Emerging data in the neoadjuvant setting and the expansion of testing in standard practice have presented questions surrounding screening for and developing strategies from variant information in EGFR-mutated resectable non–small cell lung cancer.
EGFR-mutated resectable non–small cell lung cancer (NSCLC) has been proven as an actionable disease with several approved targeted therapies for those with advanced disease. Survival data from pivotal trials, namely from the phase 3 ADAURA study (NCT02511106), have also afforded investigators tools to develop optimal postsurgical treatment strategies in early-stage disease. However, emerging data in the neoadjuvant setting and the expansion of testing in standard practice have presented questions surrounding screening for and developing strategies from variant information.1-3
Expressed in most epithelial cancers, EGFR is associated with poor prognosis in NSCLC.1 With nearly 2 decades of research in evaluating targeted agents against chemotherapy and identifying resistance mechanisms, investigators have touched the tip of the iceberg of the role of EGFR in patients as they seek to enhance outcomes and identify early and accurately those who will benefit from targeted therapy in the adjuvant and neoadjuvant settings.4
These gaps in knowledge necessitate multidisciplinary conversations at various points of care with interventional pulmonologists, surgeons, and clinical oncologists. Addressing the various aspects of testing—obtaining tissue or liquid biopsy samples, turnaround times, interpretation of results, assessing for FDA-approved pathways or clinical trial enrollment—not only determines a pathway of care, but adds to mounting pressure to ensure that the patient understands the rationale behind the results and anticipated approach to care.
“The success of the treatment plan is directly proportional to the amount of time and conversation we have with the patient the first time we shake their hands,” Gustavo Cumbo-Nacheli, MD, FCCP, DAABIP, an interventional pulmonologist said in a recent OncLive Peer Exchange featuring a panel of thoracic experts across disciplines. “This is a high-level discussion with experts, and you all provide treatment, whereas the surgeon and I, we do the pretreatment, which is diagnostic. The patient doesn’t know when they show up in the clinic, what’s going on, and what’s going to happen. They don’t know anything about personalized medicine or nodule risk stratification. At the very beginning, we must prepare the patient that the biopsy [result] is going to put a first name into the tumor they have, then there’s testing that is going to put the last name. Next, there’s going to be waiting time to know what preoperative treatment will be, and even if they were to have surgery, there may be other treatments that may happen before and after. Communication with the patient is very important to secure the success of a program.”
Ensuring the patient informed sets expectations for coming opportunities for testing and what those results may mean for the likelihood of success with targeted treatments. Cumbo-Nacheli added, “We are going to have  opportunities of checking whether there are any type of genetic rearrangements or DNA abnormalities that would allow [the patient] to get a specific treatment. We set the stage for them to go in-depth with discussions about different treatment options, and then, hopefully, we will be able to explain what true value care is provided to them, which is good-quality medication, better outcomes, and lower cost overall. We are openers for [medical oncologists] to be closers, if you will, but it’s very important that...after a positive-screening CT [result], after a preresection, the pulmonologist or the interventional pulmonologist discuss that testing is coming. It may take a little bit, but give us the time to get the chest set right for us to move forward.”
For patients who receive a diagnosis of early-stage and late-stage disease, barriers to timely testing are similar—turnaround times delay starting treatment. Reflex testing is commonplace in most institutions, but even immunohistochemistry (IHC) analysis for markers such as PD-L1 are taking weeks to return, moderator Zofia Piotrowska, MD, MHS, explained anecdotally. “When you have a patient [with] newly diagnosed [disease] and you’ve just told them they have stage IV cancer, they’re super anxious to start treatment, as we all would be,” she said. “Now you’re telling them, ‘I’m going to order molecular testing [for] PD-L1 [analysis]; this might take on average a couple of weeks for that to come back.’ ”
She also noted that in addition to the timeline, there is an evolution in testing. “The landscape is changing. We used to order a single gene PCR [polymerase chain reaction] test, and that’s no longer sufficient,” Piotrowska said.
Jay Moon Lee, MD, said that although PD-L1 is often the first result to arrive, acting on it prior to receiving next-generation sequencing (NGS) results may not be in the best interest of the patient. Ashish Saxena, MD, PhD, agreed: “Because it’s an IHC, [PD-L1 results] will come back first, [expression] can be high and there’s a reflex: If there is high PD-L1 [expression], the patient should get immunotherapy right away. But I try to tell my patients that we’re also looking for targeted therapies and the immunotherapy likely won’t work well no matter what your PD-L1 [expression] is if you’re [showing] EGFR-positive or ALK-positive [results].”
Lee added, “When you look at the IMMUNOTARGET registry, over 50% of patients with actionable, targeted mutations and oncogeneaddicted NSCLC [have] PD-L1 positive [results].... We shouldn’t be fooled by PD-L1 and the meaning of PD-L1 in an oncogene-addicted NSCLC is totally different. You look at patients with EGFR-mutated [disease] who have high PD-L1 [expression] and the data that are coming out about what’s happening in the tumor microenvironment, those are exhausted T cells. You can throw as much immunotherapy at them [as you want], but most of the time, it’s not going to work.”5
Whenever systemic therapy is in consideration, Lee said that NGS is important. “Even though we don’t have FDA-[approved] targeted therapies for [most] of those targets, [as a surgeon], what I use the data for is when I am considering a chemotherapy/immunotherapy regimen in the neoadjuvant setting,” he said. “If a patient has an oncogene-addicted NSCLC, we know from the metastatic disease setting that their response rates are very low. This affects my thinking in terms of whether that patient is going to be medically inoperable in the sense that that [the] tumor that I could have taken out by a lobectomy, may turn into a bilobectomy or a pneumonectomy and the patient becomes no longer operable or it affects the resectability [of the tumor]. Maybe I lose that window to do that sleeve resection. Or maybe it’s close to that vertebral body and close to that T4 structure…where the resection and complete resection rates are lower. Knowing the metastatic disease data, if the response rates are marginal then I’m going to say [to] take that patient to surgery first and forgo the neoadjuvant chemotherapy/immunotherapy, even though we don’t have approval in that setting yet.”
Saxena noted that ordering NGS testing as soon as possible affords opportunities to act soon and provide a roadmap of the patient’s disease throughout treatment. “[Results] are not something that you’re going to get in a day. Whether it’s [for a patient with] early-stage [or] late-stage disease, having it done at the time of diagnosis is important. I often test again and look for possible mechanisms of resistance [at relapse or in a patient with later-stage disease]. This is also where I think liquid biopsies make things easier because especially in the advanced-stage setting where a patient might not be amenable to getting another biopsy, that information can help. Sometimes you find another targetable mutation or, more likely, maybe there’s a clinical trial, such as a basket trial, open and available for patients who have different mechanisms of resistance.”
Lee said that although every institution has a different workflow, testing early and with NGS is “critically important. At UCLA [in Los Angeles, California] we do reflexive NGS testing from the biopsy specimen. Often, at the time of biopsy, we don’t know what stage of NSCLC the patient may have. The NGS testing has relevance if the institution has clinical trials tied to targeted therapies in the earlystage setting. It has relevance in terms of efficacy or lack of efficacy with immunotherapy in the early-stage setting. And finally, [NGS results] can determine which patients should go [to the] ADAURA trial regimen.”
Among those with resectable disease, adjuvant treatment with the third-generation EGFR tyrosine kinase inhibitor (TKI) osimertinib (Tagrisso) elicited clinically meaningful disease-free survival outcomes for patients with stage II to IIIA EGFR-mutated NSCLC.2,3 Findings from the ADAURA trial, which randomly assigned patients 1:1 to receive osimertinib at 80 mg or placebo until disease recurrence supported the approval of the agent for patients with EGFR exon 1 deletions or exon 21 L858R mutations in December 2020.6
In updated overall survival (OS) analysis, the 5-year OS rate for patients with stage II to IIIA disease who received osimertinib (n = 233) was 85% (95% CI, 79%-89%) vs 73% (95% CI, 66%-78%) with placebo (n = 237). The HR for death was 0.49 (95% CI, 0.33-0.73; P < .001). For all treated patients (stage IB to IIIA), the 5-year OS rate with osimertinib (n = 339) was 88% (95% CI, 83%-91%) vs 78% (95% CI, 73%-82%) with placebo (n = 343). The HR was 0.49 (95% CI, 0.34- 0.70; P < .001).7
Although an EGFR-negative result is not required for patients to receive treatment with neoadjuvant chemotherapy/immunotherapy, this treatment is optimal treatment for patients who do not present with EGFR or ALK mutations in both the neoadjuvant and adjuvant settings. For example, data from the phase 3 CheckMate 816 trial (NCT02998528) supported the approval of adjuvant nivolumab (Opdivo) plus platinum-doublet chemotherapy.7,8
Patients with EGFR and ALK mutations were excluded from the study. Event-free survival (EFS) served as the primary end point, with the chemotherapy/immunotherapy regimen (n = 179) eliciting a median EFS of 31.6 months (95% CI, 30.2-not reached) vs 20.8 months (95% CI, 14.0-26.7) with chemotherapy alone (n = 179; HR, 0.63; 95% CI, 0.43-0.91; P = .0052).6 The pathologic complete response rates were 24.0% (95% CI, 18.0%-31.0%) vs 2.2% (95% CI, 0.6%-5.6%), respectively.5 “You want a quick turnaround on PD-L1, ALK, and EGFR status, and [those results are] mainly driven by IHC; however, there are limited panels where you can get a quick turnaround on EGFR,” Lee said. “So, right now we’re in this divisive situation where we’re doing both NGS testing that takes a couple weeks and then the quick turnaround test that we need to act upon right away.
EGFR in the neoadjuvant setting is a topic of ongoing discussion as data begin to emerge. “The elephant in the room is, what do we do with neoadjuvant [EGFR TKI] therapy? Is there a role here? And I think we’ve all started to have these discussions in the tumor board,” she said, adding that having a positive EGFR result is essential, but that this setting complicates decision-making. “Let’s say that before you even get to a decision about surgery, you know that the patient has an EGFR mutation. It’s a tough question as to what to do with those patients.”
Data presented that the 2023 American Society of Clinical Oncology Annual Meeting showed that neoadjuvant osimertinib may have a role in the space9; however, as Lyudmila A. Bazhenova, MD, highlighted, “In neoadjuvant EGFR-inhibition trials, we’re just barely starting to scratch the surface.”
In the phase 2 study, patients with surgically resectable NSCLC (stage I-IIIA) with a documented EGFR mutation of either exon 19 deletion or L858R were enrolled to receive 1 to 2 months of neoadjuvant osimertinib prior to resection.9 Among 17 evaluable patients, the pathological response rate was 48% with no patients achieving pathological complete response and 15% of patients having a major pathologic response. The median duration of osimertinib treatment was 56 days (IQR, 41-62). During the treatment period, 44% (n = 4 of 9) had lymph node downstaging. The median diseasefree survival was 32.4 months (95% CI, 25.9-not reached).9
“What we have learned from neoadjuvant studies is, No. 1, that patients do respond to targeted therapy, No. 2, that in some patients, we see nodal clearance, and No. 3, the major pathologic response and complete pathologic response are nowhere near what we see with chemoimmunotherapy,” Bazhenova said. “In my practice, outside of the clinical trial, I am not giving neoadjuvant TKIs for any oncogenic-driven tumors…. What I took from [the phase 2 data] is not the efficacy data, but the tremendous ability to learn about resistance to targeted therapy. The investigators found 2 potential biomarkers, but it was for a small number of patients. We need more studies.”
Bazhenova noted that educating the patient becomes the priority following a diagnosis of any-stage disease. “Education, education, education. You tell the patient you can give them a mediocre treatment now, or the best treatment in 2 weeks,” she said, adding that setting expectations is vital. “Explain the risk: (a), it might not work, (b) if [for example,] I give the patient immunotherapy and later discover an EGFR mutation. [The treatment] increases the risk of complications [such as pneumonitis], which could make the patient not eligible for any EGFR TKI. It is important to guide them through the decision and make sure they understand why it’s important for us to wait. And if we can’t—in my experience, the patients who cannot wait are actually pretty rare, no more than 2% or 3%—then in that situation, we do have an alternative, which is a platinum-doublet [chemotherapy], which does work in patients with lung cancer.”
Although education for the patient is vital, inside and across clinics communication is equally important. “There has to be communication because there are patientrelated, system-related, and physicianrelated barriers to this,” Cumbo-Nacheli said. “[Communication must occur] whether it’s a multidisciplinary team weekly or synchronous EMRs [electronic medical records] with oncology surgeons or pulmonologists…. Collective intelligence is the highest form of decision-making. We must set the stage for success with effective communication with other team members and the patient.”
Nicholas Rohs, MD, said that within education is the idea that implementing data, such as those from ADAURA, is what is being left on the table in terms of other available options such as chemotherapy. “It’s the Wild West out here,” he said. “My unanswered question is, what are we going to do with all these data coming together all at the same time? How do we put it together and how are we selecting the right patients for this therapy?”