Randomized Phase II Study of 3 Months or 2 Years of Adjuvant Afatinib in Patients With Surgically Resected Stage I-III EGFR-Mutant Non–Small-Cell Lung Cancer

Abstract

PURPOSE

For patients with surgically resected disease, multiple studies suggest a benefit of epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) in delaying cancer recurrence. The necessary duration of therapy for benefit is unknown.

MATERIALS AND METHODS

This randomized phase II study enrolled patients with completely resected stage IA-IIIB EGFR-mutant non–small-cell lung cancer (American Joint Committee on Cancer 7th edition) after stage-appropriate standard-of-care adjuvant therapy. Patients were randomly assigned 1:1 to 3 months or 2 years of adjuvant afatinib starting at 30 mg by mouth daily. Computed tomography imaging was performed every 6 months for 3 years and then annually. The primary study end point for this planned 92-patient trial was recurrence rate at 2 years from randomization. A 20% improvement (from 70% with 3 months to 90% with 2 years) was targeted.

RESULTS

Forty-six patients enrolled and 45 were treated. The assigned course of afatinib treatment was completed by 96% (22/23) of patients in the 3-month group and only 41% (9/22) in the 2-year group. The 2-year recurrence-free survival (RFS) rates were 70% in the 3-month group and 81% in the 2-year group (P = .55). The median RFS was 42.8 months in the 3-month group and 58.6 months in the 2-year group. Side effects were consistent with those previously described for afatinib.

CONCLUSION

Recurrences at 2 years were 11% less common with 2 years versus 3 months of adjuvant afatinib. This difference did not meet the 20% primary study target, likely because of underaccrual and early drug discontinuation on the 2-year group. With the availability of osimertinib with better efficacy and tolerability than earlier-generation agents, the optimal duration of adjuvant EGFR TKI therapy remains an important question.

INTRODUCTION

Despite breakthroughs such as immunotherapy and targeted therapies in the treatment of advanced, metastatic non–small-cell lung cancer (NSCLC), the management of surgically resected, early-stage disease has remained relatively unchanged over the last 15 years. For larger primary tumors, or tumors with malignant lymph node involvement, 3 months of treatment with adjuvant cisplatin-based chemotherapy improves the chances of overall survival (OS) by 5%-10%.¹ Smaller (4 cm or less) and node-negative NSCLC is now staged as IA1-IB by the American Joint Committee on Cancer (AJCC) 8th edition staging criteria.² Although there is no level 1 evidence of benefit of adjuvant chemotherapy in stage I disease, recurrence rates are high, retrospective studies show survival advantage with adjuvant chemotherapy in some of these smaller tumors,³ and guidelines support medical oncology consultation to discuss adjuvant chemotherapy for all patients with resected NSCLC with high risk of recurrence.^4,5 In stages of resected NSCLC where cisplatin-based chemotherapy is the standard of care, efforts to add novel treatments to chemotherapy, such as bevacizumab, did not demonstrate improvement in survival.⁶ Multiple phase III trials of adjuvant checkpoint inhibitor immunotherapy are still ongoing with results anticipated over the next few years.

CONTEXT

• Key Objective

• In patients with surgically resected stage I-III epidermal growth factor receptor (EGFR)-mutant non–small-cell lung cancer, does 2 years of adjuvant afatinib improve recurrence rates relative to 3 months of adjuvant afatinib?

• Knowledge Generated

• Disease-free survival was numerically but not statistically improved for patients treated with 2 years of adjuvant afatinib relative to 3 months. Despite dose reductions, a 2-year course of adjuvant afatinib was not completed by the majority of patients.

• Relevance

• The optimal duration of adjuvant EGFR tyrosine kinase inhibitor therapy is still uncertain.

Mutations in the epidermal growth factor receptor (EGFR) gene occur in 15%-50% of NSCLC and confer exquisite sensitivity to small-molecule EGFR tyrosine kinase inhibitors (EGFR TKIs) in the metastatic setting. Afatinib, an irreversible EGFR TKI, has superior response rates, progression-free survival, and OS as compared with chemotherapy for patients with advanced EGFR-mutant NSCLC.⁷ Prior efforts to incorporate EGFR-directed therapy into the adjuvant treatment of early-stage EGFR-mutant NSCLC, including retrospective studies, prospective single-arm studies, meta-analyses, and subgroup analyses of randomized clinical trials, have been inconclusive for OS^8-13 but have consistently suggested a benefit in recurrence-free survival (RFS).

The delivery of adjuvant cytotoxic chemotherapy is limited by cumulative, irreversible, and potentially life-threatening toxicity. In contrast, with appropriate supportive care, EGFR TKIs may be delivered continuously without concern for cumulative adverse events (AEs).^14,15 Initial studies of consolidation or adjuvant EGFR TKIs were designed to deliver 2 years of therapy.^11,16,17 Stopping at 2 years in some designs was a rational limit that would, at least, cover the period during which most resected lung cancers would normally recur.^18,19 Meanwhile, randomized studies of postoperative TKI therapy for GI stroma tumors (GIST) have generally, but not consistently, demonstrated improved outcomes with a longer duration of therapy.^20-26 In the case of GIST, it is particularly interesting that high rates of disease recurrence were observed within 6-12 months after treatment discontinuation with both 1-year and 3-year durations of adjuvant imatinib, yet 3 years of therapy has recently been confirmed to improve OS. These data suggest that adjuvant TKI therapy may have a cytostatic maintenance effect on minimal residual disease, which outweighs the initial cytotoxic effect. With the hypothesis that a longer duration of adjuvant EGFR TKI therapy could improve RFS, and possibly OS, but with equipoise regarding the benefit of any adjuvant EGFR TKI therapy, we launched this randomized, 2-arm study of 3 months versus 2 years of adjuvant afatinib in patients with surgically resected EGFR-mutant NSCLC following treatment with stage-appropriate chemotherapy (ClinicalTrials.gov identifier: NCT01746251).

MATERIALS AND METHODS

Study Participants

Patients enrolled were at least 18 years of age with pathologically confirmed stage IA-IIIB (AJCC 7th edition) NSCLC with an EGFR mutation, including exon 19 deletion, exon 21 L858R, and other known activating and sensitizing EGFR mutations at the discretion of the principal investigator (L.V.S.). All patients had complete surgical resection without evidence of recurrence based on screening computed tomography (CT) scans of the chest and abdomen. Brain imaging was completed preoperatively in all patients but was not mandated postoperatively for enrollment. Patients had Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1. Neoadjuvant or adjuvant chemotherapy and radiation were allowed and required at least 3 weeks washout, but no more than 6 months elapsed since surgery. Patients treated with sequential chemotherapy and radiation could enroll up to 9 months after surgery. All patients had adequate organ function. Key exclusion criteria included a history of prior treatment with EGFR TKIs, a history of a second clinically significant malignancy, uncontrolled intercurrent illnesses, or evidence of clinically active or a history of interstitial lung disease. All patients enrolled were required to sign an informed consent form, understand and willingly participate in clinical research, and comply with protocol procedures.

This was an investigator-initiated, multicenter, independently monitored study that received institutional review board approval at each trial site. The study was conducted in accordance with the US Code of Federal Regulations governing clinical study conduct and ethical principles; these guidelines originate from the Declaration of Helsinki. The trial was registered with ClinicalTrials.gov with identifier NCT01746251.

Randomization and Treatment

Randomization was stratified by pathologic stage (stage I, II, or III). Eligible patients were randomly assigned 1:1 to either a 3-month or a 2-year course of afatinib adjuvant therapy. An oral dose of afatinib 30 mg daily was administered throughout the first cycle (28 days) of treatment. There was an optional increase to afatinib 40 mg daily in patients who tolerated the first cycle of treatment well. Dose delays and/or dose reduction to afatinib 20 mg daily were allowed for prespecified drug-related toxicities. Guidelines for managing common side effects of afatinib were included in the protocol, and participating institutions were experienced at safe delivery of afatinib.

Chest computed tomography scans were performed every 6 months for the first 3 years and annually thereafter. Additional imaging was performed if indicated on the basis of clinical assessments. AEs were monitored throughout the study and for 30 days after the last study drug administration.

Trial Design and Statistical Methods

The primary objective of this phase II study was to demonstrate that 2 years of adjuvant afatinib improved RFS compared with 3 months of adjuvant afatinib in patients with resected stage I-III NSCLC with an EGFR mutation. Secondary objectives were to determine the safety and tolerability of adjuvant afatinib, estimate OS, and collect clinical data and molecular characteristics from the enrolled patients. Baseline characteristics were compared for imbalance using the Fisher’s exact test. OS and RFS median estimates and curves were constructed using the Kaplan-Meier method, and differences in treatment groups were tested with the log-rank test. Two-year RFS was analyzed under the intent-to-treat approach using a one-sided Fisher’s exact test. The null hypothesis was a 2-year RFS rate of 70% in the 3-month arm, improving to 90% in the 2-year arm. Initial sample size estimates required 46 patients in each group to have 82.5% power to detect a difference between the group proportions of 20%. This was based on an one-sided Fisher’s exact test with a significance level of 0.1. In 2016, 36 months after initial study activation, the study size was decreased to a new goal of 30 patients per arm, with 82.5% power to detect a 26% reduction in recurrence (one-sided α = .10); the amended null hypothesis was a 2-year RFS rate of 64% in the 3-month arm.

RESULTS

Patients

From May 2013 to January 2017, 46 patients enrolled from four hospitals and were randomly assigned to 3 months (n = 24) or 2 years (n = 22) of adjuvant afatinib (Table 1; Fig 1). One patient in the 3-month group withdrew consent prior to treatment. There were no significant imbalances in baseline characteristics between arms. Both groups had more females than males and more never smokers than ever smokers. More than 70% of patients in both treatment groups had either a mutation in EGFR exon 21 L858R or an EGFR exon 19 deletion. About half of the patients in both groups had prior therapy, and disease stage was balanced across the two treatment groups. All but two patients (one in each group) with stage III lung cancer received chemotherapy prior to enrollment.

TABLE 1.

Baseline Patient Characteristics

FIG 1.

Flowchart illustrating patient enrollment and compliance.

Treatment and AEs

All patients were treated with afatinib 30 mg by mouth daily for the first month. Two patients (9%) in the 3-month arm and no patients in the 2-year arm were escalated to afatinib 40 mg daily. Dose reductions to afatinib 20 mg daily occurred in 5 patients (22%) in the 3-month and 10 patients (45%) in the 2-year arms. The median number of treatment cycles in the 3-month group was 3, range 0-3 cycles; and in the 2-year group, it was 17 cycles, range 1-26 cycles. With full data (no censoring), treatment was completed in 96% (22/23) of patients in the 3-month group and 41% (9/22) of patients in the 2-year group. A higher proportion of patients completed treatment in the 3-month group compared with the 2-year group (P = .002; Fisher's exact test). In the 2-year group, a maximum of 26 cycles were allowed per protocol, yet a median of 17 cycles were given with a range of 4-26 cycles.

Adverse Events

No grade 4 or 5 AEs were observed. Three grade 3 events were reported in the 3-month group, and 8 grade 3 AEs were reported in the 2-year group. Diarrhea (n = 3 [14%]) and paronychia (n = 2 [9%]) were the most common grade 3 AEs observed in the 2-year group and were the only grade 3 events observed in more than 5% of patients in either group. The median time to onset of grade 3 events in the 2-year group was 56 days. The most common grade 1 or 2 AEs in both groups were diarrhea, mucositis or cheilitis, and rash. Grade ≤ 2 dry skin occurred in 59% versus 17% in the 2-year versus 3-month groups and paronychia in 55% versus 8%, respectively. The median time to highest grade toxicity onset was 141 days in the 2-year group and 16 days in the 3-month group. The toxicity data are summarized in Table 2.

TABLE 2.

Adverse Events

Clinical Outcomes

The median follow-up was 61 months. RFS was 11% higher in the 2-year group than the 3-month group, but this difference was not statistically significant and did not meet the primary objective (Fig 2). Landmark RFS at 2 years was 70% in the 3-month group and 81% in the 2-year group (P = .55). The median RFS was 42.8 months in the 3-month group and 58.6 months in the 2-year group, and the median OS was not yet reached for either group (Fig 2).

FIG 2.

(A) Recurrence-free survival (log-rank test not significant, P = .40) and (B) overall survival (log-rank test not significant, P = .72) in the 3-month and 2-year treatment groups.

Recurrences occurred in 23 patients, 13 (57%) in the 3-month arm and 10 (45%) in the 2-year group. Eight of 23 recurrences (35%) included the CNS. Brain-only recurrence was seen in only 4 patients, 1 in the 3-month arm and 3 in the 2-year arm. At recurrence, 16 patients underwent biopsy with 1 of 16 patient cancers exhibiting small-cell transformation. Genotyping was performed on 14 of 15 adenocarcinoma specimens but failed in the patient with small-cell transformation. The EGFR T790M resistance mutation was not identified in any of the 14 tested samples.

DISCUSSION

This prospective, randomized phase II trial aimed to evaluate different durations of adjuvant afatinib in patients with resected, stage I-III EGFR-mutant NSCLC. In this study, patients experienced numerically improved 2-year RFS with treatment for 2 years (81%) compared with 3 months (70%). Treatment completion rates were significantly higher when 3 months of adjuvant afatinib was prescribed, with only 4% of patients discontinuing treatment early compared with 59% of patients in the 2-year group. Toxicity rates were similar between the groups and consistent with prior studies of afatinib,^27,28 including acneiform rash, diarrhea, fatigue, mucositis, paronychia, and skin fissures. Although no pneumonitis or grade 4 or higher related AEs were observed, it is evident that the toxicities, despite being anticipated and low-grade, were not acceptable to half of the patients in the 2-year arm.

Previous studies of adjuvant EGFR TKIs have been reported over the last decade. In a retrospective cohort study, the 2-year disease-free survival (DFS) was 17% better in patients treated with EGFR TKI therapy (erlotinib or gefitinib) as compared with patients not receiving this therapy (89% v 72%, hazard ratio [HR], 0.53; 95% CI, 0.28 to 1.03; P = .06).¹⁰ In our previously published 100-patient single-arm study of adjuvant erlotinib for 2 years after optional chemotherapy (SELECT), the 2-year DFS was 88% (95% stage I, 78% stage II, and 91% stage III).¹² The RADIANT study randomly assigned patients to 2 years of adjuvant erlotinib or placebo after completion of planned chemotherapy. RADIANT was not limited to those with tumors harboring EGFR mutations but did include 161 patients with EGFR-mutant NSCLC.¹¹ Among these patients, the 2-year DFS for erlotinib was 75% compared with 54% for placebo, which was not a significant difference because of statistical methodology. In the ADJUVANT/CTONG1104 trial, 222 patients were randomly assigned to adjuvant gefitinib for 2 years (no chemotherapy) versus adjuvant chemotherapy.²⁹ The median DFS was superior in those who received the EGFR TKI, 28.7 versus 18.0 months, HR 0.6 (P = .0054).¹³ Overall, these studies demonstrate that 2 years of adjuvant EGFR TKI therapy improves DFS relative to chemotherapy or placebo. However, the studies to date were not powered to demonstrate an OS benefit of this strategy and no OS benefit has been observed. Such OS benefit of TKI following chemotherapy might only be demonstrated with a larger sample size, such as that in the ongoing ALCHEMIST adjuvant erlotinib trial (n = 450) or the ADAURA adjuvant osimertinib trial (n = 682).^30-32

Similar to the previously reported SELECT trial, with a 2-year RFS of 88%, but in contrast to other reported randomized trials of adjuvant EGFR TKI therapy, this study enrolled stage IA NSCLC.¹² This inclusion was adequately accounted for in the statistical plan, as the control arm (3-month) performed as anticipated with a 2-year RFS rate of 70% (assumption 70%). In the RADIANT study, which required tumors > 3 cm (stage IB by AJCC 7th edition staging criteria), a 75% 2-year RFS was reported on the erlotinib arm.¹¹ In ADJUVANT/CTONG1104, which included only stage II and higher and did not require staging positron emission tomography (PET)/CT, 64% of patients had stage IIIA NSCLC and a 60% 2-year RFS was reported on the gefitinib arm.^13,29 In contrast, ADAURA included relatively equal numbers of stage IB, II, and IIIA patients and demonstrated a significant but diminishing benefit in lower-stage disease (DFS HR of 0.39, 0.17, and 0.12 with increasing stage).³² In the trials excluding stage I patients who overall had higher rates of recurrence than expected during the 2-year planned duration of adjuvant therapy, it is possible that a design with a shorter duration of adjuvant EGFR TKI therapy, such as 3 months, would have resulted in even more recurrences observed.

Slower-than-anticipated accrual was the primary limitation of the study. The accrual challenges were preempted with inclusion of a treatment arm for all patients instead of observation or placebo. Despite this adjustment for patient appeal and enrollment at four institutions with prior experience in enrolling patients onto adjuvant EGFR TKI clinical trials, screening and enrollment may not have captured every possible candidate. Other adjuvant trials—ALCHEMIST and ADAURA—were ongoing during the conduct of this study and potentially competed with this study (ClinicalTrials.gov identifier: NCT02193282 and NCT02511106, respectively). In addition, some patients declined to participate to avoid receiving further therapy after completion of their standard-of-care plan, and others may have received off-label adjuvant therapy prescribed directly by their treating physician, as previously described.¹⁰ Furthermore, because of the EGFR mutation status requirement at the time of enrollment, some patients may not have been eligible because testing is not universally regarded as standard of care. As a result of slower-than-anticipated accrual, the study was modified during conduct with a revised goal and statistical plan from 92 to 60 patients and anticipated RFS improvement from 20% to 26%, but even this updated goal was not met before the study was closed. For this study to have validated the observed 14% difference in recurrence and keep the other statistical considerations intact, it would have required 196 randomly assigned patients. Although the observed magnitude of difference could translate to a clinically meaningful difference between the arms in an adequately powered study, this study failed to demonstrate that the longer duration of therapy is superior. Additionally, the efficacy results are challenging to interpret given the heterogeneity of stages and EGFR mutation subtypes enrolled. Our findings are also limited by the lack of a standard-of-care comparator arm, a result of the focus of this study on duration of therapy. However, this study clearly demonstrated that long-term adjuvant afatinib therapy is difficult, even in experienced hands. In this study, only 41% of patients prescribed 2 years of afatinib were able to complete the planned therapy. The early drug discontinuations may have likewise influenced the study outcome.

Although not significant, the trend toward superiority with 2 years compared with 3 months of adjuvant afatinib corroborate the duration of therapies being studied in other ongoing EGFR TKI trials, such as the ALCHEMIST arm with 2 years of adjuvant erlotinib or ADAURA arm with 3 years of adjuvant osimertinib. The ALCHEMIST trial is powered for OS benefit, whereas ADAURA is powered for DFS. These studies, if positive, could lead to the approval of these agents in early-stage EGFR-mutant NSCLC. Indeed, the recent ADAURA presentation suggests a dramatic early DFS benefit of 3 years of adjuvant osimertinib. It is unclear if the early unblinding of the ADAURA study may impact longitudinal results as at the time of study presentation the study was fully accrued with all patients followed for at least 1 year.³² In the future, subsequent trials may investigate the role of even longer durations of therapy, particularly in patients who are at highest risk for recurrence, such as those who have tested positive for molecular residual disease after surgical resection.³³ For patients to consider taking years of adjuvant EGFR TKI therapy, improved tolerance of therapy will be essential, potentially through a low threshold for dose adjustments, the intensive use of supportive care, and the use of newer agents with fewer and less severe side effects.

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/po/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).

Joel W. Neal

Honoraria: Research to Practice, MLI Peerview, Medscape, Biomedical Learning Institute, Prime Oncology, Rockpointe, CME Matters, MJH Life Sciences

Consulting or Advisory Role: AstraZeneca, Genentech/Roche, Exelixis, Jounce Therapeutics, Takeda, Lilly, Calithera Biosciences, Amgen, Iovance Biotherapeutics, Blueprint Pharmaceuticals, Regeneron

Research Funding: Genentech/Roche, Merck, Novartis, Boehringer Ingelheim, Exelixis, Nektar, Takeda, Adaptimmune, GlaxoSmithKline

Patents, Royalties, Other Intellectual Property: Up To Date-Royalties

Daniel B. Costa

Honoraria: Pfizer, Takeda/Millennium

Consulting or Advisory Role: AstraZeneca, Takeda/Millennium, Teladoc, Blueprint Medicines

Research Funding: Takeda/Millennium, AstraZeneca, Pfizer, Merck, Merrimack, Bristol-Myers Squibb, Clovis Oncology, Genentech, Spectrum Pharmaceuticals

Alona Muzikansky

Consulting or Advisory Role: Sofregen

Joseph B. Shrager

Consulting or Advisory Role: Becton Dickinson

Research Funding: Varian Medical Systems

Patents, Royalties, Other Intellectual Property: A potential pharmacological means of preventing ventilator-induced diaphragm dysfunction. A potential use of CRISPR for EGFR mutant lung cancer therapy

Michael Lanuti

Honoraria: Emory University

Consulting or Advisory Role: Bristol-Myers Squibb, PreCisa

Research Funding: NIH, Stand up to Cancer

Kavitha J. Ramachandran

Consulting or Advisory Role: Drishti

Deepa Rangachari

Honoraria: AstraZeneca

Consulting or Advisory Role: Advance Medical, DynaMed

Research Funding: Bristol-Myers Squibb, Novocure, Abbvie/Stemcentrx

Zofia Piotrowska

Consulting or Advisory Role: AstraZeneca, Guardant Health, Spectrum Pharmaceuticals, Immunogen, Lilly, Medtronic, Incyte, Genentech, C4 Therapeutics, Blueprint Medicines, Jazz Pharmaceuticals, Janssen

Research Funding: Novartis, ARIAD/Takeda, Spectrum Pharmaceuticals, AstraZeneca, Tesaro, Johnson & Johnson, Cullinan Oncology, Daiichi Sankyo Europe GmbH

Travel, Accommodations, Expenses: AstraZeneca, ARIAD/Takeda, Genentech

Mark G. Kris

Consulting or Advisory Role: AstraZeneca, Pfizer, Daiichi Sankyo

Travel, Accommodations, Expenses: AstraZeneca, Pfizer, Genentech, Daiichi Sankyo

Other Relationship: Memorial Sloan-Kettering Cancer Center, Genentech/Roche

(OPTIONAL) Open Payments Link: https://openpaymentsdata.cms.gov/physician/markgkris/summary

Lecia V. Sequist

Consulting or Advisory Role: AstraZeneca, Genentech/Roche, Janssen Oncology

Research Funding: Boehringer Ingelheim, Genentech, Merrimack, Novartis, AstraZeneca, Johnson & Johnson, Merck, Pfizer, Guardant Health, Blueprint Medicines, Loxo

Jamie E. Chaft

Consulting or Advisory Role: Genentech/Roche, AstraZeneca/MedImmune, Merck, Bristol-Myers Squibb, Flame Biosciences

Research Funding: Genentech/Roche, Bristol-Myers Squibb, AstraZeneca/MedImmune, Merck

No other potential conflicts of interest were reported.

AUTHOR CONTRIBUTIONS

Conception and design: Joel W. Neal, Alona Muzikansky, Kavitha J. Ramachandran, Christopher G. Azzoli, Lecia V. Sequist, Jamie E. Chaft

Administrative support: Lecia V. Sequist

Provision of study materials or patients: All authors

Collection and assembly of data: Joel W. Neal, Daniel B. Costa, Joseph B. Shrager, Deepa Rangachari, Mark S. Huberman, Zofia Piotrowska, Mark G. Kris, Christopher G. Azzoli, Lecia V. Sequist, Jamie E. Chaft

Data analysis and interpretation: Joel W. Neal, Daniel B. Costa, Alona Muzikansky, Joseph B. Shrager, Michael Lanuti, James Huang, Deepa Rangachari, Mark G. Kris, Christopher G. Azzoli, Lecia V. Sequist, Jamie E. Chaft

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors