A panel of experts review two patient cases to elucidate data in the adjuvant setting of non–small cell lung cancer.
At an Around the Practice program hosted by CancerNetwork®, experts spoke about recent treatment advances for patients with non–small cell lung cancer (NSCLC). The panel was led by Charu Aggarwal, MD, MPH, associate director of the Penn Center for Precision Medicine, physician leader of airways malignancies research, and Leslye M. Heisler Associate Professor for Lung Cancer Excellence at Penn Medicine in Philadelphia.
Additional panelists were Benjamin P. Levy, MD, clinical director of medical oncology and associate professor of oncology at Johns Hopkins Sidney Kimmel Cancer Center at Sibley Memorial Hospital in Washington, DC; Harvey I. Pass, MD, Stephen E. Banner Professor of Thoracic Oncology in the Department of Cardiothoracic Surgery and professor in the Department of Surgery at NYU Langone Health in New York; and Heather Wakelee, MD, professor of medicine at Stanford University Medical Center in Palo Alto, California.
Aggarwal: At your institution, what is your approach to biomarker testing for patients with NSCLC?
Wakelee: When a patient presents with newly diagnosed disease, our solid tumor testing is our in-house version of [Cancer Personalized Profiling by Deep Sequencing]. To get faster answers, though, we still do [polymerase chain reaction] assays for EGFR and [fluorescence in situ hybridization] for ALK and ROS [gene mutations]. We have a high percentage of patients who have driver mutations, so that’s critical. We also realize that the tissue testing takes time, so we do liquid biopsies as well. We don’t yet have [Clinical Laboratory Improvement Amendments] for our in-house tests, so we tend to send [them] out. Then, for our patients with early-stage disease, we do reflex testing as well. Obviously, we’re not going to wait a couple of weeks to get the tissue testing back to make a perioperative decision in the preoperative setting. If it’s someone for whom we’re debating whether we will [give neoadjuvant therapy], we’ll try to get the rapid test back before we make a decision in someone who has a high probability [of cancer]. Most of our testing is still tissue testing for the patients with early-stage disease on the resected specimens.
Levy: [Our practice is] similar. For our patients with advanced-stage nonsquamous NSCLC, we try to do both tissue NGS [next-generation sequencing] and a liquid biopsy or plasma genotyping. We know we miss some alterations in tissues that we can detect with [circulating tumor DNA]; [we] can better capture alterations and make an informed treatment decision. We have a 3-fold panel: We do PD-L1 [testing], a fusion panel, and roughly a 300- to 500-gene NGS assay.
Aggarwal: Do you think about other patient or treatment factors when you’re ordering biomarker testing?
Wakelee: As much as we all know that any patient’s tumor can harbor a driver mutation, there are probabilities that are related to the patient, to their medical and smoking history, and to their ethnic background. At least for this idea about whether to do any of that testing before making a decision about perioperative treatment, we do take that into account. If a patient has a very heavy smoking history, the probability that they will have an EGFR or ALK [gene mutation] is relatively low, and the probability they might have actionable PD-L1 is higher. We’re thinking a little differently, but in either case, there’s still a test that might help, and that’s where this idea comes in about knowing PD-L1 and looking for EGFR and ALK [mutations]. There are still so many other drivers that you’re missing, but we don’t know what to do in that setting, nor do we know what to do with ALK or EGFR in the preoperative setting. It makes us less inclined toward a neoadjuvant immunotherapy approach in that setting.
Levy: Clinical phenotypes shouldn’t drive decision-making or testing, and so they need to be testing everyone. There’s some suggestion that we should be testing patients who have squamous cell histology, at least light smokers or never smokers. There are actionable mutations that we can unearth. For patients with advanced-stage disease, we try to do both plasma and tissue. In a plasma-positive [patient], we don’t wait for the tissue. For the patients who are more likely to harbor these alterations, the never smokers, if both the plasma and tissue are negative, I don’t give up. It’s important to recognize tissue heterogeneity and that where the needle is when the tissue is procured may not identify the alteration. There may be an opportunity to rebiopsy these patients and send [the tissue] to a different panel.
Aggarwal: How do you ensure that [consistent] communication is occurring among all the members [on a multidisciplinary team]?
Pass: We have clinic together. In other words, the way we set it up was that the surgeons were going to be at the cancer center. The busy surgeons were at the cancer center along with the pulmonologist on the same floor so that there could be instant referrals [and discussion], not next-day stuff. That gets reinforced at the multidisciplinary conference, where all the pulmonologists, surgeons, medical oncologists, and [nurse practitioners] are there. That’s the way it should be. That way, if it turns out that we have to rebiopsy, we have very talented interventionalists who know how to do robotic bronchoscopy as well as other localization techniques; we have a good track record of being able to get the data for the sequences that need to be looked at. It depends upon a lot of cooperation, but it works very well at our place, in my opinion, especially when you’re trying to develop and get people to go on protocol.
Aggarwal: Is there a typical paradigm for management of patients presenting with early-stage lung cancer?
Pass: In a case like this, you have to have a biopsy beforehand. We will send a patient like this to get interventional EBUS to do what we can to know what the stage is before we enter. This is a big lesion in my opinion, and I’ve always been concerned about those. [Lesions like this] need to be worked up completely. It’d be nice to know what the real histopathology is, but we’re not quite good enough on a biopsy to 100% to say this is micropapillary or solid. We can, however, get some hints.
Aggarwal: If you were seeing this patient today, what would you recommend? Do the results from the CheckMate 816 trial [NCT02998528] play a role?1
Pass: I would advocate strongly that this patient gets induction therapy. It could be induction immunotherapy [IO] plus chemotherapy or induction IO alone. I don’t know which is better at this point; it seems to me that induction IO plus chemotherapy is a standard that will be rapidly adopted, so I would be pushing for that. It could be done safely, as we’ve all learned from [CheckMate] 816 and other trials. The question of how you’re going to treat the patient afterward with maintenance IO is another one to discuss, but I would strongly advocate for this patient to get induction therapy.
Aggarwal: Do you favor adjuvant approaches in this setting? How would you treat this patient today?
Wakelee: This is a case we would talk about now in our tumor board, with the approval of neoadjuvant nivolumab [Opdivo] plus chemotherapy.2 There are definitely pros and cons and I don’t think we fully know [what the best approach is]. We have one randomized phase 3 study in the neoadjuvant setting, which showed a definite benefit of [nivolumab] over neoadjuvant chemotherapy. We have 2 positive adjuvant trials of IO after chemotherapy. Most of the neoadjuvant trials give neoadjuvant [therapy followed by] adjuvant, and we haven’t seen those results yet.
We would certainly talk with the patient. Many patients have strong feelings about whether they would like to go through their surgery first. We do have to acknowledge that with neoadjuvant trials, somewhere between 10% to 15% of patients enrolled don’t go to surgery, whether that’s because they weren’t destined for surgery or it’s something about that neoadjuvant [therapy]. It’s a risk to be mindful of. That’s why it would be important to have a discussion with all the details about the patient and then bring the patient into that decision-making.
Levy: Having treated several patients, both on study and off study with neoadjuvant chemotherapy plus IO, I can tell you that I’ve become a believer in what we’ve seen, at least at the level of tissue analysis at the time of surgical resection. This is shared decision-making with the patient. That’s a tough one because sometimes patients don’t want to participate; they want you to make the call. If that comes up, I encourage a chemotherapy plus IO approach. There are competing strategies here. We do have other data in the neoadjuvant chemotherapy plus IO realm with other trials that look just as promising, and now we have this phase 3 trial. This is again a call for multidisciplinary care. There are a lot of theoretical advantages to using neoadjuvant chemotherapy vs adjuvant independent of the data we’ve seen, so I would opt for chemotherapy plus IO, but I’ll hedge and say that it’s shared decision-making as well.
Aggarwal: What are your thoughts on this?
Pass: [A tumor in station] 10 is difficult. For me, it’s not surprising; it’s a large tumor. You certainly could have micrometastases that are picked up at the time of surgery. This depends upon how efficient you are at doing your lymph node dissection. It’s up to the surgeon to get those nodes so that you know what the stage is. Patients have poor 5-year survivals with T3N0. It can be around 40%, so this patient will need more therapy and to be referred. You’re going to have to look at the final pathology. It’s [adenocarcinoma], and this is one that our medical oncologist would certainly do further NGS on, to give options for treatment.
Aggarwal: Can you discuss the CheckMate 816 trial?
Levy: This was a practice-changing seminal study that looked at roughly 350 patients with newly diagnosed resectable IB to IIIA NSCLC without any known EGFR or ALK mutations or ALK alterations. They were randomized to 3 cycles of neoadjuvant chemotherapy vs 3 cycles of neoadjuvant chemotherapy plus IO, asking the question of whether the addition of IO is beneficial to chemotherapy in the neoadjuvant setting. Patients [received] surgery within 6 weeks post treatment and then they had the option of getting adjuvant chemotherapy or [radiotherapy] as well. This is where things get interesting. The primary end point was [pathological complete response; pCR], and that was defined as 0% residual viable tumor cells both in the primary tumor and the sample lymph nodes.
The [pCR] rate was 24% in the neoadjuvant chemotherapy-plus-IO arm vs only 2% in the chemotherapy-alone arm [odds ratio, 13.94; 99% CI, 3.49-55.75; P <.001]. If you looked at the patients who went on to resection, the [pCR rate] was 30%. We need to keep in mind that two-thirds of patients in this study had stage IIIA disease, which is a large number of patients. Two subsequent iterations of these data have been looking at the surgical approaches. The bottom line is that adding immunotherapy to chemotherapy didn’t seem to disrupt surgery or make surgery more difficult or lead to worse outcomes. In fact, there were more pneumonectomies in the chemotherapy arm than in the chemotherapy-plus-IO arm, and there was a higher chance of doing a lobectomy in the chemotherapy-plus-IO arm vs the chemotherapy-alone arm.
The HR was 0.63 [97.38% CI, 0.43-0.91; P = .005] for EFS [event-free survival] with neoadjuvant chemotherapy plus IO vs neoadjuvant chemotherapy. When you put this all together, this has [received] serious traction. [The pCR] rate is quite high. Based on earlier data that correlate to a better outcome, we now have early snapshots of EFS and OS [overall survival]. [Treatment] didn’t seem to disrupt the type of surgery that was done or safety with surgery. These components create a new standard of care and have allowed us to [look again] at these patients and talk about them in multidisciplinary conferences the way we’ve been discussing on this panel.
Aggarwal: Lead us through the IMpower010 trial [NCT02486718] and the implications of the study?3
Wakelee: With IMpower010, enrolled patients had completely resected stage IB and IIIA NSCLC; most had stage II disease, 12% had stage IB, and 40% or so had IIIA. The patients who were enrolled [in the trial] were stratified on several different factors. They all [received adjuvant] chemotherapy. That’s different from some of the other adjuvant trials. Those who got through chemotherapy and still wanted to keep going on treatment were randomized. More than 1000 patients got atezolizumab [Tecentriq] or best supportive care. There was no placebo in the study. The study also had a complicated statistical testing plan.
In this trial, we had initially planned to use the SP142 [assay], but then as [testing evolved], we realized that was very much the outlier. The analysis was done with the SP263 [assay]. A caveat, but an important one, is that 55% of patients enrolled in this trial had PD-L1 expression based on the SP263 assay. The first group tested had stage II to IIIA disease with PD-L1 expression, with a disease-free survival [DFS] hazard ratio of 0.66 [95% CI, 0.50-0.88; P = .0039], which is highly statistically significant. It’s in that subgroup that we have the FDA approval for atezolizumab.4
The next stratification that was looked at were patients with stage II to IIIA disease, regardless of PD-L1 expression. The DFS was still statistically significant, with a hazard ratio of 0.79 [95% CI, 0.64-0.96; P = .020], but it wasn’t as good. When we [looked at] patients with stage IB disease, there hadn’t yet been enough events at the data cutoff, [with an HR] around 0.81 [95% CI, 0.67-0.99; P = .040]. No OS data are definitive, [but] trends are looking positive and the curves are starting to separate even though the drug was stopped after 1 year.
When we broke it out further and look at patients whose tumors have no PD-L1 expression, there’s no benefit in this trial. That’s important because to me, we [want to] find a treatment that works for everybody, but if we don’t, we need to be able to figure out who gets the benefit and who doesn’t. We’ve all seen patients who’ve had significant adverse effects from IO, so you want to make sure that you’ve got a higher probability of benefit before you’re giving these drugs to everybody.
Aggarwal: Are you more likely to choose neoadjuvant chemotherapy plus IO for a patient with PD-L1 negative disease, thinking that you may not be able to come in with adjuvant IO?
Wakelee: We know from the metastatic disease setting that when we combine chemotherapy and IO, it works better than either alone. Even though PD-L1 is important, it’s less important [than in the metastatic setting], and we see this with CheckMate 816.
When you look at the subset data based on PD-L1 expression, all patients seem to benefit, but there is more benefit with higher PD-L1 levels. If I have a patient who has no PD-L1 expression, I’m going to look at the data from the IMpower010 [trial]. Adjuvant IO wasn’t very helpful there when you give it alone vs with chemotherapy. The PD-L1 isn’t as important, so maybe that’s a better approach for this patient and for the stage IIIA disease. That’s where we’re more inclined toward neoadjuvant chemotherapy, anyway. If I have a patient with stage IIIA disease and no PD-L1 expression, I’m going to be more inclined toward chemotherapy plus nivolumab in the neoadjuvant setting than I am toward an adjuvant IO approach.
Aggarwal: How would you manage this situation?
Pass: If you found that you had stage IIIA disease, I’m not so sure that you would be leaning toward giving IO since they have EGFR mutation. You may want to think about induction chemotherapy that we’ve done before, with or without radiation. It goes back to the old days for me, but certainly it shows me the importance of staging. If you find that the lymph nodes are negative, you could operate on the patient and completely resect the [tumor], and the patient would be a wonderful candidate for osimertinib [Tagrisso]. From my standpoint, [we need to receive] complete mediastinum staging by our interventional pulmonologist and go from there.
Aggarwal: Please discuss the ADAURA trial [NCT02511106].5
Levy: Osimertinib is a third-generation [tyrosine kinase inhibitor] that binds irreversibly to the intracellular domain of the tyrosine kinase that’s being constitutively activated by a mutation in EGFR either exon 19, 21, or T790M, and it can also be 1 target for osimertinib. We leverage osimertinib in the curative-stage setting or in the adjuvant setting.
The ADAURA trial looked at patients with stage IB through IIIA resected lung cancer. All had [EGFR mutations in] exon 19, L858R, or exon 21. [Tumors] were resected and [patients were] given adjuvant chemotherapy. There were roughly 700 patients, and we saw hazard ratios that we don’t normally see for the primary end point of DFS [in patients with stage II to IIIA disease], at 0.17 [99.06% CI, 0.11-0.26]. These curves diverged early on. If we look at the overall population, stage IB to IIIA disease, the hazard ratio is still very low [0.20; 99.12% CI, 0.14-0.30; P <.001], which is statistically significant.
The higher the stage, the more likely you were to benefit [from treatment]. We haven’t seen OS data yet, but we’ve seen early snapshots.