Masahiro Tsuboi, MD, a surgeon at the National Cancer Center Hospital East in Japan, discusses findings from the phase 3 ADAURA trial in 

-mutant non–small cell lung cancer (NSCLC).

In the double-blind, randomized, placebo-controlled, phase 3 trial, a total of 682 patients with completely resected stage IB to IIIA NSCLC were stratified by stage, 

 mutational status, and race. Patients were randomized 1:1 to receive 80 mg of osimertinib (Tagrisso) once daily or placebo once daily.

The primary end point of the trial, which was disease-free survival in patients with stage II to IIIA disease, was met, Tsuboi says. In these patients, treatment with osimertinib following complete tumor resection reduced the risk of disease recurrence or death by 83% versus placebo, which garnered a lot of excitement in the space, Tsuboi concludes. 

Videos

Masahiro Tsuboi, MD, a surgeon at the National Cancer Center Hospital East in Japan, discusses findings from the phase 3 ADAURA trial in EGFR-mutant non–small cell lung cancer. 

Dr. Tsuboi on Findings From the Phase 3 ADAURA Trial in EGFR+ NSCLC

 We have seen in the ADAURA study that it is possible to change the natural history of an oncogenic-addicting non–small cell lung cancer, not only in the early and late stages of the disease. This is a paradigm change in the treatment of advanced non–small cell lung cancer. These findings have a clear implication for obtaining a genomic profile in all the non–small cell lung cancer patients with non-squamous cell carcinoma, including early stage patients. We will achieve a lot of new advances by giving targeted therapy to these patients after surgical resection and after adjuvant chemotherapy if necessary. We still have many issues to resolve, especially the optimal duration of treatment. For example, should we give only 3 years of treatment, or should we maintain the drug for a longer time until relapse? This is an open and important question to address in the future.

 In 2020, adjuvant chemotherapy and platinum-based chemotherapy are still the standard of care, whatever the molecular status of the alteration, even in the 

-mutated population. Because of the ADAURA trial, disease-free survival can be considered a new standard of care in this population. There are other things, but it’s probably too early… We should keep in mind that in this population who have 

 mutations…from metastatic disease evolution. Trying to prevent the brain evolution for this population with 

 mutations is important and something that has been shown in the ADAURA trial.

One percent of patients who received osimertinib experienced brain evolution. Ten percent of patients who did not receive osimertinib achieved brain evolution. This is really important, and if you compare both arms, it was 11% of disease evolution in the osimertinib arm. It was around 40% in the placebo arm. We completely changed the landscape of evolution by reducing the probability of disease evolution and the probability of developing brain metastases in this population. There is no reason not to test patients after surgery to know if they have 

 mutations, and we should discuss this with patients who receive this type of adjuvant treatment with osimertinib.

 We still need to fully analyze and discuss the introduction of the osimertinib adjuvant therapy for patients with stage IB disease. Because stage IB in the ADAURA trial is…divided into stage IB and IIA in TNM [tumor, node, metastases staging system]. Considering the number of patients with stage IB disease who are cured by the surgery alone, it is strange to conclude that the group with different prognoses can be grouped together… [There are] 2 surrogate end points for the overall survival in this population. What is the result of the first-generation TKI [tyrosine kinase inhibitor]? It’s negative in this regard. It will be interesting to see if the scenario is similar to the chemotherapy…

Secondly, there is the patient selection. Considering there are patients who are cured by surgery alone, …we should explore the direction of…adjuvant therapy to include those who are at a higher risk of the recurrence using that…science. I’m particularly interested in the ctDNA [circulating tumor DNA] intervention studies with MRD [minimal residual disease]. We also discussed the other issues, such as combination treatment with chemotherapy, optimal treatment duration, and the cost of new drugs. There seems to be a wave of precision medicine in postoperative care. Adjuvant treatment aims to control micrometastases, which seems to be shifting from the postoperative setting alone to perioperative treatment. It is important to deliver perioperative treatment to those who can be expected to respond to the treatment based on the blood samples, biopsy specimen, or resected specimen. A high-risk benefit balance should be considered when determining the treatment plan for each patient, given that some patients may be cured by surgery.

Paradigm Shift for Early EGFR+ NSCLC

 ADAURA was a practice-changing trial. The magnitude of DFS [disease-free survival] improvement in this trial, especially with the reduction in the brain metastases rate, is enough to change our practice, even if we do not have yet the certainty of overall survival benefit. The proportion of patients experiencing recurrences in the control arm, especially for patients with stage II and III disease and despite the use of adjuvant chemotherapy for approximately 75% of these patients, is consistent with the clinical practice. This proportion of patients with recurrences is very high, and that justifies prevention of these relapses due to the clinical impact, especially for the brain metastasis and the quality of life.

Today the use of osimertinib in the adjuvant setting after surgical resection of 

-positive non–small cell lung cancer will change the treatment paradigm for these patients. As I said, I do not think we have to drop adjuvant chemotherapy if indicated because ADAURA included patients receiving adjuvant chemotherapy if they were fit to receive this treatment. In the future, we need more follow-up for this study, especially for the overall survival but also about some data on the treatment for recurrences in the control arm. What proportion of patients will receive osimertinib at the time of disease relapse vs another TKI [tyrosine kinase inhibitor] vs no TKI at all, or more treatment? In the experimental osimertinib arm, a critical point will be the genomic characteristics of the recurrent disease and the treatment of relapses.

This is a critical point for considering the overall survival benefit. This benefit will be likely if the relapses in the osimertinib arm remain mostly sensitive to osimertinib. We can make them parallel with the SELECT trial, in which the patient received 2 years of adjuvant treatment with erlotinib; approximately 40 of 100 patients included in this single-arm study had relapses. The median time for these 40 patients was approximately 2 years, and the majority of patients were treated with erlotinib at the time of disease recurrence. The median duration of treatment with erlotinib given for these relapses was around 13 months, which is more or less in the range of the usual progression-free survival for patients with metastatic non–small cell lung cancer receiving erlotinib. This suggests that at the time of relapse in the osimertinib arm of the ADAURA trial, the disease would be still sensitive on osimertinib. If that is the case, an overall survival benefit would be very likely in this trial.

Another important perspective is to better select the patient needing an adjuvant treatment. If we knew the presence of minimal residual disease after a surgical resection—for example, with highly sensitive ctDNA [circulating tumor DNA] determination detection—this knowledge would be critical in identifying patients who really need adjuvant treatment and preventing giving unnecessary and costly treatment to other patients without detection of minimal residual disease. We need very sensitive ctDNA detection to determine if there is minimal residual disease after surgical resection. We need tumor information…that may be necessary for this sensitive detection of ctDNA. This also underscores the need for genomic testing of any tumor, any non–small cell lung cancer, even in early stage of the disease. This is crucial for obtaining the genomic profile of the tumor, as well as the presence of an oncogene addiction. It could also personalize the detection of minimal residual disease with ctDNA.

Practical Implications of ADAURA in NSCLC

 The ADAURA study has osimertinib as adjuvant therapy in patients with completely resected surgical stage IB, II, and IIIA 

 mutation-positive non–small cell lung cancer. In ADAURA, 682 patients were randomized into a 1:1 ratio to the…osimertinib or placebo. The primary end point was disease-free survival [DFS] for the stage II and IIIA disease. The secondary end point was disease-free survival for the overall population. The primary end point of ADAURA was met in patients with stage II and IIIA disease. An impressive hazard ratio for the disease-free survival of 0.17 was observed, equating to 83% reduction in the risk of disease recurrence with osimertinib vs placebo. In the overall population of patients with stage IB, II, and IIIA disease, a hazard ratio of 0.20 was observed, equating to 80% reduction in the risk of disease recurrence with osimertinib. The patients who received osimertinib had fewer…than those who received the placebo, with a lower incidence of metastatic disease in those patients with recurrence, including fewer CNS [central nervous system] recurrence events.

Adjuvant osimertinib demonstrated a clinically meaningful improvement in the CNS disease-free survival compared with placebo. It had a hazard ratio of 0.18, equating to 82% reduction in the risk of disease recurrence. CNS disease recurrence was less likely with osimertinib compared with placebo, with a conditional probability of less than 1% after 18 months with osimertinib.

I conclude that the ADAURA trial demonstrated that adjuvant osimertinib was highly statistically significant with the clinical meaning for improvement in disease-free survival and reduction of risk of the local and distant recurrence, as well as improving CNS metastases and CNS disease-free survival in the patients with complete resected stage IB, II, or IIIA 

 mutation-positive non–small cell lung cancer. Also, there is a limitation that the maturity of both DFS and OS [overall survival] are low, given the reality that the patients with early stage lung cancer miss the chance to be cured once there is recurrence. I believe that prolonged DFS with osimertinib adjuvant is useful information for the patient hoping for a cure. In the future, we must consider this in the patient selection because there is some financial toxicity in each country.

 In the ADAURA trial, which compared patients with stage IB to IIIA who have been operated on, they might receive adjuvant chemotherapy, so that is the standard of care after they had been randomized to receive either osimertinib or placebo within 3 years. We have what has been currently presented and recently published. These are early data, so we need to be cautious, but they’re really impressive. The primary end point was disease-free survival, and the hazard ratio was around 0.17 for the whole population. That means you have a reduction of the risk of deaths or recurrence of the disease—around 83% in this population—if they received osimertinib within 3 years. At 3 years in the osimertinib arm, disease-free survival, for example, is around 80%; in the placebo arm, it’s around 30%. The magnitude of difference is unprecedented. Whatever the previous trial in the 

-mutated population that has been already reported, we never observed this huge improvement in disease-free survival in favor of osimertinib. These are early data, so the trial was unblinded at 3 years before it was expected because of this important magnitude of benefit.

That means we don’t have the translation, and we don’t have the result. This huge improvement in disease-free survival in the osimertinib arm will translate to an overall survival benefit. I cannot imagine that it will not translate to an improvement in overall survival for this population when you see this magnitude of benefit. Whatever does not translate in an increment in overall survival, the patient should know that at 3 years, 80% achieve disease-free survival. In the other arm, it’s only 30% disease-free survival. It’s enough to make it as a standard of care for the 

 population after thoracic surgery. It’s unprecedented with this population.

Unprecedented Results from ADAURA

 For the adjuvant chemotherapy, giving 3 to 4 cycles of a platinum-based regimen after a surgical resection of non–small cell lung cancer is often not easily tolerated in our patients. In our experience, I would say only 60% to 70% of the patients eligible for adjuvant chemotherapy receive adjuvant cisplatin-based chemotherapy. For some of these patients, we have to stop the adjuvant chemotherapy due to hematological or renal toxicity. The compliance with adjuvant chemotherapy is an issue after surgical resection of non–small cell lung cancer.

 mutations or others, it is very tempting to give targeted therapy instead of chemotherapy. We have 2 Chinese trials comparing head-to-head an EGFR TKI [tyrosine kinase inhibitor], which was gefitinib in this case, to adjuvant chemotherapy, which was based on cisplatin-vinorelbine chemotherapy. In this study, we have a majority of stage II and stage IIIA disease, so these studies were addressing patients with later stages of the disease. In both Chinese trials, the EVAN and ADJUVANT CTONG1104, the EGFR TKI achieved a clear DFS [disease-free survival] benefit over chemotherapy, but this DFS benefit did not translate into an overall survival benefit.

This lack of translation of the DFS benefit into an overall survival benefit in the ADJUVANT trial, for example, shows the follow-up is mature enough to have a clear picture of the survival of these patients. The lack of translation may have different meanings. This can first suggest that the duration of EGFR TKI adjuvant treatment was too short because a majority of recurrences were at the end of the 2-year adjuvant treatment. Another explanation may be that EGFR TKIs might only delay, rather than prevent, disease recurrence. EGFR TKIs, unlike chemotherapy, might not be able to eradicate metastatic disease; maybe the stem cells explaini that the DFS benefit did not translate into an overall survival benefit in this trial. 

Due to the impact of the adjuvant chemotherapy, we do not have to drop it for eligible patients, even if there is an 

mutation in the resected tumor sample. We have to keep in mind that adjuvant chemotherapy is able to increase the cure rates of these patients, even if you consider giving a targeted therapy, especially osimertinib, after adjuvant treatment if the patient is fit to receive a cisplatin-based adjuvant therapy.

 The biggest trial for adjuvant chemotherapy was the IALT trial, the first one, followed by the ANITA trial, which confirmed the benefit in terms of overall survival. It’s an absolute benefit, around 5% at 5 years if patients receive platinum-based chemotherapy. For the patients with stage IB and IIIA, that means you will reduce the probability of deaths who have been operated on by about 10% to 12% if they receive a platinum-based chemotherapy in 4 cycles after surgery. The most well-known association is with vinorelbine. This should reach at least around 300 mg/kg—equivalent to platinum[-based chemotherapy]—which has been done in this trial. This is the only positive trial in overall survival in non–small cell lung cancer after surgery if they receive an adjuvant chemotherapy.

Adjuvant Approaches in EGFR+ NSCLC

 What was the standard of care for adjuvant treatment after surgical resection of 

mutated non–small cell lung cancer at an early stage? The standard of care was the same as for wild-type 

 non–small cell lung cancer, given that the 

status was generally unknown in such patients after surgical resection. The adjuvant chemotherapy, mainly with cisplatin and vinorelbine, isproposed to fit patients aged less than 70 or 75, depending on whether the patient is fit or not after surgical resection of a stage IIB or greater tumor, or for positive disease.

In my country, cisplatin-vinorelbine is the most commonly used regimen in the adjuvant setting. It is the best validated regimen in this setting, providing a survival benefit of around 9% at 5 years in the subgroup analysis of the LACE meta-analysis, which is very consistent with the results of the ANITA trial. In stage II or III disease, the magnitude of benefit is greater—approximately 15% survival benefit at 5 years for the phase III disease and approximately 12% for the stage II. Even if this regimen remains rather toxic, especially in terms of hematological toxicity, it remains largely used in our clinical practice. We have to also keep in mind that, today, only adjuvant chemotherapy has demonstrated its capacity to cure some patients after surgical resection of non–small cell lung cancer, which is not the case yet for targeted therapies.

For the EGFR mutant patients we do not have yet any - - data on the prognostic impact of EGFR mutations in early stages.Well, in clinical practice an in clinical trial, especially in the control arm of the ADAURA study we can see that many patients with resected early-stage 

mutated non–small cell lung cancer will experience disease relapse even after adjuvant chemotherapy. The long-term survival data from the previous clinical trials, like the RADIANT trial, were more or less immature at the time of publication. In this RADIANT study, for example, with approximately one-third of stage IB, one-third of stage II, and one-third of stage IIIA disease—and given that half of the patients received adjuvant chemotherapy—the median overall survival was around 15 months, with 40% alive at 5 years. This is not very different from the other adjuvant clinical trials with unselected patients, including patients with wild-type 

 mutations have an adverse prognostic impact after surgical resection or not. I would say the prognosis is more or less the same as for wild-type 

 In Japan, tegafur-uracil, so-called UFT, is used for stage IA and stage IB patients, with or without 

mutation. Platinum-based chemotherapy is used in stage II and IIIA patients, with or without 

 mutations. Tegafur-uracil is only approved in Japan, so no other countries can use such drugs. Japanese guidelines recommend against the use of the EGFR TKI [tyrosine kinase inhibitor]. Right now, the most common therapeutic regimen is doublet chemotherapy, especially cisplatin-vinorelbine, in Japan.

Standard of Care for EGFR+ ES-NSCLC

 I presented this issue at ESMO [European Society for Medical Oncology meeting] in 2019. The ADAURA trial showed that frequency of the 

 mutation appears to be higher in Asia, with 28% vs 16% in non-Asia. In females, it is more frequent—31% vs 14% for males—in patients with stage IB to IIIA non–small cell lung cancer. The higher frequency in these 2 subgroups has also been observed in the advanced setting. The frequency of the more common mutation in combination with uncommon mutation is very low. The frequency of 

 mutation, 44%, should be interpreted with caution due to the difference between countries in local practice pre-screening, which may confound the frequency of the reported mutations.

 exon 19 deletion occurred more frequently in the American and European populations compared with Asia and Pacific regions. In the United States, exon 19 deletion is 70% and 62% in Europe. On the other hand, in Asia, it’s 51%. The frequency of the 

exon 20 insertion mutation may be underrepresented, as the patients who are known to be exon 20 insertion-positive have not been screened by this trial. In early-stage lung cancer, about 30% to 40% of patients in Asia have the 

 mutation. In the Caucasian population, the incidence is around 15% in such early-stage cancers.

, we test all patients with advanced-stage disease, particularly for adenocarcinoma. We have questions about squamous cell carcinoma because it’s infrequent to have squamous cell carcinoma with an 

 mutation. For adenocarcinoma, particularly with non- or light smokers, patients should be systematically tested because we have a high incidence for 

-mutated patients in this case. This is for the advanced-stage metastatic disease. For the locally advanced stage, they go to surgery. There is no recommendation because we don’t have the ability to treat these patients with targeted therapy. We just had recent data with the ADAURA trial that might push the patient with this third-generation EGFR TKI [tyrosine kinase inhibitor]. Otherwise, we should not test in this population.

For the testing in this population, there are different types of techniques, particularly regarding the tissue. If you want to only test for 

 mutation with highly sensitive and specific techniques, it can be done by PCR [polymerase chain reaction]. Otherwise, it’s with the NGS [next-generation sequencing] panel. The best is to test by NGS panel because you don’t have only the 

 testing, but also other molecular alterations at the same time. If you just want to know if the patient is 

-mutated or not, there is a specific test, and it can be done in most countries.

Testing for EGFR in NSCLC Worldwide

The question of molecular testing for patients with early-stage non–small cell lung cancer was not a real question until today because, at this point, there were no recommendations to perform molecular testing in resected non–small cell lung cancer since the treatment strategy did not depend on the molecular testing results. Generally, ALK and ROS1 assessments are done with reflex testing of the preoperative biopsy in the case of adenocarcinoma or on the resected sample. Sometimes a complete molecular profile is performed on the same biopsy or on the surgical sample, but it is unusual in my country, France, due to the reimbursement problems. 

Nevertheless, we used to perform molecular testing in nonsmokers or light smokers in order to exploit the surgical sample for a better DNA or RNA quality on archival samples at the time of possible metastatic relapse. With the advent of adjuvant therapy with osimertinib in EGFR-mutated non–small cell lung cancers, the strategy will be reviewed. It’s very likely, that the molecular testing will also be performed in early stage of the disease. The pathologist is involved in this process; we need to ask for the pathologist for molecular testing in the case of early-stage, which is not the case for advanced stages of the disease, for which reflex testing is more usual.

Molecular testing is not systematically done in most of the centers because we don’t have a targeted treatment demonstration that should be applied for this population with a specific molecular alteration as we do in advanced-stage disease, in which we test for EGFR, ALK, ROS1, BRAF, and so on. For early-stage, we know we have a nice result with the ADAURA trial, with EGFR as an example. We push to systematically test patients, particularly for EGFR, and we might expect to have other results for the other types of molecular alteration. Nowadays, there is no clinical implication except to include patients in specific clinical trials. That’s why we generally don’t perform it; it’s costly, and if you have a molecular alteration, there is no clinical application.

We try to push to have a histological examination. We try to avoid cytology except if we don’t have a choice because we know for the diagnostic histological type, it’s important, and cytology can be challenging sometimes. If you want to perform other molecular testing in the disease, you need to have a minimum amount of DNA and tumor cells. It’s better to have a good histological diagnostic. For some patients, it might be challenging, so that’s why we can do it by cytology. I would say it’s a minority of patients who have a cytologic confirmation of the diagnostic. It’s mainly a histological confirmation.

We know how to do the testing. You have the tissue, and you push the patient for surgery. You have the tumor resected, so it’s not a question of the amount of tumor tissue. It’s a question of demonstration of the benefit and what to do if you have a molecular alteration. It’s costly. That means if you spend money and there is no clinical implication, we might understand but won’t do it. That’s why we need to have specific guidelines, particularly specific demonstration that brings a benefit in terms of the patient becoming disease-free—also in terms of curability and improved overall survival for this population—to confirm that we need to do this molecular testing. Otherwise, we wait for recurrence. Unfortunately, if there is any recurrence, we might treat the patient for metastatic disease afterward. In this case, we have the testing that should be done because this is 2020. Nowadays, we need to try to test the early-stage disease.

Molecular Testing in Early-Stage NSCLC

 The most critical issue is whether complete resectability is possible or not. However, the final decision may be based on the surgical findings. In stage II, IIIA disease, there may be a difference in the experience of the surgeon. Clinical issues point to resectability. Resectability of early stage lung cancer, especially II or IIIA disease, is determined by a multidisciplinary team-based discussion based on the findings, lymph nodes, and metastases. For example, the single or multistation variety. Personally, I worry about indications of surgery when there are multiple and discrete lymph nodes and metastases. In most cases, discrete N2 disease is resectable. Complete resection does not always get to a cure, especially in stage II and IIIA disease.

Both stage IIA and IIB patients with postoperative complications and the very elderly population—over 80 years of age—to be referred to the medical oncologist for adjuvant treatment. There is high risk for recurrence in the poorly differentiated tumors, including lung neuroendocrine tumors and vascular or lymph node invasions, tumors more than 4 cm in size, visceral pleural invasion, and known lymph node metastasis. When we have such findings, we hand over to the medical oncologist to do adjuvant chemotherapy after surgery. These independent factors may not be indicated and may be considered when the timing of treatment is adjuvant. Tumor size less than 4 cm depends on the nodal involvement among stage II disease. There is no difference in strategy between stage IIA and IIB disease. Japanese guidelines recommend platinum-based chemotherapy in this population.

Determining Resectability in ES-NSCLC

: Our strategy is similar to the NCCN guidelines. As for the patient risk factors, it’s age, smoking status—especially the current one—previous cancer history, patient exposures, and other primary disease such as COPD [chronic obstructive pulmonary disease], pulmonary fibrosis, and ILD [interstitial lung disease]. As you know, in Japan, there are a lot of patients with interstitial lung disease, so we are very careful at watching this point. Also, we care about infectious disease. Other factors are the size, shape, location, and density of the primary nodules. Those are associated with primary abnormalities, such as parenchymal fibrosis or scarring, and inflammatory changes.

We have a PET [positron emission tomography]/CT scan, so we are considering if it is malignant disease or not. As for the diagnostic variation, patients with strong clinical suspicion for stage I or II lung cancer, we don’t require a biopsy before surgery. Preoperative biopsy may be appropriate if it’s a known lung cancer, especially if an infectious disease is strongly suspected; that can be diagnosed in the core biopsy or bronchoscopy.

The bronchoscopy is required before surgical resection. It depends on the case. I think in this population in the United States with early stage lung cancer, mediastinal staging is essential because there is quite a difference between stage I, II, and IIIA disease. Stage I and II, usually the surgery is prior, but in stage III disease, some combination treatment should be considered. In patients, mediastinal staging, especially by using the EBUS [endobronchial ultrasound], is commonly used. Unfortunately, Japanese surgeons are not familiar with the mediastinoscopy. Usually, EBUS is the prior procedure.

The preferred diagnosis strategy for the individual patient depends on the size and location of the tumors, the presence of the mediastinal and distant disease, the patient characterization such as primary pathology and other clinical comorbidities, and the local experience and expertise. The decision about the optimal diagnosis with a suspected stage I, II, IIIA lung cancer should be made by the multidisciplinary team discussions. It’s a very essential point. Once a week, every week, my institution discusses new patients and those complicated cases in the multidisciplinary team.

Masahiro Tsuboi, MD, National Cancer Center Hospital East

Articles