung cancer. European Society for Medical Oncology website. Published April 15, 2019. Accessed November 20, 2019.
  7. Lee CK, Man J, Lord S, et al. Clinical and molecular characteristics associated with survival among patients treated with checkpoint inhibitors for advanced non—small cell lung carcinoma: a systematic review and meta-analysis. JAMA Oncol.2018;4(2):210-216. doi: 10.1001/jamaoncol.2017.4427.

Related Content