Epidermal growth factor receptor tyrosine kinase inhibitors as first-line treatment for postoperative recurrent EGFR-mutated lung adenocarcinoma: a multi-institutional retrospective study

Abstract

Open in new tabDownload slide

OBJECTIVES

The aim of this study was to analyse the long-term survival outcomes and prognostic factors of patients receiving epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs) as first-line treatment for postoperative recurrent EGFR-mutated lung adenocarcinoma.

METHODS

Using a multi-institutional database, we performed a retrospective chart review to identify all patients who had undergone complete resection of stage I–III EGFR-mutated lung adenocarcinoma at 11 acute care hospitals between 2009 and 2016 and had received first-line EGFR-TKI treatment for postoperative recurrence. Adverse events, progression-free survival (PFS) and overall survival (OS) were investigated. Survival outcomes were assessed using Kaplan–Meier analysis. Cox proportional hazards models were used to calculate the hazard ratios (HRs) and 95% confidence intervals (CIs) for PFS and OS.

RESULTS

The study sample comprised 154 patients with a median age of 69. The total numbers of events were 101 for PFS and 60 for OS. The median PFS and OS were 26.1 and 55.4 months, respectively. In the multivariable analysis, EGFR ex 21 L858R mutation (HR: 1.71, 95% CI: 1.15–2.55) and shorter disease-free intervals (HR: 0.98, 95% CI: 0.96–0.99) were significantly associated with shorter PFS. Age (HR: 1.03, 95% CI: 1.00–1.07), smoking history (HR: 2.31, 95% CI: 1.35–3.94) and pathological N2 disease at the initial surgery (HR: 2.30, 95% CI: 1.32–4.00) were significantly associated with shorter OS.

CONCLUSIONS

First-line EGFR-TKI treatment was generally associated with favourable survival outcomes in patients with postoperative recurrent EGFR-mutated lung adenocarcinoma. EGFR ex 21 L858R mutation may be an important prognostic factor for shorter PFS.

INTRODUCTION

Epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs) have improved long-term survival outcomes in patients with EGFR mutation-positive stage IV non-small-cell lung cancer (NSCLC) [1–3]. A meta-analysis of individual patient data reported that the use of first-generation EGFR-TKIs (gefitinib or erlotinib) as first-line treatment for metastatic NSCLC had significantly better progression-free survival (PFS) than platinum-based chemotherapy, with a hazard ratio (HR) of 0.37 [4]. Furthermore, third-generation EGFR-TKI (osimertinib) was found to be associated with significantly longer PFS (HR: 0.46) when compared with earlier-generation EGFR-TKIs [3]. In addition to the types of EGFR-TKIs used in treatment, other potential prognostic factors in patients with EGFR-mutated metastatic NSCLC include sex, performance status, race (Asian and non-Asian) and EGFR mutation type (exon 19 in-frame deletion mutation [ex 19 deletion] and exon 21 Leu858Arg point mutation [ex 21 L858R mutation]) [3, 5].

The postoperative recurrence of adenocarcinoma harbouring EGFR mutations is not uncommon in patients with a primary EGFR-mutated lung adenocarcinoma, although its rate of incidence may be slightly lower than in patients with a primary EGFR–non-mutated NSCLC [6–8]. Furthermore, a study reported that the 5-year disease-free survival rate after complete resection of EGFR-mutated lung adenocarcinoma was only 52.9% [8]. When limited to patients with pathological stage III EGFR-mutated lung adenocarcinoma, the median disease-free survival duration after resection with adjuvant chemotherapy was 15.2 months [8]. Accordingly, it is important to establish effective management regimens for postoperative recurrent EGFR-mutated lung adenocarcinoma in order to improve prognoses and prolong survival.

Postoperative recurrent NSCLC has been analysed in combination with untreated stage IV disease in clinical trials on EGFR-TKIs [4]. However, a previous retrospective study from our group indicated that postoperative recurrent EGFR-mutated adenocarcinomas may include characteristics that are distinct from those of stage IV EGFR-mutated adenocarcinomas [9, 10]. However, those findings were based on a single-institution experience with a small sample size that was vulnerable to selection bias. Moreover, the small sample size precluded the implementation of rigorous multivariable analyses to identify significant prognostic factors. There is therefore a need to conduct multi-institutional studies with larger sample sizes that enable more rigorous assessments of EGFR-TKI for postoperative recurrent EGFR-mutated lung adenocarcinoma.

In this study, we aimed to analyse the long-term survival outcomes and prognostic factors of patients receiving EGFR-TKI as first-line treatment for postoperative recurrent EGFR-mutated lung adenocarcinoma, as well as to perform an exploratory subgroup analysis according to EGFR mutation type. The study was conducted using a multi-institutional database of patients who had undergone surgical resection of lung adenocarcinoma.

PATIENTS AND METHODS

Ethical statement

The study was approved by the Kyoto University Hospital Institutional Review Board (Reference number: R2504), which also granted a waiver of individual informed consent. This retrospective observational study was performed in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology Statement [11].

A retrospective chart review was performed using a multi-institutional database to identify all consecutive patients who had undergone resection of lung adenocarcinoma at 11 acute care hospitals between January 2009 and December 2016 and had received first-line EGFR-TKI treatment for postoperative recurrent lung adenocarcinoma. We also collected data on patient characteristics (age, sex, smoking history and Charlson comorbidity index), intraoperative and perioperative factors, pathological characteristics (maximum tumour diameter, invasive adenocarcinoma subtypes [12], EGFR mutation types, programmed cell death-ligand 1 expression, pathological N status and pathological stage) and follow-up information.

The study sample comprised patients who had undergone a complete resection of pathological stage I–III EGFR-mutated lung adenocarcinoma between 2009 and 2016 and had received first-line EGFR-TKI treatment for postoperative recurrence. R1/R2 resection was excluded. Patients who had EGFR mutation types other than EGFR exon 19 deletion and EGFR exon 21 L858R were excluded. Pathological staging was based on the Union for International Cancer Control TNM Classification of Malignant Tumours, 7th Edition. EGFR mutation status in exons 18–21 was analysed at the initial surgery. Patients who had received adjuvant chemotherapy with EGFR-TKIs or anti-angiogenic therapy combined with EGFR-TKIs were excluded from analysis. If an adverse event was noted without disease progression, the initial EGFR-TKI was switched to another EGFR-TKI at the discretion of the physician in charge. In Japan, osimertinib was approved as a second-line EGFR-TKI after re-biopsy for the detection of EGFR exon 20 Tyr790Met (T790M) mutation in 2016. Subsequently, it was approved as a first-line EGFR-TKI for all EGFR-mutated advanced NSCLCs in 2018.

Typical postoperative surveillance involved chest computed tomography (CT) every 3–6 months after the initial surgery, and brain magnetic resonance imaging was performed if patients were symptomatic, in line with the guidelines of Japan Lung Cancer Society, National Comprehensive Cancer Network and European Society for Medical Oncology [13–15]. Typical surveillance after the initiation of EGFR-TKI treatment for postoperative recurrence was a chest X-ray or chest CT and blood work (e.g. complete blood counts, blood chemistry tests, serum carcinoembryonic antigen levels and Krebs von den Lungen-6 levels) every 3 months. Chest CTs were usually scheduled every 3–6 months for the assessment of treatment efficacy. Adverse events were evaluated according to the Common Terminology Criteria for Adverse Events [16].

When disease progression was suspected, additional examinations were scheduled; these included the use of CT with or without positron emission tomography or brain magnetic resonance imaging. Diagnoses of postoperative recurrence or disease progression were based on radiological examination. When an acquired resistance to first- or second-generation EGFR-TKIs was suspected, the physician in charge ordered a re-biopsy using a liquid sample or a bronchoscopy-, thoracoscopy- or CT-guided re-biopsy of a tissue sample to detect T790M mutation [17–19].

Statistical analysis

The main study outcomes were PFS and overall survival (OS). PFS was defined as the interval from the date of EGFR-TKI initiation until the date of disease progression, death from any cause with censoring patients without event at the last clinic visit. OS was defined as the interval from the date of EGFR-TKI initiation until the date of death from any cause with censoring patients without event at the last clinic visit. For patients with progressive disease, we also assessed post-progression survival, which was defined as the interval from the date of disease progression until the date of death from any cause or the last clinic visit. Disease-free interval (DFI) was defined as the interval from the date of initial surgery until the date of the first recurrence.

In accordance with the statistical and data reporting guidelines for the European Journal of Cardio-Thoracic Surgery and the Interactive CardioVascular and Thoracic Surgery [20], we reported descriptive statistics using counts and percentages for categorical variables, and using medians and interquartile ranges for continuous variables. Potentially relevant factors were compared using the Mann–Whitney test for continuous variables or Fisher’s exact test for categorial variables, as appropriate. The median follow-up was calculated using the reverse Kaplan–Meier method for potential follow-up. Survival probability was estimated using the Kaplan–Meier method.

The HRs and 95% confidence intervals (CIs) for PFS and OS were calculated using Cox proportional hazards regression models. Univariable models were first used to examine the individual relationships between each candidate factor and the outcomes. Variables with potential predictors under a stepwise variable selection were included in the multivariable analysis, while variables of >5% of missing data were excluded.

Exploratory subgroup analyses were conducted to compare the incidences and HRs of PFS and OS between third-generation EGFR-TKI (osimertinib) and first- or second-generation (hereinafter referred to as ‘conventional’) EGFR-TKIs; these analyses adjusted for patient age, sex, smoking history, postoperative recurrence site, brain metastasis at initial recurrence, EGFR mutation type and adenocarcinoma subtype.

For patients who developed progressive disease during follow-up, we analysed the second-line management according to first-line EGFR-TKI treatment. Post-progression survival durations and rates were determined.

Statistical significance was set at P ≤ 0.05. All analyses were performed using JMP 14 (SAS Institute Inc., Cary, NC, USA).

RESULTS

We identified 154 eligible patients (55 males and 99 females) from the database. Table 1 summarizes the patient characteristics, tumour characteristics and first-line EGFR-TKI treatment. The median age was 69 years (interquartile range: 64.75–76). Supplementary Material, Table S1 summarizes recurrence sites at initial recurrence. Fourteen patients (9.1%) experienced grade 3 or greater adverse events following first-line EGFR-TKI treatment; these included 4 patients with interstitial pneumonia, 3 patients with diarrhoea, 2 patients with liver dysfunction and 1 patient with heart failure (Supplementary Material, Table S2). First-line EGFR-TKI was switched to another EGFR-TKI due to adverse events in 8 patients (5.2%).

Survival outcomes

The total numbers of events were 101 for PFS and 60 for OS. The median follow-up period was 44.0 months on the basis of PFS events and 46.0 months on the basis of OS events. The median PFS was 26.1 months (95% CI: 22.0–32.1 months) and the median OS was 55.4 months (95% CI: 50.0–76.4 months). The 5-year PFS and OS rates were 14.2% (Fig. 1A) and 41.8% (Fig. 1B), respectively.

Figure 1:

Kaplan–Meier survival curves. (A) Progression-free survival curve for all patients; (B) overall survival curve for all patients; (C) progression-free survival curves comparing patients with EGFR exon 19 deletion and patients with EGFR exon 21 L858R mutation; (D) progression-free survival curves comparing patients with single and those with multiple recurrences; (E) progression-free survival curves comparing patients with micropapillary and/or solid components and patients without micropapillary/solid components; and (F) progression-free survival curves comparing patients with or without receiving cytotoxic adjuvant chemotherapy. EGFR: epidermal growth factor receptor; MP: micropapillary; S: solid; TKI: tyrosine kinase inhibitor.

Open in new tabDownload slide

Table 2 presents the results of the univariable and multivariable analyses of the candidate factors of PFS. In the univariable analysis, EGFR ex 21 L858R mutation (HR: 1.56, 95% CI: 1.05–2.32, P = 0.03) (Fig. 1C), shorter DFI (HR: 0.98 per 1-month increase, 95% CI: 0.96–1.00, P = 0.02) and multiple recurrence (HR: 1.69, 95% CI: 1.00–2.87, P = 0.05) (Fig. 1D) were significantly associated with shorter PFS. Micropapillary and/or solid components showed non-significant tendency to shorter PFS in the univariable analysis (HR: 1.49, 95% CI: 0.98–2.27, P = 0.06) (Fig. 1E), while postoperative adjuvant therapy did not (Fig. 1F). In the multivariable analysis, EGFR ex 21 L858R mutation (HR: 1.71, 95% CI: 1.15–2.55, P < 0.01) and shorter DFI (HR: 0.98 per 1-month increase, 95% CI: 0.96–0.99, P < 0.01) were significantly associated with shorter PFS.

Table 3 presents the results of the univariable and multivariable analyses of the candidate factors of OS. In the multivariable analysis, age (HR: 1.03 per 1-year increase, 95% CI: 1.00–1.07, P = 0.03), smoking history (HR: 2.31, 95% CI: 1.35–3.94, P < 0.01) and pathological N2 disease at the initial surgery (HR: 2.30, 95% CI: 1.32–4.00, P < 0.01) were significantly associated with shorter OS.

Management of progressive disease after first-line epidermal growth factor receptor tyrosine kinase inhibitor treatment and post-progression survival

During the follow-up period, progressive disease was observed in 83 patients. Among these, data on second-line management were available from 24 patients. The median post-progression survival duration was 30.8 months (95% CI: 19.3–41.0 months). The 3- and 5-year post-progression survival rates were 40.4% and 31.6%, respectively. The associations between first-line EGFR-TKI treatment and second-line management are summarized in Supplementary Material, Table S3. The characteristics of patients who switched to osimertinib from their first-line TKI at disease progression (N = 7) are summarized in Supplementary Material, Table S4.

Exploratory subgroup analyses

The survival curves for the subgroup analyses of patients who received osimertinib and patients who received conventional EGFR-TKIs are shown in Supplementary Material, Fig. S1A (PFS) and B (OS). Forest plots of the HRs for PFS and OS according to subgroup are shown in Supplementary Material, Figs S2 and S3, respectively. Overall, there were no significant differences in PFS or OS between osimertinib and conventional EGFR-TKIs.

DISCUSSION

To the best of our knowledge, this study is the largest survival analysis of patients receiving EGFR-TKIs as first-line treatment for postoperative recurrent EGFR-mutated lung adenocarcinoma. In addition, our median follow-up period was longer than those of previous studies [3, 5]. Few studies have focused on the management of postoperative recurrent EGFR-mutated NSCLC, which may be due to the difficulties in acquiring detailed information of such cases from either single-institution or large multi-institutional databases [21]. This was also the case in our study, in which data retrieval was labour intensive and required extensive coordination among the hospitals. Our group has previously published 2 studies on the treatment and outcomes of patients with postoperative recurrent EGFR-mutated NSCLC [9, 10]. Our single-institution study of 64 patients with postoperative recurrent EGFR-mutated lung adenocarcinoma had revealed a favourable median OS of 61 months and a favourable median PFS of 18 months with a median follow-up of 28.5 months [9]. Those results were fairly consistent with the favourable survival outcomes (median OS of 55 months and median PFS of 26 months) seen in our present multi-institutional study with a sample size of 154 patients.

Although it is challenging to compare our findings with those from the FLAURA trial [3], the more favourable survival outcomes observed in our study patients may be attributable to their previous resection of the primary lesion and resulting lower tumour burden [5]. In previous clinical trials, patients with postoperative recurrent EGFR-mutated lung adenocarcinoma were not separately analysed but were enrolled as a subset of advanced EGFR-mutated NSCLC [3, 22] or managed in accordance with recommendations on advanced EGFR-mutated NSCLC based on the guidelines of the Japan Lung Cancer Society, National Comprehensive Cancer Network and European Society for Medical Oncology [13–15].

The selection of EGFR-TKIs for first-line treatment mainly involves a choice between conventional EGFR-TKIs and third-generation EGFR-TKI (osimertinib). The FLAURA trial demonstrated the superiority of osimertinib over conventional EGFR-TKIs for OS and PFS [3]. In contrast, our study found that the superiority of osimertinib was less pronounced and did not reach statistical significance. A possible reason may be the smaller sample size of the osimertinib subgroup (n = 25) relative to the conventional EGFR-TKI subgroup (n = 129) in our study. Another reason could be the effects of osimertinib as second-line treatment. At present, the use of osimertinib as a first-line EGFR-TKI treatment appears reasonable for patients with postoperative recurrent EGFR-mutated lung adenocarcinoma in line with major oncology guidelines [20–22].

Major findings of the ADAURA trial included that disease-free survival was significantly longer in those who received osimertinib for patients with stage IB to IIIA EGFR mutation-positive NSCLC than in those who received placebo [23]. On the basis of this finding, osimertinib as an adjuvant EGFR-TKI appears reasonable. Given the fact that superiority in OS has not been shown, currently, there is a controversy between prophylactic EGFR-TKIs and EGFR-TKIs for recurrent disease, both of which appear acceptance on the basis of ADAURA and our findings.

The favourable median post-progressive survival duration of 30.8 months could be indicative of the overall successful management of progressive disease. The selection of osimertinib for second-line use based on re-biopsy results as a switching criterion may also have contributed to the increased post-progressive survival, although the number of patients receiving osimertinib as a second-line EGFR-TKI was too limited to compare with those undergoing other EGFR-TKIs. Further research is needed to explore the use of different strategies for progressive disease management.

The identification of unfavourable prognostic factors can contribute to the early recognition of patients at higher risk of progressive disease after first-line EGFR-TKI treatment. Our study identified EGFR ex 21 L858R mutation to be a significant unfavourable factor of PFS, which is congruent with recent real-world data [5]. In addition, micropapillary and/or solid components showed a trend towards shorter PFS, although this relationship did not attain statistical significance in the multivariable analysis. This suggests that information on tumour subtypes should be included in future research, as also indicated in our previous study [9].

The overall favourable survival outcomes in our study underscore the importance of postoperative surveillance in patients with EGFR-mutated NSCLC. Given the fact that >50% of our patients had multiple lesions and 16.9% had brain metastases, the use of chest CT alone may not be adequate as a surveillance modality, although we were unable to suggest ideal postoperative surveillance protocol because of the lack of data on frequency or modality of recurrence detection in our database.

Limitations

There were several limitations inherent to this study, including the retrospective design, relatively small sample size (which could affect the power of the analysis especially subgroup analysis and multivariable analyses), patient’s characteristics with missing data and the limited information on second-line management after disease progression. In addition, our database lacked data on the timing of recurrence detection, the number of recurrent lesions or the maximum response assessed using the Response Evaluation Criteria in Solid Tumors. DFIs may depend on whether recurrent lesions were symptomatic or asymptomatic. Postoperative surveillance protocols should ideally be standardized in the future for patients with resected EGFR-mutated lung adenocarcinoma.

CONCLUSION

In conclusion, first-line EGFR-TKI treatment was generally associated with favourable survival outcomes in patients with postoperative recurrent EGFR-mutated lung adenocarcinoma. EGFR ex 21 L858R mutation may be an important prognostic factor for shorter PFS. Future studies should explore the modalities of postoperative surveillance in patients with EGFR-mutated lung adenocarcinoma and the management of progressive disease after first-line EGFR-TKI treatment.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

Conflict of interest: none declared.

Data Availability Statement

The data underlying this article will be shared on reasonable request to the corresponding author.

Author contributions

Ryo Miyata: Data curation; Investigation; Writing—original draft. Masatsugu Hamaji: Conceptualization; Data curation; Writing—review & editing. Atsushi Kawaguchi: Supervision. Yumeta Shimazu: Data curation. Masaki Ikeda: Data curation. Masashi Ishikawa: Supervision. Hidenao Kayawake: Data curation. Toshi Menju: Data curation. Masashi Kobayashi: Data curation. Norihito Okumura: Supervision. Yasuto Sakaguchi: Data curation. Makoto Sonobe: Supervision. Akira Matsumoto: Data curation. Tsuyoshi Shoji: Data curation. Hiromichi Katakura: Supervision. Ryota Sumitomo: Data curation. Cheng-Long Huang: Supervision. Mamoru Takahashi: Data curation. Akihiro Aoyama: Supervision. Yusuke Muranishi: Data curation. Tomoya Kono: Data curation. Ryo Miyahara: Supervision. Naoki Date: Data curation. Takuji Fujinaga: Supervision. Ei Miyamoto: Data curation. Tatsuo Nakagawa: Supervision. Takahisa Fukada: Data curation. Hiroaki Sakai: Supervision. Hiroshi Date: Project administration; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks David R. Jones, Giuseppe Cardillo, Federico Tacconi and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• CIs

  Confidence intervals

 
• CT

  Computed tomography

 
• DFI

  Disease-free interval

 
• EGFR-TKIs

  Epidermal growth factor receptor tyrosine kinase inhibitors

 
• HRs

  Hazard ratios

 
• NSCLC

  Non-small-cell lung cancer

 
• OS

  Overall survival

 
• PFS

  Progression-free survival

Author notes